tio photocatalysis damages li pids and proteins in...
TRANSCRIPT
1
TiO2 photocatalysis damages lipids and proteins in Escherichia 1
coli 2
3
Gaëlle Carré1,2*, Erwann Hamon3, Saïd Ennahar3, Maxime Estner1, Marie-Claire Lett4, 4
Peter Horvatovich5, Jean- Pierre Gies1, Valérie Keller2, Nicolas Keller2* and Philippe 5
Andre1 6
7
1: Laboratoire de Biophotonique et Pharmacologie, UMR 7213 CNRS and Strasbourg 8
University, 74 route du Rhin, 67400 Illkirch, France 9
2: Institut de Chimie et Procédés pour l’Energie, l’Environnement et la Santé (ICPEES), 10
UMR 7515 CNRS and Strasbourg University, 25 rue Becquerel, 67087 Strasbourg Cedex 2, 11
France 12
3: Laboratoire de Chimie Analytique des Molécules BioActives, UMR 7178 CNRS and 13
Strasbourg University, 74 route du Rhin, 67400 Illkirch, France. E. Hamon, new address : 14
Aérial, Parc d’Innovation, Rue Laurent Fries, 67400 Illkirch, France. 15
4: Laboratoire de Génétique Moléculaire, Génomique, Microbiologie, UMR 7156 CNRS and 16
Strasbourg University, 28 rue Goethe, 67083 Strasbourg Cedex, France 17
5: Department of Analytical Biochemistry, Center of Pharmacy, Groningen University, 9700 18
AD Groningen, The Netherlands 19
20
* Corresponding authors E-mail: [email protected] ; [email protected] 21
22
23
24
AEM Accepts, published online ahead of print on 14 February 2014Appl. Environ. Microbiol. doi:10.1128/AEM.03995-13Copyright © 2014, American Society for Microbiology. All Rights Reserved.
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
2
ABSTRACT 25
This study investigates the mechanisms of UV-A photocatalysis on titanium dioxide (TiO2) 26
applied to the degradation of Escherichia coli and their effects on two key cellular 27
components: lipids and proteins. The impact of TiO2 photocatalysis on E.coli survival was 28
monitored by counting on agar plate as well as by assessing lipid peroxidation and performing 29
proteomic analysis. We observed through malondialdehyde quantification that lipid 30
peroxidation occurred during the photocatalytic process, and the addition of superoxide 31
dismutase that acts as a scavenger of the superoxide anion radical (O2°-), inhibited by half this 32
effect, showing us that O2°- radicals participate to the photocatalytic antimicrobial effect. 33
Qualitative analysis of two-dimensional electrophoresis allowed to select proteins for which 34
spot modifications were observed during the applied treatments. Two-dimensional 35
electrophoresis highlighted that, among the selected protein spots, 7 and 19 spots disappeared 36
already in the dark in the presence of 0.1 g/L and 0.4 g/L TiO2, respectively, which is 37
accounted for the cytotoxic effect of TiO2. Exposure to 30 min UV-A radiation in the 38
presence of 0.1 g/L and 0.4 g/L TiO2 increased the number of missing spots to 14 and 22, 39
respectively. The proteins affected by the photocatalytic oxydation were strongly 40
heterogeneous in terms of location and functional category. We identified several porins, 41
proteins implied in stress response, in transport and in bacteria metabolism. This study reveals 42
the simultaneously effect of O2°- on lipid peroxidation and on the proteome during 43
photocatalytic treatment, and therefore participates to a better understanding of molecular 44
mechanisms on antibacterial photocatalytic treatment. 45
46
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
3
INTRODUCTION 47
Several disinfection strategies exist, some of them involving for example the use of silver (1-48
3) or copper metal ions (4). However, the development of new disinfection approaches is 49
required, due to the rapid adaptation of bacterium and development of resistant strains to these 50
metals (5). Among these new disinfection approaches, photocatalysis is a promising technique 51
which belongs to Advanced Oxidation Processes (AOP), characterized by the production of 52
Reactive Oxygen Species (ROS) (6). In 1985, Matsunaga et al. were the first to report on the 53
killing of microbial cells in water by near-ultraviolet (UV) light irradiated platinum-loaded 54
titanium dioxide (TiO2) semiconductor particles (7). This pioneering work gave rise to many 55
researches in the field of disinfection by oxidative photocatalysis (8-10), with found 56
nowadays applications in many disinfection fields (11, 12). 57
Among the various semiconductor photocatalysts studied, the wide band-gap TiO2 in 58
anatase crystalline form is the most attractive one. This material has high photocatalytic 59
efficiency, due to its high quantum yield, has high stability towards photocorrosion and 60
chemicals, is insoluble in water, and has a low toxicity and low costs. The band gap energy of 61
3.2 eV for TiO2 requires photoexcitation wavelengths less than ca. 385 nm, corresponding to 62
an irradiation with near UV light (13). 63
The activation of the TiO2 semiconductor particle with the adequate UV-A light, generates 64
electrons and holes in the conduction and valence bands, respectively. The photogenerated 65
charges take part in reduction and oxidation reactions at the particle surface (12, 14). In 66
particular, holes and electrons are respectively reacting with adsorbed water and dioxygen 67
molecules to form ROS such as the °OH hydroxyl radical and the O2°- superoxide radical 68
anion, respectively. Singlet oxygen (1O2) or hydrogen peroxide (H2O2) can also be formed 69
(15). The analogy between chemical and biological targets results from the organic nature of 70
the microorganisms constituents, that can react with the active surface species issued from the 71
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
4
TiO2 photoactivation. The resulting reactions are similar to the oxidation reactions taking 72
place at the surface of irradiated TiO2 photocatalysts with organic molecules, e.g. during 73
potabilization or depollution oxidative processes in water and air treatments (16). 74
These ROS are thus responsible for the oxidation of many organic constituents of the 75
microorganisms (17) such as lipid peroxidation (18), protein alteration (19) or DNA damage 76
(20). The direct contact between the targeted microorganisms and the TiO2 particles is 77
reported to be one of the key parameters of the photocatalytic disinfection (17, 21). Notably, 78
several transmission electron microscopy analyses revealed that the binding of E. coli to TiO2 79
particles induces cell disruption and cell debris (22-24). However, the precise molecular 80
mechanism remain still unclear and is a matter of debate. 81
E. coli, a Gram-negative bacterium which complete genome has been sequenced (25), 82
represents a suitable bacterial model to study the antimicrobial effects of photocatalysis at a 83
molecular level. 84
This paper investigates the antibacterial effect of TiO2 photocatalysis on E. coli and is 85
focused on the identification of the photocatalytic damages induced on both lipids and 86
proteins, that are key cellular components of bacteria. Both lipid peroxidation and E.coli 87
proteome modifications have been investigated in this study. The implication of the O2°- 88
superoxide radical in the lipid peroxidation has been demonstrated and the identification of 89
the proteins of E. coli ATCC 8739 strain modified by the photocatalytic treatment, performed 90
through two-dimensional electrophoresis, may offer insights into the mechanism behind TiO2 91
antibacterial effects. 92
93
MATERIALS AND METHODS 94
Bacterial strains and growth media. Before each experiment, one loop full of Escherichia 95
coli ATCC 8739 strain was seeded on a slant of tryptic soy agar (TSA) (BioRad), and grown 96
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
5
aerobically at 37°C for 24 h. Bacterial inoculum was monitored by setting the culture optical 97
density at an absorbance wavelength of 620 nm (OD620), at 0.156, corresponding to 108 98
CFU/mL. 99
Photocatalytic material. Experiments were performed with commercial TiO2 Aeroxide 100
P25 (Evonik, Frankfurt, Germany, 20 % rutile – 80 % anatase crystalline form, with a specific 101
surface of 50 m2/g). 102
Photocatalytic experimental set-up. The photocatalytic tests have been assessed using an 103
experimental set-up composed (i) either of a 96 well plate – each well acting as a single liquid 104
phase photocatalytic reactor – for determining cell viability on agar plate, (ii) or of Petri 105
dishes for both lipid and proteomic analyses, since they require large amounts of materials. In 106
each case, different amount of TiO2 were added in a range from 0.1 to 0.8 g/L. The 107
photocatalytic device is composed of four equidistant commercial BlackLight Blue lamps 108
(BLB, 40W, Actinic, Philips, Eindhoven, the Netherlands) with a spectral peak centered on 109
365 nm, and providing an irradiance of 30 W/m² to the wells or the Petri dishes. After 110
addition of the bacteria and TiO2 to the photocatalytic reactor, UV-A irradiation starts 111
immediately for two 96 well plates with an exposure time of 30 min or 60 min; one other 96 112
well plate stayed in the dark for the same durations. All the 96 well plates were placed on a 113
shaker (Heidolph Duomax 1030) operating at 40 rpm to ensure adequate mixing and contact 114
between TiO2 particles and bacteria. The pH of the photocatalytic suspensions at the end of 115
the experiment was 7.8. Each experimental condition with respect of test duration and of TiO2 116
concentration corresponded to a single well and was performed in triplicate analysis. 117
Cell viability measurement. E.coli was grown at 37°C in Mueller Hinton (MH) broth 118
(BioRad) until the culture reached an OD620 of 0.156 corresponding to 1.108 CFU/mL. After 119
centrifugation at 9000 g for 10 min, the supernatant was removed and the bacterial pellet 120
was resuspended in sterile physiological water (NaCl, 9 g/L) to obtain 108 CFU/mL. This 121
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
6
population was diluted to obtain an inoculum of 1.106 CFU/mL. Each photocatalytic well 122
reactor (96 well-plate) was filled with 90 µL of 106 CFU/mL bacterial suspensions to which 123
10 µL of a TiO2 suspension in physiological water was added for performing the tests with 124
TiO2 concentrations of 0.1 g/L; 0.2 g/L; 0.4 g/L and 0.8 g/L. The tests were carried out with 125
or without TiO2. After the different treatments, dilutions were made in physiologic water and 126
100 µL of the resulting solutions were spread on TSA. Each agar plate was incubated at 37°C 127
for 24 h and the number of bacterium was counted after incubation. 128
The logarithm of reduction, i.e. log10(C/C0), where C0 is the concentration of control 129
bacteria without TiO2 in the dark expressed CFU/mL, and C is the concentration of bacteria in 130
life expressed in CFU/mL. 131
Lipid oxidation. The thiobarbituric acid (TBA) (OxiSelect TBARS Assay kit, Euromedex, 132
Strasbourg, France) assay was used and slightly modified by increasing the bacterial 133
population to assess lipid peroxidation state. TBA reacts at 95°C with malondialdehyde 134
(MDA). MDA is a lipid oxidation product, and provide a colored MDA-TBA adduct. The 135
concentration of this adduct can be derived by measuring absorbance at 532 nm by 136
spectrophotometry. 137
E.coli was grown at 37°C in 1500 mL Mueller Hinton (MH) broth (BioRad) until the 138
culture reached a OD620 of 0.156 (~1.108 CFU/mL). After centrifugation at 9000 g for 10 139
min, the supernatant was removed and the bacterial pellet was resuspended in 150 mL of 140
physiological water to obtain a final population of 1.109 CFU/mL. 50 mL of this suspension, 141
containing 5.1010 CFU was divided into 5 Petri plates of 10 mL and further used as 142
photocatalytic reactor. In fact, this lipid quantification kit is designed for eukaryotic cells, and 143
prokaryotic cells required higher amounts of bacteriologic materials. The Petri dishes 144
containing 0.4 g/L of TiO2 were placed on a shaker plate operating at 40 rpm (Heidolph 145
Duomax 1030) during the 60 min of UV-A irradiation. Control samples were treated the same 146
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
7
way in dark with absence of TiO2 catalyst. To demonstrate the effect of O2°- in the lipids 147
oxidation, SOD 1000 Units (Sigma-Aldrich, Saint Quentin Fallavier, France), was added to 148
the reaction media before the photocatalytic treatments. To quantify the lipids oxidation, the 149
50 mL content of each Petri dish was collected and subsequently centrifuged at 9000 g for 150
10 min. This was followed by resuspension of bacterial pellets in 500 µL phosphate buffered 151
in saline. Then 5 µL butylated hydroxyl toluene 100 X antioxidant was added to 500 µL of 152
sample to prevent any further lipid oxidation. Finally, 250 µL of TBA reagent and 100 µL of 153
SDS lysis buffer were added to 100 µL of sample as well as to MDA standards as described in 154
the TBARS Assay kit Protocol. Reaction was performed in duplicate. Then samples and 155
standard were placed at 95°C for 1 h, before the final centrifugation at 1000 g for 15 min 156
took place and the resulting supernatant was transferred to a 96-well plate for recording the 157
absorbance at 532 nm by spectrophotometry. The quantity of peroxidized lipid in mg was 158
calculated with help of the previously determined MDA standard curve. 159
Comparative proteomic analysis. Like for the lipid oxidation, proteomic analysis 160
requires a large amount of materials. Therefore, the photocatalytic tests were performed using 161
Petri plates as reactor under vigorous stirring. After a 18 h growing of E. coli at 37°C on slant 162
of TSA, bacterial suspensions in physiological water were adjusted to a OD620 of 0.156 163
(~1.108 CFU/mL). Twenty mL of the bacterial suspension of 1.108 CFU/mL were divided into 164
2 Petri dishes used as reactor. The proteomic analysis was performed on fresh cells as well as 165
on cells subjected to 30 min irradiation with UV-A light using 0.1 g/L and 0.4 g/L TiO2. 166
During the experiments, plates were shaken with a shaker plate at 40 rpm (Heidolph Duomax 167
1030). Comparisons were performed with cells treated with TiO2 catalyst and kept in the dark 168
(no irradiation) as well as with cells subjected to UV-A light exposure in the absence of any 169
TiO2 catalyst. Control experiment was carried without TiO2 in the dark. For analysis, cells 170
obtained after sampling were first centrifuged (3000 g, 5 min at 4°C) and their whole 171
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
8
proteomes were subsequently extracted by cryogrinding and purified using trizol reagent 172
(Euromedex, France), as previously reported in E. Hamon et al. and separated by two-173
dimensional electrophoresis (2-DE), with a pI range of 4.0-7.0 and a mass range of 10-250 174
kDa (26). In these conditions, the estimated coverage of the theoretical proteome of E. coli 175
ATCC 8739 was 62%, as inferred from an in silico analysis of E. coli protein sequence data 176
obtained from NCBI (data not shown). Image analysis of the 2-DE gels was performed using 177
the PD Quest 8.0.1 software (Bio-Rad). For each condition, the same experiments were 178
triplicated and only spots that disappeared on the three gels were selected for inter-condition 179
comparison. Spot intensities were normalized to the sum of intensities of all valid spots in one 180
gel. Protein expression differences were identified between different operating conditions. 181
Only proteins present in samples of one condition and absent in one other were considered in 182
this analysis. 183
Spots of interest were subjected to tryptic in-gel digestion and analyzed by chip-liquid 184
chromatography-quadrupole time-of flight (chip-LC-QTOF) using an Agilent G6510A QTOF 185
mass spectrometer equipped with an Agilent 1200 Nano LC system and an Agilent HPLC 186
Chip Cube G4240A (Agilent Technologies, Santa Clara, CA, USA), as described previously 187
in E. Hamon et al. (26). After data acquisition, files were uploaded to the in-house installed 188
version of Phenyx (Geneva Bioinformatics, Geneva, Switzerland) for searching the NCBInr 189
(r. 20110608) database with the following criteria: taxonomy: E. coli; scoring model: ESI-190
QTOF; parent charge: +2, +3 (trust = medium); single round; methionine oxidation, cysteine 191
carboxyamidomethylation (cysteine treated with iodoacetamide), and phosphorylation as 192
partial modifications; trypsin as digestion enzyme; allowance of two missed cleavages; 193
cleavage mode: normal; parent ion tolerance: 0.6 Da; peptide thresholds: length ≥6, score 194
threshold ≥5.0, identification significance p-value ≤ 1.0⋅10-4, accession number score 195
threshold 6.0, coverage threshold ≥0.2, identified ion series: b; b++;y; y++; allowance of 196
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
9
conflict resolution. A publicly available MS/MS search algorithm (Open Mass Spectrometry 197
Search Algorithm, OMSSA, (27)) was used with the same search criteria as described above 198
to confirm protein identities and limit the risk of false positives. On the basis of consensus 199
scoring, only proteins recognized by both database search algorithms at a false positive rate of 200
5% were considered to be correctly identified (28). 201
Statistical analysis. Statistical analysis was performed using Prism 5 (GraphPad Software) 202
and was based on Student’s t-test. 203
204
RESULTS 205
Evidence of the bacteria inactivation by UV-A irradiated TiO2. Figure 1 shows the 206
influence of the different applied treatments on the viability of E. coli cell, determined by 207
bacterial counting on agar plate and expressed in terms of the logarithmic reduction. First, the 208
sole application of UV-A light in the absence of TiO2 did not caused any cell viability 209
reduction. Secondly, independently of the TiO2 concentration within the 0.1-0.8 g/L range, no 210
TiO2 cytotoxic effect was observed in the dark, whereas a reduction of cell viability was only 211
observed in sample irradiated with UV-A light and in the presence of TiO2. This reduction 212
was directly UV-A irradiation duration dependent, since e.g. the reduction increased from 1 to 213
3 log for a TiO2 concentration of 0.2 g/L when the duration of UV-A exposure increased from 214
30 min to 60 min. Similarly, a stronger diminution of cellular viability was observed at higher 215
TiO2 concentration up to 0.4 g/L, for which two log of reduction was recorded for an 216
irradiation of 30 min. However, the antibacterial effect was not increased at higher TiO2 217
concentration of 0.8 g/L independently from the duration of UV-A exposure. This is in 218
agreement with the usual activity pattern observed in photocatalysis applied to the oxidative 219
degradation of organic molecules, for which the reaction rate is reported to be proportional to 220
the mass of catalyst before achieving a plateau. This plateau corresponds to the maximum 221
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
10
amount of TiO2 in which all the particles are completely illuminated (29, 30). For larger TiO2 222
concentrations, a screening effect of excess particles occurs, so that part of the photosensitive 223
surface is masked. Based on our method counting the bacteria grown on agar plate, the 224
highest cellular viability reduction recorded in the studied E.coli strain was around three log 225
for an irradiation of 60 min in the presence of TiO2 (0.4 g/L). 226
227
Impact of TiO2 photocatalysis and effect of ROS in the lipid peroxidation. As shown 228
in Fig. 2, neither TiO2 at a concentration of 0.4 g/L in the dark or UV-A light irradiation for 229
60 min without TiO2 increased significantly the lipid peroxidation. By contrast, the samples 230
subjected to UV-A irradiation for 60 min with presence of TiO2 catalyst at 0.4 g/L produced 231
5.7 times more lipid peroxides than the bacterial population exposed only to UV-A irradiation 232
for 60 min. The addition of SOD (1000 U) – known to act as a ROS scavenger selective 233
towards the O2°- superoxide radical anion – partially decreased the lipid peroxidation ratio 234
from 5.7 to 3.0 corresponding to 48 % lipid peroxidation rate. Therefore the SOD inhibition 235
of the O2°-produced by irradiated TiO2 induced a half diminution of the lipid peroxidation. 236
This is in agreement with our previous works, which showed that scavenging the O2°- 237
superoxide radical by adding the SOD enzyme restored completely the cell viability (31). 238
239
Impact of TiO2 photocatalysis on E. coli whole-cell proteome. To provide a better 240
understanding of the antibacterial mechanism of TiO2 photocatalysis, the proteome of E. coli 241
ATCC 8739 has been investigated by 2-DE under different the conditions. Bacterial 242
suspensions were maintained 30 min in the dark or under UV-A radiation, without TiO2 or 243
with 0.1 or 0.4 g/L of TiO2. Duration under UV-A and TiO2 concentration were chosen to 244
obtain a sublethal stress according to our previous works (31). Indeed, we have previously 245
shown by cytometry that no significant modification on membrane integrity was observed for 246
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
11
those TiO2 concentrations in the dark or under UV-A light. In contrast cell counts on agar 247
plate showed that 30 min of UV-A irradiation with presence of 0.4 g/L TiO2 induced a 2 log 248
reduction of living bacterium, while the membrane integrity was not affected significantly. 249
This suggests a heterogeneous bacterial population, composed of live cells and ViaBle but 250
Non Culturable (VBNC) cells, where the VBNC indicates a sublethal stress (32). However, 251
the conserved membrane integrity indicates that both kinds of cells still conserved their whole 252
protein machinery, allowing the comparative proteomic analysis. 253
In order to identify the putative protein targets of TiO2 and UV-A irradiation mediated 254
oxidative stress, proteomic patterns of bacteria with different treatment (Fig. 3.B-F) and 255
control (sample without TiO2 in the dark, Fig. 3.A) were recorded. Proteins affected by 256
treatment were found by identifying spots that were observed exclusively in samples of one 257
condition, but were absent in the other condition. These discriminatory spots were labeled 258
with white (absent) and black (present) numbers in the image of 2-DE gels corresponding to 259
treatment conditions. The proteins underlying the detected discriminatory gel spots were 260
identified by database searched using Phenyx and OMSSA tools. The different treatments 261
resulted in 22 spots, which disappeared in the treated samples compared to control. Table 1 262
lists 21 different proteins corresponding to 22 spots, as one protein [DNA 263
starvation/stationary phase protection protein (Dps)] was identified in two spots (spots 12 and 264
13), suggesting the presence of protein isoforms. 265
First, Fig. 3.B shows that UV-A irradiation alone lead to the disappearance of a single spot 266
[unnamed outer membrane protein (FadL), spot 1] compared to the control (Fig. 3.A), which 267
protein is implied in lipid transport and metabolism. 268
The amount of protein degradation was dependent on TiO2 concentration in sample 269
containing TiO2 catalyst, but which were not exposed to UV-A irradiation. Indeed, Fig. 3.C 270
shows that 7 spots disappeared at TiO2 concentration of 0.1 g/L [FadL, spot 1; outer 271
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
12
membrane protein OmpW (OmpW), spot 5; hypothetical protein YgiW (YgiW), spot 7; H+ 272
ATPase F1 alpha subunit (AtpA), spot 10; indigoidine synthase A like protein (Ind-A), spot 273
16; dnaK suppressor (DnaK), spot 18; small ribosomal protein S6 (Ribosomal_S6), spot 19], 274
while 19 spots disappeared at a TiO2 concentration of 0.4 g/L (Fig. 3.E). This corresponded to 275
12 additional spots and 11 additional proteins compared to sample treated with TiO2 at 276
concentration of 0.1 g/L: outer membrane protein A (OmpA), spot 2; outer membrane protein 277
C (OmpC), spot 3; outer membrane protein F (OmpF), spot 4; maltoporin (LamB), spot 6; an 278
hypothetical protein, YggE, spot 8; a dipeptide binding protein (DppA), spot 9; an enoyl ACP 279
reductase, (FabI), spot 11; Dps, spots 12 and 13; bacterioferritin (Bfr), spot 14; two 280
component response regulator (ArcA), spot 15 and a trigger factor (Tig), spot 17. These 281
results indicate that TiO2 alone induce a cytotoxic effect in the dark in our operating 282
conditions. 283
The proteomic profile recorded after the UV-A photocatalytic treatment for 30 min at 0.1 284
and 0.4 g/L TiO2 concentration are shown in Fig. 3.D and Fig. 3.F, respectively. 14 spots were 285
missing in total at a TiO2 concentration of 0.1 g/L (Fig. 3.D), among which 7 proteins 286
(OmpA, spot 2; OmpC, spot 3; OmpF, spot 4; LamB, spot 6; YggE, spot 8; Bfr, spot 14 and 287
ArcA, spot 15) are additional compared to the experiment performed in the dark in the 288
presence of 0.1 g/L TiO2 (Fig. 3.C). These proteins have already been identified in sample 289
solely treated with 0.4 g/L TiO2 (Fig. 3.E). In contrast, the gel obtained from sample 290
irradiated during 30 min in the presence 0.4 g/L of TiO2 (Fig. 3.F) showed only 3 additional 291
disappearing proteins (putative nucleotide-binding protein (YajQ), spot 20; heat shock protein 292
(GrpE), spot 21 and Hslv-Hslu protein (protease-HslV), spot 22) compared to sample treated 293
in the dark at TiO2 concentration of 0.4 g/L (Fig. 3.E). This result corresponds in total to 22 294
missing proteins. 295
296
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
13
DISCUSSION 297
It is well known that the UV-A irradiation of TiO2 induced the production of ROS (33, 34). 298
The generation of ROS is an unavoidable aspect of life under aerobic conditions, but the 299
disruption of the balance between both generation and elimination of ROS is reported to 300
induce sever damages to important cell components such as lipids, proteins or nucleic acid 301
(35, 36). Therefore, we have studied the influence of the TiO2 photocatalysis on lipids which 302
are one of the main compounds of the bacteria membrane. We have performed this study in 303
conjunction with the implication of ROS, by recording the amount of MDA – known to be the 304
end product and main biomarker of lipid peroxidation (18, 37) – produced during the 305
membrane alterations. The lipid peroxidation extent induced by TiO2 photocatalysis was in 306
agreement with results shown by Maness et al. (18) and in the review of Dalrymple et al. 307
(13). It was noteworthy that addition of SOD decreases by half the amount of lipid 308
peroxidation products, confirming that O2°- is implied in the lipid peroxidation process. 309
However we consider that others ROS such as singlet oxygen 1O2 can be generated from 310
oxidation of O2°-. This further oxidation is thermodynamically favored as E° (O2
°-/1O2)=0.34 311
V versus Normal Hydrogen Electrode (NHE) (38) and the formation of 1O2 during 312
photocatalysis on TiO2 has been previously reported (39, 40). Consequently, we cannot 313
exclude a potential involvement of 1O2 in lipid peroxidation, knowing that 1O2 molecule is a 314
high energetic oxygen molecule, which is known to lead to the initiation of lipid peroxidation 315
reactions. This supports the direct implication of O2°- radicals as important active ROS in the 316
E. coli cell death, i.e. in the photocatalytic antimicrobial effect, as previously shown in G. 317
Carré et al. (31). 318
The contact between E.coli bacteria cells and TiO2, with or without UV-A irradiation, 319
induced the disappearance of many spots. As shown in Table 1, the targeted proteins are very 320
heterogeneous in cellular localization and belong to different function families. However it is 321
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
14
clear that most of the involved proteins belong to four main functional groups, with porins 322
(4/21), proteins involved in oxidative stress response (5/21), chaperone proteins (4/21) and to 323
a lesser extent, proteins involved in the transport and the metabolism of different molecules 324
such as inorganic ion (2/21), lipids (2/21), carbohydrates (1/21), and amino-acid (1/21). To a 325
lesser extent, we identified also one protein involved in energy production and conversion, as 326
well as one involved in translation regulation. 327
The disappearance of a spot during the applied treatment in comparison with the control 328
gel may reflect a denaturation, a cleavage, an oxidation or a change in location of the spot on 329
the gel. However, we must consider that different functional protein isoforms may be present 330
in the proteome, such as Omp proteins located at the outer membrane. In the case of OmpA 331
(spot 2), many isoforms of this protein are described in the literature, with 8 already reported 332
in the sole publication of Molloy et al. (41). Thus, the alteration of a spot does not necessarily 333
mean that the protein is no longer functional. Even if no direct information on the proteomic 334
functionality can be obtained, this proteomic analysis identifies potential proteins that are 335
affected during the cytotoxic treatment in the dark with TiO2 and by a photocatalytic 336
treatment in the presence of UV-A light irradiated TiO2. 337
Taking into account the criteria relative to the presence or absence of a spot in comparison 338
with the control gel, the modification of a selected spot should be correlated to the appearance 339
of another spots in the gel corresponding to the same protein. However, presence/absence 340
comparative analysis did not reveal the appearance of any new spots in the gels. Therefore, 341
we can assume that the spots that have been altered, have led to the formation of new 342
undetectable spots. This can be explained by three non exclusive hypothesis: (i) the cleavage 343
of the protein can result in the formation of too small spots to be observed in the gel (smaller 344
than 10 kDa); (ii) the presence of spots corresponding to proteins with a pH out of the 345
observation range of the gel (i.e. outside the 4 to 7 pH range); (iii) the superimposition of new 346
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
15
spots on the gels with initially present spots. Disappearance of spots is most likely due to 347
oxidative degradation of proteins and cannot be attributed to regulation of protein expression 348
as answer to oxidative stress. Indeed, oxidation reactions and protease activity to degrade 349
proteins are fast (42), and while the applied treatments may be long enough (30 min) to cause 350
transcriptional events of the genes encoding the identified proteins, but this time is too short 351
for an efficient protein translation (43). In fact, we precisely chose these conditions not to 352
observe genomic based protein regulation phenomena. 353
We showed that these alterations were not specific to any functional class, although some 354
proteins such as outer membrane proteins were the main target during the treatment with TiO2 355
in the dark and with photocatalytic treatment with TiO2 in presence of UV-A irradiation, 356
probably due to their direct contact with TiO2 particles. In particular, it is reported that about 357
70% of the surface of the membrane of Gram-negative bacteria is composed of porins. Their 358
role is to allow the passage of nutrients and in particular the permeation of hydrophilic 359
molecules (44). Some of them, OmpA and OmpC especially, have been already identified as 360
the main targets when E.coli cells are exposed to external oxidative stress (45). Recently, 361
Krishnan and Prasadarao have shown that OmpA which is a multifunctional major outer 362
membrane protein of E.coli and Enterobacteriaceae interacts with specific receptors 363
initiating pathogenesis of some Gram negative bacterial infections (46). We hypothesized that 364
the degradation of these mains OMPs by both the cytotoxic effect of TiO2 or the ROS 365
generated by the photocatalytic reaction on TiO2 disrupts the communication between bacteria 366
and the environment. 367
Five protein spots corresponding to oxidative stress proteins were altered as well by both 368
cytotoxic and photocalytic effect of TiO2. These include two proteins located in the periplasm 369
(Ygiw and YggE), three others located in the cytoplasm (ArcA, Ind-A and YajQ). The 370
cytoplasmic YajQ protein was the single additional protein modified only during the UV-A 371
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
16
photocatalysis and not with TiO2 in the dark, in comparison to the others which were already 372
modified in the dark (TiO2 cytotoxic effect). YggE gene product is involved in the reduction 373
of the intracellular ROS level, and is therefore responsible for lowering the concentration of 374
ROS until a tolerated level (47). ArcA is a two component response regulator involved in the 375
oxidative stress response (48, 49) and YajQ is involved in ROS detoxification mechanism 376
(50). 377
Members of chaperone protein family are also altered, since we have found the 378
disappearance of 2 proteins when studying the TiO2 cytotoxic effect (Tig and DksA) and 2 379
others during the photocatalytic effect at 0.4 g/L of TiO2 (GrpE and Protease-HslV). GrpE 380
heat shock protein in interaction with DnaK and DnaJ can prevent the aggregation of stress-381
denatured proteins (51) and HslV is an ATP-dependent protease which helps in the 382
degradation of denatured proteins (52). 383
Proteins involved in transport and metabolism crucial for bacterial survival and growth, are 384
also altered. 0.4 g/L TiO2 cytotoxicity caused the disappearance of three spots assigned to two 385
Dps isoforms and Bfr, which are responsible for inorganic ion transport and metabolism. Dps 386
protein protects DNA from damages resulting from oxidative stress. This protection is 387
performed by sequestering and mineralizing metal ions, especially ferrous ion, that can be 388
engaged in the Fenton reaction and known to yield hydroxyl radicals that can damage 389
macromolecules such as proteins, membrane lipids and DNA (20, 53-55). We have noted that 390
the cytotoxic effect at TiO2 concentration of 0.4 g/L induced the disappearance of two 391
isoforms of this protein (spots 12 and 13). The second protein is Bfr, which is a 392
bacterioferritin, and has specific detoxifying properties towards the damaging action of ROS 393
(56). 394
Small ribosomal proteins were also altered by TiO2 cytotoxic effect. We have observed the 395
disappearance of the spot corresponding to the Ribosomal S-6. Protein S6 is located in the 396
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
17
central domain of the small ribosomal subunit and the level of this small ribosomal protein S6 397
could be modified by the induction of the SoxRS regulon (Sox R being presented as a sensor 398
for superoxide anion) (35). However it was shown that E.coli strains with in-frame deletions 399
of the S6 ribosomal protein are still viable and show no deficiency of 30S protein composition 400
(57). 401
Although no information about metabolism regulation during the photocatalytic treatment 402
can be derived from the 2-DE analysis, we suggested that protein modification could disrupt 403
the communication ability of the bacteria with the environment. 404
We hypothesized that such potential protein alteration and lipid peroxidation participated 405
to the antibacterial activity which occurred during the UV-A photocatalytic treatment with 406
TiO2. For this study demonstrating the impact of UV-A photocatalysis on TiO2 on the 407
proteome of E.coli, we selected a qualitative analysis in order to identify the protein spots that 408
are significantly affected (presence vs. absence) by the different treatments applied. Twenty-409
two proteins are surely few and much more are certainly modified, but this approach allowed 410
to demonstrate that the different altered spots belong to several protein functional categories. 411
The hypothesis that all the E.coli proteins could be affected without any important selectivity 412
during photocatalysis as previously suggested by Goulhen-Chollet et al. (19) could be now 413
supported by our observation, since the proteins altered by UV-A photocatalysis on TiO2 414
belong to several functional categories, with a majority of them involved in oxidative stress 415
and porins. The proteomic approach was informative to identify the proteins altered in the 416
antibacterial process taking place with TiO2 nanoparticles, and may ultimately be used to 417
determinine how their expressions are controlled. Further works will concern comparative 418
genomic/transcriptomic/proteomic analyses - e.g. performed on mutant strains -, for getting 419
more insights on the development of molecular mechanism of bacteria resistance towards 420
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
18
UV-A photocatalysis with TiO2, that could occur in bacteria containing a different 421
enzymatic/proteomic arsenal. 422
423
CONCLUSION 424
The present study have identified the effects of UV-A photocatalysis on TiO2 and of TiO2 in 425
the dark on key cellular components of E. Coli such as lipid and proteins. It showed that the 426
antibacterial photocatalytic activity was accompanied by lipid peroxidation, resulting from the 427
reaction with the O2°- superoxide radical ROS. This lipid peroxidation could enhance 428
membrane fluidity and disrupt the cell integrity. The 2-DE proteomic analysis is a first study 429
to get insight on the understanding of the antimicrobial effect of TiO2 photocatalysis, and was 430
focused on the identification of potential protein targets modified during the cytotoxic 431
treatment in the dark with TiO2 and the TiO2 photocatalytic treatment in the presence of UV-432
A irradiation. The mass spectrometry protein analysis have shown that the protein alterations 433
were not specific to a functional class, although some proteins such as outer membrane 434
proteins were the main target during both treatments, probably as a result of a greater 435
exposure to the surface of TiO2 particles. The main degraded proteins were porins and 436
proteins mainly involved in the response to oxidative stress or in environmental stress 437
regulation. 438
439
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
19
ACKNOWLEDGMENTS 440
The authors are grateful to DGA (Direction Générale de l’Armement) and Alsace regional 441
council for financially supporting this work in the frame of the PhD grant of G.C and Dr 442
Nicole Glasser for its contribution to statistical analysis. 443
444
REFERENCES 445
1. Silvestry-Rodriguez N, Sicairos-Ruelas EE, Gerba CP, and Bright KR. 446 2007. Silver as a disinfectant. Rev Environ Contam Toxicol 191:23-45. 447
2. Ruparelia JP, Chatterjee AK, Duttagupta SP, and Mukherji S. 2008. Strain 448 specificity in antimicrobial activity of silver and copper nanoparticles. Acta 449 Biomater 4:707-16. 450
3. Wolska KI, Grzes K, and Kurek A. 2012. Synergy between novel 451 antimicrobials and conventional antibiotics or bacteriocins. Pol J Microbiol 452 61:95-104. 453
4. Borkow G, and Gabbay J. 2005. Copper as a biocidal tool. Curr Med Chem 454 12:2163-75. 455
5. Silver S, and Phung LT. 1996. Bacterial heavy metal resistance: new 456 surprises. Annu Rev Microbiol 50:753-89. 457
6. Malato S, Fernandez-Ibanez P, Maldonado MI, Blanco J, and Gernjak W. 458 2009. Decontamination and disinfection of water by solar photocatalysis:recent 459 overview and trends. Catalysis Today 147:1 - 59. 460
7. Matsunaga T, Tomoda R, Nakajima T, and Wake H. 1985. 461 Photoelectrochemical sterilization of microbial cells by semiconductor 462 powders. FEMS Microbiol. Lett 29:211-214. 463
8. Wei C, Lin WY, Zainal Z, Williams NE, Zhu K, Kruzic AP, Smith RL, and 464 Rajeshwar K. 1994. Bactericidal Activity of TiO2 Photocatalyst in Aqueous 465 Media: Toward a Solar-Assisted Water Disinfection System. Environ Sci 466 Technol 28:934-8. 467
9. Gelover S, Gomez LA, Reyes K, and Teresa Leal M. 2006. A practical 468 demonstration of water disinfection using TiO2 films and sunlight. Water Res 469 40:3274-80. 470
10. Ede S, Hafner L, Dunlop P, Byrne J, and Will G. 2012. Photocatalytic 471 Disinfection of Bacterial Pollutants Using Suspended and Immobilized TiO2 472 Powders. Photochem Photobiol 88:728-35. 473
11. Carp O, Huisman CL, and Reller A. 2004. Photoinduced reactivity of titanium 474 dioxide. Progress in solid state chemistry 32:33-177. 475
12. Foster HA, Ditta IB, Varghese S, and Steele A. 2011. Photocatalytic 476 disinfection using titanium dioxide: spectrum and mechanism of antimicrobial 477 activity. Appl Microbiol Biotechnol 90:1847-68. 478
13. Dalrymple OK, Stefanakos E, Trotz MA, and Goswami DY. 2010. A review 479 of the mechanisms and modeling of photocatalytic disinfection. Applied 480 Catalysis B: Environmental 98:27-38. 481
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
20
14. Josset S, Keller N, Lett MC, Ledoux MJ, and Keller V. 2008. Numeration 482 methods for targeting photoactive materials in the UV-A photocatalytic removal 483 of microorganisms. Chem Soc Rev 37:744-55. 484
15. Cho M, Chung H, Choi W, and Yoon J. 2005. Different inactivation behaviors 485 of MS-2 phage and Escherichia coli in TiO2 photocatalytic disinfection. Appl 486 Environ Microbiol 71:270-5. 487
16. Robertson PK, Robertson JM, and Bahnemann DW. 2012. Removal of 488 microorganisms and their chemical metabolites from water using 489 semiconductor photocatalysis. J Hazard Mater 211-212:161-71. 490
17. Gogniat G, Thyssen M, Denis M, Pulgarin C, and Dukan S. 2006. The 491 bactericidal effect of TiO2 photocatalysis involves adsorption onto catalyst and 492 the loss of membrane integrity. FEMS Microbiol Lett 258:18-24. 493
18. Maness PC, Smolinski S, Blake DM, Huang Z, Wolfrum EJ, and Jacoby 494 WA. 1999. Bactericidal activity of photocatalytic TiO2 reaction: toward an 495 understanding of its killing mechanism. Appl Environ Microbiol 65:4094-8. 496
19. Goulhen-Chollet F, Josset S, Keller N, Keller V, and Lett M-C. 2009. 497 Monitoring the bactericidal effect of UV-A photocatalysis: A first approach 498 through 1D and 2D protein electrophoresis. Catalysis Today 147:169-172. 499
20. Gogniat G, and Dukan S. 2007. TiO2 photocatalysis causes DNA damage 500 via fenton reaction-generated hydroxyl radicals during the recovery period. 501 Appl Environ Microbiol 73:7740-3. 502
21. Huang Z, Maness P-C, Blake DM, Wolfrum EJ, Smolinski SL, and Jacoby 503 WA. 2000. Bactericidal mode of titanium dioxide photocatalysis. Journal of 504 Photochemistry and Photobiology A: Chemistry 130:163-170. 505
22. Gumy D, Morais C, Bowen P, Pulgarin C, Giraldo S, Hajdu R, and Kiwi J. 506 2006. Catalytic activity of commercial of TiO2 powders for the abatement of the 507 bacteria (E.coli) under solar simulated light: influence of the isoelectric point. 508 Applied Catalysis B: Environmental 63:76-84. 509
23. Vacaroiu C, Enache M, Gartner M, Popescu G, Anastasescu M, Brezeanu 510 A, Todorova N, Giannakopoulou T, Trapalis C, and Dumitru L. 2009. The 511 effect of thermal treatment on antibacterial properties of nanostructured 512 TiO2(N) films illuminated with visible light. World Journal of Microbiology and 513 Biotechnology 25:27-31. 514
24. Pigeot-Rémy S, Simonet F, Errazuriz-Cerda E, Lazzaroni JC, Atlan D, and 515 Guillard C. 2011. Photocatalysis and disinfection of water: Identification of 516 potential bacterial targets. Applied Catalysis B: Environmental 104:390-398. 517
25. Lukjancenko O, Wassenaar TM, and Ussery DW. 2010. Comparison of 61 518 sequenced Escherichia coli genomes. Microb Ecol 60:708-20. 519
26. Hamon E, Horvatovich P, Izquierdo E, Bringel F, Marchioni E, Aoude-520 Werner D, and Ennahar S. 2011. Comparative proteomic analysis of 521 Lactobacillus plantarum for the identification of key proteins in bile tolerance. 522 BMC Microbiol 11:63. 523
27. Geer LY, Markey SP, Kowalak JA, Wagner L, Xu M, Maynard DM, Yang X, 524 Shi W, and Bryant SH. 2004. Open mass spectrometry search algorithm. J 525 Proteome Res 3:958-64. 526
28. Kapp EA, Schutz F, Connolly LM, Chakel JA, Meza JE, Miller CA, Fenyo 527 D, Eng JK, Adkins JN, Omenn GS, and Simpson RJ. 2005. An evaluation, 528 comparison, and accurate benchmarking of several publicly available MS/MS 529 search algorithms: sensitivity and specificity analysis. Proteomics 5:3475-90. 530
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
21
29. Herrmann JM. 1999. Heterogeneous photocatalysis: fundamentals and 531 applications to the removal of various types of aqueous pollutants. Catalysis 532 Today 53:115-129. 533
30. Herrmann JM. 2005. Heterogeneous photocatalysis: state of the art and 534 present applications. Topics in Catalysis 34:49-65. 535
31. Carre G, Benhamida D, Peluso J, Muller CD, Lett MC, Gies JP, Keller V, 536 Keller N, and Andre P. 2013. On the use of capillary cytometry for assessing 537 the bactericidal effect of TiO2. Identification and involvement of reactive 538 oxygen species. Photochemical & Photobiological Sciences. 12:610-620 539
32. Oliver JD, Dagher M, and Linden K. 2005. Induction of Escherichia coli and 540 Salmonella typhimurium into the viable but nonculturable state following 541 chlorination of wastewater. J Water Health 3:249-57. 542
33. Hoffmann MR, Martin ST, Choi W, and Bahnemann DW. 1995. 543 Environmental Applications of Semiconductor Photocatalysis. Chem. Rev. 544 95:69-96. 545
34. Ma H, Brennan A, and Diamond SA. 2012. Photocatalytic reactive oxygen 546 species production and phototoxicity of titanium dioxide nanoparticles are 547 dependent on the solar ultraviolet radiation spectrum. Environ Toxicol Chem 548 31:2099-107. 549
35. Lushchak VI. 2011. Adaptive response to oxidative stress: Bacteria, fungi, 550 plants and animals. Comp Biochem Physiol C Toxicol Pharmacol 153:175-90. 551
36. Mitoraj D, Janczyk A, Strus M, Kisch H, Stochel G, Heczko PB, and 552 Macyk W. 2007. Visible light inactivation of bacteria and fungi by modified 553 titanium dioxide. Photochem Photobiol Sci 6:642-8. 554
37. Kohen R, and Nyska A. 2002. Oxidation of biological systems: oxidative 555 stress phenomena, antioxidants, redox reactions, and methods for their 556 quantification. Toxicol Pathol 30:620-50. 557
38. Sawyer D.T. 1981. How super is superoxide? Account of Chemical Research 558 14:393-400 559
39. Daimon T, Hirakawa T, Kitazawa M, Suetake J, Nosaka Y. 2008. Formation 560 of singlet molecular oxygen associated with the formation of superoxide 561 radicals in aqueous suspensions of TiO2 photocatalysts. Applied Catalysis A: 562 General 340:169-175. 563
40. Rengifo-Herrera J.A., Pierzchala K., Sienkiewicz A., Forro L, Kiwi J, 564 Pulgarin C. 2009. Abatement of organics and Escherichia coli by N,S co-565 doped TiO2 under UV and visible light. Implications of the formation of singlet 566 oxygen (1O2) under visible light. Applied catalysis B: Environmental 88:398-567 406. 568
41. Molloy MP, Herbert BR, Slade MB, Rabilloud T, Nouwens AS, Williams KL 569 and Gooley AA. 2001. Proteomic analysis of the Escherichia coli outer 570 membrane. European Journal of Biochemistry, 267(10):2871-2881. 571
42. Krämer R. 2010. Bacterial stimulus perception and signal transduction: 572 response to osmotic stress. The Chemical Record 10(4):217-29. 573
43. Golding I, Paulsson J, Zawilski SM, Cox EC. 2005 Real-time kinetics of 574 gene activity in individual bacteria. Cell 123(6):1025-36 575
44. Jaroslawski S, Duquesne K, Sturgis J and Scheuring S. 2009. High-576 resolution architecture of the outer membrane of the Gram-negative bacteria 577 Roseobacter denitrificans. Molecular microbiology, 74(5):1211-1222. 578
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
22
45. Tamarit J, Cabiscol E, and Ros J. 1998. Identification of the major 579 oxidatively damaged proteins in Escherichia coli cells exposed to oxidative 580 stress. J Biol Chem 273:3027-32. 581
46. Krishnan S, and Prasadarao NV. 2012. Outer membrane protein A and 582 OprF: versatile roles in Gram negative bacterial infections. FEBS J.279:919-583 31. 584
47 Kim SY, Nishioka M, Hayashi S, Honda H, Kobayashi T, and Taya M. 585 2005. The gene yggE functions in restoring physiological defects of 586 Escherichia coli cultivated under oxidative stress conditions. Appl Environ 587 Microbiol 71:2762-5. 588
48. Lu S, Killoran PB, Fang FC, and Riley LW. 2002. The global regulator ArcA 589 controls resistance to reactive nitrogen and oxygen intermediates in 590 Salmonella enterica serovar Enteritidis. Infect Immun 70:451-61. 591
49. Kohanski MA, Dwyer DJ, Wierzbowski J, Cottarel G, and Collins JJ. 2008. 592 Mistranslation of membrane proteins and two-component system activation 593 trigger antibiotic-mediated cell death. Cell 135:679-90. 594
50. Amitai S, Kolodkin-Gal I, Hananya-Meltabashi M, Sacher A, and 595 Engelberg-Kulka H. 2009. Escherichia coli MazF leads to the simultaneous 596 selective synthesis of both "death proteins" and "survival proteins". PLoS 597 Genet 5:e1000390. 598
51. Harrison C. 2003. GrpE, a nucleotide exchange factor for DnaK. Cell Stress 599 Chaperones 8:218-24. 600
52. Lau-Wong IC, Locke T, Ellison MJ, Raivio TL, and Frost LS. 2008. 601 Activation of the Cpx regulon destabilizes the F plasmid transfer activator, 602 TraJ, via the HslVU protease in Escherichia coli. Mol Microbiol 67:516-27. 603
53. Nair S, and Finkel SE. 2004. Dps protects cells against multiple stresses 604 during stationary phase. J Bacteriol 186:4192-8. 605
54. Yellaboina S, Ranjan S, Chakhaiyar P, Hasnain SE, and Ranjan A. 2004. 606 Prediction of DtxR regulon: identification of binding sites and operons 607 controlled by Diphtheria toxin repressor in Corynebacterium diphtheriae. BMC 608 Microbiol 4:38. 609
55. Liu X, and Theil EC. 2005. Ferritins: dynamic management of biological iron 610 and oxygen chemistry. Acc Chem Res 38:167-75. 611
56. Chiancone E, Ceci P, Ilari A, Ribacchi F, and Stefanini S. 2004. Iron and 612 proteins for iron storage and detoxification. Biometals 17:197-202. 613
57. Shajani Z, Sykes MT, and Williamson JR. 2011. Assembly of bacterial 614 ribosomes. Annu Rev Biochem 80:501-26. 615
616
617
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
23
Figure Legends 618
619
FIG. 1 Influence of the photocatalytic treatment on E. coli cell viability, determined by 620
counting on agar plate. Bars indicate standard deviations. in the dark, under UV-A 621
irradiation. * (P<0.05), ** (P<0.01), *** (P<0.001) and **** (P<0.0001) indicate the 622
significance of the difference between sample means obtained with Student’s t test. 623
624
FIG. 2. Impact of UV-A photocatalysis with TiO2 and influence of the addition of SOD at 625
1000 U on E.coli lipid peroxidation. A duration of 60 min and a TiO2 concentration of 0.4 g/L 626
were chosen. Membrane lipid peroxidation is measured with the production of 627
malondialdehyde (MDA). MDA is a biomarker of lipid peroxidation, and is used to measure 628
the lipid peroxidation rate relative to controls (MDA concentration obtained without and with 629
treatment). Bars indicate standard deviations. in the dark, under UV-A irradiation. 630
**(P<0.01), ****(P<0.0001) and ns (not-significant) indicate the significance of the 631
difference between sample means obtained with Student’s t test. 632
633
FIG. 3. Impact of TiO2 photocatalysis on the whole cell proteome of E.coli ATCC 8739. The 634
figure shows representative 2-DE gel pictures (pH range 4-7) of whole-cell protein lysates 635
after different treatments for 30 min UV-A light irradiation. Previously, bacteria have grown 636
during 16 h-18 h on tryptic soy agar at 37°C. (A) standard conditions (without TiO2 in the 637
dark ); (B) under UV-A light irradiation; (C and E) in the dark at TiO2 concentration of 0.1 638
g/L and 0.4 g/L, respectively; (D and F) under UV-A light irradiation at a TiO2 concentration 639
of 0.1 g/L and 0.4 g/L, respectively. Spots present (black numbers) in standard conditions (gel 640
A) but not detected (white numbers) after one of the applied treatments (gels B-F). Protein 641
corresponding to these spots were identified by Chip-LC-QTOF analysis. 642
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from
24
TABLE 1. Identification of missing proteins in E. coli ATCC 8739 whole-cell proteomes 643
recorded after 30 min of UV-A treatment, TiO2 exposure in the dark and photocatalytic test. 644
645
646
a Peptides matched: a number of tryptic peptides observed contributing to the percentage of 647
sequence coverage. 648 b pI: theorical isoelectric point. 649 c mW: theorical molecular weight in Da. 650 d Treatment: nature of the treatment leading to the disappearance of some proteins in 2-DE 651
gels when compared to standard conditions. 652 e UV treatment: 30 min of UV-A irradiation. 653 f, g C1, C2: treatment with a cytotoxic effect of TiO2 at 0.1 g/L and 0.4 g/L, respectively. 654 h ,i P1; P2: photocatalytic treatment with TiO2 at 0.1 g/L and 0.4 g/L, respectively. 655
on May 11, 2018 by guest
http://aem.asm
.org/D
ownloaded from