masters research project

61
1 NAT-MN0Y 2015-16 Natural Sciences Project Multiple Reaction Monitoring Mass Spectrometry of Myoglobin Peptides to Determine Tuna Authenticity Author: 6157106 Supervisor: Dr Andrew Mayes

Upload: joshua-kai-peazer

Post on 07-Apr-2017

26 views

Category:

Documents


7 download

TRANSCRIPT

Page 1: Masters Research Project

1

NAT-MN0Y

2015-16 Natural Sciences Project

Multiple Reaction Monitoring Mass Spectrometry of

Myoglobin Peptides to Determine Tuna Authenticity

Author: 6157106

Supervisor: Dr Andrew Mayes

Page 2: Masters Research Project

2

TABLE OF CONTENTS

1. Abstract.................................................................................................................................... 4

2. Introduction.......................................................................................................................... 4

2.1. The Tuna Industry........................................................................................................... 6

2.2. Tuna Authentication Protocols................................................................................. 9

2.2.1. DNA-based Analysis.............................................................................................. 9

2.2.1.1. PCR-RFLP..................................................................................................... 10

2.2.1.2. Immunological Analysis............................................................................ 10

2.2.2. Proteomics............................................................................................................. 11

2.2.3. Mass Spectrometry.............................................................................................. 13

2.2.3.1. Separation Techniques.............................................................................. 13

2.2.3.2. Ion Sources.................................................................................................. 15

2.2.3.3. Mass Analysers.......................................................................................... 16

2.2.3.4. Tandem Mass Spectrometry.................................................................... 17

2.3. The Present Work.......................................................................................................... 20

2.3.1. Objectives & Hypotheses............................................................................... 24

3. Methodology...................................................................................................................... 25

3.1. Materials........................................................................................................................... 25

3.2. Purified Myoglobin Preparation............................................................................. 25

3.3. Raw Sample Preparation........................................................................................... 26

3.4. Processed Sample Preparation............................................................................... 26

3.5. Proteolysis........................................................................................................................ 26

3.6. Predictive Tools.............................................................................................................. 27

3.7. LC-MS/MS Workflow................................................................................................... 27

3.8. LC-MS/MS Analyses...................................................................................................... 28

Page 3: Masters Research Project

3

4. Results..................................................................................................................................... 29

4.1. Peptide Marker Selection.......................................................................................... 29

4.2. Multiplex Myoglobin Analyses................................................................................ 32

4.3. Multiplex Raw Tuna Analyses.................................................................................. 34

4.4. Multiplex Processed Tuna Analyses...................................................................... 35

5. Discussion............................................................................................................................. 36

5.1. Marker Reliability......................................................................................................... 36

5.1.1. Purified Myoglobin.............................................................................................. 37

5.1.2. Raw Samples......................................................................................................... 39

5.1.3. Processed Samples.............................................................................................. 40

5.2. Marker Specificity......................................................................................................... 41

5.2.1. Carryover................................................................................................................ 42

6. Conclusion............................................................................................................................ 45

7. Acknowledgments.......................................................................................................... 47

8. References............................................................................................................................ 47

9. Appendix............................................................................................................................... 56

9.1. Research Proposal........................................................................................................ 56

Page 4: Masters Research Project

4

1. ABSTRACT

The improper description of tuna products is of concern, due to the frequent, fraudulent

substitution of species within these commodities. This creates a demand for novel

authentication protocols to accurately identify species within tuna products. The application

of a recently developed approach to four species of tuna: yellowfin, albacore, skipjack and

bigeye is presented here. This method, known as the CPCP approach (corresponding proteins

corresponding peptides), was originally developed to rapidly distinguish between species in

red meat products. Peptides selected as markers for species identification using predictive

tools, are detected via multiple reaction monitoring mass spectrometry (MRM MS). These are

derived solely from the digestion of myoglobin (Mb) from the species of interest. In the

present work, Mb peptide markers able to distinguish between the four species of tuna are

identified. These were visible in both purified myoglobin and in raw crude extracts in all four

species. These could not be used to distinguish between species however, due to the

occurrence of carryover. Further development of the technique should focus on the

elimination of carryover within the system to restore marker specificity. The application of

this technique to heavily processed tuna samples, as found in canned tuna, should also be

studied further. Both were attempted here, though unsuccessfully.

KEYWORDS: authentication, food, fish, tandem mass spectrometry, peptide biomarker

2. INTRODUCTION

It is the right of consumers that the origins and contents of any food product are described

clearly and accurately. This offers patrons the ability to make informed choices about these

commodities that directly impact their diet, health and lifestyle. Proper description is also

important to honest practitioners in the food industry, as the fraudulent labelling of food can

provide an economic advantage to neglectful or deceitful competitors.1

Food authentication is the verification that a food corresponds to its label description. The

scale and complexity of the industry provides many avenues of exploitation for fraudulent

processors, which have long been a documented issue warranting a global response (Box 1).2

One such avenue involves the partial, or complete, substitution of an ingredient for a similar

Page 5: Masters Research Project

5

alternative. This can be

accidental or intentional,

motivated by a difference in

value between the original and

its substitute, the adulterant.2

When intentional, the adulterant

is cheaper or easier to come by

than the original, to provide

commercial gain from

substitution. This practice was

demonstrated in Europe, when

an array of beef products were

shown to contain or consist

entirely of undeclared horse

meat.3 The scale of this incident,

known as the European

horsemeat scandal of 2013,

increased consumer and

government awareness of issues

in food fraud regulation and deterrence.4

At5 the6 time7 of the horsemeat scandal, EU legislation concerning food regulation was largely

concerned with issues of food safety, not fraud.4, 8 Procedure has since been established to

facilitate increased cooperation between member states, as well as the implementation of

union wide controls in the supply chain. New agencies have also been developed, and existing

agencies bolstered, to directly combat food fraud.4 An example of this is the development of

the National Food Crime Unit (NFCU) within the Food Standards Agency (FSA). The FSA is the

government body responsible for ensuring food is of the standard expected by consumers in

the UK.9

Tuna production is another area in the food industry at risk of fraudulent practice.10 This fact

is recognised by the European Union (EU), whose regulations state the appropriate contents

COMMERCIAL TREATMENT The misrepresentation of how a food is produced or treated. Certain treatments are desirable to consumers which indicate a product of good quality (e.g. pasteurized), whilst others are treated with suspicion (e.g. frozen). Investigators found that seven leading brands of olive oil were falsely labelled as extra virgin. Extra virgin olive oils are products of high value due to the use of more time consuming and arduous processes to yield oil of better quality.5

PRODUCT EXTENSION The adulteration of a product with a base ingredient, such as water, to extend foodstuffs increasing the product yield at minimal cost. This has been observed in a wide array of both solid and liquid food products. An investigation into the authenticity of wine vinegar revealed 60% of 92 tested vinegars imported from the Italian market were shown to have been produced by diluting from a concentrate with water, making them fraudulent.6

BRAND COUNTERFEITING The replication of a food using the brand name of the original. Similarly to geographical location and commercial treatments, brand names are often associated with product quality. This provides economic benefit to the perpetrator by selling inferior foodstuffs at a higher price rate, based purely on the reputation of the imitated brand. It was observed in London 2002 that 3 out of 5 bottles of seized Johnny Walker Black Label whisky were not authentic causing both financial and health detriments as the false products contained dangerous levels of methanol.5 The authenticity of alcohol is particularly difficult to determine as most protocols involve the destruction of the product once bottled, making luxury brands at high risk of imitation.7

Box 1. Various authenticity issues. The multitude of methods and cases of

fraudulent food representation illustrates that continued vigilance is

required. This should involve the development of dependable and efficient

food authentication protocols worldwide.

Page 6: Masters Research Project

6

and labelling of tuna products.11 The current work concerns the development of a novel

method of product analysis usable in the tuna supply chain, to enable these regulations to be

properly enforced. The approach utilizes multiple reaction monitoring mass spectrometry to

discriminate between the myoglobin of different tuna species. To provide context, it is first

necessary to discuss issues in the tuna industry, as well as reviewing technologies currently

relevant to tuna authentication. The use of chromatography and mass spectrometry will also

be detailed, before the current work is expanded upon.

2.1. THE TUNA INDUSTRY

The production of raw and processed tuna is of great economic importance. Between 1952

and 2010, the annual production of tuna has increased from 0.4 million tonnes to a peak of

over 6 million tonnes.10,12 Tuna and tuna-like species used in food products are caught

globally, primarily via longline, purse seine or pole and line fishing.13 Longliners cast nets at

the greatest depths, and purse seine and pole and line fisherman harvest fish close to the

surface with nets or poles respectively. The size of both catch and individual fish tends to

increase with the depth of catch.13 This means the fish caught by pole and line or purse seine

vessels are smaller. These are most suitable for canning, as larger fish yield more intact meat.

For this reason, the larger fish caught via longline tend to be sold whole or as steaks. (Figure

1).

0° 40° E 80° E 120° E 160° E 160° W 120° W 80° W 40° W

30° S

60° S

90° N

60° N

30° N

LONGLINE

POLE & LINE, PURSE SEINE, LONGLINE

PURSE SEINE, LONGLINE

PACIFIC

OCEAN

ATLANTIC OCEAN

INDIAN

OCEAN

PACIFIC

OCEAN

ARCTIC OCEAN

1

2

3

4

5

6

6

7 7

Figure 1 Tuna stocks, catch methods and distribution. A stock in this context is a subpopulation of a particular species of fish.13

The numbers bounded by the dotted lines indicate the locations of the main tuna stocks and their boundaries. Yellowfin

(Thunnus albacares), albacore (Thunnus alalunga), skipjack (Katsuwonus pelamis) and bigeye (Thunnus obesus) are caught from

all stocks. Atlantic (Thunnus thynnus), and Pacific (Thunnus orientalis) bluefin are caught from all stocks excluding stock 5.

Southern bluefin (Thunnus maccoyii) are only caught in stocks 2 and 4.10 The main locations of tuna canneries are indicated by

blue circles.14

Page 7: Masters Research Project

7

Once14 caught,10 tuna13 are frozen prior to transportation for processing, packaging and then sale.

Dependent on the operation, a single product can be transported multiple times between or

within any stage of production.13 It is important that operations are monitored at each stage

of production, as there are more opportunities for fraudulent practice to occur in supply

chains of increased complexity. Authentication becomes more difficult as the number of

locations and practitioners involved in production increases. This is because it becomes less

clear where and how fraud occurs in the supply chain.

Authenticity issues arise due to the

deviation in value of species used as raw

material for tuna products. Of the 61 tuna

and tuna-like species within the

Scombridae family, seven are attributed

with two-thirds of the total catch.15 These

species, termed the principal market

tunas by the FAO,10 are yellowfin

(Thunnus albacares), albacore (Thunnus

alalunga), skipjack (Katsuwonus

pelamis), bigeye (Thunnus obesus),

Atlantic bluefin (Thunnus thynnus),

Pacific bluefin (Thunnus orientalis) and

southern bluefin (Thunnus maccoyii).

Within these species of tuna there is

even further deviation in commercial significance, based on the availability of each species

(Figure 2). This is reflected in the price of tuna products.10 Skipjack, the most abundantly caught

species, yields the cheapest products. Bluefin species are in danger of extinction due to

overfishing, as recognised by the IUCN, the leading authority in conservation ecology.16-18 As

a result the capture of all three species is restricted, making their products both rare and

expensive. This price differential creates financial motive for species substitution. Effective

authentication via product speciation (i.e. species identification) protocols is required to

counter this.

Figure 2 The global catch of principal market tuna. The total catch

in 2010 weighed approximately 4 million tonnes in total.12

Yellowfin, bigeye and bluefin species are used in sashimi, a

delicacy in Japanese cuisine. In this market, the three bluefin

species are the most valuable (US$30 per kg minimum) followed

by bigeye. Yellowfin fetch much lower prices on this market.10 The

canned tuna market primarily uses skipjack, albacore and

yellowfin. Canned market value are inverse to species catch

proportion, with skipjack products being the cheapest, followed

by canned yellowfin then albacore products being the highest

value. Bigeye species have been used for canning though these

products are rare. Bluefin species are not used for canning.10

Skipjack58%

Atlantic, Pacific & Southern Bluefin

1%

Yellowfin27%

Bigeye8%

Albacore6%

Page 8: Masters Research Project

8

The threat of overfishing further highlights the need for effective speciation in tuna products,

to facilitate the recovery of bluefin populations. Certain catch methods, known as Fish

Aggregation Devices (FADs), encourage fish to congregate, allowing larger numbers to be

caught.19 This can be unsustainable when endangered fish outside the target species are

attracted, leading to their capture amongst the desired fish. This is known as bycatch, and is

reported to contribute to the observed overfishing in bluefin populations, as well as damaging

other ecosystems.20 This can also lead to accidental species substitution when substitutes go

unnoticed during tuna operations. By developing new methods of tuna product

authentication, tuna operations can be better regulated, compelling organisations to fish

sustainably.20

The risk of fraudulent substitution is compounded when fish products are made. Tunas can

be sold whole, but can also be skinned, filleted and portioned to make products consisting of

only a section of the flesh. As a result, these products bear little resemblance to their species

of origin, due to a loss of most morphological characteristics during production. Some

characteristics are maintained, such as the pink flesh coloration of tuna. The exception to this

is albacore, which has white muscle tissue. This feature is translated to albacore products.

However, visual discrimination is still difficult for consumers, as the flesh of most tuna species

is similar in appearance. Species substitution occurs frequently as a result, using substitutes

both within the Scombridae family and outside of it. This is exemplified in an investigation by

Warner et al.21 The study carried out across the US between 2010 and 2012 revealed that

59% of 114 tuna samples purchased from retailers were mislabelled. In some cases, escolar

(Lepidocybium flavobrunneum) was often used in place of albacore due to their similar

morphologies (white flesh). Escolar is a species that is not digestible by humans without

negative health effects.21

It becomes impossible for consumers to verify tuna species when products are canned, as the

flesh cannot be seen before purchase. As the raw material loses even more identifying

features during the canning process, it is also harder to regulate the fraudulent substitution

of species in these products.22 In particular many authentication techniques are challenged

by two heating stages that occur during processing, once to cook the flesh and once to

Page 9: Masters Research Project

9

sterilise the canned product. The sterilisation step is particularly severe, with cans being

heated to a minimum of 115.6 °C in a retort for at least 40 mins.23 This leads to further changes

in the morphology of tuna flesh, and can sometimes result in damage to, or loss of analytes

used by authentication techniques. Additionally, tuna can be cut into smaller pieces (i.e.

slices, chunks or flakes) prior to canning. This increases the risk of partial substitution within

these products, compared to those consisting of whole portions of flesh (i.e. steaks), as

individual pieces within a product could be sourced from different species. This translates to

the frequent occurrence of canned tuna fraud. Investigation into popular brands of tinned

tuna in twelve countries (including Australia, the UK and the US) illustrates this issue. It was

revealed that 30.3% of the 165 products tested contained, or consisted of, species not

declared on the label.24

2.2. TUNA AUTHENTICATION PROTOCOLS

The cases of fraudulent practice, and the risk of its occurrence in the future, accents the need

for novel methods of tuna classification to ensure adulteration detection. Techniques most

widely used rely on DNA based analysis, though a variety of immunological, chromatographic

and spectroscopic tools are also relevant.

2.2.1. DNA-based Analysis

The analysis of species-specific DNA generally involves methods derived from the polymerase

chain reaction (PCR). This replicates specific genes of interest to create an amplicon (i.e.

amplified material) for further testing.1 The use of species-specific primers, base sequence

comparison, fragment length comparison, or other approaches can then reveal further

information about a sample.25 The level of detail obtainable from non-coding regions of DNA

makes these techniques useful.26 They have been successfully used for the commercial

identification of species and geographical origin as well in other areas of authentication for a

range of commodities.27

When determining an analyte, MT-CYB, a region of mitochondrial DNA (mtDNA) is often

selected for the differentiation of fish species.22 This gene exhibits considerable inter-species

variation due, in part, to the rapid evolution of mtDNA. Additionally, the reported

Page 10: Masters Research Project

10

thermostability of DNA relative to most proteins, combined with the abundance of

mitochondria in a cell, makes MT-CYB suitable for canned tuna analyses. This is because the

thermal treatment during canning is claimed to result in the irreversible degradation of many

analytes within a sample.22

2.2.1.1. PCR-RFLP

In terms of tuna, restriction fragment length polymorphism (RFLP) preceded by PCR

amplification is widely used for fish species identification.27 PCR-RFLP is an inexpensive and

simple technique carried out by the digestion of an amplicon, using a restriction enzyme

(endonuclease). The fragments (termed alleles) can then be separated by length using gel

electrophoresis. Fragment length polymorphism is observed when there is variation of

corresponding allele lengths between individuals. By comparing the presence of certain

alleles with the allelic frequencies observed in a given species, the origin of a sample can be

established. An investigation carried out by Lin & Hwang 28 demonstrates the ability of the

technique to differentiate between tuna. The study utilized PCR-RFLP comparison of MT-CYB

to correctly identifying eight species amongst 18 commercially canned tuna products.28

Despite the strengths of PCR-RFLP, there are shortcomings intrinsic to the technique. Firstly,

the use of PCR-RFLP is dependent on the allelic database used for species determination. As

DNA is also variable intra-specifically, substantial polymorphisms are observed between the

restriction fragment lengths of conspecific individuals. Substantial sample sizes from each

species of tuna are therefore required to account for these variations.22, 27 Another issue is

that the restriction patterns of a species are not necessarily unique. This could lead to the

occurrence of false positives when carrying out analyses on unidentified samples.27 To avoid

difficulties associated with these and other limitations, standard practice specifies that PCR-

RFLP cannot be relied upon solely for the purposes of tuna authentication, due to the risk of

unknown adulterants in tuna products.27

2.2.1.2. Immunological Analysis

Enzyme-linked immunosorbent assays (ELISA) have many applications in authentication

including allergen,29 irradiation30 and genetically modified food detection.31 They are also

Page 11: Masters Research Project

11

used for the species assignment of milk and meat products. This involves the detection of a

species-specific analyte using a complementary antibody whose presence can be revealed via

the use of a flourogenic or chromogenic substrate.32 A variety of kits for numerous species

have been developed for commercial use, containing microtiter plates or immunosticks

coated with the relevant molecules (antigens/antibodies) for simple and rapid analysis. It is

for this reason that some of these kits are used by regulatory agencies to detect meat

adulteration on site, in order to enforce legislation.31 The analyte is usually an antibody or an

antigen though other macromolecules have been used, including DNA sequences.

ELISA targeting DNA has recently been adapted to hierarchically distinguish between tuna

species.33 This version of the method is preceded by PCR, focusing on regions of the MT-CYB

gene as analytes. The method utilized five probes bound to biotin (biotinylation) specific to

T.albacares (yellowfin), T. alalunga (albacore), T. orientalis (Pacific bluefin), T thynnus (bigeye)

and K. pelamis (skipjack). A hierarchical ELISA system was developed consisting of two stages.

The first established the presence of Thunnus in the sample and the second established the

presence of one of the five aforementioned principal market tunas in the sample. After

method validation, 11 commercial samples were analysed, revealing that 2 were

mislabelled.33

With further development, this technique could be beneficial for the field screening of tuna

in industry, as it is in meat production. In its current form however, analysis is still quite time

consuming as kits are not commercially available for the plethora of species involved in the

tuna industry. The reliability of ELISA based speciation has also been criticised, as extensive

purification is required to eliminate cross-reactivity, causing fluctuations in the speed and

efficacy of analysis between species.32

2.2.2. Proteomics

The analysis of proteins for tuna authenticity is controversial, due to their reported lack of

heat stability.22 This implies they are likely to be denatured (i.e. the loss of a protein’s natural

conformation within an organism) and degraded (i.e. the breaking of covalent bonds in the

protein’s amino acid sequence) when heat is applied. This would make them unsuitable

Page 12: Masters Research Project

12

biomarkers for canned tuna which is sterilized by prolonged exposure to high temperatures.23

However, this claim is often disputed in the literature.34 Investigation into the physiochemical

mechanisms behind protein thermostability has established that this property varies greatly

between proteins. This deviation can be attributed to singular changes in the amino acid

sequence of a protein, particularly if the number of hydrogen bonds or ion pairs are

affected.35, 36 Additionally, thermal denaturation and degradation, is tolerable to a certain

extent, when the desired biomarker is one of the protein’s constituent peptides. Analyses

using peptides are only affected when the primary structure (the amino acid sequence) of

that peptide is degraded. As the amino acid

sequences of a peptide are shorter than those of

a protein, and shorter still than many DNA

sequences, the risk of critical peptide

degradation is greatly reduced.37, 38 This implies

that the selection of appropriate proteins (and

peptides) for species identification, can bypass

this issue of thermostability.

Analytes can be selected using mass

spectrometry via top-down, bottom up or

shotgun strategies (Figure 3).39 The40 shotgun41

approach is effective for this application, as

exemplified by von Bargen et al 42. The study

successfully identified 12 peptide sequences that

were characteristic of beef, horse and pork

which were later successfully detected below the

1% level in mixed samples (horse/ pork in beef)

via tandem mass spectrometry. This study

highlights the relevance of proteomic study in

authentication. It also demonstrates that the

adaptation of successful techniques used on

related commodities should be considered for

Figure 3. Workflows for proteomic study. Three

approaches to mass spectrometry: top-down (blue),

bottom-up (orange) and shotgun (green). Top-down

proteomics is the analysis of intact proteins whilst

bottom-up proteomics involves the analysis of

peptides generated by protein digestion. Protein and

peptide mixtures tend to be separated, or certain

molecules isolated, prior to digestion in order to

reduce system throughput and simplify data analysis.

Separation can be achieved via chromatography (e.g.

HPLC) or gel electrophoresis (GE) A variant of the

bottom-up approach, known as shotgun proteomics,

removes this separation step to provide information

on all the constituents of a given sample, though this

reduces the sensitivity and selectivity of analysis.40, 41

Sample

Protein

Mixture

Peptide

Mixture

Separated

Peptides

LC/MS or

LC-MS/MS

Separated

Proteins

ENZYMATICDIGESTION

EXTRACTION

HPLC/ GE

HPLC/ GE

Page 13: Masters Research Project

13

novel tuna authentication protocols. These commodities include other fish, seafood or even

meat products, as just described.42

2.2.3. Mass Spectrometry

The mass and isotopic signature of a sample and it’s constituents can be determined by mass

spectrometry (MS). Mass spectrometers utilize three different components to carry out each

process within an instrument. The ion source ionises the sample, followed by ion separation

in the mass analyser, based on their mass to charge ratio (m/z). Detectors then record when

the sorted ions pass by (or hit) its surface producing an ion signal as a function of m/z. Ions

will behave differently depending on their weight and charge, allowing their

characterization.43 For high throughput analyses (i.e. the analysis of samples containing many

proteins or peptides), MS can be improved with an appropriate separation technique in order

to produce results of better clarity and higher specificity. As the current work centres on mass

spectrometry, it is important to consider suitable MS instrument configurations for the

analysis of tuna products. Additionally, as these products are complex mixtures (i.e. contain

many proteins), it is also important to couple a suitable separation technique to enhance

analysis.

2.2.3.1. Separation Techniques

The present work uses chromatography, which will be the focus of this section. However,

other methods of separation, such as electrophoresis44 or ion mobility spectrometry (IMS)45

are also used in food authentication. Chromatography separates a sample based on the

differential interaction of it’s constituents with two phases, a stationary phase and a mobile

phase. A sample is carried through a structure containing the stationary phase, by the mobile

phase. Constituents with varying chemical structures have differing affinities for the

stationary phase, causing them to be adsorbed onto it for different lengths of time, known as

retention times (Rt). This causes them to be desorbed (termed eluted) separately. They are

then injected into the mass spectrometer for detection in chromatography-MS coupled

systems. The nature of the stationary phase, the mobile phase and the structure (or bed)

housing the stationary phase are ways in which different chromatography techniques are

classified.

Page 14: Masters Research Project

14

In gas chromatography (GC), the mobile phase is gaseous and in liquid chromatography (LC)

the mobile phase is liquid. In the current context, the former is not relevant, as liquid

chromatography is required for the analysis of biological material. This is because it is more

suitable for the analysis of non-volatile and thermally fragile molecules.46 In liquid

chromatography the stationary phase can be contained within a column or present on a

plane. Column chromatography is more efficient than planar chromatography, as the mobile

phase can be pressurised in a technique known as high-performance liquid chromatography

(HPLC). HPLC further improves upon other low pressure chromatography, by utilising smaller

adsorbent particles in the stationary phase (known as packing particles), providing a larger

surface area. This increases the resolution of separation, the degree to which two similar

compounds can be separated. 47

For the analysis of organic analytes (those containing carbon) such as proteins, hydrophobic

stationary phases are preferred. In these systems, analytes are usually dissolved in a polar

mobile phase and molecules within the sample are retained in the column via hydrophobic

interactions. The most hydrophobic molecules are late-eluting whilst the more hydrophilic

ones elute first.48 This is known as reversed-phased chromatography (RPC). A commonly used

stationary phase in (RPC) is an octadecyl carbon chain (18 carbon atoms) bound to silica.

These are found within C18 columns, and provide a large surface area and pore volume to

increase the retention of hydrophobic analytes.48 This increases the selectivity and efficiency

of separations.47 When analysing samples that consist of molecules that exhibit a range of

hydrophobic behaviour, the most hydrophobic molecules can be retained for long periods of

time, increasing overall run times.49 To overcome this gradient elution is used. This is where

the composition of the mobile phase is altered during separation to become more

hydrophobic as separation progresses. This decreases the retention times of the most

hydrophobic constituents, whilst still maintaining effective separation.49

Page 15: Masters Research Project

15

2.2.3.2. Ion Sources

An important component that can dictate the selection of entire mass spectrometry

configurations, is the selection of a sample-appropriate ion source.43 For the analysis of tuna

proteins, or other biological molecules, soft ionisation sources are required. These sources

result in minimal sample fragmentation, by reducing the amount of residual energy exerted

on a sample.43 In macromolecular analysis this is important to ensure that the molecular

(heaviest) ion is always observed, making these techniques valuable for protein identification

via top-down workflows (Figure 3). Modern spectrometry typically uses two techniques for

the soft ionisation of proteins; matrix-assisted laser desorption/ionisation (MALDI) and

electrospray ionisation (ESI).

In a MALDI system, the sample is first mixed with a UV absorbing crystalline matrix to serve

as a proton source for ionisation, as well as to protect the sample from unwanted

fragmentation. The mixture is then placed on a metal plate and vaporised before being

ionised by the gases in the chamber.50 MALDI-MS has been used to successfully identify

protein biomarkers able to identify 25 commercially sold species of fish, outside of the

Scombridae family.51 This demonstrates the technique’s ability to discriminate between

closely related fish species, though MALDI is limited to analysing larger molecules.50 This is

because of the interference caused by the matrix in MALDI analyses, limiting the sensitivity

and specificity of these systems. Additionally, the limited charged species produced by

MALDI-MS (predominately 1+ ions) are unsuitable for fragmentation, a difficulty overcome

by the variety produced by electrospray (2+, 3+, 4+ and other ions).

Electrospray ionisation (ESI) employs an electric field to disperse a liquid analyte into a fine

aerosol. Prior to the application of current, the sample is mixed with a conductivity increasing

solvent which acts as a proton source to facilitate ionisation.52 The range of charge species

produced by ESI allows fragmentation. As a result, electrospray ionisation greatly enables

tandem mass spectrometry (MS/MS), more so than MALDI. This is in part, the reason why

MALDI-MS/MS is not routinely used in proteomic study. Additionally, MALDI cannot be paired

with HPLC, removing the added specificity provided by chromatographic separation.

Page 16: Masters Research Project

16

The technique’s aptitude for speciation has been demonstrated on meats.53 ESI-MS was used

to examine purified haemoglobin and myoglobin sourced from beef, lamb, pork and horse.

Analysis revealed that enough variation was present interspecifically in their amino acid

sequences to make these proteins suitable biomarkers for species identification. This was

observed in the study by the deviation in the weights of myoglobin and haemoglobin between

species. The analytes were also deemed diagnostically viable even when subjected to high

pressure steam at 121°C (autoclaved) for 1 hour. This indicates that ESI-MS on myoglobin and

haemoglobin could be used to characterise cooked and processed products.53

2.2.3.3. Mass Analysers

The second component of a mass spectrometer is the mass analyser. This is where ions are

sorted and separated. Fourier transform ion cyclotron resonance mass spectrometry (FTICR-

MS) is a type of mass spectrometry useful for the high throughput analysis of proteins. The

technique utilizes a Penning trap mass analyser to deduce the mass to charge ratio of ions

based on their movement in a fixed magnetic field (termed ion cyclotron resonance). 54 Due

to the increased stability provided by the superconducting magnet in the mass analyser,

FTICR-MS is able to achieve levels of resolution higher than any other MS instruments, making

it ideal for detailed compositional and structural analyses of large molecules.55 However,

these instruments are among the most expensive tools used for proteomic analysis. In

addition, slow scan speeds are associated with the use of magnetic fields for mass

discrimintation.56

For the purposes of developing a rapid, widely accessible technique, quadrupole based mass

spectrometers are found more suitable. These instruments do not rely on magnetic fields, are

relatively inexpensive and are found in most academic and industrial laboratories.56 Within

quadrupole mass analysers, ions travel between four parallel metal rods. These rods have a

voltage applied to them, producing an oscillating electric field. This electric field alters the

trajectory of the ions. Ions with differing m/z values have varied trajectories: some within a

narrow m/z region are able to pass through the quadrupole and get detected, while others

are forced to collide with the quadrupole.56 Electron multipliers are often used as detectors

in conjunction with quadrupole mass analysers.57 These consist of many dynodes (electrodes

Page 17: Masters Research Project

17

in a vacuum) that create secondary electrons upon impingement to increase signal intensity.

This facilitates the detection of ions at low levels, providing increased sensitivity in MS

analyses.57

Altering the voltage applied to quadrupoles allows the detected mass region to be altered,

allowing selectivity.56 Quadrupoles can also be used to fragment molecular ions via collision

induced dissociation. It is for these reasons that they are utilised in triple quadrupole

instruments to provide additional structural information about molecules in tandem mass

spectrometry.58

2.2.3.4. Tandem Mass Spectrometry

The sensitivity and selectivity of single-stage mass spectrometry is restricted to relatively large

peptide sequences. They therefore remain vulnerable to adulteration via false negatives. The

presence of undeclared substances above 1% is considered actionable gross negligence or

adulteration.37 With the aim of complete authentication in mind, the limit of detection (LOD)

of analytes in a homogeneous sample must be below 1%. The ability of tandem mass

spectrometry (MS/MS) to meet this requirement rapidly in comparison to other

authentication protocols is where its

value lies.42 It is largely for this reason

that the present work utilises this

technology.

Tandem mass spectrometry overcome

the issues associated with MS by the

addition of a fragmentation step. In

MS/MS systems, the initial ionisation

and separation by m/z in the first mass

analyser, is followed by the selection of

ions (termed precursor ions) for

fragmentation. This is carried out by

collision induced dissociation or other

Figure 4. An expanded MS/MS peak. The shown signals were

derived from skipjack myoglobin. The transitions are indicated by

each individual signal. The weights of each peptide fragment

(product ion) are shown by the figure legend in daltons (Da). These

were all fragmented, and therefore represent, the same peptide

(precursor ion) and elute at the same retention time. The peptide

weight is 565.3 Da.

Inte

nsi

ty (

cps

x 1

06)

Retention Time (minutes)

Page 18: Masters Research Project

18

processes, to form product ions. The product ions are then separated by a second mass

analyser, followed by detection.59 A precursor ion and product ion pair is known as a

transition. Each transition produces a unique signal upon its detection, which is

representative of a peptide fragment (Figure 4). This causes MS/MS to yield very high levels

of specificty, as the likelihood of multiple peptides having the same mass and fragments of

the same mass is very small.41 This surpasses single-stage MS analyses which relies solely on

peptide masses for information.

A study by Wulff et al60 demonstrates the value of triple quadrupole tandem mass

spectrometry in authentication. The study uses a bottom-up proteomics approach to build a

database of spectra derived from the MS/MS analysis of whole proteomes. This database

consisted of 22 fish species inculding yellowfin, albacore and skipjack tunas. The comparison

of spectra produced from unknown samples to the database allowed the correct

identification of raw, cooked and processed samples.60 This demonstrates the successful use

of quadrupole instruments in tuna identification. The study also shows that tandem mass

spectrometry remains reliable when testing various product forms, making it robust.

However, the use of whole proteomes instead of selected proteins or peptides, makes this

technique vulnerable to partial substitutions. Spectra produced from a mixture of species

would not correspond to a database consisting of single-species samples. For the proper

authentication of tuna products, this method would not be reliable unless mixed spectra were

produced and included in the database. This would be arduous considering the number of

potential adulterants. The use of whole proteomes also limits the sensitivity, and therefore

the limits of detecion, of analyses.41 By selecting specific markers, partial substitution can be

detected by the presence of markers from more than one species.

Limits of detection and quantitation are improved by the use of selected reaction monitoring

mode (SRM) (Figure 5).39, 41 This43 MS mode determines the presence of peptides based on a

select number of transitions. This reduces interference caused by irrelevant ion signals, which

selected for detection, increasing spectra clarity, run efficiency and detection capabilities.61 It

is important that the reliabilty of transitions is established before disregarding other signals,

to ensure accurate results are obtained. In the context of species identification, the signals

Page 19: Masters Research Project

19

deemed relevant would function as species markers when analysing samples of unknown

origin.

The application of SRM to multiple product ions orignating from the same precursor ion can

further increase the sensitivity of analyses. This variation is referred to as multiple reaction

monitoring (MRM) in the current work.59, 62 It is important to note that the use of this term

varies in the literature, as it is not recognised by IUPAC as a standard term. The strengths

associatied with the use of MRM MS for the analysis of peptides were demonstrated by a

recently developed protocol for meat authentication.63 The adaptation of this study is central

to the work at hand.

MASS ANALYSER (Q1)

Sample

Selection

Selection

Selection

Selection

Scanning Scanning

Scanning

Scanning

Ionisation

COLLISION CELL (Q2)

MASS ANALYSER (Q3)

Figure 5. Tandem-in-space mass spectrometry modes. When tandem-in-space, two coupled mass analysers are separated

by spatially distinct controlled fractionation. In triple quadrupole instruments these processes take place in three

quadrupoles. The pathways above indicates the four experiments primarily conducted with tandem-in-space MS. In product

ion scans (red), the precursor ion is specified in the first mass analyser (Q1) and the resultant product ions from fragmentation

(in Q2) are scanned then detected after the second mass analyser (Q3). Precursor ion scans (blue) involve the selection of

the product ion in Q3 and the detection of precursor ion masses in Q1. In selected reaction monitoring (yellow), both Q1 and

Q3 are set to a specific mass. In neutral loss scanning (green), mass analysers are set to scan all ions at an offset to account

for the observed neutral loss typically observed in the analyte.43

ION SOURCE

Page 20: Masters Research Project

20

2.3. THE PRESENT WORK

A method developed by Watson et al. 63 could be effectively applied to tuna speciation if

modified. The study successfully used an intelligence-led method to identify horse, pork, beef

and lamb via MRM MS. The approach differs from previous research by using the peptides

derived from one protein to determine species identity, as opposed to the sets of proteins

used in conventional shotgun strategies. Myoglobin (Mb) was exploited due to the presence

of inter-specific variation in Mb amino acid sequences.63 Enzymatic digestion of the Mb

enabled the production of characteristic peptide sets detectable by LC-MRM MS. This use of

corresponding proteins for speciation via the analysis of representative peptide sets is

referred to as CPCP for brevity. Corresponding proteins are nominally similar proteins with

different species of origin. 63

This method is an attractive candidate for application to tuna products for a number of

reasons. The original study demonstrated that species identification via this approach is more

rapid than currently-used PCR-based approaches. An issue common to DNA-based methods

is the time required for analysis, as extraction to final testing can take a number of days at

best.42 This can be surpassed by MS analyses, particularly when analysing nominally similar

proteins as markers. The use corresponding proteins for characterisation simplifies

preparation, run times and data handling. The efficiency of analysis is further enhanced by

the multiplex capabilities of MRM MS. A multiplex method is one that is able to detect

multiple analytes in one test. This attribute allows the presence of many species in a sample

to be determined rapidly, which is especially useful when considering the complexity of the

tuna industry.

The anaylsis of peptide fragments and the peptides they are derived from is also important

to tuna anaylsis. For this tandem mass spectrometry is required. As many of the principal

market tunas primarily belong to the same genus, the differences between the structures of

their proteins are likely to be few. This will result in a small number of peptides suitable for

species identification upon digestion. In order to reliably identify substitutions of these closely

related species, the detection of multiple transitions (as in MRM MS) is a necessity to provide

high sensitivity and specificity.

Page 21: Masters Research Project

21

These qualities are also important to ensure that low level adulterations can be detected. Due

to time constraints, quantitation will not form part of the present study on tuna. However,

given the wider context of food authentication, it is beneficial to consider a technique likely

to be able to distinguish between tuna species in a homogenous mixture, for the puposes of

future research. It is also important to be able to determine the amount of the adulterant

present in these mixtures relative to the total product. This is referred to as relative

quantitation. Preliminary investigation of the CPCP approach indicates that this is within its

capabilities. When applied to red meats, levels of relative quantitation as low as 1% in various

horse, beef, pork and lamb mixtures were achieved.63 This supports the proposition that the

CPCP approach is able to effectively identify the species present in a mixture of tuna species,

using the peptides derived from corresponding proteins. This is contingent only on the

presence of interspecific variation in the amino acid sequence of said proteins, which can be

determined using an online database (such as UniProt). The likely transitions to arise from the

digestion of proteins of interest can also be predicted by software (such as Open Source

Skyline). This removes the labour intensive process of identifying sets of species-specific

proteins, as required by high sensitivity identifications using whole proteomes (as in shotgun

proteomics).42

An ideal technique for tuna species identification should also be versatile to reflect the

numerous forms in which tuna is commercially sold. To reflect this the technique would

remain robust when testing raw, cooked or heavily processed products, such as canned tuna.

Again, initial testing of the CPCP approach on these product forms is encouraging as peptide

markers for the identification of red meats were successfully recorded.63

Prior to analysis, it is first important to select an appropriate protein as a source of peptide

marker sets. As stated previously, it is necessary to select a relatively heat stable protein for

tuna authentication, to avoid the excessive thermal degradation of the analyte during

processing. Cytochrome b, a mitochondrial protein used in the electron transport chain in

aerobic respiration, initially seems like an intuitive candidate. This is due to the freuquent use

of its coding gene, MT-CYB, in the DNA-based identification of fish species. However, initial

comparison amongst principal market tunas indicates that corresponding cytochrome b

Page 22: Masters Research Project

22

proteins offer no discriminatory value, unlike its highly variable coding gene. This is due to a

lack of interspecific variation in their amino acid sequences, revealed by using the BLAST tool

on the UniProt database.64

The use of myoglobin could be beneficial, as it was in the original study by Watson et al.63 This

protein facilitates the movement of oxygen within muscle tissue, and is also responsible for

the pink colourationg of certain meats. The amino acid sequence of the most easily sourced

tunas, yellowfin, albacore, skipjack and bigeye vary between these species, allowing their

characterisation via via CPCP (Figure 6).

65 Study of myoglobin concentrations in both light (0.37-1.28mg/g) and dark (5.3-24.4mg/g)

yellowfin muscle tissue indicates that Mb abundance will be adequate for analyses in species

regardless of their flesh colouring.66 This is an important observation in the case of albacore,

the white tuna. Additionally the pigmentation, location and water solubility of Mb, make

them an attractive candidate for use in rapid analysis, as all of these properties simplify

preparatory procedure.63,167 The thermostability of various Scrombidae Mb has also been

explored. Investigation revealed that fish Mb, though more prone to degradation than horse

Mb, are still relatively thermostable, suggesting that the peptides of interest would survive

the retort stages of canned tuna production.68 There is also added value in using the same

protein as the previous study of the CPCP approach, to explore the versatility of Mb whilst

keeping already optimised preparation protocols universal between products. It is for these

reasons that myoglobin was selected as the protein of interest in this investigation.

P02205|ADFDAVLK*CWGPVEADYTTMGGLVLTR*LFK*EHPETQK*LFPK*FAGIAQADIAGNAAISAHGATVLK*K*LGELLK*A

Q9DGJ2|ADFDAVLK*CWGPVEADYTTIGGLVLTR*LFK*EHPDTQK*LFPK*FAGIAQADLAGNAAISAHGATVLK*K*LGELLK*A

Q9DGI8|ADLDAVLK*CWGAVEADFNTVGGLVLAR*LFK*DHPETQK*LFPK*FAGIT-GDIAGNAAVAAHGATVLK*K*LGELLK*A

Q76G09|ADFDAVLK*CWGPVEADYTTIGGLVLTR*LFK*EHPETQK*LFPK*FAGIAQADIAGNAAVSAHGATVLK*K*LGELLK*A

P02205|K*GSHAAILK*PLANSHATK*HK*IPINNFK*LISEVLVK*VMHEK*AGLDAGGQTALR*NVMGIIIADLEANYK*ELGFSG

Q9DGJ2|K*GSHASILK*PMANSHATK*HK*IPINNFK*LISEVLVK*VMQEK*AGLDAGGQTALR*NVMGIIIADLEANYK*ELGFTG

Q9DGI8|K*GNHAAIIK*PLANSHAK*QHK*IPINNFK*LITEALAHVLHEK**AGLDAAGQTALR*NVMGIVIADLEANYK*ELGFTG

Q76G09|K*GSHAAILK*PLANSHATK*HK*IPINNFK*LISEVLVK*VMHEK*AGLDAGGQTALR*NVMGIIIADLEANYK*ELGFSG

Figure 6. Corresponding Tuna Mb Amino Acid Sequences. The amino acid sequences of (from top to bottom) yellowfin Mb

(P02205), albacore Mb (Q9DGJ2), skipjack Mb (Q9DGI8) and bigeye Mb (Q76G09). Cleavage sites arising from tryptic digestion

are marked by “*”. The resultant peptides unique to one species are indicated by blue (yellowfin), green (albacore), red

(skipjack), or purple (bigeye) text. Deviations from the ancestral amino acid sequence are marked in yellow. Amino acid

sequences were obtained and aligned using UniProt database.65 Methionine from position 1 has been removed to reflect its

cleavage from most mature proteins.63

Page 23: Masters Research Project

23

With this in mind, the protocols used for sample preparation were largely in accordance with

those used in Watson et al.63 Two key preparatory stages are an extraction process and an

additional purification process used to produce purified myoglobin. Extraction is a necessary

preparatory step in proteomic analysis, to separate the protein and non-protein parts of a

sample. This ensures that other cellular components, known as particulate matter, do not

contaminate the results. In addition, some particulate matter is inhibits the function of the

instrumentation and must therefore be removed prior to LC-MS/MS.69

As peptides and transitions in this approach are determined theoretically as opposed to

experimentally, it is important to establish their visibilty prior to high throughput analyses on

complex samples. For this reason, prior to enzymatic digestion (termed proteolysis), an

additional purification step was employed to isolate myoglobin from the other proteins in

certain samples. This reduces the intereference caused by other peptides within an unpurified

mixture. Useful peptides were then selected as markers in these purified samples, based on

the visibility of their transitions and their discrimnatory power. A peptide that is present in

only one species offers the most discriminatory power as a species marker, whilst one that

appears in many offers little. A multiplex method capable of scanning for the presence of all

species at once can then be created by compiling these markers. As this purification is

labourious and time consuming, once the presence of markers has been established in the

purified Mb, this purification step is removed. Marker presence can then be tested in more

complex extracts from raw tunas, and commercially canned tunas. This aligns with the

intended purpose of this technique: to rapidly identify the species within a tuna product.

Again, for the purposes of establishing peptide presence, purified samples were initally

analysed using unscheduled monitoring. This is a mode in MRM MS where ion signals are

scanned for at all retention times. Once the retention times of the selected peptides have

been determined, samples can be tested using scheduled monitoring. This MRM MS mode

only scans for ion signals at specified retention times, further increasing the clarity of results

by removing background at other retention times.

Page 24: Masters Research Project

24

2.3.1. Objectives & Hypotheses

To better understand the versatility of the CPCP method, and assess the ability of the

approach to combat tuna fraud, the aims and objectives of this investigation were as follows:

1. To establish whether the CPCP approach can be used to distinguish between various

species of tuna myoglobin using MRM MS:

H1 – An unscheduled MRM method searching for predicted transitions in purified

myoglobin yields signals that do not differ sufficiently between tuna species. The

results of this analysis could not be used to select markers for a multiplex method

searching for signals at scheduled retention times. The CPCP approach cannot be used

to distinguish between various species of tuna via MRM MS.

2. To establish whether the CPCP approach can be used to rapidly distinguish between

various species of tuna myoglobin using multiplex MRM MS:

H2 – A multiplex MRM method performed on purified myoglobin produces signals that

do not differ sufficiently between tuna species at schedules retention times (Rt). The

CPCP approach cannot be used to rapidly distinguish between various species of tuna

via MRM MS.

3. To establish whether the CPCP approach can be used to rapidly distinguish between

various species of tuna myoglobin in raw tuna samples using MRM MS:

H3 – A multiplex MRM method performed on raw tuna extracts produces signals that

do not differ sufficiently between tuna species at scheduled retention times (Rt). The

CPCP apprach cannot be used to rapidly distinguish between various species of tuna in

readily available tuna samples via MRM MS.

Page 25: Masters Research Project

25

4. To establish the whether the CPCP approach can be used to rapidly distinguish between

various species of tuna myoglobin in extensively processed samples using MRM MS:

H4 – A multiplex MRM method performed on canned tuna extracts produces signals

that do not differ sufficiently between tuna species at scheduled retention times (Rt).

The CPCP apprach cannot be used to rapidly distinguish between various species of

tuna in extensively processsed tuna samples via MRM MS.

3. METHODOLOGY

The detailed workflow and protocols are based upon previous work by Watson et al.63

3.1. Materials

Methanol and acetonitrile were purchased from Fisher Scientific (Loughborough, UK). Urea,

trifluoroethanol and trypsin (from bovine pancreas, TYPCK treated) were purchased from

Sigma-Aldrich (Gillingham, UK) and formic acid from BDH Chemicals (Poole, UK).

Both raw and processed yellowfin (Thunnus albacares), albacore (Thunnus alalunga) and

skipjack (Katsuwonus pelamis) were sourced from local supermarkets or an online fishmonger

(King Catch, Reading, UK). Due to sourcing difficulty, only raw bigeye (Thunnus obesus) was

obtained for study. Prior to preparation, raw tuna was stored at -40 °C. Processed tuna

remained within cans at room temperature. Prior to the production of purified reference

myoglobin and raw sample preparation raw tuna meat was diced then frozen in liquid

nitrogen. Samples were then ground using a Waring blender and stored at -40 °C.

3.2. Purified Myoglobin Preparation

Ground tuna samples (30 g) were extracted at 4 °C in 30 mM TRIS/HCl (pH 8.4). The extract

then underwent gravity filtration and the filtrate was centrifuged (4 °C, 30 mins 14000 g). The

supernatant was then filtered through syringe filters of decreasing size (0.8 μm, 0.45 μm and

0.2 μm). Sodium azide (0.02%) and potassium sorbate were added (1 mg/ml) and then the

sample was syringe filtered (0.2 μm).

Page 26: Masters Research Project

26

A 2 mL aliquot of crude extract was then loaded onto a Superdex 75 size-exclusion column

(GE Healthcare, Chalfont, UK) attached to a BioCad Sprint HPLC system (Applied Biosystems)

for FPLC separation. This was equilibrated with 100 mM TRIS/HCl (pH 8.4) at a flow rate 1

mL/min with a fixed absorbance at 280nm. Resultant fractions were then analysed by SDS-

PAGE to identify those that contain myoglobin (Mb) by weight (circa 16kDa). Selected

fractions were then pooled and stored at 4°C. The purified samples were then loaded onto a

PD10 column (GE Healthcare) and eluted with 25mM ammonium bicarbonate prior to

proteolysis.

3.3. Raw Sample Preparation

The phosphate extraction buffer (4 mL of 0.3 M KCl, 0.3 M phosphate buffer at pH 6.5) was

added to 300 mg of ground sample and then vortexed (30 s). The sample then underwent

extraction at room temperature for 2 hrs on a Stuart SB3 tube rotator at 20 rpm. Aliquots of

sample (2 mL) were then centrifuged (5 min, 4°C, 17000 g). 200 μL of the supernatant were

then evaporated to dryness at 50°C, using a Jouan RC 1022 centrifugal evaporator.

3.4. Processed Sample Preparation

Cans of processed sample were drained of liquid and 20 g of sample weighed into a plastic

beaker. The phosphate extraction buffer (100 mL of 0.3 M KCl, 0.3 M phosphate buffer at pH

6.5) was added. The sample then underwent extraction by blending for 1 min using an Ultra

Turrax blender. Aliquots of the extract (2mL) were then centrifuged (5 min, 4°C, 17000 g) and

200 μL of the supernatant evaporated to dryness at 50°C using a Jouan RC 1022 centrifugal

evaporator.

3.5. Proteolysis

The dried samples, either raw or processed, were dissolved in 25 mM ammonium bicarbonate

prior to the rest of proteolysis. This step is the final difference in the treatments of the purified

myoglobin, the raw samples and the processed samples. Therefore, for brevity, all shall be

referred to as “the samples” or similar from this point forward.

Page 27: Masters Research Project

27

The samples were heated to 95°C for 30 mins, then cooled. Urea was added to a final

concentration of 0.5M. Trypsin solution (1mg/ mL) was added in a 1:30 enzyme to substrate

weight ratio. The samples were then vortexed and digested overnight at 37°C.

The digests were diluted 1:2 with water and desalted using a Strata-X 33μ reversed phase

cartridge (Phenomenex, Macclesfield, UK). The cartridge was washed then activated using 1

mL of methanol then equilibrated with 1 mL of 1% formic acid. The cartridge was loaded with

the diluted sample and washed with 1 mL 5% methanol/1% formic acid. Peptides were then

eluted by adding 1 mL acetonitrile/water (90:10; 0.1% formic acid). The samples were

evaporated to dryness using the centrifugal evaporator then dissolved in 250 μL

acetonitrile/water (3:97; 0.1% formic acid).

3.6. Predictive Tools

The likely peptides resulting from the tryptic digestion of yellowfin (Thunnus albacares),

albacore (Thunnus alalunga), skipjack (Katsuwonus pelamis) and bigeye (Thunnus obesus) Mb

were predicted using the PeptideCutter tool 70 set to the ‘simple model’ option. Potential

transitions produced by fragmentation during MRM-MS were then predicted using Open

Source Skyline.71 Skyline setting included the following: 0 missed cleavages, conventional

trypsin rules [KR|P], no modifications, exclude N-terminal, amino acids set to 1, No cysteine

modifications, amino acids set to 1 (to avoid initiator methionine) peptide length 6 – 50,

collision energy of ‘ABI 4000 Q Trap’.

3.7. LC-MS/MS Workflow

Skyline-predicted transitions were used to create one MS/MS method file for each tuna

species. The presence of these transitions was then determined. This was accomplished by

testing each species of purified Mb with its own method in unscheduled monitoring mode.

The peptides useful as markers were then selected based on signal quality and species

specificity. The retention times and four most intense transitions (based on peak height) for

all selected peptides were then recorded. These were used to create a single multiplex MS/MS

Page 28: Masters Research Project

28

methods file for the detection of all species at once in scheduled monitoring mode (scan time:

2 s, nominal retention window: ±50 s). New purified Mb, raw samples and processed samples

were then analysed with this multiplex method.

3.8. LC-MS/MS Analyses

Analysis was carried out via high performance liquid chromatography coupled with tandem

mass spectrometry (HPLC-MS/MS). This was conducted on an Agilent 1200 Rapid Resolution

LC system (Stockport, UK) connected to an AB Sciex 4000 QTrap triple quadrupole mass

spectrometer (Warrington, UK). A Phenomenex XB C18 reversed-phase column (100 x 2.1mm,

2.6 μ particle size, Macclesfield, UK) using a 300 μl/min flow rate at 40°C was used for

chromatographic separation. The binary gradient profile consisted of 97% A (water and 0.1%

formic acid) and 3% to 28.4% B (acetonitrile and 0.1% formic acid) to 28.4%. Injection volume

was 10 μl. Positive mode electrospray was used to detect eluted peptides using a

TurboIonSpray probe (AB Sciex) ion source. Ion source settings were a curtain gas of 25 psi,

desolvation gas (GSI) of 50 psi, and sheath gas (GS2) of 20 psi. Source temperature was 550°C.

Data was acquired on Analyst 1.6.2 software (AB Sciex) as selected-reaction monitoring

chromatograms. These were then converted into total ion current chromatograms, displaying

the sum of all signals detected at any given retention time for illustrative purposes.

Page 29: Masters Research Project

29

4. RESULTS

As this study did not involve quantitation or limits of detection study, the relative intensities

of ion signals were merely used as an indicator of peak visibility. Only the presence of a

precursor and product ion pair (transition) can be determined using these chromatograms.

No relationships can be accurately drawn between the area underneath each peak and the

quantity of an analyte in this dataset.

4.1. Peptide Marker Selection

The targeted search of the four myoglobin for the predicted transitions revealed a number of

clearly defined signals (Figure 7). This established that the peptides and their fragments within

the digests of the purified myoglobin were detectable by MRM mass spectrometry.

0

6

12

18

24

30

0 10 20 30 40

Inte

nsi

ty (

cps

x 1

06)

0

3

6

9

12

0 10 20 30 40

0

0.2

0.4

0.6

0.8

1

0

4

8

12

16

B

C

A

Retention Time (minutes)

D

858.5

565.3

450.8

336.2 423.2

832.4

1035.0

858.5

336.2 423.2

565.3

450.8 832.4

821.5 1127.1

825.4

737.4

572.3

336.2 423.2

565.3 450.8

336.2 423.2

832.4

875.5

Figure 7. Total ion current chromatograms resulting from the unscheduled screening of purified yellowfin (A), albacore (B),

bigeye (C) and skipjack (D) myoglobin. Each ‘peak’ is the sum of all signals at any given retention time. The numbers indicate

the m/z values of the peptides observed at that retention time. These can be common to more than one species (black) or

unique to one species (green). It is important to note that not all observed peaks are visible in this format without

magnification. Only the most visible or the most unique peptides are labelled here.

Page 30: Masters Research Project

30

21 peptides associated with these signals were deemed useful as markers for the

differentiation of the tuna species examined, and were selected for use in a scheduled

multiplex method for further analyses (Table 1).

Among the selected peptides, at least one was detected in each sample that was unique to

that species. These peptides offer the most discriminatory power, though the peptides unique

to albacore and bigeye produced ion signals with very low intensities. As a result it was

necessary to retain clearly visible peptides with less discriminatory power in future methods,

to increase confidence in the species identification of a sample.

It was therefore established that the signals produced by the unscheduled screening of

purified myoglobin can be used to distinguish between the analysed species of tuna, as the

differences in these ion signals allowed effective species discrimination.

Page 31: Masters Research Project

31

Ta

ble

1

Th

e d

etec

tio

n o

f tu

na

myo

glo

bin

pep

tid

es s

elec

ted

as

mar

kers

in M

RM

-MS.

Th

e p

recu

rso

r io

n (

Pep

tid

e) a

nd

pro

du

ct io

n (

Frag

men

ts)

m/z

val

ue

s ar

e s

ho

wn

to

1 d

ecim

al p

lace

fo

r

clar

ity.

Pep

tid

es w

ere

cho

sen

bas

ed o

n d

iscr

imin

ato

ry p

ow

er a

nd

vis

ibili

ty. S

pec

ies

are

cod

ed a

s YF

= y

ello

wfi

n, A

L =

alb

aco

re, S

J =

skip

jack

an

d B

E =

big

eye.

Th

e s

pec

ies

of

ori

gin

(re

ferr

ed t

o a

s

nat

ive

spec

ies

for

bre

vity

) of e

ach

pe

pti

de

is s

ho

wn

in t

he

“sp

ecie

s” c

olu

mn

. Pe

pti

de

s n

ativ

e to

mo

re s

pec

ies

hav

e le

ss d

iscr

imin

ato

ry p

ow

er. T

he

am

ino

aci

d s

equ

ence

of

each

pe

pti

de,

ob

tain

ed

fro

m S

kylin

e s

oft

war

e, a

re s

ho

wn

in s

tan

dar

d o

ne

lett

er c

od

ing.

Th

e f

ou

r m

ost

inte

nse

fra

gmen

ts w

ere

sele

cted

bas

ed

on

pea

k h

eigh

t to

rep

rese

nt

each

pep

tid

e. T

he

pe

pti

de

and

frag

men

t m

/z

valu

es, a

s w

ell a

s th

eir

rete

nti

on

tim

es (

Rt)

wer

e u

sed

to

mak

e a

sch

edu

led

rea

ctio

n m

on

ito

rin

g m

eth

od

fo

r fu

rth

er a

nal

yses

. Th

e o

bse

rved

det

ecti

on

of

pep

tid

es in

res

ult

s o

f th

e u

nsc

hed

ule

d

(A)

and

sch

edu

led

(B

) an

alys

is o

f p

uri

fie

d m

yogl

ob

in,

and

th

e sc

he

du

led

an

alys

is o

f ra

w t

un

a ex

trac

t (C

) is

sh

ow

n. D

etec

tio

n i

n n

ativ

e (g

reen

) an

d o

ther

(b

lue

) sp

ecie

s is

ind

icat

ed

by

the

fille

d

circ

les.

Lac

k o

f p

ep

tid

e d

etec

tio

n o

f in

nat

ive

spec

ies

is in

dic

ated

by

a re

d f

illed

cir

cle.

Pep

tid

e

(m/z

) Se

qu

ence

Fr

agm

ents

(m

/z)

Spe

cies

R

t A

B

C

YF

AL

SJ

BE

YF

AL

SJ

BE

YF

AL

SJ

BE

305.

1

ELG

FSG

16

3.1,

31

0.1,

36

7.2,

480

.2

YF, B

E

13.5

312.

2

ELG

FTG

17

7.1,

76.

0, 3

81.2

, 324

.2

AL,

SJ

14.1

336.

7

LGEL

LK

147.

1, 2

60.

2, 5

02.

3, 5

59.3

YF

, AL,

SJ,

BE

12

.4

422.

7

AD

LDA

VLK

65

8.4,

43

0.3,

54

5.3,

147

.1

SJ

10.5

423.

2

IPIN

NFK

40

8.2,

52

2.3,

63

5.4,

732

.4

YF, A

L, S

J, B

E

12.4

427.

7

EHP

DTQ

K

376.

2, 2

75.

2, 5

05.

3, 1

47.1

A

L, S

J 12

.5

450.

8

LISE

VLV

K

787.

5, 6

74.

4, 4

58.

3, 5

87.4

YF

, AL,

BE

15

.1

565.

3

GLD

AG

GQ

TALR

88

8.5,

77

3.4,

70

2.4,

645

.4

YF, A

L, B

E

11.3

572.

3

GLD

AA

GQ

TALR

64

5.4,

71

6.4,

90

2.5,

787

.4

SJ

12.1

737.

4

LITE

ALA

HV

LHEK

83

3.5,

12

47.7

, 762

.4, 4

13.

2

SJ

18.4

821.

5

GN

HA

AII

KLP

LAN

SHA

K

837.

5, 1

078

.6, 9

65.6

, 12

62.8

SJ

8.

4

825.

4

NV

MG

IVIA

DLE

AN

YK

1135

.6, 1

036.

5, 9

23.4

, 130

5.7

SJ

26

.1

832.

4

NV

MG

IIIA

DLE

AN

YK

1149

.6, 1

319.

7, 1

036.

5, 9

23.4

YF

, AL,

BE

26

.8

858.

5

GSH

AA

ILK

PLA

NSH

ATK

10

66.6

, 129

2.8

, 117

9.7,

938

.5

YF, B

E

9.3

875.

5

GSH

ASI

LKP

MA

NSH

ATK

95

6.5,

10

84.6

, 119

7.6

, 139

7

AL

8.4

989.

5

CW

GA

VEA

DFN

TVG

GLV

LAR

78

4.5,

35

9.2,

99

9.6,

458

.3

SJ

26.1

1026

.0

CW

GP

VEA

DYT

TIG

GLV

LTR

10

30.6

, 389

.3, 7

15.5

, 92

9.6

A

L, B

E 24

.1

1035

.0

CW

GP

VEA

DYT

TMG

GV

LTR

71

5.5,

84

6.5,

10

48.6

, 65

8.4

YF

25

.5

1063

.1

FAG

ITG

DIA

GN

AA

VA

AH

GA

TVLK

13

50.8

, 867

.5, 1

279.

7, 1

037.

6

SJ

19.3

1127

.1

FAG

IAQ

AD

IAG

NA

AV

SAH

GA

TVLK

13

66.7

, 129

5.7

, 883

.5, 5

88.4

B

E 18

.1

1134

.1

FAG

IAQ

AD

IAG

NA

AIS

AH

GA

TVLK

13

09.7

, 138

0.8

, 883

.5, 3

59.3

YF

, AL

18.7

Page 32: Masters Research Project

32

4.2. Multiplex Myoglobin Analyses

During multiplex analysis, most of the selected peptides and their fragments produced ion

signals at the scheduled retention times (Figure 8). An exception to this was the fragments of

875.5, the peptide specific to albacore. This implies that this peptide is not a reliable species

marker.

Additionally, some ion signals did not always occur in the expected species, as the transitions

from previously injected samples were detected in the next samples (Table 1). This caused

the species identity of a sample to be unclear, due to a loss of peptide specificity. It was

established that this was not due to contamination prior to LC-MS/MS analysis via the by

testing a sample known to contain no peptides, referred to as a blank sample (Figure 9). After

analysis, peptides were observed in the blank, establishing that the contamination occurred

within the LC-MS/MS system.

0

4

8

12

16

0 5 10 15 20 25 30 35 40

0

3

6

9

12

15

0 5 10 15 20 25 30 35 40

0

0.2

0.4

0.6

0.8

1

0

4

8

12

16

Inte

nsi

ty (

cps

x 1

06)

B

C

A

Retention Time (minutes) (minutes)

D

858.5

565.3

450.8

336.2 423.2

832.4

1035.0

858.5

336.2 423.2

565.3

450.8 832.4

821.5 1127.1

825.4

737.4

572.3 336.2 423.2

565.3

450.8

336.2 423.2

832.4 875.5

Figure 8. Total ion current chromatograms resulting from the multiplex analysis of purified yellowfin (A), albacore (B), bigeye

(C) and skipjack (D) myoglobin. Each ‘peak’ is the sum of all signals at any given retention time. The numbers indicate the m/z

values of the peptides observed at that retention time. These can be common to more than one species (black) or specific to

one species (green). Certain peptides (grey) were not detected during analysis. It is important to note that not all observed

peaks are visible in this format without magnification. Only the most visible or the most unique peptides are labelled here.

Page 33: Masters Research Project

33

These results indicate that the CPCP approach could not be used to rapidly distinguish

between the tested tuna species, due to the presence of within system contamination in this

dataset.

0

0.005

0.01

0.015

0.02

0 5 10 15 20 25 30 35 40

YAB (450.8) YASB (336.7) YASB (423.2)

S (565.3)

YAB (832.4)

S (989.5)

Figure 9. Total ion current chromatograms resulting from the multiplex analysis of ammonium bicarbonate. This was run

after multiplex myoglobin analyses. Each ‘peak’ is the sum of all signals at any given retention time. The species of origin of

each peptide are coded as YF = yellowfin, AL = albacore, skipjack = SJ and bigeye = BE. The numbers indicate the m/z values of

the peptides observed at that retention time. The presence of peptides in this sample indicates that peptides are retained

within the LC-MS/MS system. It is important to note that not all observed peaks are visible in this format without magnification

Only the most visible peptides are labelled here.

Inte

nsi

ty (

cps

x 1

06)

Retention Time (minutes)

Page 34: Masters Research Project

34

4.3. Multiplex Raw Tuna Analyses

Again, the scheduled multiplex analysis of many of the selected peptides and their fragments were

detected as clear ion signals at the specified retention times. More background and noise was

observed in these chromatograms (Figure 10). Therefore, it can be established that the selected

myoglobin derived peptides are detectable in raw tuna extracts. No fragments of the peptides 1035.0

or 989.5 were detected during the analysis of these samples (Table 1). This indicates that these

peptides are not reliable species markers

Some transitions from previously injected samples were again detected in the next samples (Table 1).

This was confirmed to be caused within LC-MS/MS using a blank sample (Figure 11).

The present work indicates that the CPCP approach cannot not be used to rapidly distinguish between

the species tested in raw tuna, due to the lack of specificity detectable by MRM-MS analysis in this

dataset.

0

1

2

3

4

5

0 5 10 15 20 25 30 35 40

0

1

2

3

4

5

6

0 5 10 15 20 25 30 35 40

0

1

2

3

4

5

0

0.2

0.4

0.6

0.8

1

Inte

nsi

ty (

cps

x 1

06)

B

C

A

Retention Time (minutes)

D

858.5

565.3 450.8

336.2 423.2

832.4 1035.0

858.5

336.2 423.2

565.3 450.8

832.4

821.5 1127.1 825.4

737.4

572.3

336.2 423.2

565.3

450.8

336.2 423.2

832.4 875.5 737.4

565.3

Figure 10. Total ion current chromatograms resulting from the multiplex analysis of raw yellowfin (A), albacore (B), bigeye

(C) and skipjack (D) samples. Each ‘peak’ is the sum of all signals at any given retention time. The numbers indicate the m/z

values of the peptides observed at that retention time. These can be common to more than one species (black) or specific to

one species (green). Certain peptides (grey) were not detected during analysis, whilst some peptides were observed in the

wrong species (red). It is important to note that not all observed peaks are visible in this format without magnification. Only

the most visible or the most unique peptides are labelled here.

Page 35: Masters Research Project

35

Multiplex Processed Tuna Analyses

None of the selected peptides were detected in any of the 9 samples of canned tuna in either

scheduled or unscheduled monitoring mode (Figure 12). This indicates that the analysis of myoglobin-

derived peptides using the CPCP approach cannot be used to distinguish between the species present

in heavily processed tuna using MRM analysis.

0

0.004

0.008

0.012

0.016

0.02

0 10 20 30 400

0.003

0.006

0.009

0.012

0.015

0 10 20 30 40

0

0.005

0.01

0.015

0.02

0.025

0 10 20 30 40

A B C

Inte

nsi

ty (

cps

x 1

06)

Retention Time (minutes) Figure 12. Total ion current chromatograms resulting from the multiplex analysis of canned yellowfin (A), albacore (B) and

skipjack (C) samples. Each ‘peak’ is the sum of all signals at any given retention time. It was not possible to associate signals to

any peptides no signals were observed that were discernible from background.

0

0.03

0.06

0.09

0.12

0.15

0 5 10 15 20 25 30 35 40

Inte

nsi

ty (

cps

x 1

06)

Retention Time (minutes)

YAB (450.8)

YASB (336.7, 423.2) S (565.3)

YAB (832.4)

S (989.5)

Figure 11. Total ion current chromatograms resulting from the multiplex analysis of ammonium bicarbonate. This was run

after multiplex raw tuna analyses. Each ‘peak’ is the sum of all signals at any given retention time. The species of origin of each

peptide are coded as YF = yellowfin, AL = albacore, skipjack = SJ and bigeye = BE. The numbers indicate the m/z values of the

peptides observed at that retention time. The presence of peptides in this sample indicates that peptides are retained within

the LC-MS/MS system. It is important to note that not all observed peaks are visible in this format without magnification. Only

the most visible or the most unique peptides are labelled here.

Page 36: Masters Research Project

36

5. DISCUSSION

5.1. Marker Reliability

The majority of markers selected were detected in the scheduled analysis of both purified

samples and raw tuna samples. This demonstrates that marker detection was largely

reproducible and therefore reliable. The detection of peptides in the raw tuna samples also

demonstrates the versatility of this approach. These are encouraging observations. The

largest distinction of the CPCP approach from previous work, is the minimal experimentation

required for marker identification. Laborious processes are usually used to select species-

specific markers, particularly in crude extracts, where it is difficult to find visible markers. This

is because peptides are not all equally detectable in proteomic analyses and additional

proteins can mask the signals produced by peptides of interest. Additional challenges are that

some peptides are only present with post-translational modifications (PTMs), or that

cleavages can be missed during tryptic digestion. 72, 73

The latter occurs when trypsin fails to cleave (i.e. break apart) the amino acid sequence of a

protein at an expected residue (i.e. an amino acid within a protein). This can lead to false

negatives caused by a shift in the molecular mass of uncleaved peptides.73 To be thorough,

missed cleavages were searched for initially in all purified myoglobin, though none were

detected. This is likely due to the optimised proteolysis procedure involving heat

denaturation and the addition of urea, both promotes complete digestion.63

PTMs are covalent changes made to the primary structure of a protein, often via the

introduction of new functional groups to it’s amino acid residues. These can lead to a shift in

the molecular mass of peptides, causing them to have unexpected m/z values. If unmodified

m/z values were used to search for modified peptides, this would result in false negatives.

This can be problematic in proteomic study, due to the large number of PTMs possible.72 The

occurrence of certain modifications should be considered for future analyses on different

samples. The presence of high levels of methionine in tuna is noteworthy.74 The sulfur atom

of this residue is prone to oxidation, forming methionine sulfoxide.72 Methionine is found at

the start of all amino acid sequences (at the N-terminal) to initiate protein synthesis (termed

Page 37: Masters Research Project

37

translation). It is often cleaved from the mature protein after synthesis to increase stability

and enable proper function.75 As a result only four methionine residues remain in the

selected markers and do not appear to be a source of error in the present work.

However, four markers were not consistently observed in all samples, implying that they were

not reliable (Table 1). The 1026.0 peptide was not detected in albacore in any sample, though

it was consistently observed in bigeye. This could be due to the presence of a cysteine residue

in this peptide, as cysteine is prone to modification via the formation of disulfide bonds,

bridging two peptides to form a dimer.74 The reactive SH in cysteine makes these disulfide

bridges commonplace, though methionine cannot form disulfide bridges as its sulfur atom is

in a more stable thioester (R-S-CH3) group. The consistent lack of detection of 1026.0 in

albacore compared to its consistent detection in bigeye is of interest. It could imply that

albacore Mb are more prone to this modification than bigeye Mb, however this could also be

due to a difference in Mb concentrations between the species. The inconsistent detection of

the other cysteine-containing markers (989.5 and 1035.0) supports the explanation that

disulfide bridges interfered with analyses. To reduce the occurrence of disulfide bridges, an

additional reduction step should be included prior to proteolysis using tris(2-

carboxyethyl)phosphine (TCEP) or a similar reducing agent.76 This will facilitate the digestion

of Mb via trypsin by reducing any disulfide bonds prior to proteolysis. Reduction should be

followed up with an alkylation step with iodoacetamide to inhibit the formation of any

disulfide bonds by modifying any free sulfur groups (SH).76 The MRM-MS method file would

have to be adjusted accordingly to include these peptides containing modified cysteine

residues.

The presence of the remaining unreliable markers deviated between samples. The

mechanisms causing their lack of detection are likely to be dependent on the form of the

sample. They will therefore be discussed separately for clarity.

5.1.1. Purified Myoglobin

As they were both carried out on purified Mb samples, it seems counterintuitive that

transitions detected in unscheduled analysis were not detected during multiplex analysis.

Page 38: Masters Research Project

38

Specifically, these missing transitions were all associated with the peptide uniquely found in

albacore (875.5). It is possible that this occurred due to carryover effects, the likely cause of

within system contamination. This is an issue of great complexity, and will therefore be

expanded upon at the end of this discussion. If marker unreliability was caused by carryover,

detection of these markers would be expected in subsequent samples. This was not observed,

making it unlikely that carryover was a cause of marker absence.

It is important that this peptide can be reliably detected, as it creates the specificity needed

for albacore identification. This is made more difficult by the fact that, when present, these

transitions produced ion signals with very low intensities. This implies they are only present

in low amounts within albacore. It is therefore likely that this peptide was lost due to human

error during the purification of crude extract. The isolation of myoglobin from a crude extract

is a lengthy process, with a large number of steps involved. This creates many opportunities

for a peptide to be lost as residue, particularly when the levels of that peptide are low. To

accurately establish the reliability of this marker, a greater number of replicates need to be

analysed on the same and different samples.

Additionally, false negatives could have occurred via the uneven distribution of peptides

within the sample. This would mean the aliquot injected did not contain all peptides present

in the sample, leading to the observed lack of detection. The fact that the peptides derived

from tuna myoglobin have a tendency to aggregate, supports this theory.77 This occurs due

to the slight hydrophobic behaviour of tuna Mb in solution, caused by a low number of

charged residues in their amino acid sequences.77 The disruption of these aggregations could

therefore increase the reliability of these markers. One method of doing so would be to add

a chaotropic agent to samples prior to their analysis. This is a molecule that interferes with

hydrogen bonds and other non-covalent forces within an aqueous (polar) solution to increase

entropy (disorder). This results in a reduction of the hydrophobic effect, decreasing the

formation of aggregates.78-80 Examples of chaotropes are urea or guanidine hydrochloride.

These have been used in previous investigations as a mobile phase modifier to improve

separation efficiency in liquid chromatography. Unfortunately, these are not compatible with

mass spectrometry, as they can cause the suppression of ion signals reducing the sensitivity

of analyses.81-82 They would therefore not be suitable in a coupled LC-MS/MS system.

Page 39: Masters Research Project

39

Investigation into the reduction of hydrophobic interactions between peptides in solution

could therefore be useful to further develop this approach.

5.1.2. Raw Samples

Two peptides were not detected in the raw samples of crude extract, 1035.0 and 989.5.

Though they are both unique peptides, the loss of 1035.0 is of the greatest concern, as this is

the only marker unique to yellowfin. Without it, the approach cannot effectively distinguish

yellowfin from the other three species tested. Both human error and peptide aggregation

could be responsible for this outcome as they are in purified samples. An increased number

of replications and the use of a denaturant post proteolysis are therefore still advised in future

studies. However, it is less likely that experimental error is the cause of peptide loss in the

raw samples. This is because the preparative process is far shorter for raw samples, and there

are therefore fewer opportunities for peptides to be lost in residue. It is even less likely the

cause behind the loss of 1035.0, as the fragments associated with this peptide produced ion

signals of relatively high intensity. This indicates that this peptide is fairly abundant within

yellowfin myoglobin and is therefore unlikely to be completely lost this way.

The lack of purification is likely to increase the number of peptide-peptide interactions. This

is due to the fact that there are peptides present from many proteins in crude extract, as

opposed to just those derived from isolated myoglobin. This could cause an increase in the

number of disulfide bridges and a greater amount of aggregation. Disulfide bridges are also

more likely to occur in aggregates, due to the increased proximity of cysteine residues from

separate peptides. This is the likely mechanism for the lack of detection of 1035.0 and 989.5,

as these peptides both contain cysteine residues. As discussed above, the reduction of

hydrophobic interaction and aggregation could increase the reliability of these markers.

Page 40: Masters Research Project

40

5.1.3. Processed Samples

In the present study, peptides were not detectable in the crude extracts derived from 9 cans

of tuna. However, this is not in keeping with preliminary testing of the technique on the same

system, which successfully detected peptides in other canned tunas. As well as this, the

analysis of canned red meats via this technique also allowed successful product speciation.63

This indicates that the lack of detection in the present work was due to human error. This

explanation is made more likely by the fact that the processed samples in this investigation

were only analysed once, with one sample per can. These results therefore should not be

considered a reflection of the capabilities of this technique, as the error should not reoccur

during the analysis of new samples.

As the largest difference between this work and the previous studies is the samples used, it

is likely that the error occurred during sample preparation prior to LC-MS/MS analysis. Most

errors in preparation are likely to cause incomplete extraction or partial proteolysis, if all the

reagents used are genuine and functional. However, in these instances some peptides would

be detected, as opposed to none at all. In order to cause the universal lack of detection in all

processed samples, the error must be due to the complete failure of extraction or proteolysis.

Unsuccessful proteolysis is most likely as Mb is known to be relatively resistant to digestion,

but easy to extract.63, 83 This was countered by the heat denaturation of samples and the

addition of urea prior to digestion. Despite this, proteolysis could have still failed due to the

fragility of trypsin. To maintain proper function, trypsin must be stored in the correct

conditions (i.e. below -20 °C). Failure to do so would result in unsuccessful proteolysis due to

the inactivation of trypsin, which could explain the detection of no peptides in these

samples.84 The functionality of the trypsin used should therefore be verified before using it in

further study.

The water solubility of myoglobin should also be considered in future processed analyses.

Canned products consist of tuna portions stored in an organic or aqueous (water-based)

solutions. In the current study, cans were drained of solution prior to extraction. For products

stored in aqueous mediums, this could result in a loss of any myoglobin that has diffused into

the solution. This is unlikely to be a large source of error in the current study, as processed

Page 41: Masters Research Project

41

samples within organic and aqueous solutions were analysed. However, aqueous solutions

should be retained during the extraction of future samples, to maximise the yield of

myoglobin for downstream analyses. This consideration could also be used to decrease the

total time of analysis for this approach. If sufficient myoglobin has passively diffused, direct

analysis of this liquid could be used to indicate the species composition of the product. The

myoglobin content of various aqueous mediums in canned tunas should therefore be

established to determine their utility for product authentication.

5.2. Marker Specificity

The specificity of the selected peptides, was compromised during the multiplex analysis of

consecutive samples. Despite this, the results indicate that these Mb peptides are still useful

as markers. When determining the presence of predicted transitions, ion signals unique to

yellowfin, albacore, skipjack and bigeye were produced. Discrimination between these

species was therefore possible using these peptides. This implies that the current approach

could allow the species composition of unidentified samples to be accurately determined with

these markers. In this dataset, this is only true when samples are tested separately (with

extensive LC-MS/MS cleans in between), to eliminate carryover, the cause of within system

contamination.

Once preventative measures for carryover are established, methods testing a number of

samples in succession could, theoretically, be used for rapid species assignment with the

selected markers. However, although the selected markers are able to discriminate between

these four tuna Mb, the introduction of additional species may be problematic. Of particular

concern is the fact that the amino acid sequences of the Mb in bigeye and Atlantic, Pacific and

southern bluefin are identical. Therefore, these markers would not be able to discriminate

between these species, creating a vulnerability to species substitution. In the sashimi market,

there is a large economic incentive for the substitution of valuable bluefin products with

cheaper bigeye. There is also the possibility of accidental bigeye substitution with bluefin,

potentially via bycatch, which is concerning given the danger of bigeye extinction. 16-18, 56 To

circumvent these risks, the use of alternative proteins with more interspecific variations in

Page 42: Masters Research Project

42

their amino acid residues would be necessary, to distinguish between bluefin and bigeye

species.

5.2.1. Carryover

The main limiting factor in this investigation was the removal of peptide specificity by

contamination. The fact that this was sourced within the LC-MS/MS system, confirms the

occurrence of carryover, a cause of systematic error. This can be defined as the transfer of

material from previously injected samples to the next, causing contamination.85 This is

frequently observed in contemporary mass spectrometry, since the development of systems

of greater sensitivity, with certain organic substances detectable at levels around 1 fg

(1 x 1015 g).86 This means that even small amounts of carryover can produce ion signals via LC-

MS/MS, causing false positives and negatives.87 In the current context, this could result in

incorrect species assignment, when analysing samples of unknown origin. It is therefore

necessary that the mechanisms that cause carryover be considered, before ways to reduce

and eliminate it are determined. This would aid the development of this technique for tuna

authentication, as well as enhance chromatography and tandem mass spectrometry as a

whole. It is worth noting that carryover, though present, was not a critical issue when red

meat samples were studied on the same LC-MS/MS system, with the same approach.63 This

has two implications: first, that a certain amount of carryover is an issue universal to the

system used; and second, that carryover is exacerbated by properties intrinsic to tuna

myoglobin.

The carryover specific to the system is most likely caused by the retention of residue from

previous samples. This is commonly observed within the auto-sampler of LC-MS/MS systems.

It can also be caused by analytes trapped in dead volume (i.e. empty space) elsewhere in the

system. This can occur in poorly fitted connections between components, or due to damage

within the components themselves. When a blank sample is run, carryover resulting from

residue is reported to appear as mimicry of the previous chromatogram, which dilutes with

subsequent runs.88 This is a fitting description of the carryover observed in the present study,

indicating that residual sample is retained in the system. This can be avoided by running

samples on an alternative system. The current system could also be deconstructed, followed

Page 43: Masters Research Project

43

by the thorough inspection each of its constituents for damage and dead volume. In the

present work, neither option was explored due to instrument availability and time

constraints, though these avenues should be pursued in further research.

With regards to the carryover specific to the sample, it is likely due to the adsorptive

interaction of peptides with the surface of the LC-MS/MS system. This would lead to the

retention of peptides within the system, until they are desorbed by subsequent injections.87

The amount of carryover observed in the same system would therefore vary between

samples, depending on their chemical structures. This would explain the deviation in the

amount of carryover observed in red meat Mbs and tuna Mbs. Studies have revealed that the

number of charged residues in mammalian myoglobin is far higher than the number in tuna

myoglobin, causing them to be relatively more hydrophobic in solution. 77 This could lead to

an increase in the adsorption of tuna peptides to hydrophobic components within the LC-

MS/MS system, such as any resinous materials (plastics). It could also extend their retention

on the chromatography column.88 This would manifest as an increase in the levels of

carryover in late eluting peptides. Late eluting peptides are the most hydrophobic, making

them the most hydrophobic. They would therefore require the highest concentrations of

organic solvent to elute, which occurs at the end of a gradient profile. In the present work,

the late eluting tuna peptides were generally observed to cause the most intense and

persistent carryover. This supports the theory that hydrophobic interactions were a cause of

carryover in the system.

Exclusion of parts of the LC-MS/MS configuration reinforced this, as carryover is present in

both the auto-sampler, the column and elsewhere in the system. In previous works, Vespel,

a durable polymer used in the rotor seals of sample injectors, has been criticized for its affinity

to hydrophobic molecules. The replacement of these seals with ones made of PEEK (poly

ether ether ketone) could bypass this issue, as these plastics are reported to be more resistant

to hydrophobic compound adsorption.88

Additionally, measures were applied in this investigation to counter carryover by rinsing or

flushing various parts of the LC-MS/MS system (Table 2).85 Though87 all89 reduced the levels of

carryover observed, some even eliminating it, this was not without significant time penalty.

Page 44: Masters Research Project

44

This is impractical for use in rapid identification, as a key advantage of the CPCP approach is

the speed of its analysis. It is therefore important that alternative solutions to peptide

carryover be explored.

Solution Rationale Method Efficacy

Needle rinse

Carryover can be caused by the transfer of analytes retained on the outside of the injection needle during each injection. Washing the outside of the needle can remove this residue.

The outside of the injection needle is rinsed with in an organic wash (60% ACN, 30% i-PrOH, 10% water, 0.1% FA) in a flush port within the autosampler after each injection.87

Although the time penalty for this method was low (1 min per sample), only minor reductions in carryover were observed

Blank injection

The injection of a sample containing no peptides could flush any previous residue, removing carryover from the system.

10 μl of ACN/H2O (3:97; 0.1% FA) is injected between each sample on the same method.

Using this method, the system is carryover safe after 4 injections. Unfortunately, this causes a significant time penalty increasing analysis time (52 mins per sample.

TFE injection

2,2,2-trifluoroethanol (TFE) has a strong acidic character due to the electronegativity of its trifluoromethyl (CF3) group. This gives TFE a strong affinity for peptides, enhancing their solubility.87, 89 The injection of TFE could therefore disrupt peptides that have interacted with the LC-MS/MS system, eliminating carryover.

10 μl of TFE is injected between each sample on the same method.

The system becomes carryover safe after 2 injections. Though this is less than injection with ACN/H2O (3:97; 0.1% FA) the time penalty is still substantial (76 mins per sample)

Repeated gradient elution

Targeting the occurrence of late eluting carryover, the fluctuation of the organic concentration of mobile phase can disrupt the hydrophobic interactions of the most organic signals by repeatedly creating a high organic solvent concentration that promotes their elution.

After the initial gradient profile has run, the organic solvent (B) ramps to 100% and is held for 1 mins to allow pressure to subside. This was repeated 6 times.

The system becomes carryover safe after 6 additional gradients. Time penalty is 110 mins per sample

Table 2 Methods to reduce carryover in the LC-MS/MS analysis of tuna myoglobin. As quantitative analysis was not part of the current objectives, the efficacy of each method is commented on qualitatively, based on the observed reductions in the peak heights of carryover transitions. The system can be classed as carryover safe when the effect of carryover no longer interferes with the present analyses, it does not signify that all carryover has been eliminated.85 The most effective method currently seems to be a combination of a blank TFE injection with repeated gradient elution between samples.

Page 45: Masters Research Project

45

As described, the occurrence of carryover interferes with effective sample speciation.

However, in these circumstances it has allowed some positive conclusions to be made about

the CPCP approach. Carryover, like all contamination, is essentially the unintentional creation

of mixtures. Its presence in this study therefore indicates that some of the corresponding tuna

myoglobin tested can be identified at very low levels in a homogeneous mixture. Though

many carryover transitions did not occur in the expected species, they were still detected at

the scheduled retention times. This allowed their species of origin to be identified. Based on

this, the following is highly likely: skipjack is detectable in purified and raw yellowfin, albacore

and bigeye; bigeye is detectable in raw yellowfin and albacore; yellowfin is detectable in

purified albacore. These are powerful observations in the context of tuna authenticity, as the

identification of species in a homogeneous mixture enables the detection of unlawful product

adulteration. This is especially true for the detections listed concerning skipjack and yellowfin,

as there is economic incentive for partial substitutions using these cheaper species.

6. CONCLUSION

The MRM MS analysis of four species of tuna myoglobin yielded signals that differed

significantly between each species of Mb. These results were then used to select 21 peptides

suitable for use as markers in multiplex testing of samples for yellowfin, albacore, skipjack

and bigeye. This allows the null hypothesis (H1) to be rejected. The CPCP approach can be

used to distinguish between these four species of tuna.

The multiplex analysis of purified myoglobin did not produce signals that differed significantly

between tuna species. This was because of the occurrence of carryover within the LC-MS/MS

system. Until a solution for carryover can be determined, samples cannot be analysed

consecutively, without time consuming cleaning processes in between. For this reason, the

null hypothesis (H2) that the CPCP approach cannot be used to rapidly distinguish between

various tuna species via MRM MS, cannot be rejected.

The multiplex analysis of raw tuna extracts did not produce signals that differed significantly

between tuna species. This was again due to the occurrence of carryover within the system,

removing the specificity of marker peptides. To allow rapid speciation, an efficient solution

Page 46: Masters Research Project

46

for carryover must be determined. Therefore, the null hypothesis (H3) cannot currently be

rejected. The CPCP approach cannot be used to rapidly distinguish between various species

of tuna in readily available samples.

The multiplex analysis of extracts derived from 9 commercially canned tuna samples

produced no signals at scheduled retention times. This is likely to have occurred due to failed

proteolysis caused by human error. The null hypothesis (H4) cannot be rejected. The results

support the conclusion that the CPCP approach cannot be used to rapidly distinguish between

various species of tuna in extensively processed tuna samples. This support is not strong

however, due to the high likelihood of error.

To further develop this approach it is important to investigate ways to reduce hydrophobic

interactions within tuna myoglobin. Based on these results and previous studies, the relative

hydrophobicity of tuna Mb is likely to contribute to inconsistent marker detection (via peptide

aggregation), as well as cause carryover with the LC-MS/MS system. Reducing these

interactions could therefore allow discrimination between consecutively tested tuna samples,

as well improve the reliability of the selected markers by disrupting aggregates. To further

improve the latter, the addition of reduction and alkylation steps to sample preparation may

be beneficial, to reduce the occurrence of disulfide bridges.

It is challenging to find methods to reduce hydrophobicity that do not impede

chromatographic separation, nor inhibit mass spectrometry. It is also important to identify a

solution that does not cause a significant time penalty, to facilitate the intended purpose of

this approach: to rapidly identify the constituent species in a tuna product. The myoglobin

concentration in aqueous mediums should also be determined, to establish their utility for

more rapid product authentication in relevant canned products.

Despite the improvements to be made, the CPCP approach could still be a valuable

contribution to tuna authenticity. Although accurate discrimination in tuna is not yet possible,

the identification of markers visible in both purified Mb and raw tuna extracts indicates that

this approach remains reliable in complex mixtures. Additionally, the fact that peptides were

not detected in the processed extracts is not of great concern, as this is likely to have occurred

Page 47: Masters Research Project

47

due to human error. It is therefore not expected to reoccur in further testing, a theory

supported by previous preliminary analysis of canned tuna with the same approach. The

indication that low levels of tuna species will be identifiable in a homogeneous mixture, is

also encouraging. This is based on the detection of carryover. For these reasons investigation

into the application of the CPCP approach to tuna will continue at the Institute of Food

Research. An aim of this will be to produce further publications, which the present work will

be a part of.

7. ACKNOWLEDGEMENTS

I acknowledge the financial support of the Institute of Food Research and the University of

East Anglia. I thank Kate Kemsley, Neil Rigby, Mark Philo, Yvonne Gunning, Andrew Watson,

Stephen Day, Andrew Mayes and Rebecca Lewis for their guidance and feedback throughout

this project.

8. REFERENCES

1. Reid LM, O'Donnell CP, Downey G (2006); Recent technological advances for the determination of food authenticity, Trends in Food Science & Technology, 17(7): 344-353.

2. Dennis MJ (1998); Recent developments in food authentication, Analyst, 123(9): 151R-156R.

3. Directorate General for Health and Consumers (2013); Horse Meat Issue, European Comission (Accessed: 15/05/2016) http://ec.europa.eu/food/food/horsemeat/tests_results_en.htm

4. Avery J (2014); Fighting food fraud, European Parliamentary Research Service (Accessed: 28/03/2016) http://www.europarl.europa.eu/RegData/bibliotheque/briefing/2014/130679/LDM_BRI(2014)130679_REV1_EN.pdf

5. Squires N (2015); Italian olive oil scandal: seven top brands 'sold fake extra-virgin', The Telegraph (Accessed: 15/05/2016)

Page 48: Masters Research Project

48

http://www.telegraph.co.uk/news/worldnews/europe/italy/11988947/Italian-companies-investigated-for-passing-off-ordinary-olive-oil-as-extra-virgin.html

6. Camin F, Bontempo L, Perini M, Tonon A, Breas O, Guillou C, Moreno-Rojas JM, Gagliano G (2013); Control of wine vinegar authenticity through delta O-18 analysis, Food Control, 29(1): 107-111.

7. Meier-Augenstein W, Kemp HF, Hardie SML (2012); Detection of counterfeit scotch whisky by H-2 and O-18 stable isotope analysis, Food Chemistry, 133(3): 1070-1074.

8. Council Regulation (EC) 178/2002 of 28 January 2002; Laying down the general principles and requirements of food law, establishing the European Food Safety Authority and laying down procedures in matters of food safety, Official Journal, L031: 1-24

9. Food Standards Agency (2016); Food We Can Trust, FSA (Accessed: 15/05/2016) https://www.food.gov.uk/about-us/about-the-fsa

10. Food and Agriculture Organisation of the United Nations (2005); Review of the State of the world marine fishery resources, FAO Fisheries technical Paper, 457: 163-174.

11. Council Regulation (EEC) 1536/92 of 9 June 1992: Laying down common marketing standards for preserved tuna and bonito, Official Journal, L163: 1-4

12. Food and Agriculture Organisation of the United Nations (2010); Global Tuna Catches by Stock, FAO (Accessed: 23/03/2016) http://www.fao.org/fishery/statistics/tuna-catches/en

13. Miyake MP, Guillotreau P, Sun CW, Ishimura G (2010); Recent deveopments in the tuna industry; Stocks, fisheries, management, processing, trade and markets, FAO Fisheries and Aquaculture Technical Paper, 543.

14. Begg GA, Friedland KD, Pearce JB (1999); Stock identification and its role in stock assessment and fisheries management: an overview, Fisheries Research, 43(1-3): 1-8.

15. Southern Pacific Conference (2010); Tuna Species caught in the WCPO, SPC Oceanic Fisheries Programme (Accessed: 15/05/2016) http://www.spc.int/oceanfish/en/tuna-fisheries/tuna-species

16. Collette B, Chang S-K, Di Natale A, Fox W, Juan Jorda M, Miyabe N, Nelson R, Uozumi Y, Wang S (2011); Thunnus maccoyii, The IUCN Red List of Threatened Species (Accessed: 03/05/2016) http://www.iucnredlist.org/details/21858/0

Page 49: Masters Research Project

49

17. Collette B, Amorim AF, Boustany A, Carpenter KE, de Oliveira Leite Jr N, Di Natale A, Die D, Fox W, Fredou FL, Graves J, Viera Hazin FH, Hinton M, Juan Jorda M, Kada O, Minte Vera C, Miyabe N, Nelson R, Oxenford H, Pollard D, Restrepo V, Schratwieser J, Teixeira Lessa RP, Pires Ferreira Travassos PE, Uozumi Y (2011); Thunnus thynnus, The IUCN Red List of Threatened Species (Accessed: 03/05/2016) http://www.iucnredlist.org/details/21860/0

18. Collette B FW, Juan Jorda M, Nelson R, Pollard D, Suzuki N, Teo S, (2014); Thunnus orientalis, The IUCN Red List of Threatened Species (Accessed: 03/05/2016) http://www.iucnredlist.org/details/170341/0

19. Dempster T, Taquet M (2004); Fish aggregation device (FAD) research: gaps in current knowledge and future directions for ecological studies, Reviews in Fish Biology and Fisheries, 14(1): 21-42.

20. Teo SLH, Block BA (2010); Comparative Influence of Ocean Conditions on Yellowfin and Atlantic Bluefin Tuna Catch from Longlines in the Gulf of Mexico, Plos One, 5(5).

21. Warner K, Timme W, Lowell B, Hirshfield M (2013); Oceana study reveals seafood fraud nationwide, Oceana (Accessed: 20/03/2016) http://oceana.org/sites/default/files/reports/National_Seafood_Fraud_Testing_Results_FINAL.pdf

22. Mackie IM, Pryde SE, Gonzales-Sotelo C, Medina I, Perez-Martin R, Quinteiro J, Rey-Mendez M, Rehbein H (1999); Challenges in the identification of species of canned fish, Trends in Food Science & Technology, 10(1): 9-14.

23. Food and Agriculture Organisation of the United Nations (1988); Manual on Fish Canning, FAO (Accessed: 01/04/2016) http://www.fao.org/docrep/003/t0007e/t0007e05.htm#4.2

24. International G (2010); Genetic testing on tinned and pouched tuna products, Greenpeace (Accessed: 16/05/2016) http://www.greenpeace.org/international/Global/international/publications/oceans/2010/briefing_paper_genetic_tuna.pdf

25. Pardo MA, Perez-Villareal B (2004); Identification of commercial canned tuna species by restriction site analysis of mitochondrial DNA products obtained by nested primer PCR, Food Chemistry, 86(1): 143-150.

26. Lockley AK, Bardsley RG (2000); DNA-based methods for food authentication, Trends in Food Science & Technology, 11(2): 67-77.

Page 50: Masters Research Project

50

27. Rasmussen RS, Morrissey MT (2008); DNA-Based methods for the identification of commercial fish and seafood species, Comprehensive Reviews in Food Science and Food Safety, 7(3): 280-295.

28. Lin WF, Hwang DF (2007); Application of PCR-RFLP analysis on species identification of canned tuna, Food Control, 18(9): 1050-1057.

29. Poms RE, Agazzi ME, Bau A, Brohee M, Capelletti C, Norgaard JV, Anklam E (2005); Inter-laboratory validation study of five commercial ELISA test kits for the determination of peanut proteins in biscuits and dark chocolate, Food Additives and Contaminants, 22(2): 104-112.

30. Tyreman AL, Bonwick GA, Smith CJ, Coleman RC, Beaumont PC, Williams JHH (2004); Detection of irradiated food by immunoassay - development and optimization of an ELISA for dihydrothymidine in irradiated prawns, International Journal of Food Science and Technology, 39(5): 533-540.

31. Stave JW (2002); Protein immunoassay methods for detection of biotech crops: Applications, limitations, and practical considerations, Journal of Aoac International, 85(3): 780-786.

32. Asensio L, Gonzalez I, Garcia T, Martin R (2008); Determination of food authenticity by enzyme-linked immunosorbent assay (ELISA), Food Control, 19(1): 1-8.

33. Santaclara FJ, Velasco A, Perez-Martin RI, Quinteiro J, Rey-Mendez M, Pardo MA, Jimenez E, Sotelo CG (2015); Development of a multiplex PCR-ELISA method for the genetic authentication of Thunnus species and Katsuwonus pelamis in food products, Food Chemistry, 180: 9-16.

34. von Bargen C, Brockmeyer J, Humpf HU (2014); Meat Authentication: A New HPLC-MS/MS Based Method for the Fast and Sensitive Detection of Horse and Pork in Highly Processed Food, Journal of Agricultural and Food Chemistry, 62(39): 9428-9435.

35. Vogt G, Woell S, Argos P (1997); Protein thermal stability, hydrogen bonds, and ion pairs, Journal of Molecular Biology, 269(4): 631-643.

36. Blaber M, Zhang XJ, Matthews BW (1993); Structural Basis of Amino-acid Alpha-Helix Propensity, Science, 260(5114): 1637-1640.

37. Premanandh J (2013); Horse meat scandal - A wake-up call for regulatory authorities, Food Control, 34(2): 568-569.

Page 51: Masters Research Project

51

38. Sentandreu MA, Sentandreu E (2011); Peptide biomarkers as a way to determine meat authenticity, Meat Science, 89(3): 280-285.

39. Herrero M, Simo C, Garcia-Canas V, Ibanez E, Cifuentes A (2012); Foodomics: MS-based strategies in modern food science and nutrition, Mass Spectrometry Reviews, 31(1): 49-69.

40. Jungblut PR, Holzhutter HG, Apweiler R, Schluter H (2008); The speciation of the proteome, Chemistry Central Journal, 2.

41. Gallien S, Duriez E, Domon B (2011); Selected reaction monitoring applied to proteomics, Journal of Mass Spectrometry, 46(3): 298-312.

42. von Bargen C, Dojahn J, Waidelich D, Humpf HU, Brockmeyer J (2013); New Sensitive High-Performance Liquid Chromatography Tandem Mass Spectrometry Method for the Detection of Horse and Pork in Halal Beef, Journal of Agricultural and Food Chemistry, 61(49): 11986-11994.

43. Anonymous de Hoffmann E, Stroobant V (2007); Mass Spectrometry: Priciples and Applications, Chichester: Wiley

44. Martinez I, Friis TJ (2004); Application of proteome analysis to seafood authentication, Proteomics, 4(2): 347-354.

45. Alonso R, Rodriguez-Estevez V, Dominguez-Vidal A, Ayora-Canada MJ, Arce L, Valcarel M (2008); Ion mobility spectrometry of volatile compounds from Iberian pig fat for fast feeding regime authentication, Talanta, 76(3): 591-596.

46. Agilent Technologies (2007); Metabolomics: Considerations for Selecting GC/MS or LC/MS for Metabolomics, Agilent Technologies (Accessed: 07/04/2016) https://www.agilent.com/cs/library/selectionguide/Public/5989-6328EN.pdf

47. Kirkland JJ (1973); POROUS SILICA MICROSPHERE COLUMN PACKINGS FOR HIGH-SPEED LIQUID-SOLID CHROMATOGRAPHY, Journal of Chromatography, 83(AUG29): 149-167.

48. Walter TH, Iraneta P, Capparella M (2005); Mechanism of retention loss when C-8 and C18HPLC columns are used with highly aqueous mobile phases, Journal of Chromatography A, 1075(1-2): 177-183.

Page 52: Masters Research Project

52

49. Snyder LR, Dolan JW, Gant JR (1979); GRADIENT ELUTION IN HIGH-PERFORMANCE LIQUID-CHROMATOGRAPHY .1. THEORETICAL BASIS FOR REVERSED-PHASE SYSTEMS, Journal of Chromatography, 165(1): 3-30.

50. Yang Y, Zhang S, Howe K, Wilson DB, Moser F, Irwin D, Thannhauser TW (2007); A comparison of nLC-ESI-MS/MS and nLC-MALDI-MS/MS for GeLC-based protein identification and iTRAQ-based shotgun quantitative proteomics, Journal of biomolecular techniques : JBT, 18(4): 226-237.

51. Mazzeo MF, De Giulio B, Guerriero G, Ciarcia G, Malorni A, Russo GL, Siciliano RA (2008); Fish Authentication by MALDI-TOF Mass Spectrometry, Journal of Agricultural and Food Chemistry, 56(23): 11071-11076.

52. Fenn JB, Mann M, Meng CK, Wong SF, Whitehouse CM (1989); ELECTROSPRAY IONIZATION FOR MASS-SPECTROMETRY OF LARGE BIOMOLECULES, Science, 246(4926): 64-71.

53. Taylor AJ, Linforth R, Weir O, Hutton T, Green B (1993); Potential of Electrospray Mass Spectrometry for Meat Pigment Identification, Meat Science, 33(1): 75-83.

54. Marshall AG, Hendrickson CL, Jackson GS (1998); Fourier transform ion cyclotron resonance mass spectrometry: A primer, Mass Spectrometry Reviews, 17(1): 1-35.

55. Shi SDH, Drader JJ, Freitas MA, Hendrickson CL, Marshall AG (2000); Comparison and interconversion of the two most common frequency-to-mass calibration functions for Fourier transform ion cyclotron resonance mass spectrometry, International Journal of Mass Spectrometry, 195: 591-598.

56. Miller PE, Denton MB (1986); THE QUADRUPOLE MASS FILTER - BASIC OPERATING CONCEPTS, Journal of Chemical Education, 63(7): 617-622.

57. Batey JH (2014); The physics and technology of quadrupole mass spectrometers, Vacuum, 101: 410-415.

58. Wells JM, McLuckey SA, Collision-induced dissociation (CID) of peptides and proteins. In Biological Mass Spectrometry, Burlingame, A. L., Ed. 2005; Vol. 402, pp 148-185. 59. Murray KK, Boyd RK, Eberlin MN, Langley GJ, Li L, Naito Y (2013); Definitions of terms relating to mass spectrometry (IUPAC Recommendations 2013), Pure and Applied Chemistry, 85(7): 1515-1609.

Page 53: Masters Research Project

53

60. Wulff T, Nielsen ME, Deelder AM, Jessen F, Palmblad M (2013); Authentication of Fish Products by Large-Scale Comparison of Tandem Mass Spectra, Journal of Proteome Research, 12(11): 5253-5259.

61. de Hoffmann E (1996); Mass Spectrometry: a Primer, Journal of Mass Spectrometry, 31: 129-137.

62. Lange V, Picotti P, Domon B, Aebersold R (2008); Selected reaction monitoring for quantitative proteomics: a tutorial, Molecular Systems Biology, 4.

63. Watson AD, Gunning Y, Rigby NM, Philo M, Kemsley EK (2015); Meat Authentication via Multiple Reaction Monitoring Mass Spectrometry of Myoglobin Peptides, Analytical Chemistry, 87(20): 10315-10322.

64. Uniprot Consortium (2002); BLAST: Thunnus obesus, Uniprot (Accessed: 22/04/2016) http://www.uniprot.org/blast/uniprot/B201604226605O6NHNA

65. Uniprot Consortium (2002); Align: P02205, Q9DGJ2, Q9DGI8, Q76G09, Uniprot (Accessed: 22/04/2016) http://www.uniprot.org/align/A2016042256C4S9ZQCQ

66. Brown W (1962); The Concentration of Myoglobin and Hemoglobin in Tuna Flesh, Journal of Food Science, 27(1): 26-28.

67. Nurilmala M, Ushio H, Kaneko G, Ochiai Y (2013); Assessment of Commercial Quality Evaluation of Yellowfin Tuna Thunnus albacares Meat Based on Myoglobin Properties, Food Science and Technology Research, 19(2): 237-243.

68. Ueki N, Ochiai Y (2004); Primary structure and thermostability of bigeye tuna myoglobin in relation to those of other scombridae fish, Fisheries Science, 70(5): 875-884.

69. Rappsilber J, Ishihama Y, Mann M (2003); Stop and go extraction tips for matrix-assisted laser desorption/ionization, nanoelectrospray, and LC/MS sample pretreatment in proteomics, Analytical Chemistry, 75(3): 663-670.

70. SIB Swiss Institute of Bioinformatics (2011); Peptide Cutter, ExPASy Bioinformatics Resource Portal (Accessed: 23/04/2016) http://web.expasy.org/peptide_cutter/

71. MacLean B, Tomazela DM, Shulman N, Chambers M, Finney GL, Frewen B, Kern R, Tabb DL, Liebler DC, MacCoss MJ (2010); Skyline: an open source document editor for creating and analyzing targeted proteomics experiments, Bioinformatics, 26(7): 966-968.

Page 54: Masters Research Project

54

72. Silva AMN, Vitorino R, Domingues MRM, Spickett CM, Domingues P (2013); Post-translational Modifications and Mass Spectrometry Detection, Free Radical Biology and Medicine, 65: 925-941.

73. Siepen JA, Keevil EJ, Knight D, Hubbard SJ (2007); Prediction of missed cleavage sites in tryptic peptides aids protein identification in proteomics, Journal of Proteome Research, 6(1): 399-408.

74. Pexa A, Fischer K, Deussen A, Henle T (2008); Homocysteine in food, European Food Research and Technology, 226(4): 933-935.

75. Adachi K, Yamaguchi T, Yang Y, Konitzer PT, Pang J, Reddy KS, Ivanova M, Ferrone F, Surrey S (2000); Expression of functional soluble human alpha-globin chains of hemoglobin in bacteria, Protein Expression and Purification, 20(1): 37-44.

76. Collins DS, Unanue ER, Harding CV (1991); REDUCTION OF DISULFIDE BONDS WITHIN LYSOSOMES IS A KEY STEP IN ANTIGEN PROCESSING, Journal of Immunology, 147(12): 4054-4059.

77. Watts DA, Rice RH, Brown WD (1980); THE PRIMARY STRUCTURE OF MYOGLOBIN FROM YELLOWFIN TUNA (THUNNUS-ALBACARES), Journal of Biological Chemistry, 255(22): 916-924.

78. Hatefi Y, Hanstein WG (1969); SOLUBILIZATION OF PARTICULATE PROTEINS AND NONELECTROLYTES BY CHAOTROPIC AGENTS, Proceedings of the National Academy of Sciences of the United States of America, 62(4): 1129-&.

79. Zangi R, Zhou RH, Berne BJ (2009); Urea's Action on Hydrophobic Interactions, Journal of the American Chemical Society, 131(4): 1535-1541.

80. Arakawa T, Ejima D, Tsumoto K, Obeyama N, Tanaka Y, Kita Y, Timasheff SN (2007); Suppression of protein interactions by arginine: A proposed mechanism of the arginine effects, Biophysical Chemistry, 127(1-2): 1-8.

81. Takahashi T, Hoshino H, Yotsuyanagi T (2001); The use of water structure breakers, urea and guanidium chloride, new mobile phase modifiers in reversed-phase partition high-performance liquid chromatography, Analytical Sciences, 17(7): 847-851.

82. Annesley TM (2003); Ion suppression in mass spectrometry, Clinical Chemistry, 49(7): 1041-1044.

Page 55: Masters Research Project

55

83. Park ZY, Russell DH (2000); Thermal denaturation: A useful technique in peptide mass mapping, Analytical Chemistry, 72(11): 2667-2670.

84. Sipos T, Merkel JR (1970); AN EFFECT OF CALCIUM IONS ON ACTIVITY, HEAT STABILITY, AND STRUCTURE OF TRYPSIN, Biochemistry, 9(14): 2766-&.

85. Haeckel R (1991); PROPOSALS FOR THE DESCRIPTION AND MEASUREMENT OF CARRY-OVER EFFECTS IN CLINICAL-CHEMISTRY (RECOMMENDATIONS 1991), Pure and Applied Chemistry, 63(2): 302-306.

86. Agilent Technologies (2014); Why Instrument Detection Limit (IDL) is a Better Metric for Determining The Sensitivity of Triple Quadrupole LC/MS Systems, Agilent Technologies (Accessed: 28/04/2016) http://www.agilent.com/cs/library/technicaloverviews/public/5991-4089EN.pdf

87. Mitulovic G, Stingl C, Steinmacher I, Hudecz O, Hutchins JRA, Peters JM, Mechtler K (2009); Preventing Carryover of Peptides and Proteins in Nano LC-MS Separations, Analytical Chemistry, 81(14): 5955-5960.

88. Hughes NC, Wong EYK, Fan J, Bajaj N (2007); Determination of carryover and contamination for mass spectrometry-based chromatographic assays, Aaps Journal, 9(3): E353-E360.

89. deLaureto PP, Scaramella E, DeFilippis V, Bruix M, Rico M, Fontana A (1997); Limited proteolysis of ribonuclease A with thermolysin in trifluoroethanol, Protein Science, 6(4): 860-872.

Page 56: Masters Research Project

56

9. Appendix

9.1. Research Proposal

Analysis of Tuna via the CPCP method using Multiple Reaction Mode Mass Spectrometry

Author: Joshua Kai Peazer Advisor: Dr Stephen Day

BACKGROUND AND RATIONALE Events over the past few years have brought the issue of food contamination into the media spotlight. In 2013, it was discovered that an array of beef products in Europe contained or consisted entirely of undeclared horse meat. The implications of this were catastrophic for financial, health, ethical and religious reasons, accenting a flaws in current food authentication protocol (von Bargen et al 2013).

The most widely used food authentication techniques currently rely on DNA-based testing. Unfortunately, these methods are slow, have a limited number of identifiable species, and have poor quantitative ability. This lead the Analytical Sciences Unit at the IFR (Institute of Food Research) to propose a new method of meat authentication using Multiple-Reaction-Mode Mass Spectrometry or MRM MS (Watson et al 2015). The approach determines the species composition of a sample, by detecting its constituent ratio of peptide markers. These peptides correspond to proteins characteristic of a single species, allowing identification. This process can be referred to as the CPCP method for concision. The technique has proven effective in identifying beef, pork, lamb and horse, both raw and cooked. The approach was also able to quantify adulterants in these meats at levels below 1%, surpassing traditional DNA-based authentication (Watson et al 2015).

Similarly to the undeclared use of horsemeat, the fraudulent labelling of species as tuna is harmful both economically and medically. There are also ethical and environmental implications associated with tuna fraud. It occurs frequently, as it is unclear where the deception takes place in the supply chain. An investigation carried out by Warner et al. (2013) revealed that 59% of tuna was mislabelled from 2010 to 2012. Escolar (Lepidocybium flavobrunneum) was often used in its place, a species that is indigestible by humans without negative health effects (Warner et al 2013). As the CPCP method using MRM MS is a method still under development, its efficacy on many species has still not been tested. For this reason and the aforementioned issues associated with Tuna fraud, it would be both interesting and potentially profitable to test this novel method on tuna samples, both processed and raw. This would provide an interesting challenge for the method, to see if it is effective even on highly processed samples.

Page 57: Masters Research Project

57

OBJECTIVES & HYPOTHESES To better understand the versatility of the CPCP method coupled with MRM MS, the aims and objectives of the proposed investigation are as follows: 1. To establish the whether the CPCP method coupled with MRM MS can distinguish

between different species of tuna:

H0 - MRM MS performed on purified myoglobin produces no signals that do not differ

significantly between tuna species. CPCP coupled with MRM MS cannot be used to

distinguish between different the species of tuna.

H1 - MRM MS performed on purified myoglobin produces signals that differ significantly

between tuna species. CPCP coupled with MRM MS can be used to distinguish between

different the species of tuna.

2. To establish the whether the CPCP method coupled with MRM MS can be used to detect

different species in readily available tuna samples:

H0 - MRM MS performed on tuna steaks produces no signals corresponding to reference

peaks. CPCP coupled with MRM MS cannot be used to detect species composition of readily

available tuna.

H1 - MRM MS performed on tuna steaks produces signals corresponding to reference peaks.

CPCP coupled with MRM MS can be used to be used to detect species composition of readily

available tuna.

3. To establish the whether the CPCP method coupled with MRM MS can be used to detect different species in tuna samples that have undergone extensive processing:

H0 - MRM MS performed on tuna steaks produces no signals corresponding to reference

peaks. CPCP coupled with MRM MS cannot be used to detect species composition of

processed tuna.

H1 - MRM MS performed on tuna steaks produces signals corresponding to reference peaks.

CPCP coupled with MRM MS can be used to be used to detect species composition of

processed tuna.

METHODOLOGY It is important to note that all processes and will be repeated as many times as possible on as many species of tuna as time and sourcing allows. This will provide increased reliability of results.

Predictive Tools for Comparison To predict the peptides produced from digestion, PeptideCutter will be used. Though the tool has received criticism for its limited dataset in respect to certain enzymes, its tryptic digest dataset has proven accurate and convenient (Wang et al. 2013; Watson et al. 2015; Afzal et al 2011).

Page 58: Masters Research Project

58

The Open Source Skyline software will be used to predict MRM transition candidates for the ABI 4000 QTrap. This has proven to an analytical tool of great utility for MRM assay design on specific instruments (Liebler & Zimmerman 2013).

Materials As in Watson et al (2015), methanol and acetonitrile will be purchased from Fisher Scientific (Loughborough, UK). Urea and trypsin (from bovine pancreas) will be purchased from Sigma-Aldrich (Gillingham, UK) and formic acid from BDH Chemicals (Poole, UK). Tuna products of various species, namely yellowfin (Thunnus albacares), albacore (Thunnus alalunga), skipjack (Katsuwonus pelamis) and bigeye (Thunnus obesus), will preferably be purchased from local supermarkets and fishmongers to avoid waste. If this is not feasible due to product availability, wholesalers and specialised fish suppliers will be utilized and tuna bought in bulk.

Purified Myoglobin Preparation To test the first hypothesis, purified myoglobin will be prepared prior to proteolysis and LCMS analysis. This will also serve as a screening process, to identify biomarkers that can be used in further experimentation and to test hypotheses 2 and 3. Tuna samples (30g) will be extracted at 4°C for using 100mM TRIS buffer adjusted to pH 8.4 with HCl. The temperature and pH must be carefully controlled to reduce the risk of denaturing or otherwise affecting the purity of the desired myoglobin (Watson et al. 2015; Yee & Peyton 1995; Klis et al. 2007). The sample will then be filtered and centrifuged at 4°C for 30mins and the supernatant separated and filtered once more. Sodium azide and potassium sorbate act as biocides for anaerobic and aerobic respirators respectively. They will therefore be added for their utility as preservatives (Watson et al. 2015; Chefetz et al. 2006; Quintavalla & Vicini 2002). This crude extract will then be separated via FPLC, using a Superdex 75 size exclusion column (GE Healthcare, Chalfont, UK) attached to a BioCad Sprint HPLC system (Applied Biosystems). This was equilibrated with 100mM TRIS/HCl (pH 8.4) at a flow rate 1 mL/min. The use of FPLC over HPLC in protein identification and quantitation has proven favourable in past studies and slightly more cost effective (Tangvarasittichai et al 2007). A SuperloopTM will also be used as opposed to a manual sample loop to enable the purification process to become automated, increasing efficiency (Eisele et al. 2012). To monitor the eluent for protein, the absorbance will be fixed to 280 nm and the flow rate maintained at a low speed to avoid decreases in spectra resolution. Resultant fractions will then be analysed by SDS-PAGE to establish which contain the purified myoglobin. These will then be pooled and stored at 4°C.

Tuna Sample Preparation Samples prepared in this section will be analysed via LCMS to test hypothesis 2. This is contingent on the success of hypothesis 1. If no species specific biomarkers are identified in the screening process, then H0 must be accepted for all hypotheses.

Page 59: Masters Research Project

59

Samples of ground tuna will be prepared and extracted using a phosphate extraction buffer (0.3 M KCl, 0.3 M phosphate at pH 6.5) again to prevent damage to the desired myoglobin by controlling pH. Phosphate buffers in previous study, have been shown to produce more extractable protein from meat samples (Nahar et al 2013). A 2 mL aliquot of the sample will then be centrifuged briefly at high speed and 200 μL of the supernatant evaporated to dryness at 50°C. This will be carried out using a Jouan RC 1022 centrifugal evaporator (Watson et al. 2015).

Preparation of Processed Sample In order to test hypotheses 3 a canning process will be mimicked using known species of tuna. This will be kept as close to conditions in industrial canning as possible. Tuna will be steamed at 100°C for 60 min or until its temperature reaches between 60-85°C. The fish will then be water cooled and placed in a sealed container in a brine medium, then vacuum sealed to replicate the conditions of a can. The container will then be heated once more for 2 hours then preferably pressure cooled (Warne 1988).

Sample Proteolysis The tuna samples either raw or processed will be dissolved in 25 mM ammonium bicarbonate prior to the rest of the proteolysis process. This step is the final difference in the treatments of the purified myoglobin, the raw samples and the processed tuna samples. Therefore for brevity, both shall be referred to as “the samples” or similar from this point forward. The samples will be heated to 95°C for 30 mins, then cooled. Urea will then be added to a final concentration of 0.5M (Watson et al. 2015). Low concentrations of urea has been shown to promote the unfolding which is beneficial in preparation for proteolysis (Schnarr & Maurizot 1981). The solution will then be digested by trypsin in a 1:30 enzyme to substrate weight ratio. The samples will undergo digestion overnight at 37°C as is standard protocol (Watson et al. 2015). Accelerated digestion times could result in an improper digest due to the heterogeneity of the samples (López-Ferrer D et al 2006). High concentrations of salt and urea are incompatible with analysis by mass spectroscopy (Gundry et al 2009). The tryptic digests will therefore undergo dilution with water and desalting using a Strata-X 33μ reversed phase cartridge (Phenomenex, Macclesfield, UK), as per standard operating procedure (Watson et al 2015). The desalted sample will then be evaporated to dryness via centrifugal evaporation. Finally, the samples will then be made ready for LC-MS/MS analysis by dissolving then in acetonitrile/ H2O (Watson et al 2015).

LC-MS/MS Analysis Analysis will be carried out via high performance liquid chromatography coupled with tandem mass spectrometry (HPLC-MS/MS). This will be conducted on an Agilent 1200 Rapid Resolution LC system (Stockport UK) connected to an AB Sciex 400 QTrap triple quadrupole mass spectrometer (Warrington UK). The capillary column used for chromatographic separations will be a Phenomenex XB C18 reversed-phase column using a 300 μl/min at 40°C (Watson et al 2015). Temperature must be controlled at relatively high levels during chromatographic separation to ensure a higher quality output (Blacker et al 2008). The

Page 60: Masters Research Project

60

binary gradient used will consist of 97% A (water and 0.1% formic acid) and 3% B (acetonitrile and 0.1% formic acid) to 24%. This will be increased for a period to 100% B at the end to ensure the extraction of all compounds of interest (Skoog & West 1980). Using the predictive tools, a MS/MS methods file will be created encompassing Myoglobin from all species of Tuna. Output from MS will be viewed on Analyst 1.6.2 software (AB Sciex). Positive electrospray in dynamic monitoring mode will be used to detect eluted peptides with the four MRM transitions of the highest intensity used to determine peak height (Watson et al 2015). The electrospray ionisation mass spectrometry (ESI-MS) can provide less structural information than other MS methods, the use of tandem MS negates this limitation (Ho et al 2003).

REFERENCES Blacker AR, Speers AE, Wu CC (2008); Chromatographic benefits of elevated temperature for the proteomic analysis of membrane proteins, Proteomics, 8(19): 3956-64 Chefetz B, Stimler K, Shechter M, Drori Y (2006); Interactions of sodium azide with triazine herbicides: Effect on sorption to soils, Chemosphere, 65(2): 352-57 Eisele T, Stressler T, Kranz B, Fischer L (2012); Automated multi-step purification protocol for

Angiotensin-I-Converting-Enzyme (ACE), Journal of Chromatography B, 911: 64-70

Gundry RL, White MY, Murray CI, Kane LA, Fu Q, Stanley BA, Van Eyk JE (2009); Preparation of Proteins and Peptides for Mass Spectrometry Analysis in a Bottom-Up Proteomics Workflow, Current Protocols in Molecular Biology, 88: 10.25 Ho CS, Chan MHM, Cheung RCK, Law LK, Lit LCW, Ng KF, Suen MWM, Tai HL (2003); Electrospray

Ionisation Mass Spectrometry: Principles and Clinical Applications, Clinical Biochemist Review, 24(1):

3–12

Klis FM, de Jong M, Brul S, Groot PW (2007); Extraction of cell surface-associated proteins from living yeast cells, Yeast, 24(4): 253-8 Liebler DC, Zimmerman LJ (2013); Targeted Quantitation of Proteins by Mass Spectrometry,

Biochemistry, 52(22): 3793-806

López-Ferrer D, Cañas B, Vázquez J, Lodeiro C, Rial-Otero R, Moura I, Capelo JL (2006); Sample treatment for protein identification by mass spectrometry-based techniques, Trends in Analytical Chemistry, 25(10): 996- 1005 Nahar MK, Zakaria Z, Hashim U (2013); Effect of Buffer on the Protein Extraction for Chicken Meat, Advanced Materials Research, 795: 206-10 Quintavalla S, Vicini L (2002); Antimicrobial food packaging in meat industry, Meat Science, 62(3):

373-80

Schnarr M, Maurizot JC (1981); Unfolding of lac repressor and its proteolytic fragment by urea: headpieces stabilize the core within lac repressor, Biochemistry, 20(21): 6164-9

Page 61: Masters Research Project

61

Skoog D, West DM (1980); Principles of Instrumental Analysis, Philadelphia : Saunders College Pub

Tangvarasittichai S, Tangvarasittichai O, Jermnim N (2009); Comparison of fast protein liquid

chromatography (FPLC) with HPLC, electrophoresis & microcolumn chromatography techniques for

the diagnosis of β-thalassaemia, Indian Journal of Medical Research, 129: 242-8

von Bargen C, Dojahn J, Waidelich D, Humpf HU, Brockmeyer J (2013); New Sensitive High-Performance Liquid Chromatography-Tandem Mass Spectrometry Method for the Detection of Horse and Pork in Halal Beef, Journal of Agricultural and Food Chemistry, 61: 11986-994 Wang M, Zhao XM, Tan H, Akutsu T, Whisstock JC, Song J (2013); Cascleave 2.0 a new approach for

predicting caspase and granzyme cleavage targets, Bioinformatics, 30(1): 71-80

Warne D (1988); Manual On Fish Canning, Food and Agricultural Organization Of The United Nations,

Italy: Rome

Watson AD, Gunning Y, Rigby NM, Philo M, Kemsley EK (2015); Meat Authentication via Multiple Reaction Monitoring Mass Spectrometry of Myoglobin Peptides, Analytical Chemistry, Article ASAP Yamazai I, Yokota K, Shikama K (1964); Preparation of Crystalline Oxymyoglobin from Horse Heart, The Journal of Biological Chemistry, 239(12): 4151-53 Yee S, Peyton DH (1995); Variable-temperature study of the heme-reorientation process in equine myoglobin, Biochimica et Biophysica Acta, 1252(2):295-9