FACULTY OF NATURAL SCIENCES

DIVISION OF CELL AND MOLECULAR BIOLOGY

Influence of Human Gut Microbiota on the

Metabolic Fate of Glucosinolates

Vijitra Luang-In

SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

MARCH 2013

2

ABSTRACT Glucosinolates (GSLs) are secondary metabolites predominantly found in

cruciferous vegetables such as broccoli, brussel sprout, cabbage and cauliflower which

upon chopping and chewing will release the indigenous plant myrosinase enzyme that

catalyzes the hydrolysis of GSLs. This hydrolysis releases a range of breakdown products

including isothiocyanates (ITCs), which have been implicated in the cancer-protective

effects of cruciferous vegetables. Certain human gut bacteria are able to metabolize GSLs

and produce ITCs for human health benefits. In this work, six GSL-metabolizing bacterial

strains were isolated from human faecal sample and identified. Most bacteria were

capable of producing both nitriles (NITs) and ITCs from the metabolism of GSLs however

Enterococcus sp. C213 and Enterococcus faecium KT4S13 produced only NITs.

Enterococcus casseliflavus NCCP-53, Escherichia coli O83:H1 NRG 857C and Lactobacillus

agilis R16 were able to metabolize different types (allyl, aromatic, methylthioalkyl,

methylsulfinylalkyl and indolyl) of GSLs differently over 24 h of in vitro anaerobic

fermentations. For all GSLs, ITC production seemed to peak between 4 and 8 h of

incubation and then declined due to the inherent instability of ITCs in culture broths and

buffers. In contrast, NIT productions gradually increased over time and remained relatively

constant. The total percentage products from each GSL metabolism in all three bacteria

never reached 100%. Interestingly, E. coli O83:H1 NRG 857C produced methylthioalkyl

ITCs and NITs from methylsulfinylalkyl GSLs while E. casseliflavus NCCP-53 produced only

methylsulfinylalkyl ITCs from the same GSLs. This difference was due to reductase activity

in E. coli O83:H1 NRG 857C intact cells and cell-free extracts that biotransforms the

sulfoxide groups of methylsulfinylalkyl GSLs to the sulfide groups. The reductase enzyme is

yet to be identified at the gene and protein level, however it has been characterized using

cell-free extracts in this work. This reductase is inducible by GSLs, oxygen-independent

and requires Mg2+ ion and NADP(H) as co-factors for its activity with optimum pH and

temperature  at  pH  7.0  and  37˚C,  respectively.  Arylsulfatase activity was also detected in

this bacterium. The corresponding recombinant SUL2 enzyme (57 kDa) of E. coli O83:H1

NRG 857C expressed in BL21(DE3) exhibited arylsulfatase activity by desulfating synthetic

p-nitrocatachol sulfate substrate  with  optimum  pH  and  temperature  at  pH  6.0  and  30˚C,  

respectively. In addition, GSL-sulfatase activity was detected in crude extracts by being

3

able to desulfate different intact GSLs to produce desulfo-glucosinolates (DS-GSLs) with

less efficiency in comparison with the commercially available snail sulfatase from Helix

pomatia. The catalytic efficiency of recombinant SUL2 enzyme for GSLs in descending

order is as follows; sinigrin > glucoerucin > gluconasturtiin > glucoiberin. The DS-GSLs

(except DS-glucoraphanin) then act as substrates for the recombinant GH3 enzyme

defived from E. casseliflavus NCCP-53 to produce the corresponding NIT products in NB

broths and the buffer with the presence of 5 mM Fe2+ ions. This enzyme (79 kDa) showed

β-O-glucosidase activity for p-nitrophenyl β-D-glucopyranoside with optimum pH and

temperature   at   pH   7.0   and   37˚C,   respectively.   NIT   productions only occurred from the

metabolism of intact GSLs in bacterial culture broths, but not in the buffers unless 5 mM

Fe2+ ions are present as co-factors. Putative bacterial GSL-degrading enzymes responsible

for ITC and NIT productions from GSL metabolisms are inducible by GSL in resting cells

experiments. By using two-dimensional gel electrophoresis (2-DE) and liquid

chromatography mass spectrometry (LC-MS/MS) for the comparative analysis between

proteins obtained from cultures of L. agilis R16 and E. coli O83:H1 NRG 857C with and

without GSL supplementation, upregulated/distinct proteins that may be involved in the

metabolism of GSLs by these bacteria were identified. These proteins belong to (sugar)

transport system, carbohydrate metabolism especially kinases and oxidoreduction process.

To date, bacterial GSL-degrading enzyme is yet to be identified.

4

ACKNOWLEDGEMENTS

Firstly, I would like to thank my primary supervisor Dr. John Rossiter for giving me

such an interesting PhD project. His kindness, guidance and consistent support have

always made me feel very fortunate to be under his supervision. Secondly, I would like to

thank my co-supervisor Prof. Martin Buck for his constructive criticism and

encouragement. Thirdly, I would like to thank my collaborators, Prof. Richard Mithen and

Dr. Arjan Narbad at the Institute of Food Research (IFR, Norwich) for their kind support,

advice and stimulating discussion for the progress of my work.

I also want to thank the late Dr. Judit Nagy for guiding me through proteomics

materials, Dr.   Alex   Jones   (Sainsbury’s   laboratory,   Norwich)   and Mr. Mark Bennette for

advice on LC-MS analysis, Dr. Ellen James for teaching me the dark art i.e. PCR and

molecular cloning, Dr. Nan Zhang for assisting me with protein purification techniques and

for some dirty-joke entertainment and Dr. Carmen Naneu-Palop for providing me with her

fecal sample as a source of human gut bacteria to study from and that gave rise to my new

nickname  “Vinny  the  Poo’’.  I  want  to  give  a  massive  thank  you to all the past and present

members of the JR group who helped me on my experiments and lifted my spirit up during

some difficult times during my PhD study.

I am blessed to have all wonderful people entering my life during my 10-year stay

in the UK. This certainly has made my brief stay as a human being an incredible journey.

Friends have made me a better person. Moreover, I would not be where I am today

without my beloved parents who have been devoted their lives to get me the best

education, the best well-being and always support me on whatever I am determined to do.

I feel utmost grateful to them for all this lifetime. Above  all,  I’d  pay  my  highest  respect  to  

the timeless teachings of the Lord Buddha which keep me sane and shine the light on me

during the  darkest  hours  in  my  life.  I  will  continue  to  devote  my  life  to  follow  the  Buddha’s  

disciplines till the end of time.

5

DECLARATION OF AUTHORSHIP

I   certify   that   this   thesis   entitled   “Influence   of   Human   Gut   Microbiota   on   the  

Metabolic Fate of Glucosinolates”   is  written  entirely  by  myself,  and  that   the  research  to  

which it refers to is my own. Any ideas or quotations from the studies by other people,

which were published or otherwise, are fully acknowledged in accordance with the

standard referencing practices of the discipline.

Vijitra Luang-In

6

TABLE OF CONTENTS

CONTENT PAGE

ABSTRACT 2

ACKNOWLEDGEMENTS 4

DECLARATION OF AUTHORSHIP 5

TABLE OF CONTENTS 6

LIST OF FIGURES 14

LIST OF TABLES 21

ABBREVIATIONS 25

ABBREVIATIONS FOR AMINO ACIDS 32

Chapter 1 Introduction

1.1 GSL structure, properties, occurrence and biological roles in plants 33

1.2 Biosynthesis of GSLs 39

1.3 Degradation of GSL and its degradation products 42

1.4 Biochemistry of myrosinases 44

1.5 Specifier proteins 46

1.6 Importance of ITCs to human health 47

1.6.1 Cancer Chemoprevention 47

1.6.2 Prevention of diseases 55

1.6.3 Genotoxicity of ITCs 56

1.7 Bioavailability of GSL degradation products in humans 57

1.8 Human gut microbiota 62

1.9 Cruciferous vegetables can alter human gut microbiota communities 65

1.10 Hypotheses 66

1.11 Objectives 66

Chapter 2 Metabolism of GSLs and DS-GSLs by human gut microbiota

2.1. Introduction 67

7

2.1.1 GSL degradation by human gut microbiota 67

2.1.2 Metabolic diversity of the intestinal microbiota 68

2.1.3 Characterization of human gut microbiota 70

2.1.3.1 Enrichment culture technique 72

2.1.3.2 16S rRNA gene analysis 73

2.1.3.3 Polymerase Chain Reaction (PCR) 74

2.1.4 Analytical methods for GSLs and their degradation products 74

2.1.5 Hypotheses 78

2.1.6 Objectives 79

2.2 Materials and Methods 80

2.2.1 Preparation of GSL substrates 80

2.2.2 Preparation of sulfatase 82

2.2.3 Desulfation of GSLs 82

2.2.4 HPLC analytic conditions for DS-GSLs detection 83

2.2.5 Preparation of DS-GSL substrates 84

2.2.6 Authentic ITC and NIT standards 85

2.2.7 Isolation of GSL-degrading bacteria 85

2.2.8 PCR amplification and identification of isolates 86

2.2.9 Culturing conditions and sample collection 87

for HPLC and GC-MS analyzes

2.2.10 Sample preparation for HPLC analysis and 88

quantification of GSL from HPLC results

2.2.11 Sample preparation for GC-MS analysis 88

2.2.12 GC-MS analytical conditions for the detection of GSL 89

degradation products

2.2.13 Determination of percentage product 94

2.2.14 Determination of stability and solubility of ITC/NIT standards 94

2.2.15 Resting cell experiments 94

2.2.16 Determination of metal ion dependency on 95

NIT production from GSL metabolism in bacterial resting cells

2.2.17 Cell-free extract experiments 96

8

2.2.18 Determination of co-factor dependancy for reductase 96

activity in cell-free extracts

2.2.19 Determination of reductase activity in cell-free 97

extracts in the conversion of sulforaphane to erucin

2.2.20 Protein quantification 98

2.2.21 Denaturing sodium dodecyl sulfate polyacrylamide 99

gel electrophoresis (SDS-PAGE)

2.2.22 Native gel electrophoresis 100

2.2.23 Statistical analysis 100

2.3 Results 101

2.3.1 Screening for GSL-metabolising human gut bacteria 101

2.3.2 Isolation and purification of GSL substrates 102

2.3.3 Time-course degradation product profiles of intact 104

GSLs in individual bacterial fermentations

2.3.4 Stability of ITC/NIT degradation products 122

2.3.5 Time-course degradation product profiles of DS-GSLs 128

by individual bacterial fermentation

2.3.6 Resting cell experiments 133

2.3.7 ITC and NIT production by bacterial resting cells 135

in the buffer and the media

2.3.8 Cell-free extract experiments from E. coli O83:H1 NRG 857C 140

2.3.9 Determination of GSL-degrading enzyme activity from bacterial 140

whole cell lystaes on the native gels

2.3.10 Sulfoxide reduction of glucoiberin and glucoraphanin 142

by reductase activity in E. coli O83:H1 NRG 857C

2.3.11 Mg2+ - and NAD(P)H- dependent reductase activity 150

and its optimal pH and temperature

2.4 Summary of key findings 154

2.5 Discussion 155

2.5.1 Bacterial GSL-degrading activity 155

2.5.2 Bacterial reductase activity 161

9

Chapter 3: Forward proteomics approach to identify bacterial proteins potentially

involved in the metabolism of GSLs

3.1 Introduction 164

3.1.1 Forward proteomics 164

3.1.2 Two-dimensional electrophoresis (2-DE) 167

3.1.3 Work-flow of gel-based strategy 168

3.1.3.1 Protein preparation 168

3.1.3.2 Protein separation 170

3.1.3.3 Gel analysis, spot detection and quantification 171

3.1.3.4 Spot excision and digestion 171

3.1.3.5 Protein identification by mass spectrometry 171

3.1.4 Applications of 2-DE in bacterial proteomics 172

3.1.5 Hypotheses 174

3.1.6 Objectives 175

3.2 Materials and Methods 176

3.2.1 Sinigrin supplementation in media and bacterial cell collection 176

3.2.2 Cell lysis and protein extraction 176

3.2.3 Protein quantification 177

3.2.4 Two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) 178

3.2.5 Image acquisition and analysis 180

3.2.6 Estimation of pI and molecular weight (Mw) of the proteins 181

3.2.7 In-gel tryptic digestion 181

3.2.8 LC-MS/MS analysis 183

3.2.9 Database searching and protein identification 184

3.3 Results 185

3.3.1 Optimum GSL concentration to induce bacterial 185

myrosinase expression

3.3.2 Optimization of protein sample preparation for 2-DE 187

3.3.3 Comparative analysis of 2-DE maps of proteins isolated from 191

cells grown on media with and without sinigrin supplementation

3.3.4 LC-MS/MS analysis and protein identification 195

10

3.4 Discussion 200

Chapter 4: Reverse proteomics approach to identify bacterial proteins potentially involved in the metabolism of GSLs 4.1 Introduction 205

4.1.1 Reverse proteomics 207

4.1.1.1 Molecular cloning 208

4.1.1.2 Recombinant protein expression 211

4.1.1.3 Enzyme activity and assay 212

4.1.2 Hypotheses 213

4.1.3 Objectives 214

4.2 Materials and Methods 215

4.2.1 Sequence alignment and bioinformatic analysis 215

4.2.2 Genomic DNA extraction 215

4.2.3 Primers 216

4.2.4 Bacterial strains and plasmids 216

4.2.5 Polymerase chain reaction (PCR) 217

4.2.6 PCR product purification 218

4.2.7 Ligation 219

4.2.8 Preparation of competent cells with ligation mixture 220

4.2.9 Preparation of competent cells 220

4.2.10 Selection of transformants 220

4.2.11 Colony PCR experiment 221

4.2.12 Restriction enzyme digestion 222

4.2.13 Agarose gel electrophoresis 222

4.2.14 Plasmid extraction 223

4.2.15 DNA sequencing and sequence analysis 223

4.2.16 Recombinant protein expression and purification 223

4.2.17 SDS-PAGE analysis 224

4.2.18 Desalting recombinant enzymes 225

4.2.19 GOD-PERID assay 225

4.2.20 Substrates used in GOD-PERID assay 227

11

4.2.21 β-O-glucosidase activity assay 228

4.2.22 Arylsulfatase activity assay 230

4.3 Results 232

4.3.1 BLAST searches and sequence analysis of putative bacterial 232

myrosinases/sulfatase

4.3.2 Cloning of putative bacterial GSL-degrading enzyme/sulfatases 239

4.3.3 Recombinant protein expressions by IPTG induction 242

4.3.4 Enzyme activity assays 243

4.3.4.1 Myrosinase activity using GOD-PERID assay 243

4.3.4.2  β-O-glucosidase activity assay 246

4.3.4.3 Arylsulfatase activity assay 247

4.4. Discussion 248

Chapter 5: Characterization of the recombinant SUL2 enzyme from E. coli O83:H1 NRG

857C and the recombinant GH3 enzyme from E. casseliflavus NCCP-53

5.1 Introduction 250

5.1.1 Sulfatases in nature 251

5.1.2 Bacterial sulftases 253

5.1.3 The GH1 myrosinases 256

5.1.4 The GH3 β-glucosidases 261

5.1.5 Desulfation of intact GSLs and NIT production from DS-GSLs 263

5.1.6 Protein purification techniques 265

5.1.6.1 Desalting and buffer exchange 265

5.1.6.2 Ultrafiltration 265

5.1.6.3 Ion exchange chromatography 266

5.1.6.4 Ni2+-affinity Chromatography 269

5.1.7 Hypotheses 269

5.1.8 Objectives 270

5.2 Materials and Methods 271

5.2.1 Bioinformatics tools 271

5.2.2 Inducibility of a native SUL2 enzyme of E. coli O83:H1 NRG 857C 271

12

5.2.3 Reverse transcriptase polymerase chain reaction (RT-PCR) 271

5.2.4 Purification of recombinant enzymes 272

5.2.4.1 Affinity column chromatography 273

5.2.4.2 Ion-exchange column chromatography 274

5.2.5 Determination of pH and temperature optima 274

5.2.6 Enzyme activities measurements 275

5.2.6.1 Enzyme activities of the recombinant SUL2 enzyme 275

for intact GSLs determined by a discontinuous assay

using HPLC analysis

5.2.6.2 Enzyme activities of the recombinant SUL2 enzyme 276

for pNCS determined by a discontinuous assay

using a spectrophotometric method

5.2.7 Effects of various compounds on arylsulfatase activity of 276

the recombinant SUL enzyme

5.2.8 DS-GSLs as substrates for the recombinant GH3 enzyme 277

5.2.9 Intact GSL as substrates in a reaction containing both 277

the recombinant SUL2 enzyme and the recombinant GH3 enzyme

5.3 Results 278

5.3.1 Bioinformatics results of a native SUL2 enzyme of 278

E. coli O83:H1 NRG 857C

5.3.2 Inducibility of a native SUL2 enzyme of E. coli O83:H1 NRG 857C 281

5.3.3 Expression and purification of the recombinant SUL2 enzyme 283

5.3.4 Temperature and pH optima of the recombinant SUL2 enzyme 286

5.3.5 Desulfation of intact GSLs by the recombinant SUL2 enzyme 287

5.3.6 Enzyme activities of the recombinant SUL2 enzyme 291

5.3.7 Effects of compounds on arylsulfatase activity of the recombinant 295

SUL2 enzyme

5.3.8 Purification of the recombinant GH3 enzyme from 296

E. casseliflavus NCCP-53

5.3.9 Temperature and pH optima of the recombinant GH3 enzyme 297

5.3.10 Effects of metal ions on β-O-glucosidase activity of the 297

recombinant GH3 enzyme

13

5.3.11 NIT production from DS-GSLs by the recombinant GH3 enzyme 298

5.3.12 NIT production from intact GSLs by sequential action of 300

the recombinant SUL2 enzyme and the recombinant GH3 enzyme

5.4 Summary of key findings 303

5.5 Discussion 303

Chapter 6 General discussion

6.1 Summary of findings 309

6.2 Future work 315

6.2.1 Identification of other GSL degradation metabolites from 315

GSL metabolism

6.2.2 Further search for the putative bacterial GSL-degrading enzyme 315

from other bacteria

6.2.3 Determination of whether GSL-6-P is a substrate for bacterial 317

myrosinase in vitro

6.2.4 Purification of bacterial reductase 317

6.3 Conclusion 318

REFERENCES 319

APPENDIX 353

Appendix I 353

Appendix II 354

Appendix III 358

Appendix IV 358

14

LIST OF FIGURES

FIGURE PAGE

Figure 1.1 Structure of GSL 34

Figure 1.2 Typical HPLC chromatogram of GSL profile in 38

Sweetheart cabbage

Figure 1.3 Biosynthesis of GSLs 41

Figure 1.4 GSL degradation under different conditions 43

Figure 1.5 Mechanisms of action of ITCs in the modulation of 50

signalling pathways involved in cancer chemoprevention

Figure 1.6 ITCs modulate a large and diverse group of proteins 51

Figure 1.7 Expected metabolic fate of glucoraphanin and AITC following 59

ingestion of cooked broccoli (Brassica oleracea var. italica)

and mustard (Sinapis alba) respectively by human volunteers

Figure 1.8 The intestine's impact on health 63

Figure 2.1 ITC standard curves from GC-MS analysis 92

Figure 2.2 NIT standard curves from GC-MS analysis 93

Figure 2.3 Representative protein calibration curve 98

Figure 2.4 HPLC chromatograms of GSL substrates used in this work 104

Figure 2.5 Growth curves and pH values of bacterial cultures incubated 108

with  individual  GSLs  anaerobically  at  37˚C  over  a  time  course

Figure 2.6 Time-course degradation product profiles of bacterial cultures 111

anaerobically  incubated  with  individual  GSLs  at  37˚C

Figure 2.7 GC-MS chromatograms of degradation products of different GSLs 114

Figure 2.8 Fingerprint fragment ions of GSL degradation products 115

generated by GC-MS analysis

Figure 2.9 GC-MS chromatograms of ITC degradation products from 116

glucoiberin and glucoraphanin metabolized by E. casseliflavus

NCCP-53 over a time course

Figure 2.10 GC-MS chromatograms of ITC/NIT degradation production 117

from glucoiberin and glucoraphanin metabolized by E. coli

15

O83:H1 NRG 857C over a time course

Figure 2.11 HPLC chromatograms of bacterial degradation 119

of glucobrassicin over a time course

Figure 2.12 Stability of 1 mM ITC/NIT standards in NB broths with/without 124

E. coli O83:H1 NRG 857C cells over a time course

Figure 2.13 Solubility of various concentrations of ITC standards 126

in distilled water

Figure 2.14 Stability of 1 mM ITC standards in various buffers without 127

E. coli O83:H1 NRG 857C cells over a time course

Figure 2.15 GC-MS chromatograms of degradation products of DS-GSLs 130

metabolized by individual three bacteria

Figure 2.16 NIT productions from DS-GSLs metabolized by individual 132

bacteria over a time course

Figure 2.17 GC-MS chromatograms of different degradation products 135

of gluconasturtiin or DS-gluconasturtiin metabolized by E. coli

O83:H1 NRG 857C induced resting cells in different incubation

conditions

Figure 2.18 Effect of the addition of metal ions on PITC/PNIT production 137

From the metabolism of gluconasturtiin in E. coli O83:H1

NRG 857C induced resting cells

Figure 2.19 GC-MS chromatograms of degradation products from 138

gluconasturtiin metabolism in E. coli O83:H1 NRG 857C

resting cells upon addition of metal ions in 0.1 M citrate

phosphate buffer pH 7.0

Figure 2.20 SDS-PAGE analysis of E. coli O83:H1 NRG 857C proteins 141

Figure 2.21 Native gel electrophoresis for GSL-degrading activity test 142

Figure 2.22 Hypothetic scheme of the putative bacterial reductase 143 in E. coli O83:H1 NRG 857C cells. A similar scheme is thought to occur in the metabolism of glucoiberin in these two bacteria. EC, E. casseliflavus NCCP-53; ECO, E. coli O83:H1 NRG 857C.

16

Figure 2.23 Reduction bioconversion of glucoiberin/glucoraphanin 144

to glucoiberverin/glucoerucin by E. coli O83:H1 NRG 857C

intact cells over a time course

Figure 2.24 HPLC chromatograms of methylsulfinylalkyl GSLs converted to 145

methylthioalkyl GSLs by E. coli O83:H1 NRG 857C cell-free extracts

(obtained from glucoraphanin- induced cells) over a time course

Figure 2.25 Reduction bioconversion of sulforaphane to erucin by 149

E. coli O83:H1 NRG 857C intact cells

Figure 2.26 GC-MS chromatograms showing the reduction bioconversion 150

of sulforaphane to erucin by E. coli O83:H1 NRG 857C induced

cell-free extracts over a time course

Figure 2.27 Different metabolic fates of glucoraphanin in E. coli O83:H1 151

NRG 857C (ECO8N) and E. casseliflavus NCCP-53 (ENTCA)

Figure 2.28 HPLC chromatograms showing the effects of co-factor(s) 153

on reductase activity in E. coli O83:H1 NRG 857C cell-free extracts

Figure 2.29 Effects of temperature, pH and aeration on reductase 154

activity in E. coli O83:H1 NRG 857C cell-free extracts

Figure 2.30 Summary of key findings in this chapter 155

Figure 3.1 Strategies for forward proteomics 166

Figure 3.2 Combined gel-LC-MS based strategy (GeLC-MS) 167

Figure 3.3 List of publications in proteomic field by means of 168

two-dimensional electrophoresis technology as of Dec 2011

Figure 3.4 Sub-cellular fractionation of Gram-negative bacterial cell culture 170

Figure 3.5 MALDI-TOF mass spectrometry 172

Figure 3.6 Tandem mass spectrometry (MS/MS) 173

Figure 3.7 BSA calibration curve 178

Figure 3.8 Protein ladders used in 2-DE work 180

Figure 3.9 Calibration curves for pI and Mw determination 181

Figure 3.10 Sinigrin degradation and AITC production from various 186

sinigrin concentrations at 8 h

Figure 3.11 Growth curves of bacteria with and without sinigrin 187

supplementation over 8 h

17

Figure 3.12 Comparison of protein patterns from two lysis methods 189

Figure 3.13 Reproducibility of 2-DE gels from L. agilis R16 190

Figure 3.14 Comparative analysis of representative 2-DE maps of 194

bacterial proteins

Figure 3.15 Representative comparison of 3D montage of expression 195

levels of protein spot

Figure 3.16 Representative MS and MS/MS spectra 196

Figure 3.17 Functional grouping of the 28 upregulated proteins identified 199

on 2-DE gels of both bacteria

Figure 4.1 Hypotheses of this chapter. See main texts for more details 205

Figure 4.2 Scheme of the reverse proteomics workflow 208

Figure 4.3 Recombinant expression mechanisms in pET expression system 211

Figure 4.4 Colour table referring to the labelling of amino acids 215

(single letter code) used in ClustalW alignments

Figure 4.5 Map of an expression vector, pET28b(+) 219

Figure 4.6 Quick-Load 1 kb DNA ladder molecular marker 222

Figure 4.7 Protein markers 223

Figure 4.8 GOD-PERID assay reaction principle 226

Figure 4.9 Calibration curve for GOD-PERID assay 227

Figure 4.10 Structures of substrates used in GOD-PERID assay 228

Figure 4.11 β-O-glucosidase activity assay reaction principle 228

Figure 4.12 Calibration  curve  for  β-O-glucosidase activity assay 229

Figure 4.13 Sulfatase assay reaction principle 230

Figure 4.14 Calibration curve for arylsulfatase activity assay 231

Figure 4.15 Alignment of eight bacterial putative myrosinase peptide 234

sequences with the Brevicoryne brassicae myrosinase `Aphid'

Figure 4.16 Alignment of two bacterial putative sulfatase peptide 237

sequences with the H. pomatia sulfatase `Snail'

Figure 4.17 Phylogenetic tree for bacterial putative myrosinases 239

and sulfatases

Figure 4.18 Agarose gel electrophoresis of genomic PCR experiments 240

18

Figure 4.19 Agarose gel electrophoresis of colony PCR experiments 241

Figure 4.20 Agarose gel electrophoresis of restriction digestion experiments 242

Figure 4.21 Recombinant protein expressions on SDS-PAGE 243

Figure 4.22 Myrosinase activity using GOD-PERID assay 244

Figure 4.23 β-O-glucosidase activity using GOD-PERID assay 245

Figure 4.24 β-O-glucosidase activity assay 246

Figure 4.25 Arylsulfatase activity assay 247

Figure 5.1 Hypothese of this chapter. See main text for more details 250

Figure 5.2 Sulfatase reactions 251

Figure 5.3 Crystal structures of P. aeruginosa arylsulfatase (PARS) 254

Figure 5.4 Proposed mechanistic scheme for the hydrolysis of sulfate 255

ester by the active-site aldehyde FGly of PARS

Figure 5.5 The overall structure of plant myrosinase 257

Figure 5.6 The ascorbate activated catalysis of GSL hydrolysis by plant 258 myrosinase.

Figure 5.7 The structure of aphid myrosinase showing the dimer 258

Figure 5.8 The catalysis of glucosinolates by aphid myrosinase 259

Figure 5.9 Proposed scheme of NIT production by desulfation of GSL 264

via  sulfatase  and  β-O-glucosidase

Figure 5.10 Ion-exchange chromatography 267

Figure 5.11 Format used for FPLC chromatography 268

Figure 5.12 Ni2+-affinity chromatography 269

Figure 5.13 Bioinformatics details on SUL2 enzyme 279

Figure 5.14 Inducibility test of a native SUL2 enzyme of E. coli 283

O83:H1 NRG 857C

Figure 5.15 Purification of the recombinant sulfatase SUL2 by 284

ion-exchange chromatography

Figure 5.16 Purification of the recombinant SUL2 enzyme expressed 285

in BL21(DE3)

Figure 5.17 Temperature and pH optima of the recombinant SUL2 enzyme 287

Figure 5.18 Kinetics of H. pomatia sulfatase activity (0.05 U/mL) in 288

desulfation of intact GSLs

19

Figure 5.19 HPLC chromatograms showing desulfation of intact GSLs 290

by crude extracts of the recombinant SUL2 enzyme and the

purified H. pomatia sulfatase on the DEAE-Sephadex column

Figure 5.20 Enzyme activities of the recombinant SUL2 enzyme 292

for pNCS substrate

Figure 5.21 Enzyme activities of the recombinant SUL2 enzyme 294

for GSL substrates

Figure 5.22 Effect of compounds on the activity of the recombinant 295

SUL2 enzyme

Figure 5.23 Purification of the recombinant GH3 enzyme expressed 296

in BL21(DE3)

Figure 5.24 Temperature and pH optima of the recombinant GH3 enzyme 297

Figure 5.25 GC-MS chromatograms showing NIT production from DS-GSLs 299

by the purified recombinant GH3 enzyme in NB broths

Figure 5.26 Reaction catalyzed by the recombinant enzymes SUL2 and GH3 303

Figure 6.1 Proposed scheme of myrosinase and reductase of E. coli 310

O83:H1 NRG 857C induction by GSL

Figure 6.2 Summarized scheme of GSL/DS-GSL metabolism by human gut 311

bacteria and by bacterial recombinant enzymes under various

conditions

Figure 6.3 Proposed schematic presentation of GSL-metabolizing 314

mechanism in human gut bacteria

Appendix IA Representative GC-MS chromatogram showing no degradation 353

products from the negative control containing only GSL

substrate without bacterial cells or only bacterial cells without

GSL incubated  in  the  culture  broths  for  24  h  at  37˚C  under  

anaerobic conditions

Appendix IB Representative GC-MS chromatogram showing no degradation 353

products from the negative control containing only DS-GSL

substrate without bacterial cells incubated in the culture broths

for  24  h  at  37˚C  under  anaerobic  conditions

Appendix II List of gene sequencing results from the recombinant plasmids 354

20

Appendix III Representative HPLC chromatograms showing no DS-GSL 358

production upon 8 h incubation of intact GSLs with crude extracts

from BL21(DE3) on the DEAE-Sephadex column at 30˚C  under  

aerobic conditions

Appendix IV Representative GC-MS chromatograms showing no NIT 358

production in the negative controls containing DS-GSL alone,

GSL alone, the GH3 enzyme alone, the SUL2 enzyme alone

or the two enzymes alone incubated in NB broth or in the buffer

for  24  h  at  37˚C under anaerobic conditions

21

LIST OF TABLES TABLE PAGE Table 1.1 Most abundant GSLs found in nature 35

Table 1.2 Expected ITC or NIT product from GSL hydrolysis catalyzed 49

by plant myrosinase

Table 1.3 Chemopreventive actions of ITCs produced from GSL hydrolysis 52

Table 2.1 Some metabolic reactions of intestinal microbiota 69

Table 2.2 Advantages and disadvantages of current techniques used to 71

characterize human gut microbiota

Table 2.3 Listing of some commonly used methods for the analysis 78

of GSLs and their breakdown products

Table 2.4 Response factors for desulfated GSLs at 229 nm in relative to 81

that of desulfo-sinigrin

Table 2.5 GSLs involved in this work as detected by HPLC analysis 84

Table 2.6 Compositions of culture media 86

Table 2.7 Mass spectral (MS) data of GSL degradation products 90

Table 2.8 Compositions of SDS-PAGE, loading buffer, running buffer, 99

staining/destaining solutions

Table 2.9 Bacterial isolates exhibiting > 50% degradation of 1 mM 102

sinigrin  in  24  h  anaerobic  incubation  at  37˚C  

Table 2.10 Detection of ITC and NIT products from GSL metabolism 106

in bacterial fermentations

Table 2.11 Bacterial growth and glucobrassicin degradation in E. coli 118

O83:H1 NRG 857C and E. casseliflavus NCCP-53 over a time course

Table 2.12 Time taken to obtain 50% degradation of each GSL 120

substrate by three bacteria

Table 2.13 Percentage products of each ITC/NIT product from all GSL 121

metabolisms by each bacterium

Table 2.14 Time taken to obtain 50% or 25% decline in each ITC or 125

NIT level, respectively in NB broth with or without the

presence of E. coli O83:H1 NRG 857C cells

22

Table 2.15 Time taken to obtain 50% decline in each ITC level in 128

various aqueous solutions without the presence of

E. coli O83:H1 NRG 857C cells

Table 2.16 Detection of NIT product from DS-GSL metabolism in 129

bacterial fermentations

Table 2.17 PITC production from gluconasturtiin metabolism in 133

E. casseliflavus NCCP-53 resting cells in different buffers

for  8  h  at  37˚C under anaerobic conditions

Table 2.18 Degradation of gluconasturtiin by bacterial resting cells 134

in 0.1 M citrate phosphate buffer pH 7.0 anaerobically

incubated for 2 h at  37˚C

Table 2.19 Effect of EDTA on ITC/NIT production from gluconasturtiin 134

metabolism in E. coli O83:H1 NRG 857C in NB broth for 16 h

anaerobic  incubation  at  37˚C

Table 2.20 Effect of Fe2+ ions (5 mM) on ITC/NIT production from 136

the metabolisms of different GSLs (0.5 mM) by E. coli O83:H1

NRG 857C induced resting cells in 0.1 M citrate

phosphate buffer  pH  7.0  for  16  h  anaerobic  incubation  at  37˚C

Table 2.21 Reduction bioconversion of glucoraphanin by cell-free extracts 139

of E. coli O83:H1 NRG 857(obtained from glucoraphanin- induced

or gluconasturtiin-induced cells) over a time course in 0.1 M citrate

phosphate  buffer  pH  7.0  at  37˚C  under  anaerobic  conditions

Table 2.20 Reduction bioconversion of sulforaphane to erucin by cell-free 146

extracts and resting cells of E. coli O83:H1 NRG 857C (induced

with 1 mM glucoraphanin overnight) over a time course in 0.1 M

citrate  phosphate  buffer  pH  7.0  at  37˚C  under  anaerobic  conditions

Table 2.22 Reduction bioconversion (%) of 0.25 mM glucoraphanin to 148

glucoerucin by the addition of 1 mM of co-factor(s) in desalted

cell-free extracts of E. coli O83:H1 NRG 857C (obtained from

glucoraphanin-induced cells) within  24  h  at  37˚C  under  anaerobic  

conditions

Table 3.1 Reagents used in in-gel tryptic digestion and their compositions 182

23

Table 3.2 Experimental results of each set of cultures (with or without 191

sinigrin supplementation) from each bacterium

Table 3.3 Protein identification of differentially abundance spots 197

(≥  2  fold  increase  in  spot  volume  ratio  and  ANOVA  p  ≤  0.05  with  

≥ 2 matched peptides) of L. agilis R16 and E. coli O83:H1 NRG

857C grown on media with 2 mM sinigrin supplementation

Table 4.1 Primers used in PCR experiments and their restriction sites are 216

underlined

Table 4.2 Bacterial strains and plasmids used in this study 217

Table 4.3 Components in one PCR reaction 217

Table 4.4 Thermal cycling conditions for Pfu DNA Polymerase- 218

mediated PCR amplification

Table 4.5 Components in one ligation reaction 219

Table 4.6 Components in one colony PCR reaction 221

Table 4.7 Thermal cycling conditions for Taq DNA Polymerase-mediated 222

colony PCR amplification

Table 4.8 Ingredients for restriction enzyme digestion 222

Table 4.9 Information on genome and proteome of bacteria under study 232

Table 4.10 List of putative bacterial GSL-degrading enzymes/sulfatases with 233

high similarity to aphid myrosinase and snail sulfatase

Table 5.1 Properties of snail and bacterial sulfatases 255

Table 5.2 Properties of plant and aphid myrosinases 260

Table 5.3 Properties of the GH1 and GH3 enzyme families in comparison 262

Table 5.4 List of primers used in RT-PCT experiments 272

Table 5.5 Steps involved in the elution of the recombinant SUL2 enzyme 274

using FPLC

Table 5.6 Six proteins from E. coli O83:H1 NRG 857C and E. coli BL21(DE3) 281

producing significant alignments with anaerobic sulfatase-

maturating enzyme homolog (AslB) from Escherichia coli strain K12

Table 5.7 Purification scheme of the recombinant SUL2 enzyme from 286

Ni2+-affinity column chromatography

Table 5.8 Specific activity and relative activity of crude extracts of the 291

24

recombinant SUL2 enzyme in desulfation of intact GSLs

Table 5.9 Enzyme activities of the recombinant SUL2 on different 293

substrates

Table 5.10 Purification scheme of the recombinant GH3 enzyme from 297

Ni2+-affinity column chromatography

Table 5.11 Effects of metal ions on β-O-glucosidase activity of the 298

recombinant GH3 enzyme

Table 5.12 NIT productions from DS-GSLs by the purified recombinant 300

GH3 enzyme

Table 5.13 NIT productions from intact GSLs by sequential action of the 301

recombinant SUL2 enzyme and the recombinant GH3 enzyme

25

ABBREVIATIONS

2-DE Two-dimensional gel electrophoresis

3MSP 3-Methylsulfinylpropyl

3MTP 3-Methylthiopropyl

4MSB 4-Methylsulfinylbutyl

4MTB 4-Methylthiobutyl

6pbg 6-phospho-β-galactosidase

ABTS 2,2'-azino-bis-3-ethylbenzthiazoline-6-sulphonic acid AITC Allyl isothiocyanate

ANIT Allyle nitrile

AP-1 Activator protein-1

APS Ammonium persulfate

ARE Antioxidant response element

ARS Arylsulfatase

ATP Adenosine triphosphate

Bcl-2 B-cell lymphoma 2)

BCRP Breast cancer resistance protein

bgl β-glucosidase BITC Benzyl isothiocyanate

BLAST Basic local alignment search tool BNIT Benzyl nitrile

bp Base pair

BSA Bovine serum albumin

C-terminus Carboxylic terminus

CCD Charge-coupled device CcpA Catabolite control protein A

CCR Carbon catabolite repression

cdc Cell division cycle

Cdk1 Cyclin-dependent kinase 1

26

cDNA Complementary deoxyribonucleic acid

CDPX Chondrodysplasia punctata

CFU Colony-forming unit

CM Carboxymethyl

COG Clusters of orthologous group

COSY Correlation spectroscopy

COX-2 Cyclooxygenase-2 Cre Catabolite responsive element

CREB cAMP response element-binding protein

CYP Cytochrome P450 enzyme

Da Dalton

DCM Dichloromethane

DEAE Diethylaminoethyl

DGGE Denaturing gradient gel electrophoresis DMSO Dimethyl sulfoxide

DNA Deoxyribonucleic acid

DS-GSL Desulfo-glucosinolate

DTT Dithiothreitol

E I Enzyme I

E II Enzyme II

EC Enterococcus casseliflavus NCCP-53

ECO Escherichia coli O83:H1 NRG 857C

EDTA Ethylenediaminetetraacetic acid

EI Electron ionization

ELISA Enzyme-linked immunosorbent assay

ER Endoplasmic reticulum

ER Enhanced resolution

ERN Erucin

ERN NIT Erucin nitrile

ESI Electrospray ionization

27

ESM Epithiospecifier modifier

ESP Epithiospecifier protein

EST Expressed sequenced tag

ETN Epithionitrile

FAB Fast atom bombardment mass spectrometry

FACS Fluorescence activated cell sorting

FADH2 Reduced flavin adenine dinucleotide

FGE Formylglycine-generating enzyme

Fgly Formylglycine

FISH Fluorescence in situ hybridization FPLC Fast protein liquid chromatography

G2/M Growth 2/Mitosis

GBS Glucobrassicin GC-MS Gas chromatography mass spectrometry

GeLC-MS Gel liquid chromatography mass spectrometry

GER Glucoerucin

GFP Green fluorescent protein

GH Glycosyl hydrolase family

GI Gastrointestinal

GIB Glucoiberin

GIV Glucoiberverin

GLC Gas liquid chromatography GlpK Glycerol kinase

GNT Gluconasturtiin

GRP Glucoraphanin

GSH Glutathione GSL Glucosinolate

GSL-6-P Glucosinolate-6-phosphate GST Glutathione S-transferase

GSTM1 Glutathione S-transferase Mu 1

28

GSTT1 Glutathione S-transferase Theta 1

GTP Glucotropaeolin

HAD Haloacid dehalogenase

HARSA Human lysosomal arylsulfatase A

HARSB Human arylsulfatase B

HDAC Histone deacetylase His Histidine

HL-60 Human promyelocytic leukemia cells HPLC High pressure liquid chromatography

HPr Histidine-containing Protein

I3C Indole-3-carbinol

I3M Indol-3-ylmethyl

IAA Iodoacetamide or Indole-3-acetic acid

IBR Iberin

IBR NIT Iberin nitrile

IBS Irritable bowel syndrome

IBV Iberverin

IBV NIT Iberverin nitrile

IEF Isoelectric focusing

IPG Immobilized pH gradient

IPTG Isopropylthio-β-galactoside ITC Isothiocyanate

ITC MA Isothiocyanate mercapturic acid

KEAP1 Kelch-like ECH-associated protein 1

Km Michaelis-Menten constant

KmR Kanamycin resistance

KO Knock out

LA Lactobacillus agilis R16

LAB Lactic acid bacterium LC-MS Liquid chromatography mass spectrometry

29

Ler Landsberg erecta

log P Octanol–water partition coefficient

LPS Lipopolysaccharide

m/z Mass to charge ratio

M+ Molecular ion

MA Mercapturic acid MALDI-TOF Matrix-assisted laser desorption/ionization time-of-flight

MAPK Mitogen-activated protein kinase

MEKK1 Mitogen-activated protein kinase kinase 1

MES 2-(N-morpholino)ethanesulfonic acid

MMP-9 Matrix metallopeptidase 9

mol mole

miRNA Micro RNA

mRNA Messenger ribonucleic acid

MRS de Man, Rogosa and Sharpe

MRP-1 Multidrug resistance protein 1

MS Mass spectrometry

MS/MS Tandem mass spectrometry

MSD Multiple sulfatase deficiency

Mw Molecular weight

N-terminus Amino acid terminus

NA Not available

NADH Nicotinamide adenine dinucleotide

NADPH Reduced nicotinamide adenine dinucleotide phosphate

NB Nutrient broth

ND Not detected

n.d. Not determined

NF-κB Nuclear factor kappa B

NIR Near infra-red reflectance

NIT Nitrile

30

NL Non-linear

NMR Nuclear magnetic resonance NQO1 NAD(P)H:quinone oxidoreductase 1

Nrf2 NF-E2-related factor-2

NSP Nitrile specifier protein

NTA Nitroloacetic acid P-His-HPr Histidyl-phosphorylated form of HPr

P-Ser-HPr Seryl-phosphorylated form of HPr

PAPS 3’  Phosphoadenosine  5’-phosphosulfate

PARS Pseudomonas aeruginosa arylsulfatase

pBgl Periplasmic  β-glucosidase

PBS Phosphate buffered saline PCR Polymerase chain reaction

Pgp P-Glycoprotein

pI Isoelectric point

PITC Phenethyl isothiocyanate

pKa Ionisability constant

PMF Peptide mass fingerprinting

pNC p-Nitrocatechol

pNCS p-Nitrocatechol sulfate

PNIT Phenethyl nitrile

pNP p-Nitrophenol

pNPG p-Nitrophenyl-β-D-glucopyranoside PRD Phosphotransferase system-regulatory domain

psi Pound per square inch

PTM Post-translational modification PTS Phosphotransferase system

qPCR Quantitative polymerase chain reaction

QR Quinone reductase R The side chain group of an amino acid

31

RNA Ribonucleic acid

RNAi Ribonucleic acid interference

RNA-Seq RNA-sequencing

ROS Reactive oxygen species

rRNA Ribosomal RNA

RT-PCR Reverse transcriptase poly chain reaction

SCFA Short-chained fatty acid

SDS-PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis SF Sulforaphane

SFN NIT Sulforaphane nitrile

SIGEX Substrate induced gene expression

SIM Selected ion monitoring

SIP Stable isotope probing

SNG Sinigrin

SRB Sulfate-reducing bacteria

STAT3 Signal transducer and activator of transcription factor 3

SUL Sulfatase

TBP Tributylphosphine

TEMED N,N,N',N'-Tetramethylethylenediamine

TFP Thiocyanate forming protein

TGGE Temperature gradient gel electrophoresis

TH Thiohydroximates

TLR4 Toll-like receptor 4

TNF-α Tumor necrosis factor alpha

TOCSY Total correlated spectroscopY

TPA Tissue plasminogen activator

TR Retention time

tRFLP Terminal restriction fragment length polymorphism

Tris-Cl Tris(hydroxymethyl)amino methane-hydrochloric acid

UGT UDP-glucuronosyl transferase

32

UV Ultraviolet

VEGF Vascular endothelial growth factor

Vmax Maximum velocity

WC Wilkins Chalgren

XRF X-ray fluorescence spectroscopy

ABBREVIATIONS FOR AMINO ACIDS

Amino acid Three-letter abbreviation One-letter symbol

Alanine Ala A

Arginine Arg R

Asparagine Asn N

Aspartate Asp D

Cysteine Cys C

Glutamine Gln Q

Glutamate Glu E

Glycine Gly G

Histidine His H

Isoleucine Ile I

Leucine Leu L

Lysine Lys K

Methionine Met M

Phenylalanine Phe F

Proline Pro P

Serine Ser S

Threonine Thr T

Tryptophan Trp W

Tyrosine Tyr Y

Valine Val V

33

Chapter 1: Introduction The regular consumption of broccoli along with other vegetables of the Brassica

family including cauliflowers, watercress, rocket salads and mustards have been

associated with a lower incidence of cancer. These vegetables are rich in glucosinolates

(GSLs),   secondary   metabolites   that   can   be   degraded   by   myrosinase   (β-D-

thioglucohydrolase, E.C. 3.2.1.147) to biologically active molecules. Myrosinase is naturally

found in all Brassica plants and in the specialist aphids Brevicoryne brassicae and Lipaphis

erysimi. To date, one class of degradation products from GSLs, the isothiocyanates (ITCs)

have been the most-studied and implicated in the cancer chemopreventive properties of

Brassica vegetables. However, plant myrosinases are rapidly denatured by cooking, and

thus the health properties of these vegetables are entirely dependent upon the putative

bacterial GSL-degrading activity of human gut microbiota. Although plant and aphid

myrosinases are well-characterized, bacterial GSL-degrading enzyme still remains elusive.

Therefore, it is essential that the influence of human gut microbiota on the metabolic fate

of GSLs is researched, and this has become the focus of this PhD project.

1.1 GSL structure, properties, occurrence and biological importance in plants GSLs are naturally occurring secondary metabolites abundant throughout 15

botanical families of the order Capparales, such as the Brassicaceae, Capparaceae, and

Resedaceae (Rodman et al., 1996). Representatives of the Brassicaceae are of particular

importance as vegetables (e.g. cabbage, broccoli, cauliflower, Brussels sprouts), root

vegetables (e.g. radish, turnip, swede), leaf vegetables (e.g. rocket salad), and relishes (e.g.

wasabi, mustard) (Stoewsand et al., 1995, Fahey et al., 2001; Rosa et al., 1997) for human

diets. Interestingly, GSLs are also present in the genus Drypetes of the family

Euphorbiaceae, a genus completely unrelated to the other GSL-containing families (Halkier

& Gershenzon, 2006).

GSLs are   substituted   β-thioglucoside N-hydroxysulfates which consist of two

moieties, a glycone and a variable aglycone (Rosa et al., 1997). The  glycone  (β-D-glucose)

and aglycone (thiohydroximate-O-sulfonate)   moieties   are   joined   by   a   β-thioglucoside

linkage. The aglycone moiety is highly variable due to the side chain (R, Figure 1.1) derived

34

from any one of eight amino acids namely alanine (Ala), valine (Val), leucine (Leu),

isoleucine (Ile), phenylalanine (Phe), methionine (Met), tyrosine (Tyr) and tryptophan (Trp)

(Halkier & Gershenzon, 2006).

Figure 1.1 Structure of GSL. A GSL consists of a glycone and an aglycone joined by a β-thioglucosidic bond (red line). R is a side-chain of an amino acid.

Over 120 different GSLs have been identified to date (Fahey et al., 2001). Based on

the amino acid precursor of the side chain and the types of modification to the R group,

most GSLs can be assigned to one of three major structural groups (Hopkins et al., 2009).

Compounds derived from Ala, Leu, Ile, Met, or Val amino acids are known as aliphatic GSLs

constituting about 50% of the known structures, those derived from Phe or Tyr as

aromatic GSLs (10%), and those from Trp, indole GSLs (10%). The remaining 30% of known

structures are derived from various amino acids, or their biosynthetic origin is yet to be

known (Fahey et al., 2001; Mithen, 2001). The R groups of most GSLs are extensively

modified from these precursor amino acids, with methionine undergoing an especially

wide range of transformations (Fahey et al., 2001). Most of the R groups are elongated by

one or more methylene moieties. Both elongated and non-elongated R groups are subject

to a wide variety of transformations, including hydroxylation, O-methylation, desaturation,

glycosylation, and acylation. Known GSLs occur within structural series of variable side-

chain length, which are derived from corresponding chain-elongated amino acids. Thus, it

is likely that more than 170 additional GSLs exist in nature, but they have not yet been

discovered (Clarke, 2010). The structures of the most abundant GSLs and their classes are

shown in Table 1.1

35

Table 1.1 Most abundant GSLs found in nature

GSL class Semi-systematic name Trivial name R group

Aliphatic Allyl (2-propenyl) Sinigrin

3-Butenyl Gluconapin

2-Hydroxy-3-butenyl (R) Progoitrin

4-Pentenyl Glucobrassicanapin

Aromatic Benzyl Glucotropaeolin

2-Phenethyl Gluconasturtiin

36

GSL class Semi-systematic name Trivial name R group

Methylthioalkyl 3-(Methylthio)propyl Glucoiberverin

4-(Methylthio)butyl Glucoerucin

Methylsulfinylalkyl 3-(Methylsulfinyl)propyl Glucoiberin

4-(Methylsulfinyl)butyl Glucoraphanin

Indolic Indol-3-ylmethyl Glucobrassicin

1-Methoxyglucobrassicin Neoglucobrassicin

37

Aqueous solubility, ionisability (pKa), and lipophilicity (octanol–water partition

coefficient, log P) determine the dissolution of GSLs. Log P is a crucial factor indicating

passive membrane partitioning and therefore affects membrane permeability, but is

inversely related to solubility (i.e. increasing log P enhances permeability but reduces

solubility). The presence of the sulfate group (very low pK value of the sulfonic acid group)

confers strongly acidic properties on intact GSLs. Because of the sulfate group and

thioglucose moiety, intact GSLs are always water-soluble. Because of their negative log P

values, GSLs are unlikely to be able to cross cell membranes, and would have to be enabled

by active transport or through aqueous pores. The structure of the side chain influences log

P value of GSL degradation product (Holst & Williamson, 2004). Log P values were found

within a range of 0.23 to 4.37; thus most GSLs degradation products, such as ITC and indolyl

products, are relatively hydrophobic (Cooper et al., 1997). Several factors such as species,

cultivar, tissue type physiological age, plant health, environmental factors (agronomic

practice, climatic conditions), insect attack, and microorganism intrusion can influence GSL

occurrence and concentrations (Rosa et al., 1997; Mithen et al., 2000; Fenwick et al., 1989;

Ciska et al., 2000). GSLs are present in all organs of the plants, but their concentrations and

profiles vary. There are huge differences in GSL profiles in vegetative tissues and those in

flowers and seeds, where the total amount of the former can be 10 times higher than the

latter and can account for up to 10% of the dry matter. Also, there can be as many as fifteen

different GSLs in the same plant, but usually only three or four predominate. Phenethyl GSL

(a.k.a gluconasturtiin) is found at high levels in some minor crops such as radishes and

watercress, and hydroxybenzyl GSLs are the major component of white mustard, Sinapis

alba. The highest concentrations are predominant in the seeds, except for indol-3-ylmethyl

and N-methoxyindol-3-ylmethyl GSLs, which are rarely found in seeds (Brown et al., 2003). A

typical HPLC chromatogram with the peaks corresponding to different GSLs is shown in

Figure 1.2.

38

Figure 1.2 Typical HPLC chromatogram of GSL profile in Sweetheart cabbage. The peaks are glucoiberin, 1, sinigrin, 2, glucobrassicin, 3, 4-methyoxyglucobrassicin, 4 and neoglucobrassicin, 5. This figure was taken from Nurul Huda Abd Karim, PhD thesis. The potential roles of GSLs in plants are involved in several systems as follows:

Plant growth regulation by GSL metabolism. Indole GSLs have been proposed as

precursors for the plant hormone indole-3-acetic acid (IAA). The indole GSLs are typically

hydrolyzed to indole acetonitrile, which could be hydrolyzed further to IAA by nitrilase

(Searle et al., 1982; Bartel & Fink, 1994). Interestingly, mutation of the genes involved in the

late steps of biosynthesis of indole GSLs in Arabidopsis mutants leads to high levels of IAA

and a corresponding dwarf phenotype (including adventitious root). It was demonstrated

that disruption of the conversion of indole-3-acetaldoxime to indole GSLs causes increased

flux into IAA (Mikkelson et al., 2004). One of the more complex interactions of GSLs/ITCs is

their activity as allelochemicals, compounds that affect successive plant communities and/or

those growing simultaneously, in close proximity (Brown & Morra, 1995).

The role of the GSL-myrosinase system in plant-insect/herbivore interactions. GSLs

are well known as defensive compounds against generalist herbivores and are likely to play a

role in host plant recognition by specialist predators, therefore acting both as an insecticide

and as an insect feeding attractant (Louda & Rodman, 1983; Mithen et al., 1986; Louda et al.,

39

1987; Tsao et al., 1996; Rask et al., 2000). For example, GSLs serve as important feeding cues

to insects including Pieris sp. caterpillars and other specialist feeders e.g. Plutella sp., seed

weevils, flea beetles which are differentially stimulated to feed by various GSLs (Renwick et

al., 1992).

The role of the GSL-myrosinase system in plant/pathogen interactions. The role of

GSLs in defense against pathogens is less clear than that for herbivores. Plants recognize the

main signal molecules which are derivatives of jasmonic acid, salicylic acid and ethylene.

These molecules mediate the plant response resulting in the activation of distinct sets of

defensive genes (Reymond & Farmer, 1998; Turner et al., 2002). Another indication for a

putative defensive role of GSLs came from numerousl studies showing changes in GSL

pattern after treatment with signal molecules. The toxicity of GSL hydrolysis products to soil-

borne fungal and bacterial plant pathogens in vitro has been reported (Smolinska et al.,

2003; Mari et al., 2002). In addition, GSLs have been shown as nematicides (Lazzeri et al.,

1993; Mayton et al., 1996) and as a feeding deterrent to snails, caddisflies and amphipods

(Newman et al., 1992). To date, the mechanism of GSLs induction, the signaling pathways

involved   and   the   plant’s   potential benefit from GSLs abundance still remain to be

determined.

1.2 Biosynthesis of GSLs

Since the 1960s, the pathway of biosynthesis of GSLs has been elucidated, and many

intermediates, enzymes and genes involved have been identified. The biosynthesis of GSLs

was reviewed extensively (Botti et al., 1995; Halkier & Gershenzon, 2006; Sønderby et al.,

2010). Knowledge of biosynthetic pathways of GSLs has progressed as research advanced

from conventional in vivo feeding studies with radiolabelled precursors and biochemical

characterization of the enzymes, identification and characterization of the biosynthetic genes

encoding the involved enzymes. The publication of the Arabidopsis thaliana genome has

enabled the complete elucidation of the core biosynthetic pathway.

Biosynthesis of GSLs proceeds through three separate steps: (i) chain elongation of

selected precursor amino acids (only Met and Phe), (ii) formation of the core GSL structure,

and (iii) secondary modifications of the amino acid side chain (Figure 1.3). Both side-chain

elongation and secondary modifications are responsible for the >120 known glucosinolate

40

structures (Fahey et al., 2001), of which Arabidopsis has about 40, mainly derived from Met

and Trp (Kliebenstein et al., 2001). More details of the three biosynthetic steps are as

follows:

(i) Before entering the core structure pathway, Met undergoes chain elongation in a

similar process that the branched-chain amino acid Val is converted to its chain-elongated

homolog Leu. The first step involves a deamination by a branched-chain amino acid

aminotransferase (BCAT), which yields a 2-oxo acid. The 2-oxo acid then enters a cycle of

three successive transformations: condensation with acetyl-CoA by a methylthioalkylmalate

synthase (MAM), isomerization by an isopropylmalate isomerase (IPMI), and oxidative

decarboxylation by an isopropylmalate dehydrogenase (IPM-DH). The product of these three

reactions is a 2-oxo acid that has been elongated by a single methylene group (–CH2–

theNext, the molecule can either be transaminated by a BCAT to yield homoMet and enter

the core glucosinolate structure pathway or proceed through another round of chain

elongation. Accordingly, the overall process yields not only homoMet, but an array of chain-

elongated derivatives of Met.

(ii) The chain-elongated amino acids are then converted to aldoximes by cytochromes

P450 of the CYP79 family. CYP79B2 and CYP79B3 both metabolize Trp, CYP79A2 uses Phe as

a substrate (Wittstock & Halkier, 2000), CYP79F1 converts all chain-elongated Met

derivatives, and CYP79F2 only converts the long-chained Met derivatives. After that,

aldoximes are oxidized to activated compounds (either nitrile oxides or aci-nitro compounds)

by cytochromes P450 of the CYP83 family. CYP83B1 metabolizes both the Trp-derived and

Phe-derived acetaldoximes, and CYP83A1 converts aliphatic aldoximes. Following

conjugation of the activated aldoximes to a sulfur donor, glutathione (Dixon et al., 2010),

which can happen non-enzymatically, the produced S-alkyl-thiohydroximates are converted

to thiohydroximates by the C-S lyase SUR1 (Mikkelsen et al., 2004). Thiohydroximates are in

turn S-glucosylated by glucosyltransferases of the UGT74 family to form

desulfoglucosinolates. The glucosylation gives rise to desulfoglucosinolates, which are finally

sulfated by the sulfotransferases ST family to form a core GSL structure.

(iii) The core GSL structure can undergo side-chain modifications involving different

reactions, such as oxidation, desaturation, hydroxylation, removal of a methylsulfinyl group,

41

and addition of methoxy group. The flavin monooxygenase FMOGS-OX1 was identified as a

candidate for S-oxygenation (Hansen et al., 2007). AOP2 catalyzes the conversion of S-

oxygenated glucosinolates to alkenyl glucosinolates, whereas AOP3 catalyzes the conversion

to hydroxyalkyl glucosinolates (Kliebenstein et al., 2001). The overview of biosynthetic

pathways of GSLs is shown in Figure 1.3

Figure 1.3 Biosynthesis of aliphatic GSLs in Arabidopsis Col-0. (A) The chain-elongation machinery. Methionine enters the chain-elongation cycle via deamination by BCAT4 and is subsequently condensed with acetyl-CoA in a reaction catalyzed by MAM1 and MAM3. MAM1 can catalyze one to four condensation cycles, whereas MAM3 catalyzes one to six cycles. Subsequently, an isomerization and oxidation-decarboxylation step occurs and the molecule can re-enter the cycle or enter the core pathway following a transamination step. (B) Synthesis of the core methylthio GSL structure. The first enzymatic step has side chain specificity with CYP79F1 that converts both short- and long- chained methionine   derivatives   (n = 1–6) to the corresponding aldoxime whereas CYP79F2 only takes the long-chained   methionine   derivatives   (n = 5–6). See main texts for more details. (C) Secondary modifications. Short- and long-chained methylthio glucosinolates can be secondarily modified to methylsulfinyl glucosinolates in Col-0 leaves. In the seeds, the short-chained methylsulfinyl glucosinolates can be further modified to hydroxy form by AOP3 and benzoyloxy form by BZO1. Characterized enzymes in the pathway are noted next to the reaction arrows. This figure was adapted from Sønderby et al., (2007). Most genes in the pathway have been identified and characterized in the model plant

Arabidopsis thaliana (Sønderby et al., 2010). Modification of the levels of specific GSLs in

crop plants has been a strong interest as certain GSLs have desirable properties in flavour,

insect protection, biofumigation, and cancer prevention, whereas others have undesirable

properties. To date, metabolic engineering of GSL profiles has included altering the

expression of one or more CYP79 enzymes. Identification of the CYP79s as the enzymes

catalyzing the conversion of amino acids to aldoximes has provided important molecular

42

tools for modulating the profile of GSLs (Mikkelsen et al., 2002). Recently, the introduction

of the seven-step pathway of indolyl GSL from Arabidopsis thaliana to Saccharomyces

cerevisiae resulted in the first successful production of GSLs in a microbial host (Mikkelsen et

al., 2012). This production of indolyl GSL serves as a proof-of-concept for the expression

platform, and establishes a basis for large-scale microbial production of GSLs of interest.

Recently, stable genetic transfer of the six-step benzyl GSL pathway from A. thaliana to

Nicotiana tabacum (tobacco) leads to the production of benzyl GSL without causing

morphological alterations (Moldrup et al., 2012). Crucifer-specialist insect herbivores, like

the economically important pest Plutella xylostella (diamondback moth), frequently use GSLs

as oviposition stimuli. The transfer of a GSL biosynthetic pathway to a non-crucifer is likely to

stimulate oviposition on an otherwise non-attractive plant (Moldrup et al., 2012). In addition,

broccoli with a greater content of the glucoraphanin precursor of anti-carcinogenic ITC called

sulforaphane has been bred successfully using plant breeding methods (Mithen et al., 2003).

Three segments of the genome of a wild relative of broccoli Brassica villosa containing

greater concentrations of aliphatic GLS than commercial broccoli cultivars were introgressed

into commercial broccoli cultivar Marathon. More recently, a commercial-quality broccoli

(Beneforté®; Seminis Vegetable Seeds, Inc.) was bred to include Brassica villosa genetics

with two- to three-fold higher glucoraphanin content compared with commercial standard

hybrids (Vissavajjhala et al., 2011). A benefit of this high-glucoraphanin variety is that it

enables nutritional intervention studies that concentrate on a specific phytonutrient from a

whole food (James et al., 2012).

1.3 Degradation of GSL and its degradation products Most GSLs are chemically and thermally stable and thus degradation is mainly

enzymatically driven. Myrosinase co-occurs with GSLs and is involved in the enzymatic

degradation of these compounds. In A. thaliana, myrosinase is found in specialised cellular

compartments called myrosin cells separated from GSLs that are stored primarily in sulfur

rich cells (S-cells) located in close vicinity to the phloem (Koroleva et al., 2000; Koroleva et al.,

2010). It still remains unknown whether this type of compartmentation is common in other

cruciferous plants. The location of myrosinase in the cytoplasm of specialised myrosin cells

scattered throughout the plant tissue was proven by histochemical and immunological

studies (Kelly et al., 1998). Upon tissue disruption, myrosinase and GSL come into contact,

43

causing degradation of the β-thioglucosidic bond, and subsequently releasing glucose and an

unstable aglycone intermediate, the thiohydroxamate-O-sulfonate (Figure 1.4). The default

products are the isothiocyanates (ITCs) unless specifier proteins are present in which case

nitriles (NITs) and elemental sulfur, thiocyanate (TC), epithionitrile (ETN) can be formed

(Kissen et al., 2009; Bones et al., 2006). The scheme of GSL degradation under various

conditions is shown in Figure 1.4.

Figure 1.4 GSL hydrolysis under different conditions. In the first step, myrosinase-catalyzed degradation yields glucose and an unstable aglycone. In the absence of specifier proteins, the aglycone rapidly rearranges to an ITC. By contrast, the formation of epithionitriles, nitriles and thiocyanates depends on the chemical nature of the GSL side chain, and involves the action of an additional protein under physiological conditions. ESP, epithiospecifier protein; NSP, nitrile specifier protein; R, variable side chain; TFP, thiocyanate forming protein. This figure was modified from Tripathi & Mishra (2007).

The GSL degradation product formed by the reaction depends largely on the

chemical structure of the side chain. For example, indole GSLs produce only NITs and

unstable ITCs that rapidly form non-volatile indolylcarbinols (Bones & Rossiter, 1996; Burow

et al., 2006; Mithen, 2001; De Vos et al., 2005). On the other hand, aliphatic GSL degradation

yields volatile and pungent ITCs. The pH, the concentration of ferrous ions, and the presence

of epithiospecifier protein (ESP) or epithiospecifier modifier protein (ESM) are also

important factors contributing to the outcome of the myrosinase-GSL reaction (Burow et al.,

2006; Lambrix et al., 2001).

At pH 6-7, ITC production is favoured and is derived from the aglycone intermediate

by a Lossen rearrangement involving the migration of the side chain from the oxime carbon

44

to the adjacent nitrogen. The formation of NITs in vitro is favoured at a pH 3-4 in the

presence of Fe2+ ions (Uda et al., 1986). However, protein factors may be involved in NIT

formation in vivo, such as epithiospecifier modifier protein (ESM) and/or nitrile specifier

protein (NSP) (Foo et al., 2000; Bernardi et al., 2000; Kissen & Bones, 2009). When the GSL

side chain has a terminal double bond, epithiospecifier protein (ESP) promotes the reaction

of  the  sulfur  atom  of  the  β-thioglucosidic bond with the double bond to form a thirane ring,

giving an ETN (Figure 1.4). The formation of thiocyanate (TC) can only be derived from three

GSLs namely sinigrin, glucotropaeolin and glucoraphanin and is thought to be facilitated by

thiocyanate forming protein (TFP) (Wittstock & Burow, 2007).

Many endogenous factors can affect plant myrosinase activity and product formation

or the activity of ESP. Myrosinase activity was increased in some plant species with low

concentrations of ascorbic acid, but it was inhibited at higher concentrations (Botti et al.,

1995). During storage and processing of Brassica vegetables, cell damage occurs,

accompanied by competing processes of degradation and de novo biosynthesis of specific

GSLs. This multiple set of parameters affecting the outcome of the degradation gives rise to

a complex profile of degradation products in these plants. In addition, the chemical structure

of the GSL products is important for their biological activity. Small changes to side-chain

structures can have significant effects. For example, while methylthioalkyl GSLs produce

volatile and pungent ITCs (the major flavour compound in salad rocket is erucin),

methylsulfinylalkyl GSLs (the next products in the biochemical pathway) produce non-

volatile ITCs with relatively mild flavours, such as those found in broccoli. Removal of the

methylsulfinyl group and addition of a double bond results in a volatile ITC. Finally, addition

of a hydroxyl group to 3-butenyl and 4-pentenyl GSLs results in the spontaneous cyclisation

of the unstable ITC and the production of a non-volatile product (Holst & Williamson, 2004).

1.4 Biochemistry of myrosinases Myrosinases,  as   the  only  group  of  β-thioglucosidases known in nature, use GSLs as

substrates. Most myrosinases including myrosinase MYR1 from Brassica napus hydrolyze

multiple GSL substrates (Chen & Halkier, 1999) however some myrosinases from Brassica

napus and Crambe abyssinica are highly specific (Bernardi et al., 2003; MacLeod & Rossiter,

1986). All plant myrosinases are encoded by a multigene family. Six genes encoding classical

myrosinases have been identified in A. thaliana genome of which only four are functional

45

(TGG1-2 and TGG 4-5), and two (TGG3 and TGG6) are encoded by pseudogenes (Xu et al.,

2004; Wang et al., 2009). Twenty myrosinase genes or more were found in Brassica napus

and Sinapis alba and were grouped into families A and B (Rask et al., 2000). Some of which

have non-overlapping expression patterns (James & Rossiter, 1991; Lenman et al., 1993; Xue

et al., 1993). Biochemical characterisation of two myrosinases (previously known as

myrosinase I and II) identified from Brassica napus revealed the different features between

them in terms of their molecular weight and glycosylation level (myrosinase II had a smaller

number of glycosylated residues as compared with myrosinase I) (James & Rossiter, 1991).

Myrosinase is an abundant constitutively expressed protein in some cruciferous seeds, and is

easily obtainable by simple ammonium sulfate fractionation followed by ion-exchange

chromatography.

Classical myrosinases belong to a phylogenetically distinct group within glycosyl

hydrolase family 1 (GH1) and are thought to have evolved from -O-glucosidase ancestors

(Xu et al., 2004). This enzyme is a homodimer that contains three disulfide bridges, two of

which are important for stabilizing the N-terminus of myrosinase (Rask et al., 2000). In

addition, salt bridges, hydrogen bond and high carbohydrate content up to 20% are also

believed to provide stability to the myrosinases enabling the enzyme to carry out GSL

degradation in an extracellular environment without being inactivated by its reactive

products. A unique feature of plant myrosinase is its activation by ascorbic acid. Early work

had shown that ascorbic acid created an allosteric effect on the activity of the enzyme

(Ohtsuru & Hata, 1973; Ohtsuru & Hata, 1979). However, more recent studies have shown

that ascorbic acid is rather a catalytic base, and GSL degradation can be effectively promoted

in the absence of ascorbic acid, but at a much slower rate (Burmeister et al., 1997; Bones &

Rossiter, 2006).

The elucidation of the mechanism of plant myrosinase has been aided by x-ray

structural analysis of white mustard myrosinase (Burmeister et al., 1997). The use of a 2-F-2-

deoxybenzyl GSL inhibitor to irreversibly inhibit myrosinase by forming a covalent link

between the C-1 of the glycosyl unit and Glu 409 has enabled the identification of a catalytic

Glu 409 and a Gln 187 in the active site (Cottaz et al., 1996). This myrosinase is a dimer

linked by a zinc atom and has a characteristic (β/α)8-barrel structure which act through a

mechanism that gives retention of the anomeric configuration (Cottaz et al., 1996).

46

The myrosinase-GSL system has been extensively studied in plants (Wallsgrove &

Bennett, 1995; Halkier & Du, 1997). It was not until 2001 when the non-plant myrosinase

was first fully characterized (Jones et al., 2001). The cabbage aphid B. brassicae was shown

to produce a myrosinase capable of hydrolysing several common plant GSLs e.g. sinigrin and

glucotropaeolin (Francis et al., 2002; Jones et al., 2002). This myrosinase was purified, and

an x-ray structural determination was carried out (Huseby et al., 2005; Jones et al., 2002;

Jones et al., 2001). Sequencing has shown that this myrosinase has significant sequence

similarity (35%) to plant myrosinases and other members of GH1 (Jones et al., 2002). In

common with plant myrosinase, aphid myrosinase has the characteristic (β/α)8-barrel

structre. The residues acting as a proton donor and a nucleophile in GSL degradation by

aphid myrosinase are identified as Glu 167 and Glu 374, respectively. Gln 187 and Glu 409

are the equivalent residues in plant myrosinase, and Glu 183 and Glu 397 for the cyanogenic

β-glucosidase. Assumingly, a proton donor is necessary for GSL degradation in the case of

aphid myrosinase, but not for plant myrosinases. Unlike plant myrosinase, aphid myrosinase

does not require ascorbic acid for its activity and it appears to be more similar to animal β-O-

glucosidases than to plant myrosinase. This assessment was determined by sequence

similarity and phylogenetic techniques (Jones et al., 2002).

1.5 Specifier proteins

Except for ITCs, the biological roles of the other GSL degradation products are not

well understood. However, previous studies indicate that they may act in direct and indirect

defense (Wittstock et al., 2003; Wittstock & Burow, 2010). Other than ITCs, other products

namely NITs, ETNs and organic TCs can be formed by the rearrangement of the aglycone

promoted  by  supplementary  proteins  called  ‘specifier  proteins’  (Tookey, 1973).

Thus far, nine plant specifier proteins have been identified and characterized

biochemically with different substrate and product specificities at the molecular level

(Kuchernig et al., 2012). Specifier proteins were first discovered in Crambe abyssinica

(Brassicaceae) and found to have effect on the degradation products without having

hydrolytic activity on GSLs themselves (Tookey, 1973). A vast stride of progress has been

made within the past decade in identifying the role of specifier protein. The gene that

encodes for epithiospecifier protein (ESP) was first identified in A. thaliana ecotype

Landsberg erecta (Ler) (Lambrix et al., 2001). This stable ESP was found to redirect the

47

conversion of aglycones from ITCs by myrosinases towards the formation of simple NITs and

ETNs (Burow et al., 2006) rather than being an allosteric co-factor of myrosinase as proposed

earlier (Petroski & Kwolek, 1985). Thiocyanate forming protein (TFP), which has been

identified from Lepidium sativum, shares 63-68% amino acid sequence identity with known

ESPs and up to 55% identity with myrosinase-binding proteins from A. thaliana (Burow et al.,

2006). Despite the similarities in sequence identity between ESP and TFP, it is not known

whether both of these specifier proteins are derived from a common ancestor (Burow &

Wittstock, 2009). Recently, a recombinant TFP enzyme from Thlaspiarvense arvense

(Brassicaceae) was purified in active form in E. coli (Kuchernig et al., 2011). Interestingly, this

protein promotes the formation of allyl thiocyanate as well as the corresponding ETN upon

myrosinase-catalyzed degradation of allyl GSL, the major GSL of Thlaspiarvense arvense. All

other GSLs tested were converted to their simple NITs when hydrolyzed in the presence of

this protein. In contrast with A. thaliana ESP, TFP in vitro activity is not strictly dependent on

Fe²⁺ addition to the assay mixtures (Kuchernig et al., 2011).

Another supplementary protein namely nitrile specifier protein (NSP), capable of

redirecting degradation of GSLs to NITs, has also been identified. It was first found in the

larvae of the butterfly Pieris rapae and was also identified in A. thaliana (Burow et al., 2006;

Kissen & Bones, 2009). NSP promote simple NIT formation at physiological pH values, but do

not catalyse ETN or TC formation (Burow & Wittstock, 2009; Kissen & Bones, 2009). In A.

thaliana, six specifier proteins have been identified so far including one ESP (Lambrix et al.,

2001) and five NSPs (Burow & Wittstock, 2009; Kissen & Bones, 2009).

1.6 Importance of GSLs and ITCs to human health

1.6.1 Chemopreventive effects of GSLs

It has been shown that cancer is 30–40% preventable over time by appropriate food

and nutrition (American Institute for Cancer Research 2007). Accumulating evidence on

plant foods and cancer prevention suggests the potentially beneficial effect of GSLs which

are abundant in cruciferous vegetables (Lynn et al., 2007; Ambrosone & Tang, 2009; Mithen

et al., 2010; Traka & Mithen, 2009; Bosetti et al., 2012). Epidemiologic studies published

prior to 1996 revealed that the majority (67%) of 87 case-control studies showed an inverse

48

correlation between some type of cruciferous vegetable intake and cancer risk (Verhoeven

et al., 1996). In the past decade, results of large prospective cohort studies and studies

taking into account individual genetic variation suggest that the relationship between

cruciferous vegetable intake and the risk of several types of cancer is more complex than

previously thought (Higdon et al., 2007). To date, the inverse association appeared to be

most consistent for cancers of the lung and digestive tract. Evidence of chemopreventive

effects of cruciferous vegetables for lung, colorectal and prostate cancer are summarized

below:

Lung Cancer: Several case-control studies demonstrated that people diagnosed with

lung cancer had significantly lower intakes of cruciferous vegetables than people in cancer-

free control groups (Verhoeven et al., 1996). Prospective studies of Finnish men (Miller et al.,

2004) and Dutch men and women (Feskanich et al., 2000) indicated that higher intakes of

cruciferous vegetables (more than three weekly servings) were associated with significant

reductions in lung cancer risk.

Colorectal Cancer: A small clinical trial found that the consumption of 250 g/d (9

oz/d) of broccoli and 250 g/d of Brussels sprouts significantly increased the urinary excretion

of a potential carcinogen found in well-done meat, namely 2-amino-1-methyl-6-

phenylimidazo[4,5-b]pyridine (PhIP) (Walters et al., 2004). A prospective study of Dutch

adults found that men and women with the highest intakes of cruciferous vegetables

(averaging 58 g/d) were significantly less likely to develop colon cancer than those with the

lowest intakes (averaging 11 g/d) (Voorrips et al., 2000).

Prostate Cancer: To date, epidemiological studies provide only modest support for

the speculation that high intakes of cruciferous vegetables reduce prostate cancer risk

(Kristal & Lampe, 2002). Four out of eight case-control studies published since 1990 found

that some measure of cruciferous vegetable intake was significantly lower in men diagnosed

with prostate cancer than men in a cancer-free control group (Cohen et al., 2000; Jain et al.,

1999; Joseph et al., 2004; Kolonel et al., 2000).

At present, the putative role on cancer chemoprevention of cruciferous vegetables is

attributed to the bioactivity of GSL degradation products, mainly ITCs. ITCs have been shown

to protect against the most common cancer types, such as breast, lung, colon and prostate

49

cancers in both in vivo and in vitro studies (Bianchini & Vainio, 2004; Hayes et al., 2008;

Keum et al., 2005; Wu et al., 2011; Nakamura, 2009; Yang et al., 2010; Lai et al., 2010;

Cheung & Kong, 2009; Kim et al., 2011; Sehrawat & Singh, 2011; Cheung et al., 2008;

Powolny et al., 2011; Wang et al., 2010; Yin et al., 2009). Following metabolism of GSLs into

ITCs in vivo, the structural difference of GSLs is conferred to that of the cognate ITCs. For

example, glucoraphanin is hydrolyzed by plant myrosinase to sulforaphane (SFN), sinigrin to

allyl ITC (AITC), gluconasturtiin to phenethyl ITC (PITC), glucotropaeolin to benzyl ITC (BITC),

glucoerucin to erucin (ERN), glucoiberin to iberin (IBR) and glucoiberverin to iberverin (IBV).

The summary table of the expected ITC or NIT degradation products from GSL hydrolysis

catalyzed by plant myrosinase is shown in Table 1.2.

Table 1.2 Expected ITC or NIT product from GSL hydrolysis catalyzed by plant myrosinase

GSL substrate ITC product NIT product

Sinigrin (SNG) Allyl isothiocyanate (AITC) Allyl nitrile (ANIT)

Glucotropaeolin (GTP) Benzyl isothiocyanate (BITC) Benzyl nitrile (BNIT)

Gluconasturtiin (GNT) Phenethyl isothiocyanate (PITC) Phenethyl nitrile (PNIT)

Glucoerucin (GER) Erucin (ERN) Erucin nitrile (ERN NIT)

Glucoiberin (GIB) Iberin (IBR) Iberin nitrile (IBR NIT)

Glucoraphanin (GRP) Sulforaphane (SFN) Sulforaphane nitrile (SFN NIT)

Several mechanisms for the activities of ITCs in different stages in cancer

chemoprevention have been proposed (Figure 1.5).

50

Figure 1.5 Mechanisms of action of ITCs in the modulation of signalling pathways involved in cancer chemoprevention. CYP = Cytochrome; GST = Glutathione-S-transferase; UGT = UDP-glucuronosyl transferase; NF-κB = Nuclear factor kappa B; HDAC = Histone deacetylase; miRNA = micro RNA. This figure was modified from Navarro et al., (2011).

This includes inhibition of phase I carcinogen activating enzymes such as CYP

enzymes (Yoxall et al., 2005; Conaway et al., 1996), induction of phase II carcinogen

detoxification enzymes such as glutathione-S-transferase (Talalay, 2000; Nakamura et al.,

2000), induction of cell cycle arrest (Gamet-Payrastre et al., 2000; Singh et al., 2004) and

apoptosis (Xiao et al., 2003). ITCs are known to modulate a large number of important

cancer-related proteins (Figure 1.6) through various mechanisms.

51

Figure 1.6 ITCs modulate a large and diverse group of proteins. The modulation of many proteins by ITCs involves direct reaction of the –N=C=S groups of ITCs with cysteine thiols of the proteins. Macrophage migration inhibitory factor is modulated via its amino group. The mechanisms for modulation of many other proteins are not yet known. See main texts for details. This figure was modified from Zhang (2011).

For examples, inhibition of cytochrome P450 (CYP) enzymes (von Weymarn et al.,

2007), induction of phase II enzymes via activation of NF-E2-related factor-2 (Nrf2) (Zhang &

Hannink, 2003; Thimmulappa et al., 2002), inhibition of histone deacetylases (HDACs) (Wang

et al., 2008; Myzak et al., 2006), inhibition of membrane drug transporters (Callaway et al.,

2004, Ji & Morris, 2005), modulation of cell cycle regulators and Bcl-2 family proteins (Geng

et al., 2011; Xiao et al., 2006; Zhang & Tang, 2007) activation of caspases (Park et al., 2007;

Wu et al., 2005), downregulation of α-/β-tubulins and/or inhibition of tubulin polymerization

(Mi et al., 2009; Mi & Chung, 2010), downregulation of vascular endothelial growth factor

(VEGF) (Boreddy et al., 2011) and its receptor and inhibition of nuclear factor kappa B (NF-

κB) (Xu et al., 2005), activator protein-1 (AP-1) (Li & Zhang, 2005; Gopalakrishnan & Tony

Kong, 2008), mitogen-activated protein kinase kinase 1 (MEKK1) (Cross et al., 2007), signal

52

transducer and activator of transcription factor 3 (STAT3) (Gong et al., 2009) and Toll-like

receptor 4 (TLR4) (Youn et al., 2010).

Most of the activities mentioned above are shared by different ITCs (Munday &

Munday, 2004). The reactivity of ITC lies in its electrophilic nature and its tendency to

undergo facile addition reactions with N-, O-, or S-based nucleophiles. At least some of their

chemopreventive mechanisms are all activated through direct reaction of the carbon atom

of the –N=C=S group of ITCs with the cysteine sulfhydryl groups of glutathione (GSH) and

proteins (Drobnica et al., 1975; Zhang, 2000). The side chains of ITCs may play secondary

roles, by influencing the electrophilicity of the –N=C=S group, altering the access to the

reactive carbon through steric effects and controlling the lipophilicity of the whole molecule

(Zhang, 2011). This explains the intriguing capability of ITCs to target a diverse group of

proteins, and the phenomenon that different ITCs often share similar biological activities and

metabolic profiles. The summary of the expected ITC products from the hydrolysis of GSLs by

plant myrosinase and corresponding chemopreventive properties of ITCs is shown in Table

1.3.

Table 1.3 Chemopreventive actions of ITCs produced from GSL hydrolysis

ITC Chemopreventive mechanisms of ITC

AITC from sinigrin

Source: brussels sprouts,

cabbage, cauliflower,

kale, mustard,

horseradish and wasabi

Proliferation inhibition of various types of human cancer cells

through cell cycle arrest and/or induction of apoptosis (Zhang &

Hannink, 2003; Xiao et al., 2003; Musk & Johnson, 1993; Tang &

Zhang, 2004), inhibition of cell adhesion, migration and invasion

(Hwang & Lee, 2006) and stimulation of histone acetylation (Lea

et al., 2001), activation of Nrf2, phase II genes (Munday et al.,

2006; Zhang et al., 1998), and also inhibition of Helicobacter

pylori (Shin et al., 2004) and Escherichia coli O157:H7 (Luciano et

al., 2009).

BITC from

glucotropaeolin

Anti-cancer effects in both in vivo and in vitro experimental

models (Hwang et al., 2008; Nakamura, 2009; Hecht, 1999),

activation of DNA damage, causes growth 2/mitosis (G2/M) cell

53

Source: Papaya (Carica

papaya) and garden

cress

cycle arrest and apoptosis (Zhang et al., 2006), inhibition of tissue

plasminogen activator (TPA)-induced oxidative stress through

inhibition of reduced nicotinamide adenine dinucleotide

phosphate (NADPH) oxidase and leukocyte infiltration (Nakamura

et al., 2004), inhibition of tumor necrosis factor alpha (TNF-α)-

induced matrix metallopeptidase 9 (MMP-9) secretion by

downregulation of NF-κB and AP-1 (Lee et al., 2009).

PITC from

gluconasturtiin

Source: Watercress and

garden cress

Activation or inhibition of various cellular signaling pathways such

as phase I/II detoxification modification enzymes (Munday &

Munday, 2004), Nrf2 and Kelch-like ECH-associated protein 1

(Keap1) (Itoh et al., 2003) and NF-κB   (Jeong et al., 2004),

induction of cell cycle arrest by reduction of cyclin-dependent

kinase 1 (cdk1) and cell division cycle 25c (cdc25c) (Xiao et al.,

2004).

SFN from glucoraphanin

Source: Broccoli,

Broccoli sprouts,

cabbage, Brussel sprouts

Strong antitumor activities in vitro and in vivo (Conaway et al.,

2005; Fimognari & Hrelia, 2007; Chiao et al., 2002; Singh et al.,

2004; Jakubikova et al., 2005). Induction of phase II detoxification

gene expression through the Nrf2 or the antioxidant response

element (ARE) pathway (Sibhatu et al., 2008), reduction of the

number of polyps through suppressing mitogen-activated protein

kinase (MAPK) signalling (Bertl et al., 2006), suppression of

lipopolysaccharide (LPS)-induced cyclooxygenase-2 (COX-2)

expression, downregulating NF-κB, CCAAT/enhancer binding

protein (C/EBP), cAMP response element-binding protein (CREB)

and AP-1 (Woo & Kwon, 2007). Treatments for gastritis and

stomach cancer caused by Helicobacter pylori (Fahey et al., 2002).

ERN from glucoerucin

Source: Rocket slads

ERN may also exert its potential protective effects against human

cancer through multiple mechanisms similar to those triggered by

SFN (Melchini & Traka, 2010) e.g. induction of phase II enzymes

e.g. QR and GST in rat and human tissues (Hanlon et al., 2009;

Munday & Munday, 2004; Zhang et al., 1992), upregulation of

54

phase III detoxification system e.g. multidrug resistance-

associated protein 1/2 (MRP-1/2) in human colonic cancer CACO-

2 cells (Munday & Munday, 2004; Harris et al., 2008), induction of

tumour suppressor proteins (p53, p21), cell cycle arrest, pro-

apoptotic signals (Melchini et al., 2009; Fimognari et al., 2004;

Jakubikova et al., 2005).

IBR from glucoiberin

Source: Horseradish,

mustard

There are fewer studies on both IBR and IBV in comparison to

SFN. IBR increased glutathione S-transferase (GST) and quinone

reductase (QR) activities in the urinary bladder of the rats

demonstrating protective effects against chemical carcinogenesis

(Staack et al., 1998). IBR also led to induction of phase II enzymes

e.g. thioredoxin reductase (Barrera et al., 2012; Wang et al.,

2005), induction of apoptosis and cell cycle arrest (Jadhav et al.,

2007; Jakubikova et al., 2006).

IBV from glucoiberverin

Source: Horseradish,

mustard

Induction of phase II enzymes e.g. QR and GST in a variety of rat

tissues (Munday & Munday, 2004; Kim & Singh, 2009). However,

the anticancer effects of IBR and IBV on the tumor cells have not

been investigated in detail.

The biological effect of the above ITCs varies due to the side-chain structure. For

example, in vitro studies have shown that SFN is taken into cells faster, kept intracellularly

longer, and at higher accumulations than several other ITCs (Zhang & Talalay, 1998; Ye &

Zhan, 2001). SFN also has the highest potency of inducing the expression of two phase II

enzymes, QR and GST (Zhang & Talalay, 1998; Vermeulen, 2009). In contrast, AITC was

shown to be most effective in causing HL60 (human promyelocytic leukemia cells) cell cycle

arrest (Jakubikova et al., 2005) while PITC and BITC were the most effective in inducing

apoptosis, among six different ITCs (Munday et al., 2008). Based on the study of the effect of

ten synthetic ITC analogues on pro-inflammatory NF-κB  activity  in vitro, it was reported that

subtle changes in ITC structure had a profound impact on inhibition potential (Prawan et al.,

2009). Therefore, in addition to the amount consumed, the variety of cruciferous vegetables

ingested may also influence biological response. However, differential effects of cruciferous

55

vegetables with diverse GSL profiles have not been directly compared in vivo in humans

(Navarro et al., 2011).

To date, a major impediment to our understanding of the chemopreventative

mechanisms stimulated by GSLs is that relatively little is known about the biological effects

of GSL breakdown products other than ITCs and the indole-containing derivatives.

Specifically, there are little data about chemopreventative activities of TCs, NITs, cyano-

epithioalkanes and oxazolidine-2-thiones. Our group showed that 3-butenyl-ETN showed

some toxicity at high dose (0.1 mM) on MCL-5 cells and cHo1 cells. Other products including

2-propenyl-ETN, 2-propenyl NIT and 3-butenyl NIT were, at most, marginally toxic (Nurul

Huda Binti Abd Kadir, PhD thesis). ANIT has also been shown to induce antioxidant and

detoxification enzymes (Tanii et al., 2008). The question of whether ETN and NIT products of

GSL degradation also have chemopreventative properties still remains unanswered. Also, it

is unclear whether the formation of TCs, NITs, cyano-epithioalkanes and oxazolidine-2-

thiones from GSLs, at the expense of forming ITCs, is undesirable from a cancer

chemoprevention perspective (Hayes et al., 2008). These are areas that warrant further

examination.

1.6.2 Preventive effects against diseases

Unlike most small molecule pharmacological agents that affect single targets, the

intracellular targets of ITCs are multiple. For example, activation of transcription factor Nrf2

alone caused by ITC exposure leads to an orchestrated upregulation of an extremely large

network of genes with cytoprotective, antioxidant, and anti-inflammatory functions. It is this

ability to induce versatile and long-lasting responses, which ultimately protects against

oxidative stress, electrophilic stress, and chronic inflammation (the three main underlying

causes of most chronic diseases) that makes ITCs exceedingly efficient protective agents

(DinkovaKostova et al., 2012). Not only chemopreventive ability, but ITCs especially SFN also

has the potential to reduce the risk of diabetes (Cui et al., 2012; Miao et al., 2012; Xue et al.,

2008), atherosclerosis (Kwon et al., 2012; Kivela et al., 2010), respiratory diseases (Ritz et al.,

2007; Riedl et al., 2009), neurodegenerative disorders (Ping et al., 2010; Dash et al., 2009),

ocular disorders (Kong et al., 2007; Gao et al., 2004), and cardiovascular diseases (Zakkar et

al., 2009; Angeloni et al., 2009).

56

It is known that oxidative stress is a link between cardiovascular risk factors and

vascular disease, and thus a target for cardiovascular prevention. SFN exhibits cytoprotective

effects i.e. increase in cell viability and decline in DNA fragmentation in neonatal cardiac

myocytes. This is mediated by the increase in the expression of multiple antioxidant proteins

and reduction of reactive oxygen species (ROS) production (Angeloni et al., 2009). In

addition, SFN was shown to induce the activity of antioxidants and phase II enzymes such as

catalase, superoxide dismutase, glutathione peroxidase, glutathione reductase, glutathione

S-transferase, NAD(P)H:quinone oxidoreductase-1 and glutathione in rat aortic smooth

muscle cells and isolated mitochondria of aortic smooth muscle cells (Zhu et al., 2008). Other

ITCs have exhibited neuroprotective activity in either in vitro or in vivo models of neuronal

cell death or neurodegeneration, respectively (Kelsey et al., 2010). They are directly able to

scavenge free radicals or indirectly increase endogenous cellular antioxidant defenses via

activation of the Nrf2 transcription factor pathway. Interestingly, SFN was shown to prevent

metabolic dysfunction in an in vitro model of hyperglycemia by preventing the increasing

cellular accumulation of the glycating agent methylglyoxal (Xue et al., 2008).

1.6.3 Genotoxicity of ITCs

Genotoxicity refers to the ability of chemicals to damage DNA and/or cellular

components regulating the genome fidelity such as the topoisomerases, spindle apparatus,

DNA polymerases and DNA repair systems (Wobus & Löser, 2011). All adverse effects on

genetic information are also included (Fimognari et al., 2011).

While normal consumption of cruciferous vegetables seems to be beneficial to

human health, any large increase in intake could conceivably lead to undesirable effects. A

partial overlapping between the chemopreventive doses of ITCs and those exhibiting

genotoxic potential can be seen for most ITCs (Fimognari et al., 2011). For example, BITC

exerts in vitro protective effects against cancer development in the range 0.01–50  μM  and  

genotoxicity even at   the  dose  0.1  μM   (Kassie et al., 1999). PITC exhibits cancer protective

effects in the interval 0.1–100  μM  and  the  genotoxic  effects  in  the  interval  2–613  μM  (Musk

et al., 1995). The same is true for SFN with the in vitro cancer protective effects appear in

the interval 0.1–2,000   μM,   the   neuroprotection   in   the   range   0.01–10,000   μM   and   the  

genotoxic effects in the interval 10–140  μM  (Musk et al., 1995). As far as the in vivo studies

57

go, only AITC exhibits protective effects at levels of doses lower than those that resulted in

genotoxicity (Kassie et al., 2000). Particular attention should be paid to BITC and PITC as they

show a clear genotoxic potential at doses endowed with a chemopreventive potential. The

mechanisms of genotoxicity are complex and some of them very specific and linked to the

presence of the ITC group. SFN analogues with oxidized sulfur (e.g. erysolin) are potent

inducers of ROS. However, ERN, at the same dose levels, had no such effect (Kim et al., 2010).

It was concluded that some ITCs possess a genotoxic activity, but not all. The

potential threats associated with the intake of different ITCs are characterized by a different

toxicological profile. These findings need to be verified and extended by further studies. The

genotoxic effects of ITCs in vivo are needed to be studied as thus far in vivo data are only

available for AITC, BITC, PITC and methyl ITC. However, it is highly unlikely that such

toxicities would occur in humans, because dietary consumption levels of those ITCs and ITC

exposure in humans appear to be several orders of magnitude lower than the doses found in

the animal studies (µmol vs mmol, repectively) (Fimognari et al., 2011).

1.7 Bioavailability of GSL degradation products in humans

When brassica vegetables are cooked, a partial or total inactivation of myrosinase

can occur. Various factors also lead to variations in bioavailability of GSLs and GSL

degradation products (Dekker et al., 2000). These changes are primarily resulted from the

type of vegetable matrix, the extent of its cellular disruption, the duration and method of

cooking and the chemical structure of the GSL precursors (Rungapamestry et al., 2007).

Interestingly, it was shown that the overall average bioavailability of ITCs is 61% and 10% for

raw and cooked cruciferous vegetables, respectively (Vermeulen et al., 2006; Melchini &

Traka, 2010). After consumption of cooked brassica with heat inactivated plant myrosinase,

GSLs are hydrolyzed in the colon action by the gut microbiota. Feeding trials with human

subjects have shown that degradation of GSLs and absorption of ITCs are greater following

consumption of raw brassica with active plant myrosinase than after ingestion of the cooked

plant with inactive myrosinase (Rungapamestry et al., 2007). The digestive fate of GSLs may

be further influenced by the extent of cell rupture during ingestion, gastrointestinal (GI)

transit time, meal composition, individual genotype and differences in colonic microbiota.

The differences in epidemiological evidence relating preventive effects of brassica

58

consumption against cancer in man, especially in cohort studies (International Agency for

Research on Cancer, 2004) may be partly explained by these causes of variation, coupled

with differences between individuals in the metabolism of isothiocyanates (Seow et al.,

2005). A better knowledge of the digestive and absorptive fate of dietary GSLs and their ITC

metabolites has emerged mostly from mechanistic studies in animal models such as rats and

hamsters (Brusewitz et al., 1977; Mennicke et al., 1983; Michaelsen et al., 1994; Duncan et

al., 1997; Elfoul et al., 2001). The low recoveries of intact GSLs and their metabolites in

faeces of animals fed GSLs or ITCs suggest that a substantial proportion of ingested GSLs and

ITCs are metabolized in vivo (Slominski et al., 1988; Bollard et al., 1997; Rouzaud et al.,

2003). However, there is little information on the degradation of GSLs during and after

consumption of brassica vegetables by human subjects. The complexity of the GSL–

myrosinase system and the host of factors are likely to influence the degradation of GSLs in

vivo, this information is important in understanding the physiological consequences of

brassica consumption. Studies with human subjects have used urinary biomarkers to assess

the absorption of ITCs after the intake of GSLs from brassica vegetables (Getahun et al.,

1999; Conaway et al., 2000; Shapiro et al., 2001; Rouzaud et al., 2004). Following their

absorption into the intestinal epithelium, ITCs are released into the systemic circulation and

metabolized by the mercapturic acid pathway in the liver (Rungapamestry et al., 2007). ITCs

initially form conjugates with glutathione, then undergo enzymic modification and are

excreted in urine as their corresponding N-acetylcysteine (NAC) conjugates i.e. mercapturic

acid (MA) (Brusewitz et al., 1977; Mennicke et al., 1983). Urinary isothiocyanate mercapuric

acid (ITC MA) excretion therefore partially reflects ITC absorption in vivo, although variation

in pre- and post-absorptive recovery may also be important. The metabolic fate of GSLs and

ITCs following consumption of brassica within experimental meals by volunteers is shown in

Figure 1.7. Thus far, only the ITC MAs have been studied as urinary biomarkers of GSL

degradation in vivo. This approach has provided a reasonable understanding of the overall

uptake of ITCs after consumption of brassica by human subjects (Chung et al., 1998;

Mennicke et al., 1988).

59

Figure 1.7 Expected metabolic fate of glucoraphanin and AITC following ingestion of cooked broccoli (Brassica oleracea var. italica) and mustard (Sinapis alba) respectively by human volunteers. Following absorption, the ITC are metabolized by the mercapturic acid pathway. Initially, they are conjugated with glutathione in a glutathione transferase (GST)-catalyzed reaction and finally N-acetylcysteine conjugates (mercapturic acids) are formed. This figure was modified from Rungapamestry et al., (2007).

The metabolism of GSL degradation products and their bioavailability may be

influenced by inter-individual variations. This hypothesis is generally poorly understood and

controversial. The absorption of ITCs has been shown to vary, although not markedly,

between individuals. Low inter-individual variation has been observed after the ingestion of

AITC in the form of mustard, with recovery of 60–90% of the administered doses of AITC

(Rouzaud et al., 2004). Inter-individual variation is 1.5-fold greater after the ingestion of

cooked watercress than after raw watercress (Getahun et al., 1999), indicating that there

may be inter-individual differences in the colonic degradation of GSLs, as well as other

factors such as gastrointestinal transit time, extent of chewing and genotype.

Polymorphisms in genes coding for the activity of GST isoforms have been described (Seow

et al., 2005), and they may explain inter-individual variation in the metabolism and excretion

of ITCs. The role of GSTs involves detoxification of a wide range of electrophiles by

conjugation with glutathione. Particularly, glutathione S-transferase Mu 1 (GSTM1) was

shown to use ITCs as substrates (Kolm et al., 1995). Null genotypes for GSTM1 and

60

glutathione S-transferase Theta 1 (GSTT1) lead to the absence of their respective enzymes.

Therefore, ITC may be metabolized more slowly among GSTM1-null and GSTT1-null

individuals and consequently increase the tendency of upregulation of other GST isoenzymes

(Ketterer et al., 1998; Lin et al., 1998). Individuals with GSTM1- and/or GSTT1-null genotypes

have been shown to be at reduced risk of developing colorectal and lung cancers after the

intake of ITCs (Keum et al., 2004). The low GST activity and a slower rate of excretion of ITCs

in these individuals may help retain more ITCs at target tissues (e.g. colon) and provide a

pro-apoptotic effect in situ (Johnson, 2004).

When myrosinase is deactivated by cooking, the ionised nature of GSLs may be

expected to enable them to reach the distal gut where they could be transformed by

bacterial enzymes. This speculation was first examined and confirmed by studies in which

antibiotic treatments were used to reduce the large bowel microbiota (Slominski et al.,

1987; Shapiro et al., 1998). Reducing the microbial digestion by a combination of mechanical

cleansing and antibiotics caused a dramatic reduction (8.7-fold) of ITCs as dithiocarbamate

excretion from 11.3 ± 3.1% of the initial dose to 1.3 ± 1.3% (Shapiro et al., 1998). As these

are human data, there seems to be no doubt about the importance of the gut microbiota in

digestive ITC formation.

More direct evidence was eventually obtained from gnotobiotic experiments. A

whole faecal microflora from rats or humans was introduced into initially germfree rats, and

resulted in the disappearance of intact GSLs in the cecal and colonic contents, coupled with

the emergence of systemic effects reflecting GSL degradation (Nugon-Baudon et al., 1988;

Rabot et al., 1993; Campbell et al., 1995). It appears that the ability to degrade GSLs is widely

distributed among intestinal bacteria (Oginsky et al., 1965). Representatives of various

genera (e.g. Bacteroides, Peptostreptococcus, Enterococcus, Escherichia, Proteus) have been

isolated from human faeces. These bacteria were able to degrade progoitrin and sinigrin in

vitro (Rabot et al., 1995). Bifidobacterium sp., B. pseudocatenulatum, B. adolescentis, and B.

longum showed the ability to digest GSLs, sinigrin and glucotropaeolin in vitro (Cheng et al.,

2004).

The formation of AITC from sinigrin in the digestive tract of rats mono-associated

with a human gut strain of Bacteroides thetaiotaomicron has been reported with the

possible involvement of bacterial myrosinases (Elfoul et al., 2001; Rouzoud et al., 2004). The

61

degradation of GSLs by the human colonic microbiota has been studied in some detail in a

dynamic in vitro large-intestinal model (Krul et al., 2002). The study was carried out with

sinigrin which was shown to be degraded to AITC. The degree of conversion depended on

the concentration of sinigrin used and the nature of the inoculum. Since not all the ITC was

accounted for, possibly further metabolism was indicated to unidentified metabolites.

As yet, little is known of the structure of microbial GSL derivatives. A detailed analysis

of GSL degradation by human digestive microbiota has been carried out using 1H NMR on

both sinigrin and glucotropaeolin. By using this technique, it was shown that allylamine and

benzylamine were exclusively produced from these GSLs, respectively (Combourieu et al.,

2001). However, this appears to conflict with the later report where only ITCs were detected

(Krul et al., 2002). Upon anaerobic incubation of cooked watercress juice with human faeces,

18% of GSLs are hydrolyzed to ITCs in 2 h (Getahun et al., 1999). The contribution of the

digestive microbiota to ITC production, in vivo and in the distal gut, has been ascertained.

Following gavage with 50 µmol sinigrin, substantial amounts of AITC (up to 100 nmol at 12 h

after dosing) were measured in the cecal and colonic contents of gnotobiotic rats harbouring

a human digestive strain of Bacteroides, while no allyl nitrile (ANIT) could be detected (Elfoul

et al., 2001). It may be speculated that NITs were formed, but were not detectable because

they were readily transformed into other metabolites. This hypothesis is supported by

observations on ANIT degradation in sheep rumen fluid (Duncan & Milne, 1992) and by

reports on the ability of various microorganisms to convert NITs to organic acids and

ammonia (Kobayashi & Shimizu, 1994).

The formation of other derivatives, e.g. desulfo-glucosinolates (DS-GSLs) and TCs, has

scarcely been investigated, and studies are often not conclusive, chiefly because of analytical

impediments (Slominski et al., 1988; Rowan et al., 1991). Nevertheless, the versatility of

microbial enzymatic activities would be expected to lead to a wider array of metabolites

than those so far identified. The post-absorptive fate of GSL derivatives other than ITCs has

received comparatively little attention. TCs may be converted to cyanide and thiol

derivatives by GSTs (Ohkawa et al., 1972) and ETNs may be excreted in the form of MAs

(Brocker et al., 1984).

To date, GSL degradation by the colonic microbiota, have been much less

investigated than GSL degradation by plant myrosinase in plant material or food products.

Therefore, microbial digestions of GSLs need to be fully addressed to improve an

62

understanding of bacterial GSL-degrading enzyme occurrence/activity and the bioavailability

of GSL degradation products in the human gut.

1.8 Human gut microbiota The human gut microbiota is a complex ecosystem (Tap et al., 2009) that is estimated

to be composed of approximately 1014 bacterial cells—which is ten times more than the

total number of human cells in the body (Savage et al., 1977). There are more than 800

species of bacteria in the human gut with 30–40 species dominating this community,

comprising up to 99% of the total population (Rouzaud et al., 2004). The gut microbiome, a

term for the collective community of bacteria and their total genome capacity in the human

gut, is approximately 150 times larger than the human gene complement. With an estimated

3.3 million microbial genes, it has been referred to as the 'forgotten organ' and a

'superorganism' (Qin et al., 2010; Gill et al.,   2006;   O’Hara   et al., 2006). Seven phyla

constitute the bulk of the gut microbiota, namely Firmicutes, Bacteroidetes, Proteobacteria,

Fusobacteria, Verrucomicrobia, Cyanobacteria and Actinobacteria. The Firmicutes and

Bacteroidetes phyla accommodate the most abundant species and constitute over 90% of

the human gut microbiota (Eckburg et al., 2005; Backhed et al., 2005; Tap et al., 2009). As

early as 1907, it was hypothesized that replacing or diminishing 'putrefactive' bacteria in the

gut with lactic acid bacteria could improve bowel health and prolong life (Metschnikoff et al.,

1907). Since the 1990s, interest has grown in the role that the human gut microbiota has in

disease. To date, metagenomics and human microbiome research have arrived at the

forefront of biology mainly due to major technical and conceptual developments (Devaraj et

al., 2013). One of the most important objectives in human microbiome research is to

understand the symbiotic relationship between gut microbes and their host and to find any

correlation between microbes and diseases. Large-scale projects such as the US Human

Microbiome Project (HMP) (The Human Microbiome Project Consortium, 2012) and the

European Metagenomics of the Human Intestinal Tract (MetaHIT) (Qin, J. et al., 2010) have

made a vast stride towards this goal.

Accumulating evidence since the burst of human gut microbiome research and

metagemomics suggest that the gut microbiota play a role in the regulation of several host

metabolic pathways. That results in interactive host-microbiota metabolic, signaling, and

63

immune-inflammatory axes that physiologically connect the gut, muscle, liver and brain

(Nicholson et al., 2012; Holmes et al., 2012). The gut microbiota fights against

enteropathogens (Fukuda et al., 2011), extracts nutrients and energy from our food

(Yatsunenko et al., 2012), and boosts normal immune function (Olszak et al., 2012).

Disruptions to the normal balance between the gut microbiota and the host have been

correlated with obesity (Kallus & Brandt, 2011), malnutrition (Smith et al., 2013; Trehan et

al., 2013), inflammatory bowel disease (IBD) (De Cruz et al., 2012; Mann & Saeed, 2012)

neurological disorders (Bercik et al., 2012; Dinan & Cryan, 2012) and cancer (Kostic et al.,

2012; Tjalsma et al., 2012; Chen et al., 2012). Moreover, a growing body of evidence

indicates that the gut microbiota can communicate with the central nervous system possibly

through endocrine, neura and immune pathways and therefore affects brain function and

behavior (see Review by Cryan et al., 2012). The impact of intestine on human health is

summarised in Figure 1.8.

Figure 1.8 The intestine's impact on health. The gastrointestinal (GI) tract contributes to health by ensuring digestion and absorption of nutrients, minerals and fluids; by induction of mucosal and systemic tolerance; by defence of the host against infectious and other pathogens; and by signalling from the periphery to the brain. This figure was taken from Bischoff (2011).

64

Importantly, the diet is considered a primary factor that influences gut bacterial

diversity, and thus may alter its functional relationships with the host (Ley et al., 2008). A

variety of enzymatic activities from human gut microbiota with crucial influence on human

health through biotransformation of secondary plant products and xenobiotic compounds

have been reported (McBain et al., 1998; Heavey & Rowland, 2004; Blaut & Clavel, 2007).

Among a plethora of genes that have been identified in the human gut microbiome,

those that encode carbohydrate-active enzymes (CAZymes) are of particular importance, as

these enzymes are required to digest most of dietary polysaccharides to fermentable

monosaccharides. Most enzymes that cleave glycosidic bonds between carbohydrates or

between a carbohydrate and a non-carbohydrate moiety by hydrolysis (addition of water)

are categorized into enzyme family glycoside hydrolase (GH)(Henrissat 1991; Koropatkin et

al., 2012). At present, GHs are classified into 130 families with the conservation of amino

acid sequences, catalytic residues, molecular mechanism and stereochemical outcome

among members of a given family. GH classification uses the IUB Enzyme Nomenclature

(1984) based on the type of reaction that GH enzymes catalyse and on their substrate-

specificity. For GHs (EC 3.2.1.x), the first three digits indicate enzymes hydrolyzing O-glycosyl

linkages whereas the last number indicates the substrate and sometimes reflects the

molecular mechanism. However, such a classification does not always reflect the structural

features of GH enzymes (Henrissat 1991).

The CAZy database is a knowledge-based resource specialized in the enzymes that

build and breakdown complex carbohydrates and glycoconjugates (Cantarel et al., 2009).

The substrates hydrolyzed by members of a CAZy family are structurally diverse however

they display conserved characters such as the orientation of the glycosidic bond (axial or

equatorial) and often have a functional commonality such as being present in animal or

plant cell wall carbohydrates (Kaoutari et al., 2013). Therefore, assigning several GH families

to broad substrate categories that can facilitate the annotation of the encoding genes is

possible (Cantarel et al., 2012). If a primary degrader was designated as a bacterium that is

able to digest a complex carbohydrate due to enzymatic capacity that is missing in other

species, each complex carbohydrate may well have many different primary degraders. In

healthy adults, it was reported Bacteroidetes members in the faecal microbiota can vary

from ~15% to ~90% in the proportional representation, and that of Firmicutes members

65

varies from ~70% to ~5% of the microbiota (The Human Microbiome Project, 2012). Such

compositional variations would give rise to substantial differences in the capacity of the

microbiota to degrade complex polysaccharides provided that there is a strong difference in

CAZyme content and diversity between these two phyla. Thus, CAZymes can serve as a

useful biomarker of the functional diversity of the human gut microbiota.

1.9 Cruciferous vegetables can alter human gut microbiota communities

In spite of minor fluctuation over a short period of time, the gut bacterial community

of healthy adults is considered relatively stable (Delgado et al., 2004). However, this

community was altered by short-term dietary shifts, commonly found in controlled dietary

interventions (Gibson et al., 1995; Langlands et al., 2004; Smith et al., 2006; Costabile et al.,

2008). Several dietary components, e.g. dietary fibers such as cellulose, hemicelluloses, and

pectin (Bourquin et al., 1993), and other compounds such as lignans (Milder et al., 2007) and

GSLs (Fahey et al., 2001) present in cruciferous vegetables can be used as metabolic

substrates for certain human gut bacteria. For example, cellulose can be transformed to

short-chained fatty acid (SCFA) by Bacteroidetes (Robert et al., 2007). Lignans such as

secoisolariciresinol can be degraded in vitro by certain Eggerthella and Peptostreptococcus

isolated from human feces (Clavel et al., 2005). GSLs can also be metabolized in vitro by E.

coli, E. faecalis, B. thetaiotaomicron, E. faecium, certain Bifidobacterium spp. and

Peptostreptococcus spp. (Brabban et al., 1994; Rabot et al., 1995; Elfoul et al., 2001; Cheng

et al., 2004).

It was thought that constituents of high-cruciferous vegetable diets are likely to

influence the growth of certain bacteria in the human gut bacterial community and

ultimately modify the community composition (Li et al., 2009). Terminal restriction fragment

length polymorphism (tRFLP) fragments technique revealed that Alistipes putredinis,

Eubacterium hallii, Phascolarctobacterium faecium and Eggerthella spp. were associated

with cruciferous vegetable intake, and Burkholderiales was associated with the cruciferous

vegetable-free diet (Li et al., 2009). This substantial difference in the bacterial community

structure among individuals was also reported that is in agreement with other previous

studies (Ley et al., 2006; Eckburg et al., 2005). Bacterial species with the same metabolic

functions associated with cruciferous vegetable intake may not be closely related

66

phylogenetically. For example, incubations with GSLs in vitro showed that bacteria able to

hydrolyze GSLs come from several different phylogenetic families, including Actinobacteria,

Firmicutes, and Bacteroidetes (Brabban et al., 1994; Rabot et al., 1995; Elfoul et al., 2001;

Cheng et al., 2004). Phylogenetic analysis of the glycosidase gene family shows that the

activity of hydrolyzing GSLs is not conserved within one discrete phylogenetic group of

bacteria (Mian, 1998). Since different individuals may have different types of distantly

related bacteria having GSL-degrading activity, it is likely that GSL consumption can either

trigger or suppress different species in the gut bacterial community (Li et al., 2009).

1.10 Hypotheses

The hypotheses of this PhD project are as follows:

Certain human gut bacteria can metabolise GSLs to ITCs and/or NITs like plant and

aphid myrosinases and cetain human gut bacteria can modify the GSL side chain and the

nature of GSL degradation products.

Different GSL-degrading bacteria may degrade identical or different GSLs at different

rates

Several enzymes may be involved in GSL metabolism in GSL-degrading bacteria.

1.11 Objectives To test the above hypotheses, the objectives are set as follows:

To isolating and identifying GSL-degrading bacteria from human faecal sample.

To identify the degradation products of different types of GSLs and certain DS-GSLs

metabolized by individual bacteria. Cell-free extract and resting cells experiments

were performed to detect bacterial GSL-degrading activity in vitro and to test its

inducibility (Chapter 2).

To identify bacterial enzymes potentially involved in the metabolism of GSLs via

forwards proteomics approach by using two-dimension electrophoresis (2-DE) gels

(Chapter 3).

To identify and characterize bacterial enzymes potentially involved in the metabolism

of GSLs via reverse proteomics approach by using Basic Local Alignment Search Tool

(BLAST) searches, molecular cloning, protein purification techniques and enzyme

activity assays (Chapter 4 and 5).

67

Chapter 2: Metabolism of different GSLs and DS-GSLs by human gut bacteria 2.1. Introduction:

2.1.1 GSL degradation by human gut microbiota

To date, a number of microorganisms including bacteria and fungi have been

reported for their GSL-degradation properties. The better understanding of myrosinase

occurrence is important for both biological and biotechnological aspects of food and feed

industries. Most work has concentrated on the characterization of GSL degradation by intact

microbial cells (Maheshwari et al., 1981; Palop et al., 1995; Rakariyatham & Sakorn, 2002)

and several reports have shown myrosinase production from Aspergillus sp. (Ohtsuru & Hata,

1973; Sakorn et al., 1999). Previous in vitro studies have observed that bacteria able to

hydrolyze GSLs belong to several different phylogenetic families including Actinobacteria,

Firmicutes and Bacteroidetes (Brabban & Edwards, 1994; Cheng et al., 2004; Elfoul et al.,

2001; Palop et al., 1995; Rabot et al., 1995). Phylogenetic analysis of various β-glucosidase

genes also showed that they are not solely attached to one bacterial phylogeny (Mian, 1998).

Bacteria from different phylogenetic groups may have relevant genes to code enzymes

involved in releasing glucose from complex molecules in the gut environment for energy

utilisation. Degradation of GSLs in the colon or caecum has been known for some time

(Rabot et al., 1995). The gastrointestinal microflora of rats and poultry has the ability to

hydrolyze GSLs (Krul et al., 2002; Nugon-Baudon et al., 1990; Nugon-Baudon et al., 1988;

Slominski et al., 1988). Once cruciferous vegetables are consumed, humans rely on gut

bacteria in GSL conversion to ITCs since vigoruous cooking tends to deactivate degradative

enzymes. The significance of human gut bacteria in producing ITCs was demonstrated in a

previous feeding study showing a dramatic decline in urinary ITC excretion after cruciferous

vegetable consumption when volunteers were pre-treated with antibiotics and bowel

cleansing (Shapiro et al., 1998).

Out of above 800 bacterial species in the human gut community, up to 99% of the

total population come from dominant 30–40 species (Bäckhed et al., 2005). A distinct

68

combination of bacteria species from each person may ultimately contribute to inter-

individual differences in metabolism of dietary constituents and hence health status of the

host. To date, in vitro experiments incubating mixed or pure cultures of bacteria with GSLs

have confirmed that several bacterial species residing in the human gut, such as E. coli, B.

thetaiotaomicron, E. faecalis, E. faecium, L. agilis, certain Peptostreptococcus spp. and

Bifidobacterium spp., have the ability to metabolize GSLs in culture (Brabban & Edwards,

1994; Cheng et al., 2004; Elfoul et al., 2001; Palop et al., 1995; Rabot et al., 1995). In addition

to the amount of consumed cruciferous vegetables, the composition of gut bacterial

community may influence exposure to bioactive ITCs and ultimately determine cancer risk of

the hosts.

2.1.2 Metabolic diversity of the intestinal microbiota

The GI tract is a habitat for a large number of bacteria with a contribution to normal

digestive function. Most intestinal microbiota residing in the large intestine has the primary

function to anaerobically degrade and ferment organic matter i.e. protein and carbohydrate

into absorbable energy. The main production of this process is SCFA including butyrate,

propionate and acetate and also gases e.g. H2, CO2 and CH4 (in some cases). Especially,

butyrates are known to have a crucial role in gut environment and health (Pryde et al.,

2002). The contribution of different functional groups of microbiota linked in a trophic chain

is required to operate this complex microbial fermentative process. These bacteria can

produce a variety of hydrolytic enzymes that transform complex substrates into smaller

fragments. Further fermentation of these fragments is carried out by these hydrolytic

bacteria and also by other bacterial communities able to utilize the released breakdown

products. In comparison with other bacterial communities, human gut bacteria have more

diverse enzymes that enable them to metabolize drugs and other xenobiotics to a much

further extent (Scheline, 1973; Abu Shamat, 1993; Mikov, 1994). It has been suggested that

the gut microbiota can function as an organ with a metabolic capacity at least equivalent to

the liver (Scheline, 1973). However, the important differences between hepatic and bacterial

metabolism are that the liver is primarily responsible for oxidative and conjugative

metabolisms that produce polar high molecular weight metabolites while reductive and

hydrolytic reactions are prevalent in the gut microbiota generating non-polar low molecular

weight byproducts. Thus far, at least thirty commercially available drugs were substrates for

69

bacterial enzymes (Sousa et al., 2008) and presumably there are many more from the new

and existing drugs (with the potential for contact with the distal gut) to be discovered. Some

metabolic reactions of intestinal microbiota from either humans or rats are presented in

Table 2.1.

Table 2.1 Some metabolic reactions of intestinal microbiota

Reactions Example References

Reductions

Nitro compounds Clonazepam, nitrazepam Elmer & Remmel 1984; Takeno et al., 1990; Rafii et al., 1997

Sulfoxides Sunlindac, sulfinylpyrazone Strong et al., 1987; Strong et al., 1984b 21-Hydroxycorticoids Aldosterone Miyamori et al., 1988 Double bonds Digoxin, daidzein Reuning et al., 1985; Rafii et al., 2007 Azo compounds Prontosil Gingell et al., 1971 Amides Zonisamide Kitamura et al., 1997

Degradation

Nitrate esters

Glyceryl trinitrate, isosorbide dinitrate

Abu Shamat & Beckett 1983; Abu Shamat, 1993

Sulfate esters Sodium picosulfate Jauch et al., 1975 Succinate esters Carbenoxolone Iveson et al., 1971 Amides Methotrexate, chloramphenicol Valerino et al., 1972; Holt, 1967 Glucuronides Morphine glucuronide Walsh & Levine, 1975; Schneider et al., 1999

Glucosides sennosides, quercetin-3-glucoside Hersperidin Lee et al., 2004

Arabinofuranocyl Sorivudine Okuda et al., 1998

Proteolysis Hormones Insulin, calcitonin Tozaki et al., 1997

Removal of functional groups

N-Dealkylation Methamphetamine Caldwell & Hawksworth, 1973 Deamination Flucytosine Vermes et al., 2003 N-oxide bond cleavage Ranitidine Basit & Lacey, 2001

Other reactions Heterocyclic ring

fission Levamisole Shu et al., 1991 Side-chain cleavage Steroids Cerone-McLernon et al., 1981 Acetylation 5-Aminosalicylic acid van Hogezand et al., 1992 Isoxazole scission Risperidone Meuldermans et al., 1994

70

2.1.3 Characterization of human gut microbiota

Our understanding of the gut microbiota, how it interacts with the host and causes

human disease, has been enhanced by advances in culture-independent techniques for

phylogenetic investigation and quantification. Comprehensive knowledge of the entire gut

microbiota is neccessary to understand relationships between the gut microbiota and

disease. Culture and biochemical typing were the gold standards for the identification of

bacterial species for many years. Since the 1990s, however, culture-independent techniques

have transformed our knowledge of the gut microbiota as they are able to give a more

representative 'snapshot' of this niche (Rajilić   et al., 2007; Zoetendal et al., 2006). These

techniques are based on sequence divergences of the small subunit ribosomal RNA (16S

rRNA), and are able to demonstrate the following: first, the microbial diversity of the gut

microbiota; second, qualitative and quantitative information on bacterial species; and third,

changes in the gut microbiota in relation to disease. Examples of these techniques include

terminal restriction fragment length polymorphism (tRFLP), denaturing gradient gel

electrophoresis (DGGE), fluorescence in situ hybridization (FISH), DNA microarrays, and next-

generation sequencing of the 16S rRNA gene or its amplicons. The techniques currently used

to characterize the gut microbiota with their advantages and limitations are presented in

Table 2.2.

71

Table 2.2 Advantages and disadvantages of current techniques used to characterize human

gut microbiota

Technique Description Advantages Disadvantages

Culture Isolation of bacteria on selective media Cheap, semi-quantitative

Labour intensive, <30% of gut microbiota have been cultured to date

qPCR

Amplification and quantification of 16S rRNA. Reaction mixture contains a compound that fluoresces when it binds to double-stranded DNA

Phylogenetic identification, quantitative, fast

PCR bias, unable to identify unknown species

DGGE/TGGE Gel separation of 16S rRNA amplicons using denaturant/temperature

Fast, semi-quantitative, bands can be excised for further analysis

No phylogenetic identification, PCR bias

tRFLP

Fluorescently labelled primers are amplified and then restriction enzymes are used to digest the 16S rRNA amplicon. Digested fragments separated by gel electrophoresis

Fast, semi-quantitative, cheap

No phylogenetic identification, PCR bias, low resolution

FISH

Fluorescently labelled oligonucleotide probes hybridize complementary target 16S rRNA sequences. When hybridization occurs, fluorescence can be enumerated using flow cytometry

Phylogenetic identification, semi-quantitative, no PCR bias

Dependent on probe sequences—unable to identify unknown species

DNA microarrays

Fluorescently labelled oligonucleotide probes hybridize with complementary nucleotide sequences. Fluorescence detected with a laser

Phylogenetic identification, semi-quantitative, fast

Cross hybridization, PCR bias, species present in low levels can be difficult to detect

Cloned 16S rRNA gene sequencing

Cloning of full-length 16S rRNA amplicon, Sanger sequencing and capillary electrophoresis

Phylogenetic identification, quantitative

PCR bias, laborious, expensive, cloning bias

Direct sequencing of 16S rRNA amplicons

Massive parallel sequencing of partial 16S rRNA amplicons for example, 454 Pyrosequencing® (Roche Diagnostics GMBH Ltd, Mannheim, Germany) (amplicon immobilized on beads, amplified by emulsion PCR, addition of luciferase results in a chemoluminescent signal)

Phylogenetic identification, quantitative, fast, identification of unknown bacteria

PCR bias, expensive, laborious

Microbiome shotgun sequencing

Massive parallel sequencing of the whole genome (e.g. 454 pyrosequencing® or Illumina®, San Diego, CA, USA)

Phylogenetic identification, quantitative

Expensive, analysis of data is computationally intense

Abbreviations: DGGE, denaturing gradient gel electrophoresis; FISH, fluorescence in situ hybridization; qPCR, quantitative PCR; TGGE, temperature gradient gel electrophoresis; tRFLP, terminal restriction fragment length polymorphism. The table was taken from Fraher et al. (2012).

72

In this work, enrichment culture technique, 16S rRNA gene sequencing and PCR were

employed to characterize human gut bacteria, therefore only these techniques are reviewed

as follows.

2.1.3.1 Enrichment culture technique

Metabolic and genetic characteristics need to be effectively studied in order to

determine the role an organism plays in its environment. Culturing bacteria in a laboratory is

the best way to do this however certain organisms may be found in low abundance in its

environment and cannot be isolated easily using general all-purpose media (Schlegel &

Jannasch, 1967). Since these bacteria are likely to be outgrown by more numerous species, it

is nearly impossible to obtain a pure bacterial culture. Therefore, the knowledge of

nutritional and environmental conditions (including specific energy sources and dependency

on oxygen source) which favor bacterial growth is essential to isolate bacteria (Schlegel &

Zaborosch, 1993).

As pioneered by Winogradsky and Beijerinck (Winogradsky, 1890), enrichment

culture technique was designed to enable a particular type of microorganism to outgrow all

others. For example, a GSL-metabolizing microorganism can be isolated from a soil solution

by using a GSL substrate as the only energy source in liquid media incubated under

anaerobic conditions. This approach has been used to isolate GSL-metabolizing soil

bacterium Citrobacter in our laboratory (Abdulhadi Albaser, PhD thesis). Selective media can

also be used to suppress the growth of unwanted microorganisms so that only the desired

one can grow. In contrast, enrichment media are used to favor the growth of the desired

ones without deliberately inhibiting the others. The effects of these two medium types are

occasionally the same while other organisms might not grow due to the specific makeup of

the enrichment medium (Jennings & Isaac, 1995). Three general strategies used in an

enrichment culture are as follows;

(i) Chemical strategies: include using a specific energy source such as a GSL substrate

that enriches for microorganisms with specific capacity to metabolize it as an energy source.

Acidophilic organisms can also be selected for by controlling the pH to between 4.0 and 5.4.

(ii) Physical strategies: include incubating at high or low temperature which would

select for thermophilic (heat loving) or psychrophilic (cold loving) microorganisms. Anaerobic

organisms can be selected for by incubating the cultures without oxygen.

73

(iii) Biological strategies: include using a live host to enrich for a particular type of

bacterial virus.

Enrichment cultures can be used to enumerate microbial populations in a sample. By

using a dilution endpoint series, the inoculum is serially diluted in a sterile medium, and

numerous tubes of medium are inoculated with aliquots of each successive dilution. The

purpose of this is to minimize the probability of introducing even one individual into a given

tube in a series of tubes with a dilute microbial suspension. Thus, only the organism of

interest will grow in the last dilution at the endpoint (Claus, 1995). This technique is

relatively easy to be used in the search for new microbial types by selecting for organisms

with specific capabilities. However, contamination is likely to occur and the numbers of

microorganisms isolated from the enrichment cultures never represent the true numbers

found in their environment.

Knowledge of the gut microbiota was limited to culture-based techniques, an

approach that has been used since the early 20th century. Since then, advances have been

made in the phenotyping of isolates on the basis of their fermentation profiles and in vitro

growth requirements. Although bacterial identification by culture is fairly cheap, it is labour

intensive, and culture alone gives a limited view of the diversity of the gut microbiota

because < 30% of gut microbiota members have been cultured to date (Eckburg et al., 2005;

Guarner & Malagelada, 2003). It is important to remember that uncultured organisms in the

gut microbiota are not necessarily unculturable. They might, in fact, be culturable but

permissive growth conditions for these organisms have not yet been developed or

discovered.

2.1.3.2 16S rRNA gene analysis

Ribosomes (70S) are dispersed throughout the cytoplasm of a bacterial cell and made

up of two subunits: 30S and 50S. The 50S subunit contains two RNA molecules: 5S and 23S.

The 30S subunit (or small subunit) contains one RNA molecule: 16S ribosomal RNA (16S

rRNA). One of the functions of 16S rRNA is the initiation and extension of protein synthesis.

As rRNA (5S, 16S and 23S) is highly conserved between bacterial species, yet contains

variable regions that yield a phylogenetic signal, it is a useful target for phylogenetic

identification i.e. bacterial identification. Of the three bacterial rRNA genes, the 16S rRNA

74

gene provides the most tractable combination of conserved sites for PCR primers. Its

variable regions as evolutionary chronometers are usually used in preference to 5S or 23S

rRNA genes for phylogenetic identification (Clarridge, 2004; Olsen et al., 1986; Rajendran et

al., 2011). Most (but not all) contemporary culture-independent techniques for the analysis

of the gut microbiota are thus based on analysis of the 16S rRNA gene.

2.1.3.3 Polymerase chain reaction (PCR)

Although PCR has been a huge technical advance across the medical field, it has

limitations; each physical, chemical, and biological step—from retrieving a sample to the

resulting 16S rRNA amplicons—represents a potential source of bias (Wintzingerode et al.,

1997). For example, differential lysis of microbial cells can affect the final apparent

microbiota composition. Gram-positive organisms typically require rigorous conditions to

lyse the bacterial cell wall (which is thicker than in Gram-negative bacteria), while these

same conditions may cause excessive fragmentation of Gram-negative chromosomal DNA

(Olsen et al., 1986). Another major limitation of PCR is that primers must be designed to

target all phyla.

2.1.4 Analytical methods for GSLs and their degradation products

The abundance and structural variety of the GSLs and the fact that each can produce

different breakdown products makes their analysis very complicated (McGregor et al., 1983;

Verkerk et al, 1998). Because GSLs coexist with myrosinase in the plant, cutting or grinding

of fresh tissue in the presence of water will lead to rapid degradation of the parent

compounds, and this adds greatly to the complexity of the problem. In general, the analytical

approach can be divided into methods for total GSLs, individual GSLs and the degradation

products.

For analysis of intact GSLs, inhibition of myrosinase-like activity is essential. Before

disruption of the material, samples should be freeze-dried or frozen in liquid nitrogen to

complete dryness. The use of aqueous methanol for extraction, in combination with high

temperatures, also inhibits myrosinase (Heaney & Fenwick, 1993). Total GSLs yield

equimolar amounts of glucose upon degradation with myrosinase, and methods based on

the measurement of released glucose proved to be relatively rapid and simple to apply

(Heaney et al., 1988). The total GSL content of a food sample can be measured by

75

determining the quantity of glucose released after treatment with the enzyme, but account

must be taken of any endogenous glucose. To achieve this, extraction of GSLs can be

performed followed by a selective clean-up to eliminate free glucose and other interfering

compounds, after which controlled enzymatic release of bound glucose is possible (Mithen

et al., 2000). Several titrimetric and gravimetric methods have been described for the

quantification of the bisulfate ion (unstable aglycone, which after a Lossen rearrangement

produces and equimolar quantity of bisulfate) generated after degradation of GSLs by

myrosinase. There is also a method in which the bisulfate liberated after sulfation is

precipitated with barium chloride, and residual barium is measured by X-ray emission

spectroscopy (Schnug et al., 1987).

The traditional method for the identification and quantification of the individual

derivatised GSLs is gas liquid chromatography (GLC)(Underhill & Kirkland, 1971). Originally,

the GSLs were extracted with boiling water, derivatised and separated by isothermal

chromatography, but substantial improvements have subsequently been made (Thies,

1976). In particular, ion exchange purification of GSL extracts to remove carbohydrates and

other impurities before derivatisation has increased the sensitivity of this method.

Several common liquid chromatography ultraviolet (LC/UV) and liquid

chromatography mass spectrometry (LC/MS) techniques employing ion-pair reagents to

neutralize the charge on the sulfate group of the GSL molecule are used for determining

intact GSLs. Thus, the separation is a property of the functionality of the variable R groups.

Common ion-pair reagents used include tetra-alkyl ammonium bromide in combination with

phosphate buffers, triethylamine in combination with formic acid, and ammonium acetate

buffers (Arguello et al.,  1999;  Hrnčiřík  et al., 1998; Prestera et al., 1996; Zrybko et al., 1997).

The separation and peak shape obtained from intact GSLs are often very poor as tetra-alkyl

ammonium bromide reagents are difficult to dissolve. Other ion-pair reagents produce

marginally better peak profiles, but the separation is still imperfect (Mellon et al., 2002).

More robust and universally applicable LC/MS (and LC/UV) methods are required.

A major breakthrough in GSL analysis has been achieved with the introduction of

enzymatic on-column desulfation using sulfatase (Thies, 1976; Thies, 1978). The introduction

of a desulfation step before derivatisation was performed to eliminate sulfate that

76

interfered with GC analysis. Desulfation was elegantly carried out on the ion exchange

column, using a commercially available sulfatase isolated from an edible snail, H. pomatia.

Free sulfate in the GSL extract, which could inhibit the sulfatase, was precipitated by an

addition of barium acetate and removed by centrifugation before addition of the extract to

the ion exchange column (Vallejo et al., 2004; Mithen et al., 2000).

Since some GSLs, particularly the indoles, are thermally unstable, high performance

liquid chromatography (HPLC) has become the preferred method with the advantage of

direct determination of GSLs. One of the most commonly used method, reversed-phase

HPLC was developed for quantitative analysis of DS-GSLs (Fenwick et al., 1983). In this

method, an on-column enzymatic desulfation treatment of plant extracts is followed by

HPLC detection of the resultant DS-GSLs. Adaptation of the sulfohydrolase (sulfatase)

desulfation method as an HPLC method, although the most widely used method for GSL

separation, is still subject to difficulties in interpretation because of the effects of pH, time

and enzyme activity of the desulfation products (Fahey et al., 2001). Typically, this method

uses response factors determined with purified DS-sinigrin and uses DS-glucotropaeolin as

an internal standard (Brown et al., 2003). Corresponding DS-GSL times, and comparison to

standardized rapeseed extracts, are typically used to validate chromatographic profiles.

Unfortunately, the biological activity of these molecules is compromised by the removal of

the sulfate. After desulfation, they can no longer serve as substrates for myrosinase, and

thus their cognate ITCs are not available for bioassay or direct measurement by

cyclocondensation—key tools in the study of the pharmacokinetics, pharmacodynamics and

bioactivity of these compounds.

HPLC systems using an ultraviolet (UV) detector are very sensitive; levels of DS-GSLs

in the nano-molar range can be detected. Whilst spectral data of individual DS-GSLs will

allow initial confirmation of structural class, the addition of MS detection further improves

the discriminatory power of the technique. DS-GSLs are commonly separated using end-

capped C18 reverse phase columns eluted with water:acetonitrile gradients, whilst isocratic

elution with water:methanol phases has also been reported for the separation of both DS-

GSLs and intact GSLs. Reverse-phase C18 HPLC methods are preferable and more accurate for

determining GSL content. These methods are especially robust, powerful, and selective

77

when they form a component of an optimised negative ion mass spectrometry LC–MS

method (Bennett et al., 2004). One of the major issues in the analysis of GSLs has been the

lack of suitable standards. The only commercially available GSLs are glucotropaeolin and

sinigrin. Sinigrin is not a suitable internal standard due to its presence in most brassicacious

plants, but glucotropaeolin is not normally present in Brassica and has been used as internal

standard. Several mass spectrometric techniques have been investigated for structure

elucidation of the various DS-GSLs e.g. direct probing electron impact, chemical ionisation,

and fast atom bombardment (FAB). Considerable structural information can be obtained

with these techniques. Many years ago, a novel enzyme-linked immunosorbent assay (ELISA)

procedure for the determination of sinigrin and progoitrin in Brussels sprouts extracted with

phosphoric acid, using antisera raised against hemisuccinate-linked GSL conjugates has been

described (van Doorn et al., 1999). The method tended to overestimate GSL content in

comparison to HPLC methods but seems to offer great potential advantages at lower cost

and shorter time for routine analysis in breeding programmes (Mithen et al., 2000).

The volatility of many compounds presents a drawback in the investigation of GSL

breakdown products using HPLC method. Moreover, TCs and NITs are not detectable

spectrometrically, and ITCs/NITs can be analyzed by GLC. Oxazolidinethiones and indoles

may be also analyzed using HPLC with UV detection. A method for analysing

oxazolidinethiones in biological fluids with a high degree of selectivity was developed

(Quinsac et al., 1992). However, HPLC finds most use in the analysis of intact GSLs or DS-

GSLs. For identification and confirmation of structures, both HPLC and GLC can be coupled to

mass spectrometry (MS). Mass spectroscopy has proved to be invaluable in the identification

and structural elucidation of GSLs and their breakdown products. Positive ion FAB mass

spectrometry has yielded mass spectra characterised by abundant protonated and

cationised molecular ions with relatively little fragmentation (Fenwick et al., 1982). In the

negative ion mode, FAB produces an abundant molecular ion (of the GSL anion). This proved

especially advantageous in the analysis of crude plant extracts and mixtures of purified GSLs.

Many years ago, a spectroscopic quantitation of organic ITCs was developed (Zhang et al.,

1992). Almost all organic ITCs react quantitatively with excessive vicinal dithiols under mild

conditions to form five-membered cyclic condensation products with a release of the

corresponding free amines (R-NH2) The method can be used to measure 1 nmol or less of

78

ITCs in crude mixtures and pure ITCs. Some of the methods used for analysis of GSLs and

their degradation products are summarized in Table 2.3

Table 2.3 Listing of some commonly used methods for the analysis of GSLs and their

breakdown products

Compound Method

Total GSL Palladium chloride and thymol assays

Glucose- and sulfate-release enzyme assays

Enzyme-linked immunosorbent assay (ELISA)

Near infra-red reflectance (NIR) spectroscopy; alkaline degradation and thioglucose detection

High resolution nuclear magnetic resonance (NMR) spectroscopy

Individual intact GSL Reverse phase HPLC

Thermospray LC with tandem MS; high performance capillary electrophoresis; capillary GC–MS, GC–MS, GC–MS/MS

DS-GSL Reverse phase HPLC

Degradation products X-ray fluorescence spectroscopy (XRF); GC or GC–MS; HPLC

2.1.5 Hypotheses

The hypotheses on which this chapter is based are as follows:

GSL metabolism in certain human gut bacteria is mediated by bacterial GSL-

degrading activity to produce ITCs and/or NITs like plant and aphid myrosinases.

Certain human gut bacteria can modify the GSL side chain and the nature of the

degradation products.

Different GSLs are metabolized with different efficiencies by the same bacteria and

different bacteria.

The diversity of gut bacteria myrosinases may partly explain the inter-individual

variation in GSL metabolism.

79

2.1.6 Objectives

Based on the above hypotheses, the objectives of this study are as follows:

To isolate and identify GSL-degrading bacteria from a human faecal sample using

enrichment culture technique and 16S rDNA gene analysis, respectively.

To identify and quantify GSL degradation products from the metabolism of different

types of GSLs in individual bacterial fermentations in vitro over a time course using HPLC and

GC-MS analyses.

To identify and characterize bacterial GSL-degrading activity in cell-free extract

experiments and to test its activity on the native gel.

To examine the inducibility of bacterial GSL-degrading activity in resting cells

experiments.

80

2.2 Materials and Methods

2.2.1 Preparation of GSL substrates

Pure sinigrin (SNG) was purchased from Apin Chemicals (UK) or purified from

Brassica nigra. Gluconasturtiin (GNT)(ca. 97% purity) from land cress seeds (Barbarea verna),

glucotropaeolin (GTP)(ca. 95% purity) from watercress seeds (Lepidium sativum), glucoerucin

(GER) (ca. 96% purity) from rocket seeds (Eruca sativa), and glucoiberin (GIB) (ca. 95% purity)

from candytuft seeds (Iberis umbellata) were extracted by previously reported method

(Thies, 1988). Glucoraphanin (GRP) (ca. 97% purity) was prepared by further oxidation of

GER following the previous report (Lori et al., 1999). Glucobrassicin (GBS) (ca. 90% purity)

was prepared from wild cabbage seeds (Brassica oleracea).

For each GSL extraction, the previous method (Thies, 1988) was used. The seed

source (100 g) was ground using a coffee grinder (Wahl James Martin ZX595 coffee grinder)

to a fine powder. The seed powder was defatted with petroleum ether (60-80 fractions) by

repeated extraction (7 Xs) after which the residual seed was allowed to dry completely in a

fume hood. The defatted seed was added to boiling methanol (250 mL; 80%) for 20 min to

extract the GSLs, and this step was repeated. The combined methanol extracts were filtered

using filter aid and concentrated using a Rotavapor-210 (Büchi, UK) to dryness at less than

40C under vacuum. The dried extract was re-dissolved in distilled water (120 mL) and

subjected to protein precipitation by adding 9 mL of a Pb(OAc)2:Ba(OAc)2 (1:1; each 0.5 M).

The mixture was cooled at -20C for 15 min and centrifuged at 16,000g for 15 min. The

supernatant was loaded onto a pre-packed diethylaminoethyl (DEAE) - Sephadex A25 in

Econo-Pac column (Bio-Rad, UK). To prepare the column, 1.43 g DEAE-Sephadex A25 (Sigma-

Aldrich, UK) was swollen with 6 M imidazole/0.3% (v/v) overnight, and the excess buffer

decanted and mixed with water (repeated several times). The slurry was poured onto an

Econo-Pac column and washed with water (3 Xs 20 mL). A solution of formic acid/i-

propanol/water (3:2:5) was poured into the column (2 Xs 5 mL) followed by water (4 Xs 5

mL). The 0.5 M K2SO4 solution (25 mL) was poured into the column, and the eluate was

allowed to drop in 25 mL ethanol contained in a 100 mL beaker. The cloudy mixture was

then agitated and cooled for 10 min at 4C. The mixture was then centrifuged at 16,000g for

15 min to remove precipitated potassium sulfate, and the supernatant was then evaporated

81

nearly to complete dryness in a 250 mL round bottomed flask. The dried residues were

dissolved in absolute methanol (8 mL), transferred to a 15 mL Falcon tube and cooled at -20

C for 15 min after which it was centrifuged for 15 min at 4C. Only the supernatant was

evaporated in a 100 mL round bottomed flask nearly to dryness. The dried residues in a 100

mL flask were dissolved in 6 mL water and frozen at -80C for 2 h before being freeze-dried

overnight. Appropriate amount of the freeze-dried powder of GSL was weighed out in a 1.5

mL Eppendorf tube. This was then dissolved in 1 mL Milli-Q water to make a 10 mM stock

solution. A pure sinigrin stock solution (10 mM) was prepared in the same manner. A

solution of sinigrin (100 µL of 1 mM final concentration), as an internal standard, was added

into a solution of GSL (100 µL of 1 mM final concentration). The mixture was then desulfated

(Section 2.2.3) and analyzed by HPLC (Section 2.2.4). The purity (%) of the isolated GSL can

be compared with the pure sinigrin standard of the same amount using the formula below:

Purity (%) = Area of GSL x Response factor* x 100% Area of sinigrin standard *The response factor (RF), determined empirically as Area for standard: Area for equimolar

amount of GSL (Brown et al., 2003), is shown for each GSL in Table 2.4.

Table 2.4 Response factors for desulfated GSLs at 229 nm relative to that of desulfo-

sinigrin

GSL Relative response factor

Sinigrin 1.0

Glucoiberin 1.2

Glucoraphanin 0.9

Glucoiberverin 0.8

Glucoerucin 0.9

Glucotropaeolin 0.8

Gluconasturtiin 1.0

Glucobrassicin 0.3* *Determined by Buchner (1987). This table was taken from Brown et al., (2003).

82

For the preparation of GRP substrate, GER (50 mg) previously obtained from the

above extraction method was reconstituted in 1 mL Milli-Q water and oxidised to GRP

following the procedure of Lori et al. (1999). Hydrogen peroxide (50 µL) was added and the

solution was incubated at 60C for 30 min. The mixture was loaded onto DEAE Sephadex A25

column, eluted and dried using a Rotavapor. The dried extract was re-dissolved in Milli-Q

water and separated on a pre-packed (pre-swollen with water overnight) G10 Sephadex

(Sigma-Aldrich, UK) column (26 mm x 800 mm), and fractions were eluted with Milli-Q water

at 0.5 mL/min for 10 mL/fraction. Fractions containing GSLs were collected and freeze-dried

overnight in a freeze dryer (HETO Dry Winner). The freeze-dried material was desulfated and

analyzed by HPLC to confirm the purity. All GSLs involved in this work are shown in Table 2.5.

2.2.2 Preparation of sulfatase

Sulfatase type H-1 from H. pomatia (10,000 units/g) (Sigma-Aldrich, UK) (0.7 g) was

dissolved in 30 mL of Milli-Q water, followed with the addition of 30 mL of cold absolute

ethanol. After centrifugation at 10,000g in Avanti J-26 XP centrifuge (Beckman Coulter) for

15 min, the supernatant was collected, and 1.5 volume of cold ethanol was added to the

supernatant. The precipitate acquired following centrifugation of the mixture at 10,000g for

15 min was re-dissolved in 20 mL of Milli-Q water. The crude extract was passed through a

pre-packed DEAE Sephadex A25 (0.2 g dry weight pre-swollen with 0.5 M sodium acetate pH

5.0 overnight, and washed with water). The eluted extract was then passed through a pre-

packed CM Sephadex C25 (Sigma-Aldrich, UK) (0.2 g dry weight pre-swollen with distilled

water overnight). The resulting solution containing sulfatase at 0.3 U/mL was stored  at  −20°C  

in aliquots until required.

2.2.3 Desulfation of GSLs

A mini column (Bio-Rad, UK) was pre-packed with 1 mL of a pre-equilibrated DEAE

Sephadex A25 suspension (bed volume 50% of total volume) in 20% ethanol and washed

with 1 mL of Milli-Q water (2 Xs) and allowed to drain. Sample containing GSL was well-

mixed with 10 µL of Pb(OAc)2:Ba(OAc)2 (1:1; each 0.5 M) solution in a 100 µL sample volume

to precipitate any impurities in the sample that may interfere with HPLC analysis. The

mixture was centrifuged at 16,200g for 2 min, and the clear supernatant was loaded onto

83

the column and washed with 1 mL of water (2 Xs) followed by 0.5 mL of 20 mM sodium

acetate buffer pH 5.0 (2   Xs).   Sulfatase   (75   μL)   was   added   to   the   column   and   was   left  

overnight. DS-GSLs were eluted from the column with 0.5 mL Milli-Q water (3 Xs). Depending

on the amount of material applied to the column, the eluate was either freeze-dried in a

freeze dryer and reconstituted in 200 µL of Milli-Q water or analyzed directly in a 1.5 mL

eluted solution by HPLC analysis.

2.2.4 HPLC analytic conditions for DS-GSLs detection

DS-GSLs (Section 2.2.3) were analyzed by HPLC. Injections (10 μL) were made onto a

pre-equilibrated C18 reversed-phase column, Synergy 4u Hydro-RP (150mm x 2mm, 4μm

particles) (Phenomenex, UK). The column was designed for high water loads with high

efficiency of elution of a sample prepared in high water content. This column was further

fitted with a SecurityGuard  ™  Universal  HPLC guard column (Phenomenex, UK) to filter solid

parts stemming from pump seals or injection rotors that otherwise shorten the lifetime of a

column dramatically. The entire column was connected to Agilent Technology 1200 binary

pump series (Agilent, UK) and auto sampler Agilent 1100 series (Agilent, UK). Individual HPLC

run was carried out for 36 min using a water (solvent A), acetonitrile (solvent B) gradient,

and the following program was used: 2% (v/v) acetonitrile for 15 min, a gradient of 2% - 25%

(v/v) acetonitrile for 2 min, a gradient of 25% - 70% (v/v) acetonitrile for 2 min, 70% (v/v)

acetonitrile for 2 min, a gradient of 70% - 2% (v/v) acetonitrile for 2 min and a final 15 min

wash in 2% (v/v) acetonitrile. All runs were carried out at 35C at a flow rate of 0.2 mL/min.

The DS-GSLs were detected at 229 nm using a UV detector (Waters, UK). GSLs present in the

samples were identified based on their retention times with known standards (Table 2.5).

84

Table 2.5 GSLs involved in this work as detected by HPLC analysis

Type of GSL Trivial name Semisystematic name Molecular

Weight TR (min)*

Alkyl Sinigrin (SNG) 2-Propenyl (PRO)

358.37 6.00

Aromatic Gluconasturtiin (GNT) 2-Phenethyl (PHE)

409.43 16.23

Aromatic Glucotropaeolin (GTP) Benzyl (BEN)

423.46 13.37

Methylthioalkyl Glucoiberverin (GIV) ** 3-Methylthiopropyl (3MTP)

406.47 11.34

Methylthioalkyl Glucoerucin (GER) 4-Methylthiobutyl (4MTB)

421.51 13.23

Methylsulfinylalkyl Glucoiberin (GIB) 3-Methylsulfinylpropyl (3MSP)

453.00 3.63

Methylsulfinylalkyl Glucoraphanin (GRP) 4-Methylsulfinylbutyl (4MSB)

436.50 5.56

Indolyl

Glucobrassicin (GBS)

Indol-3-ylmethyl (I3M) 447.46 15.21

*Retention time at which GSL was eluted from the C18 reversed-phase column and detected by a UV detector. **Glucoiberverin (GIV) was not used as a substrate in bacterial fermentation, however it was identified on HPLC chromatograms as compared with GSL chromatographic profiles obtained from Arabidopsis thailiana (Nurul Huda Binti Abd Kadir, PhD thesis). Quantification of each GSL from the area of HPLC peak was achieved by using a

response factor (Table 2.4) for each GSL relative to the external standard, sinigrin (SNG) that

included in every HPLC run along with the samples. The concentration of each GSL, given

that its response factor (RF) was known, was calculated from the following formular:

Amount of GSL (mol) = (Area of GSL/Area of SNG) x RF x Amount of SNG (mol)

2.2.5 Preparation of DS-GSL substrates

DS-GSLs as substrates from the corresponding intact GSLs were prepared as follows.

Each GSL (30 mg) was dissolved in 10 mL of 0.02 M sodium acetate buffer pH 5.0 containing

10 mL of sulfatase type H-1 from H. pomatia (Sigma-Aldrich, UK) (0.3 U/mL) at 37°C for 16 h.

85

The reaction mixture was combined with 37.5 mL of absolute ethanol, and the precipitate

was removed by centrifugation at 16,100g for 10 min. Residual ethanol was removed from

the supernatant under vacuum and the residue dissolved in 2 mL Milli-Q water and then

passed through a mini-column (10mm × 80mm) filled with 2 mL bed volume of

diethylaminoethyl (DEAE) - Sephadex A25 (Sigma-Aldrich, UK) (which had been swollen with

0.5 M sodium acetate pH 5.0 overnight and equilibrated with 2 X 1 mL Milli-Q water). The 2

mL flow-through was collected and the column was rinsed with 2 mL of water, and the

eluate (4 mL in total) was freeze-dried yielding approximately 16 mg of DS-GSL as white

powder. Identification and purity of DS-GSL of over 90% were determined using a response

factor method (Brown et al., 2003) as described in Section 2.2.1.

2.2.6 Authentic ITC and NIT standards

Phenethyl isothiocyanate (PITC), phenethyl nitrile (PNIT), benzyl isothiocyanate (BITC),

benzyl nitrile (BNIT), allyl isothiocyanate (AITC), allyl nitrile (ANIT), 3-methylthiopropyl

isothiocyanate (3MTP-ITC) or iberverin (IBV) and 4-methylsulfinylbutyl isothiocyanate

(4MSB-ITC) or sulforaphane (SFN) were purchased from Sigma-Aldrich (UK). Other standards

of   ≥   97%  purity   including 4-methylthiobutyl isothiocyanate (4MTB-ITC) or erucin (ERN), 5-

methylthiopentyl nitrile (5MTP-NIT) or erucin nitrile (ERN NIT), 4-methylthiobutyl nitrile

(4MTB-NIT) or iberverin nitrile (IBV NIT) were synthesised in our laboratory. The remaining

standard, 5-methylsulfinylpentyl nitrile (5MSP-NIT) or sulforaphane nitrile (SFN NIT)  of  ≥  97%  

purity was a kind gift from the Institute of Food Research (IFR, Norwich). Note that iberin

(IBR), iberin nitrile (IBR NIT), iberverin nitrile (IBV NIT) were not purchased or available in this

work.

2.2.7 Isolation of GSL-degrading bacteria

A faecal sample from a healthy volunteer was homogenized in phosphate buffered

saline (PBS) solution, pH 7.0 using a Stomacher 400 (Seward, UK) operating at 180g for 45 s.

Faecal homogenate (100 µL) was inoculated into the culture medium (900 µL) containing 1

mg of sinigrin. Three different media without glucose addition were used; Wilkins Chalgren

(WC), Nutrient broth (NB) and de Man, Rogosa and Sharpe (MRS). The composition of each

86

medium is shown in Table 2.6. Note that L. agilis R16 used in this work was obtained from

Palop et al (1995), not from enrichment culture experiment.

Table 2.6 Compositions of culture media

Media Compositions in 1 L pH

MRS (no glucose)

10 g Peptone, 10 g Meat extract, 5 g Yeast extract, 1 g Tween-80, 2 g K2HPO4, 5 g Na-acetate, 2 g (NH4)2 citrate, 0.2 g MgSO4-7H2O, 0.05 g MnSO4-H2O

7.1 ± 0.2

WC (no glucose)

10 g Tryptone, 10 g Gelatin peptone, 5 g Yeast extract, 5 g NaCl, 1 g L-Arginine, 1 g Sodium pyruvate, 0.005 g Hemin, 0.0005 g Vitamin K

6.5 ± 0.2

NB 5 g Peptone, 1.5 g Beef extract, 1.5 g Yeast extract, 5 g NaCl 6.5 ± 0.2

Aliquots were serially diluted (10-fold) in those media every two days until the

sixteenth day in an anaerobic cabinet (MACS-MG-1000-anaerobic workstation, DW

Scientific) under an atmosphere of 5% CO2, 10% H2 and 85% N2. At day sixteen, each culture

medium (100 L) was plated onto selective corresponding media agar (1.5% agar added to

liquid broth) containing 1 mM sinigrin and incubated in the anaerobic cabinet at 37°C till

colonies were visible. From each selective media, colonies with different morphologies were

sub-cultured in 1 mL of corresponding broths containing 1 mM sinigrin overnight. Overnight

cultures were centrifuged at 16,000g for 5 min at room temperature and the clear

supernatant was screened for the presence of degradation product by GC-MS analysis

(Sections 2.2.11 and 2.2.12). The GSL-metabolizing bacterial colonies were sub-cultured in

their corresponding media till OD600nm reached ~ 0.6 and then stored at – 80°C in glycerol

(40% v/v).

2.2.8 PCR amplification and identification of isolates

DNA was extracted directly from bacterial pellet using QIAamp DNA Mini Kit (Qiagen,

UK). Universal 16S primers designed by Wang et al. (1996) were used, and their sequences

were: AmpF 5’- GAGAGTTTGATYCTGGCTCAG- 3’  and  AmpR 5’-AAGGAGGTGATCCARCCGCA -

3’.  All amplification reactions were carried out in a Thermocycler PCR sprint Hybaid (Thermo

Electron). The thermocycle programme used for the amplifications in each PCR reaction (50

87

μL  total volume) consisted of one cycle of 96°C for 4 min followed by 25 cycles of 96°C for 30

s, 50°C for 30 s and 72°C for 60 s, and then one cycle of 72°C for 4 min as previously

described (Nueno-Palop & Narbad, 2011). The PCR products were resolved by

electrophoresis in a 1.2% (w/v) agarose (Sigma-Aldrich-Aldrich, UK) gel. Visualization of the

gene products on the gel was enabled by an addition of SYBRE Safe (Invitrogen, UK) staining

(1  μL/100  mL  gel  volume).  PCR  amplified  products  were  cleaned-up using Wizard PCR Preps

DNA   Purification   System   (Promega,   UK)   according   to   the   manufacturer’s   instructions   and  

used as a template in the DNA sequencing reactions using an ABI Prism BigDye v3.1

Terminator Cycle Sequencing Ready Reaction kit (Nueno-Palop & Narbad, 2011). Sequences

obtained were examined by BLAST search (Altschul et al., 1990) using the NCBI database.

The identities of the isolates were determined on the basis of the highest matching score.

2.2.9 Culturing conditions and sample collection for HPLC and GC-MS analyzes

Bacterium from glycerol stock was sub-cultured in a corresponding 5 mL liquid

medium overnight. The next day, 100 µL of overnight culture was sub-cultured in 900 µL

fresh medium containing 1 mM GSL substrate. Note that 1 mM glucose was added to MRS

media to promote growth of L. agilis R16 at 37°C while its optimum growth temperature is

at 30°C. Biological triplicates were incubated for each time interval; 0, 2, 4, 6, 8, 10, 16, and

24 h at 37°C in an anaerobic generation system using a 2.5 L Anaerogen jar supplied with an

AnaeroGen sachet (Oxoid, UK) to simulate the human gut conditions. At each time interval,

the pH and the OD600nm values of the culture broths were recorded using Corning pH meter

model 240 and LKB Novaspec II spectrophotometer, respectively before being centrifuged at

16,000g for 5 min at room temperature. Clear supernatant of 100 µL was transferred to a 1.5

mL Eppendorf tube for HPLC analysis (Section 2.2.3, desulfation of GSLs), and the other 900

µL of the supernatant was transferred to a 2 mL Eppendorf tube for GC-MS analysis (Section

2.2.11, extraction of degradation products). AnaeroGen sachets (Oxoid, UK) were

replenished at each time interval to maintain an effective anaerobic environment.

Appropriate controls include (i) incubations of each GSL without bacterial cells, (ii) bacterial

incubations without GSLs, (iii) incubations of liquid broths without GSLs and bacterial cells.

All experiments involving bacterial cells and cell-free-extracts were anaerobically incubated

in triplicayes unless otherwise stated.

88

2.2.10 Sample preparation for HPLC analysis and quantification of GSL from HPLC results

Sample preparation for HPLC analysis was carried out as previously described

(Section 2.2.3) and analyzed under current HPLC conditions (Section 2.2.4).

Concentration of GSL present in a 100 μL   of   sample   supernatant   was   determined  

using a response factor method (Brown et al., 2003). Since 100 μL  was taken from the total

sample volume of 1000 μL,  the corrected concentration of GSL present in the total volume of

1 mL was calculated as follows:

The volume of the supernatant used in extraction = 100  μL

From Brown et al., (2003) method, the amount of GSL (μmol) in  100  μL = X

Amount of GSL (μmol)  in  1000  μL  bacterial  culture = X × (1000/100)

Absolute concentration (μmol/mL)  of  GSL  in  1000  μL  bacterial  culture = X × 10

2.2.11 Sample preparation for GC-MS analysis

The degradation products were extracted from 900 µL supernatant in a 2 mL

Eppendorf tube. The same volume of dichloromethane (DCM) was added in the same tube

and the mixture was vortexed briefly and centrifuged at 16,000g for 2 min. This step resulted

in the separation of the mixture into two layers; the upper layer containing media broth and

the lower layer containing DCM with any ITC or NIT degradation products dissolved in it.

Only the lower layer was then transferred to a new 1.5 mL Eppendorf tube using a

hypodermic syringe. The extracts were dried over 0.5 g magnesium sulfate (BDH, UK) to

remove any remaining water residues that may interfere with downstream GC-MS analysis.

After that, the solutions were centrifuged at 16,000g for 10 min, and only the clear

supernatants were transferred to a 1.5 mL Agilent sample vial to be processed by GC-MS

directly. Sample concentration was avoided as this causes the massive losses of sample

through volatilization. For quantification, equal recovery was assumed, and an internal

standard was not considered necessary.

89

2.2.12 GC-MS analytical conditions for the detection of GSL degradation products

A Hewlett Packard 6890 series system, with the Hewlett Packard 5973 mass selective

detector (HP, UK) was used for GC-MS analysis. Two capillary columns; (i) Restek 200MS

crossbond trifluoro propyl methyl polysiloxane  (30m  ×  0.25mm  i.d.;  film  thickness,  0.25  μm)  

for ANIT (polar column necessary for analysing volatile ANIT) and (ii) Agilent SMS 5% phenyl

methyl siloxane capillary (30m   ×   0.25mm   i.d.;   film   thickness,   0.25   μm)   for the remaining

degradation products, were used with helium as the carrier gas (splitter inlet pressure, 40

kPa). For the Restek 200MS, the temperature was held constant at 50°C for the total 4 min

run. For Agilent SMS column, the temperature was kept at 50°C for 5 min and ramped to

150°C at 5°C min−1 for 25 min, and then ramped to 250°C at 5°C min−1 for 15 min. The total

45 min run was carried out with a flow rate of 1 mL/min, average velocity of 36 cm/s,

pressure of 7.56 psi and injection volume of 1 μL. Mass spectra were obtained by electron

ionization  (EI)  over  a  range  of  50−550  atomic  mass  units.  Ion  source  temperature  was  230°C,  

and the electron multiplier voltage was 70.1 eV. Peaks were identified by comparing

retention times, mass spectra and fragment ion fingerprints with those obtained from the

authentic standards. Mass spectral data and retention times of GSL degradation products

obtained from current GC-MS conditions are shown in Table 2.7.

90

Table 2.7 Mass spectral (MS) data of GSL degradation products

GSL substrate Expected degradation

productsa MS spectra data m/z

TR

(min)*

Sinigrin (SNG) Allyl-ITC (AITC) 99 (M+), 72, 65 6.9

Allyl-NIT (ANIT) 67 (M+), 52 2.8

Gluconasturtiin

(GNT) Phenethyl-ITC (PITC) 163 (M+), 105, 91, 77, 72 24.7

Phenethyl-NIT (PNIT) 131 (M+), 91, 62 18.6

Glucotropaeolin

(GTP) Benzyl-ITC (BITC) 149 (M+), 91, 65 22.0

Benzyl-NIT (BNIT) 117 (M+), 91, 62 15.5

Glucoiberverin (GIV) Iberverin (IBV) 147 (M+), 126, 101, 72, 61 20.6

Iberverin nitrile (IBV NIT)** 115 (M+), 75, 68, 61 13.8

Glucoerucin (GER) Erucin (ERN) 161 (M+), 115, 72, 61 23.8

Erucin nitrile (ERN NIT) 129 (M+), 87, 61, 55 17.4

Glucoiberin (GIB) Iberin (IBR)** 163 (M+), 130, 116, 100, 72 29.2

Iberin nitrile (IBR NIT) 131 (M+), 115, 69, 61 NA

Glucoraphanin (GRP) Sulforaphane (SFN) 177 (M+), 160, 115, 72 33.4

Sulforaphane nitrile (SFN NIT) 145(M+), 128, 82, 55 26.2

Glucobrassicin (GBS) Indole-3-carbinol (I3C) n.d. n.d.

*Retention time at which degradation product was eluted as detected by GC-MS analysis. **These compounds were not available as authentic standards however they were detected in the reactions during bacterial fermentations. Their identifications were made according to previously reported retention time and fingerprint profiles (Vaughn et al., 2005). n.d., not determined.

91

The external standard method was used for quantitative analysis of GSL degradation

products. Appropriate amount of authentic ITC/NIT standard was weighed out in a pre-

weighed 1.5 mL Agilent vial and appropriate volume of absolute ethanol was added to make

up a 100 mM stock solution. Various concentrations of each ITC/NIT standard in a range of

0.01- 2 mM diluted in DCM solutions were chromatographed by GC-MS separately from the

samples. Each concentration of each ITC/NIT standard was made in triplicates. The

calibration curves of various concentrations of each ITC/NIT standard versus peak areas were

generated (Figure 2.1 and 2.2), respectively. Note that both iberin (IBR) and iberverin nitrile

(IBV NIT) were detected as degradation products by GC-MS analysis, but corresponding

authentic standards were not purchased or synthesized. Quantification of their presence in

the samples was determined by using calibration curves of iberverin (IBV) and erucin nitrile

(ERN NIT), respectively which have similar structures, and assumingly have similar responses

under the same GC-MS conditions. Therefore, the approximate concentrations, rather than

absolute concentrations, were obtained for these two products.

Concentration of ITC/NIT degradation product present in a 900 μL   of   sample  

supernatant was determined using a calibration curve of each ITC/NIT standard (Figure 2.1

and Figure 2.2). Since 900 μL   was taken from the total sample volume of 1000 μL,   the

corrected concentration of ITC/NIT present in the total volume of 1 mL was calculated as

follows:

The volume of the supernatant used in extraction = 900  μL

From the standard curve, the amount  of  ITC/NIT  (nmol)  in  900  μL = Y

The amount of ITC/NIT (nmol)  in  1000  μL = Y × (1000/900)

Absolute concentration (nmol/mL) of ITC/NIT in  1000  μL = Y × 1.11

92

Figure 2.1 ITC standard curves from GC-MS analysis. (A) Allyl isothiocyanate (AITC). (B) Benzyl isothiocyanate (BITC). (C) Phenethyl isothiocyanate (PITC). (D) Erucin (ERN). (E) Iberverin (IBV). (F) Sulforaphane (SFN). Values are means of triplicates.

93

Figure 2.2 NIT standard curves from GC-MS analysis. (A) Allyl nitrile (ANIT). (B) Benzyl nitrile (BNIT). (C) Phenethyl nitrile (PNIT). (D) Erucin nitrile (ERN NIT). Values are means of triplicates.

94

2.2.13 Determination of the percentage product

Using X and Y values from sections 2.2.10 and 2.2.12, respectively, the percentage

product from GSL degradation can be calculated as follows:

Concentration of GSL at time 0 h = X0

Concentration of GSL at time of interest = Xt

Percentage product (%) = Y/(X0-Xt) x 100%

Note that the units of X and Y must be the same for the calculation. 2.2.14 Determination of stability and solubility of ITC/NIT standards

Since the decline of ITC degradation products over a time course was observed in all

GSL metabolisms in all bacterial fermentations, it was speculated that ITC products may be

unstable in aqueous solutions. Therefore, the control experiments were conducted. The NB

medium, Milli-Q water and different buffers containing each authentic ITC/NIT standards

without bacterial culture were anaerobically incubated   at   37˚C   over   a   time   course.   The  

supernatants were extracted for GC-MS analysis as previously described (Section 2.2.11) to

determine the stability of ITC/NIT in the corresponding medium/buffer over a time course at

the experimental conditions. Three types of buffers; (i) 0.1 M citrate phosphate buffer pH

7.0, (ii) 0.1 M PBS buffer pH 7.0 and (iii) 0.1 M Tris-Cl buffer pH 7.0 were tested in this

experiment.

Authentic AITC and PITC standards in a range of 0.05 – 1 mM were also tested for

their solubility in 0.1 M citrate buffer pH 7.0. If the linear regression was obtained at all

concentrations, this means these ITCs are easily dissolved in aqueous solution.

2.2.15 Resting cell experiments

(i) Myrosinase induction: Each of three bacteria was anaerobically cultured overnight

in 1 mL corresponding medium supplemented with 1 mM sinigrin (induced cells) and

without it (control cells) at  37˚C.  Both  samples  were  centrifuged  at  16,000g  for  15  min  at  4˚C.  

The supernatants were discarded, and the cells were washed twice in 1 mL 0.1 M citrate

95

phosphate buffer pH 7.0 to remove any GSL/ITC/NIT traces. After centrifugation, the washed

cells were re-suspended in 1 mL of the same buffer containing 1 mM gluconasturtiin and

incubated  at  37˚C  for  2 h. Afterwards, the cells were removed by centrifugation for 5 min at

16,000g, and 100 µL of the supernatant was prepared for HPLC analysis (Section 2.2.10) for

GSL detection, and 900 µL was prepared for GC-MS analysis (Section 2.2.11) for ITC/NIT

detection.

(ii) Reductase induction: E. coli O83:H1 NRG 857C cells were anaerobically cultured

overnight in 1 mL NB medium supplemented with either 1 mM glucoraphanin or

gluconastutiin (induced cells) and without any GSL supplementation   (control  cells)  at  37˚C.  

The remaining steps were performed as for myrosinase induction, but 1 mM sulforaphane

was added to the buffer containing resting cells in place of 1 mM gluconasturtiin. The

reductase activity was monitored over a time course by HPLC analysis.

2.2.16 Determination of metal ion dependency on NIT production from GSL metabolism in

bacterial resting cells

Since NIT production was not detected from GSL metabolism in bacterial resting cells

in 0.1 M citrate phosphate buffer, it was thought that metal ions may be required in the

buffer for NIT production. To test this hypothesis, chelating agent

ethylenediaminetetraacetic acid (EDTA) of 1, 5, 10 mM concentrations were individually

added to E. coli O83:H1 NRG 857C bacterial cultures containing 1 mM gluconasturtiin in NB

media that were incubated at 37˚C  anaerobically  for  16  h. The pH values of sample with and

without the addition of EDTA at T0h and T16h were recorded. The sample supernatants were

analyzed by GC-MS as previously described (Section 2.2.11) to determine whether NIT

production by bacterial fermentation in culture broth was inhibited by EDTA addition.

Additional experiment was conducted in which either 5 mM of CoCl2, CaCl2, MgCl2,

FeSO4, NiCl2, or MnCl2 (Sigma-Aldrich, UK) was individually added to bacterial resting cells

(induced by 1 mM gluconasturtiin overnight) in 0.1 M citrate phosphate buffer pH 7.0

containing 0.5 mM of a GSL substrate (as per section 2.2.15 (i) Myrosinase induction). The

reactions were anaerobically incubated at 37˚C  for  16  h.  Appropriate  control  samples  (i)  GSL-

containing buffer without bacterial cells or metal ions, (i) GSL-containing buffer plus each

96

metal ion without bacterial cells, (iii) GSL-containing buffer plus bacterial cells without any

metal ions were also included. The sample supernatants were analyzed by GC-MS as

previously described (Section 2.2.11) to determine whether NIT production by bacterial

resting cells in the buffer was promoted by metal ion addition.

2.2.17 Cell-free extract experiments

(i) Myrosinase induction: Each of three bacteria was anaerobically grown overnight at

37˚C  in  100  mL  corresponding  medium  supplemented  with  1  mM  sinigrin (induced cells) and

without it (controlled cells). The bacterial  culture  was  centrifuged  at  4,000g  for  15  min  at  4˚C,  

and the pellets were washed twice with 0.1 M citrate phosphate buffer pH 7.0 to remove

traces of GSL/ITC/NIT products if any. The bacterial suspensions in the same buffer (with 20

µL protease inhibitor cocktails (Melford, UK) added) were then disrupted using a cell

disruption machine (Constant Systems, UK) with two shots at 30k psi. Whole cell lysates

were centrifuged  at  16,000g  for  30  min  at  4˚C, and the clear supernatant referred to as cell-

free extracts were obtained and filtered sterile. The quantity of protein was determined

using   Bradford’s   reagent   (Bio-Rad, UK) (Section 2.2.20). Cell-free extracts (300 µL) were

anaerobically incubated with 1 mM gluconasturtiin (100 µL of 10 mM stock solution) in 600

µL  of  0.1  M  citrate  phosphate  buffer  pH  7.0  at  37˚C  for  16  h.  After  that,  sample  supernatant  

(100 µL) was prepared for HPLC analysis (Section 2.2.10) for GSL detection, and other

supernatant (900 µL) was prepared for GC-MS analysis (Section 2.2.10) for ITC/NIT product

detection.

(ii) Reductase induction: E. casselfiflavus NCCP-53 and E. coli O83:H1 NRG 857C cells

were anaerobically grown overnight in 100 mL NB medium supplemented with 1 mM

glucoraphanin (induced cells) and without it (controlled cells). The remaining steps were

performed as the above (i), but 1 mM glucoraphanin or glucoiberin was added to the buffer

containing cell-free extracts in place of 1 mM gluconasturttin. The reductase activity of the

cell-free extracts was monitored over a time course.

2.2.18 Determination of co-factor dependancy for reductase activity in cell-free extracts

To determine whether co-factors required for reductase activity, cell-free extracts of

E. coli O83:H1 NRG 857C were desalted using Econo-Pac 10DG desalting columns (Bio-Rad,

97

UK). The appropriate buffer used in the preparation of cell-free extracts (20 mL) was added

to the column to equilibrate it. The cell-free extract supernatant (3 mL) was then added to

the column, and the effluent was discarded. The elution buffer of choice (4 mL) was added

to the column to elute the higher molecular weight component(s) and the eluate (4 mL)

containing proteins was collected. The column was then washed with elution buffer (20 mL)

to  elute  remaining  salt  ions  from  the  column.  This  fraction  (20  mL)  collected  as  ‘salt  ions’ was

freeze-dried overnight in a freeze dryer (HETO Dry Winner). The obtained powder was re-

dissolved in 1 mL Milli-Q water  and  then  was  used  as  ‘freeze-dried  factors’.  

The reaction mixture (0.2 mL) contained desalted cell-free extract (100 µL), 0.25 mM

glucoraphanin (5 µL of 10 mM stock solution) and either 1 mM of CoCl2, CaCl2, MgCl2, FeSO4,

NiCl2, or MnCl2 (Sigma-Aldrich, UK) solution, 1 mM FAD, 1 mM NADH, 1 mM NADPH (20 µL of

10  mM  stock  solution)  or  a  combination  of  two  factors  or  all  factors  together  or  ‘freeze-dried

factors’   (100  µL)   in 0.1 M citrate phosphate buffer pH 7.0. This mixture was anaerobically

incubated  at  37˚C   for 16 h. The supernatant (100 µL) was then subjected to HPLC analysis

(Section 2.2.10) for the detection of reduction conversion of glucoraphanin. All chemicals

used in this experiment were purchased from Sigma-Aldrich, UK.

2.2.19 Determination of reductase activity in cell-free extracts in the bioconversion of

sulforaphane to erucin

The non-desalted cell-free extracts (300 µL) of E. coli O83:H1 NRG 857C (induced by 1

mM glucoraphanin overnight) were filtered sterile and then were anaerobically incubated

with 1 mM sulforaphane (Sigma-Aldrich, UK) (50 µL of 20 mM stock solution in absolute

ethanol)  in  650  µL  of  0.1  M  citrate  phosphate  buffer  pH  7.0  at  37˚C  for  5 and 22 h. After that,

sample supernatant (1 mL) was prepared for GC-MS analysis (Section 2.2.11) for the

detection of sulforaphane bioconversion to erucin.

2.2.20 Determination of pH and temperature optima for reductase activity in cell-free

extracts in the bioconversion of glucoraphanin to glucoerucin

To determine temperature and pH optima, temperature and pH conditions were

varied under the same experimental conditions using the non-desalted cell-free extracts.

The reaction mixture (0.2 mL) contained non-desalted cell-free extract (100 µL), 0.25 mM

98

glucoraphanin (5 µL of 10 mM stock solution) in 0.1 M citrate phosphate buffer. This mixture

was  anaerobically  incubated  at  37˚C  for 16 h. The supernatant (100 µL) was then subjected

to HPLC analysis for the detection of reduction bioconversion of glucoraphanin.

2.2.21 Protein quantification

Protein  quantification  was  performed  according  to  Bradford  (1976)  using  Bradford’s  

reagent (Sigma-Aldrich, UK). The total assay volume (1.55 mL) contained 0.05 mL of the

protein sample (0.1–1.4 mg/mL protein sample) and 1.5 mL of the Bradford Reagent

(B6916)(Sigma-Aldrich, UK) per Eppendorf tube which was gently mixed and incubated at

room temperature for 15 min. To create the calibration curve, 2 mg/mL BSA protein

standards (Sigma-Aldrich, UK) were serially diluted to produce a range from 0.1–1.4 mg/mL

of BSA. The samples (in triplicates) were transferred into cuvettes and the absorbance at 595

nm was recorded within 1 h. The protein sample concentration was determined by

comparison of the unknown samples to the standard curve prepared using the protein

standards (Figure  2.3).   ‘Blank’   controls   containing  no  protein  but  an  equivalent  amount  of  

Bradford’s   reagent  were used to zero the LKB Novaspec II spectrophotometer (Pharmacia,

UK).

Figure 2.3 Representative protein calibration curve. Various amount of BSA were plotted against absorbance at 595nm. Values are means of triplicates.

99

2.2.22 Denaturing sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE)

Proteins were analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis

(SDS-PAGE) (Laemmelli, 1970). Compositions to make 12.5% SDS-PAGE, loading buffer,

running buffer, staining/destaining solutions are shown in Table 2.8. Protein samples (20 µg)

were mixed with 2x SDS PAGE loading buffer. The mixture was boiled at 100°C for 2 min

prior to be resolved on a 12.5% SDS-PAGE gel in 1X SDS running buffer using Mini-PROTEAN

Tetra Cell apparatus (Bio-Rad, UK). The gel was electrophoresed at 180 V for 50 min.

Afterwards, it was stained with Coomassie Brilliant Blue R-250 staining solution for 15 min,

and washed once with Milli-Q water before being destained with destaining solution till the

gel background was clear, and the protein bands became visible.

Table 2.8 Compositions of SDS-PAGE, loading buffer, running buffer, staining/destaining

solutions

12.5% SDS-PAGE (for 2 gels) 12.5% separating gel (10 mL): 3.3 mL of Milli-Q

water, 4.0 mL of 30% (w/v) acrylamide, 0.8%

(w/v) bis-acrylamide stock solution (37.5:1) (Bio-

Rad, UK), 2.5 mL of Tris-Cl (1.5 M, pH 8.8), 100 µL

of 10% SDS (Sigma-Aldrich), 100 µL of 10%

ammonium persulfate (APS) (Sigma-Aldrich), 4 µL

of N,N,N',N'-Tetramethylethylenediamine

(TEMED) (Sigma-Aldrich)

4% stacking gel (3 mL): 2.1 mL of Milli-Q water,

0.5 mL of 30% (w/v) acrylamide, 0.8% (w/v) bis-

acrylamide stock solution (37.5:1) (Bio-Rad, UK),

0.38 mL of Tris-Cl (1.0 M, pH 6.8), 30 µL of 10%

SDS, 30 µL of 10% ammonium persulfate (APS)

(Sigma-Aldrich), 3 µL of N,N,N',N

Tetramethylethylenediamine (TEMED) (Sigma-

Aldrich)

SDS PAGE loading buffer 2x: 0.5 M Tris-HCl (pH 6.8), 4.4% (w/v) SDS, 20% (v/v) glycerol (GE

Healthcare), 2% (v/v) 2-mercaptoethanol (Sigma-Aldrich), and 0.05% bromophenol blue (Sigma-

Aldrich) in Milli-Q water

10X SDS Running Buffer (1 L): 30 g Tris-base (Sigma-Aldrich), 144 g glycine (Sigma-Aldrich), 10 g SDS

dissolved in Milli-Q water

Coomassie staining solution (1 L): 2.5 g Coomassie blue R-250 (Rio-Rad), 450 mL methanol, 450 mL

Milli-Q water, 100 mL acetic acid

Destaining solution (1 L): 300 mL methanol, 100 mL acetic acid and 600 mL Milli-Q water

100

2.2.23 Native gel electrophoresis

The native gels were made in a similar procedure as denatured SDS-PAGE gels

(section 2.2.21), except SDS was not added in any solutions shown in Table 2.8. Protein

samples (20 µg) were well-mixed with 2x native gel loading buffer. The Mini-PROTEAN Tetra

Cell gel apparatus was placed in the ice-filled box to prevent any protein denaturation

caused by heat during the run. The native gel was electrophoresed at 100 V for 2 h. After

that, one portion of the gel was stained with Coomassie Brilliant Blue R-250 solution, and the

other was incubated for 1 h at 37°C with substrate solution (sufficient to cover the gel).

Substrate solution contained 50 mM sodium acetate buffer (pH 7.0), 10 mM sinigrin, 0.5 mM

ascorbic acid, and 50 mM barium chloride (Sigma-Aldrich). GSL-degrading activity was

indicated as a white precipitate of barium sulfate on the gel. The positive control, the

purified myrosinase from S. alba (5 µg) was also loaded on the native gel to assess the

validity of native gels for GSL-degrading activity detection.

2.2.24 Statistical analysis

Data analysis including calculation of average values, percentage products, linear

regression and standard deviations were performed using Microsoft Excel 2010 or GraphPad

Prism 6. The data were analyzed by one-way analysis of variance (ANOVA). Differences were

considered significant if p < 0.05.

101

2.3 Results 2.3.1 Screening for GSL-metabolising human gut bacteria It is well-known that plant myrosinase can metabolize GSL to ITC or NIT product

depending upon the conditions of the hydrolysis. Accumulating evidence suggests that

certain bacteria may exhibit GSL-degrading activity (or myrosinase-like activity) as ITC and/or

NIT products were detected upon GSL incubation with bacterial culture in vitro. Thus, the

hypothesis of this experiment is that certain human gut bacteria may be able to metabolize

GSL to ITC or/and NIT product like plant myrosinase.

In order to develop the predominant GSL-degrading component of the microbial

community from a human fresh faecal sample, the use of enrichment culture for 16 days

with sinigrin as a selective carbon source in different growth media was implemented. Note

that M9 minimal media were used initially to grow bacteria with sinigrin supplementation,

but bacteria did not grow well (data not shown) and thus rich media were used instead. A

mixture of bacteria capable of metabolizing 1 mM sinigrin had a degradation capacity of 80%,

80% and 50% in MRS, WC and NB broths, respectively within 24 h anaerobic incubation at

37˚C.   The  negative   control   containing 1mM sinigrin without any faecal sample showed no

degradation suggesting that sinigrin is stable. Its degradation must be due to human gut

microbiota in faecal sample. Twenty bacterial colonies from the three agar plates were

individually sub-cultured in their corresponding liquid broths containing 1 mM sinigrin for 24

h  anaerobic   incubation  at  37˚C.  Sinigrin degradation by each culture was assessed by HPLC

analysis. Six colonies with a degradation capacity higher than 50% were subjected to 16S

rDNA gene analysis for strain identification. L. agilis R16 previously reported to have a high

capacity to degrade sinigrin (Palop et al., 1995) was not isolated from the enrichment culture

experiment, but was also included in this study. The strain identification, degradation rates

and degradation products obtained are shown in Table 2.9.

It was found that six GSL-degrading bacteria include three Gram-positive

Enterococcus, two Gram-negative Escherichia coli, and one clone SEW-E-011. Most of them

produced both AITC and ANIT from sinigrin degradation except Enterococcus sp. C213 and

Enterococcus faecium KT4S13 that only produced ANIT without AITC at all (Table 2.9).

Although clone SEW-E-011 yielded the highest degradation of sinigrin (Table 2.9), it was not

chosen for further study due to unavailability of its genome/proteome database.

102

L. agilis R16 (obtained from Palop et al., 1995), Enterococcus casseliflavus NCCP-53,

and Escherichia coli O83:H1 NRG 857C were chosen for further experiments as they all

produced both AITC and ANIT as degradation products with high degradation rates.

Importantly, accessibility to the genome/proteome database of relative E. casseliflavus

strains and E. coli O83:H1 NRG 857C would facilitate molecular cloning work.

Table 2.9 Bacterial isolates exhibiting > 50% degradation of 1 mM sinigrin in 24 h anaerobic

incubation  at  37˚C  

Broth Bacterial strain Type* OD600nm Degradation (%)

Degradation products

WC Clone SEW-E-011 Gram + 0.557 82.3 ± 1.05 AITC and ANIT

WC Enterococcus casseliflavus NCCP-53 Gram + 0.382 78.7 ± 2.13 AITC and ANIT

WC Enterococcus sp. C213 Gram + 0.487 75.1 ± 1.12 ANIT

MRS Lactobacillus agilis R16** Gram + 0.771 71.7 ± 0.98 AITC and ANIT

NB Escherichia coli O83:H1 NRG 857C Gram - 0.511 57.8 ± 3.11 AITC and ANIT

NB Escherichia coli UMNF18 Gram - 0.474 57.4 ± 2.45 AITC and ANIT

MRS Enterococcus faecium KT4S13 Gram + 0.524 50.1 ± 1.89 ANIT

Values are mean ± SD, n = 3, but only means are shown for OD600nm values. *Type of bacteria; Gram +, Gram-positive; Gram -, Gram-negative. **L. agilis R16 was obtained from Palop et al. (1995), not from the enrichment culture.

2.3.2 Isolation and purification of GSL substrates

Since GSLs are not generally commercially available except for sinigrin, most GSL

used in later bacterial fermentation experiments were purified from seed sources as

mentioned in section 2.2.1. For each GSL extraction, the previous method (Thies, 1988) was

used. The seed source from specific Cruciferous plant was ground and defatted and dried.

The defatted seed was boiled in methanol to extract the GSLs. The methanol extracts were

filtered and concentrated to dryness. The dried extract was re-dissolved in distilled water

and subjected to protein precipitation. The supernatant was loaded onto a pre-packed

diethylaminoethyl (DEAE)-Sephadex A25 column for GSL purification. The solution of K2SO4

was used to elute GSL from the column into ethanol and the potassium sulphate precipitate

103

was removed. The clear supernatant was then evaporated nearly to complete dryness. The

dried residues of GSL were dissolved in absolute methanol and were evaporated to dryness.

The dried residues were dissolved water and freeze-dried overnight. Appropriate amount of

the freeze-dried powder of GSL was weighed out and dissolved in water. The GSL solution

was then desulfated (Section 2.2.3) and analyzed by HPLC (Section 2.2.4). The purity (%) of

the isolated GSL was determined by comparing with pure sinigrin standard of the same

amount that was ran along side with the GSL sample. All isolated GSL substrates were obtained with above 90% purity. The HPLC

chromatograms of all purified GSLs used in this work are shown in Figure 2.4.

104

Figure 2.4 HPLC chromatograms of GSL substrates used in this work. (A) Glucoiberin with the retention time at 3.7 min. (B) Glucoraphanin with the retention time at 5.4 min. (C) Sinigrin with the retention time at 6.2 min. (D) Glucoerucin with the retention time at 13.4 min. (E) Glucotropaeolin with the retention time at 13.7 min. (F) Gluconasturtiin with the retention time at 16.4 min. (G) Glucobrassicin with the retention time at 15.4 min. Residual peaks within 5 min are dirts eluted from the column. Small peaks detected at early retention time represent unknown residues eluted from C18 reverse-phase column. 2.3.3 Time-course degradation product profiles of intact GSLs in individual bacterial

fermentations

From section 2.3.1, certain bacteria isolated from human faecal sample were capable

of metabolizing sinigrin to AITC and/or ANIT. Three bacteria; L. agilis R16, E. casseliflavus

105

NCCP-53, and E. coli O83:H1 NRG 857C that have never been reported for GSL-degrading

capacity (except Lactobacillus agilis R16 with sinigrin) were studied to determine whether

they were able to metabolize different GSLs (with different side chains) differently in terms

of types of products generated and degradation rates. To do so, each bacterial culture was

grown  on  a  GSL   substrate   in   their   corresponding  broths   anaerobically   at   37˚C  over   a   time  

course. Six different intact GSL substrates including sinigrin, glucotropaeolin, gluconasturtiin,

glucoerucin, glucoiberin, and glucoraphanin (1 mM) were used in this experiment (Figure

2.4) while glucobrassicin (0.1 mM) was used as a substrate in the later experiment. The

amount of degraded GSL over time was determined by HPLC analysis using authentic sinigrin

as external standard (section 2.2.10) and the amount of ITC and/or NIT product generated

upon GSL degradation was determined by GC-MS analysis using external authentic standard

calibration curve (section 2.2.12).

It was found that all three bacteria were capable of metabolizing most GSLs to

corresponding ITCs and/or NITs (Table 2.10) in the same way that plant myrosinase

hydrolyzed these GSLs. This suggests that human gut bacteria exhibited GSL-degrading

activity like plant myrosinase during GSL metabolism.

106

Table 2.10 Detection of ITC and NIT products from GSL metabolism in bacterial

fermentations

GSL Corresponding ITC and

NIT* LA EC ECO

Sinigrin

AITC √ √ √ ANIT √ √ √

Glucotropaeolin

BITC √ √ √ BNIT √ √ √

Gluconasturtiin

PITC √ √ √ PNIT X √ √

Glucoerucin

ERN √ √ √ ERN NIT √ √ √

Glucoiberin

IBR X √ X (IBV) IBR NIT X X X (IBV NIT)

Glucoraphanin

SFN X √ X (ERN) SFN NIT X X X (ERN NIT)

*Corresponding GSL degradation products typically produced by plant myrosinase and detected in bacterial fermentations. See abbreviations in Table 2.7. Brackets indicate the detection of the unexpected products instead. LA; L. agilis R16; EC, E. casseliflavus NCCP-53; ECO, E. coli O83:H1 NRG 857C.

Growth curves and pH values of bacterial cultures incubated with individual GSLs

were recorded (Figure 2.5). L. agilis R16 grown on any GSL substrates in modified MRS broth

containing 1 mM glucose reached stationary phase at 16 h of incubation (Figure 2.5A). Note

that 1 mM glucose was added to support the growth of this bacterium before it began to

metabolize GSL. Without glucose addition, this bacterium would grow slowly and not reach

its optimal growth (Palop et al., 1995). Highest OD600nm values of approximately 1.0 were

observed in L. agilis R16 when glucoiberin and glucoraphanin were substrates while lower

OD600nm values of 0.8-0.9 were observed among other GSL substrates (Figure 2.5A). E.

casseliflavus NCCP-53 grown in WC broth reached stationary phase at 5 h with the highest

OD600nm values of approximately 0.48-0.60 when glucotropaeolin, glucoiberin and

glucoraphanin were substrates while lower OD600nm values of 0.37-0.51 were observed

among other GSL substrates (Figure 2.5B). Similarly, E. coli O83:H1 NRG 857C reached

stationary phase at 8 h of incubation, but with the highest OD600nm values of approximately

0.50 found among all GSL substrates (Figure 2.5C).

107

Since L. agilis R16 is a lactic acid-producing bacterium (LAB) generating lactic acid

upon metabolism of glucose under anaerobic conditions, the pH decline over a time course

was expected. The pH values declined from 7.0 to 4.0 over a time course was observed in all

GSL substrates (Figure 2.5D). In contrast, E. casseliflavus NCCP-53 showed an increasing

trend in pH values initially from pH 6.5 to pH 7.0 or 7.5 after 6 h in all GSL substrates (Figure

2.5E). On the other hand, E. coli O83:H1 NRG 857C showed a decreasing trend in pH values

initially from pH 6.5 to below pH 6.0 for all GSLs (Figure 2.5F). Since GSLs alone did not

degrade spontaneously when incubated at 37°C for 24 h and bacteria did not accumulate

any GSL degradation products unless they were cultivated in medium containing GSLs

(Appendix I), the presence of any GSL degradation products can be ascribed only to bacterial

metabolism.

108

Figure  2.5  Growth  curves  and  pH  values  of  bacterial  cultures  incubated  with  individual  GSLs  anaerobically  at  37˚C  over  a  time course. (A) Growth kinetics (in log scale) of L. agilis R16 (LA) in modified MRS broth containing 1 mM glucose, (B) Growth curves of E. casseliflavus NCCP-53 (EC) in WC broth, (C) Growth curves of E. coli O83:H1 NRG 857C (ECO) in NB broth, (D) pH values of LA, (E) pH values of EC and (F) pH values of ECO. Values are means of triplicates. SNG, Sinigrin; GTP, Glucotropaeolin; GNT, Gluconasturtiin; GER, Glucoerucin; GIB, glucoiberin and GRP, glucoraphanin.

109

The graphs of time-course GSL degradation of each GSL substrate and GSL

degradation product formation by each bacterium are generated (Figure 2.6).

For sinigrin, complete degradation was detected within 8 h in E. casseliflavus NCCP-

53 while gradual degradation was observed over a time course in both L. agilis R16 and E.

coli O83:H1 NRG 857C with absolute degradation predicted to occur beyond 24 h incubation

(Figure 2.6A). This means sinigrin was more favoured by E. casseliflavus NCCP-53 than the

other two bacteria. Higher concentrations of both allyl isothiocyanate (AITC) and allyl nitrile

(ANIT) products were detected by GC-MS analyses in E. casseliflavus NCCP-53. The

decreasing trend in AITC production was observed in all bacteria after it peaked at 4 or 8 h

while NIT production seemed to increase till it peaked at 6 or 8 h and afterwards it remained

fairly constant (Figure 2.6A).

When glucotropaeolin was used as a substrate, complete degradation occurred in

only E. coli O83:H1 NRG 857C within 24 h while the other two bacteria yielded 96%

degradation at 24 h. The trends in production of both benzyl isothiocyanate (BITC) and

benzyl nitrile (BNIT) in all three bacteria were similar to those found in sinigrin degradation

(Figure 2.6B).

For gluconasturtiin, complete degradation at 6 h was found in both E. casseliflavus

NCCP-53 and L. agilis R16 and at 16 h in E. coli O83:H1 NRG 857C. The trends in production

of both phenethyl isothiocyanate (PITC) and phenethyl nitrile (PNIT) in all three bacteria

were similar to the previous two GSLs. Higher concentrations of total products were found in

E. casseliflavus NCCP-53 > E. coli O83:H1 NRG 857C > L. agilis R16 (Figure 2.6C). Surprisingly,

no PNIT product was detected in L. agilis R16 despite several attempts were repeated. The

reason for this is unknown.

For glucoerucin, complete degradation was found at 24 h in all three bacteria (Figure

2.6D). The trends in production of erucin (ERN) and erucin nitrile (ERN NIT) were also similar

in all bacteria and similar to those of other ITC/NIT production from the previous GSLs.

For glucoiberin and glucoraphanin, the results are rather different from other GSL

substrates. Glucoiberin was 40% degraded at 24 h in E. casseliflavus NCCP-53 and 50% at 8 h

in E. coli O83:H1 NRG 857C while 10% degradation was found in L. agilis R16 at 24 h (Figure

2.6E). The ITC product found in a very low concentration in E. casseliflavus NCCP-53 was

iberin (IBR) while iberverin (IBV) was found in a much higher concentration in E. coli O83:H1

110

NRG 857C. Interestingly, no corresponding NIT product was found in E. casseliflavus NCCP-53

while iberverin nitrile (IBV NIT) was found in in E. coli O83:H1 NRG 857C. No ITC/NIT

production was detected in L. agilis R16 (Figure 2.6E).

Similarly, glucoraphanin was 50% degraded at 24 h in E. casseliflavus NCCP-53 and at

4 h in E. coli O83:H1 NRG 857C while 11% degradation at 24 h was found in L. agilis R16

(Figure 2.6F). The ITC product found in a very low concentration in E. casseliflavus NCCP-53

was sulforaphane (SFN) while ERN was found in a much higher concentration in E. coli

O83:H1 NRG 857C. Interestingly, no NIT product was found in E. casseliflavus NCCP-53 while

ERN NIT was found in E. coli O83:H1 NRG 857C. No ITC/NIT production was detected in L.

agilis R16 (Figure 2.6F)

For glucoiberin and glucoraphanin substrates, they were degraded in descending

rates by E. coli O83:H1 NRG 857C > E. casseliflavus NCCP-53 > L. agilis R16. E. casseliflavus

NCCP-53 metabolized these GSLs to the expected ITC products without any NIT products. In

contrast, E. coli O83:H1 NRG 857C metabolized these GSLs to ITC/NIT degradation products

that are reduced analogues of the expected degradation products. The reason for this was

further investigated in section 2.3.10.

111

Figure 2.6 Time-course  degradation  product  profiles  of  bacterial  cultures  anaerobically  incubated  with  individual  GSLs  at  37˚C.  The left, middle and right panels show GSL degradation, ITC production and NIT production, respectively. (A) Sinigrin. (B) Glucotropaeolin. (C) Gluconasturtiin. (D) Glucoerucin. (E) Glucoiberin. (F) Glucoraphanin. LA, L. agilis R16; EC, E. casseliflavus NCCP-53; ECO, E. coli O83:H1 NRG 857; AITC, allyl isothiocyanate; ANIT, allyl nitrile; BITC, benzyl isothiocyanate; BNIT, benzyl nitrile; PITC, phenethyl isothiocyanate; PNIT, phenethyl nitrile. Values are means ± SD, n = 3.

112

Figure 2.6 Time-course degradation product profiles of bacterial cultures  anaerobically  incubated  with  individual  GSLs  at  37˚C.  The left, middle and right panels show GSL degradation, ITC production and NIT production, respectively. (A) Sinigrin. (B) Glucotropaeolin. (C) Gluconasturtiin. (D) Glucoerucin. (E) Glucoiberin. (F) Glucoraphanin. LA, L. agilis R16; EC, E. casseliflavus NCCP-53; ECO, E. coli O83:H1 NRG 857C; IBR, Iberin; IBV, Iberverin; IBV NIT, Iberverin nitrile; SFN, Sulforaphane; ERN, Erucin; ERN NIT, Erucin nitrile. Values are means ± SD, n = 3.

113

The GC-MS chromatograms of ITC/NIT degradation products from the metabolisms

of GSLs are shown in Figure 2.7. Fingerprint fragment ions of GSL degradation products are

shown in Figure 2.8.

114

Figure 2.7 GC-MS chromatograms of degradation products of different GSLs. (A) SNG was metabolized to ANIT 1, at 2.83 min and AITC 2, at 6.98 min. (B) GTP was metabolized to BNIT 3, at 15.5 min and BITC 4, at 22.0 min BNIT. (C) GNT was metabolized to PNIT 5, at 18.6 min and PITC 6, at 24.7min. (D) GIV was metabolized to IBV NIT 7, at 13.8 min and IBV 8, at 20.6 min. (E) GER was metabolized to ERN NIT 9, at 17.4 min and ERN 10, at 23.8 min. (F) GIB was metabolized to IBR 11, at 29.2 min. (G) GRP was metabolized to SFN 12, at 33.4 min. Referred to Table 2.7.

115

Figure 2.8 Fingerprint fragment ions of GSL degradation products generated by GC-MS analysis. (1) ANIT, (2) AITC, (3) BNIT, (4) BITC, (5) PNIT, (6) PITC, (7) IBV NIT, (8) IBV, (9) ERN NIT, (10) ERN, (11) IBR and (12) SFN. Referred to Table 2.7.

116

Since ITC productions from glucoiberin and glucoraphanin metabolized by E.

casseliflavus NCCP-53 over a time course were rather low (Figure 2.6E and 2.6F, respectively),

the corresponding GC-MS chromatograms showed high noise signals (data not shown).

Therefore, selected ion monitoring (SIM) mode was used to generate representative

chromatograms of ITC products of these GSLs in this bacterium (Figure 2.9A and 2.9B,

respectively).

Figure 2.9 GC-MS chromatograms of ITC degradation products from glucoiberin and glucoraphanin metabolized by E. casseliflavus NCCP-53 over a time course. (A) Glucoiberin was metabolized to iberin, 1 at 29.23 min. (B) Glucoraphanin was metabolized to sulforaphane, 2 at 33.40 min. Selected ion monitoring (SIM) mode was used to generate these chromatograms in order to minimize noise signals.

117

The GC-MS chromatograms of ITC/NIT productions from glucoiberin and

glucoraphanin metabolized by E. coli O83:H1 NRG 857C over a time course are shown in

Figure 2.10A and 2.10B, respectively.

Figure 2.10 GC-MS chromatograms of ITC/NIT degradation productions from glucoiberin and glucoraphanin metabolized by E. coli O83:H1 NRG 857C over a time course. (A) Glucoiberin was metabolized to iberverin nitrile, 1 at 13.8 min and iberverin, 2 at 20.6 min. (B) Glucoraphanin was metabolized to erucin nitrile, 3 at 17.4 min and erucin, 4 at 23.8 min.

118

These results show that ITC productions from glucoiberin and glucoraphanin

metabolized by both bacteria declined after 8 h. However, NIT productions from E. coli

O83:H1 NRG 857C increased over time and remained fairly constant at 8 and 16 h as

previously seen in the trends of NIT productions from other GSL substrates in all bacteria.

Glucobrassicin, as another GSL substrate, was used at a much lower concentration

(0.1 mM) due to low materials of purified glucobrassicin. Only two bacteria, E. coli O83:H1

NRG 857C and E. casseliflavus NCCP-53, were tested on this GSL. Bacterial growth and

glucobrassicin degradation in the two bacteria were monitored over a time course (Table

2.11).

Table 2.11 Bacterial growth and glucobrassicin degradation in E. coli O83:H1 NRG 857C and

E. casseliflavus NCCP-53 over a time course

Time (h)

ECO EC

OD600nm GBS (µM)a Degradation (%)b OD600nm GBS (µM)a Degradation

(%)

0 0.12 ± 0.01 100 ± 2 0 0.10 ± 0.02 100 ± 4 0

2 0.21 ± 0.02 76 ± 3 24 ± 5 0.18 ± 0.02 65 ± 5 35 ± 7

4 0.32 ± 0.02 61 ± 5 39 ± 6 0.35 ± 0.03 62 ± 4 38 ± 6

6 0.44 ± 0.03 58 ± 2 42 ± 3 0.42 ± 0.02 58 ± 3 42 ± 5

8 0.48 ± 0.02 55 ± 5 45 ± 6 0.51 ± 0.01 35 ± 2 65 ± 3

16 0.51 ± 0.01 53 ± 4 47 ± 5 0.56 ± 0.02 27 ± 4 73 ± 5 aRemaining glucobrassicin (GBS) in the reaction solution, 100 µM was an initial concentration. bDegradation (%) of GBS = the amount of GSL degraded in (%) relative to the initial amount. Values are mean ± SD, n = 3. ECO, E. coli O83:H1 NRG 857C; EC, E. casseliflavus NCCP-53

Both bacteria were able to degrade glucobrassicin with 47% degradation at 16 h by E.

coli O83:H1 NRG 857C and 73% degradation by E. casseliflavus NCCP. The degradation

products of glucobrassicin by these bacteria were not detected under current GC-MS

conditions. The products may be extremely volatile, and thus LC-MS analysis instead of GC-

MS analysis may be required for further analysis. The HPLC chromatograms of glucobrassicin

metabolism in E. coli O83:H1 NRG 857C and E. casseliflavus NCCP-53 over a time course are

shown in Figure 2.11.

119

Figure 2.11 HPLC chromatograms of bacterial degradation of glucobrassicin over a time course. (A) Sinigrin external standard at 6 min. (B) E. coli O83:H1 NRG 857C gradually degraded glucobrassicin (14.6 min) over time with 47% degradation at 16 h. (C) E. casseliflavus NCCP-53 gradually degraded glucobrassicin (14.6 min) with 73% degradation at 16 h. The arrows indicate the insets showing enlarged peaks of glucobrassicin. Peaks no. 1 to 3 found in all figures are unknown residues, no. 4 and 5 are probably polar GSLs emerged during the metabolism. These figures are representatives of triplicates with similar results.

120

The time taken to obtain 50% decline of each GSL substrate by three bacteria is

shown in Table 2.11. Among all GSL substrates tested so far, the degradation rates are in

descending order: gluconasturtiin > sinigrin > glucoerucin > glucotropaeolin > glucoiberin or

glucoraphanin in L. agilis R16 (Table 2.12). E. casseliflavus NCCP-53 degraded these

substrates in a slightly different descending order: gluconasturtiin > glucoerucin > sinigrin >

glucotropaeolin > glucobrassicin > glucoraphanin > glucoiberin. In contrast, E. coli O83:H1

NRG 857C showed rather different descending order: glucoerucin > glucoraphanin >

gluconasturtiin > sinigrin > glucoiberin > glucotropaeolin > glucobrassicin.

Table 2.12 Time taken to obtain 50% decline of each GSL substrate by three bacteria

GSL Time (h) of 50% decline*

LA EC ECO

Sinigrin 4 3.4 6

Glucotropaeolin 7 6 12

Gluconasturtiin 1.8 1.8 5

Glucoerucin 6 3 3.5

Glucoiberin > 24 > 24 8

Glucoraphanin > 24 24 4

Glucobrassicin NA 15 > 16 *Time taken to observe 50% decline from each GSL. Values are means of triplicates. LA, L. agilis R16; EC, E. casseliflavus NCCP-53; ECO, E. coli O83:H1 NRG 857C.

From these results, it was speculated that the polarity and the size of GSL side chain

may influence degradation rate in which GSL with the less polar side chain was more easily

degraded by L. agilis R16 and E. casseliflavus NCCP-53 as the polarity of GSL substrates in

descending order is: glucoiberin > glucoraphanin > sinigrin > glucotropaeolin > glucoerucin >

gluconasturtiin > glucobrassicin. However, this speculation does not hold true for E. coli

O83:H1 NRG 857C. The percentage product of each product from each bacterial GSL

metabolism is shown in Table 2.13.

121

Table 2.13 Percentage products of each ITC/NIT product from all GSL metabolisms by each

bacterium

GSL Bacteria Product Percentage product (%)*

2 h 4 h 6 h 8 h 10 h 16 h 24 h

SNG ECO AITC 27 ± 4 28 ± 5 20 ± 4 19 ± 3 13 ± 2 7 ± 1 6 ± 0.2

ANIT 18 ± 5 25 ± 4 30 ± 6 14 ± 3 12 ± 5 11 ± 2 14 ± 3

EC AITC 32 ± 3 6 ± 0.5 15 ± 2 12 ± 1 9 ± 3 10 ± 2 7 ± 1

ANIT 21 ± 7 11 ± 1 22 ± 3 28 ± 6 28 ± 4 22 ± 3 16 ± 4

LA AITC 13 ± 4 10 ± 2 11 ± 3 23 ± 1 5 ± 2 2 ± 0.3 1 ± 0.1

ANIT ND ND ND ND 4 ± 2 3 ± 0.6 4 ± 1

GTP ECO BITC 9 ± 2 18 ± 3 17 ± 2 20 ± 4 11 ± 3 3 ± 0.2 1 ± 0.2

BNIT 5 ± 1 14 ± 4 13 ± 2 18 ± 2 22 ± 9 15 ± 6 7 ± 1

EC BITC 3 ± 0.3 18 ± 2 30 ± 4 23 ± 3 21 ± 2 10 ± 1 6 ± 0.4

BNIT 11 ± 3 18 ± 5 22 ± 2 19 ± 3 20 ± 4 15 ± 2 11 ± 3

LA BITC ND 6 ± 2 11 ± 3 14 ± 5 7 ± 1 4 ± 2 2 ± 0.9

BNIT 16± 2 15 ± 5 12 ± 4 14 ± 4 13 ± 3 15 ± 4 14 ± 3

GNT ECO PITC ND 12 ± 3 13 ± 2 22 ± 6 10 ± 1 4 ± 0.3 2 ± 0.1

PNIT 5 ± 1 10 ± 3 19 ± 3 8 ± 3 6 ± 2 8 ± 2 5 ± 1

EC PITC ND 6 ± 0.4 11 ± 4 13 ± 3 7 ± 3 2 ± 1 0.6 ± 0.2

PNIT ND 20 ± 6 27 ± 4 24 ± 9 21 ± 6 13 ± 5 16 ± 3

LA PITC 9 ± 2 13 ± 5 16 ± 3 11 ± 3 7 ± 1 5 ± 2 3 ± 1

PNIT ND ND ND ND ND ND ND

GER ECO ERN ND 4 ± 2 27 ± 8 34 ± 9 35 ± 17 20 ± 5 5 ± 1

ERN NIT ND 10 ± 3 6 ± 2 23 ± 6 22 ± 4 16 ± 1 10 ± 0.5

EC ERN 27 ± 6 40 ± 14 33 ± 6 24 ± 10 25 ± 6 9 ± 4 7 ± 2

ERN NIT 11 ± 3 24 ± 8 24 ± 7 23 ± 2 18 ± 4 12 ± 1 12 ± 1

LA ERN ND ND 16 ±4 15 ± 3 9 ± 1 6 ± 0.2 4 ± 0.2

ERN NIT ND 7 ± 2 5 ± 0.6 6 ± 1 6 ± 0.2 3 ± 1 2 ± 0.6

GIB ECO IBV 14 ± 8 12 ± 2 8 ± 2 8 ± 1 5 ± 1 4 ± 1 1 ± 0.2

IBV NIT 16 ± 6 9 ± 2 7 ± 2 5 ± 1 4 ± 0.4 5 ± 0.4 3 ± 0.7

EC IBR ND 2 ± 1 3 ± 0.5 3 ± 1 1 ± 0.1 1 ± 0.3 0.4 ± 0.1

GRP ECO ERN 4 ± 2 5 ± 1 6 ± 1 10 ± 1 6 ± 2 3 ± 0.5 1 ± 0.2

ERN NIT 8 ± 2 4 ± 1 5 ± 0.6 6 ± 3 9 ± 2 8 ± 1 6 ± 1

EC SFN 6 ± 1 2 ± 0.8 3 ± 1 3 ± 0.6 2 ± 0.3 1 ± 0.2 0.3 ± 0.1

*The amount of product (mol) in (%) relative to the digested amount of GSL (mol). SNG, sinigrin; GTP, glucotropaeolin; GNT, gluconasturtiin; GER glucoerucin; AITC, allyl isothiocyanate; ANIT, allyl nitrile; BITC, benzyl isothiocyante; BNIT, benzyl nitrile; PITC, phenethyl isothiocyanate; PNIT, phenethyl nitrile; ERN, erucin; ERN NIT, erucin nitrile; GIB, glucoiberin; IBV, iberverin; IBR, iberin; IBV NIT, iberverin nitrile; GRP, glucoraphanin; SFN, sulforaphane; ND, Not detected. Values are means ± SD, n = 3. LA, L. agilis R16; EC, E. casseliflavus NCCP-53; ECO, E. coli O83:H1 NRG 857C.

These results showed that the total percentage products of both ITC and NIT

122

products from each GSL metabolism in all three bacteria never reached 100% (Table 2.13).

The highest total percentage products from each GSL in all bacteria were found between 4

and 8 h where ITC productions peaked. E. coli O83:H1 NRG 857C showed the highest total

percentage products of sinigrin (53%) at 4 h, glucotropaeolin (38%) at 8 h, glucoerucin (57%)

at 8-10 h, gluconasturtiin (32%) at 6 h, glucoiberin (34%) at 2 h and glucoraphanin (16%) at 8

h. E. casseliflavus NCCP-53 showed the highest total percentage products of sinigrin (40%) at

8 h, glucotropaeolin (42%) at 8 h, glucoerucin (64%) at 4 h, gluconasturtiin (38%) at 6 h,

glucoiberin (3%) at 6-8 h and glucoraphanin (3%) at 6-8 h. L. agilis R16 showed the highest

total percentage products of sinigrin (23%) at 8 h, glucotropaeolin (28%) at 8 h, glucoerucin

(21%) at 8 h, gluconasturtiin (16%) at 6 h. The total percentage products from the

metabolisms of glucoiberin and glucoraphanin in all bacteria were much lower than those

obtained from other GSLs.

2.3.4 Stability of ITC/NIT degradation products

It was noticeable that ITC productions from metabolisms of all GSL substrates

declined over a time course following their peaks and the total percentage product of both

ITC and NIT formation from each GSL metabolism never reached 100% (Table 2.13).

Therefore, the study of the stability of ITC/NIT in culture broths was carried out to determine

whether these ITC/NIT products were stable under current fermentation conditions. To

achieve this, each authentic ITC/NIT standard was added to NB media with and without E.

coli O83:H1 NRG 857C cells incubated over  a  time  course  at  37˚C  under  anaerobic  conditions.  

The levels of all ITC standards tested including AITC, BITC, PITC, IBV and ERN declined sharply

from 1 mM (at time 0 h) to 0.2 – 0.4 mM within 8 h in NB broths without bacterial cells

(Figure 2.12A). These results suggest that ITCs had short lives in NB broths. Assumingly, the

same occurrence may be applied to MRS and WC broths even without bacterial cells. When

bacterial cells were added to those ITC standards in NB broths, the decline of ITCs levels

seemed to be faster (from 1 mM to 0.1 – 0.3 mM within 8 h) (Figure 2.12B). That explains

why the decreasing trends in ITC production from all GSL metabolisms in all bacterial cells

were observed (Figure 2.6). However, it is not clear whether these authentic ITC standards

were degraded to other metabolites. ITCs may be conjugated to the medium components

and/or cellular components. This indeed needs further investigation. The same test was

123

performed to determine the stability of NIT standards including ANIT, BNIT, PNIT, IBV NIT

and ERN NIT in NB broths. The levels of these NIT standards in NB media without bacterial

cells were rather stable over time, except for ANIT with a decline from 1 to 0.7 mM after 2 h

(Figure 2.12C). With cells, slight declines of NITs from 1 to 0.8-0.9 mM were observed while

ANIT declined sharply from 1 to 0.68 mM within 2 h (Figure 2.12D).

124

Figure 2.12 Stability of 1 mM ITC/NIT standards in NB broths with/without E. coli O83:H1 NRG 857C cells over a time course. (A) ITC standards without cells. (B) ITC standards with cells. (C) NIT standards without cells. (D) NIT standards with cells. Values are means of triplicates.

125

The time taken to obtain 50% or 25% decline in each ITC or NIT level, respectively in

NB broths with or without the presence of bacterial cells is shown in Table 2.14.

Table 2.14 Time taken to obtain 50% or 25% decline in each ITC or NIT level, respectively in

NB broths with or without the presence of E. coli O83:H1 NRG 857C cells

ITC Time (h) of 50% decline

NIT Time of 25% decline

No cells Cells No cells Cells

AITC 6.2 2.3 ANIT 4 1.8

BITC 5.9 3.6 BNIT > 24 > 24

PITC 6.2 2.4 PNIT > 24 > 24

IBV 7.5 3.2 IBV NIT > 24 > 24

ERN 5.2 3.7 ERN NIT > 24 > 24

SFN 5.1 1.8 Values are means of triplicates. This experiment was carried out at 37˚C under anaerobic conditions over a time course.

The time taken to obtain 50% or 25% decline in each ITC or NIT level was shortened

by the presence of bacterial cells in NB broths. This suggests the possibility of interactions of

ITCs or NITs with bacterial cellular components, but with less pronounced effect on NIT

levels.

To test the solubility of ITCs in aqueous solution, various concentrations of AITC and

PITC were extracted in distilled water. The linear regressions were obtained from both ITCs

(Figure 2.13) indicating that ITCs of physiological concentrations were easily dissolved in

distilled water. Therefore, the decline of ITC productions over a time course is not due to

insolubility of ITCs in aqueous solutions.

126

Figure 2.13 Solubility of various concentrations of authentic ITC standards in distilled water. Values are means of triplicates.

The stability of ITCs was further investigated in aqueous solutions over a time course;

distilled water, 0.1 M citrate phosphate buffer pH 7.0, 0.1 M PBS buffer pH 7.0 and 0.1 M

Tris-Cl buffer pH 7.0. It was shown that all ITCs tested in aqueous solutions declined over

time with different rates indicating the instability of ITCs in aqueous solutions (Figure 2.14).

127

Figure 2.14 Stability of 1 mM authentic ITC standards in various buffers without E. coli O83:H1 NRG 857C cells over a time course. Several aqueous buffer solutions of 0.1 M and pH 7.0 were used. (A) Tris-Cl. (B) PBS. (C) Citrate phosphate. (D) Distilled water. Values are means of triplicates.

128

The time taken to obtain 50% decline of each ITC in different aqueous solutions is

shown in Table 2.15.

Table 2.15 Time taken to obtain 50% decline in each ITC level in various aqueous solutions

without the presence of E. coli O83:H1 NRG 857C cells

ITC Time (h) of 50% decline

Water Citrate Phosphate PBS Tris Cl

AITC 7.1 1.7 4.8 4.6

BITC 11 5 7.2 7.5

PITC 18 5.6 7.3 7.2

IBV 15 3.8 12.5 7.3

ERN 15 3.7 5.2 7.1

SFN 6 n.d. n.d. n.d.

Values are means of triplicates. n.d., not determined.

These results showed that the levels of ITCs declined fastest in citrate phosphate

buffer and slowest in distilled water. PBS and Tris-Cl buffers had less pronounced effects

than citrate phosphate buffer. It was assumed that the presence of citrate, phosphates and

chlorides in these buffers may be attributed to faster decline of ITC levels in comparion with

water molecules alone.

2.3.5 Time-course degradation product profiles of DS-GSLs in individual bacterial

fermentation

Previously, it was reported that rat intestinal microbiota digest DS-sinigrin to form

allyl nitrile (ANIT) and 1-cyano-2,3-epithiopropane (Lu et al., 2011). To determine whether

the three bacteria can metabolize any DS-GSLs to NIT products or other metabolites,

different DS-GSLs of 1 mM including DS-sinigrin, DS-glucotropaeolin, DS-gluconasturtiin, DS-

glucoerucin, and DS-glucoraphanin were used as substrates in the same fermentation

129

conditions as in the previous experiments. The negative controls containing only DS-GSLs

without bacterial cells showed no degradation products at 24 h suggesting DS-GSLs were

stable under experimental conditions (Appendix II).

The generation of NIT products from the metabolisms of DS-GSLs in all three bacteria

is summarized in Table 2.16. These results indicate that DS-GSLs are precursors to pure NIT

production whereas intact GSLs are precursors to both ITC and NIT production during the

same bacterial fermentation in culture broths.

Table 2.16 Detection of NIT product from DS-GSL metabolism in bacterial fermentations

DS-GSL Corresponding NIT product LA EC ECO

DS-Sinigrin ANIT √ √ √

DS-Glucotropaeolin BNIT √ √ √

DS-Gluconasturtiin PNIT √ √ √

DS-Glucoerucin ERN NIT √ √ √

DS-Glucoraphanin SFN NIT X X X (ERN NIT)

*Corresponding NIT products expected to be produced from DS-GSL metabolism. See abbreviations in Table 2.7. Brackets indicate detection of the unexpected products instead. LA; L. agilis R16; EC, E. casseliflavus NCCP-53; ECO, E. coli O83:H1 NRG 857C. A tick (√) means the product was detected and a cross (X) means not detected.

The results showed that E. coli O83:H1 NRG 857C was able to metabolize all DS-GSLs

to the corresponding NIT products without any ITC products (Figure 2.15).

130

Figure 2.15 GC-MS chromatograms of degradation products of DS-GSLs metabolized by individual three bacteria. (A) DS-sinigrin was metabolized to allyl nitrile (ANIT), 1 (2.83 min). (B) DS-glucotropaeolin was metabolized to benzyl nitrile (BNIT), 2 (15.5 min). (C) DS-gluconasturtiin was metabolized to phenethyl nitrile (PNIT), 3 (18.6 min) by E. casseliflavus NCCP-53 and E. coli O33:H1 NRG 857C. (D) DS-glucoerucin was metabolized to erucin nitrile (ERN NIT), 4 (17.4 min). (E) DS-glucoraphanin was metabolized to pure erucin nitrile, 5 (17.47 min) by E. coli O33:H1 NRG 857C only. These figures are representatives of triplicates with similar results.

131

However, E. casseliflavus NCCP-53 was able to metabolize all DS-GSLs except for DS-

glucoraphanin while L. agilis R16 was not able to metabolize DS-gluconasturtiin to PNIT or

DS-glucoraphanin to ERN NIT. These results are in accordance with the previous results

showing absence of ERN NIT product from glucoraphanin metabolism in E. casseliflavus

NCCP-53 (Figure 2.6F) and the absence of PNIT product from gluconasturtiin metabolism in L.

agilis R16 (Figure 2.6C). It is clear that L. agilis R16 was unable to produce PNIT from either

gluconasturtiin or DS-gluconasturtiin. The reason for this is unknown.

NIT production profiles from DS-GSLs metabolisms in individual bacteria over a time

course are shown in Figure 2.16. Concentrations of NIT products among all three bacteria

increased over a time course with slight higher NIT productions in comparison with those

obtained from the metabolisms of intact GSLs. At 24 h, E. coli O83:H1 NRG 857C and L. agilis

R16 showed 56% ANIT production in (%) relative to the initial dose of DS-sinigrin and 52%

from E. casseliflavus NCCP-53. All three bacteria showed approximately 30% BNIT

production from DS-glucotropaeolin at 24 h. Similarly, 25-30% PNIT productions from DS-

gluconasturtiin were detected in all bacteria except L. agilis R16 (no PNIT production from

the metabolism of gluconasturtiin either). For ERN NIT productions from DS-glucoerucin, 38,

34 and 30% productions were observed in E. coli O83:H1 NRG 857C, E. casseliflavus NCCP-53

and L. agilis R16, respectively at 24 h. However, the concentrations of ERN NIT from DS-

glucoraphanin metabolism in E. coli O83:H1 NRG 857C (i.e. 13% production) were much

lower than other NIT products from the metabolisms of other DS-GSL substrates. The other

two bacteria were unable to produce ERN NIT from this substrate. This indicates that DS-

glucoraphanin substrate was not favoured by these bacteria.

132

Figure 2.16 NIT productions from DS-GSLs metabolisms in individual bacteria over a time course. (A) DS-sinigrin was metabolized to allyl nitrile (ANIT). (B) DS-glucotropaeolin was metabolized to benzyl nitrile (BNIT). (C) DS-gluconasturtiin was metabolized to phenethyl nitrile (PNIT) by ECO and EC. (D) DS-glucoerucin was metabolized to erucin nitrile (ERN NIT). (E) DS-glucoraphanin was metabolized to erucin nitrile (ERN NIT) by ECO only. LA, L. agilis R16; EC, E. casseliflavus NCCP-53; ECO, E. coli O83:H1 NRG 857C. Values are means ± SD, n = 3.

133

2.3.6 Resting cell experiments

From the previous results (section 2.3.3), the three bacteria were capable of

metabolizing GSLs to ITCs and/or NITs like plant myrosinase suggesting that these bacteria

may exhibit GSL-degrading activity. To determine whether putative GSL-degrading enzyme

activity in three bacteria are inducible or constitutively expressed (like plant myrosianse),

resting cell experiments were carried out. Firstly, E. casseliflavus NCCP-53 was used in a trial

using different buffers to find the optimal buffer for ITC production from GSL metabolism.

Washed suspensions of non-induced cells (grown without 1 mM GSL in WC media overnight)

and induced cells (grown on 1 mM sinigrin) were anaerobically incubated with 1 mM

gluconasturtiin in various types of buffers for 8 h at 37˚C. As a result, both induced and non-

induced cells of E. casseliflavus NCCP-53 in all buffers produced PITC products without PNIT

products (Table 2.17) even in the buffers containing no ascorbic acid (i.e. an activator for

plant myrosinases). However, higher PITC products were observed in induced cells suggesting

the inducibility of bacterial GSL-degrading enzymes. Several buffer conditions supported PITC

production in the resting cell experiments, but citrate phosphate buffer pH 7.0 was chosen to

be used in further experiments as it gave the highest PITC production from the metabolism of

gluconasturtiin.

Table 2.17 PITC production from gluconasturtiin metabolism in E. casseliflavus NCCP-53 resting  cells  in  different  buffers  for  8  h  at  37˚C under anaerobic conditions

No. Buffer Sample PITC  products  (μM)

1 0.03 M MES + 6 mM MgCl2 + 2 mM ascorbic acid + 1 mM GNT N 73 ± 12 I 116 ± 14

2 0.1 M Citrate phosphate buffer pH 7.0 + 1 mM GNT N 123 ± 8

I 276 ± 15

3 0.1 M Citrate phosphate buffer pH 6.0 + 1 mM GNT N 118 ± 11

I 197 ± 21

4 0.05 M PBS buffer pH 7.0 + 1 mM GNT N 113 ± 18

I 203 ± 10

5 0.1 M Tris Cl buffer pH 7.0 + 1 mM GNT

N 83 ± 14 I 176 ± 20

GNT, gluconasturtiin; I, Induced cells; N, Non-induced cells. Values are means ± SD, n = 3

134

A further resting cell experiment was carried out in which either non-induced or

induced resting cells of individual three bacteria was anaerobically incubated with 1 mM

gluconasturtiin in 0.1 M citrate phosphate buffer pH 7.0 for only 2 h at 37˚C. The putative

bacterial GSL-degrading activity from L. agilis R16 is most likely to be inducible as the induced

cells degraded gluconasturttin to completion while the non-induced cells showed 71%

degradation. Higher degradation of gluconasturtiin was also detected in E. casseliflavus NCCP-

53 in spite of no major difference in bacterial growth. In contrast, E. coli O83:H1 NRG 857C

showed complete degradation in both induced and non-induced cells (Table 2.18). The reason

for this is still not known. However, PITC products were only detected in induced cells and not

in non-induced cells (Table 2.18). There was no detection of NIT products in any bacteria. It

seems likely that putative bacterial GSL-degrading activity from E. casseliflavus NCCP-53 and E.

coli O83:H1 NRG 857C is inducible since their induced cells produced PITC products in buffers,

but the non-induced cells did not.

Table 2.18 Degradation of gluconasturtiin by bacterial resting cells in 0.1 M citrate

phosphate buffer pH 7.0 anaerobically incubated for  2  h  at  37˚C

Properties LA EC ECO

I N I N I N

OD600nm 0.723 ± 0.004

0.738 ± 0.014

0.513 ± 0.008

0.498 ± 0.012

0.425 ± 0.002

0.448 ± 0.007

GNT degradation

(%)* 100 71 ± 12 80 ± 8 71 ± 5 100 100

PITC (µM) 75 ± 11 ND 41 ± 15 ND 44 ± 13 ND

*The amount of GNT disappearance in (%) relative to the initial dose (1 µmol/mL) of GNT. GNT, Gluconasturtiin; LA, L. agilis R16; EC, E. casseliflavus NCCP-53; ECO, E. coli O83:H1 NRG 857C; I, Induced; N, Non-induced; ND, Not detected; PITC, Phenethyl isothiocyanate. Values are means ± SD, n = 3

135

2.3.7 Determination of metal ion dependency on NIT production from GSL metabolism in

the buffer and culture broths

To determine the conditions required for PITC/PNIT production from gluconasturtiin

and DS-gluconasturtiin metabolism in resting cells of E. coli O83:H1 NRG 857C induced by 1

mM sinigrin overnight, both buffer and NB broth were used in this experiment. It was found

that gluconasturtiin and DS-gluconasturtiin were stable in the buffer as no degradation

products were found without bacterial induced resting cells (Figure 2.17A). Both PNIT and

PITC were produced when induced resting cells were incubated in NB broths (Figure 2.17B)

while only PITC was produced when induced resting cells were incubated in 0.1 M citrate

phosphate buffer pH 7.0 (Figure 2.17C). Only PNIT was produced from DS-gluconasturtiin in

NB broths with induced resting cells (Figure 2.17D). This suggests that there may be

something present in NB broths and absent in the buffer that are responsible for PNIT

production.

Figure 2.17 GC-MS chromatograms of different degradation products of gluconasturtiin or DS-gluconasturtiin metabolized by E. coli O83:H1 NRG 857C induced resting cells in different incubation conditions. (A) No product was found in 0.1 M citrate phosphate buffer pH 7.0 containing either gluconasturtiin or DS-gluconasturtiin without resting cells. (B) Phenethyl nitrile (PNIT), 1 (18.60 min) and phenethyl isothiocyanate (PITC), 2 (24.69 min) were produced from gluconasturtiin in NB broth with induced resting cells. (C) Only PITC, 2 was produced from gluconasturtiin in 0.1 M citrate phosphate buffer pH 7.0 with induced resting cells. (D) Only PNIT, 1 was produced from DS-gluconasturtiin in NB broth with induced resting cells. These figures are representatives of triplicates.

136

To determine whether metal ions are required for NIT production in NB broth, three

concentrations (1, 5, 10 mM) of ethylenediaminetetraacetic acid (EDTA) as a metal ion-

chelating agent were added into E. coli O83:H1 NRG 857C bacterial cultures in NB broth

containing 1 mM gluconasturtiin anaerobically incubated at 37˚C  for  16  h. The results showed

that EDTA had a significant inhibitory effect on both PITC and PNIT production from

gluconasturtiin metabolism in E. coli O83:H1 NRG 857C in NB broth (Table 2.19). The pH

values dropped from 6.5 to 5.3-5.6 upon EDTA addition, but it was not the cause of NIT

product inhibition since NIT production was still detected in this pH range during bacterial

fermentation experiments.

Table 2.19 Effect of EDTA on ITC/NIT production from gluconasturtiin metabolism in E. coli

O83:H1 NRG 857C in NB broth for 16 h anaerobic incubation  at  37˚C

EDTA (mM) PNIT (mM) PITC (mM) pH

0 0.185 ± 0.031 0.083 ± 0.015 6.5 ± 0.1

5 0.004 ± 0.001 ND 5.6 ± 0.2

10 ND ND 5.4 ± 0.1

20 ND ND 5.3 ± 0.1 ND, Not detected; PITC, Phenethyl isothiocyanate; PNIT, Phenethyl nitrile. Values are means ± SD, n = 3 An additional experiment was conducted in which either metal ions, CoCl2, CaCl2,

MgCl2, FeSO4, NiCl2, MnCl2 of the same concentration (5 mM) was individually added to E. coli

O83:H1 NRG 857C resting cells (induced by 1 mM sinigrin in NB broth overnight). The reaction

mixture was anaerobically incubated in 0.1 M citrate phosphate buffer pH 7.0 containing 0.5

mM gluconasturtiin at 37˚C   for   16   h.   Appropriate   control   samples   (i)  GSL-containing buffer

without bacterial cells or metal ions, (i) GSL-containing buffer plus each metal ion without

bacterial cells, (iii) GSL-containing buffer plus bacterial cells without any metal ions were also

included.

The results showed that the control, GSL-containing buffer plus Fe2+ without bacterial

cells gave NIT products (4.5 µM) while the other controls with the addition of other metal ions

showed no NIT products indicating the existence of non-enzymatic NIT production promoted

137

by Fe2+ addition (Figure 2.18). The addition of Co2+, Ca2+, Ni2+, Mn2+ had an inhibitory effect on

PITC production and no PNIT production was detected. However, the addition of Fe2+ ions

resulted in PNIT production, but with reduced PITC production. This indicates that NIT

production in bacterial resting cells in the buffer was promoted by 5 mM Fe2+ ions (Figure

2.18). Interestingly, the addition of Mg2+ had stimulatory effect on PITC production without

PNIT production.

Figure 2.18 Effects of metal ions on PITC/PNIT production in E. coli O83:H1 NRG 857C induced resting cells. Each metal ion solution (5 mM) was incubated in 0.1 M citrate phosphate buffer pH 7.0 containing 0.5 mM gluconasturtiin with induced resting cells of E. coli O83:H1 NRG 857C at 37˚C for 16 h. Control 1, cells and 5 mM Fe2+ without GSL; Control 2, GSL and 5 mM Fe2+ without cells; Control 3, cells and GSL without any metal ions. Values are means of triplicates.

The GC-MS chromatograms of degradation products from gluconasturtiin metabolism

in E. coli O83:H1 NRG 857C resting cells upon addition of metal ions in 0.1 M citrate

phosphate buffer pH 7.0 are shown in Figure 2.19.

0 50 100 150 200 250Control 1Control 2Control 3

2+Co2+Ca2+Ni2+Mn2+Mg2+Fe3+Fe

PNITPITC

Concentration (M)

138

Figure 2.19 GC-MS chromatograms of degradation products from gluconasturtiin metabolism in E. coli O83:H1 NRG 857C resting cells upon addition of metal ions. (A) PITC, 1 (24.69 min) production was promoted by Mg2+ ions and inhibited by Mn2+ ions. There was no PNIT production. (B) PNIT, 2 (18.60 min) production was promoted by Fe2+ ions with reduced PITC production. Each metal ion solution (5 mM) was incubated in 0.1 M citrate phosphate buffer pH 7.0 containing 0.5 mM gluconasturtiin with induced resting cells of E. coli O83:H1 NRG 857C at 37˚C for 16 h. These figures are representatives of triplicates.

The same experiments were carried out with other four GSLs, glucotropaeolin,

glucoerucin, glucoiberin and glucoraphanin to determine whether Fe2+ ions would promote

NIT production from the metabolism of GSLs in the buffer. Similarly, Fe2+ ions promoted NIT

139

production from all GSLs tested, but with reduced ITC production compared to the contols (no

addition of Fe2+ ions) (Table 2.20). The non-enzymatic NIT productions in trace amounts were

also observed in the control samples containing only GSLs and 5 mM Fe2+ without bacterial

resting cells (Table 2.20).

Table 2.20 Effect of Fe2+ ions (5 mM) on ITC/NIT production from the metabolisms of GSLs

(0.5 mM) by E. coli O83:H1 NRG 857C induced resting cells anaerobically incubated in 0.1 M

citrate phosphate buffer pH 7.0 for 16 h at  37˚C

GSL GTP GER GIB GRP

Products  (μM) BITC BNIT ERN ERN NIT IBV IBV

NIT ERN ERN NIT

No cells + Fe2+ ND 6 ± 2 ND 4 ± 1 ND 5 ± 2 ND 6 ± 1

Cells + no ions 52 ± 12 ND 55 ± 10 ND 42 ± 8 ND 48 ± 6 ND

Cells + Fe2+ 36 ± 5 69 ± 9 34 ± 8 57 ± 7 16 ± 4 46 ± 9 19 ± 5 49 ± 8 GTP, Glucotropaeolin; GER, glucoerucin: GIB, glucoiberin; GRP, glucoraphanin; BITC, Benztl isothiocyanate; BNIT, Benzyl nitrile; ERN, Erucin; ERN NIT, Erucin nitrile; IBV, Iberverin; IBV NIT, Iberverin nitrile; ND, Not detected. Values are means ± SD, n = 3.

Since NIT production was not detected in the negative control sample containing only

GSLs without bacterial cells in the culture broths MRS, WC and NB used during bacterial

fermentations at 24 h. It was postulated that the concentration of Fe2+ ions present in the

culture broths may be so low (possibly lower than 5 mM used in the current experiment) that

it was not sufficient to promote the non-enzymatic NIT production from GSL in the culture

broths. Therfore, NIT production in the culture broths with the presence of bacterial cells was

thought to be mainly enzymatically-driven with the aid of low concentration of Fe2+ ions.

140

2.3.8 Cell-free extract experiments from all bacteria

Since putative bacterial GSL-degrading enzyme activity from all three bacteria is likely

to be inducible (Section 2.3.6), these bacterial cells were induced with 1 mM sinigrin in

corresponding broths overnight to allow the induction of GSL-degrading enzyme activity to be

studied in cell-free extract experiments. To determine whether the GSL-degrading enzyme

activity is in the soluble fraction, cell-free extracts of all three bacteria after cell disruption

were incubated with 1 mM gluconasturtiin in 0.1 M citrate phosphate buffer pH 7.0 for 16 h

and the detection of ITC/NIT formation was carried out by GC-MS analysis. No ITC/NIT

products were observed from any cell-free extracts. This result indicates that no myrosinase-

like activity could be detected under the conditions tested. It is thought that bacterial GSL-

degrading enzyme activity may be inactive in this in vitro activity assay. Since bacterial GSL-

degrading activity was only detected in bacterial anaerobic fermentation experiments, it was

postulated that bacterial GSL-degrading activity may be sensitive to air exposure in cell-free

extract experiments. Another possibility is that bacterial GSL-degrading activity may be part of

the protein complex or multiple component protein system that once disrupted (upon cell

breaking up during cell-free extraction in this case) renders GSL-degrading enzyme activity

inactive.

2.3.9 Determination of GSL-degrading enzyme activity from bacterial whole cell lystaes on

the native gels

Since no bacterial GSL-degrading enzyme activity were detected from cell-free extracts

of all bacteria, it was hypothesized that bacterial GSL-degrading enzyme might be

extracellular (i.e. secreted into the culture broths). Interestingly, several groups report

extracellular β-glucosidase secreted by fungi such as A. fumigatus Z5 (Liu et al., 2012),

Daldinia eschscholzi (Aphichart et al., 2007) and from the algal lytic bacterium Sinorhizobium

kostiense AFK-13 (Kim & Lee, 2007). To test this hypothesis, the overnight culture broths from

both sinigrin-induced and non-induced E. coli O83:H1 NRG 857C cultures were centrifuged,

and the clear broths were analyzed on SDS-PAGE to see whether there were any proteins

secreted into the culture broths. No protein bands were found (Figure 2.20). This indicates

that bacterial GSL-degrading enzyme was still in the whole cell lysate.

141

Figure 2.20 SDS-PAGE analysis of E. coli O83:H1 NRG 857C proteins. Whole cell lysates, supernatants and culture broths (after removal of cell pellets) collected from both induced (I) and non-induced (N) cultures of E. coli O83:H1 NRG 857C were analyzed on 4-12% SDS-PAGE. Lane M is PageRuler prestained protein marker (ThermoScientific, UK).

Native PAGE gels have been a successful assay for GSL-degrading enzyme activity

(Shikita et al., 1999; Thangstad et al., 2004; Ahuja et al., 2011). Myrosinase can be located

after native polyacrylamide gel electrophoresis by incubating the gel with sinigrin. The sulfate

released from sinigrin by myrosinase action reacts with barium ion in the incubation solution

to produce white precipitate of barium sulfate. This activity gel assay was used to determine

whether there is bacterial GSL-degrading enzyme activity in the whole cell lysates. The

positive control, purified plant myrosinase from white mustard (Sinapis alba), was also

included to test the validity of the assay. Only the positive control displays GSL-degrading

enzyme activity as indicated by white precipitate. However, no GSL-degrading enzyme activity

was detected from bacterial whole cell lysates or supernatants (Figure 2.21).

142

Figure 2.21 Native gel electrophoresis for GSL-degrading enzyme activity test. The whole cell lysates and supernantants of non-induced (N) and induced (I) samples of E. coli O83:H1 NRG 857C were analyzed on the native gel. (A) Coomassie-stained native gel showing the band of purified plant myrosinase (boxed) (B) Activity native gel showed white precipitate of barium sulfate (boxed) from the activity of plant myrosinase indictaing the presence of GSL-degrading enzyme activity towards sinigrin. Other white precipitates between the lanes are false positive. Other white precipitates between the lanes are false positive because once the bands

excised and incubated with a solution of gluconasturtiin (i.e. aromatic GSL), phenethyl

isothiocyanate (PITC) product was not detected while the plant myrosinase gave this product.

This result confirms that putative bacterial GSL-degrading enzyme activity was not detected in

cell-free extracts.

2.3.10 Sulfoxide reduction of glucoiberin and glucoraphanin by reductase activity in E. coli

O83:H1 NRG 857C

From the previous time-course study during bacterial fermentation experiments

(Section 2.3.3), E. coli O83:H1 NRG 857C and E. casseliflavus NCCP-53 were able to metabolize

glucoiberin and glucoraphanin to different products (referred to Table 2.10). E. casseliflavus

NCCP-53 produced only iberin (IBR) from glucoiberin and produced only sulforaphane (SFN)

143

from glucoraphanin without any NIT products. In contrast, E. coli O83:H1 NRG 857C produced

iberverin (IBV) and iberverin nitrile (IBV NIT) from glucoiberin, erucin (ERN) and erucin nitrile

(ERN NIT) from glucoraphanin. Since no corresponding products i.e. iberin (IBR) and

sulforaphane (SFN) were detected from metabolisms of glucoiberin and glucoraphanin in E.

coli O83:H1 NRG 857C, instead the reduced sulfide ITC and NIT analogues were produced. It

was suspected that this bacterium exhibits reductase activity that can reduce the sulfoxide

groups on methylsulfinylalkyl GSLs i.e. glucoiberin and glucoraphanin to produce

methylthioalkyl GSLs i.e glucoiberverin and glucoerucin with the sulfide groups and thus

produced IBV, IBV NIT and ERN, ERN NIT as products. The hypothetic scheme of the putative

bacterial reductase in E. coli O83:H1 NRG 857C is shown in Figure 2.22. Thus, the aim of this

section was to identify the bacterial reductase from E. coli O83:H1 NRG 857C intact cells and

cell-free extract.

Figure 2.22 Hypothetic scheme of the putative bacterial reductase in E. coli O83:H1 NRG 857C cells. A similar scheme is thought to occur in the metabolism of glucoiberin in these two bacteria. EC, E. casseliflavus NCCP-53; ECO, E. coli O83:H1 NRG 857C.

It was found that glucoiberin and glucoraphanin disappeared over time during E. coli

O83:H1 NRG 857C fermentation. Concomitantly, their reduced analogues, glucoiberverin and

glucoerucin increased until 4 or 8 h and then declined (Figure 2.23) due to subsequent

metabolism catalyzed by possibly bacterial GSL-degrading enzyme to IBV, IBV NIT and ERN,

144

ERN NIT. The differences in degradation products from the metabolism of glucoiberin and

glucoraphanin in E. coli O83:H1 NRG 857C and E. casseliflavus NCCP-53 could be explained by

the reduction of methylsulfinylalkyl GSLs i.e. glucoiberin and glucoraphanin to (reduced)

methylthioalkyl GSLs i.e. glucoiberverin and glucoerucin (Figure 2.23), respectively over time

by a putative reductase enzyme present in E. coli O83:H1 NRG 857C intact cells.

Figure 2.23 Reduction bioconversion of glucoiberin/glucoraphanin to glucoiberverin/glucoerucin during E. coli O83:H1 NRG 857C fermentation over a time course. The cultures were anaerobically fermented   at   37˚C   and the presence of GSLs was detected by HPLC analysis. Values are means of triplicates.

To determine whether reductase enzyme of E. coli O83:H1 NRG 857C was inducible by

GSLs, E. coli O83:H1 NRG 857C was grown overnight without glucoraphanin supplementation

in NB broth and the following day the cell-free extracts were obtained. The cell-free extracts

were anaerobically incubated with either 1 mM glucoiberin or glucoraphanin in 0.1 M citrate

phosphate  buffer  pH  7.0  at  37˚C  over  a  time  course.  There  was  no  reduction  of  these GSLs.

However, when a bacterial culture of E. coli O83:H1 NRG 857C was induced with 1 mM

glucoraphanin in NB medium overnight before isolating the cell-free extracts, the reduction of

the sulfoxide groups of glucoiberin and glucoraphanin was observed. The HPLC

chromatograms displaying the reduction bioconversion of glucoiberin and glucoraphanin to

glucoiberverin and glucoerucin by E. coli O83:H1 NRG 857C cell-free extracts (obtained from

145

glucoraphanin-induced cells) are shown in Figure 2.24. The reduction bioconversion (%) from

glucoiberin/glucoraphanin to glucoiberverin/glucoerucin was found to be 67 and 56%,

respectively at 16 h (Table 2.19A). These results strongly suggest the existence of active,

inducible, soluble, cytosolic reductase enzyme in E. coli O83:H1 NRG 857C.

Figure 2.24 HPLC chromatograms of methylsulfinylalkyl GSLs converted to methylthioalkyl GSLs by E. coli O83:H1 NRG 857C cell-free extracts (obtained from glucoraphanin-induced cells) over a time course. This experiment was carried out in   0.1   M   citrate   phosphate   buffer   pH   7.0   at   37˚C   under  anaerobic conditions and samples were taken at 4 and 16 h. (A) Glucoiberin, 1 (3.62 min) was converted to glucoiberverin, 2 (11.12 min). (B) Glucoraphanin, 3 (5.33 min) was converted to glucoerucin, 4 (13.8 min). These figures are representatives of triplicates.

To determine whether the reductase enzyme of E. coli O83:H1 NRG 857C is inducible

by other GSL substrates other than glucoraphanin, 1 mM gluconasturtiin was added to E. coli

O83:H1 NRG 857C culture overnight and the following day its cell-free extracts were

incubated with 1 mM glucoraphanin over a time course. It was found that gluconasturtiin-

induced cell-free extracts showed similar trends in the bioconversion of 0.25 mM

146

glucoiberin/glucoraphanin to glucoiberverin/glucoerucin (Table 2.21B) when compared with

those from glucoraphanin-induced cell-free extracts (Table 2.21A). This indicates that

bacterial reductase of E. coli O83:H1 NRG 857C is likely to be induced by both methylsulfinyl

GSL i.e. glucoraphanin and also aromatic GSL i.e. gluconasturtiin.

Table 2.21 Reduction bioconversion of glucoraphanin by cell-free extracts of E. coli O83:H1

NRG 857C (obtained from gluconasturtiin/glucoraphanin-induced cells) over a time course

GSL conversion Bioconversion (%)

4 h 16 h A) Glucoraphanin-induced From glucoiberin to glucoiberverin

18 ± 5 67 ± 9

From glucoraphanin to glucoerucin

23 ± 4 56 ± 8

B) Gluconasturtiin-induced From glucoiberin to glucoiberverin

15 ± 6 60 ± 15

From glucoraphanin to glucoerucin

21 ± 8 61 ± 9

Values are means SD, n = 3. This experiment was carried out in 0.1 M citrate phosphate buffer pH 7.0  at  37˚C  under  anaerobic  conditions with 100 µL cell-free extracts using 0.25 mM GSL substrate.

E. casseliflavus NCCP-53 cell-free extracts were also tested for reduction bioconversion

of methylsulfinylalkyl GSLs as previously described with the cell-free extracts of E. coli O83:H1

NRG 857C. It was found that there was no reduction bioconversion of glucoraphanin from any

cell-free extracts of E. casseliflavus NCCP-53, indicating a lack of reductase activity in this

bacterium (data not shown) under the conditions tested.

Not only E. coli O83:H1 NRG 857C cells were able to convert methylsulfinylalkyl GSLs

to methylthioalkyl GSLs, they were also able to convert methylsulfinylalkyl ITC i.e.

sulforaphane to a reduced methylthioalkyl ITC i.e. erucin within 5 h during anaerobic

fermentations at  37˚C  in  NB  broths (Figure 2.25).

147

Figure 2.25 GC-MS chromatograms showing reduction bioconversion of sulforaphane to erucin by E. coli O83:H1 NRG 857C intact cells. (A) Sulforaphane, 1 (33.40 min) at 0 h. (B) A decline of sulforaphane, 1 led to an increase of erucin, 2 (23.86  min)  after  5  h  anaerobic  incubation  at  37˚C in NB broths. These figures are representatives of triplicates. Similar experiments were carried out using cell-free extracts and resting cells of E. coli

O83:H1 NRG 857C (induced with 1 mM glucoraphanin overnight) incubated with 1 mM

sulforaphane over a time course in   0.1   M   citrate   phosphate   buffer   pH   7.0   at   37˚C   under  

anaerobic conditions. It was found that 28.4 and 11.3% of initial sulforaphane was converted

to erucin at 5 and 22 h in cell-free extracts (Table 2.22). Similarly 21.6 and 8.6% bioconversion

of initial sulforaphane to erucin was detected in induced resting cells of E. coli O83:H1 NRG

857C at 5 and 22 h, respectively (Table 2.22). The control sample containing only 1 mM

sulforaphane without cell-free extracts or resting cells in the buffer showed the decline of 49

and 80% in sulforaphane level at 5 and 22 h (Table 2.22). The levels of both sulforaphane and

erucin observed in cell-free extracts and resting cells also declined between 5 and 22 h (Table

2.22). This was possibly due to the instability of sulforaphane and erucin that may be

conjugated to the buffer components (referred to section 2.3.4) or sulforaphane/erucin

degradation or further metabolism to unknown and/or undetected metabolites.

5 10 15 20 25 30 350

100000

200000

300000 5 10 15 20 25 30 350

200000

400000

600000To

tal I

on C

urre

nt

Retention Time (min)

B

A

2

Tota

l Ion

Cur

rent

1

148

Table 2.22 Reduction bioconversion of sulforaphane to erucin by cell-free extracts and

resting cells of E. coli O83:H1 NRG 857C (induced with 1 mM glucoraphanin overnight) over

a time course

Time (h)

Cell-free extracta Resting cells in bufferb Controlc

SFN  (μM) ERN  (μM) SFN  (μM) ERN  (μM) SFN  (μM)

0 1000 (100) ND 1000 (100) ND 1000

5 245 ± 13 (24.5) 284 ± 9 (28.4) 23 ± 5 (2.3) 216 ± 10 (21.6) 512 ± 13 (51.2)

22 83 ± 10 (8.3) 113 ± 11 (11.3) 4 ± 0.6 (0.4) 86 ± 7 (8.6) 203 ± 9 (20.3)

aCell-free extracts (300 µL) was added to a 1 mL reaction containing 1000 μM sulforaphane at 0 h. bResting cells of OD600nm = 0.5 was added to a 1 mL reaction mixture. cThe control sample containing only 1 mM sulforaphane without cell-free extracts or resting cells in the buffer. Values representing the ITC concentrations remained in the solution are means ± SD, n = 3. This experiment was carried out in 0.1 M citrate phosphate   buffer   pH   7.0   at   37˚C   under   anaerobic  conditions. ND, Not detected; SFN, Sulforaphane; ERN, Erucin. The GC-MS chromatograms showing the reduction bioconversion of sulforaphane to

erucin by E. coli O83:H1 NRG 857C induced cell-free extracts over a time course in 0.1 M

citrate  phosphate  buffer  pH  7.0  at  37˚C  under  anaerobic  conditions  are  shown  in  Figure  2.26.

149

Figure 2.26 GC-MS chromatograms showing the reduction bioconversion of sulforaphane to erucin by E. coli O83:H1 NRG 857C induced cell-free extracts over a time course. (A) Sulforaphane, 1 (33.40 min) at 0 h and the control sample containing only sulforaphane without cell-free extract incubated for 22 h. (B) A decline of sulforaphane, 1 led to an increase of erucin, 2 (23.86 min) after 5 h. (C) The levels of both sulforaphane, 1 and erucin, 2 declined at 22 h. Reactions were performed in 0.1 M citrate  phosphate  buffer  pH  7.0  at  37˚C  under  anaerobic  conditions. These figures are representatives of triplicates. From these results, the scheme of glucoraphanin metabolism in E. casseliflavus NCCP-

53 and E. coli O83:H1 NRG 857C has been proposed in Figure 2.27. This is thought to be the

same for the metabolism of glucoiberin.

150

Figure 2.27 Different metabolic fates of glucoraphanin in E. coli O83:H1 NRG 857C (ECO) and E. casseliflavus NCCP-53 (EC). Erucin is a reduced analogue of sulforaphane. A similar scheme is thought to occur in the metabolism of glucoiberin in these two bacteria.

2.3.11 Mg2+- and NAD(P)H- dependent reductase activity and its optimal pH and

temperature

The aim of this section was to characterize E. coli O83:H1 NRG 857C reductase in cell-

free extracts. In general, reductase enzymes require reducing co-factors such as NADH or

NADPH for activity (De Jongh et al., 1987; Lamed & Zeikus, 1980; Panagiotou &

Christakopoulos, 2004; Stenbaek et al., 2008). Dehydrogenase enzymes require both a

divalent metal ion and a reducing co-factor for oxidoreductase activity (Goulian & Beck, 1966;

Hektor et al., 2002; Hori et al., 1967). Some reductase enzymes can use either NADH or

NADPH for activity (Corvest et al., 2012; Verduyn et al., 1985; Vermeulen et al., 2006).

To determine whether E. coli O83:H1 NRG 857C reductase require such factors, the

cell-free extracts were desalted to remove any factors present in the cell-free extracts. The

resulting desalted cell-free extracts were added with 1 mM of either Mg2+, FAD, NADH or

NADPH or a combination of two factors. Other metal ions e.g. Mn2+, Fe2+, Fe3+, Ca2+, Co2+, Ni2+,

used in section 2.3.7 were also tested in this experiment in addition to Mg2+. The reaction

mixture was incubated with 0.25 mM glucoraphanin for 24 h and was analyzed by HPLC for

the detection of the bioconversion to glucoerucin. As a result, both Mg2+ and either of NADH

151

or NADPH were required as co-factors for reductase activity (Table 2.23). Other metal ions

with/without combination of reducing reagents gave negative results.

Table 2.23 Reduction bioconversion (%) of 0.25 mM glucoraphanin to glucoerucin by the

addition of 1 mM of co-factor(s) in desalted cell-free extracts of E. coli O83:H1 NRG 857C

(obtained from glucoraphanin-induced cells) within  24  h  at  37˚C  under  anaerobic  conditions

Treatmentsa % Conversion of glucoraphanin to glucoerucinb

Cell-free extract* 71 ± 5

Freeze-dried factors fraction** 85 ± 8

FAD or NADH or NADPH ND

Mg2+, Mn2+, Fe2+, Fe3+, Ca2+, Co2+, or Ni2+ ND

FAD + Mg2+ ND

NADH + Mg2+ 52 ± 4

NADPH + Mg2+ 58 ± 9

NAD(P)H + other metal ion ND

aDesalted cell-free extracts were mixed with co-factor(s) indicated. bConversion (%) to glucoerucin (mol) as (%) relative to the initial dose of glucoraphanin (mol) as analyzed from HPLC chromatograms. Values are means ± SD, n = 3. *Cell-free extracts (non-desalted) alone without any addition of co-factors. **Salt fraction (from non-desalted cell-free ectracts) collected during desalting step and freeze-dried and then added to desalted protein extracts. ND, Not detected.

The HPLC chromatograms showing the effects of co-factor(s) on reductase activity in

the desalted E. coli O83:H1 NRG 857C cell-free   extracts   at   24   h   at   37˚C   under   anaerobic  

incubations are shown in Figure 2.28.

152

Figure 2.28 HPLC chromatograms showing the effects of co-factor(s) on reductase activity in E. coli O83:H1 NRG 857C cell-free extracts. (A) Desalted cell-free extracts with 0.25 mM glucoraphanin, 1 (5.33 min) without any factors. (B) Desalted cell-free extracts added with a solution of freeze-dried factors showed bioconversion from 1 to glucoerucin, 2 (13.8 min). (C) Desalted cell-free extracts added with 1 mM of each FAD, NADPH, and, NADH factors. The peak, 3 at 15.3 min indicates the presence of FAD without bioconversion. (D) Desalted cell-free extracts added with 1 mM MgCl2. (E) Desalted cell-free extracts added with 1 mM of both FAD and MgCl2. (F) Desalted cell-free extracts added with 1 mM of both NADPH and MgCl2. (G) Desalted cell-free extracts added with 1 mM of both NADH and MgCl2. The reactions were performed in   0.1   M   citrate   phosphate   buffer   pH   7.0   at   37˚C   under  anaerobic conditions for 24 h. The figures are representatives of triplicates.

The optimal temperature and pH of reductase activity in cell-free extracts in the

bioconversion of glucoraphanin to glucoerucin as detected by HPLC analysis were found at

37˚C  and pH 7.0, respectively (Figure 2.29A and 2.29B). It is interesting to note that previous

study on reduction of an uricosuric drug called sulphinpyrazone by human or rabbit faeces

153

gave greater reduction under anaerobic than under aerobic conditions (Strong et al., 1987). In

this work, however, both aerobic and anaerobic conditions resulted in similar rates with no

significant difference in the bioconversion of glucoraphanin to glucoerucin by reductase in E.

coli O83:H1 NRG 857C cell-free extracts over a time course (Figure 2.29C). Similarly, the

recent study showed that the presence of oxygen did not influence the bioconversion of

methylsulfinylalkyl GSLs in Escherichia coli Nissle 1917 and Enterobacter cloacae ATCC13047

(Mullaney et al., 2013).

Figure 2.29 Effects of temperature, pH and aeration on reductase activity in E. coli O83:H1 NRG 857C cell-free extracts (obtained from glucoraphanin-induced cells in the bioconversion (%) of glucoraphanin to glucoerucin. (A) Optimal temperature was found at  37˚C. (B) Optimal pH was found at pH 7.0. The reactions were performed in 0.1 M citrate phosphate buffer for 24 h under anaerobic conditions. (C) The effect of aerobic and anaerobic conditions upon reductase activity. Values are means of triplicates.

From these results, it was found that reductase enzyme in E. coli O83:H1 NRG 857C

cell-free extracts is inducible by GSLs, oxygen-independent, Mg2+- and NAD(P)H-dependent

for its sulfoxide reductivity towards methylsulfinylalkyl GSLs.

154

2.4 Summary of findings

The summary of key findings in this chapter is shown in Figure 2.30.

Figure 2.30 Summary of key findings in this chapter. (A) Bacterial fermentations in culture broths. (B) Bacterial resting cells in 0.1 M citrate phosphate buffer pH 7.0. (C) Bacterial reductase from cell-free extract in 0.1 M citrate phosphate buffer pH 7.0. LA, L. agilis R16; ECO, E. coli O83:H1 NRG 857C; EC, E. casseliflavus NCCP-53.

155

2.5 Discussion:

2.5.1 Bacterial GSL-degrading activity

This report showed GSL-degrading capacity from six strains of human gut bacteria that

have never been published before. Enterococcus sp. C213 and E. faecium KT4S13 are NIT

producers. However, clone SEW-E-011, E. casseliflavus NCCP-53, L. agilis R16 (obtained from

Palop et al. (1995), E. coli UMNF18 and E. coli O83:H1 NRG 857C are both ITC and NIT

producers. This suggests the presence of GSL-degrading activity or bacterial myrosinase-like

activity, similar to plant myrosinase, in these human gut bacteria. This is the first report to

show the time-course degradation product profiles from the metabolisms of different GSLs

including sinigrin, glucotropaeolin, gluconasturtiin, glucoerucin, glucoiberin, glucoraphanin,

glucobrassicin and certain corresponding DS-GSLs by the two chosen bacteria Enterococcus

casseliflavus NCCP-53, E. coli O83:H1 NRG 857C and a bacterium L. agilis R16. A general trend

of ITC and NIT production over a time course was shown for the first time upon different GSL

metabolism in the three bacteria. For most GSLs, once ITC production peaked during bacterial

fermentations, it gradually declined while NIT production gradually increased and remained

fairly constant or slightly declined for certain GSLs in certain bacteria. The stability test of

several authentic ITC/NIT standards in culture broths with and without the presence of

bacterial cells and in different aqueous solutions have been carried out for the first time. The

results showed that all ITC levels decline over a time course in all solutions whereas NIT levels

stayed fairly constant except ANIT.

This is the first finding of ITC products detected upon the metabolism of glucoraphanin

and glucoiberin in bacterial fermentations. E. coli O83:H1 NRG 857C produced erucin and

erucin NIT products from glucoraphanin, and iberverin and iberverin NIT from glucoiberin. It is

likely that this bacterium cannot metabolize these GSLs directly. However, E. casseliflavus

NCCP-53 produced sulforaphane and iberin from glucoraphanin and glucoiberin, respectively

without NIT production. Interestingly, both bacteria were found to metabolize glucoerucin

more readily and produced the corresponding erucin and erucin NIT products at higher levels

than when glucoraphanin and glucoiberin were used as substrates. This result is different

from the previous study showing only NIT products were produced from the metabolisms of

glucoraphanin and glucoiberin in Lactobacillus plantarum KW30, Lactococcus lactis subsp.

lactis KF147, Escherichia coli Nissle 1917 and Enterobacter cloacae (Mullaney et al., 2013). The

different hydrolysis products obtained from the same GSLs in different bacterial strains

156

underscore the diversity of gut bacterial GSL-degrading enzymes that may partly explain the

inter-individual variation in GSL metabolism and bioavailability observed in epidemiological

studies (Ambrosone et al., 2004; Higdon et al., 2007; Seow et al., 2005; Steck et al., 2007).

Since there is very few evidence in the metabolism of DS-GSLs in bacterial

fermentation, the time-course degradation product profiles from the metabolisms of

different DS-GSLs by the three bacteria were performed for the first time. The results showed

that pure NITs were produced from the metabolisms of most DS-GSLs during bacterial

fermentations. This supports the speculation that DS-GSL is a pre-cursor to NIT production

(Wathelet et al., 2001). This finding is in accordance with previous work where DS-GSL, the

intermediate, hypothesized by Smits et al. (1993) explains why ANIT formation continues

after total disappearance of GSL in their experiments. Thus, it is probable that these bacteria

may exhibit sulfatase activity that desulfates intact GSLs to produce DS-GSLs and hence NIT

production, in addition to ITC production, from the metabolism of intact GSLs. Interestingly,

this is the first finding to show that bacterial resting cells produced only ITCs (without any

NITs) from the metabolism of GSLs in citrate phosphate buffer pH 7.0. Also, NIT production

only occurred in the culture media during bacterial fermentations. It was found that Fe2+ ions

(present in the culture broths used) are required for NIT productions from bacterial

metabolisms of GSLs in citrate phosphate buffer pH 7.0.

Our results confirm that human gut microbiota is diverse in its capacity to metabolize

GSLs to different products. As previously reported, certain human gut bacteria such as B.

pseudocatenulatum, B. adolescentis and B. longum produced NITs predominantly from

sinigrin and glucotropaeolin metabolism (Cheng et al., 2004) while some bacteria such as L.

agilis R16 and B. thetaiotaomicron produced ITCs predominantly from sinigrin metabolism

(Elfoul et al., 2001). There may be different enzymes/mechanisms involved in the metabolism

of the same or different GSLs by different bacteria. In addition, the structure of GSL substrate

e.g. the size and the high polarity of sulfoxide group of glucoraphanin and glucoiberin may

render β-thioglucosidic bonds of these GSLs inaccessible to bacterial GSL-degrading enzymes,

and thus influencing the metabolic capacity of each bacterium. That may be the reason why L.

agilis R16 cannot metabolize glucoraphanin or and glucoiberin to ITC/NIT products. Our result

supports the previous reports suggesting that individual GSLs are differently affected by

157

digestive enzymes in vitro (Maskell & Smithard, 1994). Clearly GSL metabolism in the human

gut is a much more complex process than previously thought.

It was found in this work that ITCs are unstable in aqueous solutions. This effect was

more pronounced when bacterial cells were present in the medium. In contrast, the levels of

most NITs except for ANIT in broths remained fairly constant without bacterial cells with a

slight decline in the presence of bacterial cells. The fact that NITs are less labile than ITCs may

be a rationale for rather constant NIT production from all GSL metabolisms during bacterial

fermentations. Also, both ITC and NIT production occurred in the pH range of 3.7-7.6. This

contradicts with the previous thought that NIT production in culture broths was due to low pH

conditions (Cheng et al., 2004; Mullaney et al., 2013).

Thus far, our results are in agreement with the previous findings from previous studies

reporting that ITCs are unstable in aqueous solutions. The electrophilic character of the

functional isothiocyanic group has enabled ITCs to react with some nucleophilic agents

including amino, hydroxyl, thiol, carboxylic acids from small peptides, amino acids and water

(Zhang et al., 1996; Cejpek et al., 1998) and probably flavonols to potentially generate new

compounds (Cejpek et al., 2000; Kawakishi & Kaneko, 1987; Luciano et al., 2008). The

reactions of AITC with alanine, glycine, and several peptides in model systems have been

described (Cejpek et al., 2000). In the previous report, most sulforaphane was lost at 24 h

incubation with bacteria, and almost 50% loss was also found without bacteria (Basten et al.,

2002; Lai et al., 2010). This was also the case with AITC (Combourieu et al., 2001; Ye et al.,

2002; Zhang, 2004). It has been proposed that the composition of the media or the buffer

may react with sulforaphane (or any ITCs), and thus depleted ITCs in the culture broths

(during bacterial fermentations) and the buffer used in this work. Our results indicate that the

ITC instability in the media/buffer is currently underestimated. Thus, the total percentage

products obtained from GSL metabolism represented in this work and previous works may be

underestimated. However, it still remains unclear whether ITC/NIT products generated during

bacterial fermentations behave in the same way as those purchased ITC/NIT standards did in

the stability test. AITC has been reported to be unstable and is gradually decomposed to other

compounds having a garliclike odor in the presence of water at both room temperature and

37°C (Kawakishi & Namiki, 1969). AITC is also sensitive to temperature and pH. The high

temperature of 37°C and alkaline conditions accelerate the decomposition of AITC (Ina et al.,

158

1981). The decomposition rate of AITC in a low concentration range (below 1.6 mM) obeys

the first-order rate equation, and its degradation kinetics can be explained by the nucleophilic

attack of hydroxide ions and water molecules on the AITC molecule (Ohta et al., 1995). In

aqueous solution, pH and especially temperature have an influence on the decomposition of

AITC (Ohta et al., 1995). In addition, Combourieu et al., (2001) also tested the stability of AITC

and BITC incubated in the buffer in the absence of cells for 48 h at 37°C under the same

conditions as those employed with bacterial incubations. Both allylamine and benzylamine

were detected by 1D 1H NMR spectroscopy indicating the high sensitivity of these ITCs to

hydrolysis (Combourieu et al., 2001). In the previous report, after 17 h incubation of sinigrin

with L. agilis R16, a 45% decline in AITC concentration was observed as well as the control

incubation of AITC in broth without bacteria (Palop et al., 1995). This indicates a spontaneous

chemical transformation of AITC, and the chemical nature of any ITC conversion product

therefore remains to be investigated. Other ITCs have also been reported for their instability

in aqueous solutions. For examples, PITC is unstable in aqueous media and rapidly degraded

to phenethylamine at low pH (Negrusz et al., 1998). In our study, the total percentage

products from the metabolisms of all GSLs by all three bacteria never reached 100% in spite of

high degradation rates up to 100% of most GSLs over 24 h. This suggests that there might be

other metabolite products that were not detected by current methods or ITCs may be further

degraded by bacteria (Kliebenstein et al., 2001; Palop et al., 1995; Tang et al., 1972).

Therefore, in theory, higher GSL degradation rate by bacteria does not necessarily always

translate into a higher ITC yield and higher ITC exposure for the host.

Although, conversion of GSLs to ITCs by plant myrosinase (S. alba) was shown to have

a yield of up to 90% (Kawakishi & Muramatsu, 1966; Piotrowski, 2004). Other previous studies

showed that the percentage products of most bacterial GSL metabolisms were far from 100%

(Cheng et al., 2004; Combourieu et al., 2001; Giamoustaris & Mithen, 1996; Hall et al., 2001;

Krul et al., 2002; Mithen et al., 1995; Palop et al., 1995; Parkin et al., 1994). This suggests that

the efficiency of bacterial GSL-degrading enzyme is lower compared to the action of plant

myrosinase (Krul et al., 2002) or metabolites other than ITCs may be formed during the

metabolism. By using 1D and 2D 1H NMR spectroscopy, Combourieu et al. (2001) have clearly

shown that sinigrin and glucotropaeolin are converted by the human fecal microbiota into

allylamine and benzylamine, respectively. This finding is in contradiction with most studies

159

showing that human gut bacteria produced ITCs from GSLs (Shapiro et al., 1998; Getahun and

Chung, 1999; Elfoul et al., 2001), and low amounts of ITCs were recovered accounting for 10

to 20% of the initial amounts of GSLs. From our results, amines or alcohols as derivatives of

ITC/NIT products were not detected on GC-MS chromatograms suggesting no such

degradation of ITC/NIT products occurred in the three bacteria tested. However, the

unaccounted for metabolite products, if any, should be further investigated.

This work demonstrated that Fe2+ ions are required for NIT productions from GSL

metabolisms by bacterial resting cells in the buffer. Similarly, plant myrosinases also requires

Fe2+ ions as co-factors for NIT production from GSL metabolism in vitro (Sørensen, 1990;

Burow et al., 2006; de Torres Zabala et al., 2005). However, GSL degradation to NIT products

can also occur via a non-enzymatic mechanism catalyzed by Fe2+ (Bellostas et al., 2007;

Bellostas et al., 2009). From our results, NITs were produced only in trace amounts from

intact GSLs in the buffer containing 5 mM Fe2+ ions without bacterial cells suggesting a non-

enzymatic NIT production caused by Fe2+ ions. It is likely that myrosinase-catalyzed NIT

production by gut bacteria was also promoted by Fe2+ ions present in culture broths. An

increased NIT production in the presence of Fe2+ has been previously reported both in the

presence (Agerbirk et al., 1998; de Torres Zabala et al., 2005) and the absence of myrosinase

(Bellostas et al., 2008; Austin et al., 1968). Similar to our results, inhibition of ITC production

in the presence of myrosinase and Fe2+ has been observed as an outcome of enhanced NIT

production (Agerbirk et al., 1998; de Torres Zabala et al., 2005; Bellostas et al., 2009). It has

been hypothesized that NIT production by Fe2+ in the absence of myrosinase involves the

formation of a GSL–Fe2+ complex, which leads to the formation of the NIT (Bellostas et al.,

2008). NIT formation from GSLs through myrosinase-mediated degradation has been known

to be influenced by the low pH of the media (Gil & MacLeod, 1980; VanEtten et al., 1966)

and/or the presence of Fe2+ ion (Austin et al., 1968; Tookey & Wolff, 1970; Uda et al., 1986).

The previous study showed that ANIT production from sinigrin occurred above pH 6.30 at 80

μM of Fe2+ ion in the incubation mixtures (Lu et al., 2011). In comparison with the reported

data (Austin et al., 1968; Bellostas et al., 2007; Uda et al., 1986), this concentration was

significantly lower to promote ANIT formation from sinigrin the digestive incubation mixture.

This suggests that the human gut bacteria may utilize another mechanism for NIT formation

instead of the promotive effect of Fe2+ ion and/or a lower medium pH (Lu et al., 2011).

160

Whether human gut bacteria utilizes the ESP-involving route for the formation of NIT

production from DS-GSL remains to be determined by further investigations (Lu et al., 2011).

The bacterial GSL-degrading activity (a.k.a. myrosinase-like activity) of the three

bacteria responsible for producing ITCs and NITs were found to be inducible by GSL substrate

in resting cells experiments. This finding supports the speculation that bacterial GSL-

degrading activity is inducible by GSLs. For examples, bacterial myrosinases of Enterobacter

cloacae and L. agilis R16 were produced only in the presence of the inducer such as sinigrin

and mustard extract, respectively (Tani et al. 1974, Palop et al., 1995). From the cell-free

extracts experiment, glucoiberin/glucoraphanin was directly converted to

glucoiberverin/glucoerucin by E. coli O83:H1 NRG 857C cells without being further

metabolized suggesting that there was no GSL-degrading activity in cell-free extracts. It is

speculated that GSL-degrading activity may be in the membrane/debris fraction or is an

insoluble enzyme or part of a protein complex. Thus far, GSL-degrading activity, as indicated

by ITC/NIT production from GSL metabolism, was only detected in bacterial intact cells during

fermentation experiments, not in cell-free extracts. This is in accordance with the previous

report that sinigrin-degrading activity of L. agilis R16 was always associated with the presence

of intact cells (Palop et al., 1995; Prescott & John, 1996) indicating that intracellular or cell-

associated activity is always involved. Also, fungal myrosinase of Aspergilus niger was

reported to be remarkably unstable and could not be protected by stabilizing agents

(Kliebenstein et al., 2001; Ohtsuru & Hata, 1973). The instability of bacterial GSL-degrading

enzyme may be the issue with the bacteria studied in this work. To date, the only bacteria

that showed myrosinase activity in cell-free extracts include Gram-positive Bifidobacterium

(Cheng et al., 2004), Gram-negative Enterobacter cloacae (Tani et al. 1974) and Gram-

negative Citrobacter (Abdulhadi Albaser, PhD thesis). Interestingly, for the case of Citrobacter,

ITC production from the metabolism of GSLs in broths and citrate phosphate buffers was

barely detected by GC-MS analysis, but a glucose release from GSL degradation to glucose

was detected by GOD-PERID assays (personal communications). A similar result was also

obtained from Bifidobacterium (Cheng et al., 2004) where ITCs were not detected from the

metabolism of GSLs. These findings are different from our results obtained from bacterial

fermentations showing both ITC and NIT productions from most GSL metabolisms. This

161

suggests the diversity in enzymes/mechanisms involved in the metabolism of the same or

different GSLs by different bacteria.

2.5.2 Bacterial reductase activity

The bacterial reductase involved in the GSL metabolism was partially characterized for

the first time in this work. The reductase from E. coli O83:H1 NRG 857C cell-free extract is

inducible by GSLs, oxygen-independent and requires Mg2+ ion and NADP(H) as co-factors for

its  activity  with  optimum  pH  and  temperature  at  pH  7.0  and  37˚C,  respectively.  This reductase

was shown to be able to biotransform methylsulfinylalkyl GSLs (sulfoxide) to methylthioalkyl

GSLs (sulfide) and also biotransform methylsulfinylalkyl ITC to methylthioalkyl ITC. This result

supports the previous finding of the bioconversion of methylsulfinylalkyl GSLs i.e. glucoiberin

and glucoraphanin by a reduction reaction in human gut bacteria (Lai et al., 2010; Mullaney et

al., 2013; Saha et al., 2012).

It was speculated that the sulfoxide groups on glucoiberin and glucoraphanin may

present steric effects that prevent bacterial GSL-degrading enzyme to gain access to the β-

thioglucosidic bonds of these GSLs or prevent the transportation of GSLs into bacterial

cytoplasm in Gram-positive bacteria or periplasm in Gram-negative bacteria. The transport of

GSL and GSL-degrading enzyme activity may be tightly linked. E. coli O83:H1 NRG 857C seems

to overcome this problem by expressing a reductase enzyme to bioconvert the sulfoxide

groups of methylsulfinylalkyl GSLs i.e. glucoiberin/glucoraphanin to the sulfide groups of

methylthioalkyl GSLs i.e. glucoiberverin/glucoerucin before the latter being metabolized by

bacterial GSL-degrading enzyme to corresponding ITC and NIT products. It was thought that E.

coli O83:H1 NRG 857C cannot metabolize methylsulfinylalkyl GSLs directly as no

corresponding oxidized product series were detected. Due to the bacterial reductase, higher

levels of products were obtained from glucoiberin/glucoraphanin metabolism in E. coli

O83:H1 NRG 857C than those obtained in E. casseliflavus NCCP-53. From our results, the

interconversion of iberverin and iberin products are expected to be in a manner analogous to

that of sulforaphane and erucin as previously reported (Bheemreddy & Jeffery, 2007;

Kassahun et al., 1997; Lai et al., 2010; Saha et al., 2012). Since bioconversion of GSL can be

achieved by human gut bacteria, one could assume that the consumption of Brassica

vegetables rich in glucoerucin (e.g. rocket salad) or glucoiberverin may give rise to the same

162

active components in vivo as glucoraphanin-containing broccoli species. Although this

reduction bioconversion may be part of bacterial oxidoreduction process that can be

reversible, there was no proof of a reversible oxidation reaction as there was no conversion to

sulforaphane from erucin production during the metabolism of glucoerucin in E. coli O83:H1

NRG 857C fermentation. Also, the amounts of the reduced methylthioalkyl GSL always

increased over a time course and above the oxidized methylsulfinylalkyl species in E. coli

O83:H1 NRG 857C cell-free extracts.

It has been long known that under the anaerobic conditions in the human gut, the

principal reaction of sulfoxides is a reduction to the corresponding sulfides. Sulfoxides that

have been tested extensively both in vivo and in vitro are the xenobiotics sulphinpyrazone and

sulindac. Sulphinpyrazone is a uricosuric drug that is metabolized to a sulfide analogue with a

platelet anti-aggregatory activity, ten times more active than its sulfoxide counterpart (Del

Maschio et al., 1984). Studies in the rat (Renwick et al., 1982; Kashiyama et al., 1994), rabbit

(Strong et al., 1984b) and in humans (Strong et al., 1984a) have shown that intestinal

microflora are exclusively responsible for reduction bioconversion of sulphinpyrazone to the

active sulfide metabolites. Sulindac is a sulfoxide prodrug that requires reduction to the active

anti-inflammatory sulfide analogue. The human gut bacteria were shown to reduce sulindac in

vitro (Galletti et al., 2001; Strong et al., 1987). Studies with over 200 isolated strains of human

bacteria showed significant sulfoxide reduction by several facultatively anaerobic bacteria,

such as E. coli, Klebsiella oxytoca and KIebsiella pneumoniue under anaerobic conditions

(Strong et al., 1987). Interestingly, it was reported that the sulfoxide R(+)-flosequinan was

reduced by these bacteria to the sulfide. This stereoselectively demonstrates that chiral

inversion at the sulfoxide position of flosequinan enantiomers via stereoselective reduction of

sulfoxide was enabled by human gut bacteria (Eiji et al., 1994; Kashiyama et al., 1994).

Sulfoxides were also found to be reduced via aldehyde dehydrogenase or a hepatic

thioredoxin-dependent system in the presence of an electron donor (Lee & Renwick, 1995). At

least three different soluble enzymes in E. coli capable of reducing sulindac are present. One

of which appeared to be a NADPH-linked thioredoxin system; only one of the enzymes was

capable of reducing the more hindered sulfoxide groups in sulphinpyrazone or flosequinan

(Botti et al., 1995; Lee & Renwick, 1995). From the same study, NADH, NADPH and

dithiothreitol (DTT) were shown to act as co-factors for sulfoxide reductase activity in E. coli

cytosolic fraction that had reduction activity with sulindac, diphenyl sulfoxide and

163

sulphinpyrazone. There is a possibility that multiple sulfoxide reductase systems are present

in E. coli cytosol (Chen & Halkier, 1999; Lee & Renwick, 1995). Similar to the above report, our

results showed that cytosolic E. coli O83:H1 NRG 857C reductase is oxygen-independent and

NAD(P)H- and Mg2+ -dependent for its sulfoxide reduction towards methylsulfinylalkyl GSLs.

The search for the corresponding gene or protein responsible for sulfoxide reduction in E. coli

O83:H1 NRG 857C is underway in our group.

Interestingly, 6-phospho-β-glucosidase (NCBI Ref: YP_006120095.1) of E. coli O83:H1

NRG 857C is the only β-glucosidase (out of six) with the predicted gene ontology in

oxidoreduction activity (Consortium, 2012) and contains sugar binding site, NAD binding site,

and divalent metal binding site. This protein is a member of glycoside hydrolase family 4

(GH4) with the unique requirement for NAD(H) and a divalent metal for oxidoreduction

activity (Henrissat et al., 1995; Thompson et al., 1998; Thompson et al., 1999). It remains to

be determined whether this 6-phospho-β-glucosidase is capable of reducing the sulfoxide on

methylsulfinylalkyl GSL or ITC/NIT series. However, the recent report proposed that the

enzyme methionine sulfoxide reductase A (MsrA) EC 1.8.4.11 (ExPASy) may be responsible for

the reduction bioconversion (Mullaney et al., 2013). Previously, the membrane-associated

and soluble peptide methionine sulfoxide reductases in E. coli were found to require NADPH

for its reduction activity (Lori et al., 1999; Spector et al., 2003). Also, sulfoxide reductase

catalytic subunit YedY of E. coli O127:H6 strain E2348/69 was able to catalyze the reduction of

a variety of substrates e.g. trimethylamine N-oxide, dimethyl sulfoxide, L-methionine

sulfoxide and phenylmethyl sulfoxide with the requirement of molybdenum as a cofactor

(Iguchi et al., 2008). In E. coli O83:H1 NRG 857C, ten sulfoxide reductases remain to be tested

for reduction activity upon methylsulfinylalkyl GSLs.

Thus far, our findings from this work underline the significant influence of human gut

bacteria on the metabolic fate of GSLs. ITC products generated from these metabolisms may

promote chemoprevention in human health. In addition, bacterial reductase showing the

reduction bioconversion of methylsulfinylalkyl GSLs to methylthioalkyl GSLs provides a link to

the sulfoxide reduction of xenobiotics in humans. It remains to be determined whether this

reductase can also act upon xenobiotics.

164

Chapter 3: Forward proteomics approach to identify bacterial proteins potentially involved in the metabolism of GSLs 3.1 Introduction

In Chapter 2 (section 2.3.6), resting cells experiment demonstrated that the putative

bacterial GSL-degrading enzyme activity from all three bacteria tested is likely to be inducible

by GSL supplementation as the resting cells from GSL-induced overnight cultures yielded

higher GSL degradation and corresponding ITC products in the buffer as opposed to those

cells from overnight cultures without GSL supplementation that produced no ITC product with

less GSL degradation. It is therefore hypothesized that GSL supplementation is likely to induce

GSL-degrading enzyme and possibly other enzymes involved in bacterial GSL metabolism.

Thus, forward proteomics approach using 2-DE technique was performed to compare the

protein patterns on 2-DE gels obtained from protein extracts isolated from the cells with and

without GSL supplementation. The distinctively expressed or upregulated proteins found in

bacterial cultures grown on media upon GSL supplementation are expected to be involved in

bacterial GSL metabolism.

3.1.1 Forward proteomics

The   term   ‘proteome’   coined   by   Marc   Wilkins   in   1994   describes   the   entire   set   of  

proteins encoded by the genome of an organism, cell or tissue at a given time (Wilkins et al.,

1996).  Therefore,  the  term  ‘proteomics’   is   the  study  of  proteomes  which  may  entail  several  

different aspects related to the study of proteins at a global or cellular level. This includes the

analysis of protein expression patterns, biological function, macromolecular protein structure,

spatiotemporal intracellular distribution, stability and turnover rates, posttranslational

modification (PTM) as well as protein–protein interactions. Thus far, proteomics has been

mainly considered as an approach allowing for either protein identification from complex

mixtures of proteins or the characterization of changes at the level of their expression/PTM

levels.  This  has  been  termed  ‘forward  proteomics’  which  primarily  depends  on  the  power  of  

sample preparation technologies, mass spectrometry (MS) analysis and bioinformatics to

165

achieve satisfactory goals (Cristoni & Bernardi, 2004).

Classical forward proteomics work involves two dominant strategies; a gel-based

approach and a gel-free based approach (Figure 3.1). The gel-based approach involves a

separation step, usually by a means of two-dimensional gel electrophoresis (2-DE) followed by

imaging analysis, spot excision, protein digestion and protein identification using mass

spectrometry (MS) (Figure 3.1) (Andersen & Mann, 2000). Proteins resolved by 2-DE can be

identified by in-gel trypsin digestion via peptide mass fingerprinting (PMF) using MS or

tandem mass spectrometry (MS/MS) (Link et al., 1999). The gel-free based approach is based

on the use of stable isotope tagging and liquid chromatography (LC-MS) (Aebersold & Mann,

2003). The perturbed and non-perturbed protein extracts are differentially labeled with

different stable isotopes (12C/13C, 14N/15N and 1H/2H). This enables the same peptide from

two different samples to exhibit the same chemical behavior, with a difference in mass

detectable by MS techniques. Peptide peak intensities can be used for relative quantification

of these peptides. The steps of this approach are as follows: (1) differential isotopic labelling;

(2) digestion of combined protein samples to obtain peptide mixtures; (3) chromatographic

fractionation of mixed peptide samples; (4) analysis of the separated peptides by MS/MS; and

(5) processing of the MS results to obtain relative protein abundance as well as protein

identification by database searching (Figure 3.1). This strategy is also known as shotgun

proteomics.

166

Figure 3.1 Strategies for forward proteomics. Gel-based proteomics (on the left) mainly involves the use of 2-DE, protein digestion and MS. LC-MS driven proteomics or gel-free based proteomics (on the right) involves the use of stable isotope tagging and liquid chromatography (LC-MS). This figure was taken from Roepstorff (2012).

In  this  work,  a  preferred  strategy  called  ‘GeLC-MS’  was  used  to  separate the proteins

by 2-DE technique, the gel spots were sliced and digested followed by LC-MS/MS analysis

(Figure 3.2).

167

Figure 3.2 Combined gel-LC-MS based strategy (GeLC-MS). This figure was taken from adapted from Roepstorff (2012).

3.1.2 Two-dimensional electrophoresis (2-DE)

Two-dimensional electrophoresis (2-DE) was developed almost 40 years ago (Klose,

1975; O' Farrell, 1975). Two different separation methods are combined in 2-DE: isoelectric

focusing   (IEF)   and   ‘conventional’   sodium-dodecyl sulfate polyacrylamide gel electrophoresis

(SDS-PAGE) in order to separate proteins on the basis of pI and molecular weight, respectively.

Therefore, individual protein components can be spatially resolved and subsequently be

visualized  as  ‘spots’  on  the  gels.

Any single-dimension method cannot resolve more than 80–100 different protein

components. However, it was shown that 1,100 different proteins from lysed E. coli cells were

resolved on a single 2-DE map (O' Farrell, 1975). Theoretically, 2-DE is capable of resolving up

to 10,000 proteins simultaneously (Klose, 1975), with approximately 2,000 proteins being

routine, and detecting and quantifying protein amounts of less than 1 ng per spot (Lopez,

2007).

168

The limitations of the 2-DE approach are well known, i.e. poor solubility of membrane

proteins, limited dynamic range and difficulties in displaying and identifying low-abundance

proteins. However, 2-DE still remains as one of the major separation techniques for the next

years because its resolution and the advantage of storing the isolated proteins in the gel

matrix until further analysis is unrivalled by any other alternative technique. To date, there

are more than 5,000 publications reported using the 2-DE technique for analyzing protein

patterns in a plethora of biological systems (Figure 3.3). The increasing trend was observed

from 1996 to 2011.

Figure 3.3 List of publications in proteomic field by means of two-dimensional electrophoresis technology as of Dec 2011. The list is generated using PubMed   search   with   keywords   ‘two-dimensional  electrophoresis,’ ‘2-DE,’   ‘2-D  electrophoresis,’   ‘2-D  gel   electrophoresis,’ ‘2D-PAGE,’   ‘2D-DIGE’  or   ‘DIGE’  and  each  of   ‘proteome,’   ‘proteomic’ or  ‘proteomics.’  A  growing  trend   is  observed   in  the late 1990s to early 2000s. The number of publications is levelled in recent years, suggesting an importance of gel-based approaches in proteomics is still evident at present.

3.1.3 Work-flow of gel-based strategy

3.1.3.1 Protein preparation

When cultured cells are used, highly consistent conditions are required. As for any

experimental approach, sample preparation is the most critical part of proteomics

experiments. This step involves tissue/cell homogenization and/or lysis.

0

100

200

300

400

500

600

700

800

900

Num

ber o

f pub

licat

ions

Year

169

In general, proteins can be found in cytoplasm, inner membrane, outer membrane,

periplasm (in the case of Gram-negative bacteria) or extracellularly (secreted). Various

procedures can be used to obtain specific cell fractions, enabling their protein content to be

determined (Figure 3.4). For Gram-negative bacteria, the periplasmic extract can be isolated

by cell spheroblasting, followed by differential centrifugation, or by a cold osmotic shock

method. After separation of periplasmic proteins, cell lysis can be used to isolate the

cytoplasmic proteins, with the membrane fraction separated by centrifugation. The inner and

outer membranes of Gram-negative bacteria can be isolated by selective cell lysis using

lysozyme or ethylenediaminetetraacetic acid (EDTA) treatment, by mechanical methods such

as the use of a French Press, or by using commercially available chemical lysis reagents

(mostly detergents) (Thein et al., 2010). After lysis, distinct inner and outer membrane

vesicles are formed, which can be separated from each other by density centrifugation (with

the inner membrane vesicles having a higher density). Selective detergent treatment, such as

with Triton X-100, which preferentially dissolves cytoplasmic membranes, can also be used to

separate inner and outer membrane protein fractions. In Gram-positive bacteria, cell surface

proteins have been successfully isolated for analysis. The methodology that is known as

surface shaving liberates surface exposed protein domains by digesting the cell surface with

enzymes such as trypsin or proteinase K. Choosing the appropriate fractionation method is

essential to the success of subcellular proteomics as cross-contamination of the various

fractions will seriously affect the results.

170

Figure 3.4 Sub-cellular fractionation of Gram-negative bacterial cell culture. Centrifugation was utilized to separate intact bacterial cells from culture medium and extracellular proteins. Protein samples in the supernatant constitute the extracellular fraction. Periplasm can be isolated by cell spheroblasting or cold osmotic shock treatment followed by ultracentrifugation. The cytosolic frac- tion and membrane fraction can be isolated separately via various cell lysis procedures, detergent treatment and ultracentrifugation. This figure was taken from Curreem et al., (2012).

3.1.3.2 Protein separation

The 2-DE technique combines two dimensions of physical protein separation by their

chemical properties (Klose, 1975; O' Farrell, 1975). Firstly, the proteins are arranged according

to their content of basic or acidic amino acids on a linear gel with an immobilized pH-gradient

(Bjellqvist et al., 1982; Klose, 1975), and are separated by their isoelectric point (pI). Secondly,

pI-separated proteins are separated by their molecular size similar to a conventional SDS-

PAGE (Klose, 1975; Laemmli, 1970). These two combined techniques gives rise to a high-

resolution separation of single protein types on a two-dimensional arrangement of proteins.

The separated proteins can be visualized on the gel by different staining methods, for

171

example, Coomassie brilliant blue (Meyer & Lamberts, 1965), silver (Rabilloud et al., 1988), or

fluorescent staining (Berggren et al., 1999). Each spot on a 2-DE gel represents one protein

species, and a specific pattern of the spots on the gel corresponds to each cell or a cellular

condition. Thus, changes of the cellular proteome such as its gene activity and metabolism

under healthy or diseased conditions can be determined by changes in this spot pattern

(Klose, 1975).

3.1.3.3 Gel analysis, spot detection and quantification

Following the previous step, the gel images were captured by using laser imaging

devices, scanners, or charge-coupled device (CCD) camera-based (Berth et al., 2007). The

obtained digitized computer images of the gel can be displayed with common image analysis

software. The quantitative information of the gel is transformed into computer-readable data

via image capture step. Next, most 2-DE programs follow these steps for the evaluation of 2-

DE gel images: spot detection, spot filtering, spot editing, background correction, gel

matching, normalisation, comparison, quantification and reporting and exporting of data. One

of the crucial steps is normalisation. Before the gels are compared for differences in the spot

pattern, the spot volumes of the different gels have to be adjusted by normalisation. This step

corrects for different protein loads and staining effectiveness. Gels are normalised according

to the total spot volume or the volume of a single prominent reference protein (Berth et al.,

2007).

3.1.3.4 Spot excision and digestion

After gel analysis, the next procedure involves several steps: (1) excision of the spot of

interest; (2) treatment with an appropriate protease (e.g. trypsin); (3) extraction (purification)

of the peptide fragments produced; followed by (4) mass spectrometric analysis.

3.1.3.5 Protein identification by mass spectrometry

There are two main approaches for the identification of proteins using MS-based

techniques: (i) peptide mass fingerprinting and (ii) peptide sequencing.

(i) Peptide mass fingerprinting (PMF) is typically performed using matrix assisted laser

desorption/ionization (MALDI) for ionization of the peptides and a time-of-flight (TOF) mass

172

analyzer to characterize peptides (Figure 3.5). Identification of proteins using PMF takes

advantage of the high mass accuracy of the method (Yates et al., 2009). Spots excised from

the 2-DE  gel  are  digested  with  trypsin,  generating  a   ‘fingerprint’  of  tryptic  peptides  that  can  

serve as a signature for a protein. Protein identification is made by comparing the

experimentally obtained molecular weights of the peptides with theoretically calculated ones

from proteins in a database (or of those derived from theoretical translation of DNA sequence

in the database).

Figure 3.5 MALDI-TOF mass spectrometry. Sample is co-deposited with matrix on a stainless steel plate and advanced to the mass spectrometer by ionization with a laser. The TOF analyzer separates ions based on the time it takes them to travel the length of the flight tube. This figure was taken from Delahunty & Yates (2006).

(ii) Peptide sequencing is achieved by tandem mass spectrometry (MS/MS) using

electrospray ionization (ESI) and either a quadrupole or ion trap mass spectrometer (Figure

3.6). Like PMF, proteins separated on a 2-DE gel are excised and digested with trypsin prior to

analysis. Unlike MALDI, ESI is performed at atmospheric pressure and can be linked with liquid

chromatographic (LC) methods for the introduction of the sample into the mass spectrometer

(LC-MS/MS). The additional chromatographic separation helps to resolve co-eluting proteins

and increase the number of identifications that can be made for a protein mixture. The

peptide fragmentation data collected from MS/MS are specific for each peptide because they

yield the actual amino acid sequence of the peptide as well as its molecular weight. The

increased specificity makes MS/MS capable of identifying multiple proteins from a gel spot of

173

protein mixtures. Amino acid sequences of the peptides are obtained by de novo

interpretation of the spectra. These sequences are then BLAST searched against the protein

database to identify the protein of origin.

Figure 3.6 Tandem mass spectrometry (MS/MS). (A) Electrospray ionization (ESI) ionizes the analytes out of a solution and is therefore readily coupled to liquid-based. This figure was taken from Lane (2005). (B) Tandem mass spectrometry with two mass analyzers separated by an ion activation device. A mixture of peptides is ionized by ESI process and then separated in the first segment by their mass to charge (m/z) ratios. Selected ions are then advanced to the ion activation device where they are fragmented. This figure was taken from Delahunty & Yates (2006).

3.1.4 Applications of 2-DE in bacterial proteomics

The use of proteomics, especially 2-DE,  is  of  central  importance  to  study  a  ‘proteome  

map’  of  a  large  proportion  of  the  entire protein complements of the cell. In combination with

MS, and the plethora of bacterial whole-genome sequence data currently available, it is often

possible to identify almost every single protein spot on the 2-DE gel.

A better understanding of bacterial cellular physiology can be achieved by profiling the

174

proteome of bacterial grown under certain conditions. Since protein expression is an

energetically costly process, only the specific proteins directly required for growth, survival or

pathogenicity are expressed under a particular physiological condition. Consequently,

cataloging all the proteins expressed under a specific set of conditions (including the

identification of postranslational modifications) can lead to identification of the components

responsible for maintaining certain cellular activities. A bacterium, such as E. coli, which has

approximately 4,300 genes, typically expresses 1,000–1,500 proteins under standard

laboratory growth conditions. This is ideally suited to 2-DE, in which it is possible to

simultaneously visualize and quantify a comparable number of protein spots on a single gel

(Kalia & Gupta, 2005). Using 2-DE expression profiling under various culture conditions, it is

often possible to identify sets of proteins involved in basic cellular metabolism versus those

expressed  in  response  to  various  cellular  stresses  or  external  stimuli.  The  majority  of  ‘house-

keeping’   proteins   functioning   in   central   metabolic   pathways   are   expressed   at   relatively  

constant levels during the active growth stage however many proteins involved in stress

response  or  adaptation  are  significantly  upregulated  from  a  lower  ‘basal’   level  under  certain  

growth conditions. To investigate the putative physiological roles or biological functions of

proteins, various fractionation approaches may be used to establish their predominant

cellular localizations. In such approaches, whole cell proteomes may be compared with

various   ‘sub-proteomes,’   which   may   include   extracellular   proteins,   membrane   or   cell   wall  

associated proteins, DNA/RNA associated proteins, or cytoplasmic fractions.

3.1.5 Hypotheses

Although several groups have identified bacterial strains associated with GSL

degradation capacity, very little is known about bacterial proteins involved in the metabolism

of GSLs. In Chapter 2, putative bacterial GSL-degrading enzyme activity is shown to be likely

inducible by GSL. Therefore, it may be possible to identify these inducible proteins expressed

during GSL supplementation in bacterial cultures. The hypotheses of this chapter are as

follows:

GSL supplementation to bacterial cultures may produce different proteome maps on

2-DE gels when compared to those obtained from bacterial cultures without GSL

supplementation.

175

Distinctively expressed or upregulated proteins found on 2-DE gels ontained from

bacterial cultures grown on media with sinigrin supplementation may be involved in

bacterial metabolism of GSL.

3.1.6 Objectives

The objective of this chapter is as follows:

To identify distinctively expressed or upregulated proteins in bacterial cultures grown

on media with sinigrin supplementation by comparatively analyzing 2-DE proteome

maps of bacterial cells isolated from cultures with- versus without- GSL

supplemetation and performing subsequent MS/MS analysis for protein identification.

176

3.2 Materials and Methods

3.2.1 Sinigrin supplementation in media and bacterial cell collection

In the previous report, GSL-degrading enzyme activity L. agilis R16 was shown to be

inducible by sinigrin (Palop et al., 1995) and the results from Chapter 2 support this report.

The same result was obtained for E. coli O83:H1 NRG 857C. Due to time and resource

restriction, only two bacteria, L. agilis R16 (LA) and E. coli O83:H1 NRG 857C (ECO), were

studied in this chapter. The optimal GSL concentration to be used to induce bacterial GSL-

degrading enzyme was determined by incubating various concentrations of sinigrin in 1 mL

culture broths containg or E. coli O83:H1 NRG 857C cells for 8 h at 37°C under anaerobic

conditions. The GSL degradation and degradation production were assessed from each

substrate concentration. The sinigrin concentration that yielded the highest degradation of

sinigrin and highest AITC production was used for the next experiment.

An overnight culture (1 mL) of LA or ECO culture was sub-cultured into 40 mL modified

MRS broth (glucose omitted for LA) and NB broth (for ECO) containing 2 mM sinigrin in a 50

mL sterile falcon tube. The cultures were grown anaerobically using AnaeroGen sachets

(Oxoid, UK) in an AnaeroGen anaerobic-generating system (Oxoid, UK) for 8 h at 30°C for LA

and 37°C for ECO without shaking. For the control sample, 2 mM glucose was added instead

of sinigrin. After 8 h, the bacterial cultures were centrifuged in an Eppendorf 5415 D

centrifuge at 3,300g for 30 min at 4°C to pellet bacterial cells. The bacterial pellets were

washed twice with 30 mL PBS buffer and stored at - 80°C until analysis. The supernatants from

an overnight culture were subjected to HPLC analysis for the detection of sinigrin degradation,

and to GC-MS analysis for the detection of AITC/ANIT production as previously described in

Chapter 2.

3.2.2 Cell lysis and protein extraction

Two different protocols used for protein isolation were as follows.

(i) Extraction using R3 lysis reagent (Bio-Rad). Pelleted cells were suspended in 2 mL

of 2D lysis buffer [7 M urea, 2 M thiourea, 4% CHAPS, 2% IPG Buffer, 40 mM DTT], and 25 µL

protease inhibitor was added into a solution. Then bacterial cells were lysed using a one-shot

cell disruption system (Bugbuster, Constant system Ltd, UK) under 30k PSI (2000 bar). Cell

177

lysates (2 mL) were aliquoted into a 2 mL eppendorf which was then centrifuged at 13.2 K

rpm at 4°C for 10 min. The clear supernatant 1 was carefully transferred into a 2 mL

Eppendorf tube. The remaining pellet was resuspended with 2 mL 2D lysis buffer by vigorous

vortexing for 2 min. The suspension was centrifuged at 16,100g at 4°C for 10 min. The

supernatant 2 was carefully transferred into a new 2 mL Eppendorf tube. Supernatants 1 and

2 (250 µL each) were mixed to make a total volume of 500 µL which was then concentrated

using Millipore Amicon Ultracentrifugal filters 0.5 mL 10K according   to   the   company’s  

instructions. The final volume of 100 µL concentrated supernatant was obtained and cleaned

using 2D clean-up kit (Bio-Rad) as per manual. Finally, the protein supernatant was adjusted

reach concentration of 200 µg/50 µL with 2D rehydration buffer (7 M urea, 2 M thiourea, 2%

CHAPS, 0.5/2% Pharmalyte or IPG Buffer, 0.002% bromophenol blue, 7 mg DTT per 2.5-ml

aliquot was added just prior to use) and stored at - 80°C until analysis.

(ii) Lysis by Enzymatic Treatment with Lysostaphin. After harvesting bacteria, cells

were washed twice with 5 mL of 0.1 M PBS buffer and once with 5 mL of digestion buffer (10

mM Tris-HCl, pH 7.6, 1 mM EDTA, 5 mM MgCl2). Cells were resuspended in 3 mL of digestion

buffer containing 10 µL of 100X protease inhibitor cocktail for bacterial cells (Melford, UK)

and 5U of lysostaphin (Sigma-Aldrich, UK). After incubation at 37°C for 30 min, the clear

supernatant was obtained by centrifugation at 8000g for 15 min at 4°C. The supernatant was

then treated with DNase (0.75 mg/mL) and RNase (0.5 mg/mL) by incubating for 15 min on ice

to remove nucleic acid contaminants. The supernatant was further treated with 20% v/v TCA

in ice for 30 min and the precipitate was collected by centrifugation at 16, 100g for 20 min at

4°C. The precipitated protein was washed with 200 µL of acetone to remove traces of TCA and

finally acetone was removed by speed vacuum treatment. Precipitated protein was re-

suspended in 2D rehydration buffer and  stored  at  −80°C for later use.

3.2.3 Protein quantification

Since many of the reagents used in the preparation of 2-DE samples, including

detergents (SDS), reductants (DTT), chaotropes (thiourea, urea) and carrier ampholytes, are

incompatible with Bradford protein assay, a 2-DE Quant Kit (GE Healthcare, UK) was used to

determine the accurate quantity of protein in the samples. The procedure works by

quantitatively precipitating proteins while leaving interfering substances in the solution. The

178

assay is based on the specific binding of copper ions to a protein. Precipitated proteins are re-

suspended in a solution containing copper, and unbound copper is determined with a

colorimetric agent. The color density inversely corresponds to the protein concentration.

Different amounts of bovine serum albumin (BSA, 2 mg/mL, provided in the kit) in a

range of 0.5–50   μg   were   added   to   1.5   mL   sterile   Eppendorf   tubes.   Protein   samples   were  

prepared in duplicate, and different volumes (1–50   μL)   were   used   to   ensure   that   protein  

concentration fell within a useful range of the assay (0.5–50   μg).   The   precipitant   (500   μL,

provided in the kit) was added to each tube and was thoroughly mixed by brief vortexing and

incubated at room temperature for 3 min. The co-precipitant  (500  μL,  provided  in  the  kit)  was  

added to each tube and was mixed by inversion briefly and centrifuged at 9,000g for 5 min.

The  supernatant  was  discarded,  and  the  pellet  was  dissolved   in  a  mixture  of  100  μL  copper  

solution  and  400  μL  of  distilled  water.  The  color  reagent   (1  mL)  was  added  to  sample  tubes  

and incubated at room temperature for 15 min. The absorbance at 480 nm was measured

spectrophotometrically using distilled water as a reference. A calibration curve of BSA was

plotted (Figure 3.7), and the protein concentration can be determined from this.

Figure 3.7 BSA calibration curve. Various amounts (0.5–50   μg)   of   BSA   from   2-D Quant Kit (GE Healthcare, UK) were plotted against absorbance at 480 nm.

y = -0.0159x + 1.3952 R² = 0.9795

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0 10 20 30 40 50 60

Abso

rban

ce a

t 480

nm

BSA (µg)

179

3.2.4 Two-dimensional polyacrylamide gel electrophoresis (2D-PAGE)

For the first-dimension electrophoresis, immobilized pH non-linear (NL) gradient (IPG)

Immobiline Drystrip gels (11 cm, pH 3-11 NL) (GE Healthcare, UK) were rehydrated in

Immobilline Drystrip dehydration tray (GE Healthcare, UK) for 17 h at room temperature.

These strips with the pH 3-11 NL gradient have an usual expansion from a middle range pH 5-

7 resulting in a sigmoidal pH gradient which means this range is less narrow separated than in

linear strips. Since most crude lysates from all species contain many polypeptides in this

middle pI range, a non-linear (NL) range should be used for a start. The following day, a total

of 250 µg of protein sample (in 50 μL) was applied onto each strip using a cup loader.

Isoelectro focusing (IEF) was carried out using IPGPhor III machine (GE Healthcare, UK) at 500

V (step and hold) for 6 h, 500-1000 V (gradient) for 1 h, then 1000-8000 V (gradient) for 5 h

and then 8000 V (gradient) for 5 h. The average run took approximately 52000 Vh in total. To

enhance the efficiency in protein transfer from the first to the second dimension, IPG strips

were incubated in reducing buffer (1.0% (w/v) DTT, 75 mM Tris-Hl pH 8.8, 6 M urea, 29.3%

(v/v) glycerol, 2% SDS, 0.002% bromophenol blue) for 15 min, followed by 15 min incubation

in alkylation buffer (2.5% (w/v) iodoacetamide, 75 mM Tris-Cl pH 8.8, 6 M urea, 29.3% (v/v)

glycerol, 2% SDS, 0.002% bromophenol blue). Strips were then overlaid on 4-12% pre-cast

Criterion XT Bis-Tris-SDS gels (13.3 x 8.7 cm (W x L), 1 mm thick) (Bio-Rad, UK) and sealed with

agarose solution (0.5% agarose, 1% bromophenol blue in a Laemmli SDS electrophoresis

buffer). Protein marker (7 µL), either Rainbow marker (GE healthcare, UK) (Figure 3.8A) or

Low Range unstained marker (Sigma-Aldrich) (Figure 3.8B), was loaded onto the marker well.

The second-dimension gel electrophoresis was carried out in 1X MES SDS running buffer (50

mM MES, 50 mM Tris-base, 3.47 mM SDS, 1.0 mM EDTA, pH 7.3) at 40 V for 10 min, and then

at 150 V until the dye front reached the bottom of the gel (~1 h). Gels were stained in ~ 25 mL

of InstantBlue Coomassie® stain (Expedeon, UK) untill the bands became visible. Gels were

washed with ultrapure water twice and kept in ultrapure water at 4°C till further analysis.

180

Figure 3.8 Protein ladders used in 2-DE work. (A) Rainbow marker (GE healthcare, UK). (B) Low Range unstained marker (Sigma-Aldrich). 3.2.5 Image acquisition and analysis

Stained gels were scanned using the 2D gel CCD image analyzer Dyversity (Syngene,

UK) at a resolution of 100 microns for 1100 ms. The exposure time was adjusted to achieve a

value of ~55,000-63,000 pixel intensity as previously described (Brobey et al., 2006).

Subsequently the images were analyzed using Nonlinear Dynamics Progenesis v4 software

(Newcastle upon Tyne, UK). For the purpose of this experiment, each set of gel replicates

from either culture grown on media with or without sinigrin supplementation was combined

into average gels. The average gel represents reproducible spots present on both sets of the

replicate gels. The gel with the greatest number of spots was automatically selected as the

image for the reference gel, and unmatched spots from the other gels were added to this

image to give a comprehensive reference gel for matching spots on the different gels. Gel

images were aligned by automated calculation of six manually assigned alignment landmark

vectors. Scanned gels were analyzed by intra-gel (difference in-gel) and inter-gel (biological

variance) analysis. The automatically detected spots by the software were visually inspected

as well. Any artefacts were removed by manual spot filtering and editing to correct for spots

that did not split properly or were not automatically detected by the software. The spot

intensity levels were normalised by expressing the intensity of each protein spot in a gel as a

proportion of the total protein intensity detected for the entire gel (relative volume, %

volume) to study quantitative changes. A 2-fold increase threshold (spot volume ratio change

and ANOVA p < 0.05 and a false discovery rate of q < 0.05) was chosen as a criterion in the

181

identification of differentially expressed protein candidates. Since the primary objective was

to search for inducible proteins or upregulated proteins with more likelihood to be putative

myrosinases, any spots with changes in downregulated expression upon sinigrin

supplementation were omitted from further analysis.

Each spot that met this criterion was visually verified to confirm whether it was

correctly detected, and the matching of the spots between the replicate sets was accurate.

Protein bands and spots were excised manually from the gels and subjected to in-gel tryptic

digestion.

3.2.6 Estimation of pI and molecular weight (Mr) of the proteins

By plotting the pH of an IPG strip as a function of its length, the pI of a protein can be

determined from its focused position on that strip (Figure 3.9A). This estimation method has

been routinely used in literature without the need to use IEF pH 3-11 marker (Furuhashi et al.,

2010) (Figure 3.9B). Likewise, by plotting the molecular weight (Mr) of proteins from Low

Range unstained marker (Sigma-Aldrich) as a function of migration distance, the Mr of a

protein can be determined from the distance it migrated on the gel (Figure 3.9C).

Figure 3.9 Calibration curves for pI and Mw determination. (A) The graph of pH 3-11 NL range versus the % length of IPG strip. (B) Representative 2-DE gel with pI values (3–11 NL) indication above the gel This figure was taken from Furuhashi et al. (2010). (C) The graph of Mr of proteins from Low Range unstained marker (Sigma-Aldrich) versus distance of migration.

182

3.2.7 In-gel tryptic digestion

For mass spectrometric identification, protein extracts were digested with sequencing

grade trypsin as described by (Bouwman et al., 2004). The reagents used and their

compositions are shown in Table 3.1.

Table 3.1 Reagents used in in-gel tryptic digestion and their compositions

Reagents Compositions

100 mM ammonium bicarbonate (BDH, UK)

395.3 mg ammonium bicarbonate, up to 50 mL of HPLC water

100 mM ammonium bicarbonate in 50% acetonitrile

395.3 mg ammonium bicarbonate, 25 mL of HPLC acetonitrile, up to 50 mL of HPLC water

10 mM DTT (Melford, UK) 7.71 mg dithiothreitol (DTT), up to 5 mL of 100 mM ammonium bicarbonate *MAKE FRESH*

50 mM IAA (Sigma-Aldrich, UK) 56 mg iodoacetamide (IAA) reconstituted in 6.06 mL of ammonium bicarbonate *MAKE FRESH*

20 mM ammonium bicarbonate in 50% acetonitrile

10 mL of 100mM ammonium bicarbonate, 15 mL of HPLC water, 25 mL of HPLC acetonitrile

40 mM ammonium bicarbonate in 10% acetonitrile

20 mL of 100mM ammonium bicarbonate, 5mL of HPLC acetonitrile, 25mL of HPLC water

50 mM acetic acid (VWR) 144  μL  of  acetic  acid,  up  to  50  mL  of  HPLC  water  

Trypsin Gold (Promega) solution

100   μL   of   50  mM   acetic   acid,   up   to   5  mL   of   40  mM  ammonium bicarbonate in 10% acetonitrile aliquot into  500  μL  Eppendorf  Protein  Lobind  Tubes  and  store  at -80°C.

50% acetonitrile/5% TFA 25 mL of HPLC acetonitrile, 2.5 mL Trifluoroacetic acid (TFA) (Sigma-Aldrich, UK), up to 50 mL HPLC water

The band was excised from the gel using a cutting blade and forceps (rinsed with 70%

ethanol before and after each band) into 1 x 1 mm pieces and place in a 1.5 mL Lobind

Eppendorf tube. All washing steps were carried out at a room temperature on a shaker unless

183

otherwise   stated.   The   gel   pieces   were   rinsed   with   300   μL   of   HPLC   water   for   15   min   and  

washed  with  300  μL  of  HPLC  acetonitrile  for  15  min,  and  the  supernatant  discarded.  Gel  pieces  

were  washed  with  300  μL  of  100  mM  ammonium  bicarbonate  for  15  min,  and the supernatant

discarded.  The  gel  pieces  were  washed  with  300  μL  of  100  mM  ammonium  bicarbonate  in  50%  

acetonitrile for 15 min, and the supernatant was discarded. The gel pieces were washed with

100   μL   of   HPLC   acetonitrile   for   5  min   and   the   supernatant  was discarded. Gel pieces were

dried in a speed vacuum Heto VR-1 (Heto-Holten, Denmark) for 5 min before 50  μL  of  10  mM  

DTT was added and incubated for 1 h at 60 °C. The supernatant was discarded before gel

pieces   were   added   with   50   μL   of   50   mM   IAA   and   incubated for 30 min in the dark. The

supernatant  was  discarded  before  gel  pieces  were  washed  with  300  μL  of  100  mM  ammonium  

bicarbonate for 15 min. The supernatant was discarded before gel pieces were washed with

300  μL  of  20  mM  ammonium  bicarbonate  in  50%  acetonitrile for 15 min. The supernatant was

discarded   before   gel   pieces   were   washed   with   100   μL   of   HPLC   acetonitrile   for   5   min.   The  

supernatant was discarded, and gel pieces were dried in a speed vacuum for 5 min. The gel

pieces  were  added  with  20  μL  of  1  μg/μL  trypsin  solution  and  incubated  for  1  h.  After  that,  40  

mM ammonium bicarbonate in 10% acetonitrile was added to overlay the gel pieces and

incubated overnight at 37°C.  The  following  day,  the  reaction  tubes  were  added  with  150  μL  of  

HPLC water, and incubated at 37°C for 10 min. The supernatants were transferred to a 1.5 mL

Lobind  Eppendorf  tube.  The  gel  pieces  were  extracted  twice  with  50  μL  of  50%  acetonitrile/5%  

TFA for 60 min each time. The extracts were pooled and dried in a speed vacuum. The dried

extracts  were  dissolved  in  20  μL  of  0.5%  formic  acid  immediately  before  LC-MS/MS analysis.

3.2.8 LC-MS/MS Analysis

The samples were analyzed on an Applied Biosystems QTrap MS coupled to an Agilent

1100 LC stack. The Agilent stack consisted of a binary pump, capillary pump, well plate auto-

sampler and a column oven with integrated 6-port valve. The samples were loaded onto a

trap column (Agilent Zorbax SB 5 µm x 0.3 mm x 35 mm) using the binary pump; the trap

column was washed and then switched into the capillary flow; peptides were separated on a

capillary column (Agilent SB 5 μm 0.5 mm x 150 mm column). The LC was interfaced to the MS

with a turbo ion spray source. The loading/washing solvent was milli-Q H2O containing 0.2%

COOH, 0.02% TFA at a flow rate 150 μL/min and the resolving solvent was a linear gradient

184

system of 0% B to 40% B over 60 min at a flow rate of 10 μL/min [(A) 94.9% H2O, 5% CH3CN,

0.1% COOH; (B) 94.9% CH3CN, 5% H2O, 0.1% COOH]. The column oven temperature was set to

40°C, and the valve was switched to direct the flow from the trap into the resolving column

after a 5 min wash. The MS parameters were normally set to Curtain Gas 10 psi, GS1 20 psi,

GS2 20 psi, interface heater on, temperature at 150°C, DP 65. The acquisition method

consisted of an enhanced mass spectrum (EMS) survey scan (350-1200m/z) followed by an

enhanced resolution (ER) scan and then enhanced product ion scans (100-1500 m/z) of

selected ions. The information dependant acquisition (IDA) was set for the top 4 most intense

peaks after dynamic background subtraction of the survey scan. The criteria include; m/z >

325 < 1200 with a charge state of 2-3, the ER scan was used to determine charge state, rolling

collision energy was used, +1 precursors were excluded for the dependant scans. Former

target ions were excluded after two occurrences for 2 min.

3.2.9 Database searching and protein identification

Mass spectrometric data was analyzed by the database search engine ProteinPilot

using the Paragon algorithm (4.0.0.0, 459) (Applied Biosystems). The sample parameters were

set to: trypsin digestion, cysteine alkylation set to iodoacetamide, urea denaturation and

acetylation emphasis. The C-terminal cleavage at lysine (Lys) and arginine (Arg) was selected

for   trypsin   specificity.   “Biological   modification”   was   set   for   processing   parameters,   and   a  

thorough ID search effort was selected. The peptide and fragment mass tolerances were < 10

ppm. During the search by Protein Pilot, an automatic mass recalibration of the data sets

based on highly confident peptide spectra was carried out. A first search iteration was

specifically performed to select high confidence peptide identifications. These were then used

to recalibrate both the MS and MS/MS data, which is automatically re-searched. Tandem MS

data were searched against Lactobacillus and E. coli O83:H1 NRG 857C database from UniProt

release 2012_01. All reported proteins were identified with at least two peptides having a

confidence  (Conf.)  interval  of  ≥  95% (p < 0.05)  as  determined  by  ProteinPilot™  Unused  scores  

(≥ 1.3) with the corresponding false positive discovery rate below 1%. The winner protein

selected to represent the spot would have the highest unused ProScore with the theoretical

pI and Mr values close to the calculated ones (based on the spot location) and this protein was

found in all triplicates.

185

3.3 Results

3.3.1 Optimum GSL concentration to induce bacterial GSL-degrading activity In Chapter 2, the putative bacterial GSL-degrading activity from all three bacteria

tested is likely to be inducible by GSL addition. It was also found that ITC and NIT production

peaked at 8 h following addition of most GSL substrates in cultures of L. agilis R16 and E. coli

O83:H1 NRG 857C. Therefore, bacterial cells were pelleted at 8 h so as to avoid possible

degradation or reduced expression of putative bacterial GSL-degrading activity, if any, rather

than collecting them later. In this chapter, the aim was to identify inducible proteins in

bacterial cells grown on GSL supplementation using the 2-DE approach as these identified

proteins may potentially be involved in the metabolism of GSLs in human gut bacteria.

Firstly, the optimal concentration of sinigrin to be supplemented during bacterial

fermentation to promote GSL-degrading activity was determined by measuring the

degradation percentage of several sinigrin concentrations tested and accordingly the AITC

production over 8 h in two bacteria, L. agilis R16 and E. coli O83:H1 NRG 857C. The highest

degradation of sinigrin and AITC formation in each bacterium were resulted from 2 mM

sinigrin (Figure 3.10). Therefore, 2 mM sinigrin was supplemented in bacterial cultures to

maximize myrosinase–like activity in both bacteria for 2-DE gel experiments.

186

Figure 3.10 Sinigrin degradation and AITC production from various sinigrin concentrations at 8 h in bacterial fermentation. (A) Sinigrin degradation (%) of several sinigrin concentrations in L. agilis R16 anaerobic incubatiosn at 30˚C for 8 h. (B) AITC production (%) over digested sinigrin in L. agilis R16. (C) Sinigrin degradation (%) of several sinigrin concentrations in E. coli O83:H1 NRG 857C anaerobic incubatiosn at 37˚C for 8 h. (D) AITC production (%) over digested sinigrin by E. coli O83:H1 NRG 857C. Values are means ± SD, n = 3. Degradation (%) was calculated from the number of moles of sinigrin degraded over the number of moles of initial sinigrin in percentage. AITC production (%) was calculated from the number of moles of AITC produced over the number of moles of degraded sinigrin in percentage.

Comparative analysis of growth curves between cultures of both L. agilis R16 and E.

coli O83:H1 NRG 857C with and without 2 mM sinigrin supplementation over 8 h showed no

significant difference of growth kinetics in the same bacterium (Figure 3.11). Both bacteria

reached a stationary phase approximately the same time at 6 h. This result suggests that

sinigrin supplementation did not alter bacterial growths when compared with the control

cultures without its supplementation. The similarity in growth kinetics of the two cultures

(with and without sinigrin supplementation) of same bacterium would make the comparative

analysis of protein patterns from these bacteria justified.

At different growth phases, different proteins are produced and predominate e.g.

ribosomal proteins are predominant at log phase where there is high activity of protein

synthesis to support bacterial growth while proteolytic proteins are predominant at

187

stationary phase where proteins are degraded and recycled. This would have posed a

problem in comparative analysis in interpreting protein expression patterns if the growth

rates of bacteria are far different.

Figure 3.11 Growth kinetics of bacteria with and without sinigrin supplementation over 8 h. Cultures of L. agilis R16 (LA) and E. coli O83:H1 NRG 857C (ECO) were grown on media either with sinigrin supplementation (S) or without it (N) for 8 h of anaerobic incubation at 30°C for LA and 37°C for ECO. The OD600nm values were plotted in log scale. Values are means, n = 3.

3.3.2 Optimization of protein sample preparation for 2-DE

Employing proteomics technologies to search for putative bacterial GSL-degrading

enzymes and possible key bacterial proteins involved in the metabolism of GSLs was set as the

ultimate goal. This could be achieved by establishing 2-DE as a tool to study protein

expression patterns of bacterial cells grown on media with and without sinigrin

supplementation. The key step to obtain a high quality 2-DE gel is to optimize the conditions

for protein extraction and protein content to be loaded on the 2-DE gel. There are several

ways for bacterial protein isolation using different lysis reagents or lysis methods. However,

the preliminary results suggest that cell disruption by a cell disruption machine led to a

greater release of protein content from a Gram-positive bacterium L. agilis R16 in comparison

with a method of sonication (Abdulhadi Albaser, PhD thesis). Therefore, a cell disruption

machine was used throughout for cell lysis.

188

Initially, proteins from cells grown on media without sinigrin supplementation were

prepared by two different cell lysis procedures: (i) using R3 extraction reagent alone and (ii)

enzymatic lysostaphin treatment. Lysostaphin, as an endopeptidase, cleaves the pentaglycine

cross-bridges of the staphylococcal cell wall rapidly lysing the bacteria (Kusuma & Kokai-Kun,

2005; Kumar et al., 2008). This enzyme has been used to lyse cell walls of Gram-positive

bacteria and isolate membrane proteins from Gram-positive Staphylococcus aureus bacterium (Nandakumar, et al., 2005).

The results showed that method (i) yielded 1.04 µg/µL of isolated protein while

method (ii) yielded 7.35 µg/µL (which is approximately seven-fold higher than method (i)).

There was a similarity in protein distribution between these two samples, with the majority of

the proteins ranging from pI 4 to pI 6. Method (i) produced well-separated spots (Figure

3.12A) while method (ii) gave suboptimal resolution at the acidic end (pI < 5) (Figure 3.12B).

After careful automated and manual editing to correct for artifacts using Progenesis v4

software (Section 3.2.5), 210 spots were detected from method (i) (Figure 3.12A) and 523

spots from method (ii) (Figure 3.12B). Method (ii) yielded more than a two-fold increase of

the number of detectable spots. This indicates that addition of lysostaphin (5U) markedly

improved efficiency in protein isolation from Gram-positive bacterium L. agilis R16. Therefore,

enzymatic treatment with lysostaphin was used throughout to release as many proteins as

possible. It was also found that 250 µg was the optimal protein content for loading on the 2-

DE gel (Figure 3.12B).

189

Figure 3.12 Comparison of protein patterns from two cell lysis methods. L. agilis R16 cells were grown in MRS broths without sinigrin supplementation and the cell-free extracts were analyzed on 2-DE gels. (A) Proteins prepared by ProteoPrep membrane extraction kit (Sigma-Aldrich) with 80 µg of protein loaded on the gel. (B) Proteins prepared by lysostaphin treatment with 250 µg of protein loaded on the gel. Approximate pI value (3–11 NL) is indicated above gel photo. Protein ladders (7 µL), Rainbow marker (GE healthcare, UK) was loaded in (A), Low Range unstained marker (Sigma-Aldrich) was loaded in (B). Only one replicate from each method was carried out.

In addition, three biological replicates of 2-DE gels from each set of L. agilis R16 grown

on media with and without sinigrin supplementation were compared to determine the

reproducibility of 2-DE technique. It was shown that 2-DE gels were fairly reproducible with a

similar number of protein spots, and expression patterns (Figure 3.13).

190

Figure 3.13 Reproducibility of 2-DE gels from L. agilis R16. (A) Triplicates of proteins from cells grown on media without sinigrin. (B) Triplicates of proteins from cells grown on media with sinigrin.

191

3.3.3 Comparative analysis of 2-DE maps of proteins isolated from cells grown on media

with and without sinigrin supplementation

To identify inducible proteins upon sinigrin supplementation in the cultures of L.

agilis R16 and E. coli O83:H1 NRG 857C, bacteria were grown on media with and without 2

mM sinigrin for 8 h and 2-DE maps obtained from these cultures were used for comparative

analysis. The 2-DE gels of each set of cultures (with or without sinigrin supplementation)

from each bacterium were produced in triplicates for L. agilis R16 and duplicates for E. coli

O83:H1 NRG 857C. These gels were subjected to gel analysis using Progenesis v4 software

(Section 3.2.5) for spot counting and identification of spots with changes in abundance.

Information about bacterial growth, spot numbers on 2D gels of each set of cultures (with or

without sinigrin supplementation) from each bacterium are shown in Table 3.2.

Table 3.2 Experimental results of each set of cultures (with or without sinigrin

supplementation) from each bacterium

Bacteriaa ORF no.b Spot no.c OD600nm Sinigrin degraded  (μM)d

AITC + ANIT production

(μM)e

Percentage product

(%)

LA (S) 3,088

523 ± 19 0.758 ± 0.010 1000 654 ± 18 65

LA (N) 561 ± 25 0.714 ± 0.012 ND ND ND

ECO (S) 4,575

363 ± 27 0.531 ± 0.009 1000 611 ± 26 61

ECO (N) 354 ± 15 0.529 ± 0.011 ND ND ND aLA, L. agilis R16; ECO, E. coli O83:H1 NRG 857C; S, supplemented with sinigrin; N, without sinigrin. Bacterial fermentations were carried out at 30°C for LA and 37°C for ECO under anaerobic conditions for 8 h. bNumber of open reading frame (ORF) predicted in L. plantarum strain JDM1 and in E. coli O83:H1 NRG 857C by UniProt. cNumber of protein spots on 2D gels counted by Progenesis v4 software. d Determined by HPLC nalysis. e Determined by GC-MS nalysis. ND, Not detected; AITC, Allyl isothiocyanate; ANIT, Allyl nitrile. Values are means ± SD, n = 3 (for LA) and n = 2 (for ECO).

192

Sinigrin was not completely metabolized to AITC and NIT products as shown by the

percentage product of 65% and 61% in L. agilis R16 and E. coli O83:H1 NRG 857C,

respectively (Table 3.2). This suggests that some metabolite products may be unaccounted

for in these bacteria, and they were not detected by current GC-MS analysis. Another reason

is possibly due to the instability of AITC in the culture broths (Chapter 2, section 2.3.4), and

thus it declined over time during bacterial fermentations. Higher numbers of protein spots

were observed in cells grown on media with sinigrin supplementation from both bacteria

(Table 3.2). However, more spots were detected in 2D maps of L. agilis R16. The number of

spots account for a sixth, and a tenth of open reading frames (ORFs) predicted for gene

products in L. agilis R16 and E. coli O83:H1 NRG 857C, respectively (Table 3.2). Since the

genome and proteome database of L. agilis R16 is not available, this analysis was carried out

using ORF data of the relative bacterium L. plantarum strain JDM1.

Representatives of 2-DE gels of each set of cultures from each bacterium were

shown in Figure 3.14. A majority of spots were found in a more acidic pI range of 4-5.5 in L.

agilis R16 (Figure 3.14A and 3.14B) which is a lactic acid bacterium (LAB) and a pI range of

4.5-6 in E. coli O83:H1 NRG 857C (Figure 3.14C and 3.14D). In E. coli O83:H1 NRG 857C, the

spots between 29-66 kDa and pI of 4.5-5.5 are not well-separated possibly due to non-

protein impurities at the acidic end (Figure 3.14C and 3.14D).

There are 32 and 35 spots in L. agilis R16 and E. coli O83:H1 NRG 857C with at least

two-fold increase in abundance in response to sinigrin were detected (Figure 3.14B and

3.14D, respectively). Any spots with changes in downregulated expression upon sinigrin

supplementation were omitted from further analysis (although these proteins may involve in

GSL metabolism) since the primary objective was to search for inducible proteins or

upregulated proteins which are more likely to be putative bacterial GSL-degrading enzymes.

These spots with at least two-fold increased intensity were excised from the gels

and enzymatically digested with trypsin (Section 3.2.7) prior to LC-MS/MS analysis for

protein identification. Unfortunately, some spots were lost during in-gel trypsin digestion

procedure involving multiple steps, and some protein samples were contaminated with

human skin proteins and/or dust that produced poor results in protein identification that did

not meet the criteria for protein identification used in ProteinPilot™   Software 4.0. These

unidentified spots were indicated in blue circles in Figure 3.14.

193

A strict cut-off for protein identification was applied to minimize false positive

results.  Unused  ProtScore  ≥  1.3,  which  corresponds  to  a  confidence  (Conf.)  limit  of  95%, and

at least two peptides with 95% confidence were considered for protein identification

(Section 3.2.9). Therefore, only the protein spots with the highest unused ProScore with the

theoretical pI and Mr values close to the calculated ones and good % coverage were

presented in the next section (red circles in Figure 3.14).

194

Figure 3.14 Comparative analysis of representative 2-DE maps of bacterial proteins. The 2-DE proteome maps obtained from cells of (A) L. agilis R16 (LA) grown without sinigrin supplementation for 8 h at 30˚C. (B) L. agilis R16 culture with sinigrin. (C) E. coli O83:H1 NRG 857C (ECO) culture without sinigrin supplementation for 8 h at 37˚C. (D) E. coli O83:H1 NRG 857C culture with sinigrin. Both red and purple circles indicate protein spots with at least two-fold increase in abundance. Red circles with numbers indicate successfully identified proteins while purple circles indicate unidentified spots with low Conf. value (Progenesis v4 software). These gels are representatives of three replicates from LA and duplicates from ECO.

195

The increase in protein abundance was determined by using Progenesis v4

software (Section 3.2.5). The montage 3D comparison of the spot at the same

position between two 2-DE gels (with and without sinigrin supplementation) was

generated as an example (Figure 3.15).

Figure 3.15 Representative comparison of 3D montage of expression levels of protein spot. Intensity levels of spot no. 4 (oxidoreductase) from 2-DE maps of (A) L. agilis R16 culture grown on media without sinigrin supplementation and (B) L. agilis R16 culture grown on media with sinigrin supplementation were compared using Progenesis v4 software. A three-fold increase in spot intensity in (B) was shown. 3.3.4 LC-MS/MS analysis and protein identification

All upregulated proteins spots were subjected to in-gel digestion and

analyzed by LC-MS/MS. Both representative precursor MS and peptide sequence of

D-lactate dehydrogenase from L. agilis R16 (Figure 3.14B, spot no. 10) are shown in

Figure 3.16. Each peptide mass fingerprint obtained was used to search for the

positive protein using UniProt database (Consortium, 2012) via ProteinPilot™  

software v4.0. The unused score (a confidence percentage) was calculated, and it

reflects   the   probability   of   a   hit   being   a   “false   positive”.   There   is   a   false   positive  

identification probability of about 5% at the 95% confidence level. There are 28

upregulated proteins (12 from L. agilis R16 and 16 from E. coli O83:H1 NRG 857C)

that were successfully identified (Table 3.3).

196

Figure 3.16 Representative MS and MS/MS spectra. The identified protein D-lactate dehydrogenase of L. agilis R16 (Figure 3.14B, spot no. 10) produced (A) MS spectrum of matched   peptide   sequence   ‘AWHSSSETTAK’  with   corresponding   precursor  MS   region.   The  sequences were determined with a 99 score confidence (Conf.). (B) MS/MS spectrum of the matched peptide. Both spectra were generated by ProteinPilot software v4.0.

Since the proteome and genome database of L. agilis R16 is unavailable, the

identification of upregulated proteins in this bacterium relied on the homology to

proteins belonging to other Lactobacillus species including L. plantarum strain ATCC

BAA-793/strain JDM1. Thus, the observed Mr/pI of spots was different from their

theoretical Mr/pI (Table 3.3). This also re-occurred in E. coli O83:H1 NRG 857C that

has its genome/proteome available in the database. This might result from an

unusual posttranslational modification. In addition, when the proteins are focused

too long, cysteins become oxidised and the pI of the proteins change. Some proteins

become unstable at their isoelectric point. The modified proteins have a different pI

and start to migrate again with the horizontal streaks radiating from the spots (Lopez,

2007). This may also explain the differences in theoretical and experimental Mr and

pI values.

197 Table 3.3 Protein identification of upregulated spots  (≥  2  fold  increase  in  spot  volume  ratio  and  ANOVA  p  ≤  0.05  with  ≥  2  matched  peptides)  of  L. agilis R16 (LA) and E. coli O83:H1 NRG 857C (ECO) anaerobically grown on media with 2 mM sinigrin supplementation for 8 h at 30˚C for LA and at 37˚C for ECO

Spot no.

Uniprot Accession

no. Protein description

Theoretical Mr

(kDa)/pI

Calculated Mr

(kDa)/pI

Unused ProtScorea

% Coverageb

Peptide matched

Relative spot

volume ratioc

L. plantarum (strain ATCC BAA-793/strain JDM1) Purine metabolism

1 C6VN56 Deoxyguanosine kinase 24.3/4.76 20.3/4.51 7.95 37.2 5 2.2 Proteolysis

2 C6VM88 ATP-dependent Clp protease proteolytic subunit 21.5/4.87 20.8/4.63 17.62 58.7 7 2.9 Oxidoreduction

4 C6VJP1 Oxidoreductase 20.4/5.02 17.6/5.05 16.92 68.2 9 3.3 6 C6VM58 Oxidoreductase 31.7/5.16 26.1/4.63 12.18 76.1 13 2.2 8 C6VNQ6 Oxidoreductase 35.9/5.19 35.8/5.43 11.32 72.7 10 2.3

Hydrolysis 5 C2FJX3 Haloacid dehalogenase (HAD) superfamily hydrolase 25.7/4.9 25.5/4.95 5.9 56.7 6 2.3

Carbohydrate metabolism 3 C6VMV8 Phosphoglycolate phosphatase 23.5/4.94 18.6/4.83 4.82 38.9 3 2.4 7 C6VNI4 Ribokinase 32.2/4.82 30/4.74 9.89 43.4 5 2.1 9 O32755 Glyceraldehyde-3-phosphate dehydrogenase 36.6/5.62 36.2/5.49 9.66 43.8 4 2.5

10 C0LJH4 D-lactate dehydrogenase 37.2/4.88 38/4.75 8.31 43.4 4 2.1 11 Q88V41 Acetate kinase 43.5/5.96 48/5.85 9.19 56.9 9 2.0

Transport 12 Q88ZZ2 Putative ABC transporter ATP-binding protein lp_0149 62.4/5.05 66/5.05 11.56 47.3 6 2.2

a Unused ProtScore is calculated using only peptides from th spectra that have not already been used to justify more confident proteins. It is a true indicator of protein confidence. b % coverage represents % of the no. of amino acids matching to ≥ 1 identified peptide divided by the total no. of amino acids in the protein sequence. C Increase in protein expression of the spots from the cultures with sinigrin versus control cultures without sinigrin as analyzed by ProteinPilot software v 4.0.

198

Spot no.

Uniprot Accession

no. Protein description

Theoretical Mr

(kDa)/pI

Calculated Mr

(kDa)/pI

Unused ProtScorea

% Coverageb

Peptide matched

Relative spot

volume ratioc

E. coli O83:H1 NRG 857C

Sugar Transport 13 E4P223 Glucose-specific PTS system component 18.3/4.73 17.3/4.72 5.58 18.3 3 3.1

23 E4P2U5 Fused putative sugar transporter subunits of ABC superfamily: ATP-binding components 55.9/5.61 51.2/5.71 10.15 25.7 5 2.1

25 E4P9F3 N-acetyl glucosamine specific PTS system components IIABC 68.3/5.78 66.1/5.74 7.81 32.1 4 2.2

Carbohydrate metabolism 14 E4P7M3 Adenylate kinase 23.5/5.75 25.6/5.78 15.31 60.8 11 2.2 17 E4P1Z9 Glucokinase 34.7/5.95 37.2/5.85 4.69 15.9 2 2.1 19 E4PDZ4 Acetate kinase 43.3/5.85 43.3/5.66 20.23 50 8 2.5 20 E4PD26 Glucose-1-phosphatase/inositol phosphatase 45.7/5.60 46.4/5.60 5.53 18.9 3 2.8 22 E4P9L3 Glucose-6-phosphate dehydrogenase 55.7/5.56 51.5/5.68 10.98 45.3 5 2.4

Hydrolysis 15 E4PCB5 Putative hydrolase 29.28/5.57 29.0/5.81 19.15 58.3 12 2.2

Oxidoreduction 16 E4PD28 Flavoprotein WrbA 20.8/5.59 20.5/5.72 34.71 85.4 26 2.3

21 E4P7U4 Uncharacterized protein (Putative Oxidoreductase) 45.9/5.63 44.2/5.65 26.79 66.2 15 2.3

26 E4P8L1 Alkyl hydroperoxide reductase subunit C 20.8/5.03 16.7/4.87 20.38 59.6 12 2.5

27 E4P713 Superoxide dismutase 21.3/5.58 22.0/5.68 22.47 74.6 17 2 Proteolysis 28 E4P7I7 ATP-dependent Clp protease proteolytic subunit 23.2/5.52 23.8/5.70 26.08 50.7 10 2.4

Others 18 E4P96X Putative uncharacterized protein 39.6/6.0 41.3/5.77 11.92 48.5 6 2.3 24 E4PAW9 Uncharacterized protein 58/5.72 53.6/5.83 6.98 52.4 5 2.2

199

A majority of upregulated proteins in L. agilis R16 belong to carbohydrate metabolism

and oxidoreduction system while the minority belong to purine metabolism, hydrolysis, and

proteolysis (Figure 3.17). Similarly, most upregulated proteins in E. coli O83:H1 NRG 857C were

found in carbohydrate metabolism, oxidoreduction system and sugar transport while the rest

found in hydrolysis, proteolysis and others (Figure 3.17).

Figure 3.17 Functional grouping of 28 upregulated proteins identified on 2-DE gels of L. agilis R16 (LA) and E. coli O83:H1 NRG 857C (ECO)

3.57%

7.14%

25%

7.14%

35.71%

14.29%

7.14%

Carbohydrate metabolism

Oxidoreduction

Sugar transport

Proteolysis

Hydrolysis

Others

Purine metabolism LA ECO

200

3.4 Discussion

In this chapter, the usage of 2-DE combined with LC-MS/MS led to identification of

proteins involved in the transport system, proteolysis, oxidoreduction and carbohydrate

metabolism which potentially are important in GSL metabolism in L. agilis R16 and E. coli

O83:H1 NRG 857C. Since 2-DE analysis showed no identification of β-(thio)glucosidases (as

putative myrosinases) on any gels of either bacterium, it is speculated that bacterial GSL-

degrading enzymes may not have amino acid sequences similar to those found in aphid or plant

counterparts or they were not detected on our gels.

Due to the drawback of 2-DE, only the most abundant proteins in complex mixtures are

visualized. Additionally, a number of specific classes of proteins are relatively incompatible with

the isoelectric focusing step in 2-DE, including large, hydrophobic membrane proteins whose

limited solubility leads to protein precipitation and aggregation as well as very acidic or basic

proteins. If this is a case for bacterial GSL-degrading enzymes as a membrane protein that

precipitated or aggregated, it would not appear as a well-defined spot on 2-DE gels and thus it

was overlooked for further analysis.

It is also possible that bacterial GSL-degrading enzymes are low in abundance, yet

exhibit high degading activity towards sinigrin. Low abundance proteins, which often are crucial

to understand some biological changes and do the most important regulatory functions in a cell,

may be present in only a few copies per cell. If this is the case, then one may not be able to

detect a protein spot corresponding to bacterial GSL-degrading enzymes on 2-DE gels. There

are currently three major approaches to overcome the limitation when detecting low copy

number  proteins  in  the  presence  of  highly  abundant  “housekeeping”  proteins.

(i) Ultrazoom gels, i.e. IPG strips which cover a series of narrow, overlapping ranges of pI

(e.g. IPG 4–5, 4.5–5.5, 5–6) for higher resolution and improved detection of low copy number

proteins. Furthermore, computer-aided image analysis and protein identification by MS are

simplified due to the smaller number of co-migrating protein species and the more reliable

database search results (Westbrook et al., 2001). Ultrazoom gels allow the detection of proteins

down to 300 copies due to their higher sample loading capacity (Hoving et al., 2000).

201

(ii) Ultrasensitive protein stains such as silver staining which improves protein detection

up to five-fold as opposed to Coomassie brilliant blue R-250 that detects approximately 0.1 µg

of protein (Hoving et al., 2000).

(iii) Pre-fractionation steps to reduce the complexity of the sample and enrich low copy

number proteins.

Bacterial myrosinase was first successfully purified years ago (Tani et al., 1974). It was

active in-vitro, but its amino sequence was not identified due to lack of technology back in the

day. Since then, most bacterial myrosinases reported in literature were only active in intact

cells, but inactive in-vitro (Elfoul et al., 2001; Palop et al., 1995; Rabot et al., 1995). The finding

that bacterial myrosinase activity was mostly found in vivo renders purification and

identification of bacterial myrosinase more difficult.

Interestingly, myrosinase activity in vitro was detected in cell-free extracts of a Gram-

negative bacterium Citrobacter isolated from soil using enrichment culture containing only GSL

as an only source of carbon in M9 medium (Abdulhadi Albaser, PhD thesis). Although its cell-

free extracts exhibited myrosinase activity as tested positive using GOD-PERID assay, ITC/NIT

products were not detected from GSL degradation in resting cell experiments and bacterial

fermentation experiments. These findings were different from the results obtained from the

two bacteria (tested in this chapter) which did not exhibit myrosinase activity in vitro, but

produced ITC/NIT from GSL degradation in vivo (Chapter 2). All the results thus far suggest that

bacteria may have different mechanisms/enzymes to degrade GSLs and produce corresponding

products. The bacterial GSL-degrading enzyme system may be more complicated than

previously thought.

In addition to myrosinase, other proteins may play a role in the metabolism of GSL in

human gut bacteria. Thus, it is important to identify the proteins that are distinctively

expressed or upregulated upon GSL supplementation in bacterial cultures which may suggest

their involvement in GSL metabolism. The results showed that sinigrin supplementation

resulted in upregulated expression of several proteins in contrast to the control cultures

(without sinigrin supplementation) in both bacteria. Those putative uncharacterized proteins

202

and putative hydrolase found in E. coli O83:H1 NRG 857C from 2-DE analysis may also be

important and involved in GSL metabolism as well as other proteins. It is perhaps best to keep

an open mind as very little is known about the GSL metabolism in human gut bacteria.

L. agilis R16 showed increased abundance in deoxyguanosine kinase. This enzyme

belongs to the family of transferase or kinases that specifically transfer phosphorus-containing

groups to an alcohol group as acceptor. This enzyme participates in metabolism of purines that

are respective building-blocks of DNA and RNA (Gower et al., 1979). The increased expression

of deoxyguanosine kinase may be associated with increased turnover of mRNA transcripts of

genes upregulated by sinigrin supplementation.

ATP-dependent Clp protease proteolytic subunit also had increased expression upon

sinigrin supplementation in both bacteria. This enzyme catalyzes the hydrolysis of proteins into

small peptides in the presence of ATP and Mg2+ ions (Gottesman et al., 1998). The increased

expression of this enzyme may be due to the high turnover of bacterial GSL-degrading enzyme

or its instability. L. agilis R16 was thought to have a very unstable myrosinase (Palop et al.,

1995). The increased abundance in ATP-dependent Clp protease proteolytic subunit may

indicate the presence of abnormal proteins or specific unstable proteins during sinigrin

metabolism.

Interestingly, a very recent study reported that iberin, an ITC derivative of glucoiberin,

significantly and highly upregulated the expression of an efflux transporter, an outer membrane

protein, components of ABC transporters and two (probable) oxidoreductases in Pseudomonas

aeruginosa (Jakobsen et al., 2012). This finding is in accordance with our results on increased

expressions of ABC transporter subunits and oxidoreductases found in L. agilis R16 upon

sinigrin supplementation. It was suggested that both the GSL metabolism and the formation of

its degradation products could affect the transport system supporting influx of GSL and efflux of

its degradation products out of bacterial cells. Oxidoreductases may be required to

biotransform either GSL substrate or ITC/NIT product or the unaccounted for metabolites to the

new more reduced or more oxidized product since 100% percentage product was never

achieved upon metabolism of any GSLs thus far (Chapter 2).

It has been known that the electrophilic feature of ITCs has facilitated its interaction

203

with some nucleophilic agents including amino, hydroxyl, thiol, carboxylic acids from small

peptides, amino acids and water (Zhang et al., 1996; Cejpek et al., 1998). For examples, dietary

ITCs namely sulforaphane, erucin, and iberin are found to play a vital role in the regulation of

redox status through the induction of thioredoxin reductase 1 (TrxR1) in human breast cancer

MCF-7 cells (Wang et al., 2005). AITC, after its formation by sinigrin hydrolysis, appeared to

inhibit bacteria by its interactions with amino acids and proteins and interference with the

action of important enzymes e.g. glutathione reductase, thioredoxin and acetate kinase

(Luciano et al., 2008; Luciano et al., 2009). Our findings with upregulated expressions of acetate

kinase and oxidoreductases may be a result from AITC products as inducers of these proteins

from sinigrin degradation by the studied bacteria.

Desulfation of GSL by putative bacterial sulfatase that consequently produces sulfate

and desulfo-glucosinolate (DS-GSL) may involve oxidoreductases in the metabolic fate of sulfate

and DS-GSL. In addition, GSL can be converted to NIT via a possible redox-cycling step of iron (II)

and iron (III) ions and involves an oxidoreductase (Hanschen et al., 2012). Therefore, increased

expressions of enzymes involved in oxidoreduction system e.g. flavoprotein WrbA, alkyl

hydroperoxide reductase subunit C and superoxide dismutase with oxidoreduction activity

were also observed in E. coli O83:H1 NRG 857C.

GSL, as a β-glucoside, may require phosphorylation modification prior to its degradation

by certain 6-phospho-β-glucosidases (Witt et al., 1993). Thus, some proteins involved in

phosphotransferase system (PTS), sugar transport subunits and kinases such as glucose-specific

PTS system component, N-acetyl glucosamine specific PTS system component IIABC and

glucokinase were identified. These proteins were upregulated upon sinigrin metabolism in E.

coli O83:H1 NRG 857C. Despite intensive studies of PTS permeases, the mechanism that couples

sugar translocation to phosphorylation in E. coli and the nature of the translocation apparatus

are not well-understood (Yagur-Kroll et al., 2009).

Once sinigrin is degraded by putative bacterial GSL-degrading enzyme, glucose and

aglycone are released. It is likely that some proteins involved in sugar/glucose metabolism were

upregulated upon sinigrin supplementation. For example, phosphoglycerate phosphatase,

glyceralaldehyde-3-phosphate dehydrogenase, D-lactate dehydrogenase, acetate kinases from

204

L. agilis R16, and glucokinases, adenylate kinase, glucose-1-phosphatase, glucose-6-phosphate

dehydrogenase from E. coli O83:H1 NRG 857C may be affected at an expressional level.

From our results, greater numbers of protein spots were detected from 2-DE gels of L.

agilis R16 grown on media with and without sinigrin supplementation than E. coli counterparts.

The differences in protein expression between these two bacteria may be caused by differences

in inherent metabolisms of different bacterial strains and/or stimulating effects from different

compositions in growth media. Some ingredients found in modified MRS (glucose omitted)

media for L. agilis R16 e.g. Tween-80, K2HPO4, Na-acetate, (NH4)2 citrate, MgSO4 7H2O and

MnSO4-H2O (in addition to peptone, beef extract, yeast extract, and NaCl in NB media used for

E. coli O83:H1 NRG 857C growth) may stimulate/increase the expression of certain proteins.

These ingredients in bacterial broths may also be the cause of the increased expression of

certain proteins identified in this work.

To conclude, a majority of identified proteins with at least two fold increased

abundance upon sinigrin supplementation in both bacteria belong to transport system,

carbohydrate metabolism and oxidoreduction system. This work provides, for the first time,

identification of proteins that are potentially involved in the metabolism of GSL in human gut

bacteria. However, proteins that are expressed in response to the accumulation of GSL

degradation products including sulfate and ITC/NIT should be accounted for in addition to those

upregulated upon GSL metabolism as performed in this work. For a more comprehensive study,

the proteome change in response to the addition of ITC and NIT into bacterial cultures should

be investigated in the future.

205

Chapter 4: Reverse proteomics approach to identify bacterial proteins potentially involved in the metabolism of GSLs 4.1 Introduction

In parallel to forward proteomics approach (Chapter 3), reverse proteomics

approach was used in this chapter to identify bacterial proteins potentially involved in the

metabolism of GSLs such as bacterial GSL-degrading enzymes, β-O-glucosidases or sulfatases.

Since the availability of genome database is a pre-requisite for reverse proteomics, only two

bacteria E. casseliflavus NCCP-53 and E. coli O83:H1 NRG 857C with accessible

genome/proteome database were studied in this chapter. Based on the results found in

literature and in chapter 2, the hypotheses of this chapter are shown in Figure 4.1.

Figure 4.1 Hypotheses of this chapter. See main texts for more details.

206

The putative bacterial enzymes potentially involved in GSL metabolism to be searched

for in this chapter are as follows:

Bacterial GSL-degrading enzymes: From chapter 2, intact GSLs were metabolized

to ITC and/or NIT products by bacteria isolated from human faecal sample suggesting the

existence of bacterial GSL-degrading activity (a.k.a myrosinase-like activity) in these human gut

bacteria like in plant and aphid. Therefore, in this chapter, the putative bacterial GSL-degrading

enzyme was searched for using the sequence of the well-characterized myrosinase from B.

brassicae (cabbage aphid) as a reference. Aphid myrosinase was successfully characterized

using sequence alignment with white mustard myrosinase (Jones et al., 2002). Aphid

myrosinase has significant sequence similarity (35%) to plant myrosinases and other members

of glycosyl hydrolase family 1 (GH1) (Jones et al., 2002). Based on sequence similarity and

phylogenetic   techniques,   aphid   myrosinase   appears   to   be   more   similar   to   animal   β-O-

glucosidases than to plant myrosinases. Aphid myrosinase is most similar   to   insect   β-O-

glucosidases from the mosquito Anopheles gamblae (47%), the fruit fly Drosophila

melanogaster (45%), and the cockroach Leucophaea maderae (48%). These results strongly

suggest that myrosinase activity has evolutionarily diverged from  β-O-glucosidases in plants and

animals (Jones et al.,  2002).  Aphid  myrosinase  is  also  similar  to  various  bacterial  β-glucosidases

Bacillius halodurans (37%), Clostridium acetobutylicum (35%), Thermoanaerobacter

tengcongensis (35%), and 6-phospho-β-glucosidase from E. coli (29%).

Myrosinase  is  a  type  of  β-glucosidases.  β-glucosidases (EC 3.2.1.21) catalyze the

hydrolysis of the glucosidic linkage of aryl- and alkyl-β-glucosides, and liberate terminal non-

reducing   β-glucosyl residues from oligosaccharides. Based on amino acid sequence and

structural   similarity,   known   β-glucosidases have been placed in Glycoside Hydrolase (GH)

Families GH1, GH3, GH5, GH9, GH30, and GH116 (Ketudat Cairns & Esen, 2010; Cantarel et al.,

2009; Sansenya et al., 2011). Therefore, one can expect bacterial GSL-degrading enzyme or

myrosinas-like enzyme to fall into one of these six GH families. At this writing, all characterized

myrosinases in plants and aphid come from GH1 family only. There is no report of myrosinase

that comes from other GH family. However, BLASTp results showed a few GH3 enzymes that

have certain degree of sequence similarity to aphid myrosinase (Table 4.10).

207

Bacterial sulfatases: From chapter 2, DS-GSL was found as a precursor to NIT

production in bacterial fermentation whereas intact GSL was a precursor to both ITC and NIT

production. This raised the question whether these bacteria may exhibit sulfatase activity that

transform GSL to DS-GSL by desulfation like in H. pomatia (Roman snail) and also they may

exhibit β-glucosidases that can transform DS-GSL to NIT products. In this chapter, the putative

bacterial sulfatase was searched for using the sequence of sulfatase from H. pomatia as a

reference.

4.1.1 Reverse proteomics

In classical proteomics, the starting materials are isolated proteins which are analyzed

via MS techniques, and proteins are identified using complete genome sequences. In contrast,

in reverse proteomics, the starting point is the genome sequence of an organism. First, the

transcriptome and proteome are predicted in silico and subsequently this information is used to

generate reagents for gene functional analysis. This reverse proteomics strategy includes

several steps such as cDNA cloning, protein expression (several systems from E. coli, S.

cerevisiae to mammals are available) and specific functional assays based on gene function

annotations (Figure 4.2). When reverse proteomics is applied downstream of a forward

proteomics pipeline, the experimental design will be focused on the genome subset

corresponding to the previously identified gene products.

208

Figure 4.2 Scheme of the reverse proteomics workflow. This figure was taken from Palcy & Chevet

(2006).

4.1.1.1 Molecular cloning

Molecular cloning is used to assemble recombinant DNA molecules and to direct their

replication within host organisms (Watson, 2007). The method entails the replication of a single

DNA molecule in a single cell as a starting point to generate a large population of cells

containing identical DNA molecules. The steps of molecular cloning are described as follows:

(1) Choosing cloning vector and host organism

E. coli and plasmid vectors are commonly used as they are versatile, technically

sophisticated, widely available, and allow rapid growth of recombinant organisms with minimal

equipment (Brown, 2006). The vector used has to contain four DNA segments that are of

209

significance to its function and experimental utility, (i) an origin of DNA replication is crucial for

the recombinant sequences adjacent to the vector to replicate inside the host organism, (ii) at

least one unique restriction endonuclease recognition sites are present where foreign DNA may

be inserted, (iii) a selectable genetic marker gene that enables the survival of cells that have

taken up vector, and (iv) an additional gene to be used for screening cells harbouring foreign

DNA (Brown, 2006).

(2) Vector DNA preparation

Restriction endonuclease is used to cleave the cloning vector at the specific restriction

site(s) where foreign DNA will be inserted. To generate compatible ends, the vector DNA and

foreign DNA are cleaved with the same restriction enzyme. Most modern vectors e.g. pGEM

(Promega) contain a variety of unique restriction sites that are located within a gene

(frequently β-galactosidase). This gene inactivation by foreign DNA insertion can be used to

distinguish recombinant from non-recombinant organisms during the screening process (Russell

& Sambrook, 2001).

(3) DNA insert preparation

The DNA to be cloned is extracted from the genomic DNA of an organism of interest.

Contaminating proteins, RNA (ribonuclease) and smaller molecules are removed from the DNA

prior to DNA amplification by polymerase chain reaction (PCR). DNA for cloning experiments

may also be obtained from complementary DNA (cDNA) cloning via RNA starting material using

reverse transcriptase, or in the form of synthetic DNA. The purified DNA is then digested with a

restriction enzyme to produce fragments with ends compatible with those of the vector (Russell

& Sambrook, 2001).

(4) Generation of recombinant DNA with DNA ligase

The vector DNA and foreign DNA insert are simply mixed together at appropriate

concentrations with DNA ligase that covalently links the ends together which is termed

‘ligation’.

210

(5) Introduction of recombinant DNA into the host organism

Various methods are used to introduce recombinant DNA into cells e.g. transformation,

transduction, transfection and electroporation (Howe, 2007). Competent host cells with a

physiological state that can take up DNA are introduced with recombinant DNA using either

above method. Competent cells are usually prepared through a special growth regime and

chemical treatment process depending on the specific species and cell types that are used.

Once cells have taken up and replicated DNA from their local environment, the process is

termed transformation (Lederberg et al., 1994).

(6) Selection of organisms containing vector sequences

The introduction of recombinant DNA into the host organism is usually a low efficiency

process. When host organisms are bacterial cells, the selection marker is usually a gene that

confers resistance to certain antibiotics e.g. ampicillin and kanamycin that would otherwise kill

the cells. Upon addition of antibiotics, cells harboring the vector will survive however those that

did not take up vector sequences will die (Brown, 2006).

(7) Screening for clones habouring desired DNA inserts and biological properties

Colonies of transformed cells can be distinguished from those containing the parental

vector (i.e. vector DNA with no recombinant sequence inserted) by using the blue-white

screening system. This system is enabled by the use of modern bacterial cloning vectors e.g.

pUC19 (Promega, UK) and more recent derivatives including the pGEM vectors (Howe, 2007).

Foreign DNA is introduced into a sequence encoding an essential part of β-galactosidase in

these vectors. This enzyme activity leads to formation of a blue-coloured colony on the agar

plate. Insertion of the foreign DNA into the β-galactosidase coding sequence disables the

function of the enzyme, so that colonies (clones) containing recombinant plasmids remain

colorless (white). A very broad range of experimental methods, including the use of polymerase

chain reaction, antibody probes, restriction fragment analysis, nucleic hybridization and/or DNA

sequencing can be used to confirm the presence of desired DNA construct in a number of

different clone (Howe, 2007).

211

4.1.1.2 Recombinant protein expression

Since pET28b+ vector was used in this work, only pET expression system was reviewed

as follows. The pET expression system makes use of multiple cloning sites for the insertion of

different fusion partners, hybrid promoters and restriction sites, along with a high number of

genetic backgrounds modified for various expression purposes (Dubendorf & Studier, 1991;

Studier et al., 1990). Expression requires a BL21(DE3) host strain lysogenized by a DE3 phage

fragment encoding the T7 RNA polymerase (bacteriophage T7 gene 1) under the control of the

isopropyl   β-D-1-thiogalactopyranoside (IPTG) inducible lacUV5 promoter (Sørensen &

Mortensen, 2005) (Figure 4.3). Both lacUV5 promoter and the T7/lac hybrid promoter encoded

by the pET expression plasmid are repressed by LacI repressor. The lacI gene has a copy in the E.

coli genome and in the pET plasmid. During IPTG induction, tetrameric LacI is released from the

lac operator upon IPTG binding, and T7 RNA polymerase is transcribed. As little as 50–100  μM  

IPTG is usually sufficient to achieve full induction.

Figure 4.3 Recombinant expression mechanisms in pET expression system. A general pET plasmid configuration is shown on the right and a genomic configuration of BL21(DE3) host is shown on the left. This figure was taken from http://life.nthu.edu.tw/~b871614/protocols/pet.html.

212

Transcription of the target gene from the T7/lac hybrid promoter (also repressed by

LacI) on the pET plasmid is subsequently initiated by T7 RNA polymerase produced by

BL21(DE3) host (Sørensen & Mortensen, 2005). The T7 promoter with a 20-nucleotide sequence

is not recognized by the E. coli RNA polymerase. T7 RNA polymerase transcribes five times

faster than E. coli RNA polymerase (50 nucleotides per second) at the maximum rate of 230

nucleotides per second. This system leads to the synthesis of large amounts of mRNA, and, in

most cases, the concomitant accumulation of the desired protein at very high concentrations

(40–50% of the total cell protein) (Baneyx, 1999).

4.1.1.3 Enzyme activity and assay

Enzyme activity is measured as the quantity of active enzyme present under defined

conditions (Passonneau & Lowry, 1993). It is performed in vitro under conditions that often do

not closely resemble those in vivo. The conditions used should be at the optimum pH,

‘saturating’   substrate   concentrations,   and   at   the   optimum   temperature.   The   factors   e.g.  

substrate concentrations(s), pH, ionic strength and nature of salts present, and temperature

can affect the activity of an enzyme. In some cases, the opposite direction to that of the

enzyme’s  natural  function  is  measured  for  the  enzyme  activity.  A  complete   in vitro study of the

parameters affecting enzyme activity should enable one to extrapolate to the activity expected

to be occurring in vivo (Scopes, 1993).

Typically, enzyme assays measure either the appearance of product or the

disappearance of substrate over time. Several methods have been developed to determine the

concentration of substrates or products in a reaction. However, all enzyme assays can be

categorized into two types: (i) discontinuous and (ii) continuous assays.

(i) The discontinuous assay measures enzyme concentration in fixed periods of time. The

amount of substrate consumption or product production is determined from samples removed

from an enzyme reaction at intervals (Bergmeyer, 1974). Methods for stopping the reaction

include those which denature the enzyme, such as a strong acid, alkali or detergent, heat, or

treatments with irreversible inhibitors such as heavy metal ions. In some cases, the enzyme can

213

be stopped by the addition of a complexing agent such as ethylenediaminetetraacetic acid

(EDTA), which chelates or removes metal ions essential for activity. Stopped assays should be

checked at least once with varying times of incubation to ensure that the rate is linear through

the period selected for the standard method. Examples of discontinuous assays include

radiometric assays that measure the incorporation of radioactivity into substrates or its release

from substrates, chromatographic assays that measure product generation by separating the

reaction mixture into its components by HPLC analysis.

(ii) The continuous assay follows the progress of the reaction as it occurs. This method is

much more convenient in that the result is seen immediately, and any deviation of the initial

rate from linearity can be observed. However, not all enzymes have an assay method that can

be observed continuously. The simplest continuous assay is designed to follow the action of the

enzyme itself by changes in absorbance (e.g. NAD(P)H at 340nm with dehydrogenases),

viscosity, pH, fluorescence or one of several other possible physical parameters. In many cases

of hydrolase assays, an artificial substrate which releases a coloured or fluorescent product is

used. Unfortunately, most enzymes do not produce any change in a readily detectable physical

parameter by their activity. This can be overcome using a coupled continuous method in which

the product is acted on further (usually by other enzymes that are added to the mixture) until

an ultimate product is formed which can be observed physically.

4.1.2 Hypotheses

Since very little is known about the enzymes involved in bacterial GSL metabolism, the

aim was to try as many techniques as possible to identify theses key proteins involved in GSL

metabolism such as GSL-degrading enzymes, β-O-glucosidases or sulfatase. In Chapter 3, the

sensitivity of forward proteomics approach may be limited if the protein of interest is low

abundance. Therefore, a complementary reverse proteomics approach was used in this chapter.

The hypothesis is as follows.

The bacterial genes/proteins involved in GSL metabolism can be identified using BLAST

search and can be characterized using molecular biology techniques.

214

4.1.3 Objectives

To test the above hypothesis, objectives were set out as follows;

To use BLAST search in a search for putative bacterial GSL-degrading enzymes or

myrosinase-like enzymes, β-O-glucosidases or sulfatases based on known sequences of

the well-characterized proteins from cabbage aphid and Roman snail.

To clone and express genes of interest and subsequently perform enzyme activity assays

on those recombinant enzymes.

Since the availability of genome database is a pre-requisite for reverse proteomics, only

two bacteria E. casseliflavus NCCP-53 and E. coli O83:H1 NRG 857C with accessible

genome/proteome database were studied in this chapter. The characterization of enzymes of

interest found in this chapter will be investigated in Chapter 5.

215

4.2 Materials and Methods

4.2.1 Sequence alignment and bioinformatic analysis:

Sequence similarity search was performed by BLASTX Network Service (NCBI) (Altschul

et al., 1990) (http://blast.ncbi.nlm.nih.gov/Blast.cgi). Detailed information on the proteins of

interest were retrieved from UniProt (Consortium, 2012) at www.UniProt.org. Multiple

sequence alignment was generated using ClustalW2 (Larkin et al., 2007)(http://www.ebi-

ac.uk/ClustalW). The default colour table for amino acids is shown in Figure 4.4. Amino acids

with similar properties are given similar colours.

Figure 4.4 Colour table referring to the labelling of amino acids (single letter code) used in ClustalW alignments.

The accessory application within ClustalW2 with default parameters (gap open penalty,

10; gap extension penalty, 0.05) (Thompson, 1995) was used to generate the alignments.

4.2.2 Genomic DNA extraction

Bacterial culture (2 mL) in corresponding culture broth (i.e. NB medium for E. coli

O83:H1 NRG 857C and WC medium for E. casseliflavus NCCP-53) was anaerobically grown

overnight at 37°C. Subsequently, the culture was centrifuged at 16,000g for 5 min to give a

bacterial pellet. Unless otherwise stated, an Eppendorf 5415 D centrifuge was used throughout

this work for centrifugation steps of small volumes (up to 2 mL per sample). Genomic DNA was

extracted directly from the bacterial pellet using the Wizard® Genomic DNA Purification Kit

(Promega) as per instructions provided with the kit.

216

4.2.3 Primers

Primers with designated restriction sites for PCR experiments were synthesized by

Sigma-Aldrich primer synthesis service. These primers are listed in Table 4.1.

Table 4.1 Primers used in PCR experiments and their restriction sites are underlined

No. Name* Primer  sequence  (5’-3’) Restriction site

1 EC_bgl-F GGTTTGCCATATGTTTCACACAAACT NdeI 2 EC_bgl-R AGGGAGCTCTCATGTTTCACTTGTC SacI 3 EC_GH3#1-F GGTCATATGGAACAGCAGAAATTAACCGA NdeI 4 EC_GH3#1-R GTTGAGCTCTTACCTAACTAATTGCAGGG SacI

5 EC_GH1-F GGTTTGCCATATGGATCATAAACAACT NdeI 6 EC_GH1-R GTTGAGCTCCTAGCACTCTTGC SacI 7 EC_GH3#2-F GGTCATATGAAAAATCAAACACTGGTA NdeI 8 EC_GH3#2-R GTTGAGCTCTTACGTTCGACTGCC SacI 9 EC_GH3#3-F GGTGCTAGCATGAAAAATCAAACACTGG NheI

10 EC_GH3#3-R GTTGAGCTCTCATAGAAGTTCGAAAGTCG SacI 11 EC_6pbg1-F GGTTTGCCATATGTACATGCTTAAATTACC NdeI

12 EC_6pbg1-R CCAGAGCTCTTAAATAATCGTTTTGGTT SacI 13 EC_pBgl-F GGTCATATGGAGAAGCATATGATTGAG NdeI

14 EC_pBgl-R GTTGAGCTCTCATTCTTTTGCTCCTTT SacI 15 EC_SUL1-F GGTCATATGAAAAAAAATAAAGTATCCACC NdeI 16 EC_SUL1-R GTTGAGCTCTTATTCTCCGCTATCTTG SacI 17 ECO_SUL2-F GGTCATATGAAACGCCCCAATTTTCT NdeI 18 ECO_SUL2-R GTTGAGCTCTCAGAACTTCTGTTTTTTCT SacI 19 ECO_6pbg2-F GGTGCTAGCATGAGCCAGAAATTA NdeI 20 ECO_6pbg2-R GTTGAGCTCATGTGCTTTTTTAAGC SacI

*Referred to Table 4.10 for gene annotations. EC, E. casseliflavus NCCP-53; ECO, E. coli O83:H1 NRG

857C

4.2.4 Bacterial strains and plasmids

E. coli DH5α was used for gene cloning and E. coli BL21(DE3) was used for protein

217

expression. Gene inserts of interest were ligated into pET28b(+) vectors. A list of bacterial

strains and plasmids used in this study are shown in Table 4.2.

Table 4.2 Bacterial strains and plasmids used in this study

Strain /plasmid Relevant characteristics Reference

Strains

DH5α F-, endA1, glnV44, thi-1 , recA1, relA1, gyrA96, deoR , nupG , Φ80dlacZΔM15,  Δ(lacZYA-argF)U169, hsdR17(rK

- mK+),  λ– Promega

BL21(DE3) F-, dcm, ompT, hsdS(rB- mB

-), gal  λ(DE3) Stratagene Plasmids pET28b(+) Expression vector, IPTG-inducible, T7 promoter, His-tag, KmR Novagen

pET28b-bgl 1470 bp gene insert from E. casseliflavus ligated into pET28b+ vector This work

pET28b-GH3#1 2151 bp gene insert from E. casseliflavus ligated into pET28b+ vector This work

pET28b-GH1 1437 bp gene insert from E. casseliflavus ligated into pET28b+ vector This work

pET28b-GH3#2 1488 bp gene insert from E. casseliflavus ligated into pET28b+ vector This work

pET28b-GH3#3 2211 bp gene insert from E. casseliflavus ligated into pET28b+ vector This work

pET28b-6pbg1 1407 bp gene insert from E. casseliflavus ligated into pET28b+ vector This work

pET28b-pBgl 2256 bp gene insert from E. casseliflavus ligated into pET28b+ vector This work

pET28b-SUL2 1494 bp gene insert from E. coli O83:H1 ligated into pET28b+ vector This work

pET28b-6pbg2 1353 bp gene insert from E. coli O83:H1 ligated into pET28b+ vector This work

4.2.5 Polymerase chain reaction (PCR)

The genomic DNA was used as a template for gene amplification by Pfu DNA polymerase

(Promega) for 28 cycles. Each PCR reaction was mixed in a sterile, nuclease-free

microcentrifuge tube containing the components as shown in Table 4.3.

218

Table 4.3 Components in one PCR reaction

Components Volume (µL)

Final Concentration

10X Pfu DNA Polymerase Buffer with MgSO4 5.0 1X

10 mM dNTP mix (Promega) 1.0 200 µM

10 µM Forward primer 1.0 0.2 µM

10 µM Reverse primer 1.0 0.2 µM

DNA template 1.0 < 0.5 µg/50 µL

Pfu DNA Polymerase, 2-3 U/µL (Promega) 0.5 1.25 U/50 µL

Nuclease-Free water to final volume of 40.5

Total volume 50.0

The PCR reaction carried out in Eppendorf Thermal Cycler MasterCycler Personal 5332

includes thermal cycling conditions as shown in Table 4.4.

Table 4.4 Thermal cycling conditions for Pfu DNA Polymerase-mediated PCR amplification

Step Temperature (°C) Time Cycle(s)

Initial denaturation 95 1 min 1

Denaturation 95 30 s

28 Annealing 58 30 s

Extension 72 2 min/kb

Final extension 72 5 min 1

Hold 4 Indefinite 1

4.2.6 PCR product purification The PCR products analyzed on a 0.8% agarose gel (Section 4.2.11) were excised using a

sterile blade and subsequently purified according to the manual of QIAquick PCR Purification Kit

(Qiagen). The DNA concentration was determined using a Nanodrop ND-1000

spectrophotometer (Thermo Scientific, UK).

219

4.2.7 Ligation Ligations  were  performed  using  T4  DNA  Ligase   (Promega)  according   to   the  producers’  

specifications. A 3:1 ratio of desired PCR insert to pET28b(+) vector (50 ng) was well-mixed by

pipetting in a microcentrifuge tube and incubated at 4°C overnight. Components in a ligation

mixture (10 µL) are shown in Table 4.5. A map of pET28b(+) vector is shown in Figure 4.5.

Table 4.5 Components in one ligation reaction

Component Volume (µL)

10X T4 Ligase Buffer (Promega) 1.0

T4 Ligase enzyme, 3 U/mL (Promega) 1.0

Cut pET28b+ vector 2.0

PCR product or gene insert 6.0

Total volume 10.0

Figure 4.5 Map of an expression vector, pET28b(+)

220

4.2.8 Bacterial transformation with ligation mixture

A ligation mixture of 50 ng (approximately 2 µL) was added to a 50 µL aliquot of freshly

thawed competent E. coli DH5𝛼 cells. These were incubated on ice for 20 min then at 42°C for

90 s and stored on ice for 2 min. Subsequently, 950 µL of LB was added to suspend the cells

which were incubated in a shaker at 225 rpm, 37°C for 1 h. Following this, cells were

centrifuged at 3,300g for 1 min. The pellet was re-suspended in 200 µL of fresh LB broth and

spread on a selection plate (Section 4.2.9) containing 50 µg/mL kanamycin which allows

bacterial cells harbouring pET28b+ plasmids to grow.

4.2.9 Preparation of competent cells

The competent E. coli DH5α  and  BL21(DE) cells were prepared by diluting an overnight

bacterial culture 1:100 in 200 mL LB. When the OD600 reached 0.4 - 0.5, the cultures were

centrifuged at 3,220g for 20 min at 4°C using Eppendorf 5810 R centrifuge. An Eppendorf 5810

R centrifuge was used throughout this work for centrifugation step of large volume (up to 50

mL per sample). The resulting bacterial pellet was gently re-suspended in 5 mL of ice-cold 100

mM CaCl2. The suspension was kept on ice for 15 min followed by centrifugation at 3,220g for 5

min at 4°C. The bacterial pellet was gently re-suspended in a 1 mL mixture containing 700 µL of

cold 100 mM CaCl2 and 300 µL of 50% sterile glycerol stock (to make a final concentration of

15% glycerol). A 50 µL aliquot of cells was dispensed into a pre-cooled 1.5 mL Eppendorf tube

for each transformation and frozen at - 80°C. The competent BL21(DE3) cells were made the

same way without adding any antibiotics.

4.2.10 Selection of transformants To make selection plates, 15 g agar (Oxoid, UK) was added to 1 L of Luria-Bertani (LB)

medium (In 1 L: 10 g tryptone (Oxoid, UK), 5 g yeast extract (Oxoid, UK), 10 g NaCl (Sigma-

Aldrich, UK)) and was well-mixed before being autoclaved at 121°C for 15 min. The medium was

allowed to cool to 50°C before adding kanamycin (Sigma-Aldrich, UK) to a final concentration of

50  μg/mL.  The  transformant colonies harbouring recombinant gene-pET28b plasmids grown on

221

the selection plates were selected. Subsequently, the selected colonies were subjected to a

colony PCR experiment to determine whether these colonies were of the right size for the gene

fragment of interest.

4.2.11 Colony PCR experiment This method was designed to quickly screen for plasmid inserts directly from

transformant colonies. To each PCR tube containing the PCR reaction (total volume of 25 µL) a

small amount of the colony was added as well as other components (Table 4.6). To do this, a

fine yellow pipette tip was used to touch the edge of the colony and the reaction was pipetted

up and down to mix. For colony PCR, Taq polymerase already included in 5X Crimson Tag

Reaction Buffer (Table 4.6) was used instead of Pfu polymerase. The vector forward primer, T7

promoter primer with the sequence   5’   TAATACGACTCACTATAGGG 3’   and   the   gene-specific

reverse primer (Table 4.1) were used to amplify both the gene region on the pET28b+ vector

and the gene of interest from the selected colonies in this experiment.

Table 4.6 Components in one colony PCR reaction

Component Volume (µL) Final Concentration

5X Crimson Taq Reaction Buffer (Biolabs, UK) 5.000 1X 10 mM dNTP (Promega) 0.500 200 µM

10 µM Forward Primer 0.500 0.2 µM

10 µM Reverse Primer 0.500 0.2 µM

Template DNA from a colony See text <1,000 ng

Crimson Taq DNA Polymerase (Biolabs, UK) 0.125 1.25 U/50 µL PCR

Nuclease-free water 18.375

Total volume 25.000

The colony PCR reaction carried out in Eppendorf Thermal Cycler MasterCycler Personal

5332 includes thermal cycling conditions as shown in Table 4.7.

222

Table 4.7 Thermal cycling conditions for Taq DNA Polymerase-mediated colony PCR

amplification

Step Temperature (°C) Time Cycle(s)

Initial denaturation 95 30 s 1

Denaturation 95 30 s

30 Annealing 56 1 min

Extension 68 1 min/kb

Final extension 68 5 min 1

Hold 4 Indefinite 1

4.2.12 Restriction enzyme digestion

All restriction enzymes, NdeI, SacI and NheI (Fast Digest) were purchased from

Fermentas (UK). The total reaction volume of 30 µL contained ingredients shown in Table 4.8.

Table 4.8 Ingredients for restriction enzyme digestion

Ingredient Volume (µL) PCR product/pET28b+ vector 20 ddH2O 5 10X Tango buffer 3 NheI/NcoI 1 SacI 1 Total volume 30

4.2.13 Agarose gel electrophoresis

All PCR products and digested DNA fragments were resolved on a 0.8% agarose gel

(Sigma-Aldrich, UK). The gel (40 mL) was stained with 0.5 µL of SYBR® Safe DNA gel stain

(Invitrogen). To assign the molecular weight of the sample, 7 µL of Quick-Load 1 kb DNA ladder

molecular marker (BioLabs, UK) (Figure 4.6) was loaded into a separate well. Gels were

223

visualised by Safe   Imager™   blue   light   transilluminator   (Invitrogen) and gel pictures were

captured by Quantity One® 1-D Analysis Software (Bio-Rad).

Figure 4.6 Quick-Load 1 kb DNA ladder molecular marker (BioLabs, UK) used in this chapter. 4.2.14 Plasmid extraction Transformant colonies were grown in 5 mL LB media supplemented with kanamycin at a

final   concentration   of   50   μg/mL overnight at 37°C. Subsequently, the overnight culture was

centrifuged at 16,100g for 5 min to collect a cell pellet. Plasmids of transformant colony were

isolated from the pellet using a Qiagen Plasmid Miniprep Kit as per manual.

4.2.15 DNA sequencing and sequence analysis

The extracted plasmid was  sent  for  sequencing  to  the  ‘GATC’  sequencing  company (UK)

using the specific gene primer as a forward primer and the pET28b+ T7-TER primer as a reverse

primer. The obtained gene sequences were checked whether they are the same as the known

genes using BLAST search (Altschul et al., 1990)

4.2.16 Recombinant protein expression A pre-culture of transformed cells (5 mL) was grown overnight at 37C in LB broth

supplemented with 50 µg/mL kanamycin. On the following day, the pre-culture was inoculated

224

into 1 L LB broth supplemented with kanamycin as in the start culture. The recombinant

BL21(DE3) cells were anaerobically grown to an OD600 of 0.6 and induced to express the

recombinant  protein  by   the  addition  of  0.5  mM   isopropyl  β-D-1-thiogalactopyranoside (IPTG)

for overnight at 25C at 200 rpm. The cells from the induced liquid culture were centrifuged for

20 min at 3,220 at 4C in an Eppendorf 5810 R centrifuge. The pellets were re-suspended in 10

mL buffer (100 mM Tris-Cl pH 7.0) supplemented with 100 µL protease inhibitor cocktail (100X,

Melford, UK). The cells were lysed through two shots of a 30k psi disruption cycle in a tissue

disrupter (Constant Cell Disruption Systems, UK). The supernatants were recovered after

centrifugation at 16,100g at 4C for 30 min. These cell-free extracts were desalted against

buffer (0.1 M mM Tris-Cl, pH 7.0) on an Econo Pac 10 DG column (Bio-Rad, UK) (Section 4.2.18).

The   protein   concentrations   of   the   eluted   proteins   were   assessed   using   Bradford’s   reagent  

(Sigma, UK) as previously described (Chapter 2, section 2.2.20), and the purity was analyzed by

SDS-PAGE (Chapter 2, section 2.2.11). The desalted supernatant was stored at 4C until enzyme

activity assays.

4.2.17 SDS-PAGE analysis

SDS-PAGE analysis was carried out as previously described (Chapter 2, section 2.2.11).

The protein markers used in this chaper are shown in Figure 4.7.

Figure 4.7 Protein markers. (A) PageRuler pre-stained protein ladders (Thermoscientific, UK), (B) EZ-Run unstained ladders (Fisher, UK), (C) Broad Range pre-stained ladders (BioLabs) and (D) Low Range unstained marker (Sigma-Aldrich).

225

4.2.18 Desalting recombinant enzymes

The appropriate buffer (20 mL of 0.1 M mM Tris-Cl, pH 7.0) was added to the Econo Pac

10 DG column (Bio-Rad, UK) to equilibrate it. The cell-free extract containing the recombianat

enzyme (3 mL) was added to the column, and the flow-through was discarded. The elution

buffer (4 mL of 0.1 M mM Tris-Cl, pH 7.0) was added to the column to elute the higher

molecular weight component(s) and the flow-through   (4  mL)   collected   as   ‘desalted   cell-free

extract’  for  further  analysis.  

4.2.19 GOD-PERID assay

This  assay  was  used  to  determine  glucose  release  upon  GSL  or  β-glucoside breakdown

by  myrosinase  activity  or  β-thioglucosidase/β-O-glucosidase activity, respectively (Bones, 1990).

The GOD-PERID reagent consists of 2,2'-azino-bis(3-ethylbenzthiazoline-6-sulphonic acid) or

ABTS (1 mg/mL), glucose oxidase (8 U/mL) and peroxidase (0.35 U/mL) dissolved in Tris buffer

(1.2 g/100 mL) pH 7.2. To make 250 mL GOD-PERID reagent, 3 g of Tris was dissolved in distilled

water with adjustment to pH 7.2 by using HCl, 12.7 mg of glucose oxidase (157.5 U/mg) was

dissolved in 10 mL of Milli-Q water and then added to the Tris buffer, 4.7 mg peroxidase (148

U/mg) was dissolved in 20 mL Milli-Q water, and a 2.5 mL aliquot was added to Tris buffer, 250

mg ABTS was added and stirred to dissolve, and finally the mixture was made up to 250 mL with

Milli-Q water. This reagent was stored in dark cold place until use. All the chemicals were

purchased from Sigma-Aldrich, UK. GSL was added to GOD-PERID reagent as a substrate for

myrosinase, and if present, the breakdown of GSL would produce AITC and/or NIT and D-

glucose. This D-glucose acts as a substrate for glucose oxidase in the GOD-PERID reagent which

in turn leads to the formation of a green, soluble end-product catalyzed by peroxidase. The

assay scheme is shown in Figure 4.8. The green dye absorbance maximum of 420 nm light (ε =

3.6 × 104 M–1 cm–1) can easily be followed with a spectrophotometer (Shin & Lee, 2000), and

therefore one can easily determine whether an enzyme has myrosinase activity.

226

Figure 4.8 GOD-PERID assay reaction principle. (1) GSL is hydrolyzed by myrosinase to produce an unstable aglucone and D-glucose. (2) D-glucose is converted by glucose oxidase to form D-glucono-1,4-lactone and also produced hydrogen peroxide (H2O2). (3) Peroxidise, with the help of hydrogen peroxide, converts 2,2'-azino-bis(3-ethylbenzthiazoline-6-sulphonic acid) or ABTS into a dark green and soluble end-product.

The reaction mixture (300 µL) contained protein solution (100 µL, ~ 200 µg), 10 mM

sinigrin (60 µL) and 0.1 M citrate phosphate buffer (140 µL) were mixed and aerobically

incubated at 37°C for 30 min. Subsequently, the mixture was boiled at 100°C for 5 min to

deactivate the enzymes and GOD-PERID solution (1 mL) was added to the mixture and

incubated at 37°C for 15 min. The mixture was then transferred to the cuvette, and the

absorbance was measured at 420 nm (A420nm) using LKB Novaspec II spectrophotometer

(Pharmacia, UK). Glucose release from the reaction can be determined by using a calibration

curve of known glucose amounts versus absorbance at 420 nm (Figure 4.9).

227

Figure 4.9 Calibration curve for GOD-PERID assay. A graph of A420nm versus various amounts of glucose was plotted. Values are means of triplicates.

One unit of myrosinase activity was defined as the amount of enzyme liberating 1 µmol

of glucose per min. Total activity and specific activity of enzyme can be determined from the

following formulae:

Total activity = Average of A420nm x Dilution factor x F Mr of glucose x 30 min

= X µmol glucose/min

(where F = gradient from calibration curve (i.e. extinction coefficient), Mr = molecular weight)

Specific activity = Total activity Total protein

= Y µmol glucose/min/mg

4.2.20 Substrates used in GOD-PERID assay

In addition to GSL, other substrates (with α- or β- glycosidic bonds) were also used in

GOD-PERID assay. The substrates including cellobiose, trehalose dihydrate, salicin, and methyl

β-D-glucopyranoside were dissolved in Milli-Q water to make up 10 mM solution stocks. All

228

chemicals were purchased from Sigma-Aldrich, UK. Cellulose is a disaccharide consisting of two

glucose molecules   linked  by  a  β(1→4) bond. It can be hydrolyzed to glucose enzymatically or

with acid. Trehalose is a natural alpha-linked disaccharide formed  by  an  α,α-1,1-glucosidic bond

between  two  α-glucose units. Salicin is an alcoholic β-glucoside.  Methyl  β-D-glucopyranoside is

a synthetic substrate. The GOD-PERID assays using these substrates were prepared as section

4.2.17. The structures of these compounds are shown in Figure 4.10.

Figure 4.10 Structures of substrates used in GOD-PERID assay. (A) Cellobiose. (B) Trehalose. (C) Salicin. (D) Methyl  β-D-glucopyranoside.

4.2.21 β-O-glucosidase activity assay

The β-O-glucosidase activity was assayed by measuring the increase in absorbance at

400 nm due to the release of yellow p-nitrophenol (pNP) from colorless p-nitrophenyl-β-D-

glucopyranoside (pNPG) substrate (Figure 4.11).

Figure 4.11 β-O-glucosidase assay reaction principle. The colorless substrate, p-nitrophenyl-β-D-glucopyranoside (pNPG),   is  hydrolyzed  by  β-O-glucosidase to produce D-glucose and a yellow-coloured p-nitrophenol (pNP) product.

229

The reaction mixture (1 mL) contained 10 µL of protein solution (~ 20 µg), 1 mM pNPG

(100 µL of 10 mM pNPG stock solution) and 100 mM citrate phosphate pH 7.0 (890 µL). After

incubation at 37°C for 5 min, the reaction was stopped by adding 5 mL of 0.1 M NaOH. The

absorbance was measured at 400 nm (A400nm) by LKB Novaspec II spectrophotometer

(Pharmacia, UK). The amount of pNP product from the reaction can be determined by using a

calibration curve of known pNP amounts versus A400nm (Figure 4.12). To make a calibration

curve, different dilutions of pNP ranging from 0.02 µmol to 0.6 µmol from a 10 mM pNP stock

solution dissolved in 100 mM citrate phosphate pH 7.0 were prepared in 1 mL and 5 mL of 0.1

M NaOH was added. The corresponding A400nm values of the total volume of 6 mL reaction

mixtures were plotted against those dilutions. All chemicals were purchased from Sigma-

Aldrich, UK. One unit of β-O-glucosidase activity was defined as the amount of enzyme

liberating 1 µmol of pNP per min. Total activity and specific activity of the enzyme can be

determined in a similar manner as shown in Section 4.2.18.

Figure 4.12 Calibration   curve   for   β-O-glucosidase activity assay. A graph of A400nm versus various amounts of pNP was plotted. Values are means of triplicates.

230

4.2.22 Arylsulfatase activity assay

Arylsulfatase activity was assayed by the method of Roy (1953). The arylsulfatase

activity was assayed by measuring the increase in absorbance at 510 nm (A510nm) due to the

release of p-nitrocatechol (pNC) from p-nitrocatechol sulfate (pNCS) substrate. The assay

scheme is shown in Figure 4.13.

Figure 4.13 Sulfatase assay reaction principle. The yellow substrate, p-nitrocatechol sulfate dipotassium (pNCS), is hydrolyzed by sulfatase to produce sulfate and a red-coloured p-nitrocatechol product (pNC) which was visible seen upon 0.2 M NaOH addition.

The reaction mixture (1 mL) contained 50 mM sodium acetate buffer pH 5.0 (890 μL), 1

mM pNCS (100 μL of 10 mM pNCS stock solution), and protein solution (10 μL, ~ 20 µg). The

mixture was incubated at 37°C for 5 min after which it was added with 5 mL of 0.1 M NaOH to

stop the reaction. The absorbance was measured at 510 nm by LKB Novaspec II

spectrophotometer (Pharmacia, UK). The amount of pNC product from the reaction can be

determined by using a calibration curve of known pNC amounts as a function of absorbance at

510nm (Figure 4.14). To make a calibration curve, different dilutions of pNC ranging from 0.02

µmol to 1 µmol from a 10 mM pNC stock solution dissolved in 50 mM sodium acetate buffer

(pH 5.0) were prepared in 1 mL and prepared in 1 mL and 5 mL of 0.1 M NaOH was added. The

corresponding A400nm values of the total volume of 6 mL reaction mixtures. The

corresponding A400nm values of the total volume of 6 mL reaction mixtures were plotted

against those dilutions. All chemicals were purchased from Sigma-Aldrich, UK. One unit of

arylsulfatase activity was defined as the amount of enzyme liberating 1 µmol of pNC per min.

231

Total activity and specific activity of the enzyme can be determined in a similar manner as

shown in Section 4.2.18.

Figure 4.14 Calibration curve for arylsulfatase activity assay. A graph of A510nm versus various amounts of pNC was plotted. Values are means of triplicates.

232

4.3 Results

4.3.1 BLAST searches and sequence analysis of putative bacterial GSL-degrading

enzymes/sulfatases

In this chapter, the aim was to identify putative bacterial βaglucosidases with GSL-

degrading activity and bacterial sulfatases from the isolated GSL-degrading bacterial strains that

have accessible genome/proteome database (see Figure 4.1 for hypotheses). Thus, bacterial

proteins with some degrees of similarity to the existing well-characterised aphid myrosinase

from B. brassicae (UniProt accession no. Q95X01) were searched for. This is because aphid

myrosinase is assumingly more closely related to bacterial myrosinases than plant myrosinase

(Jones et al., 2002). In a search for bacterial sulfatases, the snail sulfatase from H. pomatia

(UniProt accession no. Q9NJU7) was used as a reference sequence. The genome/proteome

database of E. coli O83:H1 NRG 857C is available, but that of E. casseliflavus NCCP-53 is not.

However, the genome/proteome database of the close relative bacteria E. casseliflavus strain

EC10/EC20/EC30/ATCC 12755 is accessible. According to UniProt database, there are eleven β

leven ng to Unin each E. casseliflavus strain EC10/EC20/EC30/ATCC 12755 and six βix ain

EC10/EC20/E. coli O83:H1 str. NRG 857C (Table 4.9).

Table 4.9 Information on genome/proteome and the number of proteins of interest from the bacteria under study

Bacterium Size (Mb) GC % Gene no.

Protein no.

Bgl no. GH no. SUL

no.

E. coli O83:H1 NRG 857C 4.89 50.7 4690 4582 6 40 15

E. casseliflavus EC20* 3.42 42.5 3341 3292 11 44 4 *This strain was used as a representative of E. casseliflavus NCCP-53 GH = Glycosyl Hydrolase family; Bgl = β-glucosidases; SUL = Sulfatase

There are seven candidate GSL-degrading enzymes (25-50% similarity to aphid

myrosinase with different Max Scores) and one candidate sulfatase enzyme (25% similarity to

233

snail sulfatase) of E. casseliflavus EC10/EC20/EC30 to be cloned from genomic DNA of E.

casseliflavus NCCP-53 (Table 4.10). Also, there are one β-glucosidase protein (36%) and one

putative sulfatase protein (38%) from E. coli O83:H1 str. NRG 857C to be cloned (Table 4.10).

Table 4.10 List of putative bacterial GSL-degrading enzymes/sulfatases with high similarity to aphid myrosinase and snail sulfatase

UniProt assession no. Gene name Assigned

name Gene family

Gene length

(bp)

Protein size

(kDa) pI

% Sequence identity*

E. casseliflavus EC10/EC20/EC30

C9CJJ3 Glycoside Hydrolase GH3#1 GH3 2211 81 5.51 50 (16.9)

C9AZA4 Periplasmic  β-glucosidase pBgl GH3 2256 83 5.93 33 (18.1)

C9AY94 Glycoside Hydrolase GH3#3 GH3 2151 79 4.73 32 (17.7)

C9AW70 Glycoside Hydrolase GH3#2 GH3 1488 54 5.15 25 (12.3)

C9ABS9 β-glucosidase bgl GH1 1470 56 5.21 34 (244)

C9AXB6 Glycoside Hydrolase GH1 GH1 1437 55 5.18 33 (234)

C9AZJ8 6-phospho-β-galactosidase 6pbg1 GH1 1407 54 5.18 30 (172)

C9ACP4 Sulfatase SUL1 SUL 2124 81 4.83 25 (31)

E. coli O83:H1 NRG 857C

E4P7X8 6-phospho-β-glucosidase 6pbg2 GH4 1353 50 5.90 36 (20)

E4P283 Sulfatase/phosphatase YidJ SUL2 SUL 1494 57 5.11 38 (74)

*Sequence identity (%) was from BLASTp search. Numbers in brackets are Max Score.

The genes chosen to be cloned from the two bacteria (Table 4.10) have met the

selection criteria; (i) Based on BLASTp searches, these candidate proteins showed at least 25%

sequence similarity to aphid myrosinase or snail sulfatase, (ii) The chosen proteins are ranked

within top 10 according to Max score (BLASTp search) from each GH1, GH3 and GH4 family that

have similar sequences to aphid myrosinase. Thus, other enzymes e.g. periplasmic   β-

glucosidase and 6-phospho-β-glucosidase from GH3 and GH4 families, respectively were also

selected for cloning although they are not GH1 enzymes and have lower Max scores on BLASTp

search when compared with other GH1 enzymes. The rationale was that the putative bacterial

234

GSL-degrading enzymes may come from any GH families than GH1 or they may be periplasmic

enzyme or 6-phospho-β-glucosidase.

Multiple sequence alignments among the above three putative bacterial β-glucosidases

from GH1 family (i.e. bgl, GH1 and 6pbg1) with the GH1 aphid myrosinase show the well-

conserved catalytic acid/base Glu (E) residues and the catalytic nucleophile Glu (E) residues and

several other conserved residues (Figure 4.15A). Four putative bacterial β-glucosidases from

GH3 family (i.e. GH3#1, GH3#2, GH3#3, pBgl) show high degree of alignment similarity among

one another containing the well-conserved catalytic nucleophile Asp (D) residues in place of Glu

(Figure 4.15B), but the second acid/base residues are not easily identified due to high variability.

However, their alignments (of GH3 enzymes) with the GH1 aphid myrosianse did not show

many conserved residues with the GH1 aphid myrosinase. Likewise, the alignments of the GH4

enzyme (i.e. 6pbg2) with the GH1 aphid myrosinase show less conserved residues with the GH1

aphid myrosinase when compared with other GH1 enzymes. The active residues of GH4 enzyme

include Tyr (Y) and Arg (R) (Figure 4.15C).

A GH1

bgl MMFHTNLDPFPENFLWGAASAAYQIEGAWAEDGKGPSIWDTYAQIPGNTFEE-TNGKVAI 59 GH1 -MDHKQLKEFPNDFLWGSASAAYQVEGAWQEDGKGASVWDDFVRIPGKTFKA-TNGDVAV 58 6pbg1 ----MYMLKLPEDFIFGGATAAYQVEGATKEGGKGAVAWDDFLEEQGR-----FSPDPAS 51 Aphid -----MDYKFPKDFMFGTSTASYQIEGGWNEDGKGENIWDRLVHTSPEVIKDGTNGDIAC 55 :*::*::* ::*:**:**. *.*** ** . . . . * bgl DHYHRYKEDVALMKQMGLKGYRFSVAWSRILPDGEG-AVNEAGVAFYEKLVDELLRQGVE 118 GH1 DHYHRFKEDVALMKEQGLKTYRFSIAWTRIFPEGRG-EVNQAGLDFYLALIDELIKAGIE 117 6pbg1 DFYHQYAKDIELCERFGVNGLRLSIAWSRIFPDGDG-EPNPEGIAFYHRVFEECAKRNVT 110 Aphid DSYHKYKEDVAIIKDLNLKFYRFSISWARIAPSGVMNSLEPKGIAYYNNLINELIKNDII 115 * **:: :*: : : .:: *:*::*:** *.* : *: :* :.:* : .: bgl PILTLYHWDLPQALQDKYLGWEGRETAEAFERYCRILFERLGKKVTYWVTMNEQNVFTSL 178 GH1 PMVTLYHWDLPRALQEEYGGWESRKIIEDFTNYAAVLFEAFRGKVHYWVSLNEQNIFTSL 177 6pbg1 PFVTLHHFDTPKRLFDQ-GDFLNRETIEAFVSYAIFCFHEF-KEVKVWSTFNEIYPVATN 168 Aphid PLVTMYHWDLPQYLQDL-GGWVNPIMSDYFKEYARVLFTYFGDRVKWWITFNEP-IAVCK 173 *::*::*:* *: * : .: . : * *. . * : .* * ::** . bgl GYRWAAHPPGLK-DLKRMYAANHIINLANAKAINLFHELVPQGKIG-PSFGYGPMYPFSC 236 GH1 GYLLAAHPPGVT-DPKRMYEVNHIANLANASVINKFHEMKIPGKIG-PSFAYSPNYPINS 235 6pbg1 QYLLGVFPPGIKYDFTKIIACLHNMMVAHARVVNYFKENELPGEIG-VVHSLETKYAATD 227 Aphid GYSIKAYAPNLNLKTTGHYLAGHTQLIAHGKAYRLYEEMFKPTQNGKISISISGVFFMPK 233 * ...*.:. . . * :*:. . . :.* : * . : bgl DPE---DVLAAENGEAFNNAWFLDVYCKGEYPKFVYKQLAKVGLAPEVT-----PEDQAL 288 GH1 DPK---NILAAENAEDLMAHYWLDVYLWGEYPIAAMNYLKEQGIAPTIE-----PGDMDL 287 6pbg1 APE---DKHAAFLDDALSIRFLLDATYLGYYSTETLTALDEICEANQASY-HFPEEDFVE 283

235

Aphid NAESDDDIETAERANQFERGWFGHPVYKGDYPPIMKKWVDQKSKEEGLPWSKLPKFTKDE 293 .: : :* : : : . * *. . : : bgl LKQAKP--DFLGINYYHGGTAQQNNLQKQSAEKKEFSKVDPYLMQAAAGEFSPEETMFAT 346 GH1 LRSAKP--DFLGINYYQTATNAYNPLDGVGAGKMNTTGKK------GSSEETGTPGMFKK 339 6pbg1 LKKASTRNDYLGINHYQ--CHFVKAYDGENAIHHNGTGEK-------GTSVYKVKGIGER 334 Aphid IKLLKGTADFYALNHYSSRLVTFGSDPN--------------------PNFNPDASYVTS 333 :: . *: .:*:* . bgl AENPHLKKTDWGWEID-PVGFRVALRRIQANYDLP--IFITENGLGAIDQLTEDKQIHDP 403 GH1 AENPFVERTNWDWEID-PQGLRIALRRITSRYRVP--VIITENGLGEYDKLTDDHQIHDQ 396 6pbg1 IYKEGIPRTDWDWLIY-PEGLYDLLLRIKSDYPHYNKIYITENGMGYKDQFEDG-IIMDQ 392 Aphid VDEAWLKPNETPYIIPVPEGLRKLLIWLKNEYGNP-QLLITENGYG------DDGQLDDF 386 : : .: : * * *: * : * : ***** * :. : * bgl YRITYLQEHLVELQKAITDG-VELIGYCAWSFTDLLSWLNGYKKRYGFVYVDRDNQSERQ 462 GH1 YRIDYLAGHVHAIKEAISDG-AEVLGYCTWSFTDLLSWLNGYQKRYGFVYVDQDETQEGS 455 6pbg1 PRIDYLRVYLESLSKAITAG-VNVKGYFLWSLMDLFSWTNGYNKRYGLFYVDFETQK--- 448 Aphid EKISYLKNYLNATLQAMYEDKCNVIGYTVWSLLDNFEWFYGYSIHFGLVKIDFNDPQR-- 444 :* ** :: :*: . :: ** **: * :.* **. ::*:. :* : . bgl LARIPKDSFYWYQEVIRTNGASLTSET 489 GH1 LARYKKDSFYWYQELIKTNGQEC---- 478 6pbg1 --RYPKESAYWYKLVSETKTII----- 468 Aphid -TRTKRESYTYFKNVVSTGKP------ 464 * ::* ::: : *

B GH3 GH3#2 MKNQTLVQLVNQLTLDEKIGQLVQ-LSGEFFHG-SDLSLGPQQKLGIEQQTIDVVGSVLN 58 GH3#3 MKNQTLVQLVNQLTLDEKIGQLVQ-LSGEFFHG-SDLSLGPQQKLGIEQQTIDVVGSVLN 58 GH3#1 MEQQKLTELLSEMTLDEKIDQLLQ-LAAAFYSDKAEEKTGPMGDLGLTQENINNAGTTLG 59 pBgl MEKHMIETLLRQMTLKEKIGQLNQRLYGWEVYEKTNGKIMLTETFKKEVARFGSLGWIYG 60 Aphid --------MDYKFPKDFMFGTSTASYQIEGGWNEDGKGENIWDRLVHTSPEVIKDGTNGD 52 : ::. . :. : . * . GH3#2 VTGAQ------------------VTRKIQTDYLRKSRHKIPLLFMADIIYGYR----TVF 96 GH3#3 VTGAQ------------------VTRKIQTDYLRKSRHKIPLLFMADIIYGYR----TVF 96 GH3#1 VSGAK------------------EAIRVQKEYIENNRLNIPTILMADIIHGFR----TIF 97 pBgl VFRADPWSGRNQQTGLTTAESYELSLMIQTYLQEHTRLGIPAFLSEECPHGHQGLEATTF 120 Aphid IACDS-----------------------------YHKYKEDVAIIKDLNLKFYR-----F 78 : . : : : . * GH3#2 PIPLGLGATWNPALIQSAYQAAAQEARAAGAHVTYAPMVDLVRDARWGRCLESTGEDPLL 156 GH3#3 PIPLGLGATWNPALIQSAYQAAAQEARAAGAHVTYAPMVDLVRDARWGRCLESTGEDPLL 156 GH3#1 PIPLGLGSSWDLAAAEKMAEVSAKEAAVSGLHVTFSPMVDLVRDPRWGRVMESTGEDPYL 157 pBgl PVNFSVGSSWNPDLYQAAQTITAQEIRAKGAHVGLVSALDIARDPRWGRTEECFSEDPFL 180 Aphid SISWARIAPSGVMNSLEPKGIAYYNNLIN--ELIKNDIIPLVTMYHWDLPQYLQDLG--- 133 .: . :. . : : .: : :. :*. . . GH3#2 NADFAKAMVEGIQQEKGGTLLG-IAACVKHFAAYGASEGGRDYNTVDMSERKLRQDYLSG 215 GH3#3 NADFAKAMVEGIQQEKGGTLLG-IAACVKHFAAYGASEGGRDYNTVDMSERKLRQDYLSG 215 GH3#1 NSRFAEAFVKGYQGDDLRTDFNRVAACVKHFAAYGAAIGGRDYNTVNMSERQLRESYLPG 217 pBgl TSSFTKAAVRGLQGLKTTIEKQNVLAVLKHFAAQGAGMGGHNAGPVAIGDREFREIHLPP 240 Aphid --GWVNPIMSDYFKEYARVLFTYFGDRVKWWITFNEP--------IAVCKGYSIKAYAPN 183 :.:. : . . :* : : . : : . : : . GH3#2 YKAAVEAGCKLVMTSFNTYDGIPATANQFLIKQILREEWQFDGVVISDYAAVQELIPHGI 275 GH3#3 YKAAVEAGCKLVMTSFNTYDGIPATANQFLIKQILREEWQFDGVVISDYAAVQELIPHGI 275 GH3#1 YKAALDAGAKLVMTSFNTVDGIPATANRWLFRDVLREEFGFEGVVISDWAAIKEVIAHGA 277 pBgl MKAGIAAGALGCMAAYNDLDGVPCHANAYLLQEVLREESGFAGIVMADGCGLDRIADWLG 300 Aphid LNLKTTGHYLAGHTQLIAHG----KAYRLYEEMFKPTQNGKISISISGVFFMPKNAESDD 239

236

: . : . * . . : .: ::. : . GH3#2 ATDDREAAKLAIEATNDIDMKTRCYAKELRPLLESGAIDQRLIDDAVYHVLKLKKDLGLF 335 GH3#3 ATDDREAAKLAIEATNDIDMKTRCYAKELRPLLESGAIDQRLIDDAVYHVLKLKKDLGLF 335 GH3#1 AEDEKHAAELAIKAGVDIEMMTTCYTDNLKELIAEGTVEEALVDEAVLRILTLKNELGLF 337 pBgl SRS--QAAAKSLTSGVDVSLWDEVFP-VLEEAVLDGLIAETVIDEAVRRVLLLKEKLGLF 357 Aphid DIETAERANQFERGWFGHPVYKGDYPPIMKKWVDQKSKEEGLPWSKLP------------ 287 . . * . . : :. :. : . : : . : GH3#2 EDPFRGSSEEVEAQILLSEENRKLARKVASEAIVLLQNKQEVLPLTPKKEKILLVGPYGD 395 GH3#3 EDPFRGSSEEVEAQILLSEENRKLARKVASEAIVLLQNKQEVLPLTPKKEKILLVGPYGD 395 GH3#1 ENPYRGADEAAEAATVLSQEHREIARDIAKKSMVLLKN-EGVLPLQ-KTEKVAIVGPGAH 395 pBgl KEVTP------HVSLPDKEKARQASLKLAEESVVLLEN-NGILPLKKTRQKIAVIGPHVK 410 Aphid --------------------------KFTKDEIKLLKGTADFYALNHYSSRLVTFGSDPN 321 ..:.. : **:. . .* .:: .*. . GH3#2 NQ-AMIGLWAVHGKTEDVTTLKTALQNTVSEKYVHYEPGCPLLEDDSILGDFGYTASGNS 454 GH3#3 NQ-AMIGLWAVHGKTEDVTTLKTALQNTVSEKYVHYEPGCPLLEDDSILGDFGYTASGNS 454 GH3#1 SR-DLLGAWSWQGKQEEVVTLVAGAQSLGADLLIGQEP-------------FDYFAP--- 438 pBgl QLYHQLGDYTPFKEEAMCMTLWEGLTKLNTHQVDFAYEKGCEIANGTTAQRRRACQIAED 470 Aphid PN--------FNPDASYVTSVDEAWLKPNETPYIIPVPEG-------------------- 353 . :: . . GH3#2 SSAAQQDLWLKEALKAGTEADIILFAMG-EHSLQSGEAGSRT------------------ 495 GH3#3 SSAAQQDLWLKEALKAGTEADIILFAMG-EHSLQSGEAGSRTDLHLPAVQRAFIKKMTAL 513 GH3#1 SEAA-----IQEAIELVKEADKVVLALG-EQEWMSGEAASRSDIRLPQAQLSLVETLKEY 492 pBgl ADVILVTIGGSSARDFTTDFDKNGAALRGSQEMTSGENIDLATLDLPQCQLDLLFALKKR 530 Aphid ------------------------------------------------LRKLLIWLKNEY 365 GH3#2 ------------------------------------------------------------ GH3#3 GKKNILINFSGRPLVLKEETKQMDAILQAWFPGTEGAQAIVDILFGKVNPSGRLSMSFPE 573 GH3#1 NEQLIVTLYNGRPLDLQG-VDAAKAIVEAWFPGTEGGNALAQILWGEYNPSGRLSMSFPE 551 pBgl KKPLIGIVISGRPHCLAPLKEVFDGLLYAGYPGQYGGEAIARILFGETVPSGKLAVSIPD 590 Aphid GNPQLLITENG-----YGDDGQLDDFEKISYLKNYLNATLQAMYEDKCNVIGYTVWSLLD 420 GH3#2 ------------------------------------------------------------ GH3#3 DVGQLPLYYNHFNTGRPLNSKTHTGRFVSKYLDCSNEPLFPFGYGLSFGEASYHSLKLSD 633 GH3#1 TVGQVPVYYNVDNTGRPYESAPDE-KYVSKYLDVSNYAKYPFGFGLSYSPVAYSTVTLDQ 610 pBgl TVGQLPVCYNYRNT-----------AFQKDYLDQNGTPVYSFGYGLSYASFTCSAVSAEY 639 Aphid NF------------------------------------EWFYGYSIHFG----------- 433 GH3#2 ------------------------------------------------------------ GH3#3 STMN--ETLEAEITIRNNSAYSRLETVQLYIRDHVGSVVRPVKELKKYKKIPLSPYEEIT 691 GH3#1 PTMTKDQTVTASITVTNQGTAAVWETVQCYIRDLVGEVVRPVKELKGFKKIWLEAGESTT 670 pBgl AGEG----IIVSGTLENHSAVSGKEVIQCYLKEYTKAYVPRKKVLCGFQKVWVPNQGQVA 695 Aphid ---------LVKIDFNDPQRTRTKRESYTYFKNVVSTGKP-------------------- 464 GH3#2 -------------------------------------------------------- GH3#3 VLFTITKEDLFYYQKDLTFGVEPGLFTLFIGKNS-----AEGEAATFELL------ 736 GH3#1 VQFEITEELLRYVHSNQQASSDPGKFHIMIGGNS-----RDTQQTTLQLVR----- 716 pBgl FQLVIDEASVQQLAISLKDTASFCLEVETTGQQYRFVFQRSNPDRTWQVTQKGAKE 751 Aphid --------------------------------------------------------

C GH4 Aphid MDYKFPKDFMFGTSTASYQIEGGWNEDGKGENIWDRLVHTSPEVIKDGTNGDIACDSYHK 60

237

6pbg2 MSQKLKVVTIGGGSSYTPELLEGFIKRYHELPVSELWLVD---VEGGKAKLDIIFDLCQR 57 *. *: : * *: : :: *: : : : : : * . :: ** * :: Aphid YKEDVAIIKDLNLKFYRFSISWARIAPSGVMNSLEPKGIAYYNNLINELIKNDIIPLVTM 120 6pbg2 MIDNAGVPMKLYKTLDRR-------------EALKDADFVTTQLRVGQLPARELDERIPL 104 ::..: .* .: * ::*: .:. : :.:* .:: :.: Aphid YHWDLPQYLQDLGGWVN-----PIMSDYFKEYARVLFTYFGDRVKWWITFNEPIAVCKGY 175 6pbg2 SHGYLGQETNGAGGLFKGLRTIPVIFDIVKDVEELCPN------AWVINFTNPAGMVT-E 157 * * * :. ** .: *:: * .*: .: . * *.*.:* .: . Aphid SIKAYAPNLNLKTTGHYLAGHTQLIAHGKAYRLYEEMFKPTQNGKISISISGVFFMPKN- 234 6pbg2 AVYRHTGFKRFIGVCNIPIGMKMFIRDVLMLKDCDDLSIDLFGLNHMVFIKDVLVNGKSR 217 :: :: .: . : * . :* . : ::: . : : *..*:. *. Aphid -AESDDDIETAERANQFERGWFGHPVYKGDYPPIMKKWVDQKSKEEGLPWSKLPKFTKDE 293 6pbg2 FAELLDGVASGQLKASGVKNIFDLPFSEG----LIRSLNLLPCSYLLYYFKQKEMLAIEM 273 ** *.: :.: . :. *. *. :* :::. .. :.: :: : Aphid IKLLKGTADFYALNHYSSRLVTFGSDPNPNFNPDASYVTSVDEAWLKPNETPYIIPVPEG 353 6pbg2 GEYYKGGARAQVVQKVEKQLFELYKNPELKVKPKE--LEQRGGAYYSDAACEVINAIYND 331 : ** * .::: ..:*. : .:*: :.:*. : . . *: . * .: :. Aphid LRKLLIWLKNEYGNPQ------LLITENGYGDDGQLDDFEKISYLKNYLNATLQAMYEDK 407 6pbg2 KQAEHYVNIPHHGHIDNIPADWAVEMTCTLGRDG-ATPHPRITHFDDKVMGLIHTIKGFE 390 : .:*: : : * ** . :*:::.: : . :::: : Aphid CNVIGYTVWSLLDNFEWFYGYS-IHFGLVKIDFNDPQRTRTKRESYTYFKNVVSTGKP-- 464 6pbg2 IAASNAALSGEFNDVLLALNLSPLVHSDRDAELLAREMILAHEKWLPNFADCIAELKKAH 450 . . :: . :::. . * : .. . :: : ::.: . * : :: *

Figure 4.15 Alignments of putative bacterial GSL-degrading enzyme sequences with B. brassicae `Aphid' myrosinase. (A) Alignments of proteins from GH1 family. The catalytic residues, a general acid/base Glu (E) (yellow) and a nucleophile Glu (E) (blue) are highlighted. (B) Alignments of proteins from GH3 family show the conserved catalytic Asp (D) residue (green), but the other active residue is less readily identified and highly variable. (C) Alignment of a protein from GH4 family. The catalytic residues Tyr (Y) (yellow) and Arg (R) (blue) are highlighted. The assigned gene names are referred to Table 4.10. The symbols ":" means that conserved substitutions have been observed and "." means that semi-conserved substitutions are observed. Amino acid color codes are referred to Figure 4.3 (Larkin et al., 2007).

Multiple sequence alignments between the putative bacterial sulfatase SUL2 with snail

sulfatase show the consensus ‘CXPXR’ sulfatase signature as an active site near  the  N’terminus  and

a conserved stretch of polar amino acid residues around the active site (Figure 4.16).

Snail MCKCLLVLIAIITACAVADQSSASAGTRQDAGQPNIVFVLADDFGFHDVG-YHGSEIHTP 59 SUL2 ------------------------------MKRPNFLFIMTDTQATNMVGCYSGKPLNTQ 30 :**::*:::* . : ** * *. ::* Snail TLDALSASGVRLEN-YYVQPICTPTRSQLMSGRYQIHTGLQHGIINSCQPNALPNDSPTL 118 SUL2 NIDSLAAEGIRFNSAYTCSPVCTPARAGLFTGIYANQSGPWTNNVAP------GKNISTM 84 .:*:*:*.*:*::. * .*:***:*: *::* * ::* . : . :: .*: Snail ADKLKESGYATHMVGKWHLG---FYKQEYLPWNRGFDTYFGYLNAAEDYFNHNVPWRQVR 175 SUL2 GRYFKDAGYHTCYIGKWHLDGHDYFGIGECPPEWDADYWFDGANYLSELTEKEISLWRNG 144 . :*::** * :*****. :: * : . * :*. * .: ::::. :

238

Snail YLDLRDNNGPVRNETGQYSAHLFTGKAIDVVQSHN-TSKPLFLYLAYQSVHAPLEVPEKY 234 SUL2 LNSVEDLQANHIDET-FTWAHRISNRAVDFLQQPARADEPFLMVVSYDEPHHPFTCPVEY 203 .:.* :. :** ** ::.:*:*.:*. :.:*::: ::*:. * *: * :* Snail EHKYR----NITDKNRRTFAGMVSALDEGVANLTQALKDKGLWNNTVLIFSTDNGGQIHA 290 SUL2 LEKYTDFYYELGEKAEDDLANKPEHHRLWAQAMPSPVGDDGLYHHPLYFACNDFVDDQIG 263 .** :: :* . :*. . . :...: *.**:::.: : ..* .: . Snail GGNNYPLRGWKASLWEGGFHGVGFVSGGALKRSGAVSKGLIHVSDWFPTLVTLAGGNLNG 350 SUL2 RVINALTPEQRENTWVIYTSDHGEMMG-AHKLISKGAAMYDDITRIPLIIRSPQGERRQV 322 * : . * . * : * * * . : .:: : : * . : Snail TKPLDGFNQWDTISNETPSPREILLHNIDILYPQ--KGVPLYSNTWDTRVRAAIRVGDYK 408 SUL2 DTPVSHIDLLPTMMALADIEKPEILPGENILAVKEPRGVLVEFNRYEIEHDSFGGFIPVR 382 .*:. :: *: : : :* . :** : :** : * :: . : . : Snail LITGDPGNGSWVPPPDGHLY---FVPEIQESAAKNVWLFNITADPNEHNDLSSEKPLEVL 465 SUL2 CWVTDDFKLVLNLFTSDELYDRRNDPNEMHNLIDDIHFADVRSKMHDALLDYMDKIRDPF 442 . * : ....** *: .. .:: : :: :. :: :* : : Snail RLLQILVQFNNTAVPPRYP-----APDPRCDPALHGDVWGPWE------------ 503 SUL2 RSYQWNLRPWRKDAQPRWMGAFRPRPQDGYSPVVRDYDTGLPTQGVKVEEKKQKF 497 * * :: .. . **: *: .*.::. * Figure 4.16 Alignments of a putative bacterial sulfatase protein sequence with H. pomatia`Snail' sulfatase. The catalytic sulfatase signatures ‘CXPXR’  (yellow)  at  N’  terminus  are highlighted. The active sites containing a stretch of polar residues (blue) are highlighted. The assigned gene names are referred to Table 4.10. The symbols "*" means that the residues or nucleotides in that column are identical in all sequences in the alignment, ":" means that conserved substitutions have been observed and "." means that semi-conserved substitutions are observed. Amino acid color codes are referred to Figure 4.3 (Larkin et al., 2007).

A phylogenetic tree for bacterial putative myrosinases and sulfatases was also

constructed (Figure 4.17).  It  shows  that  ‘Aphid’  myrosinase  is  more  closely related to bacterial

bgl, GH1 and 6pbg1 which all come from the same GH1 family. Snail sulfatase is more closely

related to bacterial SUL2 than SUL1.

239

Figure 4.17 Phylogenetic tree for bacterial putative myrosinases and sulfatases. ‘Aphid’  refers  to  aphid  myrosinase from B. brassicae,   ‘Snail’   refers   to   snail   sulfatase   from  H. pomatia, and the rest refers to Table 4.10. This was constructed using the program ClustalW2 (Larkin et al., 2007). 4.3.2 Cloning of putative bacterial GSL-degrading enzymes/sulfatases

Ten candidate genes (Table 4.10) were cloned from genomic DNA of either E. coli

O83:H1 NRG 857C or E. casseliflavus NCCP-53 using the designed primers flanking the gene

regions of interest. The PCR gene products were analyzed by agarose gel electrophoresis, and

the expected sizes of the product bands from all ten genes were observed (Figure 4.18).

240

Figure 4.18 Agarose gel electrophoresis of genomic PCR experiments: Lane M – 10k bp ladder (BioLabs, UK). Single PCR fragments were amplified with corresponding primers (Table 4.1) using genomic DNAs from E. casseliflavus NCCP-53 and E. coli O83:H1 NRG 857C; Lane bgl: expected length of 1470 bp; Lane GH1: expected length of 1437 bp; Lane GH3#2: expected length of 1488 bp; Lane 6pbg1: expected length of 1407 bp; Lane GH3#3: expected length of 2151 bp; Lane GH3#1: expected length of 2211 bp; Lane pBgl: expected length of 2256 bp; Lane SUL1: expected length of 2124 bp; Lane SUL1: expected length of 1494 bp; Lane 6pbg2: expected length of 1353 bp.

Each PCR product was ligated to a pET28b+ expression vector, and the resulting gene

construct  was   transformed   into  DH5α   competent   cells.   The   transformant   colonies   containing  

the desire gene construct grown on the kanamycin selection plates were randomly selected for

colony PCR experiments using the vector forward primer and the gene-specific reverse primer.

The colony PCR results showed that there were the correct sizes of gene inserts in the vectors

(Figure 4.19).

241

Figure 4.19 Agarose gel electrophoresis of colony PCR experiments: Lane M – 10k bp ladder (BioLabs, UK). Single PCR fragments were amplified with corresponding primers (Table 4.1) from two selected colonies habouring potential gene inserts; Lanes bgl: expected length of 1,470 bp; Lanes GH3#1: expected length of 1437 bp; Lanes GH3#2: expected length of 1488 bp; Lanes 6pbg1: expected length of 1407 bp; Lanes GH3#3: expected length of 2151 bp; Lanes GH3#1: expected length of 2211 bp; Lanes pBgl: expected length of 2256 bp; Lanes SUL1: expected length of 2124 bp; Lanes SUL1: expected length of 1494 bp; Lanes 6pbg2: expected length of 1353 bp.

These positive colonies were then grown in LB broth containing kanamycin overnight,

and their plasmids were isolated for restriction enzyme digestion experiments. No transformant

colonies for the SUL1 gene were obtained after several attempts, and thus the study of this

gene was discontinued. The restriction pattern of corresponding restriction enzyme confirmed

the presence of the gene inserts and pET28b+ expression vectors (Figure 4.20), showing three

genes as examples before the plasmids were sent to GATC company (UK) for gene sequencing,

and it was confirmed that all the genes had correct sequences (Appendix II).

242

Figure 4.20 Agarose gel electrophoresis of restriction enzyme digestion experiments: Lane M – 10k bp ladder (BioLabs, UK); Lane 1: Uncut circular pET28b(+) vector with 5369 bp of size; Lane 2: Cut linearized pET28b(+) vector with the expected length of 5369 bp; Lane 3: Uncut pET28b-bgl plasmid; Lane 4: Uncut pET28b-GH1 plasmid; Lane 5: Uncut pET28b-GH3#2 plasmid; Lane 6: Cut pET28b-bgl plasmid by NdeI and Sac I enzymes with a linearized pET28b(+) fragment of 5369 bp and a bgl gene insert of 1470 bp ; Lane 7: Cut pET28b-GH1 plasmid by NdeI and Sac I enzymes with a linearized pET28b(+) fragment of 5369 bp and a GH1 gene insert of 1437 bp; Lane 8: Cut pET28b-GH3#2 plasmid by NdeI and Sac I enzymes with a linearized pET28b(+) fragment of 5369 bp and a GH3#2 gene insert of 1488 bp.

4.3.3 Recombinant protein expressions by IPTG induction

To express the desired genes recombinantly, the recombinant plasmids were

transformed into E. coli BL21(DE3) host cells for recombinant protein expression. SDS-PAGE

analysis indicated that large amounts of protein from each gene of interest is expressed in the

cell supernatant after induction by 0.5 mM IPTG for 6 h at 25°C under aerobic conditions (Figure

4.21) with the band size similar to the predicted molecular weight. The negative control, non-

recombinant BL21(DE3) culture was induced by IPTG in the same manner as other recombinant

cultures. However, there was no protein band of interest expressed in this control sample as

expected. These cell-free extracts were desalted to be ready for the following assays; GOD-

PERID assay, β-O-glucosidase activity assay and arylsulfatase activity assay. Protein

concentrations of cell-free extratcs were determined by Bradford’s method (Chapter 2, section

2.2.20).

243

Figure 4.21 Recombinant protein expressions on SDS-PAGE. Recombinant cultures were aerobically induced by 0.5 mM IPTG for 6 h at 25°C once OD600nm reached 0.6. Cell-free extracts of recombinant enzymes were analyzed on SDS-PAGE. (A) Recombinant proteins from E. casseliflavus NCCP-53 with the expected sizes were shown in boxes; bgl (56 kDa), GH1 (55 kDa), GH3#2 (54 kDa), 6pbg1 (54 kDa), GH3#3 (79 kDa), pBgl (83 kDa) and GH3#1 (81 kDa). Lane M, Low Range unstained marker (Sigma-Aldrich); Lane C, Control non-recombinant cells. (B) Recombinant proteins from E. coli O83:H1 NRG 857C with the expected sizes were shown in red boxes; SUL2 (57 kDa) and 6pbg2 (50 kDa). Lane M, PageRuler pre-stained protein ladders (Thermoscientific, UK); Lane C, Control sample with proteins from non-recombinant BL21(DE3) cells. Each lane was loaded with 20 µg protein content. 4.3.4 Enzyme activity assays

The desalted cell-free extracts containing proteins of interest from the previous section

were analyzed for the myrosinase enzyme activity using the GOD-PERID assay, the β-O-

glucosidase enzyme activity and the arylsulfatase activity as follows.

4.3.4.1 Myrosinase activity using GOD-PERID assay

No myrosinase activity was detected in any of cell-free extracts of the recombinant

enzymes (Figure 4.22). AITC or/and ANIT were not detected by GC-MS analysis either. The

negative control containing cell-free extracts from BL21 (DE3) cells without recombinant

protein expression showed negative result. The positive control (with S. alba myrosinase)

244

showed development of green-coloured product from sinigrin substrate. This is an indication of

myrosinase activity with specific activity of 28.3 ± 0.12 μmol/min/mg. This positive result from

the positive control proves the validity of GOD-PERID assay for testing myrosinase activity.

Figure 4.22 Myrosinase activity using GOD-PERID assay. Cell-free extracts of each recombinant protein (~ 200 µg) was aerobically incubated with 0.5 mM sinigrin in 0.1 M citate phosphate buffer pH 7.0 at 37°C for 30 min and was boiled for 5 min before 1 mL GOD-PERID reagent was added. The mixture was aerobically incubated at 37°C for 15 min. BL21 is a negative control, cell-free extract of BL21 without recombinant enzyme; GH1 is glycosyl hydrolase 1; GH3#3 is glycosyl hydrolase family 3; bgl is β-glucosidase; C is a positive control containing purified S. alba myrosinase (5 μg). From this result, it was hypothesized that the current pH 7.0 might not be optimal for

the myrosinase to function. Therefore, different pHs of 4.0, 5.0, 6.0 and 8.0 of 0.1 M citrate

phosphate buffer and also different buffers such as 0.1 M Tris-Cl, 0.1 M PBS, and 0.1 M sodium

phosphate buffer were used with the same sinigrin concentration under the same experimental

conditions. In spite of several trials, green product was still not detected in any recombinant

enzyme or in any buffer conditions (data not shown). Since most of the recombinant proteins

come from GH family which are supposed to hydrolyze β-O-glucosides such as cellobiose, salicin,

trehalose and methyl β-D-glucopyranoside, GOD-PERID assay was carried out with these

substrates to determine whether any of the recombinant enzymes can hydrolyze them. It was

found that GH1, GH3#1, bgl and GH3#3 enzymes all were able to hydrolyze cellobiose and

trehalose with GH3#3 having the highest specific activity (Figure 4.23A and 4.23C). In contrast,

only GH3#3 was able to hydrolyze salicin and methyl β-D-glucopyranoside (Figure 4.23B and

4.23D). The specific activity of GH3#3 on substrates in descending order is cellobiose >

trehalose > salicin > methyl β-D-glucopyranoside. The negative controls (cell-free extracts of

BL21(DE3) cells without any recombinant protein expression) showed no activity on most

245

substrates except for trehalose. These results indicate that recombinant enzymes exhibit β-O-

glucosidase or glycosyl hydrolase activity upon natural β-O-glucoside substrates. This also

proves the validity of GOD-PERID assay.

(A) Apparent specific  activity  (μmol/min/mg) (B) Apparent specific  activity  (μmol/min/mg)

BL21 GH1 bgl GH3#1 GH3#3 GH1 bgl GH3#1 GH3#3

ND 1.43 0.87 1.47 6.48

ND ND ND 0.79

± 0.12 ± 0.06 ± 0.10 ± 0.11 ± 0.05

(C) Apparent specific  activity  (μmol/min/mg) (D) Apparent specific  activity  (μmol/min/mg) BL21 GH1 bgl GH3#1 GH3#3 GH1 bgl GH3#1 GH3#3 2.02 2.67 3.7 3.41 3.59

ND ND ND 0.09

± 0.07 ± 0.12 ± 0.09 ± 0.13 ± 0.08 ± 0.02 Figure 4.23 β-O-glucosidase activity using GOD-PERID assay. Cell-free extracts of each recombinant protein (~ 20 µg) was aerobically incubated with either 1 mM of (A) cellobiose, (B) salicin, (C) trehalose or (D) methyl β-D-glucopyranoside in 0.1 M citate phosphate buffer pH 7.0 at 37°C for 30 min and was boiled for 5 min before 1 mL GOD-PERID reagent was added. The mixture was then incubated at 37°C for 15 min. BL21 is a negative control, supernatant without recombinant enzyme; GH1 is glycosyl hydrolase family 1; bgl is β-glucosidase; GH3#1 is glycosyl hydrolase family 3; GH3#3 is glycosyl hydrolase family 3. Apparent specific activity of each enzyme was determined since each enzyme was not purified. Data of specific activity of each enzyme on each substrate was shown below the pictures. Values are means ± SD of triplicates. ND, Not detected.

246

4.3.4.2 β-O-glucosidase activity assay

Since GOD-PERID assay only produced positive results when β-O-glucosides were used

as substrates, it was thought that these enzymes may not exhibit β-thioglucosidase (i.e.

myrosinase) activity to hydrolyze GSLs, but they may only have β-O-glucosidase activity.

Therefore, another experiment to reassure the existence of β-O-glucosidase activity of these

enzymes was carried out using p-nitrophenyl-β-D-glucopyranoside (pNPG) as a synthetic

substrate in 0.1 M citrate phosphate buffer pH 7. If β-O-glucosidase activity is present, the

breakdown of pNPG would produce D-glucose and a yellow-coloured p-nitrophenol (pNP)

product that can be determined spectrophotometrically at 400nm. It is important to note that

plant myrosinase from GH1 can also hydrolyze pNPG in addition to GSLs. It was found that

three enzymes GH1, bgl and GH3#3 produced yellow-coloured products suggesting β-O-

glucosidase activity of these enzymes on pNPG substrate (Figure 4.24). Since the GH3#3 enzyme

showed the highest specific activity on all β-O-glucosides tested, this enzyme was to be studied

in further details in Chapter 5.

Apparent specific activity (µmol/min/mg)

BL21 bgl GH1 GH3#2 6pbg1 GH3#3 GH3#1 pBgl 6pbg2

ND ND 0.43 ± 0.11 ND ND 1.54 ± 0.06 0.39 ± 0.08 ND ND

Figure 4.24 β-O-glucosidase activity assay. Cell-free extracts of the recombinant enzymes (~ 20 µg) were aerobically incubated with 1 mM pNPG in 1 mL of 0.1 M citate phosphate buffer pH 7.0 at 37°C for 5 min after which 5 mL of 0.1 M NaOH was added to the mixture. BL21 is a negative control, supernatant without recombinant enzyme; bgl is β-glucosidase; GH1 is glycosyl hydrolase family 1; GH3#2 is glycosyl

hydrolase family 3; 6pbg1 is 6-phospho- β-galactosidase; GH3#3 is glycosyl hydrolase family 3; GH3#1 is

glycosyl hydrolase family 3; pBgl is periplasmic β-glucosidase; 6pbg2 is 6-phospho- β-glucosidase. Apparent specific activity of each enzyme was determined since each enzyme was not purified. Data on

247

specific activity of each enzyme on each substrate was shown below the pictures. Values are means ± SD of triplicates. ND, Not detected. 4.3.4.3 Arylsulfatase activity assay

To determine whether cell-free extracts of the recombinant SUL2 enzyme has

arylsulfatase activity, this enzyme was aerobically incubated with a synthetic substrate p-

nitrocatechol sulphate (pNCS). It was found that the recombinant SUL2 produced a red/orange-

coloured product of p-nitrocatechol (pNC) suggesting arylsulfatase activity that desulfated pNCS

substrate. The positive control containing the purified sulfatase from H. pomatia resulted in a

very strong red coloured pNC product indicating the validity of the assay (Figure 4.25). The

recombinant SUL2 enzyme with arylsulfatase activity was studied in further details in Chapter 5.

Specific  activity  (μmol/min/mg)        

HP SUL SUL2 BL21

20.8 ± 0.18 1.21 ± 0.07* ND

Figure 4.25 Arylsulfatase activity assay at pH 7.0. Cell-free extracts of the recombinant SUL2 protein (~ 20 µg) or the purified H. pomatia sulfatase (~ 20 µg) was aerobically incubated with 1 mM pNCS substrate in 1 mL of 0.05 M sodium acetate buffer pH 5.0 at 37°C for 5 min after which 5 mL of 0.1 M NaOH was added to the mixture. HP SUL is a purified H. pomatia sulfatase positive control; SUL2 is a sulfatase from E. coli O83:H1 NRG 857C; BL21 is a negative control, cell-free extract of BL21 without recombinant enzyme. Data on specific activity of each enzyme was shown below the pictures. Values are means ± SD of triplicates. ND, not detected. *The value is apparent specific activity of SUL2 enzyme since this enzyme was not purified.

248

4.4 Discussion

In addition to forward genetics approach (Chapter 3), a reverse genetics approach was

used to identify and express the bacterial putative genes for myrosinase/arylsulfatase activity in

this chapter. Ten genes from the two bacteria were cloned and expressed. No GSL-degrading

activity was detected in any recombinant enzymes tested. These enzymes may be inactive in

the pH buffers used or GSL substrate may need to be modified e.g. phosphorylation of the 6-OH

group before being hydrolyzed by these enzymes and only occur in intact cells where the

integrity of transport/phosphorylation system is intact. The previous study has shown that 6-

phosphoryl-β-D-glucopyranosyl hydrolase (P-β-glucosidase, EC 3.2.1.86) purified from

Fusobacterium mortiferum hydrolyzed several P-β-glucosides, including the isomeric

disaccharide phosphates cellobiose-6-phosphate, gentiobiose-6-phosphate, sophorose-6-

phosphate, and laminaribiose-6-phosphate, to yield glucose-6-phosphate and appropriate

aglycons (Thompson, 2002). These substrates had to be phosphorylated by a β-glucoside kinase

(BglK) of K. pneumonia bacteria prior to the hydrolysis by 6-P-β-glucosidases (Thompson et al.,

1997). Since GSLs are β-glucosides, it is thought that they may need to be phosphorylated by β-

glucoside kinase prior to being hydrolyzed by myrosinase which recognizes the phosphorylation

on the GSL structure. However, this hypothesis still remains untested.

Our results showed that crude extracts of some recombinant enzymes showed β-O-

glucosidase activity on some β-O-glucosides, but not GSL. This indicates that these recombinant

enzymes are at least active upon these substrates with different specific activities. Information

on β-O-glucosidases is currently limited to relatively few species of bacteria from the human

colonic ecosystem (Dabek et al., 2008). With this work, new data on β-O-glucosidases from E.

casseliflavus NCCP-53 has been provided. Among broad substrate specificity β-glucosidases

reported to date, a number of β-glucosidases have higher activity for aryl-glucosides e.g. methyl

β-D-glucopyranoside and pNPG than cellobiose (González-Pombo et al., 2008; Matsui et al.,

249

2000; Parry et al., 2001). These enzymes have only 20–50% enzyme activity against cellobiose,

comparing to that for the pNPG. Some zymes have only 20–50% enzyme activity against

cellobiose, comparin(Hashimoto et al., 1998; Marques et al., 2003). In contrast, the

recombinant GH3#1, GH1, GH3#3 and bgl enzymes from E. casseliflavus NCCP-53 in this study

have higher activity on cellobiose than aryl-glucosides including methyl ose than aryl-

glucosidespNPG. Interestingly, all crude extracts of these recombinant enzymes and also the

extracts from the control BL21(DE3) without recombinant enzyme expression exhibited activity

for trehalose containing an α,α-1,1-glucosidic bond indicating that BL21(DE3) host expressed

endogeneous trehalase enzyme to hydrolyze trehalose. Therefore, the recombinant enzymes

need to be purified in order to determine whether they can hydrolyze trehalose. The

hydrolyzing activities of crude extracts of the recombinant GH3#1, GH1, and GH3#3 enzymes

were observed for substrates containing ified in order to deterpNPG and cellobiose. However,

no activity was observed for salicin and and methyl β-D-glucopyranoside (except for GH3#3).

The basis of the vast diversity in biological function of β-glucosidases from different GH

familes is the substrate aglycone specificity differences that determine their natural substrates

(Ketudat Cairns & Esen, 2010). There is also a difference in substrate specificity within the GH1

and GH3 groups (Ketudat Cairns & Esen, 2010). Thus, different members from different GH

families and even members in the same GH family can have different activity towards the same

substrate. As demonstrated in the results, the β-O-glucosidase activity towards pNPG was

detected from GH1, GH3#3 and GH3#1 whereas there was no activity from other members of

the same families. In addition, arylsulfatase activity towards pNCS substrate was detected in the

recombinant SUL2 enzyme from E. coli O83:H1 NRG 857C. Thus far, bacterial GSL-degrading

activity has not been found despite several experiments including cell-free extract experiment,

native activity gel analysis (Chapter 2), forward and reverse proteomics approaches (Chapter 3

and 4, respectively) were carried out. It was speculated that bacterial GSL-degrading enzyme

system may be more complicated than previously thought. Since 1974 when the first bacterial

myrosinase produced by Enterobacter cloacae was purified (Tani et al., 1974), other researchers

had no success in repeating it.

250

To summarize, βoO-glucosidase activity with highest specific activity was found in the

recombinant GH3#3 enzyme, and arylsulfatase activity in the recombinant SUL2 enzyme. These

two enzymes may be of importance in GSL metabolism in human gut bacteria. Thus, these were

characterized in further details with the GH3#3 enzyme to be referred as GH3 in the next

chapter.

Chapter 5: Characterization of the recombinant SUL2 enzyme from E. coli O83:H1 NRG 857C and the recombinant GH3 enzyme from E. casseliflavus NCCP-53 5.1 Introduction

In chapter 4, β-O-glucosidase activity from the recombinant GH3 enzyme derived from E.

casseliflavus NCCP-53 and arylsulfatase activity from the recombinant SUL2 enzyme derived

from E. coli O83:H1 NRG 857C were detected in in vitro activity assays. These two enzymes may

be of importance in GSL metabolism in human gut bacteria. Based on the results from previous

chapters, the hypotheses of this chapter are shown in Figure 5.1.

251

Figure 5.1 Hypothese of this chapter. See main text for more details.

The recombinant GH3 enzyme may be involved in the hydrolysis of DS-GSLs and the

production of corresponding NITs as detected during bacterial fermentations (Chapter 2,

section 2.3.5). The recombinant SUL2 enzyme may be involved in desulfonation of intact GSLs

to produce corresponding DS-GSLs. By combining these two recombinant enzymes together in

the same reaction tube, intact GSLs may be desulfated to produce DS-GSLs and thus NIT

products may be generated. Thus, these two recombinant enzymes GH3 and SUL2 were

characterized in further details in this chapter.

5.1.1 Sulfatases in nature

Sulfatases represent a large protein family that is involved in heterogeneous processes

in humans ranging from degradation of macromolecules to hormone biosynthesis and

modulation of developmental cell signaling (Hanson et al., 2004; Hopwood & Ballabio, 2001;

Kim & Singh, 2009; Parenti et al., 1997). Sulfatases catalyze the hydrolysis of sulfate esters (CO–

252

S) as well as sulfamates (CN–S) (Figure 5.2, reactions (1) and (2), respectively) from different

sulfated substrates such as steroids, carbohydrates, proteoglycans and glycolipids.

Figure 5.2 Sulfatase reactions. Sulfatases catalyze the hydrolysis of sulfate esters such as (1) O-sulfates and (2) N-sulfates.

Their biological significance in humans is particularly emphasized by their association

with several inherited diseases such as mucopolysaccharidoses (Franco et al., 1995),

metachromatic leukodystrophy (von Figura et al., 1999), X-linked ichthyosis, chondrodysplasia

punctata (Selmer et al., 2004), and the rare multiple sulfatase deficiency syndrome (Parenti et

al., 1997; Schmidt et al., 1995). In prokaryotes, their function has been restrained to sulfate

supply. Nevertheless, sulfatases have been recently implicated in pathogenesis (Mougous et al.,

2002) notably in E. coli (Hoffman et al., 2000; Xie et al., 2004).

Sulfatases  belong  to  at  least  three  mechanistically  distinct  groups,  namely  the  Fe  (II)  α-

ketoglutarate-dependent dioxygenases (Müller et al., 2004), the recently identified group of Zn-

dependent alkylsulfatase (Hagelueken et al., 2006) and the broad family of arylsulfatases

(Hanson et al., 2004). This latter family of enzymes,   termed   “sulfatases”   in   this   chapter,   is  

certainly the most widespread among bacteria with some of them possessing more than one

hundred sulfatase genes in their genomes (Glöckner et al., 2003). The sulfatase activity on a

broad diversity of substrates leads to their classification by the International Union of

Biochemistry and Molecular Biology (IUBMB) into 17 classes (from EC 3.1.6.1 to EC 3.1.6.17).

Despite this apparent heterogeneity, sulfatases are a conserved family of enzymes sharing the

following features: (i) they have a similar size, ranging from 500 to 800 amino acids, (ii) they are

extensively glycosylated, (iii) they share significant sequence homology ranging from 20% to

60% over their entire length, and in particular in the N-terminal region carrying motif that is

unique for this class of enzymes, and (iv) the most distinguished characteristic of sulfatases is a

unique co- or post-translational modification that they undergo to produce a Cα-formylglycine

253

(FGly) residue in their active site (Clarke, 2010; Franco et al., 1995; Hanson et al., 2004; Schmidt

et al., 1995). This residue is derived from the conversion of a cysteine (in prokaryotes and

eukaryotes)  or  a  serine  (in  prokaryotes),  hence  defining  two  classes  of  sulfatases,  the  “Cys-type”  

sulfatases  and  the  “Ser-type”  (Dierks et al., 1998b; Schmidt et al., 1995). One of two different

enzymes, formylglycine-generating enzyme (FGE) (Cosma et al., 2003; Dierks et al., 2003) and

AtsB (Szameit et al., 1999), is responsible for the unique modification and conversion of Cys or

Ser to FGly, respectively. Although these systems are highly divergent, they recognize the same

consensus  motif  “(C/S)XPXR”  also  known  as  the  “sulfatase  signature”.  This  consensus  sequence  

is crucial for the unique conversion of Cys or Ser to FGly and the correct conformation of the

catalytic site of sulfatases. This consensus sequence is conserved across all known members of

the sulfatase family (Dierks et al., 1999; Dierks et al., 1997; Sardiello, 2005).

Comparative analysis of the crystal structures of one bacterial and three human

sulfatases (Boltes et al., 2001; Bond et al., 1997; Hernandez-Guzman et al., 2003; Lukatela et al.,

1998) showed a highly similarity in three-dimensional structure with a superimposable core

region that constitutes the active site of the enzymes (Boltes et al., 2001; Hopwood & Ballabio,

2001). Crystallographic studies revealed the structures of the active sites of sulfatases which

contain a divalent metal ion located within the substrate-binding pocket and a highly conserved

Cys residue, which is the target of the posttranslational modification shared by all sulfatases

and required for enzymatic activity (Fey et al., 2001). Sulfatases are found in most species from

bacteria to eukaryotes, but some lower eukaryotes and most plants lack sulfatases. The

significant sequence conservation among different species strongly suggests that sulfatases are

members of an evolutionary conserved gene family sharing a common ancestor.

5.1.2 Bacterial sulfatases

Arylsulfatase activity has been identified in several species of enterobacteria, aquatic

bacteria, pathogenic bacteria, extremophilic bacteria, and soil bacteria. Most bacterial

arylsulfatase are upregulated during sulfur starvation indicating the role in scavenging

(Dodgson et al., 1982; Kertesz, 2000). Although their preferred substrates are unknown, it is

thought they may be sulfated carbohydrates (Hanson et al., 2004). To date, only a handful of

254

bacterial sulfatases with different properties have been charcterized. For example, an

arylsulfatase (EC 3.1.6.1) extracted from Pseudomonas aeruginosa PAO1 shows maximal

activity at 57°C and pH 8.9, and a Km of 105 μM for pNCS (Beil et al., 2005).

The most thoroughly characterized bacterial sulfatase by crystal structure determined at

1.3 Å came from Pseudomonas aeruginosa (PARS) (Boltes et al., 2001). The folding and active

site region of PAS are strikingly similar to those of the known human sulfatases HARSA and

HARSB revealing a nearly spherical globular monomer with mixed α/β topology, which is

divided into two domains (Figure 5.3).

Figure 5.3 Crystal structures of Pseudomonas aeruginosa arylsulfatase (PARS). (A) Structure of PARS, characterized by two subdomains with mixed α/β topology. (B) Same structure of PARS, rotated 90°; the

strands of the large β-sheet within the N-terminal domain (numbered 1–10) and the small β-sheet in the

C-terminal domain (labeled a–d) are visible. Red cylinders: α-helices, yellow arrows: β-sheets. This picture was taken from Hanson et al., (2004).

The larger, N-terminal domain consists of α-helices surrounding a large mixed β-sheet,

which consists of 10 strands in PARS structure. The smaller, C-terminal domain contains a four-

stranded antiparallel β-sheet tightly packed against a long, solvent-exposed C-terminal α-helix.

255

As is typical for the α/β family of enzymes (Branden, 1980), the active-site cavity is nestled at

the C-terminal end of the large β-sheet, with the FGly residue located at the bottom of a

narrow cleft lined with a stretch of highly interconnected polar residues and a divalent metal

cation unambiguously characterized as Ca2+ by the geometry of its coordination sphere (Boltes

et al., 2001). The catalytic N-terminal domain of sulfatases shows a high degree of structural

similarity to that of the alkaline phosphatases but differs dramatically in sequence (Sowadski et

al., 1985). Based on the structural data, a mechanism for sulfate ester cleavage of PARS has

been proposed (Figure 5.4). This involves a nucleophilic attack from an aldehyde hydrate of

FGly residue as the functional group on the sulfur atom in the sulfate ester substrate. The

alcohol is eliminated following by the formation of a reaction intermediate containing penta-

coordinated sulfur. Subsequent elimination of the sulfate from intermolecular rearrangement

of the intermediate regenerates the aldehyde, which is again hydrated. The Ca2+ ion is involved

in stabilizing the charge and anchoring the substrate during catalysis (Boltes et al., 2001).

Figure 5.4 Proposed mechanistic scheme for the hydrolysis of sulfate ester by the active-site aldehyde FGly of PARS. A transesterification–elimination mechanism was presented for hydrated aldehyde group present in the crystal structure of PARS. This picture was taken from Hanson et al., (2004). See main texts for details.

At this writing, no crystal structures of sulfatases bound to their natural substrates have

been solved, leaving the residues that play a role in substrate specificity largely unknown. It was

thought that the recognition of natural substrates occurs outside the narrow cleft containing

the conserved active-site residues. In general, the C-terminal region of sulfatases bears the

highest structural diversity and is likely to be responsible for the variation in substrate

specificity (Hanson et al., 2004; Ghosh, 2005).

256

The properties of snail and bacterial sulfatases are summarized in Table 5.1.

Table 5.1 Properties of snail and bacterial sulfatases

Helix pomatia (snail) arylsulfatase

EC 3.1.6.1 – arylsulfatase

Reaction type: hydrolysis of sulfuric ester

Reaction: 4-nitrophenyl sulfate + H2O = 4-nitrophenol + sulfate (Roy, 1987)

Sinigrin + H2O = Desulfo-sinigrin + sulfate (Roy, 1987)

Activator: Cd2+ (Tokheim et al., 2005)

Inhibitor: Cu2+/SO42- (Tokheim et al., 2005; Roy, 1987)

pH optimum: 7.4 (Stawoska et al., 2010)

Temperature optimum: 30˚C (Stawoska et al., 2010)

Molecular weight: monomer, 1 * 66 kDa on SDS-PAGE (Roy, 1987)

Km: 1.2 mM for 4-Nitrophenyl sulfate at pH 7.1, at 30°C (Stawoska et al., 2010)

Pseudomonas aeruginosa (bacterium) arylsulfatase

EC 3.1.6.1 – arylsulfatase

Reaction type: hydrolysis of sulfuric ester

Reaction: 4-nitrophenyl sulfate + H2O = 4-nitrophenol + sulfate (Beil et al., 1995)

4-nitrocatechol sulfate + H2O = 4-nitrocatechol + sulfate (Beil et al., 1995)

4-nitrophenyl phosphate + H2O = 4-nitrophenol + phosphate (Olguin et al., 2008)

Activator: phosphate, sulphate activate hydrolysis of 4-nitrophenyl sulfate (Delisle & Milazzo, 1972)

pH optimum: 8.4 (hydrolysis of 4-nitrophenyl sulfate) (Delisle & Milazzo, 1972)

Temperature optimum: 57˚C (Beil et al., 2005)

pH range: 7.5 and 10.2 (Beil et al., 1995)

Temperature range: 40°C and 65°C (Beil et al., 1995)

Molecular weight: monomer, 1 * 57 kDa on SDS-PAGE (Beil et al., 1995)

Km: 0.55 mM for 2-methyl-4-nitrophenyl sulfate at pH 8.9, at 30°C (Hanson et al., 2006)

257

0.105 mM for 4-nitrocatechol sulfate (Beil et al., 1995)

0.042 mM for 4-nitrophenyl sulfate at pH 8.9, at 30°C (Hanson et al., 2006)

5.1.3 The GH1 myrosinases

The plant myrosinase is a dimer with subunits of 60-70 kDa each (Björkman & Janson,

1972) linked by a zinc atom and has a characteristic (on,)8-barrel structure (Figure 5.5) in

common with GH1 enzymes which act through a mechanism that gives retention of the

anomeric configuration. The difference between plant myrosinases and β-O-glucosidases is

located at the level of the active site.

Figure 5.5 The overall structure of plant myrosinase. Plant myrosinase exists as a dimer held together by a zinc (green) atom, fluoroglucose (yellow) and carbohydrate (sticks). This picture was taken from Burmeister et al., (2000).

What makes plant myrosinases differ from plant β-O-glucosidases is that the catalytic

acids differ from together by a zinc (green) atom, fluoroglucose (yellow) and carbohydrate

(sticks). This picture was taken from Burmeister plant myrosinases and β -

000220000</publication_date><number>5</number><doi>10d by a correctly positioned water

258

molecule or by an activation of a water molecule by an ascorbate molecule acting as a catalytic

base entering the active site after departure of the aglycon (Burmeister et al., 2000). The

mechanism of the plant myrosinase is shown in Figure 5.6.

Figure 5.6 The ascorbate activated catalysis of GSL hydrolysis by plant myrosinase.

Interestingly an insect myrosinase has been purified and an X-ray structural

determination carried out (Husebye et al., 2005; Jones et al., 2001; Jones et al., 2002). In

common with the plant myrosinase the aphid enzyme has the characteristic (β/α)8-barrel

structure (Figure 5.7).

259

Figure 5.7 The structure of aphid myrosinase showing the dimer. The two catalytic glutamic acid residues are shown in red (reproduced from Husebye et al., 2005).

Unlike the plant myrosinase, the aphid myrosinase is not activated by ascorbate and is

more like a β-O-glucosidase than a plant myrosinase (Husebye et al., 2005). Similar to in plant β

-O-glucosidases, aphid myrosinase has two catalytic glutamic acid residues. The only residue

specific for aphid myrosinase in proximity of the glycosidic linkage is Tyr180 which may have a

catalytic role (Husebye et al., 2005). The aglycone binding site is different from plant

myrosinase, whereas due to the presence of Trp424 in the glucose binding site, this part of the

active site is more similar to plant β-O-glucosidases, as plant myrosinases carry a phenylalanine

residue at this position (Husebye et al., 2005). The likely mechanism of the aphid enzyme is

shown in Figure 5.8.

Figure 5.8 The catalysis of glucosinolates by aphid myrosinase. This picture was taken from Bone & Rossiter (2006).

The aphid myrosinase is also a dimer as plant myrosinase, has a relatively high

temperature optimum, a low isoelectric point, and is active towards at least two structurally

different GSLs, one aliphatic and one aromatic, which is indicative of low substrate specificity

(Pontoppidan et al., 2001). This suggests coevolution of the cabbage aphid with its main food

source. The aphid has shown to employ a similar defense strategy as plants. Like its main food

source, the cabbage aphid compartmentalizes its native myrosinase and the GSL it ingests.

260

When the cabbage aphid is attacked and its tissues are damaged, its stored GSLs are activated,

deterring herbivory from attacking other aphids (Bridges et al., 2002).

The properties of plant and aphid myrosinases are summarized in Table 5.2.

Table 5.2 Properties of plant and aphid myrosinases

S. alba (plant) myrosinase

EC 3.2.1.147-thioglucosidase

Reaction type: hydrolysis of a thioglucoside or an O-glucoside

Reaction: a thioglucoside + H2O = a sugar + a thiol

Sinigrin + H2O = a sugar + AITC (or ANIT depending on pH and co-factor)

Metal ion: a dimeric enzyme is stabilized by a Zn2+-ion bound on a twofold axis (Burmeister et al., 1997)

Activator: ascorbate (Bjoerkman & Loennerdal, 1973)

Inhibitor: 2-deoxy-2-fluoroglucotropaeolin (Cottaz et al., 1996)

pH optimum: 4.5 (hydrolysis of sinigrin, in citrate and phosphate buffer) (Palmieri et al., 1982)

Temperature optimum: 60˚C (Björkman & Lönnerdal, 1973)

pI: 4.8 (Bellostas et al., 2008)

Molecular weight: dimer, 2 x 60 kDa on SDS-PAGE (Bjoerkman & Janson, 1972)

Km: 0.156 mM for sinigrin at pH 7.0, at 30°C (Palmieri et al., 1982)

61 mM for p-nitrophenyl-beta-D-glucopyranoside (Botti et al., 1995)

Brevicoryne brassicae (aphid) myrosinase EC 3.2.1.147-thioglucosidase

Reaction type: hydrolysis of a thioglucoside or an O-glucoside

Reaction: a thioglucoside + H2O = a sugar + a thiol

sinigrin + H2O = a sugar + AITC

p-nitrophenyl beta-D-glucopyranoside + H2O = p-nitrophenol + D-glucose (Pontoppidan et al., 2001) Activator: ascorbate (0.3 mM, strong activation) (Pontoppidan et al., 2001) pH optimum: 5.5 (Jones et al., 2001)

261

Temperature optimum: 40˚C (Pontoppidan et al., 2001)

pI: 4.9 (Pontoppidan et al., 2001)

Molecular weight: dimer, 2 x 53 kDa on SDS-PAGE (Jones et al., 2001) Km: 0.41 mM for sinigrin at pH 4.5, at 37°C (Pontoppidan et al., 2001)

0.52 mM for glucotropaeolin at pH 4.5, at 37°C (Pontoppidan et al., 2001)

5.1.4 The GH3 β-glucosidases

The β-glucosidases from GH3 glycoside hydrolase family is widely distributed in bacteria,

fungi and plants. GH3 enzymes perform multiple functions including cellulosic biomass

degradation, energy metabolism, plant and bacterial cell wall remodeling and pathogen

defense. The GH3 family contains > 2700 enzymes in the Carbohydrate-Active enZYmes (CAZy)

Database, including β-glucosidases (EC 3.2.1.21), glucan 1,3-β-glucosidases (EC 3.2.1.58) and

xylan 1,4-β-xylosidases (EC 3.2.1.37) (McAndrew et al., 2013). In many cases, the enzymes have

dual or broad substrate specificities with respect to monosaccharide residues, linkage position

and chain length of the substrate (Fincher et al., 2013). There are a few well-characterized

‘bifunctional’ enzymes in the family that have both β-D-xylopyranosidase activity and α-L-

arabinofuranosidase (Lee et al., 2003), and one characterized example of an N-acetyl-β-D-

glucosaminidase/β-glucosidase from Cellulomonas fimi (Nag3) (Mayer et al., 2006). At this

writing, there are only eight crystal structures of GH3 enzymes available in the Protein Data

Bank (McAndrew et al., 2013). Due to the high diversity of protein structural arrangements

found among GH3 members (see below), a robust subfamily classification is currently not

available.

The GH3 the high diversemploy the  classic  “retaining”  mechanism  (Hrmova et al., 1996;

Legler et al., 1979; Vocadlo et al., 2000) with the carboxylic acid side chains of two active-site

residues acting individually as a catalytic nucleophile and a general acid/base residue (Zechel &

Withers, 2000). The mechanism starts with the displacement of the leaving group (aglycone) by

the catalytic nucleophile, assisted by proton donation to the nucleofuge from the conjugate

acid form of the general acid/base residue. The second step involves the degradation of the

262

resulting covalent glycosyl-enzyme intermediate by an incoming water molecule, which is

activated through deprotonation by the basic form of the general acid/base residue. The

catalytic nucleophile of GH3 has been conclusively identified as a conserved Asp residue via

mechanism-based active site labeling of several members (Vacadlo et al., 2000; Hrmova et al.,

2001; Dan et al., 2000; Paal et al., 2004). However, the catalytic acid–base residue appears to

be highly variable amongst GH3 members from bacteria, fungi, and plants, therefore it is less

readily identifiable (Paal et al., 2004; Litzinger et al., 2010; Varghes et al., 2000; Pozzo et al.,

2010).

A comparison between the GH1 enzyme family where plant and aphid myrosinases

come from and the GH3 enzyme family where the recombinant GH3 enzyme (used in this

chapter) comes form is summarized in Table 5.3.

Table 5.3 Properties of the GH1 and GH3 enzyme families in comparison

GH1 GH3

e.g. plant and aphid myrosinases e.g. the recombinant GH3 enzyme in this work

Substrate specificity

β-glucosidase   (EC   3.2.1.21);   β-galactosidase (EC 3.2.1.23);   β-mannosidase   (EC   3.2.1.25);   β-glucuronidase   (EC   3.2.1.31);   β-D-fucosidase (EC 3.2.1.38); phlorizin hydrolase (EC 3.2.1.62); exo-β-1,4-glucanase (EC 3.2.1.74); 6-phospho-β-galactosidase (EC 3.2.1.85); 6-phospho-β-glucosidase   (EC   3.2.1.86);   strictosidine   β-glucosidase (EC 3.2.1.105); lactase (EC 3.2.1.108); amygdalin  β-glucosidase (EC 3.2.1.117); prunasin β-glucosidase (EC 3.2.1.118);   raucaffricine   β-glucosidase (EC 3.2.1.125); thioglucosidase (EC 3.2.1.147);   β-primeverosidase (EC 3.2.1.149); isoflavonoid 7-O-β-apiosyl-β-glucosidase (EC 3.2.1.161); hydroxyisourate hydrolase (EC 3.-.-.-); β-glycosidase (EC 3.2.1.-)

β-glucosidase (EC 3.2.1.21); xylan 1,4-β-xylosidase (EC  3.2.1.37);  β-N-acetylhexosaminidase (EC 3.2.1.52); glucan 1,3-β-glucosidase (EC 3.2.1.58); glucan 1,4-β-glucosidase (EC 3.2.1.74); exo-1,3-1,4-glucanase (EC 3.2.1.-);  α-L-arabinofuranosidase (EC 3.2.1.55).

Mechanism Retaining Retaining

Clan

263

GH-A None 3D structure

(β/α)8 (β/α)8 and  (α/β)6 Catalytic Nucleophile/Base

Glu (E) Asp (D) Catalytic Proton Donor

Glu (E) Variable See main texts for more details.

The GH1- and GH3-type enzymes are different in substrate specificity and regulation

(Ketudat Cairns & Esen, 2010). The family GH1, members all have similar (β/α)8-barrel domains

that contain their active sites which are two conserved carboxylic acid residues on β-strands 4

and 7, serving as the catalytic acid/base and nucleophile, respectively (Vuong & Wilson, 2010).

In contrast, the GH3 β-glucosidases have a two-domain structure, a (β/α)8-barrel followed by an

α/β sandwich comprising a 6-stranded β-sheet sandwiched between three α-helices on either

side (Varghese et al., 2000). The active site of GH3 enzymes is situated between the (β/α)8-

barrel and (α/β)6-barrel sandwich domains, each of which contributes one catalytic carboxylate

residue. The basis of the vast diversity in biological function of β-glucosidases from different GH

familes is the substrate aglycone specificity differences that determine their natural substrates

(Ketudat Cairns & Esen, 2010). Structures of complexes of enzymes with inhibitors and mutant

enzymes with substrates, along with mutagenesis and chimera studies comparing similar

enzymes with divergent specificities, have suggested that the basis of aglycone specificity is

complex (Tribolo et al., 2007; Berrin et al., 2003).

5.1.5 Desulfation of intact GSLs and NIT production from DS-GSLs

One of the most-widely used arylsulfatase (ARS) in research is derived from the

intestinal juice of H. pomatia (Roman snail). This snail sulfatase comprises 503 amino acids and

shows 52% identity to human arylsulfatase B (HARSB) and 27% to human lysosomal

arylsulfatase A (HARSA) (Hanson et al., 2004; Holst & Williamson, 2004; Schmidt et al., 1995;

Wittstock et al., 2000). The residues that are characteristic for the active site in eukaryotic ARSs

are also present in the sequence of snail sulfatase (Wittstock et al., 2000). Snail sulfatase

exhibits sulfatase activity upon aryl (Dodgson and Powell, 1959), steroid (Jarrige, 1963) and GSL

264

substrates (Thies, 1979). This  is  termed  ‘GSL-sulfatase  activity’.  However,  GSL-sulfatase activity

is rather different from the classical plant myrosinase (β-thioglucosidase glucohydrolase, EC

3.2.3.1), which also catalyzes sulfate release from GSLs. Snail sulfatase is a glycopolypeptide

with an apparent molecular weight of 66 kDa and a pI which ranges between 3.9 and 4.8 (Roy &

Williams, 1989). Like myrosinases, snail sulfatase is a stable enzyme that can be kept at 4–5°C

and pH 6–7 for many months without loss of activity. Snail sulfatase with GSL-sulfatase activity

has been widely used to produce DS-GSLs from intact GSLs for quantitative GSL analyzes of

plant extracts by HPLC (Sang & Truscott, 1984; Thies, 1979). DS-GSL is found to be a pre-cursor

of NIT production (Figure 5.9) in certain organisms. For example, Aspergillus flavus can

transform intact GSLs to NITs with an arylsulfatase and  a  β-O-glucosidase (Galletti et al., 2008).

Additionally,  the  recombinant  β-O-glucosidase from Caldocellum saccharolyticum can hydrolyze

DS-GSLs to produce pure NITs (Wathelet et al., 2001) and   the   recombinant   β-O-glucosidase

from the thermophilic bacterium Tp8 cloned into E. coli was able to hydrolyze a  β-thioglucosidic

bond of a DS-GSL (Plant et al., 1988).

Figure 5.9 Proposed scheme of NIT production by desulfation of GSL via sulfatase and β-O-glucosidase.

GSL is desulfated by sulfatase to produce DS-GSL which is hydrolyzed by β-O-glucosidase or β-S-glucosidase to produce D-glucose, sulfate ion and NIT. This figure was modified from Galletti et al., (2008).

GSL-sulfatase activity has also been found in the diamondback moth Plutella xylostella

(Ratzka et al., 2002) and the desert locust Schistocerca gregaria (Falk & Gershenzon, 2007). A

GSL-sulfatase catalyzes the hydrolysis of GSLs to their corresponding DS-GSLs. These are no

longer substrates for hydrolysis by myrosinase, therefore eliminating the formation of toxic GSL

hydrolysis products such as ITCs. Interestingly, no bacterial sulfatases have yet to be shown to

265

accept GSL as substrates.

In larvae of Pieris rapae, however, a different method is employed by the insect for

circumventing the toxic effects of GSLs in its diet (Wittstock et al., 2004). P. rapae larvae

possess   a   “Nitrile-specifying   protein”   (NSP)   in   their   gut.   NSP   prevents   the   formation   of   ITCs  

during GSL hydrolysis, instead directing the aglycone of the GSL to form NITs which are less

toxic than ITCs. The presence of this enzyme allows P. rapae larvae to feed extensively on plants

containing GSLs.

5.1.6 Protein purification techniques

In this chapter, the recombinant SUL2 enzyme from E. coli O83:H1 NRG 857C and the

recombinant GH3 from E. casseliflavus NCCP-53 (Chapter 4) were to be further purified for

enzymatic assays and kinetic studies. A variety of purification techniques are available, but only

those used in this chapter are presented here.

Proteins can be purified in an active form on the basis of their characteristics e.g. size,

charge, solubility, and specific binding affinity. In general, protein mixtures are separated using

a series of separations to yield a pure protein which is concentrated using ultrafiltration. At

each step in the purification, the protein is assayed to determine its specific activity and its

concentration.

5.1.6.1 Desalting and buffer exchange

The goal of desalting process is to remove buffer salts from a protein sample in

exchange for water while buffer exchange process is aimed to exchange one set of buffer salts

in a sample for another set. Typically, it is necessary to remove salts, phenol or contaminated

nucleotides from protein solutions, and to separate excess cross-linking, labeling or

derivatization reagents from conjugated proteins to enable the enzyme activity. A protein

solution can be prepared in a more appropriate buffer using buffer exchange prior to

downstream applications such as affinity chromatography, ion exchange or electrophoresis

(Stryer et al., 2002).

266

5.1.6.2 Ultrafiltration

Ultrafiltration is used to separate extremely small particles and dissolved molecules

from fluids based on molecular size. Molecules from 1K to 1000K molecular weight (MW) are

retained by certain ultrafiltration membranes, while salts and water will pass through. Not only

ultrafiltration membranes are used to collect material retained by the filter, but also to purify

material passing through the filter. In general, ultrafiltration is used to separate proteins from

buffer components for buffer exchange, concentration or desalting. The most commonly used

membranes have a nominal molecular weight limit (NMWL) of 3 kDa to 100 kDa. Advantages of

ultrafiltration over precipitation process is that it is far gentler to solutes and more efficient

because it can simultaneously concentrate and desalt solutes.

5.1.6.3 Ion exchange chromatography

This is a process to separate proteins and other molecules in a solution on the basis of

differences in net charge. A particular net charge of a protein can be achieved by dissolving a

protein in a buffer that is either below or above its isoelectric point (pI). For example, a protein

with a pI of 5 in a buffer at pH 7 will gain a net negative charge, and can bind to positively

charged molecules e.g. diethylaminoethyl (DEAE) cross-linked to a solid support e.g. Sephadex,

known as an anion exchange column. In contrast, a protein with a pI of 7 in a buffer at pH 5 will

gain a positive charge, and can bind to a negatively charged solid support e.g. carboxymethyl

(CM)-Sephadex, known as a cation exchange column. In general, salts such as sodium chloride is

used to elute the bound protein irrespective of what type of the column is being used. The

chloride anion as the counter ion is used in an anion application to exchange for and then

release the target protein. The sodium cation as the counter ion is used in a cation application.

The strength of the electrostatic interaction between a protein and a solid support is

determined by the difference in the target protein pI and the buffer used. The more

concentrated the sodium chloride is needed to elute the protein with increasing electrostatic

charge (Stryer et al., 2002). Alternatively, target proteins can be eluted by altering the pH of the

buffer. For example, a protein with a pI of 5 bound to an anion column at pH 7 will be eluted by

267

decreasing the pH to below 5 (Stryer et al., 2002). Irrespective of whether they are buffering

agents or salts, ion exchangers are different in their effectiveness for specific applications. The

scheme of how ion-exchange chromatography works is shown in Figure 5.10.

Figure 5.10 Scheme of how ion-exchange chromatography works. (A) Negatively charged proteins are bound to an anion exchanger with positively charged stationary phase at low ionic strength. The bound proteins can be eluted either by increasing the ionic strength of the buffer or by adjusting the pH. The figure was modified from www.waters.com. (B) A cation exchange containing negatively charged beads that allow negatively charged proteins to go through. This figure was taken from Stryer et al. (2002).

A more biocompatible high-resolution separation of biopolymers of proteins can be

achieved by using fast protein liquid chromatography (FPLC). This high-performance

chromatography makes use of small-diameter stationary phases to provide high resolution. It

features include biocompatible aqueous buffer systems, high loading capacity, fast flow rates,

and availability of stationary phases in most common chromatography modes e.g. reversed

phase, ion exchange, affinity, and gel filtration (Sheehan & O' Sullivan, 2004). The protein

separation by FPLC is reproducible due to the involvement of a high level of automation

including gradient program control, auto-samplers, and peak collection (Madadlou et al., 2011).

This method is also applicable to other types of biological samples including plasmids and

268

oligonucleotides. The most common FPLC mode is anion exchange of proteins e.g. Mono Q HR

16/10 Columns pre-packed  with  Mono  Q™  (GE  Healthcare).  

The most commonly used configuration for ion-exchange FPLC is shown in Figure 5.11.

The basic FPLC system consists of a program controller (LCC 500 Plus), two P-500 pumps (one

each for buffers A and B), a mixer, pre-filter, seven-port M-7 valve, assorted sample loops

(0.025–10 mL), a column, a UV-1 ultraviolet (UV) monitor (fitted with an HR-10 flow cell and

280-nm filter), and a Frac-100 fraction collector (Sheehan & O' Sullivan, 2004).

Figure 5.11 Format used for FPLC chromatography. The superloop may be incorporated using the arrangement shown within dashed lines. See text for more details. This figure was taken from Sheehan & O' Sullivan (2004).

A superloop, as associated equipment, may be incorporated to an eight-port M-8 valve

and a P-1 peristaltic pump for the sample load up to the volume of 10 mL. All the parts are

obtained from GE Healthcare. Mono Q HR5/5 anion-exchange column is obtained from GE

Healthcare and is stored in 20% ethanol. All reagents for buffer preparation are of analytic

grade, and milli-Q or HPLC-grade water should be used to prepare buffers. Examples of buffers

269

commonly used in FPLC are Buffer A: 10 mM Tris-HCl, pH 7.0 and Buffer B: 10 mM Tris-HCl, pH

7.0, 1 M NaCl. Salt is included in Buffer B to elute the desired proteins during the salt gradient.

5.1.6.4 Ni2+-affinity chromatography

This technique uses the ability of histidine amino acid (His) to bind nickel ion (Ni2+). Six

His residues at the end of a protein (either N or C terminus) are known as a 6X His tag. Ni2+ is

bound to an agarose bead by chelation using nitroloacetic acid (NTA) beads. The general

method involves mixing the NTA beads with the protein sample, and pouring the slurry of NTA

beads and proteins into the column to batch absorb the proteins onto the column. Low affinity

bound proteins are removed by using low concentrations of phosphate and imidazole. If

necessary, the imidazole can be increased to 20 mM before most His tagged proteins are eluted.

Finally, higher concentrations of imidazole are used to elute His-tagged proteins from the NTA-

beads. A scheme of how Ni2+-affinity chromatography works is shown in Figure 5.12.

Figure 5.12 Ni2+-affinity chromatography. Proteins with His tags were loaded on the Ni2–charged column. Upon increasing concentration of imidazole in the buffer, His-tagged protein is eluted from the column.

5.1.7 Hypotheses

270

Since sulfatase activity of the recombinant SUL2 enzyme from E. coli O83:H1 NRG 857C

and β-O-glucosidase activity of the recombinant GH3 enzyme from E. casseliflavus NCCP-53

were detected in Chapter 4, several hypotheses regarding these two recombinant enzymes are

proposed to be tested in this chapter as follows:

The recombinant SUL2 enzyme may be able to desulfate intact GSLs to produce DS-GSLs.

There are differences in the efficiency of sulfatase activity on intact GSL substrates

between the commercially available H. pomatia sulfatase and the recombinant SUL2 enzyme.

There are differences in substrate specificity for sulfatase activity of the recombinant

SUL2 enzyme on different types of intact GSL substrates.

The recombinant GH3 enzyme can catalyse the hydrolysis of DS-GSL substrates to NIT

products as shown in Figure 5.1.

The sequential reaction from both the recombinant SUL2 enzyme and the recombinant

GH3 enzyme, in spite of different bacterial origins, can produce NIT products from intact GSL

substrates as shown in Figure 5.8.

5.1.8 Objectives

To test the above hypotheses, the aims were as follows:

To desulfate intact GSLs by purified H. pomatia sulfatase and the recombinant SUL2

enzyme in a partially purified form or crude extracts. The amounts of DS-GSLs produced from

two different sulfatases were compared.

To desulfate different types of intact GSLs at different concentrations by the

recombinant SUL2 enzyme. The amounts of DS-GSLs produced were used to determine enzyme

activities of sulfatase activity on different substrates.

To use different DS-GSLs as substrates for the recombinant GH3 enzyme. The hydrolysis

products were identified by GC-MS analysis.

To use Intact GSLs as substrates for a sequential reaction of both the recombinant SUL2

enzyme and the recombinant GH3 enzyme. The hydrolysis products were identified by GC-MS

analysis.

271

5.2 Materials and Methods

5.2.1 Bioinformatics tools

Amino acid sequences of SUL2 was analyzed for sequence features (i.e. domains, motifs)

using PROSITE (Sigrist et al., 2013) (http://prosite.expasy.org). The sequence logo was generated

using InterProScan (Quevillon et al., 2005) (http://www.ebi.ac.uk/InterProScan/). Sequence logo

is a graphical version of a consensus sequence; where the height of the letter stack indicates

conservation and the height of each letter (within a stack) indicates the relative frequency of

that residue at each position. Amino acids are coloured according to their chemical properties:

polar amino acids (G,S,T,Y,C,Q,N) are green; basic (K,R,H) blue; acidic (D,E) red and hydrophobic

(A,V,L,I,P,W,F,M) black (Crooks et al., 2004).

5.2.2 Inducibility of a native SUL2 enzyme of E. coli O83:H1 NRG

Cell-free extracts (200 µL) from E. coli O83:H1 NRG 857C cultures (both induced and

non-induced with 1 mM gluconasturtiin overnight in 200 mL NB broths) were assayed for

sulfatase activity using 1 mM pNCS as a substrate as previously described (Chapter 4, section

4.2.21). Cell-free extract from E. coli BL21(DE3), as a negative control, was also assayed. Protein

concentrations were determined by Bradford assay (Bradford, 1976) as previously described

(Chapter 2, section 2.2.20).

5.2.3 Reverse transcriptase polymerase chain reaction (RT-PCR)

E. coli O83:H1 NRG 857C was grown in 2 mL NB and WC media, respectively

with/without 1 mM gluconasturtiin at specified time intervals within 24 h. At each time interval,

the bacterial cultures were pelleted and total RNA from the bacterial pellets extracted using

Trizol   (Invitrogen)  according   to  manufacturer’s   instructions.  RNA  was  purified  using  Pure   Link  

RNA minikit (Invitrogen, UK) with on-column DNase treatment to remove contaminating DNA.

272

The primer pairs were designed using Primer 3 software (http://frodo.wi.mit.edu/primer3/) and

1 mg of total RNA was used as the template for reverse transcription of nine selected genes

with Omniscript reverse transcriptase by OneStep RT-PCR  (Qiagen)  according  to  manufacturer’s  

instructions. PCR was carried out with the primer pairs of SUL2 gene (Table 5.1) in Eppendorf

Thermal Cycler MasterCycler Personal 5332 (50˚C  for  30 min,  95˚C  for  15 min, followed by 28

three-step cycles  of  94˚C  for  1  min,  56˚C  for  1  min  and  72˚C  for  1 min). The transcripts of 16S

ribosomal RNA from E. coli O83:H1 NRG 857C (NCBI Reference Sequence: NC_017634.1) with

constitutive expression were used as positive controls. The primer pairs used in this experiment

are shown in Table 5.4.

Table 5.4 List of primers used in RT-PCT experiments

ECO, E. coli O83:H1 NRG 857C; SUL2, Sulfatase; 16S, 16S rRNA (as a control gene)

Total RNA and double stranded cDNA were quantified with a spectrophotometer (NanoDrop

1000, Thermo Scientific). The products were checked on a 0.8% agarose gel as previously described

(Chapter 4, section 4.2.11) to verify RNA/cDNA quality and fragment lengths. To control for equal

amounts of RNA used in the RT-PCR  reactions,  1  μg  of  total  RNA  from  each  sample  was  analyzed by

agarose gel electrophoresis.

5.2.4 Purification of recombinant enzymes

After inducing BL21(DE3) bacterial cultures by IPTG, cells were lysed and cell-free

extracts containing the released recombinant enzymes were obtained as previously described

(Chapter 4, section 4.2.14). Those recombinant enzymes with His-tags were then purified for

further activity assays by two different purification methods as follows:

5.2.4.1 Ni2+-ffinity column chromatography

No. Name Primer  sequence  (5’-3’) Properties

1 ECO_SUL2-F ATGAAACGCCCCAATTTTCT Forward primer to flank SUL2 gene

2 ECO_SUL2-R TCAGAACTTCTGTTTTTTCT Reverse primer to flank SUL2 gene

3 ECO_16S-F GAGTTTGATCATGGCTCAG Forward primer to flank 16S rRNA gene

4 ECO_16S-R AAGGAGGTGACCAACCGCA Reverse primer to flank 16S rRNA gene

273

The Ni2+-charged Profininty IMAC resin column (Bio-Rad, UK) (4 ml bed volume packed

in a mini Econo-Column gravity-flow, Bio-Rad, UK) was pre-equilibrated with equilibration

buffer (50 mM sodium phosphate pH 8.0, 300 mM NaCl). The cell supernatant containing

recombinant enzymes was loaded onto the column and the column washed with wash buffer

(50 mM sodium phosphate pH 8.0, 300 mM NaCl, 5 mM imidazole) (20 column volumes) at a

flow rate of 1.5 mL/min. The column was eluted with elution buffer (50 mM sodium phosphate

pH 8.0, 300 mM NaCl, 500 mM imidazole) (1 column volume/fraction for 5 fractions) at a flow

rate of 1.5 mL/min. Each fraction was assayed using enzyme activity assay to determine which

fraction contains recombinant enzymes. The protein concentration of the fractions was

assessed  using  Bradford’s  reagent  (Chapter 2, section 2.2.20) and purity assessed by SDS-PAGE

using the protocol as previously described (Chapter 2, section 2.2.21). The gel densitometry to

determine the purity of each fraction was analyzed by ImageJ 1.46 software (NIH government,

US). Fractions containing recombinant enzymes were pooled, concentrated and desalted

against buffer (50 mM sodium acetate pH 6.0 for SUL2 enzyme and 100 mM citrate phosphate

pH 7.0 for GH3 enzyme) using Amicon Ultra-15 Centrifugal filter units with 10K MWCO

(Millipore, Watford, UK). The proteins were stored at 4C (for no more than two weeks) until

required.

5.2.4.2 Ion-exchange column chromatography

The Sepharose HiTrap Q HP column (GE healthcare, UK)(5 mL bed volume) was

connected to FPLC system consisting of Waters 600S controller and Waters 626 pump. The

column was pre-equilibrated with equilibration buffer (100 mM Tris-Cl pH 7.0). The supernatant

containing recombinant enzymes was loaded onto the column, and the column eluted with a

salt gradient of equilibration buffer, A (100 mM Tris-Cl pH 7.0) and elution buffer, B (100 mM

Tris-Cl pH 7.0, 500 mM NaCl) at a flow rate of 0.6 mL/min. The steps to elute the proteins are

shown in Table 5.5. This method was used to purify SUL2 enzyme as an alternative to test

whether purity of SUL2 elution was improved from the use of affinity column chromatography.

274

Table 5.5 Steps involved in the elution of the recombinant SUL2 enzyme using FPLC

At time (min) Flow (mL/min) Buffer A (%) Buffer B (%) Mode

0 0.6 100 0 Initial

15 0.6 100 0 Hold

60 0.6 0 100 Gradient

75 0.6 0 100 Hold

80 0.6 100 0 Gradient

105 0.6 100 0 Hold

105.1 0 0 0 Hold

The eluted fractions (3 mL) were collected using a Biologic Biofrac fraction collector (Bio-

Rad, UK). Each fraction (3 mL) was tested for sulfatase activity using sulfatase activity assay as

previously mentioned (Chapter 4, section 4.2.21) to determine which fractions contain the

recombinant enzymes. The protein concentration of the fractions was assessed using

Bradford’s   reagent  as previously described (Chapter 2, section 2.2.20) and purity assessed by

SDS-PAGE using the previous protocol (Chapter 2, section 2.2.21). Fractions containing

recombinant enzymes were pooled, concentrated and desalted against buffer (50 mM sodium

acetate pH 7.0 for the SUL2 enzyme) using Amicon Ultra-15 Centrifugal filter units with 10K

MWCO (Millipore, Watford, UK). The proteins were stored at 4C (for no more than two weeks)

until required.

5.2.5 Determination of pH and temperature optima

To investigate the pH and temperature optima for both recombinant enzymes, the

reaction mixtures consisting of typical assay conditions as previously described were carried out

in triplicates (Chapter 4, sections 4.2.20 and 4.2.21) with varying pH conditions in a range of pH

3.0 and pH 10.0 and varying temperatures in a range of 4C and 80C.

275

5.2.6 Measurements of apparent enzyme activities

5.2.6.1 Apparent enzyme activities of the recombinant SUL2 enzyme for intact GSLs

determined by a discontinuous assay using HPLC analysis

Since the partially purified recombinant SUL2 enzyme was unable to desulfate intact

GSLs, the non-purified crude extracts exhibiting GSL-sulfatase activity were used for this

experiment instead. The values determined throughout this work are apparent enzyme

activities since the recombinant SUL2 enzymes used in this experiment are not pure.

To determine the apparent specific activity towards 1 mM GSL, the crude extracts of the

recombinant SUL2 enzyme (1000 µg) were loaded on the DEAE-Sephadex column (Chapter 2,

section 2.2.3) to desulfate 1 mM GSL. In addition, desulfation of intact GSLs (1 mM) by the

purified H. pomatia (HP) sulfatase (100 µg) (0.3U/mL) was performed for comparative analysis.

Equilibrating buffer, 20 mM sodium acetate buffer pH 5.0 (for HP) or 50 mM sodium acetate

buffer pH 6.0 (for SUL2) was used throughout the on-column reactions that were incubated at

30˚C  for  8  h.  DS-GSLs produced by the recombinant SUL2 enzyme were eluted along with some

protein residues from the crude extracts from the DEAE-Sephadex column. To precipitate these

protein residues that may interfere with HPLC analysis, the 1.5 mL eluted DS-GSL solutions were

well-mixed with 150 µL of Pb(OAc)2:Ba(OAc)2 (1:1; each 0.5 M) solution. The mixture was

centrifuged at 16,100g for 2 min, and the clear supernatant was loaded onto the DEAE-

Sephadex column without desulfation step, and the 1.5 mL flow-through was collected and

analyzed by HPLC (Chapter 2, section 2.2.4).

To determine the apparent enzyme activities i.e. Km and Vmax, in vitro incubations of

crude extracts of the recombinant SUL2 enzyme (1000 µg) with varied concentrations of GSLs

(0.1-20 mM) in 1 mL of 50 mM sodium acetate buffer pH 6.0 in 1.5 mL Eppendorf tubes were

carried out at   30˚C   for   8   h.   After   that,   100   µL supernatant from 1 mL reaction mixture was

collected for the on-column desulfation by 75 µL of the purified H. pomatia (HP) sulfatase

276

(0.3U/mL) as previously described (Chapter 2, section 2.2.3). The 1.5 mL flow-through was

collected from the DEAE-Sephadex column and analyzed by HPLC to determine the amounts of

GSLs remaining in the in vitro incubations. The negative controls containing only GSLs (0.1-20

mM) without recombinant SUL2 enzymes were also processed to determine the amounts of

GSLs present in the in vitro incubations, and therefore the amounts of GSLs disappearing (as DS-

GSLs) due to GSL-sulfatase activity of the recombinant SUL2 enzyme in the in vitro incubations

were determined. The apparent enzyme activities (.e. Km and Vmax,) of the recombinant SUL2

enzyme in crude extracts in desulfating intact GSLs were estimated from Michaelis-Menten

curves by the least squared fitting method with 95% confidence interval using GraphPad Prism

6.

5.2.6.2 Apparent enzyme activities of the recombinant SUL2 enzyme for pNCS

determined by a discontinuous assay using a spectrophotometric method

To measure apparent enzyme activities, the amount of a partially purified recombinant

SUL2 (100 µg) was kept constant, and the concentration of pNCS substrate (0.1-4 mM) was

varied. The reaction mixture was aerobically incubated for 5 min at 30˚C   in   50  mM   sodium  

acetate buffer pH 6.0. A typical sulfatase activity was assayed as previously described (Chapter

4, section 4.2.21). The reaction rate of red-colored p-nitrocatechol (pNC) production from pNCS

substrate was calculated using a calibration curve (Chapter 4, Figure 4.13). The Vmax and Km

values were estimated from Michaelis-Menten curves by the least squared fitting method with

95% confidence interval using GraphPad Prism 6.

5.2.7 Effects of various compounds on arylsulfatase activity of the recombinant SUL enzyme

To measure the effects of various compounds on arylsulfatase activity of the partially

purified recombinant SUL2 enzyme for pNCS substrate, various compounds including Na2SO4,

NaHSO4, CoCl2, CaCl2, MgCl2, FeSO4, NiCl2, MnCl2, FeCl3 (Sigma-Aldrich, UK) were individually

added to a final concentration of 1.0 mM in a typical sulfatase activity assay as previously

described (Chapter 4, section 4.2.21). Each reaction contained 100 µg/mL of the recombinant

SUL2 enzyme with the control sample contained only the partially purified recombinant SUL2

277

enzyme without any compounds. The effect of each compound on the arylsulfatase activity was

determined as a percentage activity in relative to the control sample.

5.2.8 DS-GSLs as substrates for the recombinant GH3 enzyme

DS-GSLs (1 mM) obtained previously (Chapter 2, section 2.2.5) were used as substrates

for the purified recombinant GH3 enzyme (100 µL) aerobically incubated in a 1 mL buffer (100

mM citrate phosphate, pH 7.0) containing 1 mM Fe2+ or a 1 mL NB broth (pH 7.0) for 16 h at

37˚C.  The  negative  controls include a reaction mixture with the recombinant GH3 enzyme alone

and the other with DS-GSL substrate alone. After that, the reaction mixtures were extracted

with the same volume of DCM as previously described (Chapter 2, section 2.2.11) in order to

identify GSL degradation products by GC-MS analysis.

5.2.9 Intact GSL as substrates in a reaction containing both the recombinant SUL2 enzyme and

the recombinant GH3 enzyme

GSLs (1 mM) obtained previously (Chapter 2, section 2.2.1) were used as substrates for a

reaction mixture containing both the purified recombinant GH3 enzyme (100 µg) and the crude

extracts of the recombinant SUL2 enzyme (1000 µg) incubated in a 1 mL NB broth (pH 7.0) for

16 h at 30˚C  under aerobic conditions. Note  that  the  temperature  of  30˚C  was  used  instead  of  

37˚C   as   it   was   optimal   for   the   recombinant   SUL2   enzyme.   Also, crude extracts of the

recombinant SUL2 enzyme were filtered by 0.22 μm filter to obtain sterile solutions to be used

in sterile Eppendorf tubes in this experiment. The negative controls include the reaction

mixture containing either recombinant enzyme alone, the reaction without intact GSL substrate

and the other with intact GSL substrate alone. After 16 h, 900 µL of each reaction mixture was

extracted with the same volume of DCM as previously described (Chapter 2, section 2.2.11) in

order to identify the hydrolysis products by GC-MS analysis. The other 100 µL of supernatant

was desulfated as previously described (Chapter 2, section 2.2.3) and analyzed by HPLC

(Chapter 2, section 2.2.4) to determine the degradation of GSL substrate.

278

5.3 Results

5.3.1 Bioinformatics results of a native SUL2 enzyme of E. coli O83:H1 NRG 857C

To determine whether the intracellular SUL2 enzyme of E. coli O83:H1 NRG 857C is a

‘Cys-type’   or   ‘Ser-type’   sulfatase,   bioinformatics   research  was   carried   out.   SUL2 enzyme was

found to have alkaline phosphatase-like domain (Figure 5.13A). This is not surprising since all

sulfatase protein structures appear as globular monomers divided in two domains. The N-

terminal   domain,   spanning   the   active   site,   contains   α-helices   surrounded   by   a   large   β-sheet.

The structure of this domain shares significant similarities with that of alkaline phosphatases

(Bond et al., 1997). Cross-species sequence comparisons showed amino acid sequence

similarities between the active sites of sulfatases and those of phosphatases suggesting a

similar enzymatic mechanism and a common evolutionary origin between these two protein

families (Lukatela et al., 1998). There are two conserved sites for the sulfatase family located in

the N-terminal region of SUL2 enzyme (Figure 5.13B). The SULFATASE_1 site contains the

conserved arginine (R) which could be implicated in the catalytic mechanism; it is located four

residues after a position that, in eukaryotic sulfatases, is a conserved Cys residue (Galletti et al.,

2008) which has been shown to be modified to 2-amino-3-oxopropionic acid. In prokaryotes,

Cys is replaced by Ser. In spite of its prokaryotic origin, SUL2 enzyme is categorized into a ‘Cys-

type’ sulfatase (Figure 5.13B and 5.13C).

279

A

B PS00523 SULFATASE_1 Sulfatases signature 1 : PVCTPARagLFTG 50 – 62 PS00149 SULFATASE_2 Sulfatases signature 2 : GYhTcyIGK.WH 92 – 102 C

280

Figure 5.13 Bioinformatics details on SUL2 enzyme. (A) Sequence feature of SUL2 enzyme. The figure was retrieved from InterProScan (Quevillon et al., 2005). (B) Two sulfatase signatures found in SUL2 protein sequence. Yellow highlight indicates a strictly conserved sequence called “sulfatase   signature”  (C/S)XPXR where R is a putative active site residue. (C) Sequence logo for sulfatases signature 1 ‘CTPAR’ showing that SUL2 enzyme is a ‘Cys-type’  sulfatase. The information on (B) and (C) was retrieved from PROSITE (Sigrist et al., 2013). The maturation of the activity of all sulfatases require posttranslational modification

mediated by one of two different enzymes, FGE (Cosma et al., 2003; Dierks et al., 2003) and

AtsB (Szameit et al., 1999), responsible for the conversion of Cys or Ser to FGly, respectively.

Sequence analysis of many bacterial genomes reveals that both types of sulfatase modifying

factor genes, AtsB and SUMF1, are physically associated with sulfatase genes, indicating the

presence of sulfatase operons (Landgrebe et al., 2003). Therefore, the whole genomes of both E.

coli O83:H1 NRG 857C (as an origin of SUL2 enzyme) and E. coli BL21(DE3) (as a host to express

the recombinant SUL2 enzyme) were searched for the presence of AtsB or SUMF1 protein

homolog using BLAST search against the known anaerobic sulfatase-maturating enzyme

homolog, AslB from E. coli strain K12 (Uniprot accession no. P25550) (Benjdia et al., 2007). It is

important to note that anaerobic sulfatase-maturating enzyme homolog is named differently

from one bacterial strain to the next, but in most cases it is named AtsB. For example, it is

named AslB in E. coli strain K12 whereas it is named AtsB in Klebsiella pneumonia (Uniprot

accession no. Q9X758) (Szameit et al., 1999) and in other bacteria.

The results showed that there are three candidate proteins with low E-value for

sulfatase-maturating enzymes with sequence similarity to AslB in E. coli O83:H1 NRG 857C

genome (Table 5.6A). Likewise, three candidate proteins with low E-value for anaerobic

sulfatase maturation enzymes are found in E. coli BL21(DE3) genome (Table 5.6B). The

functionality of the recombinant SUL2 enzyme clearly validates the use of E. coli BL21(DE3) as

the recombinant expression system for the production of active sulfatase enzyme. The results

also logically point to the existence of the necessary sulfatase-modifying enzyme(s) encoded by

the E. coli BL21(DE3) genome. The heterologous expression of other catalytically active

sulfatases in E. coli host has also been reported (Boltes et al., 2001; Dierks et al., 1998a).

281

Table 5.6 Three proteins from E. coli O83:H1 NRG 857C and E. coli BL21(DE3) producing

significant alignments with anaerobic sulfatase-maturating enzyme homolog (AslB) from E.

coli strain K12

A Query = sp|P25550|ASLB_ECOLI Anaerobic sulfatase-maturating enzyme homolog AslB OS = E. coli strain K12, GN=aslB PE=3 SV=4 Object = E. coli O83:H1 NRG 857C Sequences producing significant alignments: Score E-value*

ref|YP_006122114.1| regulator of arylsulfatase activity 827 0.0

ref|YP_006119836.1| hypothetical protein NRG857_07390 330 4e-110

ref|YP_006122007.1| putative DNA-binding transcriptional regulator 45.1 1e-06

*The Expect value (E) is a parameter that describes the number of hits one can "expect" to see by chance when searching a database of a particular size. It declines exponentially as the Score of the match increases.

B Query = sp|P25550|ASLB_ECOLI Anaerobic sulfatase-maturating enzyme homolog AslB OS = E. coli strain K12, GN=aslB PE=3 SV=4 Object = E. coli BL21(DE3) Sequences producing significant alignments: Score E-value

ref|YP_003001361.1| anaerobic sulfatase maturation enzyme 830 0.0

ref|YP_002999279.1| anaerobic sulfatase maturation enzyme 334 1e-11

ref|YP_003001245.1| DNA-binding transcriptional regulator 44.3 3e-06

5.3.2 Inducibility of a native SUL2 enzyme of E. coli O83:H1 NRG 857C

282

It is known that GSL-sulfatase activity in desert locusts S. purpurea is induced ten-fold

when the locust are fed with GSLs after being maintained on a GSL-free diet, and activity

declines when GSLs are removed from the locust diet (Falk & Gershenzon, 2007). However,

sulfatase from H. pomatia was constitutively expressed, and bacterial sulfatase of C. perfringens

was also constitutively expressed with weak sulfatase activity of 0.25 nmol min–1mg–1 using a

synthetic substrate, p-nitrophenyl sulfate (pNPS) at neutral pH (Berteau et al., 2006). Therefore,

the inducibility of a native SUL2 enzyme of E. coli O83:H1 NRG 857C was determined in this

work.

Protein extracts from E. coli O83:H1 NRG 857C cells (both induced and non-induced with

1 mM gluconasturtiin overnight) were assayed for sulfatase activity using 1 mM pNCS as a

substrate. Protein extract from E. coli BL21(DE3), as a negative control, was also assayed. The

results showed that cell-free extracts from both non-induced and GSL-induced cells showed a

weak sulfatase activity of 0.07-0.09 µmol·min–1·mg–1 using pNCS as a substrate in 50 mM

sodium acetate buffer pH 5.0 (Figure 5.14A). In contrast, BL21(DE3) host did not exhibit active

endogenous sulfatase activity (Figure 5.14A) and hence suitable for heterologous expression of

the recombinant SUL2 enzyme from E. coli O83:H1 NRG 857C.

This constitutive expression of a native SUL2 enzyme was also reconfirmed by reverse

transcription–PCR (RT-PCR) analysis showing that the SUL2 transcript level remained constant

from 0 h to 24 h in E. coli O83:H1 NRG 857C cells with and without 1 mM gluconasturtiin

supplementation (Figure 5.14B). The 16S rRNA transcripts (as positive controls) from both types

of bacterial cells also showed constitutive expression (Figure 5.14B) supporting the validation of

RT-PCR analysis.

283

A (i) BL21(DE3) (ii) Induced (iii) Non-induced

Specific activity ND 0.09 ± 0.02 0.07 ± 0.04 (µmol·min–1·mg–1)

B

Figure 5.14 Inducibility test of a native SUL2 enzyme of E. coli O83:H1 NRG 857C. (A) Bacterial protein extracts were assayed for endogenous sulfatase activity using 1 mM pNCS as a substrate in 50 mM sodium acetate buffer pH 5.0. (i) E. coli BL21(DE3) showed no sulfatase activity (ii) E. coli O83:H1 NRG 857C induced with 1 mM gluconasturtiin overnight showed a weak sulfatase activity. (iii) E. coli O83:H1 NRG 857C non-induced with GSL also showed a weak sulfatase activity. Values are means ± SD of triplicates. (B) Reverse transcription–PCR (RT-PCR) analysis of SUL2 transcripts from E. coli O83:H1 NRG 857C cells over a time course of 24 h. (i) SUL2 transcripts from, non-induced E. coli O83:H1 NRG 857C cells (ii) SUL2 transcripts from gluconasturtiin-induced E. coli O83:H1 NRG 857C. (iii) 16S rRNA transcripts (as positive controls) from both types of bacterial cells. The experiments were independently repeated in triplicates with similar results.

5.3.3 Expression and purification of the recombinant SUL2 enzyme

284

To purify the recombinant SUL2 enzyme, two different protein purification methods;

Ni2+-affinity chromatography and ion-exchange chromatography were used. The elution profile

from the latter method is shown in Figure 5.15. The recombinant SUL2 enzyme eluted in the

fraction numbers 18, 19 and 20 during increasing salt gradient (Figure 5.15, step 2).

Figure 5.15 Purification of the recombinant SUL2 enzyme by ion-exchange chromatography. Step 1, Buffer A 100% (10 mM Tris-HCl, pH 7.0) without salt gradient; step 2 increasing salt gradient to Buffer B 100% (10 mM Tris-HCl, pH 7.0, 1 M NaCl); step 3, Buffer B 100% without salt gradient and step 4 decreasing salt gradient to Buffer A 100%. Dashed lines indicate purification steps and bold lines indicate protein concentration of each elution fraction collected. The 3 mL protein fraction was collected at a flow rate of 0.6 mL/min in a total of 105.10 min run.

The elution fractions from both purification methods were analyzed by SDS-PAGE

(Figure 5.16). The soluble recombinant SUL2 enzymes with a molecular weight of approximately

57 kDa (as predicted by UNIPROT) were found in the three eluted fractions from both methods.

Protein homogeneity a.k.a. protein purity was determined by densitometry analysis of proteins

on SDS-PAGE gels using ImageJ 1.46 software (NIH government, US). The use of Ni2+-affinity

column chromatography purification resulted in the partially purified recombinant SUL2

enzyme with 61% purity assessed by SDS-PAGE (Figure 5.16A, fraction E2), and the ion-

exchange column chromatography resulted in a protein with 65% purity (Figure 5.16B, fraction

number 19). The relative activity of each fraction from both methods is shown in Figure 5.16C

and 5.16D, respectively.

285

Figure 5.16 Purification of the recombinant SUL2 enzyme expressed in BL21(DE3). (A) The 4-12% SDS- PAGE analysis of protein fractions obtained from Ni2+-affinity chromatography; Lane C, total proteins from empty pET28b+ vector expressed in BL21(DE3) as a negative control; lane M, PageRuler pre-stained protein ladders (ThermoScientific, UK); lane S, supernatant of crude extracts from recombinant bacteria expressing SUL2 enzyme; lane FT, the flow-through fraction during purification; lane E1, elution fraction no. 1; lane E2, elution fraction no. 2; lane E3, elution fraction no. 3. (B) Relative activity of each eluted protein fraction with the same volume of 10 µL. (C) By ion-exchange chromatography, protein fractions are loaded in; Lane S, supernatant of crude extracts containing SUL2 enzyme; Lane 18, elution fraction no. 18; Lane 19, elution fraction no. 19; Lane 20, elution fraction no. 20; lane M, PageRuler pre-stained protein ladders (ThermoScientific, UK). Protein bands corresponding to SUL2 enzymes (57 kDa) are indicated by arrows. (D) Relative activity of each eluted protein fraction with the same volume of 10 µL using 1 mM pNCS as a substrate.

286

Both procedures gave similar purity levels of the eluted proteins, but Ni2+-affinity

column chromatography was used for further purification of the recombinant SUL2 enzyme as

it was much quicker. The purification scheme of the recombinant SUL2 by Ni2+-affinity column

chromatography is shown in Table 5.7.

Table 5.7 Purification scheme of the recombinant SUL2 enzyme using Ni2+-affinity column

chromatography

Purification Step Total activity (U)

Total protein (mg)

Specific activity (U/mg)

Purity (Fold) Yield (%)

Cell-free extract 268 206 1.3 1 100

Ni2+ affinity column chromatography 125 12 10.4 8 47

One unit (U) of sulfatase is defined as the amount of enzyme liberating 1 µmol min-1 of pNC product.

5.3.4 Temperature and pH optima of the recombinant SUL2 enzyme

The kinetics of the hydrolysis of small aryl substrate p-nitrophenyl sulfate (pNPS)

catalyzed by arylsulfatase from H. pomatia was studied at a wide range of temperatures as well

as at ambient and elevated pressures. It was found that pH   7.4   and   30˚C   were   pH   and  

temperature optima for H. pomatia sulfatase and the Km value of 2.5 mM for pNPS was

determined (Stawoska et al., 2010).

To determine the temperature and pH optima of the recombinant SUL2 enzyme, various

temperatures and pH conditions were tested in 50 mM sodium acetate buffer for sulfatase

activity.   The   recombinant   SUL2   enzyme   activity  was   found   to  be  optimal   at   30˚C   and   pH   6.0  

when 1 mM pNCS was a substrate in arylsulfatase activity assays (Figure 5.17A and 5.17B).

Interestingly, 40% of the maximal sulfatase activity still remained at 4˚C.

287

A B

Figure 5.17 Temperature and pH optima of the recombinant SUL2 enzyme. (A) Effect of temperature on arylsulfatase activity using 1 mM pNCS as a substrate. (B) Effect of pH. Enzyme activity was expressed as the percentage of the activity at 30˚C  and  pH 6.0 in 50 mM sodium acetate buffer which was defined as 100%. Partially purified recombinant SUL2 enzyme (100 µg) was used in each reaction mixture. Values were means of triplicates.

5.3.5 Desulfation of intact GSLs by the recombinant SUL2 enzyme

To determine whether the recombinant SUL2 enzyme can desulfate GSL as desert locust

sulfatase and H. pomatia sulfatase, several intact GSLs (1 mM) were used as substrates for the

partially purified recombinant SUL2 enzyme in the DEAE-Sephadex column desulfation

procedure as previously described (Chapter 2, section 2.2.3). In Chapter 2, the column was left

at room temperature during desulfation of intact GSLs by H. pomatia sulfatase (100 µg)

overnight (16-24h). It was reported that a minimum of 11 hours is necessary for complete

desulfatation of various GSLs by H. pomatia sulfatase (activity at pH 5.8 and 30°C, 0.05 U/mL;

one activity unit (U) corresponds to the desulfation of 1 µmol of sinigrin per min) on the DEAE-

Sephadex column (Figure 5.18) (Wathelet et al., 2008). If necessary, the desulfation time can be

reduced to 2 h if a ten times concentrated sulfatase solution (activity: 0.5 U/mL) is used for a

broad spectrum of GSLs (allyl, benzyl, indolyl, methylthio, and methylsulfinyl) (Wathelet et al.,

2008). Based on this finding (Figure 5.18), the incubation time of 8 h was used to desulfate

intact GSLs by the recombinant SUL2 enzyme. In theory, this would give similar levels of

desulfation as overnight incubation.

288

Figure 5.18 Kinetics of H. pomatia sulfatase activity (0.05 U/mL) in desulfation of intact GSLs. The experiment was carried out on the DEAE-Sephadex column at pH 5.8 and 30°C under aerobic conditions. This figure was taken from Wathelet et al. (2008).

To determine the specific activity of SUL2 enzyme towards 1 mM GSL, the crude extracts

of the recombinant SUL2 enzyme were loaded on the DEAE-Sephadex  column  at  30˚C  for  8  h  to  

desulfate GSL that was added earlier. The desulfation of intact GSLs by the purified H. pomatia

(HP) sulfatase was also carried out for comparative analysis. DS-GSLs produced by the

recombinant SUL2 enzyme were eluted along with some protein residues from the crude

extracts from the DEAE-Sephadex column. The proteins were precipitated by a mixture of

Pb(OAc)2:Ba(OAc)2. The clear supernatant was loaded onto the DEAE-Sephadex column without

desulfation step, and the flow-through was collected and analyzed by HPLC. The areas of the

peaks on HPLC chromatograms corresponding to DS-GSL products were used to determine the

amount of DS-GSL production by the enzyme.

It was found that the partially purified recombinant SUL2 enzyme was unable to

desulfate intact GSLs on the DEAE-Sephadex column. In contrast, the non-purified crude

extracts of this enzyme was successful in desulfating intact GSLs on the DEAE-Sephadex column.

This led us to postulate that co-factors or co-proteins present in the crude extracts may be

essential for GSL-sulfatase activity of the recombinant SUL2 enzyme. Interestingly, arylsulfatase

289

activity for pNCS substrate does not seem to require such factors since positive results were

detected in both the partially purified recombinant SUL2 enzyme and the crude extracts.

To compare the efficiency in GSL-sulfatase activity of these two sulfatases on intact GSLs,

1000 µg of crude extracts of the recombinant SUL2 enzyme and 100 µg of the purified H.

pomatia sulfatase were used in the DEAE-Sephadex on column desulfation procedure with

intact GSLs (1 mM). The results showed that the recombinant SUL2 enzyme was able to

desulfate intact GSLs with less efficiency than H. pomatia sulfatase (Figure 5.19). A negative

control including cell-free extracts from BL21(DE3) without recombinant enzyme expression of

SUL2 enzyme showed no desulfation of these GSLs tested as no DS-GSLs were detected on HPLC

chromatograms (Appendix III). This indicates that BL21(DE3) host did not exhibit endogenous

GSL-sulfatase activity.

290

0 5 10 15 20

50

100

150

200

250

300

0 5 10 15 20

75

150

225

300

375

450

5 10 15 20 0 5 10

Abs

orba

nce

at 2

29nm

Retention time (min)

SUL2 sulfatase HP sulfatase

Abs

orba

nce

at 2

29nm

Retention time (min)

SUL2 sulfatase HP sulfatase

350

300

250

200

150

100

Abs

orba

nce

at 2

29nm

Retention time (min)

SUL2 sulfatase HP sulfatase

0

DC

B

320

240

160

400

80

Abs

orba

nce

at 2

29nm

Retention time (min)

SUL2 sulfatase HP sulfatase

0

A

Figure 5.19 HPLC chromatograms showing desulfation of intact GSLs by crude extracts of the recombinant SUL2 enzyme and the purified H. pomatia sulfatase on the DEAE-Sephadex column. Either crude extracts of the recombinant SUL2 enzyme, 1000 µg (black line) or the purified H. pomatia (HP) sulfatase, 100 µg (red line) was used to desulfate 1 mM GSL on the DEAE-Sephadex column for 8 h at 30˚C under aerobic conditions. Intact GSLs used are as follows; (A) Sinigrin. (B) Glucoerucin. (C) Gluconasturtiin and (D) Glucoiberin. The figures are representatives of triplicates.

The apparent specific activity of the recombinant SUL2 enzyme for intact GSL substrates

plus pNCS substrate is shown in Table 5.8. Its relative activity (%) was compared with H.

pomatia sulfatase. Glucoiberin with the highest polarity (in the side chain) has the lowest

specific activity and the lowest relative activity to H. pomatia sulfatase. The specific activity was

found in the descending order according to the degree of side chain polarity of GSLs as follows;

sinigrin > glucoerucin > gluconasturtiin > glucoiberin. This suggests that the side chain property

of GSL may influence the efficiency of GSL-sulfatase activity of the recombinant SUL2 enzyme.

291

However, H. pomatia sulfatase seemed to desulfate these GSLs with the same efficiency as seen

from similar peak areas of DS-GSLs produced on-column from HPLC chromatograms (Figure

5.19).

Table 5.8 Specific activity and relative activity of crude extracts of the recombinant SUL2

enzyme in desulfation of intact GSL substrates

Substrates (1 mM) Apparent specific activity (U/mg)* Relative activity** (%)

pNCS 10.4 ± 0.25 51.0 ± 1.4

Sinigrin 0.71 ± 0.15 4.4 ± 1.1

Gluconasturtiin 0.38 ± 0.09 3.5 ± 0.7

Glucoerucin 0.66 ± 0.11 4.1 ± 0.3

Glucoiberin 0.11 ± 0.03 0.6 ± 0.2

*The values determined are apparent specific activities since the SUL2 enzymes are not pure. Aerobic incubation of the partially purified recombinant SUL2 enzyme (approximately 100 µg) with 1 mM pNCS in 50 mM sodium acetate buffer pH 6.0 was incubated for 15 min at 30˚C. One unit (U) of sulfatase was defined as the amount of enzyme liberating 1 µmol min-1 of pNC. For desulfation of 1 mM intact GSL substrate on the DEAE-Sephadex column for 8 h at 30˚C, crude extracts of the recombinant SUL2 enzyme (approximately 1000 µg) was used instead. One unit (U) of sulfatase was defined as the amount of enzyme liberating 1 nmol min-1 of DS-GSL. **Activity produced by crude extracts of the recombinant SUL2 enzyme (approximately 1000 µg) in (%) relative to that produced by the purified H. pomatia sulfatase (approximately 100 µg). Values were means ± SD of triplicates.

5.3.6 Apparent enzyme activities of the recombinant SUL2 enzyme

Since very little is known about the enzyme activities of bacterial sulfatases known to

date, the aim of this chapter was to determine these parameters of the recombinant SUL2

enzyme using pNCS and common intact GSL substrates. Note that the partially purified

recombinant SUL2 enzyme was used for pNCS substrate while crude extracts were used for GSL

substrates in this experiment. The Michaelis-Menten kinetic curve and the Lineweaver–Burk

292

plot showed that the arylsulfatase activity of the recombinant SUL2 enzyme has the Michaelis–

Menten constant (Km) of 1.09 mM and the Vmax of 25.1 U/mg for pNCS (Figure 5.20).

Figure 5.20 Apparent enzyme activities of the partially purified recombinant SUL2 enzyme for pNCS substrate. (A) Michaelis-Menten curve of the recombinant SUL2 enzyme for pNCS. (B) Lineweaver–Burk plot of graph (A). Arylsulfatase activity was measured by monitoring the release of pNC from pNCS by the recombinant SUL2 enzyme (100 µg) in 50 mM sodium acetate buffer pH 6.0 for 5 min at 30˚C. Values are means ± SD of triplicates.

In addition to pNCS substrate, the enzyme activities for GSLs tested are presented in

Table 5.9. To determine the apparent enzyme activities i.e. Km and Vmax, in vitro incubations of

crude extracts of the recombinant SUL2 enzyme with varied concentrations of GSLs in

Eppendorf   tubes   were   carried   out   at   30˚C   for   8   h.   After   that,   supernatant from a reaction

mixture was collected for the on-column desulfation by the purified H. pomatia (HP) sulfatase.

The flow-through was collected from the DEAE-Sephadex column and analyzed by HPLC to

determine the amounts of GSLs remaining in the in vitro incubations. The negative controls

containing only GSLs without recombinant SUL2 enzymes were also processed to determine the

amounts of GSLs present in the in vitro incubations, and therefore the amounts of GSLs

disappearing (as DS-GSLs) due to GSL-sulfatase activity of the recombinant SUL2 enzyme in the

in vitro incubations were calculated.

293

Table 5.9 Apparent enzyme activities* of the recombinant SUL2 on different substrates

Substrate Km (mM) Vmax (U/mg) kcat (s-1) Kcat/Km (M-1 s-1)

pNCS 1.09 25.1 232 2.13 X 105

Sinigrin 9.68 11.1 9.94 x 10-3 1.02

Gluconasturtiin 15.7 8.81 7.87 x 10-3 0.502

Glucoiberin 0.480 0.169 1.51 x 10-4 0.314

*The values determined are apparent enzyme activities since the SUL2 enzymes are not pure. Aerobic incubation of the partially purified recombinant SUL2 enzyme (100 µg) with 1 mM pNCS in 50 mM sodium acetate buffer pH 6.0 for 15 min at 30˚C. One unit (U) of sulfatase was defined as the amount of enzyme liberating 1 µmol min-1 of pNC product. For desulfation of various concentrations of intact GSL substrates on the DEAE-Sephadex column for 8 h at 30˚C, crude extracts of the recombinant SUL2 enzyme (1000 µg) was used instead. One unit (U) of sulfatase was defined as the amount of enzyme liberating 1 nmol min-1 of DS-GSL. Values of Km and Vmax are estimated with a 95% confidence using GraphPad Prism 6.

The Vmax and catalytic efficiency (Kcat/Km) of the recombinant SUL2 enzyme were found

in the same descending order; pNCS > sinigrin > gluconasturtiin > glucoiberin, whereas the

order of Km was gluconasturtiin > sinigrin > glucoiberin > pNCS (Table 5.6). The highest Vmax with

the lowest Km of the recombinant SUL2 enzyme were found for pNCS substrate indicating the

SUL2 enzyme was most efficient in desulfating this synthetic substrate. Amongst the GSLs

tested, the most preferred substrate was sinigrin, and the least favored GSL was glucoiberin.

The corresponding Michaelis-Menten curves and Lineweaver-Burk plots of crude extract

of the recombinant SUL2 enzyme for all GSLs tested are shown in Figure 5.21.

294

Figure 5.21 Apparent enzyme activities of crude extracts of the recombinant SUL2 enzyme for GSL substrates. (A) Michaelis-Menten curve of the recombinant SUL2 enzyme for sinigrin. (B) Lineweaver–Burk plot of graph (A); (C) Michaelis-Menten curve of SUL2 for gluconasturtiin; (D) Lineweaver–Burk plot of graph (C); (E) Michaelis-Menten curve of SUL2 for glucoiberin; (F) Lineweaver–Burk plot of graph (E). Values are means ± SD of triplicates. These graphs were generated using GrapPad Prism 6.

295

5.3.7 Effects of compounds on arylsulfatase activity of the recombinant SUL2 enzyme

It was reported that H. pomatia sulfatase was activated by Cd2+, but was inhibited by

Cu2+ (Tokheim et al., 2005). Surprisingly, not many reports were found regarding the effect of

certain ions on bacterial sulfatases. Therefore, various compounds including Na2SO4, NaHSO4,

CoCl2, CaCl2, MgCl2, FeSO4, NiCl2, MnCl2, FeCl3 of 1.0 mM were tested for thier effects on

arylsulfatase activity of the recombinant SUL2 enzyme pr pNCS substrate.

Essentially, no effect was observed with most of the compounds and metal ions tested,

except for Fe2+, NaHSO4 and Na2SO4 that reduced arylsulfatase activity by 20, 75, and 50%,

respectively (Figure 5.22).

Figure 5.22 Effect of compounds on arylsulfatase activity of the partially purified recombinant SUL2 enzyme. Either 1 mM of Na2SO4, NaHSO4, CoCl2, CaCl2, MgCl2, FeSO4, NiCl2, MnCl2, FeCl3 was added with the partially purified recombinant SUL2 enzyme (100 µg)  in  50  mM  sodium  acetate  buffer  pH  6.0  at  30˚C  for 15 min. Relative activity (%) was shown. The control sample contained only the recombinant SUL2 enzyme without any compounds. Values are means ± SD of triplicates.

Since GSL-sulfatase activity of the recombinant SUL2 enzyme was found in crude

extracts (not in the purified fraction), it was highly speculated that metal ions or co-proteins

present in the crude extracts may be required to activate GSL-sulfatase activity of the

recombinant SUL2 enzyme in the purified fraction. This hypothesis still remains to be tested.

296

5.3.8 Purification of the recombinant GH3 enzyme from E. casseliflavus NCCP-53

Since three of the candidate recombinant enzymes in Chapter 4 showed β-O-glucosidase

activity for pNPG substrate (Chapter 4, section 4.3.4.2), the recombinant GH3 enzyme with the

highest  β-O-glucosidase activity was chosen to be studied in more details in this chapter.

The Ni2+-affinity chromatography was used to purify the recombinant GH3 enzyme as

previously described as per the recombinant SUL2 enzyme. The elution fractions were analyzed

on SDS-PAGE (Figure 5.23A). A soluble recombinant GH3 enzyme with a molecular weight of

approximately 79 kDa (as predicted by UNIPROT) was found in two elution fractions (E1 and E2).

It was clear that the purity of each fraction was of high purity (> 90%). These fractions were

much purer than the elution fractions of the recombinant SUL2 enzyme purified by the same

method with the same buffer conditions. The reason for the differences in these results is not

known. The activity of these elution fractions was determined using β-O-glucosidase activity

assay as previously described (Chapter 4, section 4.2.19). The relative activity of these fractions

is shown in Figure 5.23B.

Figure 5.23 Purification of the recombinant GH3 enzyme expressed in BL21(DE3). (A) The 4-12% SDS-PAGE analysis of protein fractions obtained from Ni2+-affinity chromatography. Lane S, the supernatant from recombinant bacteria expressing the recombinant GH3 enzyme; lane FT, the flow-through fraction during the purification; lane W, the wash fraction; lane M, PageRuler pre-stained protein ladders (ThermoScientific, UK); lane E1 and E2, elution fraction no. 1 and 2 containing the band of GH3 enzyme (79 kDa) is indicated by an arrow. (B) Elution fractions (10 µL) with relative β-O-glucosidase activity using 1 mM pNPG substrate.

The purification scheme of the recombinant GH3 enzyme by Ni2+-affinity column

chromatography is shown in Table 5.10.

297

Table 5.10 Purification scheme of the recombinant GH3 enzyme from Ni2+-affinity column chromatography

Purification Step Total activity (U)

Total protein (mg)

Specific activity (U/mg)

Purity (Fold) Yield (%)

Crude extract 382 212 1.8 1 100 Ni2+ affinity column

chromatography 187 8 23.0 13 49

One unit (U) of GH3 enzyme was defined as the amount of enzyme liberating 1 µmol min-1 of pNP product

5.3.9 Temperature and pH optima of the recombinant GH3 enzyme

The temperature and pH optima of the recombinant GH3 enzyme for pNPG substrate

were determined using β-O-glucosidase activity assay as previously described (Chapter 4,

section 4.2.19). The β-O-glucosidase activity of the recombinant GH3 enzyme was found to be

optimal at pH 7.0 and 37°C (Figure 5.24A and 5.24B) in 0.1 M citrate phosphate buffer.

Figure 5.24 The pH and temperature optima of the recombinant GH3 enzyme. (A) Effect of pH on β-O-glucosidase activity using 1 mM pNPG as a substrate. (B) Effect of temperature. Enzyme activity was expressed as the percentage of the activity at 37˚C   and  pH 7.0 in 100 mM citrate phosphate buffer which was defined as 100%. Purified recombinant GH3 enzyme (100 µg) was used in each sample. Values were means of triplicates.

5.3.10 Effects of metal ions on β-O-glucosidase activity of the recombinant GH3 enzyme

To determine whether β-O-glucosidase activity of the purified recombinant GH3 enzyme

is influenced by metal ions, various metal ions including CoCl2, CaCl2, MgCl2, FeSO4, NiCl2, MnCl2,

FeCl3 of 1.0 mM were tested using 1 mM pNPG substrate in 0.1 M citrate phosphate buffer pH

298

7.0. Essentially, no significant effect was observed with most metal ions tested, except for Mn2+

that promoted activity to 232% while Fe2+ declined activity to 84% (Table 5.11).

Table 5.11 Effects of metal ions on β-O-glucosidase activity of the recombinant GH3 enzyme

Metal cations Relative activity (%)*

No ion 100 (23.0 U/mg)

Fe3+ 97

Fe2+ 84

Mn2+ 232

Mg2+ 95

Ca2+ 95

Co2+ 99

Ni2+ 99 *Relative activity (%) was (%) activity relative to that of the purified recombinant GH3 enzyme without any ions added. Values are means of triplicates. Either 1 mM CoCl2, CaCl2, MgCl2, FeSO4, NiCl2, MnCl2, FeCl3 was added with the purified recombinant GH3 enzyme (10 µL) using 1 mM pNPG substrate in 0.1 M citrate phosphate buffer pH 7.0 at  37˚C  for  5  min.

5.3.11 NIT production from DS-GSLs by the recombinant GH3 enzyme

It was reported that DS-GSLs were precursors for the  recombinant  β-O-glucosidase from

Caldocellum saccharolyticum to production of pure NITs (Wathelet et al., 2001). Whether this

holds true for the recombinant GH3 enzyme exhibiting β-O-glucosidase activity was to be

determined in the following experiments. It is known that NIT production from the metabolism

of intact GSLs by bacterial cells only occurred in the culture broths (Chapter 2, section 2.3.3)

and 0.1 M citrate phosphate buffer pH 7.0 with the presence of 5 mM Fe2+ ions for E. coli

O83:H1 NRH 857C resting cells (Chapter 2, section 2.3.7). Therefore, DS-glucoraphanin, DS-

glucoerucin and DS-gluconaturtiin (1 mM) were used as substrates for the recombinant GH3

enzyme of E. casseliflavus NCCP-53 (100 µg) in both NB broth and 0.1 M citrate phosphate

buffer pH 7.0 with the presence of 1 mM Fe2+ or Mn2+ (as found to promote β-O-glucosidase

activity of GH3 in Table 5.8)

299

The negative controls containing either DS-GSL alone or the GH3 enzyme alone in NB

broth or the buffer showed no NIT production suggesting that these DS-GSLs were stable in

experimental conditions and not degradable to NIT (Appendix IV). There was no NIT production

from DS-glucoraphanin in either NB broth or the buffer (Figure 5.25A). This finding is in

agreement with the previous results showing that there was no NIT production from the

metabolism of intact glucoraphanin by E. casseliflavus NCCP-53 (Chapter 2, section 2.3.3). The

lack of NIT production from glucoraphanin may be due to inability of sulfatase of E. casseliflavus

NCCP-53, if any, to desulfate glucoraphanin to produce DS- glucoraphanin as a substrate for the

GH3 enzyme to produce NIT. The presence of the sulfoxide group of glucoraphanin may present

steric effects that make its hydrolysis by the GH3 enzyme difficult.

Figure 5.25 GC-MS chromatograms showing NIT production from DS-GSLs by the purified recombinant GH3 enzyme in NB broths. (A) No NIT product from DS-glucoraphanin by GH3 enzyme. (B) Erucin NIT, 1 was produced (17.44 min) from DS-glucoerucin by GH3 enzyme. (C) Phenethyl NIT, 2 was produced (18.57 min) from DS-gluconasturtiin by GH3 enzyme. All samples were incubated with the purified recombinant GH3 enzyme (100 µg) in NB broths   (pH   7.0)   for   16   h   at   37˚C.   The   figures   are  representatives of triplicates.

It was found that erucin NIT production from DS-glucoerucin and phenethyl NIT from

300

DS-gluconasturtiin (Figure 5.25B and 5.25C) by the recombinant GH3 enzyme were observed in

both NB broth and 0.1 M citrate phosphate buffer pH 7.0 with the presence of 1 mM Fe2+ and

Mn2+ with higher NIT products in NB broth (Table 5.12). No ITC products were detected as

expected. This result suggests that the presence of either Fe2+ or Mn2+ is required for NIT

production from DS-GSL hydrolysis by the recombinant GH3 enzyme in the buffer.

Table 5.12 NIT productions from DS-GSLs by GH3 enzyme

Substrate (500 µM) Product Nitrile production (µM)

Buffera LBb Fe2+ Mn2+ No ion

Desulfo-glucoraphanin ND ND ND ND

Desulfo-erucin ERN NIT 47.1 ± 4.06 63.9 ± 2.41 66.1 ± 3.56 Desulfo-gluconasturtiin PNIT 28.4 ± 7.76 43.3 ± 9.04 80.5 ± 1.57

a100 mM citrate phosphate buffer pH 7.0 with the presence of 1 mM metal ions. bLB broth pH 7.0 without any metal ion. Values are means ± SD of triplicates. ERN NIT, Erucin nitrile; PNIT, Phenethyl nitrile; ND, Not detected. 5.3.12 NIT production from intact GSLs by sequential action of the recombinant SUL2 enzyme

and the recombinant GH3 enzyme

It was reported that intact GSL was desulfated by H. pomatia sulfatase to produce DS-

GSL which was then hydrolyzed by the recombinant β-O-glucosidase from Caldocellum

saccharolyticum to produce D-glucose, sulfate ion and corresponding NIT product (Galletti et al.,

2008; Kopycki et al., 2011). To determine whether this holds true for both the recombinant GH3

enzyme and the recombinant SUL2, intact GSLs were used as substrates in the reaction tubes

containing these two enzymes in NB broths for 16 h at 30˚C. The NB broth was used in this

experiment as assumingly it contains any co-factors necessary for NIT production. The negative

controls containing either intact GSL alone or both enzymes alone showed no NIT product

suggesting that intact GSL was stable in experimental conditions and not degradable to NIT

(Appendix IV).

301

Similar to the previous experiment (Section 5.3.10), erucin NIT production from

glucoerucin and phenethyl NIT from gluconasturtiin were observed (Table 5.13). There was no

NIT production from glucoraphanin (Table 5.13). However, the concentrations of NIT products

in this experiment (Table 5.13) were lower from those in Table 5.12. The lower production of

NIT may be due to the need for intact GSL substrates to be desulfated by the recombinant SUL2

enzyme in comparison with the readily available DS-GSL substrates used in the previous

experiments.

Table 5.13 Nitrile productions from glucosinolates by sequential action of the SUL2 enzyme and GH3 enzyme

GSL (1 mM) DS-GSL (µM)a NIT product NIT concentration (µM)b

Percentage product (%)c

Glucoraphanin ND ND ND ND

Glucoerucin 141 ± 5 (14%) Erucin nitrile 71 ± 8 50 ± 4

Gluconasturtiin 117 ± 5 (12%) Phenethyl nitrile 65 ± 9 55 ± 10

Reaction mixtures of each 1 mM glucosinolate substrate were incubated with purified GH3 enzyme (100 µg) and crude extracts of SUL2 enzyme (1000 µg) in 1 mL LB broths pH 7.0 for  16  h  at  30˚C;  ND,  Not  detected. Values are means ± SD of triplicates. aDesulfo-glucosinolates produced from desulfation of glucosinolates by crude extracts of SUL2 enzyme were determined by HPLC analysis. Numbers in brackets indicate the production of DS-GSL (mol) in (%) relative to the initial dose of GSL substrate (mol). bNitriles produced from desulfo-glucosinolates by rGH3 enzyme were determined by GC-MS anaylysis. cProduction of nitriles (nmol) in (%) product relative to the amounts of desulfo-glucosinolate intermediates (nmol) produced from glucosinolate substrates.

The DS-GSL production as a result of desulfation of intact GSL (1 mM) by crude extracts

of the recombinant SUL2 enzyme in the reaction tubes was shown to be 14 and 12% of the

initial amounts of glucoerucin and gluconasturtiin substrates, respectively (Table 5.13). This

result indicates that GSL-sulfatase activity of crude extract of the recombinant SUL2 enzyme can

be found as free enzyme in solution in a reaction tube and also as immobilized enzyme on the

DEAE-Sephadex column. The percentage NIT product from DS-GSL metabolism as a result of the

302

reaction by the purified recombinant GH3 enzyme was shown to be 50 and 55% of the

corresponding DS-glucoerucin and DS-gluconasturtiin, respectively produced from the starting

intact GSL substrates by the recombinant SUL2 enzyme (Table 5.13). This suggests that the

sequential transformation from intact GSLs to DS-GSLs by SUL2 enzyme and then from DS-GSLs

to NITs by GH3 enzyme was possible, but far from 100% efficiency.

303

5.4 Summary of key findings

The summary of key findings in this chapter is shown in Figure 5.26:

Figure 5.26 Reaction catalyzed by the recombinant enzymes SUL2 and GH3.

5.5 Discussion

In this work, bacterial recombinant sulfatase from human gut bacterium with GSL-

sulfatase activity that can desulfate intact GSLs to produce corresponding DS-GSLs was reported

for the first time. Also, it was the first report to show bacterial recombinant β-glucosidases of

GH3 glycoside hydrolase family from another human gut bacterium with hydrolytic activity

towards DS-GSLs that produces corresponding NIT products. The sequential reactions of both

recombinant sulfatase and GH3 enzymes derived from different bacterial origins to desulfate

GSLs and produce DS-GSL for NIT production were reported for the first time.

The recombinant SUL2 enzyme expressed in E. coli BL21(DE3), which was originally

cloned from a human gut bacterium E. coli O83:H1 NRG 857C, was purified as a soluble and

active enzyme with the purity of 61% and has the molecular weight of 57 kDa by SDS–PAGE as

predicted. This enzyme is a member of a highly conserved sulfatase family as defined by a

signature sulfatase domain located toward its amino terminus. This protein is soluble in the cell

crude extracts suggesting its location at the periplasmic space or in the cytosol which is

consistent with the previous observations of bacterial arylsulfatases in Alteromonas

carrageenovora (Barbeyron et al., 1995) and Pseudomonas C12B (Fitzgerald & George, 1977).

Thus far, nearly all active, recombinantly expressed sulfatase reported in the literature

304

possesses a Cys within the active site sequence. Therefore, it seems likely that a Cys-specific

modifying machinery functionally exists in E. coli host. Although E. coli BL21(DE3) carries at least

one sulfatase-related gene, the aslA gene (Sofia et al., 1994), this species has not yet been

found to express active endogenous sulfatases. Conversion of Ser to FGly is catalyzed also by E.

coli as Klebsiella sulfatase can be expressed in E. coli as an active enzyme

To date, no bacterial sulfatase with GSL-sulfatase activity has been identified and

characterized. Notably, bacterial sulfatases appear to not generally accept GSLs as substrates (U.

Wittstock and B.A. Halkier, unpublished results). Also, only a little DS-sinigrin was produced

after 6 h or 12 h of incubation of sinigrin with rat intestinal microbiota suggesting its

glucosinolate-sulfatase activity, if any, may have a very low specificity toward sinigrin (Lu et al.,

2011). However, the recombinant SUL2 enzyme in crude extracts was found to be able to

desulfate different intact GSLs to different degrees. Pure NIT production can be obtained

without the formation of ITCs from DS-GSLs hydrolyzed by the recombinant GH3 enzyme of E.

casseliflavus NCCP-53. In addition, pure NIT production can be obtained from intact GSLs

hydrolyzed by the sequential action of the recombinant SUL2 enzyme in crude extracts and the

recombinant GH3 enzyme. Although these two enzymes are of two different bacterial origins,

they seemed to work in concert to produce pure NIT products. This indicates that the functional

redundancy may exist across different strains of human gut bacteria. This sequential action of

both bacterial sulfatase and glycosyl hydrolase procedure seems to be easily applied for

producing NITs, which looks appealing both as starting building blocks for synthesizing new

bioactive structures, and as important analytical standards. For example, in a recent report, the

sequential action of two recombinant enzymes, a sulfatase from H. pomatia and   a   β-O-

glucosidase from Caldicellulosiruptor saccharolyticus on GSLs allowed synthesis of

thiohydroximates (TH) from a structurally broad array of abundant precursors including

homochiral compounds of demonstrated biological activity (Kopycki et al., 2011). This

chemoenzymatic synthetic route would allow access to many of the thiohydroximate core

structures of the 200 known naturally occurring GSLs, if not all. The enrichment of this group for

compounds can have possible pharmacological potential.

Similar to the Schistocerca gregaria (desert locust) sulfatase and H. pomatia (snail)

305

sulfatase, the recombinant SUL2 enzyme was able to desulfate both the synthetic substrate

pNCS (in its partially purified fraction) and the natural GSL substrates (in its crude extracts),

indicative of arylsulfatase activity as well as GSL-sulfatase activity. The desert locust sulfatase

displayed a sharp pH optimum at 6.5 (range tested: 2.5–11; half-maxima: 5.8 and 7.7) for

arylsulfatase activity (Falk & Gershenzon, 2007) while snail sulfatase has a pH optimum at 7.2

and   temperature   optimum   at   30˚C. Likewise, the partially purified recombinant SUL enzyme

exhibited optimal arylsulfatase   activity   at   a   similar   pH   and   temperature   at   pH   6.0   and   30˚C,  

respectively. In general, bacterial arylsulfatases can be categorized into two groups on the basis

of the pH optimum, with one group showing optimal activity at pH values of 6.5–7.1, and the

group at higher pH values of 8.3–9.0 (Kertesz et al., 1993; Kertesz, 1999). Arylsulfatases from

Salmonella typhimurium (Henderson & Milazzo, 1979) and Klebsiella pneumonia (Okamura et

al., 1977) have their optimal pH values of 6.7 and 7.5, respectively, falling into the first group,

whereas arylsulfatases from Pseudomonas aeruginosa (Beil et al., 1995) and Pseudomonas

testosterone (Tazuke et al., 1998) were classified as the second group. However, the

recombinant SUL2 enzyme in this study displayed its maximal activity at pH 6.0, which does not

fall into either group.

The arylsulfatase activity of the recombinant SUL2 enzyme was not influenced by metal

ions indicating that it does not need metal ions during the desulfating of pNCS substrate. This is

different from a bacterial sulfatase from Pseudoalteromonas carrageenovora with its

arylsulfatase activity promoted by Mg2+ ion (Kima et al., 2005). Also, arylsulfatase activity is

increased by Ca2+, a co-factor found in most of the active sites of sulfatases studied thus far

(Bond et al., 1997; Lukatela et al., 1998; Hernandez-Guzman et al., 2003; Boltes et al., 2001).

However, arylsulfatase activity of the recombinant SUL2 enzyme was inhibited by Na2HSO4 and

Na2SO4. This is in accordance with the previous report that snail sulfatase was inhibited by

K2SO4 when using p-nitrophenyl sulfate (pNPS) as a substrate (Roy & Williams, 1989). Similarly,

the desert locust sulfatase activity was inhibited strongly by Na2SO3 and slightly by Na2HPO4

and NaF (Falk & Gershenzon, 2007).

The recombinant SUL2 enzyme in crude extracts was noticeably much less efficient in

GSL desulfation than snail sulfatase and desert locust sulfatase. For GSL-sulfatase activity of the

306

latter two sulfatases, there seems to be no discrimination between different GSLs, and all are

desulfated at similar rates irrespective of the nature of the side chain. Interestingly, the

recombinant SUL2 enzyme desulfated intact GSLs tested in this work with much less efficiency

especially on GSLs with the more polar side chain such as glucoiberin. These differences may lie

in the different mechanisms in GSL metabolism in these three organisms. Desert locusts possess

a GSL-sulfatase in their alimentary canal, which catalyzes the cleavage of the sulfate group

efficiently from all GSLs tested, rendering them inert to myrosinase activity. The resulting DS-

GSLs are excreted in high quantities in the feces. GSL sulfatase was first described for larvae of

the diamondback (DMB) moth, Plutella xylostella (Ratzka et al. 2002). Both DMB and desert

locust sulfatases have broad specificity for aliphatic, indolic, and aromatic GSLs. The snail, as a

generalist herbivore, also possesses a GSL-sulfatase activity in its gut, but it is unclear whether

this enzyme is able to detoxify GSLs when the animal feeds on cruciferous plants (Ratzka et al.

2002). Human gut bacteria, however, may have a sophisticated system to deal with toxicity of

ITC products. Therefore, there is no need for bacterial sulfatase to be as efficient as sulfatases

found in locust and snail. In addition, it may well be that recombinant SUL2 enzyme was only

partially maturated (by posttranslational modifications) and thus the low GSL-sulfatase activity

of the enzyme. From the previous report, C. perfringens sulfatase expressed in E. coli BL21(DE3)

was only partially maturated as a large amount of the non-maturated peptide was detected

(Berteau et al., 2006) and this explained its low arylsulfatase activity. To determine whether the

rescombinant SUL enzyme undergoes the posttranslational modification of the Cys residue to a

FGly, matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) analysis is needed

for further experiments as previous performed (Schmidt et al., 1995; Dierks et al., 1997).

Interestingly, when the catalytic efficiencies (Kcat/Km) of various substrates were calculated for

desert locust sulfatase, the GSL-sulfatase activity is much more efficient (19.6 for sinigrin and

6.8 for (S)-2-hydroxy-3-butenyl GSL than the arylsulfatase activity (Vmax/Km = 0.017) on pNCS

substrate (Falk and Gershenzon, 2007). Snail sulfatase, on the other hand, has very similar

arylsulfatase and GSL-sulfatase activities for several substrates (Roy & Williams, 1989). In

contrast, our results showed that arylsulfatase activity for pNCS substrate (Kcat/Km = 6.36 x 105)

of the recombinant SUL2 enzyme was much more efficient than GSL-sulfatase activity (1.02 for

307

sinigrin, 0.502 for gluconasturtiin and 0.314 for glucoiberin). This also suggests that the

presence of the more polar side chain of GSL i.e. glucoiberin may result in a very low efficiency

of GSL sulfatase activity of the recombinant SUL2 enzyme in desulfating this particular GSL.

The native SUL2 enzyme was found to be constitutive. This finding is in agreement with

the previous study showing that snail sulfatase is constitutively expressed (Falk & Gershenzon,

2007) and so is bacterial sulfatase of C. perfringens (Berteau et al., 2006). However, there is no

report  on  the  inducibility  of  bacterial  β-O-glucosidases to date. The physico-chemical properties

of recombinant   β-O-glucosidase/glycosyl hydrolase are different from those of myrosinase,

which  is  a  β-thioglucosidase. Myrosinase showed weak activity on the synthetic substrate pNPG,

however it was totally inactive towards DS-GSLs (Palmieri et al., 1987). Interestingly, the

recombinant  GH3  enzyme  discovered  in  this  work  was  able  to  catalyse  both  the  β-O-glucosidic

bond on pNPG   substrate   and   also   the   β-thioglucosidic bond on certain DS-GSL substrates

obtained by sulfatase-assisted desulfation of natural intact GSLs (Chapter 2, section 2.2.4). This

finding clearly shows that this recombinant enzyme was able to catalyze the hydrolysis of not

only aryl-glycosides and disaccharides containing O-glucosidic bonds as previously reported

(Plant et al., 1988),  but  also  a  β-thioglycosidic bond, in contradiction with what was claimed in

the above-mentioned report. It was also shown that a series of DS-GSLs was transformed into

the corresponding NITs without any formation of ITCs or other products by exploiting this

glucosidase-catalyzed hydrolysis. Our result is in accordance with the reported use of

recombinant   β-O-glucosidase derived from the thermophile bacterium Caldicellulosiruptor

saccharolyticus in generating NITs from DS-GSLs (Wathelet et al. 2001). NIT production from

DS-GSLs by the recombinant GH3 enzyme was detected in NB broth and the buffer with the

presence of Fe2+ ions. In Chapter 2, NIT production was not detected when the metabolism of

intact GSL by bacterial resting cells was carried out in the citrate phosphate buffer pH 7.0 unless

Fe2+ ions are present. These results proved that the recombinant GH3 enzyme requires Fe2+ ions

to hydrolyze DS-GSLs for NIT production, and DS-GSLs seem to be intermediates in GSL

metabolism. This is different from the previous report that NIT production from DS-GSL by the

recombinant  β-O-glucosidase from C. saccharolyticum was found in 50 mM sodium phosphate

buffer pH 6.0 without the presence of Fe2+ ions (Wathelet et al., 2001). The recombinant GH3

308

enzyme  belongs  to  glycosyl  hydrolase  family  3  while  the  recombinant  β-O-glucosidase derived

from Caldocellum saccharolyticum belongs to glycosyl hydrolase family 1. The differences in the

structures and origins of these two enzymes may explain the different requirement of metal

ions in their activity/mechanism of NIT production. The recombinant   β-O-glucosidase of C.

saccharolyticus was capable of hydrolyzing a wide range of substrates (Love et al. 1988) and

able to generate NITs from DS-GSLs (Wathelet et al. 2001). Using DS-sinigrin as a substrate, the

maximum enzyme activity was found at pH 6.2 and in the temperature range of 65-70˚C.

To conclude, this is the first report of characterization of the bacterial sulfatase SUL2

with both GSL-sulfatase activity and arysulfatase activity. This is also the first report showing

that two bacterial enzymes of different origins, SUL2 enzyme from E. coli O83:H1 NRG 857C and

GH3 enzyme from E. casseliflavus NCCP-53 were able to work in concert to produce NIT

products from certain intact GSLs. The GH3 enzyme was characterized in this work and was

shown to be able to hydrolyze certain DS-GSLs to NIT products. This finding showed that human

gut microbiota is a valuable source for discovery of new enzymes.

309

Chapter 6: General Discussion 6.1 Summary of findings To date, the mechanism of GSL degradation in human gut bacteria still remains elusive

and very little is known about the proteins involved in the bacterial metabolism of GSLs.

Therefore, this study was aimed to enhance our understanding of the influence of human gut

bacteria on the metabolic fate of GSLs. It was also aimed to identify bacterial proteins involved

in the metabolism of GSLs and to characterize these proteins at a molecular and a biochemical

level. This was achieved by a series of experiments (i) time-course fermentation of different

GSLs and DS-GSLs by human gut bacteria; (ii) 2-DE-based proteomics analysis; (iii) molecular

cloning of putative genes of interest; and (iv) protein purification and in vitro enzyme activity

assays.

In Chapter 2, six GSL-metabolizing bacterial strains isolated from human faecal sample

were reported for GSL-degrading capacity. Most bacteria were capable of producing both NITs

and ITCs from GSLs however Enterococcus sp. C213 while Enterococcus faecium KT4S13

produced only NITs. The three bacteria studied in this work, L. agilis R16, E. casseliflavus NCCP-

53 and E. coli O83:H1 NRG 857C, were able to metabolize a range of GSLs with different

efficiencies. These results underscore the importance of human gut bacteria in digestive ITC

formation from GSL degradation. Inter-individual differences in the appearance of bacterial

strains exhibiting myrosinase activity may result in very different hydrolytic activities, and lead

to apparent discrepancies in ITC exposure among humans (Holst & Williamson, 2004). In the

same chapter, the putative bacterial GSL-degrading enzymes responsible for producing NITs

and ITCs are inducible by GSL in the resting cells experiments. NIT production by bacterial cells

was only found during bacterial fermentations of GSLs in culture broths. However, NIT

production by bacterial resting cells did not occur in the buffer unless Fe2+ ions as co-factors are

present. These metal ions are likely to be a pre-requisite for NIT production in the buffer. In

Chapter 4 and Chapter 5, the recombinant SUL2 enzyme in crude extracts was found to exhibit

GSL-sulfatase activity that desulfates intact GSLs to produce DS-GSLs which then act as

substrates for the recombinant GH3 enzyme of E. casseliflavus NCCP-53. The product of this

310

sequential enzymatic hydrolysis of intact GSL by these two enzymes of different bacterial

origins was pure NITs. This indicates that the mechanism of GSL metabolism may be very similar

among certain bacteria. Interestingly, E. coli O83:H1 NRG 857C produced methylthioalkyl ITCs

and NITs from methylsulfinylalkyl GSLs while E. casseliflavus NCCP-53 produced only

methylsulfinylalkyl ITCs from the same GSLs. This is explained by the presence of cytosolic

reductase enzyme in E. coli O83:H1 NRG 857C that reduces the sulfoxide group of GSLs to the

sulfide group. This GSL bioconversion is believed to facilitate further degradation to ITC and NIT

products by bacterial myrosinase-like enzyme or bacterial GSL-degrading enzyme. Since this

reductase enzyme was inducible by GSLs, it is hypothesized that there may be a sensor protein

that recognizes the structure of GSLs and leads to the expression of reductase. Likewise,

myrosinase, which is also inducible by GSL, may have a sensor protein to recognize the

structure of GSL which then leads to myrosinase expression. The proposed schematic

presentation of sulfoxide reduction on methylsulfinyl GSL by bacterial reductase and GSL

degradation by myrosinase of E. coli O83:H1 NRG 857C is shown in Figure 6.1.

Figure 6.1 Proposed scheme of myrosinase and reductase of E. coli O83:H1 NRG 857C induction by GSL. The GSL sensor protein may recognize the moiety of the GSL structure and then leads to to the transcription of myrosinase-like enzyme to degrade GSL(s) into ITC/NIT product(s) and possibly also leads to the transcription of reductase to reduce the sulfoxide group (if any) on GSL to the sulfide.

311

The schematic presentation of GSL and DS-GSL metabolism by human gut bacteria and

by the characterized bacterial recombinant SUL2 and GH3 enzymes under various conditions is

shown in Figure 6.2.

Figure 6.2 Summarized scheme of GSL/DS-GSL metabolism by human gut bacteria and by bacterial recombinant enzymes under various conditions. (A) Bacterial fermentations in culture broths. (B) Bacterial resting cells in 0.1 M citrate phosphate buffer pH 7.0. (C) Recombinant enzymes SUL2 and GH3 in buffer and broth. Putative enzymes are indicated in italic and characterized enzymes are in capital. ECO, E. coli O83:H1 NRG 857C; EC, E. casseliflavus NCCP-53.

312

From the cell-free extract experiments in Chapter 2, bacterial cell-free extracts showed

no myrosinase activity in vitro. Instead myrosinase activity was exclusive in intact cells in

fermentation/resting cell experiments suggesting that bacterial GSL-degrading enzyme activity

is membrane/cell-associated. Hypothetically, GSLs, as β-glucosides, may need to be

phosphorylated by phospho-kinases prior to degradation by myrosinase. The lack of

phosphorylation on GSL substrates used in vitro may explain the absence of myrosinase activity

in the cell-free extracts. Assumingly, bacterial phosphorylation system was no longer intact and

hence inactive.

The phosphoenolpyruvate (PEP)-dependent phosphotransferase system (PTS) is crucial

for carbohydrate acquisition in many bacteria (Postma et al. , 1993). The PTS involves several

proteins with a role in transportation of PTS-dependent carbohydrates in both Gram-positive

and Gram-negative bacteria. The PTS components consist of a heat-stable histidine protein

(HPr), a non-specific Enzyme I (EI) and a sugar-specific membrane associated Enzyme II (EII)

(Postma et al., 1993) with two cytoplasmic domains, EIIA and EIIB, and an integral membrane

domain EIIC (Saier et al., 1988). It is clear that the utilization of common dietary carbohydrates

has  been  facilitated  by  the  PTS,  but  its  role  in  β-glucoside utilization is much less clear (Cote et

al., 2000). Foods containing plant extracts tend to contain  β-glucosides, such as salicin, arbutin,

esculin and cellobiose. The translocation of these β-glucosides has been associated with the PTS

in certain bacteria including B. subtilis, E. coli, C. longisporum and E. chrysanthemi (Brown &

Thomson, 1998; Fox & Wilson, 1968; Hassouni et al., 1990; Le Coq et al., 1995).

It was discovered that BglK as a β-glucoside kinase (EC 2.7.1.85), present in many

species of bacteria including K. Pneumonia (Thompson et al., 2002), is separate and distinct

from glucokinase (EC 2.7.1.2). The role of BglK is to phosphorylate β-glucoside, and this

phosphorylated product is a substrate for the phospho-β-glucosidase (P-β-glucosidase, EC

3.2.1.86). It was also found that P-β-glucosidase purified from Fusobacterium mortiferum

hydrolyzed several P-β-glucosides, including the isomeric disaccharide phosphates, cellobiose-

6-phosphate, gentiobiose-6-phosphate, sophorose-6-phosphate, and laminaribiose-6-

phosphate, to yield glucose-6-phosphate and appropriate aglycones (Thompson et al., 1997).

313

Since GSLs are β-glucosides, it is thought that they may need to be phosphorylated by β-

glucoside kinase prior to being hydrolyzed by myrosinase which can recognize the

phosphorylation on GSL structure. There is also a possibility that myrosinase may be tightly

linked to the PTS system.

In Chapter 3, by using two-dimensional gel electrophoresis (2-DE) and liquid

chromatography mass spectrometry (LC-MS/MS) for the comparative analysis between GSL-

induced and non-induced cultures of L. agilis R16 and E. coli O83:H1 NRG 857C, upregulated

proteins that may be involved in the metabolism of GSLs by these bacteria were identified.

Some of those identified proteins belong to phosphotransferase system (PTS), glucokinase,

carbohydrate metabolism, glycosyl hydrolysis and oxidoreduction process. These results led to

the speculation that PTS system and/or β-glucoside kinase may be involved in the metabolism

of GSL and/or the activation of myrosinase in human gut bacteria.

From all the results in this work, the schematic presentation of GSL-metabolizing

mechanism in human gut bacteria has been proposed in Figure 6.3. In this scheme, the PTS

system catalyzes the synchronized uptake and phosphorylation of a GSL substrate. The PTS

comprises three proteins. In the cytoplasm, phosphoenolpyruvate (PEP) phosphorylates EI,

which then transfers the phosphoryl group to HPr. The phosphoryl group from HPr is

transferred to several EII proteins. Within EII, HPr donates the phosphoryl group to EIIA that

then transfers it to EIIB, whereupon EIIC initiates sugar translocation (Teplyakov et al., 2006).

This scheme also includes the roles of the bacterial reductase, sulfatase and glycosyl hydrolase

that were characterized in this work.

314

Figure 6.3 Proposed schematic presentation of GSL-metabolizing mechanism in human gut bacteria. (A) The phosphoenolpyruvate (PEP): sugar phosphotransferase system (PTS). See texts for details. (B) Sulfoxide GSL is reduced by reductase to sulfide GSL which is phosphorylated by putative β-glucoside kinase prior to GSL degradation to NIT and ITC products by putative myrosinase. These products are released to outside of the cells. (C) GSL is desulfated by sulfatase to produce DS-GSL that is a substrate for glycosyl hydrolase/β-O-glucosidase to produce pure NITs.

Interestingly, a transporter seems to be required in both the sequestration of the GSLs

in the haemolymph and the transport of the GSL-3-sulfate in the gut of Athalia rosae sawflies

(Opitz et al., 2011). A successful transport through the gut epithelia, presumably through the

action of modified glucose transporter(s) may be determined by the chemical properties of the

sugar moieties of these GSL metabolites (Opitz et al., 2011). Nevertheless, very little is known

about the transport of glycosidically bound metabolites in insects (Kuhn et al., 2004; Discher et

al., 2009), no transporter has yet been identified. Similar to bacteria, how specific involved

transporter(s) function in these sawflies and how the glucosides are transported remain to be

determined.

315

6.2 Future work

The results from this work have provided new interesting and important data for the

better understanding of GSL metabolism in human gut bacteria. However, there are more areas

that require further investigated as follows:

6.2.1 Identification of other GSL degradation metabolites from GSL metabolism

Since the total percentage products of all GSLs metabolized by the three bacteria never

reached 100% and the unaccounted loss was unknown, except for the instability of ITCs

(Chapter 2), this needs to be investigated further. Experiments like 1D and 2D TOCSY and gCOSY 1H NMR should be used both to elucidate the molecular structure of GSL derivatives and to

quantify the concentrations of metabolites. Different compounds of biochemical interest can

be analyzed simultaneously in NMR spectra, including GSL substrates and ITC/NIT degradation

products. Considering that the preparation of one sample takes 5 min and that recording a 1D 1H NMR spectrum and a 2D 1H NMR spectrum takes about 10 min and 60 min, respectively,

NMR spectroscopy is a very powerful technique and can be used routinely. Another advantage

is that this technique is without a priori hypothesis; consequently, unexpected metabolites,

such as the amine derivatives, can be detected. The 13C NMR can be also used for studying

carbon metabolism by the human digestive microbiota, using [6-13C] glucose/amino acid moiety

of a GSL as a substrate.

6.2.2 Further search for the putative bacterial GSL-degrading enzymes from other

bacteria

Since the gene encoding a bacterial myrosinase-like enzyme in human gut bacteria has

yet to be identified, a few approaches can be employed to identify bacterial GSL-degrading

gene/protein as follows:

(i) Mutants of GSL-degrading bacteria can be chemically or genetically created. These

mutants will be screened for lack of myrosinase activity. Based on the use of specific growth

media with sinigrin, a simple microtitre plate method (Bioscreen C system) will be developed.

316

Myrosinase activity will be determined by following the decline in sinigrin concentration at

A230nm using a UV multiwell plate reader. By using this method, it will be possible to screen over

400 bacterial colonies. Those that show no myrosinase activity will be subject to genome

sequencing to identify the mutated gene that may encode for myrosinase.

(ii) The culture and sequence-independent approach of stable isotope probing (SIP) is

worth a try. SIP involves the 13C or 14C-labeling of the compound of interest; organisms able to

degrade the compound will assimilate the labeled carbon into their biomass (DNA, RNA, or fatty

acids). This provides a way to get at sequence and functional information from candidate

degraders (Radajewski et al., 2000). The 16S rRNA-based SIP can be used to assign metabolic

activities such as glucose and starch fermentation to specific members of the gut microbiota

(Egert et al., 2007; Kovatcheva-Datchary et al., 2009). SIP can also be used to monitor the

activity of specific members of the gut microbiota in response to changes in nutrient status by

monitoring de novo RNA synthesis (Reichardt et al., 2011).

(iii) Another strategy for screening microbial communities for genes involved in GSL

metabolism is a technique called substrate induced gene expression (SIGEX). SIGEX counts on

the fact that catabolic genes are often transcriptionally activated by their substrate (Uchiyama

& Watanabe, 2008) and in this case it is GSL. In SIGEX procedure, community DNA is extracted

from environmental samples, sheared to 5–10 kb in length, and then cloned into a green

fluorescent protein (GFP) expression vector. Fluorescence activated cell sorting (FACS) can then

be used to screen the resulting metagenomic library for clones that induce GFP upon growth in

media supplemented with GSL. This approach was successfully used to isolate a previously

characterized phenol degradation operon from Ralstonia eutropha, an organism isolated from

sludge, and aromatic-hydrocarbon responsive operons from petroleum-contaminated

groundwater (Uchiyama et al., 2004).

(vi) RNA-sequencing (RNA-seq) experiment can be performed to identify the genes with

transcriptional upregulation during GSL metabolism in human gut bacteria, and that may

potentially be a myrosinase gene. The traditional low-throughput expressed sequenced tag

(EST) sequencing by Sanger technology only detects the more abundant transcripts. However,

317

this RNA-seq approach has initiated the revelation of the dynamics and complex landscape of

the transcriptome from yeast to human at an unprecedented level of sensitivity and accuracy

(Martin & Wang, 2011). A near-complete snapshot of a transcriptome, including the rare

transcripts that have regulatory roles, can be achieved by a typical RNA-seq experiment with

the massive sequencing depths (100–1,000 reads per base pair of a transcript). In contrast to

microarrays, base-pair-level resolution, a much higher dynamic range of expression levels, and

de novo annotation can also be achieved by RNA-seq (Martin & Wang, 2011).

6.2.3 Determination of whether GSL-6-P is a substrate for bacterial GSL-degrading

enzyme in vitro

It is hypothesized that phosphorylation of GSL may be a pre-requisite for its degradation

by bacterial GSL-degrading enzyme in cell-free extract experiments. To test this hypothesis, GSL

must be phosphorylated at the C6-glucose moiety either enzymatically or chemically to

generate GSL-6-P. To  achive  this,  β-glucoside kinase (EC 2.7.1.85), extracted from K. Pneumonia

(Thompson et al., 2002) will be used to phosphorylate GSLs in the same manner it was done to

enzymatically phosphorylate β-glucosides, such as cellobiose, arbutin, salicin and esculin as

previously reported (Thompson et al., 2002). Although this has never been performed on GSLs

before, it is quite likely that β-glucoside kinase will phosphorylate GSL at C6. Interestingly, GSL

has a sulfate group in the aglycone, and this may present a problem for the simple procedure(s)

that was used to obtain other phosphorylated β-O-glycosides. Alternatively, chemical synthesis

would be an option although it would represent a challenging synthesis.

6.2.4 Purification of bacterial reductase

Since co-factors and optimal operating have been determined conditions for bacterial

reductase activity of E. coli O83:H1 NRG 857C, this will facilitate the purification of bacterial

reductase from its cell-free extracts. The cell-free extracts will be purified using ion-exchange

chromatography/gel filtration via FPLC machine. The collected fractions will be identified for

reductase activity upon the addition of NAD(P)H as a reducing co-factor plus Mg2+ ions. The

disappearance of NAD(P)H, as an indicator of reductase activity, can be monitored using UV/Vis

318

spectrometer. Once the fractions containing reductase are identified, they will be analyzed and

purified by SDS-PAGE until a single band of interest is obtained. This protein band will be

subjected to protein identification. The amino acid sequence will lead to the identification of

the gene responsible for bacterial reductase activity. Subsequently, the cloning of the gene will

be carried out followed by the characterization of the protein. This reductase enzyme may be of

importance for the reduction of xenobiotics.

6.3 Conclusion

This study has identified six bacterial strains from human gut microbiota capable of

degrading GSLs and producing ITCs and/or NITs as degradation products. Different product

profiles by different bacteria highlight the importance of human gut microbiota in contribution

to promote human health due to chemopreventive effects of ITCs. The mechanism of GSL

degradation by human gut bacteria has been made clearer owing to identification and

characterization of bacterial sulfatase, glycosyl hydrolase and reductase involved in the

metabolism of GSLs. These findings provide a better understanding of the role of human gut

bacteria on the metabolic fate of GSLs. This urges further exploration of human gut microbiota

as a source for novel metabolites and novel enzymatic activity.

319

References: Abu Shamat, M. (1993). The role of the gastrointestinal microflora in the metabolism of drugs.

International Journal of Pharmaceutics, 97(1), 1–13. Abu Sharmat, M., & Beckett, A. H. (1983). Glyceryl trinitrate: metabolism by the intestinal flora. J Pharm

Pharmacol, 35, 71P. Aebersold, R., & Mann, M. (2003). Mass spectrometry-based proteomics. Nature, 422(6928), 198–207. Agerbirk, N., Olsen, C. E., & Sørensen, H. (1998). Initial and final products, nitriles, and ascorbigens

produced in myrosinase-catalyzed hydrolysis of indole glucosinolates. Journal of Agricultural and Food Chemistry, 46(4), 1563–1571.

Altschul, S. F., Gish, W., Miller, W., Myers, E. W., & Lipman, D. J. (1990). Basic local alignment search tool. Journal of Molecular Biology, 215(3), 403–410.

Ambrosone, C. B., McCann, S. E., Freudenheim, J. L., Marshall, J. R., Zhang, Y., & Shields, P. G. (2004). Breast cancer risk in premenopausal women is inversely associated with consumption of broccoli, a source of isothiocyanates, but is not modified by GST genotype. The Journal of Nutrition, 134(5), 1134–1138.

Ambrosone, C. B., & Tang, L. (2009). Cruciferous vegetable intake and cancer prevention: Role of nutrigenetics. Cancer Prevention Research, 2(4), 298–300.

American Institute for Cancer Research. (2007). World Cancer Research Fund: Food, nutrition, physical activity, and the prevention of cancer: a global perspective. World Cancer Research Fund.

Andersen, J. S., & Mann, M. (2000). Functional genomics by mass spectrometry. FEBS Letters, 480(1), 25–31.

Angeloni, C., Leoncini, E., Malaguti, M., Angelini, S., Hrelia, P., & Hrelia, S. (2009). Modulation of Phase II enzymes by sulforaphane: Implications for its cardioprotective potential. Journal of Agricultural and Food Chemistry, 57(12), 5615–5622.

Arguello, L. G., Sensharma, D. K., Qiu, F., Nurtaeva, A., & Rassi, Z. E. (1999). High-performance liquid-phase separation of glycosides analytical and micropreparative HPLC combined with spectroscopic and enzymatic methods for generating a glucosinolate library. Journal-Aoac International, 82, 1115–1127.

Artimo, P., Jonnalagedda, M., Arnold, K., Baratin, D., Csardi, G., de Castro, E., et al. (2012). ExPASy: SIB bioinformatics resource portal. Nucleic Acids Research, 40(W1), W597–W603.

Atkinson, C., Frankenfeld, C. L., & Lampe, J. W. (2005). Gut bacterial metabolism of the soy isoflavone daidzein: Exploring the relevance to human health. Experimental Biology and Medicine, 230(3), 155–170.

Austin, F. L., Gent, C. A., & Wolff, I. A. (1968). Degradation of natural thioglucosides with ferrous salts. Journal of Agricultural and Food Chemistry, 16(5), 752–755.

Bäckhed, F., Ley, R. E., Sonnenburg, J. L., Peterson, D. A., & Gordon, J. I. (2005). Host-bacterial mutualism in the human intestine. Science, 307(5717), 1915–1920.

Bak, S., & Feyereisen, R. (2001). The Involvement of two P450 enzymes, CYP83B1 and CYP83A1, in auxin homeostasis and glucosinolate biosynthesis. Plant Physiology, 127(1), 108–118.

Ballabio, A., & Shapiro, L. J. (1995). Steroid sulfatase deficiency and X-linked ichthyosis. The Metabolic and Molecular Bases of Inherited Diseases (Scriver CR, Beaudet AL, Sly WS, Valle D, Eds) 7th Ed. New York: McGraw-Hill, Inc, 2999–3022.

Baneyx, F. (1999). Recombinant protein expression in Escherichia coli. Current Opinion in Biotechnology, 10(5), 411–421.

Barbeyron, T., Potin, P., Richard, C., Collin, O., & Kloareg, B. (1995). Arylsulphatase from Alteromonas carrageenovora. Microbiology, 141(11), 2897–2904.

320

Barrera, L. N., Cassidy, A., Johnson, I. T., Bao, Y., & Belshaw, N. J. (2012). Epigenetic and antioxidant effects of dietary isothiocyanates and selenium: potential implications for cancer chemoprevention. Proceedings of the Nutrition Society, 71(02), 237–245.

Bartel, B., & Fink, G. R. (1994). Differential regulation of an auxin-producing nitrilase gene family in Arabidopsis thaliana. Proceedings of the National Academy of Sciences of the United States of America, 91(14), 6649–6653.

Basit, A. W., & Lacey, L. F. (2001). Colonic metabolism of ranitidine: implications for its delivery and absorption. International Journal of Pharmaceutics, 227(1), 157–165.

Basten, G. P., Bao, Y., & Williamson, G. (2002). Sulforaphane and its glutathione conjugate but not sulforaphane nitrile induce UDP-glucuronosyl transferase (UGT1A1) and glutathione transferase (GSTA1) in cultured cells. Carcinogenesis, 23(8), 1399–1404.

Beil, S., Kehrli, H., James, P., Staudenmann, W., Cook, A. M., Leisinger, T., & Kertesz, M. A. (1995). Purification and characterization of the arylsulfatase synthesized by Pseudomonas aeruginosa PAO during  growth  in  sulfate-­‐free  medium  and  cloning  of  the  arylsulfatase  gene  (atsA). European Journal of Biochemistry, 229(2), 385–394.

Belenguer, A., Duncan, S. H., Calder, A. G., Holtrop, G., Louis, P., Lobley, G. E., & Flint, H. J. (2006). Two routes of metabolic cross-feeding between Bifidobacterium adolescentis and butyrate-producing anaerobes from the human gut. Applied and Environmental Microbiology, 72(5), 3593–3599.

Bellostas, N., Petersen, I. L., Sørensen, J. C., & Sørensen, H. (2008). A fast and gentle method for the isolation of myrosinase complexes from Brassicaceous seeds. Journal of Biochemical and Biophysical Methods, 70(6), 918–925.

Bellostas, N., Sørensen, A. D., Sørensen, J. C., & Sørensen, H. (2007). Fe2+-catalyzed formation of nitriles and thionamides from intact glucosinolates. Journal of natural products, 71(1), 76–80.

Bellostas, N., Sørensen, A. D., Sørensen, J. C., & Sørensen, H. (2009). Type and concentration of redox reagents influencing nitrile formation upon myrosinase (Brassica carinata)-catalyzed hydrolysis of glucosibarin. Journal of Molecular Catalysis B: Enzymatic, 57(1), 229–236.

Benjdia, A., Dehò, G., Rabot, S., & Berteau, O. (2007). First evidences for a third sulfatase maturation system in prokaryotes from E. coli aslB and ydeM deletion mutants. FEBS letters, 581(5), 1009–1014.

Bennett, R. N., Fred, A., & Kroon, P. A. (2004). Screening crucifer seeds as sources of specific intact glucosinolates using ion-pair high-performance liquid chromatography negative ion electrospray mass spectrometry. Journal of Agricultural and Food Chemistry, 52(3), 428–438.

Berggren, K., Steinberg, T. H., Lauber, W. M., Carroll, J. A., Lopez, M. F., Chernokalskaya, E., et al. (1999). A luminescent ruthenium complex for ultrasensitive detection of proteins immobilized on membrane supports. Analytical Biochemistry, 276(2), 129.

Bergmeyer, H.U. (1974). Methods of Enzymatic Analysis. Academic Press, NY. Bernardi, R., Finiguerra, M. G., Rossi, A. A., & Palmieri, S. (2003). Isolation and biochemical

characterization of a basic myrosinase from ripe Crambe abyssinica Seeds, highly specific for epi-progoitrin. Journal of Agricultural and Food Chemistry, 51(9), 2737–2744.

Bernardi, R., Negri, A., Ronchi, S., & Palmieri, S. (2000). Isolation of the epithiospecifier protein from oil-rape (Brassica napus ssp. oleifera) seed and its characterization. FEBS Letters, 467(2–3), 296–298.

Berrin, J. G., Czjzek, M., Kroon, P. A., McLauchlan, W. R., Puigserver, A., Williamson, G., & Juge, N. (2003). Substrate (aglycone) specificity of human cytosolic beta-glucosidase. Biochemical Journal, 373(Pt 1), 41.

Berteau, O., Guillot, A., Benjdia, A., & Rabot, S. (2006). A new type of bacterial sulfatase reveals a novel maturation pathway in prokaryotes. Journal of Biological Chemistry, 281(32), 22464–22470.

Berth, M., Moser, F. M., Kolbe, M., & Bernhardt, J. (2007). The state of the art in the analysis of two-dimensional gel electrophoresis images. Applied Microbiology and Biotechnology, 76(6), 1223–1243.

321

Bertl, E., Bartsch, H., & Gerhäuser, C. (2006). Inhibition of angiogenesis and endothelial cell functions are novel sulforaphane-mediated mechanisms in chemoprevention. Molecular Cancer Therapeutics, 5(3), 575–585.

Bhatia,   Y.,   Mishra,   S.,   &   Bisaria,   V.   S.   (2002).   Microbial   β-glucosidases: cloning, properties, and applications. Critical Reviews in Biotechnology, 22(4), 375–407.

Bheemreddy, R. M., & Jeffery, E. H. (2007). The metabolic fate of purified glucoraphanin in F344 rats. Journal of Agricultural and Food Chemistry, 55(8), 2861–2866.

Bianchini, F., & Vainio, H. (2004). Isothiocyanates in cancer prevention. Drug Metab Rev, 36, 655–67 Bischoff,  S.  C.  (2011).  “Gut  health”:  a  new  objective  in  medicine?  BMC Medicine, 9(1), 24. Bjellqvist, B., Ek, K., & Giorgio Righetti, P. (1982). Isoelectric focusing in immobilized pH gradients:

principle, methodology and some applications. Journal of Biochemical and Biophysical Methods, 6(4), 317–339.

Björkman, R., & Janson, J. C. (1972). Studies on myrosinases: I. Purification and characterization of a myrosinase from white mustard seed (Sinapis alba, L.). Biochimica et Biophysica Acta (BBA)-Enzymology, 276(2), 508–518.

Björkman, R., & Lönnerdal, B. (1973). Studies on myrosinases III. Enzymatic properties of myrosinases from Sinapis alba and Brassica napus seeds. Biochimica et Biophysica Acta (BBA)-Enzymology, 327(1), 121–131.

Blaut, M., & Clavel, T. (2007). Metabolic diversity of the intestinal microbiota: Implications for health and disease. The Journal of Nutrition, 137(3), 751S–755S.

Blaut, M., Schoefer, L., & Braune, A. (2003). Transformation of flavonoids by intestinal microorganisms. Int J Vitam Nutr Res, 73(2), 79–87–0087.

Bollard, M., Stribbling, S., Mitchell, S., & Caldwell, J. (1997). The disposition of allyl isothiocyanate in the rat and mouse. Food and Chemical Toxicology, 35(10–11), 933–943.

Boltes, I., Czapinska, H., Kahnert, A., Bülow, von, R., Dierks, T., Schmidt, B., et al. (2001). 1.3 Å Structure of arylsulfatase from Pseudomonas aeruginosa establishes the catalytic mechanism of sulfate ester cleavage in the sulfatase family. Structure, 9(6), 483–491.

Bond, C. S., Clements, P. R., Ashby, S. J., Collyer, C. A., Harrop, S. J., Hopwood, J. J., & Guss, J. M. (1997). Structure of a human lysosomal sulfatase. Structure, 5(2), 277–289.

Bones,  A.  M.   (1990).  Distribution  of  β-thioglucosidase activity in intact plants, cell and tissue cultures and regeneration plants of Brassica napus l. Journal of Experimental Botany, 41(6), 737–744.

Bones, A. M., & Rossiter, J. T. (1996). The myrosinase-glucosinolate system, its organisation and biochemistry. Physiologia Plantarum, 97(1), 194–208.

Bones, A. M., & Rossiter, J. T. (2006). The enzymic and chemically induced decomposition of glucosinolates. Phytochemistry, 67(11), 1053–1067.

Boreddy, S. R., Sahu, R. P., & Srivastava, S. K. (2011). Benzyl isothiocyanate suppresses pancreatic tumor angiogenesis and invasion by inhibiting HIF-α/VEGF/Rho-GTPases: Pivotal role of STAT-3. PLoS ONE, 6(10), e25799.

Bosetti, C., Filomeno, M., Riso, P., Polesel, J., Levi, F., Talamini, R., et al. (2012). Cruciferous vegetables and cancer risk in a network of case–control studies. Annals of Oncology, 23(8), 2198–2203.

Botti, M. G., Taylor, M. G., & Botting, N. P. (1995). Studies on the mechanism of myrosinase investigation of the effect of glycosyl acceptors on enzyme activity. Journal of Biological Chemistry, 270(35), 20530–20535.

Bourquin, L. D., Titgemeyer, E. C., & Fahey, G. C., Jr. (1993). Vegetable fiber fermentation by human fecal bacteria: cell wall polysaccharide disappearance and short-chain fatty acid production during in vitro fermentation and water-holding capacity of unfermented residues. The Journal of Nutrition, 123(5), 860-869.

Bouwman, F., Renes, J., & Mariman, E. (2004). A combination of protein profiling and isotopomer

322

analysis using matrix-assisted laser desorption/ionization-time of flight mass spectrometry reveals an active metabolism of the extracellular matrix of 3T3-L1 adipocytes. Proteomics, 4(12), 3855–3863.

Brabban, A. D., & Edwards, C. (1994). Isolation of glucosinolate degrading microorganisms and their potential for reducing the glucosinolate content of rapemeal. FEMS Microbiology Letters, 119(1–2), 83–88.

Brändén,   C.   I.   (1980).   Relation   between   structure   and   function   of   α/β–proteins. Quarterly reviews of biophysics, 13, 317.

Bridges, M., Jones, A. M., Bones, A. M., Hodgson, C., Cole, R., Bartlet, E., et al. (2002). Spatial organization of the glucosinolate–myrosinase system in Brassica specialist aphids is similar to that of the host plant. Proc. R. Soc. Lond. B, 269(1487), 187–191.

Brobey, R. K., Mei, F. C., Cheng, X., & Soong, L. (2006). Comparative two-dimensional gel electrophoresis maps for promastigotes of Leishmania amazonensis and Leishmania major. Brazilian Journal of Infectious Diseases, 10(1), 1–6.

Brocker, E. R., Benn, M. H., Lüthy, J., & Däniken, von, A. (1984). Metabolism and distribution of 3,4-epithiobutanenitrile in the rat. Food and Chemical Toxicology, 22(3), 227–232.

Brown, G. D., & Thomson, J. A. (1998). Isolation and characterisation of an aryl-β-D-glucoside uptake and utilisation system (abg) from the gram-positive ruminal Clostridium species C. longisporum. Molecular and General Genetics MGG, 257(2), 213–218.

Brown, P. D., & Morra, M. J. (1995). Glucosinolate-containing plant tissues as bioherbicides. Journal of Agricultural and Food Chemistry, 43(12), 3070–3074.

Brown, P. D., Tokuhisa, J. G., Reichelt, M., & Gershenzon, J. (2003). Variation of glucosinolate accumulation among different organs and developmental stages of Arabidopsis thaliana. Phytochemistry, 62(3), 471–481.

Brown, T. (2006). Gene cloning and DNA analysis: an introduction. Blackwell Publisher, Cambridge, MA. Brüsewitz, G., Cameron, B. D., Chasseaud, L. F., Görler, K., Hawkins, D. R., Koch, H., & Mennicke, W. H.

(1977). The metabolism of benzyl isothiocyanate and its cysteine conjugate. Biochemical Journal, 162(1), 99.

Burmeister, W. P. (2000). High resolution x-ray crystallography shows that ascorbate is a cofactor for myrosinase and substitutes for the function of the catalytic base. Journal of Biological Chemistry, 275(50), 39385–39393.

Burmeister, W. P., Cottaz, S., Driguez, H., Iori, R., Palmieri, S., & Henrissat, B. (1997). The crystal structures of Sinapis alba myrosinase and a covalent glycosyl–enzyme intermediate provide insights into the substrate recognition and active-site machinery of an S-glycosidase. Structure, 5(5), 663–676.

Burow, M., Bergner, A., Gershenzon, J., & Wittstock, U. (2006). Glucosinolate hydrolysis in Lepidium sativum––identification of the thiocyanate-forming protein. Plant Molecular Biology, 63(1), 49–61.

Burow, M., & Wittstock, U. (2009). Regulation and function of specifier proteins in plants. Phytochemistry Reviews, 8(1), 87–99.

Caldwell, J., & Hawksworth, G. M. (1973). The demethylation of methamphetamine by intestinal microflora. Journal of Pharmacy and Pharmacology, 25(5), 422–424.

Callaway, E. C., Zhang, Y., Chew, W., & Chow, H. H. S. (2004). Cellular accumulation of dietary anticarcinogenic isothiocyanates is followed by transporter-mediated export as dithiocarbamates. Cancer Letters, 204(1), 23–31.

Campbell,   L.   D.,   Słominski,   B.   A.,   Nugon-Baudon, L., & Rabot, S. (1995). Studies on intestinal tract glucosinolate content, xenobiotic metabolizing enzymes and thyroid status in germ-free and conventional rats fed rapeseed meal. In Proceedings of 9th International Rapeseed Congress, Cambridge (UK), 1, 209–211.

Cantarel, B. L., Coutinho, P. M., Rancurel, C., Bernard, T., Lombard, V., & Henrissat, B. (2009). The

323

Carbohydrate-Active EnZymes database (CAZy): an expert resource for glycogenomics. Nucleic Acids Research, 37(suppl 1), D233–D238.

Cantarel, B. L., Lombard, V., & Henrissat, B. (2012). PLOS ONE: Complex Carbohydrate Utilization by the Healthy Human Microbiome. PLoS ONE.

Cerone-McLernon, A. M., Winter, J., Mosbach, E. H., & Bokkenheuser, V. D. (1981). Side-chain cleavage of cortisol by fecal flora. Biochimica Et Biophysica Acta (BBA) - Lipids and Lipid Metabolism, 666(3), 341–347.

Chassard, C., & Bernalier-Donadille, A. (2005). H2 and acetate transfers during xylan fermentation between  a  butyrate-­‐producing  xylanolytic  species  and  hydrogenotrophic  microorganisms  from  the  human gut. FEMS Microbiology Letters, 254(1), 116–122.

Chen, S., & Halkier, B. A. (1999). Functional expression and characterization of the myrosinase myr1 from Brassica napus in Saccharomyces cerevisiae. Protein Expression and Purification, 17(3), 414–420.

Chen, W., Liu, F., Ling, Z., Tong, X., & Xiang, C. (2012). Human intestinal lumen and mucosa-associated microbiota in patients with colorectal cancer. PLoS ONE, 7(6), e39743.

Cheng, D. L., Hashimoto, K., & Uda, Y. (2004). In vitro digestion of sinigrin and glucotropaeolin by single strains of Bifidobacterium and identification of the digestive products. Food and Chemical Toxicology, 42(3), 351–357.

Cheung, K. L., Khor, T. O., & Kong, A. N. (2008). Synergistic effect of combination of phenethyl isothiocyanate and sulforaphane or curcumin and sulforaphane in the inhibition of inflammation. Pharmaceutical Research, 26(1), 224–231.

Cheung, K. L., & Kong, A. N. (2009). Molecular targets of dietary phenethyl isothiocyanate and sulforaphane for cancer chemoprevention. The AAPS Journal, 12(1), 87–97.

Chiao, J. W., Chung, F. L., Kancherla, R., Ahmed, T., Mittelman, A., & Conaway, C. C. (2002). Sulforaphane and its metabolite mediate growth arrest and apoptosis in human prostate cancer cells. International Journal of Oncology, 20(3), 631–636.

Chung, F. L., Jiao, D., Getahun, S. M., & Yu, M. C. (1998). A urinary biomarker for uptake of dietary isothiocyanates in humans. Cancer Epidemiology Biomarkers & Prevention, 7(2), 103–108.

Ciska, E., Martyniak-Przybyszewska, B., & Kozlowska, H. (2000). Content of glucosinolates in cruciferous vegetables grown at the same site for two years under different climatic conditions. Journal of Agricultural and Food Chemistry, 48(7), 2862–2867.

Clarke, D. B. (2010). Glucosinolates, structures and analysis in food. Anal. Methods, 2(4), 310–325. Clarridge, J. E. (2004). Impact of 16S rRNA gene sequence analysis for identification of bacteria on

clinical microbiology and infectious diseases. Clinical Microbiology Reviews, 17(4), 840–862. Claus, G. W. (1995). Understanding microbes: A laboratory textbook for microbiology. W.H. Freeman and

Company, N.Y. Clavel, T., Henderson, G., Alpert, C.-A., Philippe, C., Rigottier-Gois, L., Dore, J., & Blaut, M. (2005).

Intestinal bacterial communities that produce active estrogen-like compounds enterodiol and enterolactone in Humans. Applied and Environmental Microbiology, 71(10), 6077–6085.

Cohen, J. H., Kristal, A. R., & Stanford, J. L. (2000). Fruit and vegetable intakes and prostate cancer risk. Journal of the National Cancer Institute, 92(1), 61–68.

Cole, C. B., Fuller, R., Mallet, A. K., & Rowland, I. R. (1985). The influence of the host on expression of intestinal microbial enzyme activities involved in metabolism of foreign compounds. Journal of Applied Microbiology, 59(6), 549–553.

Combourieu, B., Elfoul, L., Delort, A. M., & Rabot, S. (2001). Identification of new derivatives of sinigrin and glucotropaeolin produced by the human digestive microflora using 1H NMR spectroscopy analysis of in vitro incubations. Drug Metabolism and Disposition, 29(11), 1440–1445.

Conaway, C. C., Getahun, S. M., Liebes, L. L., Pusateri, D. J., Topham, D. K. W., Botero-Omary, M., &

324

Chung, F. L. (2000). Disposition of glucosinolates and sulforaphane in humans after ingestion of steamed and fresh broccoli. Nutrition and Cancer, 38(2), 168–178.

Conaway, C. C., Jiao, D., & Chung, F. L. (1996). Inhibition of rat liver cytochrome P450 isozymes by isothiocyanates and their conjugates: a structure-activity relationship study. Carcinogenesis, 17(11), 2423–2427.

Conaway, C. C., Wang, C. X., Pittman, B., Yang, Y. M., Schwartz, J. E., Tian, D., et al. (2005). Phenethyl isothiocyanate and sulforaphane and their N-acetylcysteine conjugates Inhibit malignant progression of lung adenomas induced by tobacco carcinogens in A/J mice. Cancer Research, 65(18), 8548–8557.

Consortium, U. (2012). UniProt Consortium: Reorganizing the protein space at the Universal Protein Resource (Uniprot). Nucleic Acids Res, 40, D71–D75.

Cooper, D. A., Webb, D. R., & Peters, J. C. (1997). Evaluation of the potential for olestra to affect the availability of dietary phytochemicals. The Journal of Nutrition, 127(8), 1699S–1709S.

Corvest, V., Murrey, D. A., Hirasawa, M., Knaff, D. B., Guiard, B., & Hamel, P. P. (2012). The flavoprotein Cyc2p, a mitochondrial cytochrome c assembly factor, is a NAD(P)H-dependent haem reductase. Molecular Microbiology, 83(5), 968–980.

Cosma, M. P., Pepe, S., Annunziata, I., Newbold, R. F., Grompe, M., Parenti, G., & Ballabio, A. (2003). The multiple sulfatase deficiency gene encodes an essential and limiting factor for the activity of sulfatases. Cell, 113(4), 445–456.

Costabile, A., Klinder, A., & Fava, F. (2008). Whole-grain wheat breakfast cereal has a prebiotic effect on the human gut microbiota: a double-blind, placebo-controlled, crossover study. British Journal of Nutrition, 99(1), 110–120.

Cottaz, S., Henrissat, B., & Driguez, H. (1996). Mechanism-based inhibition and stereochemistry of glucosinolate hydrolysis by myrosinase. Biochemistry, 35(48), 15256–15259.

Cristoni, S., & Bernardi, L. R. (2004). Bioinformatics in mass spectrometry data analysis for proteomics studies. Expert Rev Proteomics, 1(4), 469–483.

Crooks, G. E., Hon, G., Chandonia, J. M., & Brenner, S. E. (2004). WebLogo: a sequence logo generator. Genome Research, 14(6), 1188–1190.

Cross, J., Foss, F., Rady, J., Macdonald, T., & Templeton, D. (2007). The isothiocyanate class of bioactive nutrients covalently inhibit the MEKK1 protein kinase. BMC Cancer, 7(1), 183.

Cryan, J. F., & Dinan, T. G. (2012). Mind-altering microorganisms: the impact of the gut microbiota on brain and behaviour. Nature Reviews Neuroscience, 13(10), 701–712.

Cui, W., Bai, Y., Miao, X., Luo, P., Chen, Q., Tan, Y., et al. (2012). Prevention of diabetic nephropathy by sulforaphane: Possible role of Nrf2 upregulation and activation. Oxidative Medicine and Cellular Longevity, 2012, 1–12.

Curreem, S. O. T., Watt, R. M., Lau, S. K. P., & Woo, P. C. Y. (2012). Two-dimensional gel electrophoresis in bacterial proteomics. Protein & Cell, 3(5), 346–363.

Dabek, M., McCrae, S. I., Stevens, V. J., Duncan, S. H., & Louis,  P.   (2008).  Distribution  of  β-­‐glucosidase  and   β-­‐glucuronidase   activity   and   of   β-­‐glucuronidase   gene   gus   in   human   colonic   bacteria.   FEMS Microbiology Ecology, 66(3), 487–495.

Dan, S., Marton, I., Dekel, M., Bravdo, B.-A., He, S., Withers, S. G., & Shoseyov, O. (2000). Cloning, expression, characterization, and nucleophile identification of family 3, Aspergillus niger β-glucosidase. Journal of Biological chemistry, 275:4973-4980.

Dash, P. K., Zhao, J., Orsi, S. A., Zhang, M., & Moore, A. N. (2009). Sulforaphane improves cognitive function administered following traumatic brain injury. Neuroscience Letters, 460(2), 103–107.

Davies, G. J., & Sinnott, M. L. (2008). Sorting the diverse: the sequence-based classifications of carbohydrate-active enzymes. Biochemical Journal, 382.

De Cruz, P., Prideaux, L., Wagner, J., Ng, S. C., McSweeney, C., Kirkwood, C., et al. (2012).

325

Characterization of the gastrointestinal microbiota in health and inflammatory bowel disease. Inflammatory Bowel Diseases, 18(2), 372–390.

De Jongh, K. S., Schofield, P. J., & Edwards, M. R. (1987). Kinetic mechanism of sheep liver NADPH-dependent aldehyde reductase. Biochemical Journal, 242, 143–150.

Dekker, M., Verkerk, R., & Jongen, W. M. F. (2000). Predictive modelling of health aspects in the food production chain: a case study on glucosinolates in cabbage. Trends in Food Science & Technology, 11(4–5), 174–181.

Delahunty,  C.,  &  Yates,  J.  R.,  III.  (2006).  Proteomics:  A  Shotgun  Approach  without  Two-­‐dimensional  Gels.  eLS. doi:10.1038/npg.els.0006197

Delgado, S., Suárez, A., Otero, L., & Mayo, B. (2004). Variation of microbiological and biochemical parameters in the faeces of two healthy people over a 15 day period. European Journal of Nutrition, 43(6), 375–380.

Delisle, G. J., & Milazzo, F. H. (1972). Characterization of arylsulfatase isoenzymes from Pseudomonas aeruginosa. Canadian Journal of Microbiology, 18(5), 561–568.

Del Maschio, A., Livio, M., Cerletti, C., & De Gaetano, G. (1984). Inhibition of human platelet cyclo-oxygenase activity by sulfinpyrazone and three of its metabolites. European Journal of Pharmacology, 101(3–4), 209–214.

de Torres Zabala, M., Grant, M., Bones, A. M., Bennett, R., Lim, Y. S., Kissen, R., & Rossiter, J. T. (2005). Characterisation of recombinant epithiospecifier protein and its over-expression in Arabidopsis thaliana. Phytochemistry, 66(8), 859–867.

Devaraj, S., Hemarajata, P., & Versalovic, J. (2013). The human gut microbiome and body metabolism: implications for obesity and diabetes. Clinical chemistry. doi:10.1373/clinchem.2012.187617

De Vos, M., Van Oosten, V. R., Van Poecke, R. M. P., Van Pelt, J. A., Pozo, M. J., Mueller, M. J., et al. (2005). Signal signature and transcriptome changes of Arabidopsis during pathogen and insect attack. Molecular Plant-Microbe Interactions, 18(9), 923–937.

Dierks, T., Lecca, M. R., Schlotterhose, P., Schmidt, B., & Figura, von, K. (1999). Sequence determinants directing conversion of cysteine to formylglycine in eukaryotic sulfatases. The EMBO Journal, 18(8), 2084–2091.

Dierks, T., Lecca, M. R., Schmidt, B., & Figura, von, K. (1998a). Conversion of cysteine to formylglycine in eukaryotic sulfatases occurs by a common mechanism in the endoplasmic reticulum. FEBS Letters, 423(1), 61–65.

Dierks, T., Miech, C., Hummerjohann, J., Schmidt, B., Kertesz, M. A., & Figura, von, K. (1998b). Posttranslational formation of formylglycine in prokaryotic sulfatases by modification of either cysteine or serine. Journal of Biological Chemistry, 273(40), 25560–25564.

Dierks, T., Schmidt, B., Borissenko, L. V., Peng, J., Preusser, A., Mariappan, M., & Figura, von, K. (2003). Multiple   sulfatase   deficiency   is   caused   by   mutations   in   the   gene   encoding   the   human   C   α-formylglycine generating enzyme. Cell, 113(4), 435–444.

Dierks, T., Schmidt, B., & Figura, von, K. (1997). Conversion of cysteine to formylglycine: a protein modification in the endoplasmic reticulum. Proceedings of the National Academy of Sciences of the United States of America, 94(22), 11963–11968.

Dinkova-Kostova, A. T., & Kostov, R. V. (2012). Glucosinolates and isothiocyanates in health and disease. Trends in Molecular Medicine, 18(6), 337–347.

Discher, S., Burse, A., Tolzin-Banasch, K., Heinemann, S. H., Pasteels, J. M., & Boland, W. (2009). A versatile transport network for sequestering and excreting plant glycosides in leaf beetles provides an evolutionary flexible defense strategy. ChemBioChem., 10(13), 2223–2229.

Dixon, D. P., Skipsey, M., & Edwards, R. (2010). Roles for glutathione transferases in plant secondary metabolism. Phytochemistry, 71(4), 338–350.

Dodgson, K. S., White, G. F., & Fitzgerald, J. W. (1982). Sulfatases of microbial origin. CRC Press.

326

Drobnica,   Ľ.,   Podhradský,  D.,  &  Gemeiner,  P.   (1975).   Kinetics  of   the   reactions  of   isothiocyanates  with  sulphide and thiols. Collection of Czechoslovak Chemical Communications, 40(12), 3688–3697.

Dubendorf, J. W., & Studier, F. W. (1991). Controlling basal expression in an inducible T7 expression system by blocking the target T7 promoter with lac repressor. Journal of Molecular Biology, 219(1), 45–59. Duncan,  A.  J.,  &  Milne,  J.  A.  (1992).  Effect  of  long-­‐term  intra-­‐ruminal  infusion  of  the  glucosinolate   metabolite allyl cyanide on the voluntary food intake and metabolism of lambs. J. Sci. Food Agric., 5(1), 9–14.

Duncan, A. J., Rabot, S., & Nugon-Baudon, L. (1997). Urinary Mercapturic Acids as Markers for the Determination of Isothiocyanate Release from Glucosinolates in Rats Fed a Cauliflower Diet. Journal of the Science of Food and Agriculture, 73(2), 214–220.

Duncan, S. H., Barcenilla, A., Stewart, C. S., Pryde, S. E., & Flint, H. J. (2002). Acetate utilization and butyryl coenzyme A (CoA): acetate-CoA transferase in butyrate-producing bacteria from the human large intestine. Applied and Environmental Microbiology, 68(10), 5186–5190.

Duncan, S. H., Louis, P., & Flint, H. J. (2004). Lactate-utilizing bacteria, isolated from human feces, that produce butyrate as a major fermentation product. Applied and Environmental Microbiology, 70(10), 5810–5817.

Eckburg, P. B., Bik, E. M., Bernstein, C. N., Purdom, E., Dethlefsen, L., Sargent, M., et al. (2005). Diversity of the human intestinal microbial flora. Science, 308(5728), 1635–1638.

Egert, M., De Graaf, A. A., Maathuis, A., de Waard, P., Plugge, C. M., Smidt, H., et al. (2007). Identification  of  glucose-­‐fermenting  bacteria  present  in  an  in vitro model of the human intestine by RNA-­‐stable  isotope  probing.  FEMS Microbiology Ecology, 60(1), 126–135.

Eiji, K., Tsuyoshi, Y., Takashi, T., Masaaki, O., & Tetsuya, K. (1994). Chiral inversion of drug: role of intestinal bacteria in the stereoselective sulphoxide reduction of flosequinan. Biochemical Pharmacology, 48(2), 237–243.

Elfoul, L., Rabot, S., Khelifa, N., Quinsac, A., Duguay, A., & Rimbault, A. (2001). Formation of allyl isothiocyanate from sinigrin in the digestive tract of rats monoassociated with a human colonic strain of Bacteroides thetaiotaomicron. FEMS Microbiology Letters, 197(1), 99–103.

Elmer, G. W., & Remmel, R. P. (1984). Role of the intestinal microflora in clonazepam metabolism in the rat. Xenobiotica, 14(11), 829–840.

Fahey, J. W., Haristoy, X., Dolan, P. M., Kensler, T. W., Scholtus, I., Stephenson, K. K., et al. (2002). Sulforaphane inhibits extracellular, intracellular, and antibiotic-resistant strains of Helicobacter pylori and prevents benzo[a]pyrene-induced stomach tumors. Proceedings of the National Academy of Sciences, 99(11), 7610–7615.

Fahey, J. W., Zalcmann, A. T., & Talalay, P. (2001). The chemical diversity and distribution of glucosinolates and isothiocyanates among plants. Phytochemistry, 56(1), 5–51.

Falk, K. L., & Gershenzon, J. (2007). The desert locust, Schistocerca gregaria, detoxifies the glucosinolates of Schouwia purpurea by desulfation. Journal of Chemical Ecology, 33(8), 1542–1555.

Fenwick, G. R., Eagles, J., & Self, R. (1982). The fast atom bombardment mass spectra of glucosinolates and glucosinolate mixtures. Organic Mass Spectrometry, 17(11), 544–546.

Fenwick, G. R., & Heaney, R. K. (1983). Glucosinolates and their breakdown products in cruciferous crops, foods and feedingstuffs. Food Chemistry, 11(4), 249–271.

Fenwick, G. R., Heaney, R. K., & Mawson, R. (1989). Glucosinolates cheeke. Toxicants of Plant Origin. Feskanich, D., Ziegler, R. G., Michaud, D. S., Giovannucci, E. L., Speizer, F. E., Willett, W. C., & Colditz, G.

A. (2000). Prospective study of fruit and vegetable consumption and risk of lung cancer among men and women. Journal of the National Cancer Institute, 92(22), 1812–1823.

Fey, J., Balleininger, M., Borissenko, L. V., Schmidt, B., Figura, von, K., & Dierks, T. (2001). Characterization of posttranslational formylglycine formation by luminal components of the

327

endoplasmic reticulum. Journal of Biological Chemistry, 276(50), 47021–47028. Figura, von, K, Gieselmann, V., & Jaeken, J. (1999). The Metabolic and Molecular Bases of Inherited

Disease, Vol. III, 8th Edn, McGraw-Hill, New York, 3695. Figura, von, K, Gieselmann, V., & Jaeken, J. (2001). Metabolic and molecular basis of inherited diseases.

Metachromatic Leukodystrophy, 3695–3724. Fincher,   G.,   Mark,   B.,   Brumer,   H.   “Glycoside   hydrolase   family   3”   in   CAZypedia, available at URL

http://www.cazypedia.org/, accessed 22 June 2013.

Fitzgerald, J. W., & George, J. R. (1977). Localization of arylsulfatase in Pseudomonas C12B. Applied and Environmental Microbiology, 34(1), 107–108.

Fimognari, C., & Hrelia, P. (2007). Sulforaphane as a promising molecule for fighting cancer. Mutat Res, 635(2-3), 90–104.

Fimognari, C., Nüsse, M., Berti, F., Iori, R., Cantelli-Forti, G., & Hrelia, P. (2004). A mixture of isothiocyanates induces cyclin B1-and p53-mediated cell-cycle arrest and apoptosis of human T lymphoblastoid cells. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis, 554(1), 205–214.

Fimognari, C., Turrini, E., Ferruzzi, L., Lenzi, M., & Hrelia, P. (2011). Natural isothiocyanates: Genotoxic potential versus chemoprevention. Mutation Research-Reviews in Mutation Research, 750(2), 107–131

Flint, H. J., Duncan, S. H., Scott, K. P., & Louis, P. (2007). Interactions and competition within the microbial community of the human colon: links between diet and health. Environmental Microbiology, 9(5), 1101–1111.

Flores, R., Shi, J., Gail, M. H., Gajer, P., Ravel, J., & Goedert, J. J. (2012). Association of fecal microbial diversity and taxonomy with selected enzymatic functions. PLoS ONE, 7(6), e39745.

Foo, H. L., Grønning, L. M., Goodenough, L., Bones, A. M., Danielsen, B. E., Whiting, D. A., & Rossiter, J. T. (2000). Purification and characterisation of epithiospecifier protein from Brassica napus: enzymic intramolecular sulphur addition within alkenyl thiohydroximates derived from alkenyl glucosinolate hydrolysis. FEBS Letters, 468(2), 243–246.

Fox, C. F., & Wilson, G. (1968). The role of a phosphoenolpyruvate-dependent kinase system in beta-glucoside catabolism in Escherichia coli. Proceedings of the National Academy of Sciences of the United States of America, 59(3), 988-995.

Fraher, M. H., O'Toole, P. W., & Quigley, E. M. M. (2012). Techniques used to characterize the gut microbiota: a guide for the clinician. Nature Reviews Gastroenterology & Hepatology, 9(6), 312–322.

Francis, F. D. R., Lognay, G., Wathelet, J. P., & Haubruge, E. (2002). Characterisation of aphid myrosinase and degradation studies of glucosinolates. Archives of Insect Biochemistry and Physiology, 50(4), 173–182.

Franco, B., Meroni, G., Parenti, G., Levilliers, J., Bernard, L., Gebbia, M., et al. (1995). A cluster of sulfatase genes on Xp22. 3: mutations in chondrodysplasia punctata (CDPX) and implications for warfarin embryopathy. Cell, 81(1), 15–25.

Fukuda, S., Toh, H., Hase, K., Oshima, K., Nakanishi, Y., Yoshimura, K., et al. (2011). Bifidobacteria can protect from enteropathogenic infection through production of acetate. Nature, 469(7331), 543–547.

Furuhashi, T., Miksik, I., Smrz, M., Germann, B., Nebija, D., Lachmann, B., & Noe, C. (2010). Comparison of aragonitic molluscan shell proteins. Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology, 155(2), 195–200.

Gadelle, D., Raibaud, P., & Sacquet, E. (1985). Beta-glucuronidase activities of intestinal bacteria determined both in vitro and in vivo in gnotobiotic rats. Applied and Environmental Microbiology, 49(3), 682–685.

328

Galletti, S., Bernardi, R., Leoni, O., Rollin, P., & Palmieri, S. (2001). Preparation and biological activity of four epiprogoitrin myrosinase-derived products. Journal of Agricultural and Food Chemistry, 49(1), 471–476.

Galletti, S., Sala, E., Leoni, O., Cinti, S., & Cerato, C. (2008). Aspergillus flavus transformation of glucosinolates  to  nitriles  by  an  arylsulfatase  and  a  β-thio-glucosidase. Soil Biology and Biochemistry, 40(9), 2170–2173.

Gamet-Payrastre, L., Li, P., Lumeau, S., Cassar, G., Dupont, M.-A., Chevolleau, S., et al. (2000). Sulforaphane, a naturally occurring isothiocyanate, induces cell cycle arrest and apoptosis in HT29 human colon cancer cells. Cancer Research, 60(5), 1426–1433.

Gao, X., & Talalay, P. (2004). Induction of phase 2 genes by sulforaphane protects retinal pigment epithelial cells against photooxidative damage. Proceedings of the National Academy of Sciences of the United States of America, 101(28), 10446–10451.

Geng, F., Tang, L., Li, Y., Yang, L., Choi, K. S., Kazim, A. L., & Zhang, Y. (2011). Allyl isothiocyanate arrests cancer cells in mitosis, and mitotic arrest in turn leads to apoptosis via Bcl-2 protein phosphorylation. Journal of Biological Chemistry, 286(37), 32259–32267.

Getahun, S. M., & Chung, F. L. (1999). Conversion of glucosinolates to isothiocyanates in humans after ingestion of cooked watercress. Cancer Epidemiology Biomarkers & Prevention, 8(5), 447–451.

Ghosh,  D.  (2005).  Three-­‐dimensional  structures  of  sulfatases.  Methods in Enzymology, 400, 273–293. Giamoustaris, A., & Mithen, R. (1996). Genetics of aliphatic glucosinolates. IV. Side-chain modification in

Brassica oleracea. Theoretical and Applied Genetics, 93(5-6), 1006–1010. Gibson, G. R., Beatty, E. R., Wang, X., & Cummings, J. H. (1995). Selective stimulation of bifidobacteria in

the human colon by oligofructose and inulin. Gastroenterology, 108(4), 975-982. Gil, V., & MacLeod, A. J. (1980). The effects of pH on glucosinolate degradation by a thioglucoside

glucohydrolase preparation. Phytochemistry, 19(12), 2547–2551. Gill, S. R., Pop, M., DeBoy, R. T., Eckburg, P. B., Turnbaugh, P. J., Samuel, B. S., et al. (2006).

Metagenomic analysis of the human distal gut microbiome. Science, 312(5778), 1355–1359. Gingell, R., & Bridges, J. W. (1971). Intestinal azo reduction and glucuronide conjugation of prontosil.

Biochemical Journal, 125(2), 24P. Glöckner, F. O., Kube, M., Bauer, M., Teeling, H., Lombardot, T., Ludwig, W., et al. (2003). Complete

genome sequence of the marine planctomycete Pirellula sp. strain 1. Proceedings of the National Academy of Sciences of the United States of America, 100(14), 8298–8303.

Gong, A., He, M., Krishna Vanaja, D., Yin, P., Karnes, R. J., & Young, C. Y. F. (2009). Phenethyl isothiocyanate inhibits STAT3 activation in prostate cancer cells. Molecular Nutrition & Food Research, 53(7), 878–886.

González-Pombo, P., Pérez, G., Carrau, F., Guisán, J. M., Batista-Viera, F., & Brena, B. M. (2008). One-step purification and characterization   of   an   intracellular   β-glucosidase from Metschnikowia pulcherrima. Biotechnology Letters, 30(8), 1469–1475.

Goodman, I., Fouts, J. R., Bresnick, E., Menegas, R., & Hitchings, G. H. (1959). A mammalian thioglycosidase. Science, 130(3373), 450–451–451.

Gopalakrishnan, A., & Tony Kong, A. N. (2008). Anti-carcinogenesis by dietary phytochemicals: Cytoprotection by Nrf2 in normal cells and cytotoxicity by modulation of transcription factors NF-κB  and AP-1 in abnormal cancer cells. Food and Chemical Toxicology, 46(4), 1257–1270.

Gottesman, S., Roche, E., Zhou, Y. N., & Sauer, R. T. (1998). The ClpXP and ClpAP proteases degrade proteins with carboxy-terminal peptide tails added by the SsrA-tagging system. Genes & Development, 12(9), 1338–1347.

Goulian, M., & Beck, W. S. (1966). Purification and properties of cobamide-dependent ribonucleotide reductase from Lactobacillus leichmannii. Journal of Biological Chemistry, 241(18), 4233–4242.

Gower, W. R., Jr, Carr, M. C., & Ives, D. H. (1979). Deoxyguanosine kinase. Distinct molecular forms in

329

mitochondria and cytosol. Journal of Biological Chemistry, 254(7), 2180–2183. Graser, G., Oldham, N. J., Brown, P. D., Temp, U., & Gershenzon, J. (2001). The biosynthesis of benzoic

acid glucosinolate esters in Arabidopsis thaliana. Phytochemistry, 57(1), 23–32. Grubb, C. D., Zipp, B. J., Ludwig-Müller, J., Masuno, M. N., Molinski, T. F., & Abel, S. (2004). Arabidopsis

glucosyltransferase UGT74B1 functions in glucosinolate biosynthesis and auxin homeostasis. The Plant Journal, 40(6), 893–908.

Guarner, F., & Malagelada, J. R. (2003). Gut flora in health and disease. Lancet (London, England), 361(9356), 512–519.

Hagelueken, G., Adams, T. M., Wiehlmann, L., Widow, U., Kolmar, H., Tümmler, B., et al. (2006). The crystal structure of SdsA1, an alkylsulfatase from Pseudomonas aeruginosa, defines a third class of sulfatases. Proceedings of the National Academy of Sciences of the United States of America, 103(20), 7631–7636.

Halkier, B. A., & Du, L. (1997). The biosynthesis of glucosinolates. Trends in Plant Science, 2(11), 425–431. Halkier, B. A., & Gershenzon, J. (2006). Biology and biochemistry of glucosinolates. Annual Review of

Plant Biology, 57(1), 303–333. Halkier, B. A., Hansen, C. H., Mikkelsen, M. D., Naur, P., & Wittstock, U. (2002). Chapter Thirteen: The

role of cytochromes P450 in biosynthesis and evolution of glucosinolates. In J. T. R. A. R. A. Dixon (Eds.), Recent Advances in Phytochemistry (Vol. 37, pp. 223–248).

Hall, C., McCallum, D., Prescott, A., & Mithen, R. (2001). Biochemical genetics of glucosinolate modification in Arabidopsis and Brassica. Theoretical and Applied Genetics, 102(2-3), 369–374.

Hanlon, N., Poynton, C. L., Coldham, N., Sauer, M. J., & Ioannides, C. (2009). The aliphatic isothiocyanates   erucin   and   sulforaphane   do   not   effectively   up-­‐regulate   NAD(P)H:   quinone  oxidoreductase (NQO1) in human liver compared with rat. Molecular Nutrition & Food Research, 53(7), 836–844.

Hanschen, F. S., Bauer, A., Mewis, I., Keil, C., Schreiner, M., Rohn, S., & Kroh, L. W. (2012). Thermally induced degradation of aliphatic glucosinolates–identification of intermediary breakdown products and proposed degradation pathways. Journal of Agricultural and Food Chemistry, 60(39), 9890-9899

Hansen, Carsten Horslev, Du, L., Naur, P., Olsen, C. E., Axelsen, K. B., Hick, A. J., et al. (2001). CYP83B1 is the oxime-metabolizing enzyme in the glucosinolate pathway in Arabidopsis. Journal of Biological Chemistry, 276(27), 24790–24796.

Hanson, S. R., Best, M. D., & Wong, C.H. (2004). Sulfatases: Structure, mechanism, biological activity, inhibition, and synthetic utility. Angewandte Chemie International Edition, 43(43), 5736–5763.

Hanson, S. R., Whalen, L. J., & Wong, C. H. (2006). Synthesis and evaluation of general mechanism-based inhibitors of sulfatases based on (difluoro) methyl phenyl sulfate and cyclic phenyl sulfamate motifs. Bioorganic & Medicinal Chemistry, 14(24), 8386–8395.

Harris, K. E., & Jeffery, E. H. (2008). Sulforaphane and erucin increase MRP1 and MRP2 in human carcinoma cell lines. The Journal of Nutritional Biochemistry, 19(4), 246–254.

Harvey, A. J., Hrmova, M., De Gori, R., Varghese, J. N., & Fincher, G. B. (2000). Comparative modeling of the  three-­‐dimensional  structures  of  family  3  glycoside  hydrolases.  Proteins: Structure, Function, and Bioinformatics, 41(2), 257–269.

Hashimoto, W., Miki, H., Nankai, H., Sato, N., Kawai, S., & Murata, K. (1998). Molecular cloning of two genes for D-glucosidase in Bacillus sp. GL1 and identification of one as a gellan degrading enzyme1 . Archives of Biochemistry and Biophysics, 360(1), 1–9.

Hassouni, el, M., Chippaux, M., & Barras, F. (1990). Analysis of the Erwinia chrysanthemi arb genes, which mediate metabolism of aromatic beta-glucosides. Journal of Bacteriology, 172(11), 6261–6267.

Hayes, J. D., Kelleher, M. O., & Eggleston, I. M. (2008). The cancer chemopreventive actions of phytochemicals derived from glucosinolates. European Journal of Nutrition, 47(S2), 73–88.

330

Heaney, R. K., & Fenwick, G. R. (1993). Alkaloïds and sulphur compounds Methods in Plant Biochemistry, 8, 531-550.

Heaney, R. K., Spinks, E. A., & Fenwick, G. R. (1988). Improved method for the determination of the total glucosinolate content of rapeseed by determination of enzymically released glucose. Analyst, 113(10), 1515–1518.

Heavey, P. M., & Rowland, I. R. (2004). Gastrointestinal cancer. Best Practice & Research Clinical Gastroenterology, 18(2), 323–336.

Heber, D., & Bowerman, S. (2001). Applying science to changing dietary patterns. The Journal of Nutrition, 131(11), 3078S–3081S.

Hecht, S. S. (1999). Chemoprevention of cancer by isothiocyanates, modifiers of carcinogen metabolism. The Journal of Nutrition, 129(3), 768S–774S.

Hektor, H. J., Kloosterman, H., & Dijkhuizen, L. (2002). Identification of a magnesium-dependent NAD(P)(H)-binding domain in the nicotinoprotein methanol dehydrogenase from Bacillus methanolicus. Journal of Biological Chemistry, 277(49), 46966–46973.

Henderson, M. J., & Milazzo, F. H. (1979). Arylsulfatase in Salmonella typhimurium: detection and influence of carbon source and tyramine on its synthesis. Journal of Bacteriology, 139(1), 80–87.

Henrissat, B. (1991). A classification of glycosyl hydrolases based on amino acid sequence similarities. Biochemical Journal, 280(Pt 2), 309.

Henrissat, B., Callebaut, I., Fabrega, S., Lehn, P., Mornon, J. P., & Davies, G. (1995). Conserved catalytic machinery and the prediction of a common fold for several families of glycosyl hydrolases. Pnas, 92(15), 7090–7094.

Hernandez-Guzman, F. G., Higashiyama, T., Pangborn, W., Osawa, Y., & Ghosh, D. (2003). Structure of human estrone sulfatase suggests functional roles of membrane association. Journal of Biological Chemistry, 278(25), 22989–22997.

Higdon, J. V., Delage, B., Williams, D. E., & Dashwood, R. H. (2007). Cruciferous vegetables and human cancer risk: epidemiologic evidence and mechanistic basis. Pharmacological Research: the Official Journal of the Italian Pharmacological Society, 55(3), 224.

Hoffman, J. A., Badger, J. L., Zhang, Y., Huang, S.-H., & Kim, K. S. (2000). Escherichia coli K1 aslA contributes to invasion of brain microvascular endothelial cells in vitro and in vivo. Infection and Immunity, 68(9), 5062–5067.

Holmes, E., Kinross, J., Gibson, G. R., Burcelin, R., Jia, W., Pettersson, S., & Nicholson, J. K. (2012). Therapeutic modulation of microbiota-host metabolic interactions. Science Translational Medicine, 4(137), 137rv6–137rv6.

Holst, B., & Williamson, G. (2004). A critical review of the bioavailability of glucosinolates and related compounds. Natural Product Reports, 21(3), 425.

Holt, R. (1967). The bacterial degradation of chloramphenicol. Lancet, 1(7502), 1259–1260. Hooper, L. V., Bry, L., Falk, P. G., & Gordon, J. I. (1998). Host–microbial symbiosis in the mammalian

intestine: exploring an internal ecosystem. Bioessays, 20(4), 336–343. Hopkins, R. J., van Dam, N. M., & van Loon, J. J. A. (2009). Role of glucosinolates in Insect-plant

relationships and multitrophic interactions. Annual Review of Entomology, 54(1), 57–83. Hopwood, J. J., & Ballabio, A. (2001). Multiple sulfatase deficiency and the nature of the sulfatase family.

Metabolic and Molecular Basis of Inherited Diseases, 3, 3725–4261. Hori, S. H., Kamada, T., & Matsui, S. (1967). Electrophoretic separation of the magnesium-dependent

glucose 6-phosphate dehydrogenases in rats. Journal of Histochemistry & Cytochemistry, 15(7), 419–420.

Hoving,   S.,   Voshol,   H.,   &   van   Oostrum,   J.   (2000).   Towards   high   performance   two-­‐dimensional   gel  electrophoresis using ultrazoom gels. Electrophoresis, 21(13), 2617–2621.

Howe, C. J. (2007). Gene Cloning and Manipulation (2nd ed.). Cambridge University Press.

331

Hrmova, M., Harvey, A. J., Wang, J., Shirley, N. J., Jones, G. P., Stone, B. A., et al. (1996). Barley-D-glucan exohydrolases with-D-glucosidase activity: Purification, characterization, and determination of primary structure from a cDNA clone. Journal of Biological Chemistry, 271(9), 5277–5286.

Hrmova, M., Varghese, J. N., De Gori, R., Smith, B. J., Driguez, H., & Fincher, G. B. (2001). Catalytic mechanisms   and   reaction   intermediates   along   the   hydrolytic   pathway   of   a   plant   β-D-glucan glucohydrolase. Structure, 9(11), 1005–1016.

Hrnčiřík,   K.,   Velíšek,   J.,   &   Davídek,   J.   (1998).   Comparison   of   HPLC   and   GLC   methodologies   for  determination of glucosinolates using reference material. Zeitschrift FR Lebensmitteluntersuchung Und -Forschung A, 206(2), 103–107.

Husebye, H., Arzt, S., Burmeister, W. P., Härtel, F. V., Brandt, A., Rossiter, J. T., & Bones, A. M. (2005). Crystal structure at 1.1Å resolution of an insect myrosinase from Brevicoryne brassicae shows its close  relationship  to  β-glucosidases. Insect Biochemistry and Molecular Biology, 35(12), 1311–1320.

Huycke, M. M., & Gaskins, H. R. (2004). Commensal bacteria, redox stress, and colorectal cancer: Mechanisms and models. Experimental Biology and Medicine, 229(7), 586–597.

Hwang, E. S., & Lee, H. J. (2006). Allyl Isothiocyanate and its N-acetylcysteine conjugate cuppress metastasis via inhibition of invasion, migration, and matrix metalloproteinase-2/-9 activities in SK-Hep1 human hepatoma cells. Experimental Biology and Medicine, 231(4), 421–430.

Hwang, E. S., & Lee, H. J. (2008). Benzyl isothiocyanate inhibits metalloproteinase-2/-9 expression by suppressing the mitogen-activated protein kinase in SK-Hep1 human hepatoma cells. Food and Chemical Toxicology, 46(7), 2358–2364.

Iguchi, A., Thomson, N. R., Ogura, Y., Saunders, D., Ooka, T., Henderson, I. R., et al. (2008). Complete genome sequence and comparative genome analysis of enteropathogenic Escherichia coli O127:H6 strain E2348/69. Journal of Bacteriology, 191(1), 347–354.

Ina, K., Nobukuni, M., Sano, A., Kishima, I. (1981). Stability of allyl isothiocyanate. Nippon Shokuhin Kogyo Gakkaishi, 12, 627-631.

International Agency for Research on Cancer. (2004). IARC Handbooks of Cancer Prevention: Cruciferous Vegetables, Isothiocyanates and Indoles. Lyon: IARC Press.

Itoh, K., Wakabayashi, N., Katoh, Y., Ishii, T., O'Connor, T., & Yamamoto, M. (2003). Keap1 regulates both cytoplasmic-nuclear shuttling and degradation of Nrf2 in response to electrophiles. Genes to Cells, 8(4), 379–391.

Iveson, P., Lindup, W. E., Parke, D. V., & Williams, R. T. (1971). The metabolism of carbenoxolone in the rat, xenobiotica, informa healthcare. Xenobiotica, 1(1), 79–95.

Jadhav, U., Vaughn, S. F., Berhow, M. A., & Sanjeeva, M. (2007). Iberin induces cell cycle arrest and apoptosis in human neuroblastoma cells. International Journal of Molecular Medicine, 19(3), 353-361.

Jain, M. G., Hislop, G. T., Howe, G. R., & Ghadirian, P. (1999). Plant foods, antioxidants, and prostate cancer risk: findings from case-control studies in canada. Nutrition and Cancer 34(2), 173–184.

Jakobsen, T. H., Bragason, S. K., Phipps, R. K., Christensen, L. D., van Gennip, M., Alhede, M., et al. (2012). Food as a source for quorum sensing inhibitors: Iberin from horseradish revealed as a quorum sensing inhibitor of Pseudomonas aeruginosa. Applied and Environmental Microbiology, 78(7), 2410–2421.

Jakubikova, J., Bao, Y., & Sedlak, J. (2005). Isothiocyanates induce cell cycle arrest, apoptosis and mitochondrial potential depolarization in HL-60 and multidrug-resistant cell lines. Anticancer Research, 25(5), 3375–3386.

James, D., Devaraj, S., Bellur, P., Lakkanna, S., Vicini, J., & Boddupalli, S. (2012). Novel concepts of broccoli sulforaphanes and disease: induction of phase II antioxidant and detoxification enzymes by enhanced-glucoraphanin broccoli. Nutrition Reviews, 70(11), 654–665.

James, D. C., & Rossiter, J. T. (1991). Development and characteristics of myrosinase in Brassica napus

332

during early seedling growth. Physiologia Plantarum, 82(2), 163–170. Jarrige, P. (1963). Purification and properties of sulfatases from the digestive juice of H. pomatia.

Bulletin De La Société De Chimie Biologique, 31(45), 761–782. Jauch, R., Hankwity, R., Beschke, K., Pelzer, H. (1975). Bis-(phydroxyphenyl)-pyridyl-2-methane: the

common laxative principle of bisacodyl and sodium picosulfate, Arzneim Forsch, 25(11), 1796–1800. Jennings, D., and Isaac, S. (1995). Microbial cultures. Bios Scientific Publishers Ltd, Oxford, UK. Jeong, W. S., Kim, I. W., Hu, R., & Kong, A. N. T. (2004). Modulatory properties of various natural

chemopreventive agents on the activation of NF-κB   signaling   pathway.  Pharmaceutical Research, 21(4), 661–670.

Ji, Y., & Morris, M. E. (2005). Membrane transport of dietary phenethyl isothiocyanate by ABCG2 (breast cancer resistance protein). Mol. Pharmaceutics, 2(5), 414–419.

Johnson, I. T. (2004). New approaches to the role of diet in the prevention of cancers of the alimentary tract. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis, 551(1–2), 9–28.

Jones, A., Bridges, M., Bones, A. M., Cole, R., & Rossiter, J. T. (2001). Purification and characterisation of a non-plant myrosinase from the cabbage aphid Brevicoryne brassicae(L.). Insect Biochemistry and Molecular Biology, 31(1), 1–5.

Jones, A., Winge, P., Bones, A. M., Cole, R., & Rossiter, J. T. (2002). Characterization and evolution of a myrosinase from the cabbage aphid Brevicoryne brassicae. Insect Biochemistry and Molecular Biology, 32(3), 275–284.

Joseph, M. A., Moysich, K. B., Freudenheim, J. L., Shields, P. G., Bowman, E. D., Zhang, Y., et al. (2004). Cruciferous vegetables, genetic polymorphisms in glutathione s-transferases M1 and T1, and prostate cancer risk. Nutrition and Cancer, 50(2), 206–213.

Kalia, A., & Gupta, R. P. (2005). Proteomics: A paradigm shift. Critical Reviews in Biotechnology, 25(4), 173–198.

Kallus, S. J., & Brandt, L. J. (2011). The intestinal microbiota and obesity. Journal of clinical gastroenterology, 46(1), 16–24.

Kaoutari A.E., Armougom F., Gordon J.I, Raoult D, Henrissat B. (2013). The abundance and variety of carbohydrate-active enzymes in the human gut microbiota. Nat Rev Microbiol. doi: 10.1038/nrmicro3050.

Kashiyama, E., Todaka, T., Odomi, M., Tanokura, Y., Johnson, D. B., Yokoi, T., et al. (1994). Stereoselective pharmacokinetics and interconversions of flosequinan enantiomers containing chiral sulphoxide in rat. Xenobiotica, 24(4), 369–377.

Kassahun, K., Davis, M., Hu, P., Martin, B., & Baillie, T. (1997). Biotransformation of the naturally occurring isothiocyanate sulforaphane in the rat: identification of phase I metabolites and glutathione conjugates. Chemical Research in Toxicology, 10(11), 1228–1233.

Kassie, F., & Knasmüller, S. (2000). Genotoxic effects of allyl isothiocyanate (AITC) and phenethyl isothiocyanate (PEITC). Chem Biol Interact, 127(2), 163–180.

Kassie, F., Pool-Zobel, B., Parzefall, W., & Knasmüller, S. (1999). Genotoxic effects of benzyl isothiocyanate, a natural chemopreventive agent. Mutagenesis, 14(6), 595–604.

Kawakishi, S., & Kaneko, T. (1987). Interaction of proteins with allyl isothiocyanate. Journal of Agricultural and Food Chemistry, 35(1), 85–88.

Kawakishi, S., & Muramatsu, K. (1966). Studies on the decomposition of sinalbin. Part. 1 the decomposition products of sinalbin. Agricultural and Biological Chemistry, 30, 688–692.

Kawakishi, S., & Namiki, M. (1969). Decomposition of allyl isothiocyanate in aqueous solution. Agricultural and Biological Chemistry, 33(3), 452–459.

Kelly, P. J., Bones, A., & Rossiter, J. T. (1998). Sub-cellular immunolocalization of the glucosinolate sinigrin in seedlings of Brassica juncea. Planta, 206(3), 370–377.

333

Kelsey, N. A., Wilkins, H. M., & Linseman, D. A. (2010). Nutraceutical antioxidants as novel neuroprotective agents. Molecules, 15(11), 7792–7814.

Kertesz, M. A. (2000). Riding the sulfur cycle–metabolism of sulfonates and sulfate esters in Gram-­‐negative  bacteria.  FEMS microbiology reviews, 24(2), 135–175.

Kertesz, M. A., Leisinger, T., & Cook, A. M. (1993). Proteins induced by sulfate limitation in Escherichia coli, Pseudomonas putida, or Staphylococcus aureus. Journal of Bacteriology, 175(4), 1187–1190.

Ketterer, B. (1998). Dietary isothiocyanates as confounding factors in the molecular epidemiology of colon cancer. Cancer Epidemiology Biomarkers & Prevention, 7(8), 645–646.

Ketudat   Cairns,   J.   R.,   &   Esen,   A.   (2010).   β-Glucosidases. Cellular and Molecular Life Sciences, 67(20), 3389–3405.

Keum, Y. S., Jeong, W. S., & Kong, A. N. (2005). Chemopreventive functions of isothiocyanates. Drug News Perspect. 18(7), 445-51

Keum, Y. S., Jeong, W. S., & Tony Kong, A. N. (2004). Chemoprevention by isothiocyanates and their underlying molecular signaling mechanisms. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis, 555(1–2), 191–202.

Kim, M. J., Kim, S. H., & Lim, S. J. (2010). Comparison of the apoptosis-inducing capability of sulforaphane analogues in human colon cancer cells. Anticancer Research, 30(9), 3611–3619.

Kim, S. H., Nagalingam, A., Saxena, N. K., Singh, S. V., & Sharma, D. (2011). Benzyl isothiocyanate inhibits oncogenic actions of leptin in human breast cancer cells by suppressing activation of signal transducer and activator of transcription 3. Carcinogenesis, 32(3), 359–367.

Kim, S. H., & Singh, S. V. (2009). D,L-Sulforaphane causes transcriptional repression of androgen receptor in human prostate cancer cells. Molecular Cancer Therapeutics, 8(7), 1946–1954.

Kima, D. E., Kima, K. H., Baea, Y. J., Leea, J. H., Jangb, Y. H., & Nam, S.W. (2005). Purification and characterization of the recombinant arylsulfatase cloned from Pseudoalteromonas carrageenovora. Protein Expression and Purification, 39, 107–115.

Kissen, R., & Bones, A. M. (2009). Nitrile-specifier proteins involved in glucosinolate hydrolysis in Arabidopsis thaliana. Journal of Biological Chemistry, 284(18), 12057–12070.

Kitamura, S., Sugihara, K., Kuwasako, M., & Tatsumi, K. (1997). The role of mammalian intestinal bacteria in the reductive metabolism of zonisamide. Journal of Pharmacy and Pharmacology, 49(3), 253–256.

Kivelä, A. M., Mäkinen, P. I., Jyrkkänen, H.-K., Mella-Aho, E., Xia, Y., Kansanen, E., et al. (2010). Sulforaphane inhibits endothelial lipase expression through NF-κB   in   endothelial   cells.  Atherosclerosis, 213(1), 122–128.

Kjaer, A. (1976). Glucosinolates in the Cruciferae. Academic Press, London. Kliebenstein, D. J., Kroymann, J., Brown, P., Figuth, A., Pedersen, D., Gershenzon, J., & Mitchell-Olds, T.

(2001). Genetic control of natural variation in Arabidopsis glucosinolate accumulation. Plant Physiology, 126(2), 811–825.

Klose, J. (1975). Protein mapping by combined isoelectric focusing and electrophoresis of mouse tissues. A novel approach to testing for induced point mutations in mammals. Humangenetik, 26(3), 231–243.

Kobayashi, M., & Shimizu, S. (1994). Versatile nitrilases: nitrile-hydrolysing enzymes. FEMS Microbiology Letters, 120(3), 217–223.

Kolonel, L. N., Hankin, J. H., Whittemore, A. S., Wu, A. H., Gallagher, R. P., Wilkens, L. R., et al. (2000). Vegetables, fruits, legumes and prostate cancer: a multiethnic case-control study. Cancer Epidemiology Biomarkers & Prevention, 9(8), 795–804.

Kolm, R. H., Danielson, U. H., Zhang, Y., Talalay, P., & Mannervik, B. (1995). Isothiocyanates as substrates for human glutathione transferases: structure-activity studies. Biochemical Journal, 311(Pt 2), 453-459.

Kong, L., Tanito, M., Huang, Z., Li, F., Zhou, X., Zaharia, A., et al. (2007). Delay of photoreceptor

334

degeneration in tubby mouse by sulforaphane. Journal of Neurochemistry, 101(4), 1041–1052. Kopycki, J., Schmidt, J., Abel, S., & Douglas Grubb, C. (2011). Chemoenzymatic synthesis of diverse

thiohydroximates from glucosinolate-utilizing enzymes from H. pomatia and Caldicellulosiruptor saccharolyticus. Biotechnology Letters, 33(5), 1039–1046.

Koroleva, O. A., Davies, A., Deeken, R., Thorpe, M. R., Tomos, A. D., & Hedrich, R. (2000). Identification of a new glucosinolate-rich cell type in Arabidopsis flower stalk. Plant Physiology, 124(2), 599–608.

Koroleva, O. A., Gibson, T. M., Cramer, R., & Stain, C. (2010). Glucosinolate-accumulating S-cells in Arabidopsis leaves and flower stalks undergo programmed cell death at early stages of differentiation. The Plant Journal, 64(3), 456–469.

Koropatkin, N. M., Cameron, E. A., & Martens, E. C. (2012). How glycan metabolism shapes the human gut microbiota. Nature Reviews Microbiology. doi:10.1038/nrmicro2746

Kostic, A. D., Gevers, D., Pedamallu, C. S., Michaud, M., Duke, F., Earl, A. M., et al. (2012). Genomic analysis identifies association of Fusobacterium with colorectal carcinoma. Genome Research, 22(2), 292–298.

Kovatcheva-Datchary,  P.,  Egert,  M.,  Maathuis,  A.,  Rajilić-Stojanović,  M.,  De  Graaf,  A.  A.,  Smidt,  H.,  et  al.  (2009). Linking phylogenetic identities of bacteria to starch fermentation in an in vitro model of the large  intestine  by  RNA-­‐based  stable  isotope  probing. Environmental Microbiology, 11(4), 914–926.

Krul, C., Humblot, C., Philippe, C., Vermeulen, M., Van Nuenen, M., Havenaar, R., & Rabot, S. (2002). Metabolism of sinigrin (2-propenyl glucosinolate) by the human colonic microflora in a dynamic in vitro large-intestinal model. Carcinogenesis, 23(6), 1009–1016.

Kuchernig, J. C., Backenköhler, A., Lübbecke, M., Burow, M., & Wittstock, U. (2011). A thiocyanate-forming protein generates multiple products upon allyl glucosinolate breakdown in Thlaspi arvense. Phytochemistry, 72(14-15), 1699–1709.

Kuchernig, J. C., Burow, M., & Wittstock, U. (2012). Evolution of specifier proteins in glucosinolate-containing plants. BMC Evolutionary Biology, 12(1), 127.

Kuhn, J., Pettersson, E. M., Feld, B. K., Nie, L., Tolzin-Banasch,   K.,   M’Rabet,   S.   M.,   et   al.   (2006).  Sequestration of plant-derived phenolglucosides by larvae of the leaf beetle Chrysomela lapponica: thioglucosides as mechanistic probes. Journal of Chemical Ecology, 33(1), 5–24.

Kumar, J. K. (2008). Lysostaphin: an antistaphylococcal agent. Applied Microbiology and Biotechnology, 80(4), 555–561.

Kushad, M. M., Brown, A. F., Kurilich, A. C., Juvik, J. A., Klein, B. P., Wallig, M. A., & Jeffery, E. H. (1999). Variation of glucosinolates in vegetable crops of Brassica oleracea. Journal of Agricultural and Food Chemistry, 47(4), 1541–1548.

Kwon, J. S., Joung, H., Kim, Y. S., Shim, Y.-S., Ahn, Y., Jeong, M. H., & Kee, H. J. (2012). Sulforaphane inhibits restenosis by suppressing inflammation and the proliferation of vascular smooth muscle cells. Atherosclerosis, 225(1), 41–49.

Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227(5259), 680–685.

Lai, K. C., Huang, A. C., Hsu, S. C., Kuo, C. L., Yang, J. S., Wu, S. H., & Chung, J. G. (2010). Benzyl isothiocyanate (BITC) inhibits migration and invasion of human colon cancer HT29 cells by inhibiting matrix metalloproteinase-2/-9 and urokinase Plasminogen (uPA) through PKC and MAPK Signaling Pathway. Journal of Agricultural and Food Chemistry, 58(5), 2935–2942.

Lambrix, V., Reichelt, M., Mitchell-Olds, T., Kliebenstein, D. J., & Gershenzon, J. (2001). The Arabidopsis epithiospecifier protein promotes the hydrolysis of glucosinolates to nitriles and influences Trichoplusia ni herbivory. The Plant Cell, 13(12), 2793–2807.

Lamed, R. J., & Zeikus, J. G. (1980). Novel NADP-linked alcohol-aldehyde/ketone oxidoreductase in thermophilic ethanologenic bacteria. Biochemical Journal, 195(1), 183–190.

Landgrebe, J., Dierks, T., Schmidt, B., & Figura, von, K. (2003). The human SUMF1 gene, required for

335

posttranslational sulfatase modification, defines a new gene family which is conserved from pro- to eukaryotes. Gene, 316, 47–56.

Lane, C. S. (2005). Mass spectrometry-based proteomics in the life sciences. CMLS Cellular and Molecular Life Sciences, 62(7-8), 848–869.

Langlands, S. J., Hopkins, M. J., Coleman, N., & Cummings, J. H. (2004). Prebiotic carbohydrates modify the mucosa associated microflora of the human large bowel. Gut, 53(11), 1610–1616.

Larkin, M. A., Blackshields, G., Brown, N. P., Chenna, R., McGettigan, P. A., McWilliam, H., et al. (2007). Clustal W and Clustal X version 2.0. Bioinformatics, 23(21), 2947–2948.

Lazzeri, L., Tacconi, R., & Palmieri, S. (1993). In vitro activity of some glucosinolates and their reaction products toward a population of the nematode Heterodera schachtii. Journal of Agricultural and Food Chemistry, 41(5), 825–829.

Lea, M. A., Randolph, V. M., & Lee, J. E. (2001). Induction of histone acetylation in mouse erythroleukemia cells by some organosulfur compounds including allyl isothiocyanate. International Journal of Cancer, 92(6), 784–789.

Le Coq, D., Lindner, C., Krüger, S., Steinmetz, M., & Stülke, J. (1995). New beta-glucoside (bgl) genes in Bacillus subtilis: the bglP gene product has both transport and regulatory functions similar to those of BglF, its Escherichia coli homolog. Journal of Bacteriology, 177(6), 1527–1535.

Lederberg, J. (1994). The transformation of genetics by DNA: an anniversary celebration of Avery, MacLeod and McCarty (1944). Genetics, 136(2), 423.

Lee, E. J., Kim, D. I., Kim, W. J., & Moon, S. K. (2009). Naringin inhibits matrix metalloproteinase-­‐9  expression   and   AKT   phosphorylation   in   tumor   necrosis   factor-­‐α-­‐induced   vascular   smooth  muscle  cells. Molecular Nutrition & Food Research, 53(12), 1582–1591.

Lee, N. K., Choi, S. H., Park, S. H., Park, E. K., & Kim, D. H. (2004). Antiallergic activity of hesperidin is activated by intestinal microflora. Pharmacology, 71(4), 174–180.

Lee, R. C., Hrmova, M., Burton, R. A., Lahnstein, J., & Fincher, G. B. (2003). Bifunctional family 3 glycoside hydrolases   from   barley   with   α-l-arabinofuranosidase   and   β-d-xylosidase activity. Journal of Biological Chemistry, 278(7), 5377–5387.

Lee, S. C., & Renwick, A. G. (1995). Sulphoxide reduction by rat intestinal flora and by Escherichia coli in vitro. Biochemical Pharmacology, 49(11), 1567–1576.

Legler, G., Roeser, K. R., & Illig, H. K.  (1979).  Reaction  of  β-­‐D-­‐Glucosidase  A3 from Aspergillus wentii with D-­‐Glucal.  European Journal of Biochemistry, 101(1), 85–92.

Lenman, M., Falk, A., Xue, J., & Rask, L. (1993). Characterization of a Brassica napus myrosinase pseudogene:   myrosinases   are   members   of   the   BGA   family   of   β-glycosidases. Plant Molecular Biology, 21(3), 463–474.

Leoni, O., Iori, R., Palmieri, S., Esposito, E., Menegatti, E., Cortesi, R., & Nastruzzi, C. (1997). Myrosinase-generated isothiocyanate from glucosinolates: Isolation, characterization and in vitro anti-proliferative studies. Bioorganic & Medicinal Chemistry, 5(9), 1799–1806.

Ley, R. E., Lozupone, C. A., Hamady, M., Knight, R., & Gordon, J. I. (2008). Worlds within worlds: evolution of the vertebrate gut microbiota. Nature Reviews Microbiology, 6(10), 776–788.

Ley, R. E., Turnbaugh, P. J., Klein, S., & Gordon, J. I. (2006). Microbial ecology: Human gut microbes associated with obesity. Nature, 444(7122), 1022-1023.

Li, F., Hullar, M. A. J., Schwarz, Y., & Lampe, J. W. (2009). Human gut bacterial communities are altered by addition of cruciferous vegetables to a controlled fruit- and vegetable-free diet. Journal of Nutrition, 139(9), 1685–1691.

Li, J., Yao, S., & Zhang, Y. (2005). The role of c-Jun in the AP-1 activation induced by naturally occurring isothiocyanates. Food and Chemical Toxicology, 43(9), 1373–1380.

Lin, H. J., Probst-Hensch, N. M., Louie, A. D., Kau, I. H., Witte, J. S., Ingles, S. A., et al. (1998). Glutathione transferase null genotype, broccoli, and lower prevalence of colorectal adenomas. Cancer

336

Epidemiology Biomarkers & Prevention, 7(8), 647–652. Link, A. J., Eng, J., Schieltz, D. M., Carmack, E., Mize, G. J., Morris, D. R., et al. (1999). Direct analysis of

protein complexes using mass spectrometry. Nature Biotechnology, 17(7), 676–682. Litzinger, S., Fischer, S., Polzer, P., Diederichs, K., Welte, W., & Mayer, C. (2010). Structural and kinetic

analysis of Bacillus subtilis N-acetylglucosaminidase reveals a unique Asp-His dyad mechanism. Journal of Biological Chemistry, 285(46), 35675–35684.

Lopez, J. L. (2007). Two-dimensional electrophoresis in proteome expression analysis. J Chromatogr B Analyt Technol Biomed Life Sci, 849(1-2), 190–202.

Lori, R., Bernardi, R., Gueyrard, D., Rollin, P., & Palmieri, S. (1999). Formation of glucoraphanin by chemoselective oxidation of natural glucoerucin: A chemoenzymatic route to sulforaphane. Bioorganic and Medicinal Chemistry Letters, 9(7), 1047–1048.

Louda, S. M., Farris, M. A., & Blua, M. J. (1987). Variation in methylglucosinolate and insect damage to Cleome serrulata (Capparaceae) along a natural soil moisture gradient. Journal of Chemical Ecology, 13(3), 569–581.

Louda, S. M., & Rodman, J. E. (1983). Concentration of glucosinolates in relation to habitat and insect herbivory for the native crucifer Cardamine cordifolia. Biochemical Systematics and Ecology, 11(3), 199–207.

Lovelock, D. W., and Davies, R. (1978). Techniques for the study of mixed populations. Academic Press, N. Y.

Lu, M., Hashimoto, K., & Uda, Y. (2011). Rat intestinal microbiota digest desulfosinigrin to form allyl cyanide and 1-cyano-2,3-epithiopropane. Frin, 44(4), 1023–1028.

Luciano, F. B., & Holley, R. A. (2009). Enzymatic inhibition by allyl isothiocyanate and factors affecting its antimicrobial action against Escherichia coli O157:H7. International Journal of Food Microbiology, 131(2-3), 240–245.

Luciano, F. B., Hosseinian, F. S., Beta, T., & Holley, R. A. (2008). Effect of free-SH containing compounds on allyl isothiocyanate antimicrobial activity against Escherichia coli O157:H7. Journal of Food Science, 73(5), M214–M220.

Lukatela, G., Krauss, N., Theis, K., Selmer, T., Gieselmann, V., Figura, von, K., & Saenger, W. (1998). Crystal  structure  of  human  arylsulfatase  A:   The  aldehyde  function  and  the  metal  Ion  at  the  active  site suggest a novel mechanism for sulfate ester hydrolysis. Biochemistry, 37(11), 3654–3664.

Lynn, A., Collins, A., Fuller, Z., Hillman, K., & Ratcliffe, B. (2007). Cruciferous vegetables and colo-rectal cancer. Proceedings of the Nutrition Society, 65(1), 135–144.

MacLeod, A. J., & Rossiter, J. T. (1986). Isolation and examination of thioglucoside glucohydrolase from seeds of Brassica napus. Phytochemistry, 25(5), 1047–1051.

Maheshwari, P. N., Stanley, D. W., & Gray, J. I. (1981). Detoxification of rapeseed production. Journal of Food Protection, 44, 413–444.

Manach, C., Scalbert, A., Morand, C., Rémésy, C., & Jiménez, L. (2004). Polyphenols: food sources and bioavailability. The American Journal of Clinical Nutrition, 79(5), 727–747.

Mann, E. A., & Saeed, S. A. (2012). Gastrointestinal infection as a trigger for inflammatory bowel disease. Current opinion in gastroenterology, 28(1), 24–29.

Mari,  M.,  Leoni,  O.,  Iori,  R.,  &  Cembali,  T.  (2002).  Antifungal  vapour-­‐phase  activity  of  allyl-­‐isothiocyanate  against Penicillium expansum on pears. Plant Pathology, 51(2), 231–236.

Markowitz, V. M., Chen, I. M. A., Palaniappan, K., Chu, K., Szeto, E., Grechkin, Y., et al. (2010). The integrated microbial genomes system: an expanding comparative analysis resource. Nucleic Acids Research, 38(suppl 1), D382–D390.

Marques, A. R., Coutinho, P. M., Videira, P., Fialho, A. M., & Sá-Correia, I. (2003). Sphingomonas paucimobilis beta-glucosidase Bgl1: a member of a new bacterial subfamily in glycoside hydrolase family 1. Biochemical Journal, 370(Pt 3), 793-804.

337

Martin, J. A., & Wang, Z. (2011). Next-generation transcriptome assembly. Nature Reviews Genetics, 12(10), 672-682.

Maskell, I., & Smithard, R. (1994). Degradation of glucosinolates during in vitro incubations of rapeseed meal with myrosinase (EC 3.2.3 1) and with pepsin (EC 3.4.23.l)–hydrochloric acid, and contents of porcine small intestine and caecum. British Journal of Nutrition, 72(03), 455–466.

Matsui, I., Sakai, Y., Matsui, E., Kikuchi, H., Kawarabayasi, Y., & Honda, K. (2000). Novel substrate specificity of a membrane-bound   β-glycosidase from the hyperthermophilic archaeon Pyrococcus horikoshii. FEBS Letters, 467(2), 195–200.

Mayer, C., Vocadlo, D. J., Mah, M., Rupitz, K., Stoll, D., Warren, R. A. J., & Withers, S. G. (2006). Characterization of a beta-N-acetylhexosaminidase and a beta-N-acetylglucosaminidase/beta-glucosidase from Cellulomonas fimi. FEBS Journal, 273(13), 2929–2941.

Mayton, H. S., Olivier, C., Vaughn, S. F., & Loria, R. (1996). Correlation of fungicidal activity of Brassica species with allyl isothiocyanate production in macerated leaf tissue. Phytopathology, 86(3), 267–271.

McAndrew,  R.  P.,  Park,  J.  I.,  Heins,  R.  A.,  &  Reindl,  W.  (2013).  From  soil  to  structure,  a  novel  dimeric  β-glucosidase belonging to glycoside hydrolase family 3 isolated from compost using metagenomic analysis. Journal of Biological Chemistry, 288:14985-14992.

McBain, A. J., & Macfarlane, G. T. (1998). Ecological and physiological studies on large intestinal bacteria in relation to production of hydrolytic and reductive enzymes involved in formation of genotoxic metabolites. Journal of Medical Microbiology, 47(5), 407–416.

McGregor, D. I., Mullin, W. J., & Fenwick, G. R. (1983). Analytical methodology for determining glucosinolate composition and content. J Assoc Off Anal Chem, 66, 825–849.

McIntosh, F. M., Maison, N., Holtrop, G., Young, P., Stevens, V. J., Ince, J., et al. (2012). Phylogenetic distribution  of  genes  encoding  β-glucuronidase activity in human colonic bacteria and the impact of diet on faecal glycosidase activities. Environmental Microbiology, 14(8), 1876–1887.

Melchini, A., Costa, C., Traka, M., Miceli, N., Mithen, R., De Pasquale, R., & Trovato, A. (2009). Erucin, a new promising cancer chemopreventive agent from rocket salads, shows anti-proliferative activity on human lung carcinoma A549 cells. Food and Chemical Toxicology, 47(7), 1430–1436.

Melchini, A., & Traka, M. (2010). Biological Profile of Erucin: A New Promising Anticancer Agent from Cruciferous Vegetables. Toxins, 2(4), 593–612.

Mellon, F. A., Bennett, R. N., Holst, B., & Williamson, G. (2002). Intact glucosinolate analysis in plant extracts by programmed cone voltage electrospray LC/MS: performance and comparison with LC/MS/MS methods. Analytical Biochemistry, 306(1), 83–91.

Mennicke, W. H., Görler, K., & Krumbiegel, G. (1983). Metabolism of some naturally occurring isothiocyanates in the rat. Xenobiotica, 13(4), 203–207.

Mennicke, W. H., Görler, K., Krumbiegel, G., Lorenz, D., & Rittmann, N. (1988). Studies on the metabolism and excretion of benzyl isothiocyanate in man. Xenobiotica, 18(4), 441–447.

Methé, B. A., Nelson, K. E., Pop, M., Creasy, H. H., Giglio, M. G., Huttenhower, C., et al. (2012). A framework for human microbiome research. Nature, 486(7402), 215–221.

Metschnikoff, E., & Chalmers, P. (1907). The prolongation of life. optimistic studies. W. Heinemann, London;  G.  P.  Putnam’S  Sons,  New  York.

Meuldermans, W., Hendrickx, J., Mannens, G., Lavrijsen, K., Janssen, C., Bracke, J., et al. (1994). The metabolism and excretion of risperidone after oral administration in rats and dogs. Drug Metabolism and Disposition, 22(1), 129–138.

Meyer, T. S., & Lamberts, B. L. (1965). Use of coomassie brilliant blue R250 for the electrophoresis of microgram quantities of parotid saliva proteins on acrylamide-gel strips (M. TS & L. BL, Eds.). Biochimica Et Biophysica Acta, 107(1), 144–145.

Mi, L., Gan, N., Cheema, A., Dakshanamurthy, S., Wang, X., Yang, D. C. H., & Chung, F. L. (2009). Cancer

338

preventive   isothiocyanates   induce   selective   degradation   of   cellular   α- and   β-tubulins by proteasomes. Journal of Biological Chemistry, 284(25), 17039–17051.

Mi, L., Gan, N., & Chung, F. L. (2010). Isothiocyanates inhibit proteasome activity and proliferation of multiple myeloma cells. Carcinogenesis, 32(2), 216–223.

Mian, I. S. (1998). Sequence, structural, functional, and phylogenetic analyzes of three glycosidase families. Blood Cells Mol Dis, 24(2), 83–100.

Miao, X., Bai, Y., Sun, W., Cui, W., Xin, Y., Wang, Y., et al. (2012). Sulforaphane prevention of diabetes-induced aortic damage was associated with the up-regulation of Nrf2 and its down-stream antioxidants. Nutrition & Metabolism, 9(1), 84.

Michaelsen, S., Otte, J., Simonsen, L.-O., & Sørensen, H. (1994). Absorption and degradation of individual intact glucosinolates in the digestive tract of rodents. Acta Agriculturae Scandinavica, Section a - Animal Science, 44(1), 25–37.

Mikkelsen, M. D., Buron, L. D., Salomonsen, B., Olsen, C. E., Hansen, B. G., Mortensen, U. H., & Halkier, B. A. (2012). Microbial production of indolylglucosinolate through engineering of a multi-gene pathway in a versatile yeast expression platform. Metabolic Engineering, 14(2), 104–111.

Mikkelsen, M. D., Petersen, B. L., Olsen, C. E., & Halkier, B. A. (2002). Biosynthesis and metabolic engineering of glucosinolates. Amino Acids, 22(3), 279–295.

Mikov, M. (1994). The metabolism of drugs by the gut flora. European Journal of Drug Metabolism and Pharmacokinetics, 19(3), 201–207.

Milder, I. E. J., Arts, I. C. W., Putte, B. V. de, Venema, D. P., & Hollman, P. C. H. (2007). Lignan contents of Dutch plant foods: a database including lariciresinol, pinoresinol, secoisolariciresinol and matairesinol. British Journal of Nutrition, 93(3), 393–402.

Miller, A. B., Altenburg, H.-P., Bueno-de-Mesquita, B., Boshuizen, H. C., Agudo, A., Berrino, F., et al. (2004). Fruits and vegetables and lung cancer: Findings from the European Prospective Investigation into Cancer and Nutrition. International Journal of Cancer, 108(2), 269–276.

Mithen, R. F. (2001). Glucosinolates – biochemistry, genetics and biological activity. Plant Growth Regulation, 34(1), 91–103.

Mithen, R. F., Armah, C., & Traka, M. (2010). Cruciferous vegetables – and biological activity of isothiocyanates and indoles. Dordrecht, Springer Netherlands, 1–30.

Mithen, R. F., Clarke, J., Lister, C., & Dean, C. (1995). Genetics of aliphatic glucosinolates. III. Side chain structure of aliphatic glucosinolates in. Heredity, 74(21), 0–215.

Mithen, R. F., Dekker, M., Verkerk, R., Rabot, S., & Johnson, I. T. (2000). The nutritional significance, biosynthesis and bioavailability of glucosinolates in human foods. Journal of the Science of Food and Agriculture, 80(7), 967–984.

Mithen, R. F., Faulkner, K., Magrath, R., Rose, P., Williamson, G., & Marquez, J. (2003). Development of isothiocyanate-enriched broccoli, and its enhanced ability to induce phase 2 detoxification enzymes in mammalian cells. Theor Appl Genet, 106(4), 727–734–734.

Mithen, R. F., Lewis, B. G., & Fenwick, G. R. (1986). In vitro activity of glucosinolates and their products against Leptosphaeria maculans. Transactions of the British Mycological Society, 87(3), 433–440.

Miyamori, I., Koshida, H., Matsubara, T., Ikeda, M., Takeda, Y., Takeda, R., & Vecsei, P. (1988). Role of intestinal bacteria in the metabolism of aldosterone in man. Hormone Research, 29(4), 147–150.

Møldrup, M. E., Geu-Flores, F., de Vos, M., Olsen, C. E., Sun, J., Jander, G., & Halkier, B. A. (2012). Engineering of benzylglucosinolate in tobacco provides proof-of-concept for dead-end trap crops genetically modified to attract Plutella xylostella (diamondback moth). Plant Biotechnology Journal, 10(4), 435–442.

Mougous, J. D., Green, R. E., Williams, S. J., Brenner, S. E., & Bertozzi, C. R. (2002). Sulfotransferases and sulfatases in mycobacteria. Chemistry & Biology, 9(7), 767–776.

Mullaney, J. A., Kelly, W., Mcghie, T. K., Ansell, J., & Heyes, J. A. (2013). Lactic acid bacteria convert

339

glucosinolates to nitriles efficiently yet differently to Enterobacteriaceae. Journal of Agricultural and Food Chemistry, DOI: 10.1021/jf305442j

Müller, I., Kahnert, A., Pape, T., Sheldrick, G. M., Meyer-Klaucke, W., Dierks, T., et al. (2004). Crystal structure   of   the   alkylsulfatase   AtsK:   insights   into   the   catalytic   mechanism   of   the   Fe   (II)   α-ketoglutarate-dependent dioxygenase superfamily. Biochemistry, 43(11), 3075–3088.

Munday, R., Mhawech-Fauceglia, P., Munday, C. M., Paonessa, J. D., Tang, L., Munday, J. S., et al. (2008). Inhibition of urinary bladder carcinogenesis by broccoli sprouts. Cancer Research, 68(5), 1593–1600.

Munday, R., & Munday, C. M. (2004). Induction of phase II detoxification enzymes in rats by plant-derived isothiocyanates: Comparison of allyl isothiocyanate with sulforaphane and related compounds. Journal of Agricultural and Food Chemistry, 52(7), 1867–1871.

Munday, R., Zhang, Y., Fahey, J. W., Jobson, H. E., Munday, C. M., Li, J., & Stephenson, K. K. (2006). Evaluation of isothiocyanates as potent inducers of carcinogen-detoxifying enzymes in the urinary bladder: critical nature of in vivo bioassay. Nutrition and Cancer, 54(2), 223–231.

Murooka, Y., Ishibashi, K., Yasumoto, M., Sasaki, M., Sugino, H., Azakami, H., & Yamashita, M. (1990). A sulfur-and tyramine-regulated Klebsiella aerogenes operon containing the arylsulfatase (atsA) gene and the atsB gene. Journal of Bacteriology, 172(4), 2131–2140.

Musk, S. R. R., & Johnson, I. T. (1993). Allyl isothiocyanate is selectively toxic to transformed cells of the human colorectal tumour line HT29. Carcinogenesis, 14(10), 2079–2083.

Musk, S. R. R., Smith, T. K., & Johnson, I. T. (1995). On the cytotoxicity and genotoxicity of allyl and phenethyl isothiocyanates and their parent glucosinolates sinigrin and gluconasturtiin. Mutation Research Letters, 348(1), 19–23.

Myette, J. R., Shriver, Z., Claycamp, C., McLean, M. W., Venkataraman, G., & Sasisekharan, R. (2003). The heparin/heparan sulfate 2-O-sulfatase from Flavobacterium heparinum. Journal of Biological Chemistry, 278(14), 12157–12166.

Myzak, M. C., Hardin, K., Wang, R., Dashwood, R. H., & Ho, E. (2006). Sulforaphane inhibits histone deacetylase activity in BPH-1, LnCaP and PC-3 prostate epithelial cells. Carcinogenesis, 27(4), 811–819.

Nadal, I., Donant, E., & Ribes-Koninckx, C. (2007). Imbalance in the composition of the duodenal microbiota of children with coeliac disease. Journal of Medical Microbiology, (56), 1669–1674.

Nakamura, Y. (2009). Chemoprevention by isothiocyanates: Molecular basis of apoptosis induction. In Forum of Nutrition (Vol. 61, pp. 170–181), Basel.

Nakamura, Y., Miyoshi, N., Takabayashi, S., & Osawa, T. (2004). Benzyl isothiocyanate inhibits oxidative stress in mouse skin: Involvement of attenuation of leukocyte infiltration. BioFactors, 21(1-4), 255–257.

Nakamura, Y., Ohigashi, H., Masuda, S., Murakami, A., Morimitsu, Y., Kawamoto, Y., et al. (2000). Redox regulation of glutathione S-transferase induction by benzyl isothiocyanate: Correlation of enzyme induction with the formation of reactive oxygen Intermediates. Cancer Research, 60(2), 219–225.

Navarro, S. L., Li, F., & Lampe, J. W. (2011). Mechanisms of action of isothiocyanates in cancer chemoprevention: an update. Food & Function, 2(10), 579–587.

Negrusz, A., Moore, C. M., McDonagh, N. S., Woods, E. F., Crowell, J. A., & Levine, B. S. (1998). Determination of phenethylamine, a phenethyl isothiocyanate marker, in dog plasma using solid-phase extraction and gas chromatography-mass spectrometry with chemical ionization. Journal of Chromatography B: Biomedical Sciences and Applications, 718(1), 193–198.

Neufeld, K. M., Kang, N., Bienenstock, J., & Foster, J. A. (2010). Reduced anxiety-like behavior and central neurochemical change in germ-free mice. Neurogastroenterology & Motility, 23(3), 255–e119.

Newman, R. M., Hanscom, Z., & Kerfoot, W. C. (1992). The watercress glucosinolate-myrosinase system: a feeding deterrent to caddisflies, snails and amphipods. Oecologia, 92(1), 1–7.

340

Nicholson, J. K., Holmes, E., Kinross, J., Burcelin, R., Gibson, G., Jia, W., & Pettersson, S. (2012). Host-gut microbiota metabolic interactions. Science, 336(6086), 1262–1267.

Nueno-Palop, C., & Narbad, A. (2011). Probiotic assessment of Enterococcus faecalis CP58 isolated from human gut. International Journal of Food Microbiology, 145(2-3), 390–394.

Nugon-Baudon, L., Rabot, S., Wal, J. M., & Szylit, O. (1990). Interactions of the intestinal microflora with glucosinolates in rapeseed meal toxicity: First evidence of an intestinal lactobacillus possessing a GSL-degrading activity in vivo. Journal of the Science of Food and Agriculture, 52(4), 547–559.

Nugon-Baudon, L., Szylit, O., & Raibaud, P. (1988). Production of toxic glucosinolate derivatives from rapeseed meal by intestinal microflora of rat and chicken. Journal of the Science of Food and Agriculture, 43(4), 299–308.

O' Farrell, P. H. (1975). High resolution two-dimensional electrophoresis of proteins. J Biol Chem, 250(10), 4007–4021.

Oginsky, E. L., Stein, A. E., & Greer, M. A. (1965). Myrosinase activity in bacteria as demonstrated by the conversion of progoitrin to goitrin. Proceedings of the Society for Experimental Biology and Medicine. Society for Experimental Biology and Medicine (New York, NY), 119(2), 360–364.

O'Hara, A. M., & Shanahan, F. (2006). The gut flora as a forgotten organ. EMBO Reports, 7(7), 688–693. Ohkawa, H., Ohkawa, R., Yamamoto, I., & Casida, J. E. (1972). Enzymatic mechanisms and toxicological

significance of hydrogen cyanide liberation from various organothiocyanates and organonitriles in mice and houseflies. Pesticide Biochemistry and Physiology, 2(1), 95–112.

Ohtsuru, M., & Hata, T. (1973). General characteristics of the intracellular myrosinase from Aspergilh niger. Agricultural and Biological Chemistry, 37, 2543–2548.

Ohtsuru, M., & Hata, T. (1979). The interaction of L-ascorbic acid with the active center of myrosinase. Biochimica Et Biophysica Acta (BBA)-Enzymology, 567(2), 384–391.

Okamura, H., Murooka, Y., & Harada, T. (1977). Tyramine oxidase and regulation of arylsulfatase synthesis in Klebsiella aerogenes. Journal of Bacteriology, 129(1), 59–65.

Okuda, H., Ogura, K., Kato, A., Takubo, H., & Watabe, T. (1998). A possible mechanism of eighteen patient deaths caused by interactions of sorivudine, a new antiviral drug, with oral 5-fluorouracil prodrugs. Journal of Pharmacology and Experimental Therapeutics, 287(2), 791–799.

O' Leary, W. (1989). Practical Handbook of Microbiology. CRC Press, Boston. Olguin,  L.  F.,  Askew,  S.  E.,  O’Donoghue,  A.  C.,  &  Hollfelder,  F.  (2008).  Efficient  catalytic  promiscuity  in  an  

enzyme superfamily: an arylsulfatase shows a rate acceleration of 1013 for phosphate monoester hydrolysis. Journal of the American Chemical Society, 130(49), 16547–16555.

Olsen, G. J., Lane, D. J., Giovannoni, S. J., Pace, N. R., & Stahl, D. A. (1986). Microbial ecology and evolution: A ribosomal RNA approach. Annual Review of Microbiology, 40(1), 337–365.

Olszak, T., An, D., Zeissig, S., Vera, M. P., Richter, J., Franke, A., et al. (2012). Microbial exposure during early life has persistent effects on natural killer T cell function. Science, 336(6080), 489–493.

Opitz, S. E. W., Mix, A., Winde, I. B., & Müller, C. (2011). Desulfation followed by sulfation: metabolism of benzylglucosinolate in Athalia rosae (Hymenoptera: Tenthredinidae). ChemBioChem, 12(8), 1252–1257.

Paal, K., ITO, M., & Withers, S. G. (2004). Paenibacillus sp. TS12 glucosylceramidase: kinetic studies of a novel sub-family of family 3 glycosidases and identification of the catalytic residues. Biochemical Journal, 378(Pt 1), 141.

Palcy, S., & Chevet, E. (2006). Integrating forward and reverse proteomics to unravel protein function. Proteomics, 6(20), 5467–5480.

Palmieri, S., Leoni, O., & Iori, R. (1982). A steady-state kineties study of myrosinase with direct ultraviolet spectrophotometric assay. Analytical Biochemistry, 123(2), 320–324.

Palmieri, S., Iori, R., & Leoni, O. (1987). Comparison of methods for determining myrosinase activity. Journal of Agricultural and Food Chemistry, 35(4), 617–621.

341

Palop, M. L. L., Smiths, J. P., & Brink, ten, B. (1995). Degradation of sinigrin by Lactobacillus agilis strain R16. International Journal of Food Microbiology, 26(2), 219–229.

Panagiotou, G., & Christakopoulos, P. (2004). NADPH-dependent D-aldose reductases and xylose fermentation in Fusarium oxysporum. Jbiosc, 97(5), 299–304.

Parenti, G., Meroni, G., & Ballabio, A. (1997). The sulfatase gene family. Current Opinion in Genetics & Development, 7(3), 386–391.

Park, S. Y., KIM, G. Y., Bae, S. J., Yoo, Y. H., & Choi, Y. H. (2007). Induction of apoptosis by isothiocyanate sulforaphane in human cervical carcinoma HeLa and hepatocarcinoma HepG2 cells through activation of caspase-3. Oncology Reports, 18(1), 181–187.

Parkin, I., Magrath, R., Keith, D., Sharpe, A., Mithen, R., & Lydiate, D. (1994). Genetics of aliphatic glucosinolates. II: Hydroxylation of alkenyl glucosinolates in Brassica napus. Heredity, 72, 594–598.

Parry, N. J., Beever, D. E., Owen, E., Vandenberghe, I., Van Beeumen, J., & Bhat, M. K. (2001). Biochemical characterization and mechanism of action of a thermostable beta-glucosidase purified from Thermoascus aurantiacus. Biochemical Journal, 353(Pt 1), 117.

Passonneau, J.V., Lowry, O.H. (1993). Enzymatic Analysis. A Practical Guide. Humana Press, Totowa NJ. Pearson, W. R., & Lipman, D. J. (1988). Improved tools for biological sequence comparison. Proceedings

of the National Academy of Sciences, 85(8), 2444–2448. Petroski, R. J., & Kwolek, W. F. (1985). Interactions of a fungal thioglucoside glucohydrolase and

cruciferous plant epithiospecifier protein to form 1-cyanoepithioalkanes: implications of an allosteric mechanism. Phytochemistry, 24(2), 213–216.

Ping, Z., Liu, W., Kang, Z., Cai, J., Wang, Q., Cheng, N., et al. (2010). Sulforaphane protects brains against hypoxic–ischemic injury through induction of Nrf2-dependent phase 2 enzyme. Brain Research, 1343(0), 178–185.

Piotrowski, M. (2004). Desulfo-glucosinolate sulfotransferases from Arabidopsis thaliana catalyze the final step in the biosynthesis of the glucosinolate core structure. Journal of Biological Chemistry, 279(49), 50717–50725.

Plant, A. R., Oliver, J. E., Patchett, M. L., Daniel, R. M., & Morgan, H. W. (1988). Stability and substrate specificity   of   a   β-glucosidase from the thermophilic bacterium Tp8 cloned into Escherichia coli. Archives of Biochemistry and Biophysics, 262(1), 181–188.

Pontoppidan, B., Ekbom, B., Eriksson, S., & Meijer, J. (2001). Purification and characterization of myrosinase from the cabbage aphid (Brevicoryne brassicae), a brassica herbivore. European Journal of Biochemistry, 268(4), 1041–1048.

Postma, P. W., Lengeler, J. W., & Jacobson, G. R. (1993). Phosphoenolpyruvate:carbohydrate phosphotransferase systems of bacteria. Microbiology and Molecular Biology Reviews, 57(3), 543–594.

Powolny, A. A., Bommareddy, A., Hahm, E. R., Normolle, D. P., Beumer, J. H., Nelson, J. B., & Singh, S. V. (2011). Chemopreventative potential of the cruciferous vegetable constituent phenethyl isothiocyanate in a mouse model of prostate cancer. Journal of the National Cancer Institute, 103(7), 571–584.

Pozzo,   T.,   Pasten,   J.   L.,   Karlsson,   E.  N.,  &   Logan,  D.  T.   (2010).   Structural   and   functional   analyses  of  β-glucosidase 3B from Thermotoga neapolitana: a thermostable three-domain representative of Glycoside Hydrolase 3. Journal of Molecular Biology, 397, 724–739.

Prawan, A., Saw, C. L. L., Khor, T. O., Keum, Y.-S., Yu, S., Hu, L., & Kong, A.-N. (2009). Anti-NF-κB  and  anti-inflammatory activities of synthetic isothiocyanates: Effect of chemical structures and cellular signaling. Chemico-Biological Interactions, 179(2-3), 202–211.

Prescott, A. G., & John, P. (1996). Dioxygenases: Molecular structure and role in plant metabolism. Annual Review of Plant Physiology and Plant Molecular Biology, 47(1), 245–271.

Prestera, T., Fahey, J. W., Holtzclaw, W. D., Abeygunawardana, C., Kachinski, L. J., & Talalay, P. (1996).

342

Comprehensive chromatographic and spectroscopic methods for the separation and identification of intact glucosinolates. Analytical Biochemistry, 239(2). 168-179.

Qin, J., Li, R., Raes, J., Arumugam, M., Burgdorf, K. S., Manichanh, C., et al. (2010). A human gut microbial gene catalogue established by metagenomic sequencing. Nature, 464(7285), 59–65.

Quevillon, E., Silventoinen, V., Pillai, S., Harte, N., Mulder, N., Apweiler, R., & Lopez, R. (2005). InterProScan: protein domains identifier. Nucleic Acids Research, 33(suppl 2), W116–W120.

Quinsac, A., Ribailler, D., Rollin, P., & Dreux, M. (1992). Analysis of 5-vinyl-1,3-oxazolidine-2-thione by liquid chromatography. Journal of AOAC International, 75(3), 529–536.

Rabilloud, T., Carpentier, G., & Tarroux, P. (1988). Improvement and simplification of low-background silver staining of proteins by using sodium dithionite. Electrophoresis, 9(6), 288–291.

Rabot, S., Guerin, C., Nugon-Baudon, L., & Szylit, O. (1995). Glucosinolate degradation by bacterial strains isolated from a human intestinal microflora. Proceedings From the 9th Internation Rapeseed Congress, Organized by GCIRC (Group Consultatif International De Recherches Sur Le Colza. in English: International Group of Research on Rapeseed).

Radajewski, S., Ineson, P., Parekh, N. R., & Murrell, J. C. (2000). Stable-isotope probing as a tool in microbial ecology. Nature, 403(6770), 646–649.

Radojcic Redovnikovic, I., Glivetic, T., Delonga, K., & Vorkapic-furac, J. (2008). Glucosinolates and their potential role in plant. Periodicum biologorum, 110(4), 297–309.

Rafii, F., Jackson, L. D., Ross, I., Heinze, T. M., Lewis, S. M., Aidoo, A., et al. (2007). Metabolism of daidzein by fecal bacteria in rats. Comparative Medicine, 57, 282–285.

Rafii, F., Sutherland, J. B., Hansen, E. B., Jr, & Cerniglia, C. E. (1997). Reduction of nitrazepam by Clostridium leptum, a nitroreductase-producing bacterium isolated from the human intestinal tract. Clinical infectious diseases, 25.

Rajendran, P., Ho, E., Williams, D. E., & Dashwood, R. H. (2011). Dietary phytochemicals, HDAC inhibition, and DNA damage/repair defects in cancer cells. Clinical Epigenetics, 3(1), 4.

Rajilić  Stojanović,  M.,  Smidt,  H.,  & De Vos, W. M. (2007). Diversity of the human gastrointestinal tract microbiota revisited. Environmental Microbiology, 9(9), 2125–2136.

Rakariyatham, N., & Sakorn, P. (2002). Biodegradation of glucosinolates in brown mustard seed meal (Brassica juncea) by Aspergillus sp. NR-4201 in liquid and solid-state cultures. Biodegradation, 13(6), 395–399.

Ratzka, A., Vogel, H., Kliebenstein, D. J., Mitchell-Olds, T., & Kroymann, J. (2002). Disarming the mustard oil bomb. Proceedings of the National Academy of Sciences, 99(17), 11223–11228.

Reichardt, N., Barclay, A. R., Weaver, L. T., & Morrison, D. J. (2011). Use of stable isotopes to measure the metabolic activity of the human intestinal microbiota. Applied and Environmental Microbiology, 77(22), 8009–8014.

Renu Nandakumar, R., Nandakumar, M. P., Marten, M. R., & Ross, J. M. (2005). Proteome analysis of membrane and cell wall associated proteins from Staphylococcus aureus. Journal of Proteome Research, 4(2), 250–257.

Renwick, A. G., Evans, S. P., Sweatman, T. W., Cumberland, J., & George, C. F. (1982). The role of the gut flora in the reduction of sulphinpyrazone in the rat. Biochemical Pharmacology, 31(16), 2649–2656.

Renwick, J. A. A., Radke, C. D., Sachdev-Gupta, K., & Städler, E. (1992). Leaf surface chemicals stimulating oviposition by Pieris rapae (Lepidoptera: Pieridae) on cabbage. Chemoecology, 3(1), 33–38.

Reuning, R. H., Shepard, T. A., Morrison, B. E., & Bockbrader, H. N. (1985). Formation of [20R]-dihydrodigoxin from digoxin in humans. Drug Metabolism and Disposition, 13(1), 51–57.

Roepstorff, P. (2012). Mass spectrometry based proteomics, background, status and future needs. Protein & Cell, 3(9), 641–647.

Rowan, T. G., Lawrence, T. L. J., & Kershaw, S. J. (1991). Effects of dietary copper and a probiotic on glucosinolate concentrations in ileal digesta and in faeces of growing pigs given diets based on

343

rapeseed meals. Animal Feed Science and Technology, 35(3–4), 247–258. Roy, A.B., & Williams, E. A. (1989). The sulphatase of H. pomatia. Purification and kinetic properties.

Comparative Biochemistry and Physiology Part B: Comparative Biochemistry, 93(2), 229–237. Rabot, S., Nugon-Baudon, L., Raibaud, P., & Szylit, O. (1993). Rape-seed meal toxicity in gnotobiotic rats:

influence of a whole human faecal flora or single human strains of Escherichia coli and Bacteroides vulgatus. British Journal of Nutrition, 70(01), 323.

Rask, L., Andréasson, E., Ekbom, B., Eriksson, S., Pontoppidan, B., & Meijer, J. (2000). Myrosinase: gene family evolution and herbivore defense in Brassicaceae. Plant Molecular Biology, 42(1), 93–114.

Riedl, M. A., Saxon, A., & Diaz-Sanchez, D. (2009). Oral sulforaphane increases Phase II antioxidant enzymes in the human upper airway. Clinical Immunology, 130(3), 244–251.

Ritz, S. A., Wan, J., & Diaz-Sanchez, D. (2007). Sulforaphane-stimulated phase II enzyme induction inhibits cytokine production by airway epithelial cells stimulated with diesel extract. American Journal of Physiology-Lung Cellular and Molecular Physiology, 292(1), L33–L39.

Roberfroid, M., Gibson, G. R., Hoyles, L., McCartney, A. L., Rastall, R., Rowland, I., et al. (2010). Prebiotic effects: metabolic and health benefits. British Journal of Nutrition, 104(S2), S1–S63.

Robert, C., Chassard, C., Lawson, P. A., & Bernalier-Donadille, A. (2007). Bacteroides cellulosilyticus sp. nov., a cellulolytic bacterium from the human gut microbial community. International Journal of Systematic and Evolutionary Microbiology, 57(7), 1516–1520.

Rodman, J. E., Karol, K. G., Price, R. A., & Sytsma, K. J. (1996). Molecules, morphology, and Dahlgren's expanded order Capparales. Systematic Botany, 21(3), 289–307.

Rosa, E., Heaney, R. K., & Fenwick, G. R. (1997). Glucosinolates in crop plants. Horticultural 19, 99–215. Rouzaud, G., Rabot, S., Ratcliffe, B., & Duncan, A. J. (2003). Influence of plant and bacterial myrosinase

activity on the metabolic fate of glucosinolates in gnotobiotic rats. British Journal of Nutrition, 90(2), 395–404.

Rouzaud, G., Young, S. A., & Duncan, A. J. (2004). Hydrolysis of glucosinolates to isothiocyanates after ingestion of raw or microwaved cabbage by human volunteers. Cancer Epidemiology Biomarkers & Prevention, 13(1), 125–131.

Rowan, T. G., Lawrence, T. L. J., & Kershaw, S. J. (1991). Effects of dietary copper and a probiotic on glucosinolate concentrations in ileal digesta and in faeces of growing pigs given diets based on rapeseed meals. Animal Feed Science and Technology, 35(3–4), 247–258.

Rungapamestry, V., Duncan, A. J., Fuller, Z., & Ratcliffe, B. (2007). Effect of cooking brassica vegetables on the subsequent hydrolysis and metabolic fate of glucosinolates. Proceedings of the Nutrition Society, 66(01), 69–81.

Russell, D. W., & Sambrook, J. (2001). Molecular Cloning. A Laboratory Manual. 3rd ed, Cold Spring Harbour, New York.

Saha, S., Hollands, W., Teucher, B., Needs, P. W., Narbad, A., Ortori, C. A., et al. (2012). Isothiocyanate concentrations and interconversion of sulforaphane to erucin in human subjects after consumption of commercial frozen broccoli compared to fresh broccoli. Molecular Nutrition & Food Research, 56(12), 1906-1916

Saier, M. H., Yamada, M., Erni, B., Suda, K., Lengeler, J., Ebner, R., et al. (1988). Sugar permeases of the bacterial phosphoenolpyruvate-dependent phosphotransferase system: sequence comparisons. The FASEB Journal, 2(3), 199–208.

Sakorn, P., Rakariyatham, N., Niamsup, H., & Kovitaya, P. (1999). Sinigrin degradation by Aspergillus sp. NR-4201 in liquid culture. Science Asia, 25(4), 189–194.

Sang, J. P., & Truscott, R. J. W. (1984). Liquid chromatographic determination of glucosinolates in rapeseed as desulfo-glucosinolates. Journal of the Association of Offical Analytical Chemists, 67(4), 829–833.

Sansenya, S., Opassiri, R., Kuaprasert, B., & Chen, C. J. (2011). The crystal structure of rice (Oryza sativa

344

L.) Os4BGlu12, an oligosaccharide and tuberonic acid glucoside-hydrolyzing   β-glucosidase with significant thioglucohydrolase activity. Archives of Biochemistry and Biophysics, 510, 62-72.

Santacruz, A., Marcos, A., Wärnberg, J., Martí, A., Martin Matillas, M., Campoy, C., et al. (2009). Interplay between weight loss and gut microbiota composition in overweight adolescents. Obesity, 17(10), 1906–1915.

Sardiello, M. (2005). Sulfatases and sulfatase modifying factors: an exclusive and promiscuous relationship. Human Molecular Genetics, 14(21), 3203–3217.

Sarikamis, G., Marquez, J., Maccormack, R., Bennett, R. N., Roberts, J., & Mithen, R. (2006). High glucosinolate broccoli: a delivery system for sulforaphane. Molecular Breeding, 18(3), 219–228.

Savage, D. C. (1977). Microbial ecology of the gastrointestinal tract. Annual Review of Microbiology, 31(1), 107–133.

Scheline, R. R. (1973). Metabolism of foreign compounds by gastrointestinal microorganisms. Pharmacological Reviews, (25), 452–514.

Schlegel, H. G., & Jannasch, H. W. (1967). Enrichment cultures. Annual Reviews in Microbiology, 21(1), 49–70.

Schlegel, H. G., & Zaborosch, C. (1993). General Microbiology. Cambridge University Press. Schmidt, B., Selmer, T., Ingendoh, A., & Figurat, K. (1995). A novel amino acid modification in sulfatases

that is defective in multiple sulfatase deficiency. Cell, 82(2), 271–278. Schneider, H., Schwiertz, A., Collins, M. D., & Blaut, M. (1999). Anaerobic transformation of quercetin-3-

glucoside by bacteria from the human intestinal tract. Archives of microbiology, 171(2), 81–91. Schnug, E., & Haneklaus, S. (1988). Theoretical principles for the indirect determination of the total

glucosinolate content in rapeseed and meal quantifying the sulphur concentration via X-ray fluorescence (X-RF method). Journal of the Science of Food and Agriculture, 45(3), 243–254.

Scopes, R. K. (1993). Protein Purification: Principles and Practice (3rd ed.), Springer, NY. Searle, L. M., CHAMBERLAIN, K., RAUSCH, T., & BUTCHER, D. N. (1982). The conversion of 3-

indolylmethylglucosinolate to 3-indolylacetonitrile by myrosinase, and its relevance to the clubroot disease of the Cruciferae. Journal of Experimental Botany, 33(5), 935–942.

Sehrawat, A., & Singh, S. V. (2011). Benzyl isothiocyanate inhibits epithelial-mesenchymal transition in cultured and xenografted human breast cancer cells. Cancer Prevention Research, 4(7), 1107–1117.

Sekirov, I., Russell, S. L., Antunes, L. C. M., & Finlay, B. B. (2010). Gut microbiota in health and disease. Physiological Reviews, 90(3), 859–904.

Selmer, T., Hallmann, A., Schmidt, B., Sumper, M., & Figura, K. (2004). The evolutionary conservation of a novel protein modification, the conversion of cysteine to serinesemialdehyde in arylsulfatase from Volvox carteri. European Journal of Biochemistry, 238(2), 341–345.

Seow, A., Vainio, H., & Yu, M. C. (2005). Effect of glutathione-S-transferase polymorphisms on the cancer preventive potential of isothiocyanates: An epidemiological perspective. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis, 592(1–2), 58–67.

Shapiro, T. A., Fahey, J. W., Wade, K. L., Stephenson, K. K., & Talalay, P. (2001). Chemoprotective glucosinolates and isothiocyanates of broccoli sprouts. Cancer Epidemiology Biomarkers & Prevention, 10(5), 501–508.

Shanahan, F. (2009). Therapeutic implications of manipulating and mining the microbiota. The Journal of Physiology, 587(17), 4175–4179.

Shanahan, F. (2011). The gut microbiota in 2011: Translating the microbiota to medicine. Nature Reviews Gastroenterology & Hepatology, 9(2), 72–74.

Sheehan, D., & O Sullivan, S. (2004). Fast protein liquid chromatography. Methods in Molecular Biology, 244, 253–258.

Shin, I. S., Masuda, H., & Naohide, K. (2004). Bactericidal activity of wasabi (Wasabia japonica) against Helicobacter pylori. International Journal of Food Microbiology, 94(3), 255–261.

345

Shu, Y. Z., Kingston, D. G. I., Van Tassell, R. L., & Wilkins, T. D. (1991). Metabolism of levamisole, an anti-colon cancer drug, by human intestinal bacteria. Xenobiotica, 21(6), 737–750.

Sibhatu, M. B., Smitherman, P. K., Townsend, A. J., & Morrow, C. S. (2008). Expression of MRP1 and GSTP1-1 modulate the acute cellular response to treatment with the chemopreventive isothiocyanate, sulforaphane. Carcinogenesis, 29(4), 807–815.

Sigrist, C., De Castro, E., & Cerutti, L. (2013). New and continuing developments at PROSITE. Nucleic Acids Research, 41(Database issue), D344-7.

Singh, S. V., Herman-Antosiewicz, A., Singh, A. V., Lew, K. L., Srivastava, S. K., Kamath, R., et al. (2004). Sulforaphane-induced G2/M phase cell cycle arrest involves checkpoint kinase 2-mediated phosphorylation of cell division cycle 25C. Journal of Biological Chemistry, 279(24), 25813–25822.

Slominski, B. A., Campbell, L. D., & Stanger, N. E. (1988). Extent of hydrolysis in the intestinal tract and potential absorption of intact glucosinolates in laying hens. Journal of the Science of Food and Agriculture, 42(4), 305–314.

Smith, M. I., Yatsunenko, T., Manary, M. J., Trehan, I., Mkakosya, R., Cheng, J., et al. (2013). Gut Microbiomes of Malawian Twin Pairs Discordant for Kwashiorkor. Science, 339(6119), 548–554.

Smith, S. C., Choy, R., Johnson, S. K., Hall, R. S., Wildeboer-Veloo, A. C. M., & Welling, G. W. (2006). Lupin kernel fiber consumption modifies fecal microbiota in healthy men as determined by rRNA gene fluorescent in situ hybridization. European Journal of Nutrition, 45(6), 335–341.

Smits, J. P., Knol, W., & Bol, J. (1993). Glucosinolate degradation by Aspergillus clavatus and Fusarium oxysporum in liquid and solid-state fermentation. Applied Microbiology and Biotechnology, 38(5), 696–701–701.

Smolinska, U., Morra, M. J., Knudsen, G. R., & James, R. L. (2003). Isothiocyanates produced by Brassicaceae species as inhibitors of Fusarium oxysporum. Plant Disease, 87(4), 407–412.

Sofia, H. J., Burland, V., Daniels, D. L., Plunkett, G., & Blattner, F. R. (1994). Analysis of the Escherichia coli genome. V. DNA sequence of the region from 76.0 to 81.5 minutes. Nucleic Acids Research, 22(13), 2576–2586.

Sørensen, H. (1990). Glucosinolates: Structure–Properties–Function. In Canola and Rapeseed. Production, Chemistry, Nutrition and Processing Technology. In Shahidi, F. (Ed.). Van Nostrand Reinhold Publishers, NY.

Sørensen J. C. (2001). Biorefined oilseed products – high quality products with added value. PhD Thesis. Chemistry Department, KVL, Denmark.

Sønderby, I. E., Geu-Flores, F., & Halkier, B. A. (2010). Biosynthesis of glucosinolates – gene discovery and beyond. Trends in Plant Science, 15(5), 283–290.

Sønderby, I. E., Hansen, B. G., Bjarnholt, N., Ticconi, C., Halkier, B. A., & Kliebenstein, D. J. (2007). A systems biology approach identifies a R2R3 MYB gene subfamily with distinct and overlapping functions in regulation of aliphatic glucosinolates. PLoS ONE, 2(12), e1322.

Sousa, T., Paterson, R., Moore, V., Carlsson, A., Abrahamsson, B., & Basit, A. W. (2008). The gastrointestinal microbiota as a site for the biotransformation of drugs. International Journal of Pharmaceutics, 363(1-2), 1–25.

Sowadski, J. M., Handschumacher, M. D., Krishna Murthy, H. M., Foster, B. A., & Wyckoff, H. W. (1985). Refined structure of alkaline phosphatase from Escherichia coli at 2.8 Å resolution. Journal of Molecular Biology, 186(2), 417–433.

Spector, D., Etienne, F., Brot, N., & Weissbach, H. (2003). New membrane-associated and soluble peptide methionine sulfoxide reductases in Escherichia coli. Biochemical and Biophysical Research Communications, 302(2), 284–289.

Staack, R., Kingston, S., Wallig, M. A., & Jeffery, E. H. (1998). A comparison of the individual and collective effects of four glucosinolate breakdown products from brussels sprouts on induction of detoxification enzymes. Toxicology and Applied Pharmacology, 149(1), 17–23.

346

Stawoska,  I.,  Gawęda,  S.,  Bielak-Lakomska,  M.,  Brindell,  M.,  Lewiński,  K.,  Laidler,  P.,  &  Stochel,  G.  (2010).  Mechanistic studies of the hydrolysis of p-nitrophenyl sulfate catalyzed by arylsulfatase from H. pomatia. Journal of Coordination Chemistry, 63(14-16), 2472–2487.

Steck, S. E., Gammon, M. D., Hebert, J. R., Wall, D. E., & Zeisel, S. H. (2007). GSTM1, GSTT1, GSTP1, and GSTA1 polymorphisms and urinary isothiocyanate metabolites following broccoli consumption in humans. The Journal of Nutrition, 137(4), 904–909.

Stenbaek, A., Hansson, A., Wulff, R. P., Hansson, M., Dietz, K.-J., & Jensen, P. E. (2008). NADPH-dependent thioredoxin reductase and 2-Cys peroxiredoxins are needed for the protection of Mg–protoporphyrin monomethyl ester cyclase. FEBS Letters, 582(18), 2773–2778.

Stoewsand, G. S. (1995). Bioactive organosulfur phytochemicals in Brassica oleracea vegetables—A review. Food and Chemical Toxicology, 33(6), 537–543.

Strong, H. A., Oates, J., Sembi, J., Renwick, A. G., & George, C. F. (1984a). Role of the gut flora in the reduction of sulfinpyrazone in humans. Journal of Pharmacology and Experimental Therapeutics, 230(3), 726–732.

Strong, H. A., Renwick, A. G., & George, C. F. (1984b). The site of reduction of sulphinpyrazone in the rabbit. Xenobiotica, 14(10), 815–826.

Strong, H. A., Renwick, A. G., George, C. F., Liu, Y. F., & Hill, M. J. (1987). The reduction of sulphinpyrazone and sulindac by intestinal bacteria. Xenobiotica, 17(6), 685–696.

Stryer, L., Berg, J. M., & Tymoczko, J. L. (2002). Biochemistry (5th ed.). W H Freeman & Company. Studier, W. F., Rosenberg, A. H., Dunn, J. J., & Dubendorff, J. W. (1990). Use of T7 RNA polymerase to

direct expression of cloned genes. Methods in Enzymology, 185, 60–89. Sultana, T., Savage, G. P., McNeil, D. L., Porter, N. G., Martin, R. J., & Deo, B. (2002). Effects of

fertilisation on the allyl isothiocyanate profile of above-ground tissues of New Zealand-grown wasabi. Journal of the Science of Food and Agriculture, 82(13), 1477–1482.

Szameit, C., Miech, C., Balleininger, M., Schmidt, B., Figura, von, K., & Dierks, T. (1999). The iron sulfur protein AtsB is required for posttranslational formation of formylglycine in the Klebsiella sulfatase. Journal of Biological Chemistry, 274(22), 15375–15381.

Takeno, S., Nakagawa, M., & Sakai, T. (1990). Teratogenic effects of nitrazepam in rats. Research Communications in Chemical Pathology and Pharmacology, 69(1), 59–70.

Talalay, P. (2000). Chemoprotection against cancer by induction of phase 2 enzymes. BioFactors, 12(1-4), 5–11.

Tang, L., & Zhang, Y. (2004). Dietary isothiocyanates inhibit the growth of human bladder carcinoma cells. The Journal of Nutrition, 134(8), 2004-2010.

Tang, L., Zirpoli, G. R., Guru, K., Moysich, K. B., Zhang, Y., Ambrosone, C. B., & McCann, S. E. (2008). Consumption of raw cruciferous vegetables is inversely associated with bladder cancer risk. Cancer Epidemiology Biomarkers & Prevention, 17(4), 938–944.

Tang, L., Zirpoli, G. R., Guru, K., Moysich, K. B., Zhang, Y., Ambrosone, C. B., & McCann, S. E. (2010). Intake of cruciferous vegetables modifies bladder cancer survival. Cancer Epidemiology Biomarkers & Prevention, 19(7), 1806–1811.

Tani, N., Ohtsuru, M., & Hata, T. (1974). Isolation of myrosinase producing microorganism. Agr. Biol. Chem, 38, 1617–1622.

Tanii, H., Higashi, T., Nishimura, F., Higuchi, Y., & Saijoh, K. (2008). Effects of cruciferous allyl nitrile on phase 2 antioxidant and detoxification enzymes. Medical science monitor : international medical journal of experimental and clinical research, 14(10), BR189–92.

Tap, J., Mondot, S., Levenez, F., Pelletier, E., Caron, C., Furet, J. P., et al. (2009). Towards the human intestinal microbiota phylogenetic core. Environmental Microbiology, 11(10), 2574–2584.

Tazuke, Y., Matsuda, K., Adachi, K., & Tsukada, Y. (1998). Purification and properties of a novel sulfatase from pseudomonas testosteroni that hydrolyzed 3. beta.-hydroxy-5-cholenoic acid 3-sulfate.

347

Bioscience, Biotechnology, and Biochemistry, 62(9), 1739–1744. Teplyakov, A., Lim, K., Zhu, P. P., Kapadia, G., Chen, C. C. H., Schwartz, J., et al. (2006). Structure of

phosphorylated enzyme I, the phosphoenolpyruvate:sugar phosphotransferase system sugar translocation signal protein. Pnas, 103(44), 16218–16223.

The Human Microbiome Project Consortium. (2012). Structure, function and diversity of the healthy human microbiome. Nature, 486, 207–214.

Thein, M., Sauer, G., Paramasivam, N., Grin, I., & Linke, D. (2010). Efficient subfractionation of gram-negative bacteria for proteomics studies. Journal of Proteome Research, 9(12), 6135–6147.

Thies, W. (1976). Quantitative gas liquid chromatography of glucosinolates on a microliter scale. Lipid/Fett, 78(6), 231–234.

Thies, W. (1977). Analysis of glucosinolates in seeds of rapeseed (Brassica napus L.): Concentration of glucosinolates by ion exchanges. Zeitschrift Fuer Pflanzenzuechtung, 79(4), 331–335.

Thies, W. (1978). Quantitative analysis of glucosinolates after their enzymatic desulfatation on ion exchange columns. Presented at the Proceedings of 5th Int Rapeseed Conference., Malmo. Vol. 1.

Thies, W. (1979). Detection and utilization of a glucosinolate sulfohydrolase in the edible snail, H. pomatia. Naturwissenschaften, 66(7), 364–365.

Thies, W. (1988). Isolation of sinigrin and glucotropaeolin from cruciferous seeds. Lipid/Fett, 90(8), 311–314.

Thimmulappa, R. K., Mai, K. H., Srisuma, S., Kensler, T. W., Yamamoto, M., & Biswal, S. (2002). Identification of Nrf2-regulated genes induced by the chemopreventive agent sulforaphane by oligonucleotide microarray. Cancer Research, 62(18), 5196–5203.

Thompson, J. (2002). Beta-glucoside kinase (BglK) from Klebsiella pneumoniae. purification, properties, and preparative synthesis of 6-phospho-beta-D-glucosides. Journal of Biological Chemistry, 277(37), 34310–34321.

Thompson, J. D. (1995). Introducing variable gap penalties to sequence alignment in linear space. Bioinformatics, 11(2), 181–186.

Thompson, J., Robrish, S. A., Bouma, C. L., Freedberg, D. I., & Folk, J. E. (1997). Phospho-beta-glucosidase from Fusobacterium mortiferum: purification, cloning, and inactivation by 6-phosphoglucono-delta-lactone. Journal of Bacteriology, 179(5), 1636–1645.

Tjalsma, H., Boleij, A., Marchesi, J. R., & Dutilh, B. E. (2012). A bacterial driver–passenger model for colorectal cancer: beyond the usual suspects. Nature Reviews Microbiology, 10(8), 575–582.

Tokheim, A. M., Spannaus-Martin, D. J., & Martin, B. L. (2005). Evidence for the Cd2+ activation of the aryl sulfatase from H. pomatia. BioMetals, 18(5), 537–540.

Tookey, H. L. (1973). Crambe thioglucoside glucohydrolase (EC 3.2.3.1): separation of a protein required for epithiobutane formation. Biochemistry and Cell Biology, 51(12), 1654–1660.

Tookey, H. L., & Wolff, I. A. (1970). Effect of organic reducing agents and ferrous ion on thioglucosidase activity of Crambe abyssinica seed. Biochemistry and Cell Biology, 48(9), 1024–1028.

Tozaki, H., Emi, Y., Horisaka, E., Fujita, T., Yamamota, A., & Muranishi, S. (1997). Degradation of insulin and calcitonin and their protection by various protease inhibitors in rat caecal contents: Implications in peptide delivery to the colon. Journal of Pharmacy and Pharmacology, 49(2), 164–168.

Traka, M., & Mithen, R. (2009). Glucosinolates, isothiocyanates and human health. Phytochemistry Reviews, 8(1), 269–282.

Trehan, I., Goldbach, H. S., LaGrone, L. N., Meuli, G. J., Wang, R. J., Maleta, K. M., & Manary, M. J. (2013). Antibiotics as part of the management of severe acute malnutrition. New England Journal of Medicine, 368(5), 425–435.

Tribolo, S., Berrin, J.-G., Kroon, P. A., Czjzek, M., & Juge, N. (2007). The crystal structure of human cytosolic  β-glucosidase unravels the substrate aglycone specificity of a family 1 glycoside hydrolase.

348

Journal of Molecular Biology, 370(5), 964–975. Tripathi, M. K., & Mishra, A. S. (2007). Glucosinolates in animal nutrition. Animal Feed Science and

Technology, 132(1–2), 1–27. Tsao, R., Reuber, M., Johnson, L., & Coats, J. R. (1996). Insecticidal toxicities of glucosinolate–containing

extracts from crambe seeds. J. Agric. Entomol, 13(2), 109–120. Turner, J. G., Ellis, C., & Devoto, A. (2002). The jasmonate signal pathway. The Plant Cell On-line,

14(suppl 1), S153–S164. Uchiyama, T., Abe, T., Ikemura, T., & Watanabe, K. (2004). Substrate-induced gene-expression screening

of environmental metagenome libraries for isolation of catabolic genes. Nature Biotechnology, 23(1), 88–93.

Uchiyama, T., & Watanabe, K. (2008). Substrate-induced gene expression (SIGEX) screening of metagenome libraries. Nature Protocols, 3(7), 1202–1212.

Uda, Y., Kurata, T., & Arakawa, N. (1986). Effects of pH and ferrous ion on the degradation of glucosinolates by myrosinase. Agricultural and Biological Chemistry, 50(11), 2735–2740.

Uematsu, Y., Hirata, K., Suzuki, K., Iida, K., Ueta, T., & Kamata, K. (2002). Determination of isothiocyanates and related compounds in mustard extract and horseradish extract used as natural food additives. Shokuhin Eiseigaku Zasshi, 43(1), 10–17.

Underhill, E. W., & Kirkland, D. F. (1971). Gas chromatography of trimethylsilyl derivatives of glucosinolates. Journal of Chromatography A, 57, 47–54.

Vaishampayan, P. A., Kuehl, J. V., Froula, J. L., Morgan, J. L., Ochman, H., & Francino, M. P. (2010). Comparative metagenomics and population dynamics of the gut microbiota in mother and infant. Genome Biology and Evolution, 2(0), 53–66.

Valerino, D. M., Johns, D. G., Zaharko, D. S., & Oliverio, V. T. (1972). Studies of the metabolism of methotrexate by intestinal flora—I: Identification and study of biological properties of the metabolite 4-amino-4-deoxy-N10-methylpteroic acid. Biochemical Pharmacology, 21(6), 821–831.

Vallejo, F., Tomás-Barberán, F. A., & Ferreres, F. (2004). Characterisation of flavonols in broccoli (Brassica oleracea L. var. italica) by liquid chromatography–UV diode-array detection–electrospray ionisation mass spectrometry. Journal of Chromatography A, 1054(1), 181–193.

van Doorn, H. E., van der Kruk, G. C., van Holst, G. J., Raaijmakers Ruijs, N. C., Postma, E., Groeneweg, B., & Jongen, W. H. (1999). The glucosinolates sinigrin and progoitrin are important determinants for taste preference and bitterness of Brussels sprouts. Journal of the Science of Food and Agriculture, 78(1), 30–38.

VanEtten, C. H., Daxenbichler, M. E., Peters, J. E., & Tookey, H. L. (1966). Variation in enzymatic degradation products from major thioglucosides in Crambe abyssinica and Brassica napus seed meals. Journal of Agricultural and Food Chemistry, 14(4), 426–430.

van Hogezand, R. A., Kennis, H. M., Schaik, A., Koopman, J. P., Hees, P. A. M., & Tongeren, J. H. M. (1992). Bacterial acetylation of 5-aminosalicylic acid in faecal suspensions cultured under aerobic and anaerobic conditions. European Journal of Clinical Pharmacology, 43(2), 189–192–192.

Varghese, J. N., Hrmova, M., & Fincher, G. B. (2000). Three-dimensional  structure  of  a  barley  β-D-glucan exohydrolase, a family 3 glycosyl hydrolase. Proteins: Structure, Function, and Bioinformatics, 41(2), 257–269.

Verduyn, C., Van Kleef, R., Frank, J., Schreuder, H., & Van Dijken, J. P. (1985). Properties of the NAD(P)H-dependent xylose reductase from the xylose-fermenting yeast Pichia stipitis. Biochemical Journal, 226(3), 669–677.

Verhoeven, D. T., Goldbohm, R. A., van Poppel, G., Verhagen, H., & van den Brandt, P. A. (1996). Epidemiological studies on brassica vegetables and cancer risk. Cancer Epidemiology Biomarkers & Prevention, 5(9), 733–748.

Verkerk, R., Van der Gaag, M. S., Dekker, M., & Jongen, W. (1997). Effects of processing conditions on

349

glucosinolates in cruciferous vegetables. Cancer Letters, 114(1), 193–194. Vermes, A. A. S., Kuijper, E. J., Guchelaar, H. J., & Dankert, J. (2003). An in vitro study on the active

conversion of flucytosine to fluorouracil by microorganisms in the human intestinal microflora. Chemotherapy, 49(1-2), 17–23.

Vermeulen, M. (2009). Isothiocyanates from cruciferous vegetables: kinetics, biomarkers and effects. Wageningen University. Retrieved at September 2, 2012, from the website temoa : Open Educational Resources (OER) Portal at http://www.temoa.info/node/409634

Vermeulen, M., van den Berg, R., Freidig, A. P., van Bladeren, P. J., & Vaes, W. H. J. (2006). Association between consumption of cruciferous vegetables and condiments and excretion in urine of isothiocyanate mercapturic acids. Journal of Agricultural and Food Chemistry, 54(15), 5350–5358.

Vissavajjhala P, Mero G, Bosch F, et al. (2011). Beneforte® broccoli with 2–3x phytonutrient glucoraphanin and antioxidant enzyme boosting potential. In: Dietary Bioactive Components IV. Anti-oxidant and Anti-inflammatory Functions I. FASEB J. (Meeting Abstracts), 25, 773.

Vocadlo, D. J., Mayer, C., He, S., & Withers, S. G. (2000). Mechanism of action and identification of Asp242 as the catalytic nucleophile of Vibrio furnisii N-acetyl-β-D-glucosaminidase using 2-acetamido-2-deoxy-5-fluoro-α-L-idopyranosyl fluoride. Biochemistry, 39(1), 117–126.

Voorrips, L. E., Goldbohm, R. A., van Poppel, G., Sturmans, F., Hermus, R., & van den Brandt, P. A. (2000). Vegetable and fruit consumption and risks of colon and rectal cancer in a prospective cohort study The Netherlands Cohort Study on Diet and Cancer. American Journal of Epidemiology, 152(11), 1081–1092.

von Weymarn, L. B., Chun, J. A., Knudsen, G. A., & Hollenberg, P. F. (2007). Effects of eleven isothiocyanates on P450 2A6- and 2A13-catalyzed coumarin 7-hydroxylation. Chem. Res. Toxicol., 20(9), 1252–1259.

Vrieze, A., Holleman, F., Zoetendal, E. G., De Vos, W. M., Hoekstra, J. B. L., & Nieuwdorp, M. (2010). The environment within: how gut microbiota may influence metabolism and body composition. Diabetologia, 53(4), 606–613.

Vuong, T. V., & Wilson, D. B. (2010). Glycoside hydrolases: catalytic base/nucleophile diversity. Biotechnology and bioengineering, 107(2), 195–205.

Wallsgrove, R. M., & Bennett, R. N. (1995). The biosynthesis of glucosinolates in Brassicas. Seminar Series - Society for Experimental Biology, (56), 243–259.

Walsh, C. T., & Levine, R. R. (1975). Studies of the enterohepatic circulation of morphine in the rat. Journal of Pharmacology and Experimental Therapeutics, 195, 303–310.

Walters, D. G., Young, P. J., Agus, C., Knize, M. G., Boobis, A. R., Gooderham, N. J., & Lake, B. G. (2004). Cruciferous vegetable consumption alters the metabolism of the dietary carcinogen 2-amino-1-methyl-6-phenylimidazo [4, 5-b] pyridine (PhIP) in humans. Carcinogenesis, 25(9), 1659–1669.

Wang, L. G., & Chiao, J. W. (2010). Prostate cancer chemopreventive activity of phenethyl isothiocyanate through epigenetic regulation. International Journal of Oncology, 37(3), 533–539.

Wang, L. G., Liu, X. M., Fang, Y., Dai, W., Chiao, F. B., Puccio, G. M., et al. (2008). De-repression of the p21 promoter in prostate cancer cells by an isothiocyanate via inhibition of HDACs and c-Myc. International Journal of Oncology, 33(2), 375–380.

Wang, M., Sun, X., Tan, D., Gong, S., Meijer, J., & Zhang, J. (2009). The two non-functional myrosinase genes TGG3 and TGG6 in Arabidopsis are expressed predominantly in pollen. Plant Science, 177(4), 371–375.

Wang, W., Wang, S., Howie, A. F., Beckett, G. J., Mithen, R., & Bao, Y. (2005). Sulforaphane, erucin, and iberin up-regulate thioredoxin reductase 1 expression in human NCF-7 cells. Journal of Agricultural and Food Chemistry, 53(5), 1417–1421.

Wathelet, J. P., Lori, R., Leoni, O., Rollin, P., Mabon, N., Marlier, M., & Palmieri, S. (2001). A recombinant β-O-glucosidase from Caldocellum saccharolyticum to hydrolyze desulfo-glucosinolates.

350

Biotechnology Letters, 23(6), 443–446. Wathelet, J. P., Mabon, N., & Marlier, M. (2008). Determination of glucosinolates in rapeseed

improvement of the official HPLC ISO method (precision and speed). The 10th International Rapeseed Congress, Canberra, Australia.

Watson, J. D. (2007). Recombinant DNA. W. H. Freeman. Westbrook, J. A., Yan, J. X., Wait, R., Welson, S. Y., & Dunn, M. J. (2001). Zooming-in on the proteome:

Very narrow-range immobilised pH gradients reveal more protein species and isoforms. Electrophoresis, 22(14), 2865–2871.

Wetter, L. R., & Chisholm, M. D. (1968). Sources of sulfur in the thioglucosides of various higher plants. Biochemistry and Cell Biology, 46(8), 931–935.

Wilkins, M. R., Sanchez, J. C., Gooley, A. A., Appel, R. D., Humphery-Smith, I., Hochstrasser, D. F., & Williams, K. L. (1996). Progress with proteome projects: why all proteins expressed by a genome should be identified and how to do it. Biotechnology & Genetic Engineering Reviews, 13, 19–50–50.

Winogradsky, S. (1890). Recherches sur les organismes de la nitrification. Ann Inst Pasteur, 4(257), 11–12.

Wintzingerode, V. F., Göbel, U. B., & Stackebrandt, E. (1997). Determination of microbial diversity in environmental   samples:   pitfalls   of   PCR-­‐based   rRNA   analysis.   FEMS Microbiology Reviews, 21(3), 213–229.

Wiseman, H., Casey, K., Bowey, E. A., Duffy, R., Davies, M., Rowland, I. R., et al. (2004). Influence of 10 wk of soy consumption on plasma concentrations and excretion of isoflavonoids and on gut microflora metabolism in healthy adults. The American Journal of Clinical Nutrition, 80(3), 692–699.

Witt, E., Frank, R., & Hengstenberg, W. (1993). 6-Phospho-β-galactosidases of Gram-positive and 6-phospho-β-glucosidase B of Gram-negative bacteria: comparison of structure and function by kinetic and immunological methods and mutagenesis of the lacG gene of Staphylococcus aureus. Protein Engineering, 6(8), 913–920.

Wittstock, U., Agerbirk, N., Stauber, E. J., Olsen, C. E., Hippler, M., Mitchell-Olds, T., et al. (2004). Successful herbivore attack due to metabolic diversion of a plant chemical defense. Proceedings of the National Academy of Sciences, 101(14), 4859–4864.

Wittstock, U., & Burow, M. (2007). Tipping the scales-Specifier proteins in glucosinolate hydrolysis. IUBMB Life, 59(12), 744–751.

Wittstock, U., & Burow, M. (2010). Glucosinolate breakdown in Arabidopsis: Mechanism, regulation and biological significance. The Arabidopsis Book (Vol. 8). doi: 10.1199/tab.0134

Wittstock, U., Fischer, M., Svendsen, I., & Halkier, B. A. (2000). Cloning and characterization of two cDNAs encoding sulfatases in the Roman snail, H. pomatia. IUBMB Life (International Union of Biochemistry and Molecular Biology: Life), 49(1), 71–76.

Wittstock, U., & Halkier, B. A. (2002). Glucosinolate research in the Arabidopsis era. Trends in Plant Science, 7(6), 263–270.

Wittstock, U., Kliebenstein, D. J., Lambrix, V., Reichelt, M., & Gershenzon, J. (2003). Chapter five: Glucosinolate hydrolysis and its impact on generalist and specialist insect herbivores. In J. T. Romeo (Ed.), Recent Advances in Phytochemistry (Vol. 37, pp. 101–125).

Wobus, A. M., & Löser, P. (2011). Present state and future perspectives of using pluripotent stem cells in toxicology research. Archives of Toxicology, 85(2), 79–117.

Woo, K. J., & Kwon, T. K. (2007). Sulforaphane suppresses lipopolysaccharide-induced cyclooxygenase-2 (COX-2) expression through the modulation of multiple targets in COX-2 gene promoter. International Immunopharmacology, 7(13), 1776–1783.

Wu, C. L., Huang, A. C., Yang, J. S., Liao, C. L., Lu, H. F., Chou, S. T., et al. (2011). Benzyl isothiocyanate (BITC) and phenethyl isothiocyanate (PEITC)-mediated generation of reactive oxygen species causes cell cycle arrest and induces apoptosis via activation of caspase-3, mitochondria dysfunction and

351

nitric oxide (NO) in human osteogenic sarcoma U-2 OS cells. Journal of Orthopaedic Research, 29(8), 1199–1209.

Wu, S. J., Ng, L. T., & Lin, C. C. (2005). Effects of antioxidants and caspase-3 inhibitor on the phenylethyl isothiocyanate-induced apoptotic signaling pathways in human PLC/PRF/5 cells. European Journal of Pharmacology, 518(2–3), 96–106.

Xiao, D., Johnson, C. S., Trump, D. L., & Singh, S. V. (2004). Proteasome-mediated degradation of cell division cycle 25C and cyclin-dependent kinase 1 in phenethyl isothiocyanate-induced G2-M-phase cell cycle arrest in PC-3 human prostate cancer cells. Molecular Cancer Therapeutics, 3(5), 567–576.

Xiao, D., Srivastava, S. K., Lew, K. L., Zeng, Y., Hershberger, P., Johnson, C. S., et al. (2003). Allyl isothiocyanate, a constituent of cruciferous vegetables, inhibits proliferation of human prostate cancer cells by causing G2/M arrest and inducing apoptosis. Carcinogenesis, 24(5), 891–897.

Xiao, D., Vogel, V., & Singh, S. V. (2006). Benzyl isothiocyanate–induced apoptosis in human breast cancer cells is initiated by reactive oxygen species and regulated by Bax and Bak. Molecular Cancer Therapeutics, 5(11), 2931–2945.

Xie, Y., Kim, K. J., & Kim, K. S. (2004). Current concepts on Escherichia coli K1 translocation of the blood–brain barrier. FEMS Immunology & Medical Microbiology, 42(3), 271–279.

Xu, C., Shen, G., Chen, C., Gélinas, C., & Kong, A. N. T. (2005). Suppression of NF-κB  and  NF-κB-regulated gene  expression  by  sulforaphane  and  PEITC   through   IκBα,   IKK  pathway   in  human  prostate  cancer  PC-3 cells. Oncogene, 24(28), 4486–4495.

Xu, Z., Escamilla-Treviño, L., Zeng, L., Lalgondar, M., Bevan, D., Winkel, B., et al. (2004). Functional genomic analysis of Arabidopsis thaliana glycoside hydrolase family 1. Plant Molecular Biology, 55(3), 343–367.

Xue, J., Pihlgren, U., & Rask, L. (1993). Temporal, cell-specific, and tissue-preferential expression of myrosinase genes during embryo and seedling development in Sinapis alba. Planta, 191(1), 95–101.

Xue, M., Qian, Q., Adaikalakoteswari, A., Rabbani, N., Babaei-Jadidi, R., & Thornalley, P. J. (2008). Activation of NF-E2-related factor-2 reverses biochemical dysfunction of endothelial cells Induced by hyperglycemia linked to vascular disease. Diabetes, 57(10), 2809–2817.

Yagur-Kroll, S., Ido, A., & Amster-Choder, O. (2009). Spatial arrangement of the beta-glucoside transporter from Escherichia coli. Journal of Bacteriology, 191(9), 3086–3094.

Yang, G., Gao, Y. T., Shu, X. O., Cai, Q., Li, G. L., Li, H. L., et al. (2010). Isothiocyanate exposure, glutathione S-transferase polymorphisms, and colorectal cancer risk. American Journal of Clinical Nutrition, 91(3), 704–711.

Yates, J. R., Ruse, C. I., & Nakorchevsky, A. (2009). Proteomics by mass spectrometry: approaches, advances, and applications. Annual Review of Biomedical Engineering, 11(1), 49–79.

Yatsunenko, T., Rey, F. E., Manary, M. J., Trehan, I., Dominguez-Bello, M. G., Contreras, M., et al. (2012). Human gut microbiome viewed across age and geography. Nature, 486(7402), 222–227.

Ye, L., Dinkova-Kostova, A. T., Wade, K. L., Zhang, Y., Shapiro, T. A., & Talalay, P. (2002). Quantitative determination of dithiocarbamates in human plasma, serum, erythrocytes and urine: pharmacokinetics of broccoli sprout isothiocyanates in humans. Clinica Chimica Acta, 316(1–2), 43–53.

Ye, L., & Zhang, Y. (2001). Total intracellular accumulation levels of dietary isothiocyanates determine their activity in elevation of cellular glutathione and induction of Phase 2 detoxification enzymes. Carcinogenesis, 22(12), 1987–1992.

Yin, P., Kawamura, T., He, M., Vanaja, D. K., & Young, C. Y. F. (2009). Phenethyl isothiocyanate induces cell   cycle   arrest   and   reduction   of   α- and   β-tubulin isotypes in human prostate cancer cells. Cell Biology International, 33(1), 57–64.

Youn, H. S., Kim, Y. S., Park, Z. Y., Kim, S. Y., Choi, N. Y., Joung, S. M., et al. (2010). Sulforaphane suppresses oligomerization of TLR4 in a thiol-dependent manner. The Journal of Immunology,

352

184(1), 411–419. Yoxall, V., Kentish, P., Coldham, N., Kuhnert, N., Sauer, M. J., & Ioannides, C. (2005). Modulation of

hepatic cytochromes P450 and phase II enzymes by dietary doses of sulforaphane in rats: Implications for its chemopreventive activity. International Journal of Cancer, 117(3), 356–362.

Zakkar, M., Van der Heiden, K., Luong, L. A., Chaudhury, H., Cuhlmann, S., Hamdulay, S. S., et al. (2009). Activation of Nrf2 in endothelial cells protects arteries from exhibiting a proinflammatory state. Arteriosclerosis, Thrombosis, and Vascular Biology, 29(11), 1851–1857.

Zechel, D. L., & Withers, S. G. (2000). Glycosidase mechanisms: anatomy of a finely tuned catalyst. Accounts of chemical research, 33(1), 11–18.

Zhang, D. D., & Hannink, M. (2003). Distinct cysteine residues in Keap1 are required for Keap1-dependent ubiquitination of Nrf2 and for stabilization of Nrf2 by chemopreventive agents and oxidative stress. Molecular and Cellular Biology, 23(22), 8137–8151.

Zhang, R., Loganathan, S., Humphreys, I., & Srivastava, S. K. (2006). Benzyl isothiocyanate-induced DNA damage causes G2/M cell cycle arrest and apoptosis in human pancreatic cancer cells. The Journal of Nutrition, 136(11), 2728–2734.

Zhang, Y. (2000). Role of glutathione in the accumulation of anticarcinogenic isothiocyanates and their glutathione conjugates by murine hepatoma cells. Carcinogenesis, 21(6), 1175–1182.

Zhang, Y. (2004). Phytochemicals in health and disease. In Bao, Y., Fenwick, R. (Eds). Oxidative Stress and Disease. CRC Press

Zhang, Y. (2011). The molecular basis that unifies the metabolism, cellular uptake and chemopreventive activities of dietary isothiocyanates. Carcinogenesis, 33(1), 2–9.

Zhang, Y., & Talalay, P. (1998). Mechanism of differential potencies of isothiocyanates as inducers of anticarcinogenic Phase 2 enzymes. Cancer Research, 58(20), 4632–4639.

Zhang, Y., Talalay, P., Cho, C. G., & Posner, G. H. (1992). A major inducer of anticarcinogenic protective enzymes from broccoli: Isolation and elucidation of structure. Proceedings of the National Academy of Sciences, 89(6), 2399–2403.

Zhang, Y., & Tang, L. (2007). Discovery and development of sulforaphane as a cancer chemopreventive phytochemical. Acta Pharmacologica Sinica, 28(9), 1343–1354.

Zhu, H., Jia, Z., Strobl, J. S., Ehrich, M., Misra, H. P., & Li, Y. (2008). Potent induction of total cellular and mitochondrial antioxidants and phase 2 enzymes by cruciferous sulforaphane in rat aortic smooth muscle cells: Cytoprotection against oxidative and electrophilic stress. Cardiovascular Toxicology, 8(3), 115–125.

353

APPENDIX: APPENDIX I

A) Representative GC-MS chromatogram showing no degradation products from the negative control containing only GSL substrate without bacterial cells or only bacterial cells without GSLs incubated  in  the  culture  broths  for  24  h  at  37˚C  under  anaerobic  conditions.

B) Representative GC-MS chromatogram showing no degradation products from the negative control containing only DS-GSL substrate without bacterial cells incubated in the culture broths for  24  h  at  37˚C  under  anaerobic  conditions.

5 . 0 0 1 0 . 0 0 1 5 . 0 0 2 0 . 0 0 2 5 . 0 0 3 0 . 0 0 3 5 . 0 0 4 0 . 0 0

8 0 0 0

1 0 0 0 0

1 2 0 0 0

1 4 0 0 0

1 6 0 0 0

1 8 0 0 0

2 0 0 0 0

2 2 0 0 0

2 4 0 0 0

2 6 0 0 0

2 8 0 0 0

3 0 0 0 0

3 2 0 0 0

3 4 0 0 0

3 6 0 0 0

3 8 0 0 0

4 0 0 0 0

4 2 0 0 0

T i m e - - >

A b u n d a n c e

T I C : V I N 0 2 0 2 1 3 0 8 . D

5 . 0 0 1 0 . 0 0 1 5 . 0 0 2 0 . 0 0 2 5 . 0 0 3 0 . 0 0 3 5 . 0 0 4 0 . 0 0

1 0 0 0 0

1 5 0 0 0

2 0 0 0 0

2 5 0 0 0

3 0 0 0 0

3 5 0 0 0

4 0 0 0 0

4 5 0 0 0

T i m e - - >

A b u n d a n c e

T I C : V I N 0 2 0 2 1 3 0 6 . D

354

APPENDIX II List of gene sequencing results from the recombinant plasmids used in this work SUL2 gene >9643339.seq - ID: SUL2-T7 on 2012/5/10-11:58:12 automatically edited with PhredPhrap, start with base no.: 10 Internal Params: Windowsize: 20, Goodqual: 19, Badqual: 10, Minseqlength: 50, nbadelimit: 1 ctcnnaAAAtaTTTTtGTTTtACTTTAagnaggagaTATACCatgGGcAGCAGCcatcatcatCATcATCACAGCAGCGGCctGgtGCCGCGCGGCAGCCAggtCATATgAAACGCCCCAATTTTCTGTTCATCAtgACCGATACCCAGGCCACCAATATGGTCGGtTgCTATagCggtAAGCCGctgAATACGCAAAATATTGATAGTCTGGCGGCGGAAAGTATTCGCTTTAATTCCGCCTACACctgTTCACCGGTTTGTACACCTGCACGCGCCGGACTATTCACCGGTATCTACGCTAACCAGTCCGGCCCGTgGACCAACAACGTCGCGCCGGGCAAAAACATTTCCACCATGGGGCGCTACTTTAAGGATGCGGGCTATCACACCTGCTACATCGGCAAATGGCATCTCGATGGACATGACTATTTCGGCACTGGCgaGTGTCCGCCGGAGTGGGACGCtGATTACTGGTTCGATGGAGCGAACTATCTTAGTGAACTGACGgAGAAAGAGATTAGCCTGTGGCGCAATGGCCTAAACAGCGTtgAGGATTTACAGGCGAACCATATTGACGAAACCTTCACCTGGGCGCACCGCATCAGCAATCGGGCGGTGGATTTTCTGCAACAGCCCGCGCGCGCCGACGAACCTTTCCTGATGGTGATTTCGTATGATGAGCCGCATCACCCGTTCACCTGTCCGGTGGAGTATTTAGAGAAATACACTGATTTTTACTACGAACTGGGTGAGAAAGCAGAGGATGACCTGGCGAACAAACCGGAACATCACCGCTTATGGGCGCAGGCGATGCCATCGCCAGTCGGTGATGACGGGCTTTATCACCATCCGCTCTATTTTGCCTGCAATGACTTTGTTGATGACCAAATCggACGGGTCATCAACgccTTAACGCCAGagcAACGTGAAnATACGtGGGGTTATTTATACcnccnaTCACGGCGAAaatGATGGggcGCa GH3#1 gene >6831964.seq - ID: GH3/1-T7 on 2010/8/19-5:41:18 automatically edited with PhredPhrap, start with base no.: 21 Internal Params: Windowsize: 20, Goodqual: 19, Badqual: 10, Minseqlength: 50, nbadelimit: 1 AATCGGCCAATTGGTGCAGTTATCTGGAGAATTCTTTCACngnnncGatTTGTCTTTGGGTCCTCAGCAAAAACTTGGCATCGAACAACAAACCATCGATGTTGTCGGTTCGGTATTGAACGTCACAGGTGCCCAAGTAACTCGAAAGATTCAAACAGACTACTTAAGAAAAAGTCGACACAAAATCCCACTATTGTTCATGGCAGATATCATCTATGGATACCGAACAGTCTTTCCGATCCCTCTGGGATTGGGAGCAACCTGGAATCCAGCATTGATTCAAAGCGCCTATCAAGCCGCAGCACAAGAAGCAAGAGCGGCTGGCGCACACGTAACATATGCACCAATGGTCGACTTAGTACGCGATGCTCGATGGGGGCGATGCTTGGAGTCGACAGGAGAGGATCCGCTGCTAAATGCTGATTTTGCGAAAGCAATGGTAGAAGGCATTCAGCAAGAAAAAGGGGGAACACTGCTCGGAATCGCTGCCTGTGTTAAACATTTTGCTGCTTATGGTGCAGCAGAAGGGGGACGAGATTATAATACAGTTGATATGAGCGAACGCAAACTGCGTCAAGACTACTTAAGCGGCTATAAAGCGGCTGTCGaGGCTGGATGCAAACTAGTCATGACTTCTTTTAACACGTATGATGGTATTCCCGCTACtGCTAATCAATTTTTGATCAAACAAATTTTAAGAgAAGAntGGCAGTTTGATGgAgTCGTTATTTCgGATTATGCAGCTGTtCAAGAnTTAaTTCCTCATGGGAtTGCTGCGGAtGATCgAgAAgCGGCCAAATTAgCGATCGAAgCAaCAAAtGAcaTCGATATGAAAACCcGATGttaTGCgAAAnAnCttCntCcGCTGCtGgaAagtGGt GH3#3 gene >6831966.seq - ID: GH3/3-T7 on 2010/8/19-5:41:18 automatically edited with PhredPhrap, start with base no.: 20 Internal Params: Windowsize: 20, Goodqual: 19, Badqual: 10, Minseqlength: 50, nbadelimit: 1 cGATCaGTTGCTGCAaCTgGCAGCGGCTTTTTATTCAGATaannnnnaAgAgAAAACAGGTCCGATGGGCGACTTAGGACTGACACAAGAAAACATCAACAACGCGGGAACAACGCTAGGTGTTTCTGGTGCAAAAGAAGCGATCCGCGTCCAAAAAGAGTATATCGCCAATAACCGCTTGAATATCCCGACGATATTGATGGCGGACATCATTCACGGCTTTCGGACGATTTTCCCGATTCCATTAGGATTAGGTAGTTCATGGGATTTGGCAGCAGCGGAGAAAATGGCGGAAGTATCTGCCAAAGAAGCAGCTGTTTCTGGCTTGCATGTGACCTTTTCACCGATGGTGGACTTAGTAAGAGACCCA

355

CGCTGGGGCCGTGTCATGGAATCGACGGGGGAAGATCCTTACTTGAACAGTCGCTTCGCTGAAGCCTTCGTCAAAGGCTATCAAGGGGATGATCTGCGAACGGATTTCAACCGCGTGGCTGCTTGCGTCAAACATTTTGCGGCTTACGGTGCGGCTATCGGTGGTCGCGATTACAACACGGTCAATATGTCAGAACGCCAACTGCGAGAAAGTTATTTGCCAGGCTATAAAGCAGCCCTTGATGCTGGTGCTAAGCTGGTGATGACCTCCTTTAATACGGTAGACGGCATTCCAGCAACGGCCAATCGCTGGCTTTTCCGCGATGTTTTGCGAnAAGAATTCgGGTTTGAAGGcGTTGTGATCTCTGACTGGGCAGCaATCaAAGAAGTGATCGCTCAtGGcGCAnCGganGAtGAAAAACaTGCCGCTGAACTAnnCaTCAAgCTGGGGTCgAtATCnAGAtGAtGACna GH1 gene >6831968.seq - ID: GH1-T7 on 2010/8/19-5:41:14 automatically edited with PhredPhrap, start with base no.: 12 Internal Params: Windowsize: 20, Goodqual: 19, Badqual: 10, Minseqlength: 50, nbadelimit: 1 GttgGgnntCtccaTaTGGTCgacCTGCagGCGGCCGCGnATTCACTAGTGATTCCaGagCTCTTAAATAATCGTTTTGGTTTCACTGACCAACTTGTACCAGTAAGCTGATTCTTTTGGGTAACGCTTCTGGGTTTCAAAATCAACATAAAACAAGCCATACCTTTTGTTATACCCATTTGTCCATGAGAAAAGATCCATAAGCGACCACAAAAAATACCCCTTCACATTTACCCCGGCTGTTATCACCTTGCTTAAGGATTCCAAATAGACTCTCAAATAGTCGATTCTCGGCTGATCCATAATAATGCCATCTTCAAACTGATCTTTATACCCCATCCCGTTCTCAGTGATATAGATTTTGTTGTAATGTGGGTAATCGCTTTTAATTCGCAGCAACAAGTCATATAGACCTTCTGGATAAATCAGCCAGTCCCAAACAGTTCGAGGAATTCCTTCTTTATAAATACGTTCTCCGATTCCCTTCACTTTATAAACAGAGGTCCCTTTTTCACCCGTACCGTTATGATGGATTGCATTTTCTCCATCGTAAGCTTTGACAAAATGGCATTGGTAGTGATTGATCCCTAAGTAGTCATTACGTGTGGATGCTTTTTTCAATTCAACGAAACCTTCTTCAGGAAAATGATAGCTGGCCTGATTTGCTTCACATATCTCATCAAGAGCAGTCAAGGTTTCAGTTGAATAATAGCCTAAATAGGTAGCATCTAATAAGAAACGGATCGACAATGCATCGTCTAAAAAGGCAGCATGCTTATCTTCCGGCGCGTCTGTTGCCGCATATTTCGTCTCTAATGAGTGAACTACACCAATTTCACCCGGAAGTTCGTTTTCTTTGAAATAGTTCACTACACGGGCATGAGCAACCATCATATTATGAAGACAGGCGACGATCCTTGTAAAATCATATTTGATTCCTGGTgGGAAAACACCTAACAAAtATTGATTGGTTGCAACTGGatagaTTTCATTAAAGGTACTCCacaCCTTGACTTCTTTGAATTCATGAAAAcaAAAAATGGCATAGGAAacaAATGCTTCTATTGTTtnnnGATTCAAAAatctcCATGGTCAannAACGTTTAGGGGTatcnAAATGatgca GH3#2 gene >6831971.seq - ID: GH3/2-T7 on 2010/8/19-5:41:14 automatically edited with PhredPhrap, start with base no.: 21 Internal Params: Windowsize: 20, Goodqual: 19, Badqual: 10, Minseqlength: 50, nbadelimit: 1 AnTatTTTGTTTaCTTTAnGAnGGagaTaTaCCATGGGCAGCAGCcnncAtCaTCATcATCACAGCAGCGGCCTGGTGCCGCGCGGCAGCCATATGGTGACAACCAAGCAATGATTGGTTTATGGGCAGTACACGGAAAAACAGAGGATGTCACAACATTAAAAACGGCGCTCCAAAATACAGTGTCGGAAAAATATGTTCATTACGAACCAGGTTGTCCGCTTTTAGAAGATGATTCTATACTTGGAGACTTTGGCTATACTGCCAGCGGCAATTCCTCATCAGCTGCACAGCAAGATCTTTGGCTGAAAGAAGCGTTGAAAGCTGGCACTGAGGCAGACATTATTCTTTTTGCTATGGGGGAACACAGTTTGCAAAGTGGGGAAGCTGGCAGTCGAACGTAAGAGCTCCGTCGACAAGCTTGCGGCCGCACTCGAGCACCACCACCACCACCACTGAGATCCGGCTGCTAACAAAGCCCGAAAGGAAGCTGAGTTGGCTGCTGCCACCGCTGAGCAATAACTAGCATAACCCCTTGGGGCCTCTAAACGGGTCTTGAGGGGTTTTTTGCTGAAAGGAGGAACTATATCCGGATTGGCGAATGGGACGCGCCCTGTAGCGGCGCATTAAGCGCGGCGGGTGTGGTGGTTACGCGCAGCGTGACCGCTACACTTGCCAGCGCCCTAGCGCCCGCTCCTTTCGCTTTCTTCCCTTCCTTTCTCGCCACGTTCGCCGGCTTTCCCCGTCAAGCTCTAAATCGGGGGCTCCCTTTAGGGTTCCGATTTAGTGCTTTACGGCACCTCGACCCCAAAAAaCTTGATTAGGGTGATGGTTCACGTAGTGGGCCATCGCCCTGATAGACGGTTTTTCgcCCTTTGACGTTGGAGTCCACGTTCtTTAATAGTGGACTCTTgttCCaAACTGGAACAACaCTCAACCCTATCTCGGTcTATTCTTTTGATtnataAGGGAtTTTGCcgaTtTcGGCCTaTTGGTTAAAAAaTGagcTGATTtaacaAAAATTTAAcgnnaATTTTAAcAAAanntTAAcgcTTa

356

bgl gene >6831959.seq - ID: bgl-T7 on 2010/8/19-5:41:14 automatically edited with PhredPhrap, start with base no.: 21 Internal Params: Windowsize: 20, Goodqual: 19, Badqual: 10, Minseqlength: 50, nbadelimit: 1 cntggCGGCcgcGGgnaTTCGatTGGtTTGCcaTATGTTTCACACaAACTTAGATCCTTTCCCAGAAAACTTCTTATGGGGGGCAGCTTCGGCGGCCTATCAAATTGAAGGAGCATGGGCAGAAGACGGCAAAGGTCCGTCGATTTGGGACACCTATGCCCAAATTCCTGGCAATACTTTTGAGGAAACCAACGGCAAAGTGGCGATCGATCATTACCATCGATACAAAGAAGATATTGCCTTGATGAAGCAAATGGGCTTGAAAGCCTATCGCTTCAGCGTGGCGTGGTCGCGAATCTTGCCTGATGGCGAAGGCGCGGTCAATGAAGCGGGTGTGGCGTTTTACGAAAAGCTGGTGGATGAATTGCTTCGGCAAGGAGTAGAGCCGATTTTGACGCTGTATCATTGGGACCTTCCCCAGGCTTTGCAAGACAAATACTTAGGGTGGGAAGGTCGAGAAACAGCAGAAGCATTTGAACGGTATTGCCGGATCCTTTTTGAACGCCTAGGAAAGAAAGTCACCTATTGGGTCACCATGAATGAACAAAATGTCTTCACTTCTCTTGGGTACCGTTGGGCGGCACATCCGCCGGGCTTGAAGGACTTAAAACGGATGTATGCAGCCAATCATATCATCAACCTTGCCAATGCTAAGGCGATCAATTTGTTCCATGAGCTGGTTCCTCAGGGCAAGATCGGTCCAAGTTTTGGCTATGGACCGATGTATCCGTTTAGCTGTGACCCAGAAGATGTGCTGGCAGCAGAAAATGGCGAAGCCTTCAACAACGCATGGTTTTTAGATGTCTATTGCAAAGGTGAATACCCGAAATTTGTGTACAAGCAATTAGCCAAAGTTGGCTTgGCTCCTGAAGTCACTCCAGAAGATCAAGCACTATTAAAACAGGCAAAACCTGAtTTCTTAGgAATCAATTACTATCACGgtgGGACGGcCCAGCAAAACAATTTGCAAAAGCAGTCAGCTGAAAagaAagaAtTTTCTAAAGTcGaTTcgtAtTTGATGCAagcGGCAGCTGGtgagtTctcacCCGAagaaACCATgtTTGcgacaGcnnnAAATCCTcacTTGAannnAAcGGATTGGGGCTGGGaAAtngaTCCCGTTGGTTTccnt 6pbg1 gene >6831961.seq - ID: 6pbg1-T7 on 2010/8/19-5:41:18 automatically edited with PhredPhrap, start with base no.: 25 Internal Params: Windowsize: 20, Goodqual: 19, Badqual: 10, Minseqlength: 50, nbadelimit: 1 tggcgGCcgcgGgnaTTCGaTTGGtTTGCCATATGTTTcaCACaAACTTAGATCCTTTCCCAGAAAACTTCTTATGGGGGGCAGCTTCGGCGGCCTATCAAATTGAAGGAGCATGGGCAGAAGACGGCAAAGGTCCGTCGATTTGGGACACCTATGCCCAAATTCCTGGCAATACTTTTGAGGAAACCAACGGCAAAGTGGCGATCGATCATTACCATCGATACAAAGAAGATATTGCCTTGATGAAGCAAATGGGCTTGAAAGCCTATCGCTTCAGCGTGGCGTGGTCGCGAATCTTGCCTGATGGCGAAGGCGCGGTCAATGAAGCGGGTGTGGCGTTTTACGAAAAGCTGGTGGATGAATTGCTTCGGCAAGGAGTAGAGCCGATTTTGACGCTGTATCATTGGGACCTTCCCCAGGCTTTGCAAGACAAATACTTAGGGTGGGAAGGTCGAGAAACAGCAGAAGCATTTGAACGGTATTGCCGGATCCTTTTTGAACGCCTAGGAAAGAAAGTCACCTATTGGGTCACCATGAATGAACAAAATGTCTTCACTTCTCTTGGGTACCGTTGGGCGGCACATCCGCCGGGCTTGAAGGACTTAAAACGGATGTATGCAGCCAATCATATCATCAACCTTGCCAATGCTAAGGCGATCAATTTGTTCCATGAGCTGGTTCCTCAGGGCAAGATCGGTCCAAGTTTTGGCTATGGACCGATGTATCCGTTTAGCTGTGACCCAGAAGATGTGCTGGCAGCAGAAAATGGCGAAGCCCTCAACAACGCATGGTTTTTAGATGTCTATTGCAAAGGTGAATACCCGAAATTTGTGTACAAGCAATTAGCCAAAGTTGGCTTGGCTCCTGAAGTCACTCCAGAAGATCAAGCACTATTaAAACAGGCAAAACCTGATTTCTTAGGAATCAATTACTATCACggnngGGACGGcCCAGCAAAACAATTTGCAAAAGCAGTCAGCTGAAAagaAAgAATTTTCTAAAGTCgaTCCGTATTTGATGCAAGCGGCAGCTGGTgAGTTCtcacCCGAAgaAACCATgTTTGcgacaGCanaAAATCCTCacTTGAAAanaacgganntGGgnntgGGAnatcgnTcCCGTTGgtttcc 6pbg2 gene >6831963.seq - ID: 6pbg2-T7 on 2010/8/19-5:41:18 automatically edited with PhredPhrap, start with base no.: 22 Internal Params: Windowsize: 20, Goodqual: 19, Badqual: 10, Minseqlength: 50, nbadelimit: 1 cntggCGGccgcGGgnaTTCGaTTGGTTTGCCATATGTTTCACACaAACttAGATCCTTTCCCAGAAAACTTCTTATGGGGGGCAGCTTCGGCGGCCTATCAAATTGAAGGAGCATGGGCAGAAGACGGCAAAGGTCCGTCGATTTGG

357

GACACCTATGCCCAAATTCCTGGCAATACTTTTGAGGAAACCAACGGCAAAGTGGCGATCGATCATTACCATCGATACAAAGAAGATATTGCCTTGATGAAGCAAATGGGCTTGAAAGCCTATCGCTTCAGCGTGGCGTGGTCGCGAATCTTGCCTGATGGCGAAGGCGCGGTCAATGAAGCGGGTGTGGCGTTTTACGAAAAGCTGGTGGATGAATTGCTTCGGCAAGGAGTAGAGCCGATTTTGACGCTGTATCATTGGGACCTTCCCCAGGCTTTGCAAGACAAATACTTAGGGTGGGAAGGTCGAGAAACAGCAGAAGCATTTGAACGGTATTGCCGGATCCTTTTTGAACGCCTAGGAAAGAAAGTCACCTATTGGGTCACCATGAATGAACAAAATGTCTTCACTTCTCTTGGGTACCGTTGGGCGGCACATCCGCCGGGCTTGAAGGACTTAAAACGGATGTATGCAGCCAATCATATCATCAACCTTGCCAATGCTAAGGCGATCAATTTGTTCCATGAGCTGGTTCCTCAGGGCAAGATCGGTCCAAGTTTTGGCTATGGACCGATGTATCCGTTTAGCTGTGACCCAGAAGATGTGCTGGCAGCAGAAAATGGCGAAGCCTTCAACAACGCATGGTTTTTAGATGTCTATTGCAAAGGTGAATACCCGAAATTTGTGTACAAGCAATTAGCCAAAGTTGGCTTGGCTCCTGAAGTCACTCCAGAAGATCAAGCACTATTAAAACAGGCAAAACCTGATTTCTTAGGAATCAATTACTATCACGGTgGGACGGcCCAGCAAAACAATTTGCAAAAGCAGTCAGCTGAaAagaaagaAtTTTCTAAAGTCGaTCCGTATTTGATGCAancGGCancTGGTGagtTctcacCCGAagaAACCATgTTTGcgaCAGCAgaAAATCcTcacTTGAAanaaacGGattgggGCTGGGAnAtcgnTcCCGTTGGTTTcc pBgl gene >6831962.seq - ID: pBgl-T7 on 2010/8/19-5:41:14 automatically edited with PhredPhrap, start with base no.: 22 Internal Params: Windowsize: 20, Goodqual: 19, Badqual: 10, Minseqlength: 50, nbadelimit: 1 gaattcCCTCTagAnTatTTTGTTTaCTTTAnGAnGGagaTaTaCCATGGGCAGCAGCccncntnntcatCaTCACAGCAGCGGCCTGGTGCCGCGCGGCAGCCATATGTGCCGAGAAAAAAAGTCCTGTGTGGCTTTCAAAAAGTATGGGTGCCTAATCAAGGGCAAGTAGCTTTTCAACTCGTGATCGATGAAGCATCATTGCAGCAGCTGGCAATTTCTTTGAAAGATACTGCATCGTTTTGTCTGGAAGTCGAAACAGCAGGTCAGCAGTATCGATTTTTATTTCAACGATCCAACCCTGACCGTACGTGGCAGGTAACTCAGAAAGGAGCAAAAGAATGAGAGCTCCGTCGACAAGCTTGCGGCCGCACTCGAGCACCACCACCACCACCACTGAGATCCGGCTGCTAACAAAGCCCGAAAGGAAGCTGAGTTGGCTGCTGCCACCGCTGAGCAATAACTAGCATAACCCCTTGGGGCCTCTAAACGGGTCTTGAGGGGTTTTTTGCTGAAAGGAGGAACTATATCCGGATTGGCGAATGGGACGCGCCCTGTAGCGGCGCATTAAGCGCGGCGGGTGTGGTGGTTACGCGCAGCGTGACCGCTACACTTGCCAGCGCCCTAGCGCCCGCTCCTTTCGCTTTCTTCCCTTCCTTTCTCGCCACGTTCGCCGGCTTTCCCCGTCAAGCTCTAAATCGGGGGCTCCCTTTAGGGTTCCGATTTAGTGCTTTACGGCACCTCGACCCCAAAAAaCTTGATTAGGGTGATGGTTCACGTAGTGGGCCATCGCCCTGATAGACGGTTTTTCGCCCTTTGACGTTGGAGTCCACGTTCTTTAATAGTGGACTCTTGTTCCAAACTGGAACAACACTCAACCCTATCTCGGTCTATTCtTTTGATTTataAgGGATTTTGCCGatTTCGGCCTATTGGTTaAAAAATGagCTGATTTAACAAAAATTTAACGCgaATTTTAACAAAatatTAACGCTTAcAATTTAggngg

358

Appendix III

Representative HPLC chromatograms showing no DS-GSL production upon 8 h incubation of intact GSLs with crude extracts from BL21(DE3) on the DEAE-Sephadex column at  30˚C  under  aerobic conditions. Peaks between 20 and 30 min possibly correspond to dirt residues.

Appendix IV

Representative GC-MS chromatograms showing no NIT production in the negative controls containing DS-GSL alone, GSL alone, the GH3 enzyme alone, SUL2 enzyme alone or the two enzymes alone incubated in NB broth or in the buffer for 24 h at 30˚C   under   anaerobic  conditions.

min5 10 15 20 25 30

mAU

45

50

55

60

65

ADC1 A, ADC1 CHANNEL A (F:\PHD\HPLC RAW DATA\SULFATASE\VIN160212SUL PHEN BEN\VIN16021201.D)

2.794

21.26

5 21

.647

22.53

5

24.60

9

26.02

1

5 . 0 0 1 0 . 0 0 1 5 . 0 0 2 0 . 0 0 2 5 . 0 0 3 0 . 0 0 3 5 . 0 0 4 0 . 0 0

1 5 0 0 0

2 0 0 0 0

2 5 0 0 0

3 0 0 0 0

3 5 0 0 0

4 0 0 0 0

4 5 0 0 0

5 0 0 0 0

5 5 0 0 0

6 0 0 0 0

6 5 0 0 0

7 0 0 0 0

7 5 0 0 0

8 0 0 0 0

8 5 0 0 0

9 0 0 0 0

9 5 0 0 0

1 0 0 0 0 0

1 0 5 0 0 0

1 1 0 0 0 0

1 1 5 0 0 0

T i m e - - >

A b u n d a n c e

T I C : V I N 1 7 0 2 1 3 1 5 . D