Food Biochemistry and Food Processing (Simpson/Food Biochemistry and Food Processing) || Application of Proteomics to Fish Processing and Quality

P1: SFK/UKS P2: SFK

BLBS102-c22 BLBS102-Simpson March 21, 2012 13:41 Trim: 276mm X 219mm Printer Name: Yet to Come

22Application of Proteomics to Fish

Processing and QualityHolmfrıður Sveinsdottir, Samuel A. M. Martin, and Oddur T. Vilhelmsson

Proteomics MethodologyTwo-Dimensional Electrophoresis

Basic 2DE Methods OverviewSample Extraction and CleanupFirst-Dimension ElectrophoresisEquilibrationSecond-Dimension ElectrophoresisStainingAnalysisSome Problems and Their Solutions

Identification by Peptide Mass FingerprintingSeafood Proteomics and Their Relevance to Processing andQuality

Early Development and Proteomics of FishChanges in the Proteome of Early Cod Larvae in

Response to Environmental FactorsTracking Quality Changes Using ProteomicsAntemortem Effects on Quality and ProcessabilitySpecies AuthenticationIdentification and Characterization of Allergens

Impacts of High Throughput Genomic and ProteomicTechnologiesReferences

Abstract: Proteomics involves the study of proteins, with regards toproteins, their expression by genomes, their structures and functions.The entire set of proteins or proteome expressed by a genome displayvariations in tissues and organisms, and can be used as the basis forevaluating the status and changes in the proteins in living organismsincluding fish and shellfish. This feature can be useful for developingstandards for fish and other food materials and assessing their qualityand/or safety. This chapter discusses current uses of proteomics forestablishing the attributes of fish and fish products.

Proteomics is most succinctly defined as “the study of the entireproteome or a subset thereof,” the proteome being the expressedprotein complement of the genome. Unlike the genome, the

proteome varies among tissues, as well as with time in reflec-tion of the organism’s environment and its adaptation thereto.Proteomics can, therefore, give a snapshot of the organism’sstate of being and, in principle at least, map the entirety of itsadaptive potential and mechanisms. As with all living matter,foodstuffs are in large part made up of proteins. This is espe-cially true of fish and meat, where the bulk of the food matrixis constructed from proteins. Furthermore, the construction ofthe food matrix, both on the cellular and tissue-wide levels, isregulated and brought about by proteins. It stands to reason,then, that proteomics is a tool that can be of great value to thefood scientist, giving valuable insight into the composition ofthe raw materials, quality involution within the product before,during, and after processing or storage, the interactions of pro-teins with one another or with other food components, or withthe human immune system after consumption. In this chapter,a brief overview of “classical” proteomics methodology is pre-sented, and their present and future application in relation to fishand seafood processing and quality is discussed.

PROTEOMICS METHODOLOGYUnlike nucleic acids, proteins are an extremely variegated groupof compounds in terms of their chemical and physical proper-ties. It is not surprising, then, that a field that concerns itselfwith “the systematic identification and characterization of pro-teins for their structure, function, activity and molecular inter-actions” (Peng et al. 2003) should possess a toolkit containinga wide spectrum of methods that continue to be developed ata brisk pace. While high-throughput, gel-free methods, for ex-ample, based on liquid chromatography tandem mass spectrom-etry (LC-MS/MS) (Peng et al. 2003), surface-enhanced laserdesorption/ionization (Hogstrand et al. 2002), or protein arrays(Lee and Nagamune 2004), hold great promise and are deserv-ing of discussion in their own right, the “classic” process of

Food Biochemistry and Food Processing, Second Edition. Edited by Benjamin K. Simpson, Leo M.L. Nollet, Fidel Toldra, Soottawat Benjakul, Gopinadhan Paliyath and Y.H. Hui.C© 2012 John Wiley & Sons, Inc. Published 2012 by John Wiley & Sons, Inc.

406

P1: SFK/UKS P2: SFK


22 Application of Proteomics to Fish Processing and Quality 407

Figure 22.1. An overview over the “classic approach” in proteomics. First, a protein extract (crude or fractionated) from the tissue of choice issubjected to two-dimensional polyacrylamide gel electrophoresis (2D-PAGE). Once a protein of interest has been identified, it is excised fromthe gel, subjected to degradation by trypsin (or other suitable protease) and the resulting peptides analyzed by mass spectrometry (MS),yielding a peptide mass fingerprint. In many cases, this is sufficient for identification purposes, but if needed, peptides can be dissociated intosmaller fragments and small partial sequences obtained by tandem mass spectrometry (MS/MS). See text for further details.

two-dimensional electrophoresis (2DE) followed by proteinidentification via peptide mass fingerprinting of trypsin digests(Fig. 22.1) remains the workhorse of most proteomics work,largely because of its high resolution, simplicity, and mass accu-racy. This “classic approach” will, therefore, be the main focus ofthis chapter. A number of reviews on the advances and prospectsof proteomics within various fields of study are available. Somerecent ones include: Andersen and Mann (2006), Balestrieriet al. (2008), Beretta (2009), Bogyo and Cravatt (2007), Drabiket al. (2007), Ikonomou et al. (2009), Issaq and Veenstra (2008),Jorrin-Novo et al. (2009), Latterich et al. (2008), Lopez (2007),Mamone et al. (2009), Malmstrom et al. (2007), Premsler et al.(2009), Smith et al. (2009), Wang et al. (2006), Wilm (2009),Yates et al. (2009).

Two-Dimensional Electrophoresis

2DE, the cornerstone of most proteomics research, is the si-multaneous separation of hundreds, or even thousands, of pro-teins on a two-dimensional polyacrylamide slab gel. The po-tential of a two-dimensional protein separation technique wasrealized early on, with considerable development efforts takingplace in the 1960s (Margolis and Kenrick 1969, Kaltschmidtand Wittmann 1970). The method most commonly used todaywas developed by Patrick O’Farrell. It is described in his seminaland thorough 1975 paper (O’Farrell 1975) and is outlined brieflylater. It is worth emphasizing that great care must be taken thatthe proteome under investigation is reproducibly represented onthe 2DE gels, and that individual variation in specific protein

P1: SFK/UKS P2: SFK


408 Part 3: Meat, Poultry and Seafoods

67 kDa

43 kDa

30 kDa

21 kDa

14 kDa

4 pI 7

Figure 22.2. A two-dimensional electrophoresis protein map ofrainbow trout (Oncorhynchus mykiss) liver proteins with pI between4 and 7 and molecular mass about 10–100 (S. Martin, unpublished).The proteins are separated according to their pI in the horizontaldimension and according to their mass in the vertical dimension.Isoelectro focussing was by pH 4–7 immobilized pH gradient (IPG)strip and the second dimension was in a 10–15% gradientpolyacrylamide slab gel.

abundance is taken into consideration by running gels from asufficient number of samples and performing the appropriatestatistics. Pooling samples may also be an option, depending onthe type of experiment.

Basic 2DE Methods Overview

O’Farrell’s original 2DE method first applies a process calledisoelectric focusing (IEF), where an electric field is applied toa tube gel on which the protein sample and carrier ampholyteshave been deposited. This separates the proteins according totheir molecular charge. The tube gel is then transferred ontoa polyacrylamide slab gel and the isoelectrically focused pro-teins are further separated according to their molecular massby conventional sodium dodecyl sulfate–polyacrylamide gelelectrophoresis (SDS-PAGE), yielding a two-dimensional map(Fig. 22.2) rather than the familiar banding pattern observed inone-dimensional SDS-PAGE. The map can be visualized andindividual proteins quantified by radiolabeling or by using anyof a host of protein dyes and stains, such as Coomassie blue,silver stains, or fluorescent dyes. By comparing the abundanceof individual proteins on a number of gels (Fig. 22.3), upregu-lation or downregulation of these proteins can be inferred. Al-though a number of refinements have been made to 2DE sinceO’Farrell’s paper, most notably, the introduction of immobilizedpH gradients (IPGs) for IEF (Gorg et al. 1988), the procedure

Figure 22.3. A screenshot from the two-dimensional electrophoresis analysis program Phoretix 2-D (NonLinear Dynamics, Gateshead, Tyne& Wear, UK) showing some steps in the analysis of a two-dimensional protein map. Variations in abundance of individual proteins, ascompared with a reference gel, can be observed and quantified.

P1: SFK/UKS P2: SFK



remains essentially as outlined earlier. In the following sections,a general protocol is outlined briefly with some notes of specialrelevance to the seafood scientist. For more detailed, up-to-dateprotocols, the reader is referred to any of a number of excellentreviews and laboratory manuals such as Berkelman and Stenst-edt (1998), Gorg et al. (2000, 2004), Kraj and Silberring (2008),Link (1999), Simpson (2003), Walker (2005) and Westermeierand Naven (2002).

Sample Extraction and Cleanup

For most applications, sample treatment prior to electrophoresisshould be minimal in order to minimize in-sample proteolysisand other sources of experimental artifacts. We have found directextraction into the gel reswelling buffer (7-M urea, 2-M thiourea,4% (w/v) CHAPS [3-(3-chloramidopropyl)dimethylamino-1-propanesulfonate], 0.3% (w/v) DTT [dithiothreitol], 0,5%Pharmalyte ampholytes for the appropriate pH range), sup-plemented with a protease inhibitor cocktail, to give goodresults for proteome extraction from whole Atlantic cod larvae(Guðmundsdottir and Sveinsdottir 2006, Sveinsdottir et al.2008) and Arctic charr (Salvelinus alpinus) liver (Coe andVilhelmsson 2008). Thorough homogenization is essential toensure complete and reproducible extraction of the proteome.Cleanup of samples using commercial two-dimensional samplecleanup kits may be beneficial for some sample types.

First-Dimension Electrophoresis

The extracted proteins are first separated by IEF, which is mostconveniently performed using commercial dry IPG gel strips.These strips consist of a dried IPG-containing polyacrylamidegel on a plastic backing. Ready-made IPG strips are currentlyavailable in a variety of linear and sigmoidal pH ranges. Thismethod is thus suitable for most 2DE applications and has allbut completely replaced the older and less reproducible methodof IEF by carrier ampholytes in tube gels. Broad-range linearstrips (e.g., pH 3–10) are commonly used for whole-proteomeanalysis of tissue samples, but for many applications narrow-range and/or sigmoidal IPG strips may be more appropriate asthese will give better resolution of proteins in the fairly crowdedpI 4–7 range. Narrow-range strips also allow for higher sampleloads (since part of the sample will run off the gel) and thus mayyield improved detection of low-abundance proteins.

Before electrophoresis, the dried gel needs to be reswelled toits original volume. A recipe for a typical reswelling buffer ispresented earlier. Reswelling is normally performed overnightat 4◦C. Application of a low-voltage current may speed up thereswelling process. Optimal conditions for reswelling are nor-mally provided by the IPG strip manufacturer. If the proteinsample is to be applied during the reswelling process, extractiondirectly into the reswelling buffer is recommended.

IEF is normally performed for several hours at high voltageand low current. Typically, the starting voltage is about 150V, which is then increased step-wise to about 3500 V, usuallytotaling about 10,000 to 30,000 Vh, although this will dependon the IPG gradient and the length of the strip. The appropriate

IEF protocol will depend not only on the sample and IPG strip,but also on the equipment used. The manufacturer’s instructionsshould be followed. Gorg et al. (2000) reviewed IEF for 2DEapplications.

Equilibration

Before the isoelectrofocused gel strip can be applied to thesecond-dimension slab gel, it needs to be equilibrated for 30–45minutes in a buffer containing SDS and a reducing agent such asDTT. During the equilibration step, the SDS–polypeptide com-plex that affords protein-size-based separation will form and thereducing agent will preserve the reduced state of the proteins. Atracking dye for the second electrophoresis step is also normallyadded at this point. A typical equilibration-buffer recipe is asfollows: 50 mM Tris-HCl at pH 8.8, 6-M urea, 30% glycerol,2% SDS, 1% DTT, and trace amount of bromophenol blue. Asecond equilibration step in the presence of 2.5% iodoacetamideand without DTT (otherwise identical buffer) may be requiredfor some applications. This will alkylate thiol groups and preventtheir reoxidation during electrophoresis, thus reducing verticalstreaking (Gorg et al. 1987).

Second-Dimension Electrophoresis

Once the gel strip has been equilibrated, it is applied to thetop edge of an SDS-PAGE slab gel and cemented in place us-ing a molten agarose solution. Optimal pore size depends onthe size of the target proteins, but for most applications gra-dient gels or gels of about 10% or 12% polyacrylamide areappropriate. Ready-made gels suitable for analytical 2DE areavailable commercially. While some reviewers recommend al-ternative buffer systems (Walsh and Herbert 1999), the Laemmlimethod (Laemmli 1970), using glycine as the trailing ion andthe same buffer (25-mM Tris, 192-mM glycine, 0.1% SDS) atboth electrodes, remains the most popular one. The gel is runat a constant current of 25 mA until the bromophenol blue dyefront has reached the bottom of the gel.

Staining

Visualization of proteins spots is commonly achieved throughstaining with colloidal Coomassie Blue G-250 due to its lowcost and ease of use. A typical staining procedure includes fixingthe gel for several hours in 50% ethanol/2% ortho-phosphoricacid, followed by several 30-minute washing steps in water, fol-lowed by incubation for 1 hour in 17% ammonium sulfate/34%methanol/2% ortho-phosphoric acid, followed by staining forseveral days in 0.1% Coomassie Blue G-250/17% ammoniumsulfate/34% methanol/2% ortho-phosphoric acid, followed bydestaining for several hours in water. There are, however, com-mercially available colloidal Coomassie staining kits that do notrequire fixation or destaining.

A great many alternative visualization methods are available,many of which are more sensitive than colloidal Coomassieand thus may be more suitable for applications where visual-ization of low-abundance proteins is important. These include

P1: SFK/UKS P2: SFK



radiolabeling, such as with [35S]methionine, and staining withfluorescent dyes such as the SYPRO or Cy series of dyes. Multi-ple staining with dyes fluorescing at different wavelengths offersthe possibility of differential display, allowing more than oneproteome to be compared on the same gel, such as in differencegel electrophoresis (DIGE). Patton published a detailed reviewof visualization techniques for proteomics (Patton 2002).

Analysis

Although commercial 2DE image analysis software, such asImageMaster (Amersham), PDQuest (BioRad), or Progenesis(Nonlinear Dynamics), has improved by leaps and bounds inrecent years, analysis of the 2DE gel image, including proteinspot definition, matching, and individual protein quantification,remains the bottleneck of 2DE-based proteome analysis and stillrequires a substantial amount of subjective input by the investi-gator (Barrett et al. 2005). In particular, spot matching betweengels tends to be time consuming and has proved difficult toautomate (Wheelock and Goto 2006). These difficulties arisefrom several sources of variation among individual gels, suchas protein load variability due to varying IPG strip reswellingor protein transfer from strip to slab gel. Also, gene expressionin several tissues varies considerably among individuals of thesame species, and therefore individual variation is a major con-cern and needs to be accounted for in any statistical treatmentof the data. Pooling samples may also be an option, dependingon the type of experiment. These multiple sources of varia-tion has led some investigators (Barrett et al. 2005, Karp et al.2005, Wheelock and Goto 2006) to cast doubt on the suitabil-ity of univariate tests such as the Student’s t-test, commonlyused to assess the significance of observed protein expressiondifferences. Multivariate analysis has been successfully used byseveral investigators in recent years (Gustafson et al. 2004, Karpet al. 2005, Kjaersgard et al. 2006b).

Some Problems and Their Solutions

The high resolution and good sensitivity of 2DE are what makeit the method of choice for most proteomics work, but themethod nevertheless has several drawbacks. The most signif-icant of these have to do with the diversity of proteins andtheir expression levels. For example, hydrophobic proteins donot readily dissolve in the buffers used for isoelectrofocussing.This problem can be overcome, though, using nonionic orzwitterionic detergents, allowing for 2DE of membrane- andmembrane-associated proteins (Chevallet et al. 1998, Herbert1999, Henningsen et al. 2002, Babu et al. 2004). Vilhelmssonand Miller (2002), for example, were able to use “membrane pro-tein proteomics” to demonstrate the involvement of membrane-associated metabolic enzymes in the osmoadaptive response ofthe foodborne pathogen Staphylococcus aureus. A 2DE gel im-age of S. aureus membrane-associated gels is shown in Figure22.4.

Similarly, resolving alkaline proteins, particularly those withpI above 10, on two-dimensional gels has been problematic in the

24 kDa

36 kDa

45 kDa

55 kDa

66 kDa

84 kDa

97 kDa

3 10pI

Figure 22.4. A two-dimensional electrophoresis membraneproteome map from Staphylococcus aureus, showing proteins withpI between 3 and 10 and molecular mass about 15–100(O. Vilhelmsson and K. Miller, unpublished). Isoelectrofocussing wasin the presence of a mixture of pH 5–7 and pH 3–10 carrierampholytes and the second dimension was in a 10%polyacrylamide slab gel with a 4% polyacrylamide stacker.

past. Although the development of highly alkaline, narrow-rangeIPGs (Bossi et al. 1994) allowed reproducible two-dimensionalresolution of alkaline proteins (Gorg et al. 1997), their repre-sentation on wide-range 2DE of complex mixtures such as cellextracts remained poor. Improvements in resolution and repre-sentation of alkaline proteins on wide-range gels have been made(Gorg et al. 1999), but nevertheless an approach that involvesseveral gels, each of a different pH range, from the same sam-ple is advocated for representative inclusion of alkaline proteinswhen studying entire proteomes (Cordwell et al. 2000). Indeed,Cordwell and coworkers were able to significantly improve therepresentation of alkaline proteins in their study on the rela-tively highly alkaline Helicobacter pylori proteome using bothpH 6–11 and pH 9–12 IPGs (Bae et al. 2003).

A second drawback of 2DE has to do with the extreme differ-ence in expression levels of the cell’s various proteins, whichcan be as much as 10,000-fold. This leads to swamping oflow-abundance proteins by high-abundance ones on the two-dimensional map, rendering analysis of low-abundance proteinsdifficult or impossible. For applications such as species iden-tification or study of the major biochemical pathways, wherethe proteins of interest are present in relatively high abundance,

P1: SFK/UKS P2: SFK



this does not present a problem. However, when investigating,for example, regulatory cascades, the proteins of interest arelikely to be present in very low abundance and may at timesbe undetectable because of the dominance of high-abundanceones. Simply increasing the amount of sample is usually notan option, as it will give rise to overloading artifacts in thegels (O’Farrell 1975). In transcriptomic studies, where a simi-lar disparity can be seen in the abundance of RNA transcriptspresent, this problem can be overcome by amplifying the low-abundance transcripts using the polymerase chain-reaction, butno such technique is available for proteins. The remaining op-tion, then, is fractionation of the protein sample in order to weedout the high-abundance proteins, allowing a larger sample of theremaining proteins to be analyzed. A large number of fraction-ation protocols, both specific and general, are available. Thus,Østergaard and coworkers used acetone precipitation to reducethe abundance of hordeins present in barley (Hordeum vulgare)extracts (Østergaard et al. 2002) whereas Locke and cowork-ers used preparative isoelectrofocussing to fractionate Chinesesnow pea (Pisum sativum macrocarpon) lysates into fractionscovering three pH regions (Locke et al. 2002). The fractionationmethod of choice will depend on the specific requirements ofthe study and on the tissue being studied. Discussion of somefractionation methods can be found in Ahmed (2009), Bodzon-Kulakowska et al. (2007), Butt et al. (2001), Canas et al. (2007),Corthals et al. (1997), Dreger (2003), Fortis et al. (2008), Issaq

et al. (2002), Lee and Pi (2009), Lopez et al. (2000), Milleaand Krull (2003), Pieper et al. (2003), Righetti et al. (2005a, b)Rothemund et al. (2003), von Horsten (2006).

Identification by Peptide Mass Fingerprinting

Identification of proteins on 2DE gels is most commonlyachieved via mass spectrometry (MS) of trypsin digests. Briefly,the spot of interest is excised from the gel, digested with trypsin(or another protease), and the resulting peptide mixture is ana-lyzed by MS. The most popular MS method is matrix-assistedlaser desorption/ionization time-of-flight (MALDI-TOF) MS(Courchesne and Patterson 1999), where peptides are suspendedin a matrix of small, organic, ultraviolet-absorbing molecules(such as 2,5-dihydroxybenzoic acid) followed by ionization bya laser at the excitation wavelength of the matrix molecules andacceleration of the ionized peptides in an electrostatic field intoa flight tube where the time of flight of each peptide is measured,giving its expected mass.

The resulting spectrum of peptide masses (Fig. 22.5) is thenused for protein identification by searching against expected pep-tide masses calculated from data in protein sequence databases,such as SwissProt or the National Centre for Biotechnology In-formation (NCBI) nonredundant protein sequences data base,using the appropriate software. Several programs are available,many with a web-based open-access interface. The ExPASy

Inte

nsity

Mass (m/z)

Figure 22.5. A trypsin digest mass spectrometry fingerprint of a rainbow trout liver protein spot, identified as apolipoprotein A I-1 (S. Martin,unpublished). The open arrows indicate mass peaks corresponding to trypsin self-digestion products and were, therefore, excluded from theanalysis. The solid arrows indicate the peaks that were found to correspond to expected apolipoprotein A I-1 peptides.

P1: SFK/UKS P2: SFK



Tools website (http://www.expasy.org/tools) contains links tomost of the available software for protein identification and sev-eral other tools. Attaining a high identification rate is problematicin fish and seafood proteomics due to the relative paucity of avail-able protein sequence data for these animals. As can be seen inTable 22.1, this problem is surprisingly acute for species of com-mercial importance. To circumvent this problem, it is possible totake advantage of the available nucleotide sequences, which inmany cases is more extensive than the protein sequences avail-able, to obtain a tentative identity. How useful this method iswill depend on the length and quality of the available nucleotidesequences. It is important to realize, however, that an identityobtained in this manner is less reliable than that obtained throughprotein sequences and should be regarded only as tentative inthe absence of corroborating evidence (such as two-dimensionalimmunoblots, correlated activity measurements, or transcriptabundance). In their work on the rainbow trout (Oncorhynchusmykiss) liver proteome, Martin et al. (2003b) and Vilhelmssonet al. (2004) were able to attain an identification rate of about80% using a combination of search algorithms that includedthe open-access Mascot program (Perkins et al. 1999) and alicensed version of Protein Prospector MS-Fit (Clauser et al.1999), searching against both protein databases and a databasecontaining all salmonid nucleotide sequences. In those caseswhere both the protein and nucleotide databases yielded results,a 100% agreement was observed between the two methods.

A more direct, if rather more time-consuming, way of obtain-ing protein identities is by direct sequence comparison. Untilrecently, this was accomplished by N-terminal or internal (afterproteolysis) sequencing by Edman degradation of eluted or elec-troblotted protein spots (Kamo and Tsugita 1999, Erdjument-Bromage et al. 1999). Today, the method of choice is tandemmass spectrometry (MS/MS). In the peptide mass fingerprintingdiscussed earlier, each peptide mass can potentially representany of a large number of possible amino acid sequence com-binations. The larger the mass (and longer the sequence), thehigher the number of possible combinations. In MS/MS, oneor several peptides are separated from the mixture and dissoci-ated into fragments that then are subjected to a second roundof MS, yielding a second layer of information. Correlating thisspectrum with the candidate peptides identified in the first roundnarrows down the number of candidates. Furthermore, severalshort stretches of amino acid sequence will be obtained for eachpeptide, which, when combined with the peptide and fragmentmasses obtained, enhances the specificity of the method evenfurther (Wilm et al. 1996, Chelius et al. 2003, Yu et al. 2003b, ).MS methods in proteomics are reviewed in Yates (1998, 2004)and Yates et al. (2009).

SEAFOOD PROTEOMICS ANDTHEIR RELEVANCE TO PROCESSINGAND QUALITY2DE-based proteomics have found a number of applicationsin food science. Among early examples are such applicationsas characterization of bovine caseins (Zeece et al. 1989), wheat

flour baking quality factors (Dougherty et al. 1990), and soybeanprotein bodies (Lei and Reeck 1987). In recent years, proteomicinvestigations on fish and seafood products, as well as in fishphysiology, have gained considerable momentum, as can be seenin recent reviews (Parrington and Coward 2002, Pineiro et al.2003). Herein, we consider recent and future developments infish and seafood proteomics as related to issues of concern infish processing or other quality considerations.

Early Development and Proteomics of Fish

Fishes go through different developmental stages (embryo, larva,and adult) during their lifespan that coincide with changes in themorphology, physiology, and behavior of the fish (O’Connell1981, Govoni et al. 1986, Skiftesvik 1992, Osse and van denBoogaart 1995). The morphological and physiological changesthat occur during these developmental stages are characterizedby differential cellular and organelle functions (Einarsdottir et al.2006). This is reflected in variations of global protein expres-sion and posttranslational modifications of the proteins that maycause alterations of protein function (Campinho et al. 2006).

Proteome analysis provides valuable information on the vari-ations that occur within the proteome of organisms. These vari-ations may, for example, reflect a response to biological per-turbations or external stimuli (Anderson and Anderson 1998,Martin et al. 2001, Martin et al. 2003b, Vilhelmsson et al. 2004)resulting in different expression of proteins, posttranslationalmodifications, or redistribution of specific proteins within cells(Tyers and Mann 2003). To date, few studies on fish developmentexist in which proteome analysis techniques have been applied.Recent studies on global protein expression during early devel-opmental stages of zebrafish (Tay et al. 2006) and Atlantic cod(Sveinsdottir et al. 2008) revealed that distinctive protein pro-files characterize the developmental stages of these fishes eventhough abundant proteins are largely conserved during the ex-perimental period. In both these studies, the identified proteinsconsisted mainly of proteins located in the cytosol, cytoskeleton,and nucleus. Proteome analyses in developing organisms haveshown that many of the identified proteins have multiple iso-forms (Paz et al. 2006), reflecting either different gene products(Guðmundsdottir et al. 1993) or posttranslationally modifiedforms of these proteins (Jensen 2004). Different isoforms gen-erated by posttranslational modifications are largely overlookedby studies based on RNA expression. This fact further indicatesthe importance of the proteome approach to understand cellularmechanisms underlying fish development. Studies on variousproteins have shown that during fish development sequentialsynthesis of different isoforms appear successively (Huriauxet al. 1996, Galloway et al. 1998, Focant et al. 1999, Gallowayet al. 1999, Huriaux et al. 1999, Focant et al. 2000, Huriaux et al.2003, Focant et al. 2003, Hall et al. 2003, Galloway et al. 2006,Campinho et al. 2006, Campinho et al. 2007). In this context, de-velopmental stage-specific muscle protein isoforms have gaineda special attention (Huriaux et al. 1996, Galloway et al. 1998,Focant et al. 1999, Galloway et al. 1999, Huriaux et al. 1999,Focant et al. 2000, Focant et al. 2003, Hall et al. 2003, Huriauxet al. 2003, Galloway et al. 2006, Campinho et al. 2007).

P1: SFK/UKS P2: SFK



Table 22.1. Some Commercially or Scientifically Important Fish and Seafood Species and the Availability of Proteinand Nucleotide Sequence Data as of February 18, 2010, According to the NCBI TaxBrowser(a)

ProteinSequences

NucleotideSequences

ProteinSequences

NucleotideSequences

Actinopterygii (ray-finned fishes) 257,843 1,586,862 Tetraodontiformes (puffers andfilefishes)

34,032 114,496

Elopomorpha 3,218 4,459 Pufferfish (Takifugu rubripes) 2,593 16,406Anguilliformes (eels and morays) 2,898 4,144 Green pufferfish (Tetraodon

nigroviridis)28,483 126,200

European eel (Anguilla anguilla) 227 404 Zeiformes (dories) 138 126Clupeomorpha 2,333 4,697 John Dory (Zeus faber) 102 100Clupeiformes (herrings) 2,333 4,697 Scorpaeniformes

(scorpionfishes/flatheads)4,858 7,896

Atlantic herring (Clupea harengus) 105 166 Redfish (Sebastes marines) 13 17European pilchard (Sardina

pilchardus)79 401 Lumpsucker (Cyclopterus lumpus) 20 30

Ostariophysii 104,088 211,491Cypriniformes (carps) 94,392 196,415 Chondrichthyes (cartilagenous

fishes)6,407 692,476

Zebrafish (Danio rerio) 70,151 163,995 Carcharhiniformes (ground sharks) 1,812 2,234Siluriformes (catfishes) 7,149 10,472 Lesser spotted dogfish

(Scyliorhinus canicula)261 197

Channel catfish (Ictaluruspunctatus)

1,890 1,996 Blue shark (Prionace glauca) 48 52

Protacanthopterygii 32,187 42,475 Lamniformes (mackrel sharks) 554 676Salmoniformes (salmons) 31,244 42,278 Basking shark (Cetorhinus

maximus)66 72

Atlantic salmon (Salmo salar) 16,230 19,107 Rajiformes (skates) 848 1,050Rainbow trout (Oncorhynchus

mykiss)6,216 8,361 Thorny skate (Raja radiata) 41 60

Arctic charr (Salvelinus alpinus) 257 465 Blue skate (Raja batis) 7 1Paracanthopterygii 5,396 6,397 Little skate (Raja erinacea) 201 185Gadiformes (cods) 4,517 5,573Atlantic cod (Gadus morhua) 2,106 1,964 Mollusca (mollusks) 52,622 154,100Alaska pollock (Theragra

chalcogramma)382 303 Bivalvia 14,572 25,149

Saithe (Pollachius virens) 51 53 Blue mussel (Mytilus edulis) 1,307 1,266Haddock (Melanogrammus

aeglefinus)189 151 Bay scallop (Argopecten irradians) 194 380

Lophiiformes (anglerfishes) 412 471 Gastropoda 32,489 121,832Monkfish (Lophius piscatorius) 40 104 Common whelk (Buccinum

undatum)6 11

Acanthopterygii 104,101 1,369,306 Abalone (Haliotis tuberculata) 133 279Perciformes (perch like) 44,027 619,647 Cephalopoda 3,964 5,045Gilthead sea bream (Sparus aurata) 516 956 Northern European squid (Loligo

forbesi)34 49

European sea bass (Dicentrachuslabrax)

466 36,916 Common cuttlefish (Sepiaofficinalis)

351 335

Atlantic mackrel (Scomberscombrus)

111 129 Common octopus (Octopusvulgaris)

169 204

Albacore (Thunnus alalunga) 116 227Bluefin tuna (Thunnus thynnus) 179 935 Crustacea (crustaceans) 44,586 70,954Spotted wolffish (Anarhichas minor) 38 23 Caridea 5,103 7,630Beryciformes (sawbellies) 794 616 Northern shrimp (Pandalus

borealis)19 26

P1: SFK/UKS P2: SFK



Table 22.1. (Continued)

ProteinSequences

NucleotideSequences

ProteinSequences

NucleotideSequences

Orange roughy (Hoplostethusatlanticus)

11 40 Astacidea (lobsters and crayfishes) 1,742 4,466

Pleuronectiformes (flatfishes) 4,109 11,851 American lobster (Homerusamericanus)

195 173

Atlantic halibut (Hippoglossushippoglossus)

218 2,958 European crayfish (Astacusastacus)

28 39

Witch (Glyptocephalus cynoglossus) 22 54 Langoustine (Nephropsnorvegicus)

19 35

Plaice (Pleuronectes platessa) 88 285 Brachyura (short-tailed crabs) 5,143 8,238Winter flounder

(Pseudopleuronectes americanus)139 196 Edible crab (Cancer pagurus) 37 38

Turbot (Scophthalmus maximus) 385 1,014 Blue crab (Callinectes sapidus) 123 133

Source: (a)http://www.ncbi.nlm.nih.gov/Taxonomy/taxonomyhome.html/.

The developmental changes in the composition of muscle pro-tein isoforms have been tracked by proteome analysis in Africancatfish (Heterobranchus longifilis) (Focant et al. 1999), com-mon sole (Solea solea) (Focant et al. 2003), and dorada (Bryconmoorei) (Huriaux et al. 2003). These studies demonstrated thatthe muscle shows the usual sequential synthesis of protein iso-forms in the course of development. For example, in the commonsole, 2DE revealed two isoforms (larval and adult) of myosinlight chain 2, and likewise in dorada larval and adult isoformsof troponin I were sequentially expressed during development.

Proteomic techniques have, thus, been shown to be applicablefor investigating cellular and molecular mechanisms involved inthe morphological and physiological changes that occur duringfish development.

The major obstacle on the use of proteomics in embryonic fishhas been the high proportion of yolk proteins. These interferewith any proteomic application that intends to target the cellsof the embryo proper. In a recent study on the proteome ofembryonic zebrafish, the embryos were deyolked to enrich thepool of embryonic proteins and to minimize ions and lipidsfound in the yolk prior to two-dimensional gel analysis (Tayet al. 2006). Despite this undertaking, a large number of yolkproteins remained prominently present in the embryonic proteinprofiles. Link et al. (2006) published a method to efficientlyremove the yolk from large batches of embryos without losingcellular proteins. The success in the removal of yolk proteinsby Link et al. (2006) is probably due to dechorionation prior tothe deyolking of the embryos. By dechorionation, the embryosfall out of their chorions facilitating the removal of the yolk(Huriaux et al. 1996, Westerfield 2000).

Changes in the Proteome of Early Cod Larvaein Response to Environmental Factors

The production of good quality larvae is still a challenge in ma-rine fish hatcheries. Several environmental factors can interferewith the protein expression of larvae affecting larval quality like

growth and survival rate. Proteome analysis allows us to exam-ine the effects of environmental factors on larval global proteinexpression, posttranslational modifications, and redistribution ofspecific proteins within cells (Tyers and Mann 2003), all impor-tant information for controlling factors influencing the aptitudeto continue a normal development until adult stages.

A variety of environmental factors have shown to improvethe health and survival of fish larvae, including probiotic bac-teria (Tinh et al. 2008) and protein hydrolysates (Cahu et al.1999). However, the beneficial effects of these treatments onfish larvae are poorly understood at the molecular level. Onlya few proteome analysis studies on fish larvae have been pub-lished (Focant et al. 2003, Tay et al. 2006, Guðmundsdottirand Sveinsdottir 2006, Sveinsdottir et al. 2008, Sveinsdottirand Guðmundsdottir 2008, Sveinsdottir et al. 2009), of whichtwo have focused on the changes in the whole larval pro-teome after treatment with probiotic bacteria (Sveinsdottir et al.2009) and protein hydrolysate (Sveinsdottir and Gudmundsdottir2008). These studies provide protocols for the production ofhigh-resolution two-dimensional gels of whole larval proteome,where peptide mass mapping (MALDI-TOF MS) and peptidefragment mapping (LC-MS/MS) allowed identification of ca.85% of the of the selected cod protein spots (Guðmundsdottirand Sveinsdottir 2006, Sveinsdottir and Gudmundsdottir 2008,Sveinsdottir et al. 2008, Sveinsdottir et al. 2009).

The advantages of working with whole larvae versus distincttissues is the ease of keeping the sample handling to a minimumin order to avoid loss or modification of the proteins. Neverthe-less, there are several drawbacks when working with the wholelarval proteome, like the overwhelming presence of muscle andskin proteins. These proteins may mask subtle changes in pro-teins expressed in other tissues or systems, such as the gastroin-testinal tract or the central nervous system caused by variousenvironmental factors. The axial musculature is the largest tis-sue in larval fishes as it constitutes approximately 40% of theirbody mass (Osse and van den Boogaart 1995). This is reflectedin our studies on whole cod larval proteome, where the majority

P1: SFK/UKS P2: SFK



of the highly abundant proteins were identified as muscle pro-teins (Guðmundsdottir and Sveinsdottir 2006, Sveinsdottir andGudmundsdottir 2008, Sveinsdottir et al. 2008, 2009). Also,cytoskeletal proteins were prominent among the identified pro-teins. Removal of those proteins may increase detection of otherproteins present at low concentrations. However, it may alsoresult in a loss of other proteins, preventing identification ofholistic alterations in the analyzed proteomes. Various strategieshave been presented for the removal of highly abundant proteins(Ahmed et al. 2003) or enrichment of low-abundance proteins(Oda et al. 2001, Ahmed and Rice 2005).

Tracking Quality Changes Using Proteomics

A persistent problem in the seafood industry is postmortemdegradation of fish muscle during chilled storage, which hasdeleterious effects on the fish flesh texture, yielding a tender-ized muscle. This phenomenon is thought to be primarily due toautolysis of muscle proteins, but the details of this protein degra-dation are still somewhat in the dark. However, degradation ofmyofibrillar proteins by calpains and cathepsins (Ogata et al.1998, Ladrat et al. 2000) and degradation of the extracellularmatrix by the matrix metalloproteases and matrix serine pro-teases, capable of degrading collagens, proteoglycans, and othermatrix components (Woessner 1991, Lødemel and Olsen 2003),are thought to be among the main culprits. Whatever the mecha-nism, it is clear that these quality changes are species dependent(Papa et al. 1996, Verrez-Bagnis et al. 1999) and, furthermore,appear to display seasonal variations (Ingolfsdottir et al. 1998,Ladrat et al. 2000). For example, whereas desmin is degradedpostmortem in sardine and turbot, no desmin degradation wasobserved in sea bass and brown trout (Verrez-Bagnis et al. 1999).Of further concern is the fact that several commercially impor-tant fish muscle processing techniques, such as curing, fermen-tation, and production of surimi and conserves occur under con-ditions conducive to endogenous proteolysis (Perez-Borla et al.2002). As with postmortem protein degradation during storage,autolysis during processing seems to be somewhat specific. In-deed, the myosin heavy chain of the Atlantic cod was shownto be significantly degraded during processing of “salt fish”(bacalhau) whereas actin was less affected (Thorarinsdottir et al.2002). Problems of this kind, where differences are expected tooccur in the number, molecular mass, and pI of the proteinpresent in a tissue, are well suited to investigation using 2DE-based proteomics. It is also worth noting that protein isoformsother than proteolytic ones, whether they be encoded in struc-tural genes or brought about by posttranslational modification,usually have different molecular weight or pI and can, therefore,be distinguished on 2DE gels. Thus, specific isoforms of my-ofibrillar proteins, many of which are correlated with specifictextural properties in seafood products, can be observed using2DE or other proteomic methods (Martinez et al. 1990, Pineiroet al. 2003).

Several 2DE studies have been performed on postmortemchanges in seafood flesh (Verrez-Bagnis et al. 1999, Morzelet al. 2000, Martinez et al. 2001a, Kjaersgard and Jessen 2003,2004, Martinez and Jakobsen Friis 2004, Kjaersgard et al.

2006a, b) and have demonstrated the importance and complex-ity of proteolysis in seafood during storage and processing. Forexample, Martinez and Jakobsen Friis used a 2DE approachto demonstrate different protein composition of surimi madefrom prerigor versus postrigor cod (Martinez and Jakobsen Friis2004). They found that 2DE could be used as a diagnostic toolto indicate the freshness of the raw material used for surimi pro-duction, a finding of considerable economic and public healthinterest. Kjaersgard and Jessen, who used 2DE to study changesin abundance of several muscle proteins during storage of theAtlantic cod (Gadus morhua), proposed a general model forpostmortem protein degradation in fish flesh where initially cal-pains were activated due to the increase in calcium levels inthe muscle tissue. Later, as pH decreases and ATP is depletedwith the consequent onset of rigor mortis, cathepsins and theproteasome are activated sequentially (Kjaersgard and Jessen2003).

Antemortem Effects on Quality andProcessability

Malcolm Love started his 1980 review paper on biological fac-tors affecting fish processing (Love 1980) with a lament for theeasy life of poultry processors who, he said, had the good for-tune to work on a product reared from hatching under strictlycontrolled environmental and dietary conditions “so that plasticbundles of almost identical foodstuff for man can be lined upon the shelf of a shop.” Since the time of Love’s review, theadvent of aquaculture has made attainable, in theory at least,just such a utopic vision. As every food processor knows, thequality of the raw material is among the most crucial variablesthat affect the quality of the final product. In fish processing,therefore, the animal’s own individual physiological status willto a large extent dictate where quality characteristics will fallwithin the constraints set by the species’ physical and biochem-ical makeup. It is well known that an organism’s phenotype,including quality characteristics, is determined by environmen-tal as well as genetic factors. Indeed, Huss noted in his review(Huss 1995) that product quality differences within the samefish species can depend on feeding and rearing conditions, dif-ferences wherein can affect postmortem biochemical processesin the product, which, in turn, affect the involution of qual-ity characteristics in the fish product. The practice of rearingfish in aquaculture, as opposed to wild-fish catching, thereforeraises the tantalizing prospect of managing quality characteris-tics of the fish flesh antemortem, where individual physiologicalcharacteristics, such as those governing gaping tendency, fleshsoftening during storage, etc., are optimized. To achieve thatgoal, the interplay between these physiological parameters andenvironmental and dietary variables needs to be understood indetail. With the ever-increasing resolving power of moleculartechniques, such as proteomics, this is fast becoming feasible.

In mammals, antemortem protease activities have been shownto affect meat quality and texture (Vaneenaeme et al. 1994, Kris-tensen et al. 2002). For example, an antemortem upregulation ofcalpain activity in swine (Sus scrofa) will affect postmortem pro-teolysis and, hence, meat tenderization (Kristensen et al. 2002).

P1: SFK/UKS P2: SFK



In beef (Bos taurus), a correlation was found between ante-mortem and postmortem activities of some proteases, but notothers (Vaneenaeme et al. 1994). As discussed in the earliersection, postmortem proteolysis is a matter of considerable im-portance in the fish and seafood industry and any antemortemeffects thereupon are surely worth investigating.

In a recent study on the feasibility of substituting fish meal inrainbow trout (O. mykiss) diets with protein from plant sources,2DE-based proteomics were among the techniques used (Martinet al. 2003a, b, Vilhelmsson et al. 2004). Concomitantly, variousquality characteristics of fillet and body were also measured (De

Francesco et al. 2004, Parisi et al. 2004). Among the findingswas that, according to a triangular sensory test using a trainedpanel, cooked 5-day-old trout that had been fed the plant pro-tein diet had higher hardness, lower juiciness, and lower odorintensity than those fed the fish meal-containing diet, indicatingan effect of antemortem metabolism on product texture. Thestudy identified a number of metabolic pathways sensitive toplant protein substitution in rainbow trout feed, such as path-ways involved in cellular protein degradation, fatty acid break-down, and NADPH (nicotinamide adenine dinucleotide phos-phate) metabolism (Table 22.2). In the context of this chapter,

Table 22.2. Protein Spots Affected by Dietary Plant Protein Substitution in Rainbow Trout as Judged byTwo-dimensional electrophoresis and Their Identities as Determined by Trypsin Digest Mass Fingerprinting

SpotReferenceNumber pI

MW(kDa)

NormalizedVolume

Diet FMa

NormalizedVolume Diet

PP100aFold

Difference P

Downregulated:

128 6.3 66 303 ± 57 60 ± 19 Vacuolar ATPase β-subunit 5 0.026291 6.4 42 521 ± 37 273 ± 30 β-ureidopropionase 2 0.004356 6.3 38 161 ± 37 44 ± 19 Transaldolase 4 0.031747 5.6 43 101 ± 19 19 ± 11 β-actin 2 0.040760 6.3 39 41 ± 6 21± 5 ND 2 0.040766 4.8 27 12 ± 1 6 ± 1 ND 2 0.004

Upregulated:

80 4.4 82 9 ± 4 47 ± 8 “Unknown protein” 5 0.00787 5.7 75 58 ± 14 262 ± 21 Transferrin 5 <0.001

138 5.5 67 99 ± 16 267 ± 39 Hemopexin-like 3 0.009144 5.4 63 26 ± 6 265 ± 66 l-Plastin 10 0.018190 5.9 54 6 ± 2 50 ± 9 Malic enzyme 9 0.018199 5.9 53 60 ± 16 156 ± 13 Thyroid hormone receptor 3 0.020275 6.1 45 1 ± 0.6 11 ± 0.6 NSH 9 <0.001387 5.6 35 97 ± 3 251 ± 49 Electron transferring flavoprotein 3 0.035389 5.8 35 192 ± 45 414 ± 54 Electron transferring flavoprotein 2 0.027399 6.8 33 59 ± 12 130 ± 10 Aldolase B 2 0.028457 4.7 29 26 ± 7 57 ± 5 14–3-3 B2 Protein 2 0.021461 4.7 27 75 ± 9 190 ± 12 Proteasome alpha 2 3 0.004517 4.4 22 15 ± 6 135 ± 29 Cytochrome c oxidase 9 0.013539 4.9 19 7 ± 3 18 ± 3 ND 3 0.033551 4.1 17 40 ± 11 143 ± 28 ND 4 0.018563 5.2 15 814 ± 198 3762 ± 984 Fatty acid-binding protein 5 0.039639 6.4 84 10 ± 6 28 ± 5 NSH 3 0.047648 6.1 55 17 ± 5 154 ± 46 Hydroxymethylglutaryl-CoA synthase 9 0.040678 5.3 48 26 ± 7 69 ± 15 Proteasome 26S ATPase subunit 4 3 0.044746 4.4 46 45 ± 13 107 ± 15 “Similar to catenin” 2 0.012754 4.1 15 6 ± 2 36 ± 4 ND 7 <0.001761 6.1 36 44 ± 21 204 ± 34 Transaldolase 5 0.006764 6.2 65 0 102 ± 17 NSH >10 NA770 5.0 21 4 ± 1 18 ± 4 ND 4 0.026

aValues are mean normalized protein abundance (± SE). Data were analyzed by the Student’s t test (n = 5). In the diet designated FM, protein wasprovided in the form of fish meal; in diet designated PP100, protein was provided by a cocktail of plant product with an equivalent amino acidcomposition to fish meal.NSH, no significant homology detected; ND, identity not determined; NA, not available.

P1: SFK/UKS P2: SFK



the effects on the proteasome are particularly noteworthy. Theproteasome is a multisubunit enzyme complex that catalyzesproteolysis via the ATP-dependent ubiquitin–proteasome path-way, which, in mammals, is thought to be responsible for alarge fraction of cellular proteolysis (Rock et al. 1994, Craiuet al. 1997). In rainbow trout, the ubiquitin–proteasome pathwayhas been shown to be downregulated in response to starvation(Martin et al. 2002) and to have a role in regulating proteindeposition efficiency (Dobly et al. 2004). It, therefore, seemslikely that the observed difference in texture is affected by ante-mortem proteasome activity, although further studies are neededto verify that statement.

In addition to having a hand in controlling autolysis determi-nants, protein turnover is a major regulatory engine of cellularstructure, function, and biochemistry. Cellular protein turnoverinvolves at least two major systems: (1) the lysosomal system and(2) the ubiquitin–proteasome system (Hershko and Ciechanover1986, Mortimore et al. 1989). The 20S proteasome has beenfound to have a role in regulating the efficiency with which rain-bow trout deposit protein (Dobly et al. 2004). It seems likelythat the manner in which protein deposition is regulated, par-ticularly in muscle tissue, has profound implications for qualityand processability of the fish flesh.

Protein turnover systems, such as the ubiquitin–proteasome orthe lysosome systems, are suitable for rigorous investigation us-ing proteomic methods. For example, lysosomes can be isolatedand the lysosome subproteome queried to answer the questionwhether and to what extent lysosome composition varies amongfish expected to yield flesh of different quality characteristics.Proteomic analysis on lysosomes has been successfully per-formed in mammalian (human) systems (Journet et al. 2000,Journet et al. 2002).

An exploitable property of proteasome-mediated proteindegradation is the phenomenon of polyubiquitination, wherebyproteins are targeted for destruction by the proteasome bycovalent binding to multiple copies of ubiquitin (Hershkoand Ciechanover 1986, Ciechanover 1994). By targeting theseubiquitin-labeled proteins, it is possible to observe the ubiquitin–proteasome “degradome,” that is, which proteins are beingdegraded by the proteasome at a given time or under givenconditions. Gygi and coworkers have developed methods tostudy the ubiquitin–proteasome degradome in the yeast Saccha-romyces cerevisiae using multidimensional LC-MS/MS (Penget al. 2003).

Some proteolysis systems, such as that of the matrix metal-loproteases, may be less directly amenable to proteomic study.Activity of matrix metalloproteases is regulated via a complexnetwork of specific proteases (Brown et al. 1993, Okumura et al.1997, Wang and Lakatta 2002). Monitoring of the expressionlevels of these regulatory enzymes, and how they vary withenvironmental or dietary variables, may be more convenientlycarried out using transcriptomic methods.

A well-known quality issue when farmed fish are comparedwith wild catch is that of gaping, a phenomenon caused bycleavage of myocommatal collagen cross-links that results inweakening and rupturing of connective tissue (Børresen 1992,Foegeding et al. 1996). Gaping can be a serious quality issue in

the fish processing industry as, apart from the obvious visual de-fect, it causes difficulties in mechanical skinning and slicing ofthe fish (Love 1992). Weakening of collagen and, hence, gapingis facilitated by low pH. Well-fed fish, such as those reared inaquaculture, tend to yield flesh of comparatively low pH, which,thus, tends to gape (Foegeding et al. 1996, Einen et al. 1999).Gaping is, therefore, a cause for concern with aquaculture-rearedfish, particularly of species with high natural gaping tendency,such as the Atlantic cod. Gaping tendency varies considerablyamong wild fish caught in different areas (Love et al. 1974)and, thus, it is conceivable that gaping tendency can be con-trolled with dietary or other environmental manipulations. Pro-teomics and transcriptomics, with their capacity to monitor mul-tiple biochemical processes simultaneously, are methodologieseminently suitable to finding biochemical or metabolic markersthat can be used for predicting features such as gaping tendencyof different stocks reared under different dietary or environmen-tal conditions.

Species Authentication

Food authentication is an area of increasing importance, botheconomically and from a public health standpoint. Taking intoaccount the large difference in market value of different fishspecies and the increased prevalence of processed product on themarket, it is perhaps not surprising that species authenticationis fast becoming an issue of supreme commercial importance.Along with other molecular techniques, such as DNA-basedspecies identification (Sotelo et al. 1993, Mackie et al. 1999,Martinez et al. 2001b, Pascoal et al. 2008) and nuclear magneticresonance-based techniques for determining geographical origin(Campana and Thorrold 2001), proteomics are proving to be apowerful tool in this area, particularly for addressing questionson health status of the organism, stresses or contamination lev-els at place of breeding, and postmortem treatment (Martinezet al. 2003, Martinez and Jakobsen Friis 2004). Since, unlikethe genome, the proteome is not a static entity, but changesbetween tissues and with environmental conditions, proteomicscan potentially yield more information than genomic ones, pos-sibly indicating freshness and tissue information in addition tospecies. Therefore, although it is likely that DNA-based methodswill remain the methods of choice for species authentication inthe near term, proteomic methods are likely to develop rapidlyand find commercial uses within this field. In many cases, theproteomes of even closely related fish species can be easily dis-tinguishable by eye from one another on two-dimensional gels(Fig. 22.6), indicating that diagnostic protein spots may be usedto distinguish closely related species.

From early on, proteomic methods have been recognized asa potential way of fish species identification. During the 1960s,one-dimensional electrophoretic techniques were developed toidentify the raw flesh of various species (Tsuyuki et al. 1966,Cowie 1968, Mackie 1969), which was soon followed by meth-ods to identify species in processed or cooked products (Mackie1972, Mackie and Taylor 1972). These early efforts were re-viewed in 1980 (Mackie 1980, Hume and Mackie 1980).

P1: SFK/UKS P2: SFK



67 kDa

43 kDa

30 kDa

21 kDa

14 kDa

67 kDa

43 kDa

30 kDa

21 kDa

14 kDa

4 pI 7 4 pI 7

(A) (B)

(C) (D)

Figure 22.6. Two-dimensional electrophoresis liver proteome maps of four salmonid fish (S. Martin and O. Vilhelmsson, unpublished).Running conditions are as in Figure 22.2. (A) Brown trout (Salmo trutta), (B) Arctic charr (Salvelinus alpinus), (C) rainbow trout(Oncorhynchus mykiss), (D) Atlantic salmon (Salmo salar ).

More recently, 2DE-based methods have been developed todistinguish various closely related species, such as the gadoids orseveral flat fishes (Pineiro et al. 1998, 1999, 2001). Pineiro andcoworkers have found that Cape hake (Merluccius capensis) andEuropean hake (Merluccius merluccius) can be distinguished ontwo-dimensional gels from other closely related species by thepresence of a particular protein spot that they identified, usingnanoelectrospray ionization MS, as nucleoside diphosphate ki-nase (Pineiro et al. 2001). Lopez and coworkers, studying threespecies of European mussels: (1) Mytilus edulis, (2) Mytilus gal-loprovincialis, and (3) Mytilus trossulus, found that M. trossuluscould be distinguished from the other two species on foot ex-tract two-dimensional gels by a difference in a tropomyosin spot.They found the difference to be due to a single T to D aminoacid substitution (Lopez et al. 2002). Martinez and JakobsenFriis (2004) attempted to identify not only the species present,but also their relative ratios in mixtures of several fish speciesand muscle types. They concluded that such a strategy wouldbecome viable once a suitable number of markers have beenidentified, although detection of species present in very differ-

ent ratios is problematic. Ortea et al. (2009a, b) used peptidemass fingerprinting of 2DE-separated proteins to discriminatebetween closely related shrimp species.

Recently, gel-free methods have received some attention.Mazzeo et al. (2008) found that specific protein biomarkersdetectable with MALDI-TOF MS could discriminate amongseveral commercially important fish species, such as those be-longing to Gadidae and Pleuronectiformes.

Identification and Characterizationof Allergens

Food safety is a matter of increasing concern to food producersand should be included in any consideration of product quality.Among issues within this field that are of particular concernto the seafood producer is that of allergenic potential. Aller-gic reactions to seafood affect a significant part of the popula-tion: about 0.5% of young adults are allergic to shrimp (Woodset al. 2002). Seafood allergies are caused by an immunoglobulinE-mediated response to particular proteins, including structural

P1: SFK/UKS P2: SFK



proteins such as tropomyosin (Lehrer et al. 2003). Proteomicsprovide a highly versatile toolkit to identify and characterize al-lergens. As yet, these have seen little use in the study of seafoodallergies, although an interesting and elegant approach was re-ported by Yu and coworkers (Yu et al. 2003a) at National TaiwanUniversity. These authors, studying the cause of shrimp allergyin humans, performed a 2DE on crude protein extracts fromthe tiger prawn, Penaeus monodon, blotted the two-dimensionalgel onto a PVDF membrane and probed the membranes withserum from confirmed shrimp allergic patients. The allergenswere then identified by MALDI-TOF MS of tryptic digests. Theallergen was identified as a protein with close similarity to argi-nine kinase. The identity was further corroborated by cloningand sequencing the relevant cDNA. A final proof was obtainedby purifying the protein, demonstrating that it had arginine ki-nase activity and reacted to serum IgE from shrimp allergicpatients and, furthermore, induced skin reactions in sensitizedshrimp allergic patients.

IMPACTS OF HIGH THROUGHPUTGENOMIC AND PROTEOMICTECHNOLOGIESIn recent years, proteomics has moved on from technical issuesrelated to protein separation and protein identity to highly repro-ducible gel-based or gel-free systems as discussed earlier. Thisincreased ability to separate proteins and to perform peptide se-quence analysis by MS has meant that the volume of data thatis produced and the rate of identification of proteins is ordersof magnitude greater than only a few years ago (reviewed bySeidler et al. 2010). Directly related to the increased volume ofdata generated means that extracting the relevant information isno longer a simple matter of discussing a list of protein identities,and interpretation of the proteins and their function is central toany medium- to large-scale proteome study (Malik et al. 2010).

The number of expressed sequence tags (ESTs) related tosalmonid fish is currently in the order of 800,000 sequences, rep-resenting mRNAs encoding about 30–40,000 different proteins,other commercial species including cod, sea bass, sea bream,and catfish amongst others are quickly catching up (Martin et al.2008). These sequences can be used to help identify amino acidsequences generated during proteomics studies, which meansthat now the majority of proteins can be identified rather thanthe minority.

There are now a considerable number of fish species whosewhole genome is completely sequenced; zebrafish, two speciesof puffer fish, and stickleback have their genomes sequenced.And recently the sea bass, a major aquacultured species has beenalmost fully sequenced (Kuhl et al. 2010). The technology andtools used for these projects have led the way to expand thesequence data and interpretation of sequences from commer-cially important species (Forne et al. 2010)—the data that canbe directly used for proteomic studies. A good example of this iszebrafish where 5716 proteins were identified with an estimatedfalse discovery rate of 1.34% (De Souza et al. 2009).

As more and more genes and protein sequences are depositedin databases, they are automatically annotated; the quality of

these annotations is probably one of the greatest hurdles in fullyinterpreting the output of either a transcriptomic study or pro-teomic studies. Currently, annotations include the nucleotidesequence, protein sequence, tissue distribution abundance ofmRNA, gene ontology (GO) (Gaudet et al. 2009), and KyotoEncyclopedia of Genes and Genomes (KEGG) (Okuda et al.2008) pathways. A high proportion of the genes in human, mice,yeast, among others, have many of their protein annotated to thisextent; in the near future, this will be the case for fish as well.GO and KEGG are two databases that are used to assign func-tion to a particular protein. For GO, these are three differentassignments: (1) biological process, (2) molecular function, and(3) cellular component, the same terms being used across allspecies. This gives a vocabulary of definitions for a particularprotein; in fact, there are now thousands of GO terms. When pro-teins are identified they can be matched to these different GOterms; for example, glycolysis or protein metabolism as verybasic examples. If a data set produced contains tens or hundredsof proteins identities, it is possible to use GO to show the globalchanges occurring in that tissue; if many GO terms are related tothe same biological process or function, this can help interpretthe data as opposed to a simple list of proteins. If there are twosamples to compare, statistical analysis of the GO terms can in-dicate if there are major differences in cellular activity betweenthe two samples, or by using GO cellular component, it can beindicated what are the particular organelles that are being af-fected. The KEGG pathways (http://www.genome.jp/kegg/) aresimilar to biochemical maps with interactive links to other nu-cleotide, protein, and literature databases, as with GO these helpinterpret the output data when a large number of proteins havebeen identified. New pathway analysis programs are continu-ally being developed including Reactome (Vastrik et al. 2007)that allows networks of proteins to be explored and integra-tion of data with literature through a platform called iHOP(Hoffmann and Valencia 2004). Although there is a requirementfor significant manual analysis of the data when working withfish species (including the model species), clearly as the usersbecome more familiar with the available tools and the databasesbecome more integrated, the knowledge gap between thoseworking on species with well-annotated genomes and fish willdecrease.

In future studies, data from other “omic” platforms shouldbe combined, that is, incorporating transcriptomic data frommicroarray and deep sequencing and from metabolomic data.The complementary techniques, although performed often indifferent laboratories, do ask the same questions and one of thefuture steps will be to perform meta-analysis across these highthroughput technologies.

REFERENCES

Ahmed FE. 2009. Sample preparation and fractionation for pro-teome analysis and cancer biomarker discovery by mass spec-trometry. Journal of Separation Science 32: 771–798.

Ahmed N et al. 2003. An approach to remove albumin for the pro-teomic analysis of low abundance biomarkers in human serum.Proteomics 3: 1980–1987.

P1: SFK/UKS P2: SFK



Ahmed N, Rice GE. 2005. Strategies for revealing lower abundanceproteins in two-dimensional protein maps. Journal of Chromatog-raphy B 815: 39–50.

Andersen JS, Mann M. 2006. Organellar proteomics: turning inven-tories into insights. EMBO Reports 7: 874–879.

Anderson NL, Anderson NG. 1998. Proteome and proteomics: newtechnologies, new concepts, and new words. Electrophoresis 19:1853–1861.

Babu GJ et al. 2004. Solubilization of membrane proteins fortwo-dimensional gel electrophoresis: identification of sarcoplas-mic reticulum membrane proteins. Analytical Biochemistry 325:121–125.

Bae SH et al. 2003. Strategies for the enrichment and identificationof basic proteins in proteome projects. Proteomics 3: 569–579.

Balestrieri ML et al. 2008. Proteomics and cardiovascular disease:an update. Current Medicinal Chemistry 15: 555–572.

Barrett J et al. 2005. Analysing proteomic data. International Jour-nal for Parasitology 35: 543–553.

Beretta L. 2009. Comparative analysis of the liver and plasma pro-teomes as a novel and powerful strategy for hepatocellular carci-noma biomarker discovery. Cancer Letters 286: 134–139.

Berkelman T, Stenstedt T. 1998. 2-D Electrophoresis Using Im-mobilized pH Gradients. Principles and Methods. AmershamBiosciences, Uppsala.

Bodzon-Kulakowska A et al. 2007. Methods for samples prepara-tion in proteomic research. Journal of Chromatography B 849:1–31.

Bogyo M, Cravatt BF. 2007. Genomics and proteomics: from genesto function: advances in applications of chemical and systemsbiology. Current Opinion in Chemical Biology 11: 1–3.

Bossi A et al. 1994. Focusing of alkaline proteases (subtilisins)in pH 10–12 immobilized pH gradients. Electrophoresis 15:1535–1540.

Brown PD et al. 1993. Cellular activation of the 72 kDa typeIV procollagenase/TIMP-2 complex. Kidney International 43:163–170.

Butt A et al. 2001. Chromatographic separations as a prelude to two-dimensional electrophoresis in proteomics analysis. Proteomics1: 42–53.

Børresen T. 1992. Quality aspects of wild and reared fish. In: HHHuss et al. (eds.) Quality Assurance in the Fish Industry. ElsevierScience Publishers, Amsterdam, pp. 1–17.

Cahu CL et al. 1999. Protein hydrolysate vs. fish meal in com-pound diets for 10-day old sea bass Dicentrarchus labrax larvae.Aquaculture 171: 109–119.

Campana SE, Thorrold SR. 2001. Otoliths, increments, and ele-ments: keys to a comprehensive understanding of fish popula-tions? Canadian Journal of Fisheries and Aquatic Sciences 58:30–38.

Campinho MA et al. 2007. Molecular, cellular and histologicalchanges in skin from a larval to an adult phenotype during bonyfish metamorphosis. Cell and Tissue Research 327: 267–284.

Campinho MA et al. 2006. Regulation of troponin T expression dur-ing muscle development in sea bream Sparus auratus Linnaeus:the potential role of thyroid hormones. Journal of ExperimentalBiology 209: 4751–4767.

Canas B et al. 2007. Trends in sample preparation for classicaland second generation proteomics. Journal of ChromatographyA 1153: 235–258.

Chelius D et al. 2003. Global protein identification and quan-tification technology using two-dimensional liquid chromatog-raphy nanospray mass spectrometry. Analytical Chemistry 75:6658–6665.

Chevallet M et al. 1998. New zwitterionic detergents improve theanalysis of membrane proteins by two-dimensional electrophore-sis. Electrophoresis 19: 1901–1909.

Ciechanover A. 1994. The ubiquitin-proteasome proteolytic path-way. Cell 79: 13–21.

Clauser KR et al. 1999. Role of accurate measurement (+/− 10 ppm)in protein identification strategies employing MS or MS/MS anddatabase searching. Analytical Chemistry 71: 2871–2882.

Coe JE, Vilhelmsson O. 2008. Two-dimensional polyacrylamide gelelectrophoresis of Arctic charr liver proteins: setup of equipmentand standardization of protocols. Technical Report No. TS08:08.The University of Akureyri. Akureyri, Iceland.

Cordwell SJ et al. 2000. Subproteomics based upon protein cel-lular location and relative solubilities in conjunction with com-posite two-dimensional electrophoresis gels. Electrophoresis 21:1094–1103.

Corthals GL et al. 1997. Prefractionation of protein samplesprior to two-dimensional electrophoresis. Electrophoresis 18:317–323.

Courchesne PL, Patterson SD. 1999. Identification of proteins bymatrix-assisted laser desorption/ionization mass spectrometry us-ing peptide and fragment ion masses. In: AJ Link (ed.) 2-D Pro-teome Analysis Protocols, Chapter 51. Humana Press, Totowa,NJ, pp. 487–511.

Cowie WP. 1968. Identification of fish species by thin slab poly-acrylamide gel electrophoresis. Journal of the Science of Foodand Agriculture 19: 226–229.

Craiu A et al. 1997. Two distinct proteolytic processes in the gen-eration of a major histocompatibility complex class I-presentedpeptide. Proceedings of the National Academy of Sciences of theUnited States of America 94: 10850–10855.

De Francesco M et al. 2004. Effect of long-term feeding with a plantprotein mixture based diet on growth and body/fillet quality traitsof large rainbow trout (Oncorhynchus mykiss). Aquaculture 236:413–429.

De Souza AG et al. 2009. Large-scale proteome profile of the ze-brafish (Danio rerio) gill for physiological and biomarker dis-covery studies. Zebrafish 6: 229–238.

Dobly A et al. 2004. Efficiency of conversion of ingested proteinsinto growth; protein degradation assessed by 20S proteasomeactivity in rainbow trout, Oncorhynchus mykiss. ComparativeBiochemistry and Physiology A 137: 75–85.

Dougherty DA et al. 1990. Evaluation of selected baking qualityfactors of hard red winter wheat flours by two-dimensional elec-trophoresis. Cereal Chemistry 67: 564–569.

Drabik A et al. 2007. Proteomics in neurosciences. Mass Spectrom-etry Reviews 26: 432–450.

Dreger M. 2003. Subcellular proteomics. Mass Spectrometry Re-views 22: 27–56.

Einarsdottir IE et al. 2006. Thyroid and pituitary gland developmentfrom hatching through metamorphosis of a teleost flatfish, theAtlantic halibut. Anatatomy and Embryology 211: 47–60.

Einen O et al. 1999. Feed ration prior to slaughter - a potential toolfor managing product quality of Atlantic salmon (Salmo salar).Aquaculture 178: 149–169.

P1: SFK/UKS P2: SFK



Erdjument-Bromage H et al. 1999. Characterizing proteins from2-DE gels by internal sequence analysis of peptide fragments.In: AJ Link (ed.) 2-D Proteome Analysis Protocols, Chapter 49.Humana Press, Totowa, NJ, pp. 467–472.

Focant B et al. 2000. Expression of myofibrillar proteins and parval-bumin isoforms in white muscle of the developing turbot Scoph-thalmus maximus (Pisces, Pleuronectiformes). Basic and AppliedMyology 10: 269–278.

Focant B et al. 1999. Muscle parvalbumin isoforms of Clariasgariepinus, Heterobranchus longifilis and Chrysichthys auratus:isolation, characterization, and expression during development.Journal of Fish Biology 54: 832–851.

Focant B et al. 2003. Expression of myofibrillar proteins and par-valbumin isoforms during the development of a flatfish, the com-mon sole Solea solea: comparison with the turbot Scophthal-mus maximus. Comparative Biochemistry and Physiology B 135:493–502.

Foegeding EA et al. 1996. Characteristics of edible muscle tissues.In: OR Fennema (ed.) Food Chemistry. Marcel Dekker, NewYork, pp. 879–942.

Forne I et al. 2010. Fish proteome analysis: model organisms andnon-sequenced species. Proteomics 10: 858–872.

Fortis F et al. 2008. A pI-based protein fractionation method usingsolid-state buffers. Journal of Proteomics 71: 379–389.

Galloway TF et al. 2006. Somite formation and expression of MyoD,myogenin and myosin in Atlantic halibut (Hippoglossus hip-poglossus L.) embryos incubated at different temperatures: tran-sient asymmetric expression of MyoD. Journal of ExperimentalBiology 209: 2432–2441.

Galloway TF et al. 1998. Effect of temperature on viability andaxial muscle development in embryos and yolk sac larvae of theNortheast Atlantic cod (Gadus morhua). Marine Biology 132:547–557.

Galloway TF et al. 1999. Muscle growth in yolk sac larvae of theAtlantic halibut as influenced by temperature in the egg and yolksac stage. Journal of Fish Biology 55: 26–43.

Garcia-Canas V et al. 2010. Advances in Nutrigenomics research:novel and future analytical approaches to investigate the biologi-cal activity of natural compounds and food functions. Journal ofPharmaceutical and Biomedical Analysis 51: 290–304.

Gaudet P et al. 2009. The gene ontology’s reference genome project:a unified framework for functional annotation across species.PLOS Computational Biology 5: e1000431.

Gorg A et al. 1987. Elimination of point streaking on silver stainedtwo-dimensional gels by addition of iodoacetamide to the equili-bration buffer. Electrophoresis 8: 122–124.

Gorg A et al. 1988. The current state of two-dimensional elec-trophoresis with immobilized pH gradients. Electrophoresis 9:531–546.

Gorg A et al. 1997. Very alkaline immobilized pH gradients for two-dimensional electrophoresis of ribosomal and nuclear proteins.Electrophoresis 18: 328–337.

Gorg A et al. 1999. Recent developments in two-dimensional gelelectrophoresis with immobilized pH gradients: wide pH gra-dients up to pH 12, longer separation distances and simplifiedprocedures. Electrophoresis 20: 712–717.

Gorg A et al. 2000. The current state of two-dimensionalelelctrophoresis with immobilized pH gradients. Electrophore-sis 21: 1037–1053.

Gorg A et al. 2004. Current two-dimensional electrophoresis tech-nology for proteomics. Proteomics 19: 3665–3685.

Govoni JJ et al. 1986. The physiology of digestion in fish larvae.Environmental Biology of Fishes 16: 59–77.

Guðmundsdottir A et al. 1993. Isolation and characterization of cD-NAs from Atlantic cod encoding two different forms of trypsino-gen. European Journal of Biochemistry 217: 1091–1097.

Guðmundsdottir A, Sveinsdottir H. 2006. Rynt ı proteinmengi lirfaAtlantshafsþorsks (Gadus morhua). In: GG Haraldsson (ed.)Vısindin heilla: afmælisrit til heiðurs Sigmundi Guðbjarnasyni75 ara. Haskolautgafan, Reykjavık, pp. 425–440.

Gustafson JS et al. 2004. Statistical exploration of variation in quan-titative two-dimensional gel electrophoresis data. Proteomics 4:3791–3799.

Hall TH et al. 2003. Temperature and the expression of sevenmuscle-specific protein genes during embryogenesis in the At-lantic cod Gadus morhua L. Journal of Experimental Biology206: 3187–3200.

Henningsen R et al. 2002. Application of zwitterionic detergentsto the solubilization of integral membrane proteins for two-dimensional gel electrophoresis and mass spectrometry. Pro-teomics 2: 1479–1488.

Herbert B. 1999. Advances in protein solubilization for two-dimensional electrophoresis. Electrophoresis 20: 660–663.

Hershko A, Ciechanover A. 1986. The ubiquitin pathway for thedegradation of intracellular proteins. Progress in Nucleic AcidResearch and Molecular Biology 33: 19–56.

Hoffmann R, Valencia A. 2004. A gene network for navigating theliterature. Nature Genetics 36: 664.

Hogstrand C et al. 2002. Application of genomics and proteomicsfor study of the integrated response to zinc exposure in a non-model fish species, the rainbow trout. Comparative Biochemistryand Physiology B 133: 523–535.

Hume A, Mackie I. 1980. The use of electrophoresis of the water-soluble muscle proteins in the quantitative analysis of the speciescomponents of a fish mince mixture. In: JJ Connell (ed.) Ad-vances in Fish Science and Technology. Fishing News BooksLtd., Aberdeen, pp. 451–456.

Huriaux F et al. 2003. Expression of myofibrillar proteins and par-valbumin isoforms in white muscle of dorada during develop-ment. Journal of Fish Biology 62: 774–792.

Huriaux F et al. 1996. Parvalbumin isotypes in white muscle fromthree teleost fish: characterization and their expression duringdevelopment. Comparative Biochemistry and Physiology B 113:475–484.

Huriaux F et al. 1999. Myofibrillar proteins in white muscle of thedeveloping catfish Heterobranchus longifilis (Siluriforms, Clari-idae). Fish Physiology and Biochemistry 21: 287–301.

Huss HH. 1995. Quality and quality changes in fresh fish. FAOFisheries Technical Paper No. 348. FAO. Rome.

Ikonomou G et al. 2009. Proteomic methodologies and their appli-cation in colorectal cancer research. Critical Reviews in ClinicalLaboratory Sciences 46: 319–342.

Ingolfsdottir S et al. 1998. Seasonal variations in physicochem-ical and textural properties of North Atlantic cod (Gadusmorhua) mince. Journal of Aquatic Food Product Technology 7:39–61.

Issaq HJ et al. 2002. Methods for fractionation, separation and pro-filing of proteins and peptides. Electrophoresis 23: 3048–3061.

P1: SFK/UKS P2: SFK



Issaq HJ, Veenstra TD. 2008. Two-dimensional polyacrylamide gelelectrophoresis (2D-PAGE): advances and perspectives. Biotech-niques 44: 697–698.

Jensen O. 2004. Modification-specific proteomics: characterizationof post-translational modifications by mass spectrometry. CurrentOpinion in Chemical Biology 8: 33–41.

Jorrin-Novo JV et al. 2009. Plant proteomics update (2007–2008):Second-generation proteomic techniques, an appropriate exper-imental design, and data analysis to fulfill MIAPE standards,increase plant proteome coverage and expand biological knowl-edge. Journal of Proteomics 72: 285–314.

Journet A et al. 2000. Towards a human repertoire of monocyticlysosomal proteins. Electrophoresis 21: 3411–3419.

Journet A et al. 2002. Proteomic analysis of human lysosomes:application to monocytic and breast cancer cells. Proteomics 2:1026–1040.

Kaltschmidt E, Wittmann H-G. 1970. Ribosomal proteins. VII. Two-dimensional polyacrylamide gel electrophoresis for fingerprint-ing of ribosomal proteins. Analytical Biochemistry 36: 401–412.

Kamo M, Tsugita A. 1999. N-terminal amino acid sequencing of2-DE spots. In: AJ Link (ed.) 2-D Proteome Analysis Protocols,Chapter 48. Humana Press, Totowa, NJ, pp. 461–466.

Karp NA et al. 2005. Application of partial least squared discrimi-nant analysis to two-dimensional difference gel studies in expres-sion proteomics. Proteomics 5: 81–90.

Kjaersgard IVH, Jessen F. 2003. Proteome analysis elucidating post-mortern changes in cod (Gadus morhua) muscle proteins. Journalof Agricultural and Food Chemistry 51: 3985–3991.

Kjaersgard IVH, Jessen F. 2004. Two-dimensional gel electrophore-sis detection of protein oxidation in fresh and tainted rainbowtrout muscle. Journal of Agricultural and Food Chemistry 52:7101–7107.

Kjaersgard IVH et al. 2006a. Identification of carbonylated proteinin frozen rainbow trout (Oncorhynchus mykiss) fillets and devel-opment of protein oxidation during frozen storage. Journal ofAgricultural and Food Chemistry 54: 9437–9446.

Kjaersgard IVH et al. 2006b. Changes in cod muscle proteins duringfrozen storage revealed by proteome analysis and multivariatedata analysis. Proteomics 6: 1606–1618.

Kraj A, Silberring J. 2008. Introduction to Proteomics John Wiley& Sons, Hoboken, NJ.

Kristensen L et al. 2002. Dietary-induced changes of muscle growthrate in pigs: effects on in vivo and postmortem muscle proteolysisand meat quality. Journal of Animal Science 80: 2862–2871.

Kuhl H et al. 2010. The European sea bass Dicentrarchus labraxgenome puzzle: comparative BAC-mapping and low coverageshotgun sequencing. BMC Genomics 11: 68.

Ladrat C et al. 2000. Neutral calcium-activated proteases from Eu-ropean sea bass (Dicentrachus labrax L.) muscle: polymorphismand biochemical studies. Comparative Biochemistry and Physi-ology B 125: 83–95.

Laemmli UK. 1970. Cleavage of structural proteins during the as-sembly of the head of bacteriophage T4. Nature 227: 680–685.

Latterich M et al. 2008. Proteomics: new technologies and clinicalapplications. European Journal of Cancer 44: 2737–2741.

Lee BH, Nagamune T. 2004. Protein microarrays and their applica-tions. Biotechnology and Bioprocess Engineering 9: 69–75.

Lee K, Pi K. 2009. Proteomic profiling combining solution-phaseisoelectric fractionation with two-dimensional gel electrophore-

sis using narrow-pH-range immobilized pH gradient gels withslightly overlapping pH ranges. Analytical and BioanalyticalChemistry 396: 535–539.

Lehrer SB et al. 2003. Seafood allergy and allergens: a review.Marine Biotechnology 5: 339–348.

Lei MG, Reeck GR. 1987. Two dimensional electrophoretic analysisof isolated soybean protein bodies and of the glycosylation ofsoybean proteins. Journal of Agricultural and Food Chemistry35: 296–300.

Link AJ. 1999. 2-D Proteome Analysis Protocols. Humana Press,Totowa, NJ.

Link V et al. 2006. Proteomics of early zebrafish embryos. BMCDevelopmental Biology 6: 1–9.

Locke VL et al. 2002. Gradiflow as a prefractionation tool for two-dimensional electrophoresis. Proteomics 2: 1254–1260.

Lopez JL et al. 2002. Application of proteomics for fast identifica-tion of species-specific peptides from marine species. Proteomics2: 1658–1665.

Lopez MF et al. 2000. High-throughput profiling of the mitochon-drial proteome using affinity fractionation and automation. Elec-trophoresis 21: 3427–3440.

Love RM. 1980. Biological factors affecting processing and utiliza-tion. In: JJ Connell (ed.) Advances in Fish Science and Technol-ogy. Fishing News Books, Aberdeen, pp. 130–138.

Love RM. 1992. Biochemical dynamics and the quality of freshand frozen fish. In: GM Hall (ed.) Fish Processing Technology.Blackie Academic and Professional, London, pp. 1–26.

Love RM et al. 1974. The texture of cod muscle. Journal of TextureStudies 5: 201–212.

Lopez JL. 2007. Two-dimensional electrophoresis in proteome ex-pression analysis. Journal of Chromatography B 849: 190–202.

Lødemel JB, Olsen RL. 2003. Gelatinolytic activities in muscleof Atlantic cod (Gadus morhua), spotted wolffish (Anarhichasminor) and Atlantic salmon (Salmo salar). Journal of the Scienceof Food and Agriculture 83: 1031–1036.

Mackie I. 1980. A review of some recent applications of elec-trophoresis and isoelectric focusing in the identification of speciesof fish in fish and fish products. In: JJ Connell (ed.) Advancesin Fish Science and Technology. Fishing News Books Ltd.,Aberdeen, pp. 444–450.

Mackie IM. 1969. Identification of fish species by a modified poly-acrylamide disc electrophoresis technique. Journal of the Asso-ciation of Public Analysts 5: 83–87.

Mackie IM. 1972. Some improvements in the polyacrylamide discelectrophoretic method of identifying species of cooked fish.Journal of the Association of Public Analysts 8: 18–20.

Mackie IM et al. 1999. Challenges in the identification of speciesof canned fish. Trends in Food Science and Technology 10: 9–14.

Mackie IM, Taylor T. 1972. Identification of species of heat-sterilized canned fish by polyacrylamide disc electrophoresis.Analyst 97: 609–611.

Malik R et al. 2010. From proteome lists to biological impact- toolsand strategies for the analysis of large MS data sets. Proteomics10: 1270–1283.

Malmstrom J et al. 2007. Advances in proteomic workflows forsystems biology. Current Opinion in Biotechnology 18: 378–384.

Mamone G et al. 2009. Analysis of food proteins and peptides bymass spectrometry-based techniques. Journal of Chromatogra-phy A 1216: 7130–7142.

P1: SFK/UKS P2: SFK



Margolis J, Kenrick KG. 1969. Two-dimensional resolution ofplasma proteins by combination of poly-acrylamide disc and gra-dient gel electrophoresis. Nature 221: 1056–1057.

Martin SA et al. 2002. Ubiquitin-proteasome-dependent proteolysisin rainbow trout (Oncorhynchus mykiss): effect of food depriva-tion. Pflugers Archive 445: 257–266.

Martin SAM et al. 2001. Proteome analysis of rainbow trout (On-corhynchus mykiss) liver proteins during short-term starvation.Fish Physiology and Biochemistry 24: 259–270.

Martin SAM et al. 2003a. Rainbow trout liver proteome - dietary ma-nipulation and protein metabolism. In: WB Souffrant, CC Metges(eds.) Progress in Research on Energy and Protein Metabolism.Wageningen Academic Publishers, Wageningen, pp. 57–60.

Martin SAM et al. 2003b. Proteomic sensitivity to dietary manip-ulations in rainbow trout. Biochimica et Biophysica Acta 1651:17–29.

Martin SAM et al. 2008. Genomic tools for examining immunegene function in salmonid fish. Reviews in Fisheries Science 16:112–118.

Martinez I et al. 2003. Destructive and non-destructive analyticaltechniques for authentication and composition analyses of food-stuffs. Trends in Food Science and Technology 14: 489–498.

Martinez I, Jakobsen Friis T. 2004. Application of proteome analysisto seafood authentication. Proteomics 4: 347–354.

Martinez I et al. 2001a. Post mortem muscle protein degradationduring ice-storage of Arctic (Pandalus borealis) and tropical (Pe-naeus japonicus and Penaeus monodon) shrimps: a comparativeelectrophoretic and immunological study. Journal of the Scienceof Food and Agriculture 81: 1199–1208.

Martinez I et al. 2001b. Requirements for the application of pro-tein sodium dedecyl sulfate-polyacrylamide gel electrophoresisand randomly amplified polymorphic DNA analyses to productspeciation. Electrophoresis 22: 1526–1533.

Martinez I et al. 1990. Electrophoretic study of myosin isoforms inwhite muscles of some teleost fishes. Comparative Biochemistryand Physiology B 96: 221–227.

Mazzeo MF et al. 2008. Fish authentication by MALDI-TOF massspectrometry. Journal of Agricultural and Food Chemistry 56:11071–11076.

Millea KM, Krull IS. 2003. Subproteomics in analytical chemistry:chromatographic fractionation techniques in the characterizationof proteins and peptides. Journal of Liquid Chromatography &Related Technologies 26: 2195–2224.

Mortimore GE et al. 1989. Mechanism and regulation of proteindegradation in liver. Diabetes/Metababolism Reviews 5: 49–70.

Morzel M et al. 2000. Use of two-dimensional electrophoresis toevaluate proteolysis in salmon (Salmo salar) muscle as affectedby a lactic fermentation. Journal of Agricultural and Food Chem-istry 48: 239–244.

O’Connell CP. 1981. Development of organ systems in the northernanchovy Engraulis mordax and other teleosts. American Zoolo-gist 21: 429–446.

O’Farrell PH. 1975. High resolution two-dimensional electrophore-sis of proteins. Journal of Biological Chemistry 250: 4007–4021.

Oda Y et al. 2001. Enrichment analysis of phosphorylated proteinsas a tool for probing the phosphoproteome. Nature Biotechnology19: 379–382.

Ogata H et al. 1998. Proteolytic degradation of myofibrillar com-ponents by carp cathepsin L. Journal of the Science of Food andAgriculture 76: 499–504.

Okuda S et al. 2008. KEGG Atlas mapping for global analysis ofmetabolic pathways. Nucleic Acids Research 36: W423–W426.

Okumura Y et al. 1997. Proteolytic activation of the precursor ofmembrane type 1 matrix metalloproteinase by human plasmin. Apossible cell surface activator. FEBS Letters 402: 181–184.

Ortea I et al. 2009a. Closely related shrimp species identificationby MALDI-ToF mass spectrometry. Journal of Aquatic FoodProduct Technology 18: 146–155.

Ortea I et al. 2009b. Arginine kinase peptide mass fingerprinting asa proteomic approach for species identification and taxonomicanalysis of commercially relevant shrimp species. Journal ofAgricultural and Food Chemistry 57: 5665–5672.

Osse JWM, van den Boogaart JGM. 1995. Fish larvae, development,allometric growth, and the aquatic environment. ICES MarineScience Symposia 201: 21–34.

Østergaard O et al. 2002. Initial proteome analysis of mature barleyseeds and malt. Proteomics 2: 733–739.

Papa I et al. 1996. Post mortem release of fish white muscle a-actininas a marker of disorganisation. Journal of the Science of Food andAgriculture 72: 63–70.

Parisi G et al. 2004. Effect of total replacement of dietary fish mealby plant protein sources on early post mortem changes in thebiochemical and physical parameters of rainbow trout. VeterinaryResearch Communications 28: 237–240.

Parrington J, Coward K. 2002. Use of emerging genomic and pro-teomic technologies in fish physiology. Aquatic Living Resources15: 193–196.

Pascoal A et al. 2008. Identification of shrimp species in rawand processed food products by means of a polymerase chainreaction-restriction fragment length polymorphism method tar-geted to cytochrome b mitochondrial sequences. Electrophoresis29: 3220–3228.

Patton WE. 2002. Detection technologies in proteome analysis.Journal of Chromatography B 771: 3–31.

Paz M et al. 2006. Proteome profile changes during mouse testisdevelopment. Comparative Biochemistry and Physiology D 1:404–415.

Peng J et al. 2003. A proteomics approach to understanding proteinubiquitination. Nature Biotechnology 21: 921–926.

Perkins DN et al. 1999. Probability-based protein identificationby searching sequence databases using mass spectrometry data.Electrophoresis 20: 3551–3567.

Perez-Borla O et al. 2002. Proteolytic activity of muscle in pre- andpost-spawning hake (Merluccius hubbsi Marini) after frozen stor-age. Lebensmittel-Wissenschaft und –Technologie 35: 325–330.

Pieper R et al. 2003. Multi-component immunoaffinity subtractionchromatography: an innovative step towards a comprehensivesurvey of the human plasma proteome. Proteomics 3: 422–432.

Pineiro C et al. 1998. Two-dimensional electrophoretic study ofthe water-soluble protein fraction in white muscle of gadoidfish species. Journal of Agricultural and Food Chemistry 46:3991–3997.

Pineiro C et al. 1999. The use of two-dimensional electrophoresisfor the identification of commercial flat fish species. Zeitschriftfur Lebensmitteluntersuchung und –Forschung 208: 342–348.

Pineiro C et al. 2003. Proteomics as a tool for the investigation ofseafood and other marine products. Journal of Proteome Research2: 127–135.

Pineiro C et al. 2001. Characterization and partial sequencingof species-specific sarcoplasmic polypeptides from commercial

P1: SFK/UKS P2: SFK



hake species by mass spectrometry following two-dimensionalelectrophoresis. Electrophoresis 22: 1545–1552.

Premsler T et al. 2009. Recent advances in yeast organelle andmembrane proteomics. Proteomics 9: 4731–4743.

Righetti PG et al. 2005a. Prefractionation techniques in proteomeanalysis: the mining tools of the third millennium. Electrophore-sis 26: 297–319.

Righetti PG et al. 2005b. How to bring the “unseen” proteometo the limelight via electrophoretic pre-fractionation techniques.Bioscience Reports 25: 3–17.

Rock KL et al. 1994. Inhibitors of the proteasome block the degrada-tion of most cell proteins and the generation of peptides presentedon MHC class I molecules. Cell 78: 761–771.

Rothemund DL et al. 2003. Depletion of the highly abundant proteinalbumin from human plasma using the Gradiflow. Proteomics 3:279–287.

Seidler J et al. 2010. De novo sequencing of peptides by MS/MS.Proteomics 10: 634–649.

Simpson R. 2003. Proteins and Proteomics: A Laboratory Manual.Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NewYork.

Skiftesvik AB. 1992. Changes in behaviour at onset of exogenousfeeding in marine fish larvae. Canadian Journal of Fisheries andAquatic Sciences 49: 1570–1572.

Smith MPW et al. 2009. Application of proteomic analysis to thestudy of renal diseases. Nature Reviews Nephrology 5: 701–712.

Sotelo CG et al. 1993. Fish species identification in seafood prod-ucts. Trends in Food Science and Technology 4: 395–401.

Sveinsdottir H, Gudmundsdottir A. 2008. Proteome profile compar-ison of two differently fed groups of Atlantic cod (Gadus morhua)larvae. Aquaculture Nutrition 16: 662–670.

Sveinsdottir H et al. 2008. Proteome analysis of abundant proteinsin two age groups of early Atlantic cod (Gadus morhua) larvae.Comparative Biochemistry and Physiology D 3: 243–250.

Sveinsdottir H et al. 2009. Differential protein expression in earlyAtlantic cod larvae (Gadus morhua) in response to treatment withprobiotic bacteria. Comparative Biochemistry and Physiology D4: 249–254.

Tay TL et al. 2006. Proteomic analysis of protein profiles duringearly development of the zebrafish, Danio rerio. Proteomics 6:3176–3188.

Thorarinsdottir KA et al. 2002. Changes in myofibrillar proteinsduring processing of salted cod (Gadus morhua) as determinedby electrophoresis and differential scanning calorimetry. FoodChemistry 77: 377–385.

Tinh NTN et al. 2008. A review of the functionality of probiotics inthe larviculture food chain. Marine Biotechnology 10: 1–12.

Tsuyuki H et al. 1966. Comparative electropherograms of Coregonisclupeoformis, Salvelinus namaycush, S. alpinus, S. malma and S.fontinalis from the family Salmonidae. Journal of the FisheriesResearch Board of Canada 23: 1599–1606.

Tyers M, Mann M. 2003. From genomics to proteomics. Nature422: 193–197.

Vaneenaeme C et al. 1994. Postmortem proteases activity in relationto muscle protein-turnover in Belgian blue bulls with differentgrowth-rates. Sciences des Aliments 14: 475–483.

Vastrik I et al. 2007. Reactome: a knowledge base of biologic path-ways and processes. Genome Biology 8: R39.

Verrez-Bagnis V et al. 1999. Desmin degradation in postmortemfish muscle. Journal of Food Science 64: 240–242.

Vilhelmsson O, Miller KJ. 2002. Synthesis of pyruvate de-hydrogenase in Staphylococcus aureus is stimulated by os-motic stress. Applied and Environmental Microbiology 68:2353–2358.

Vilhelmsson OT et al. 2004. Dietary plant protein substitution affectshepatic metabolism in rainbow trout (Oncorhynchus mykiss).British Journal of Nutrition 92: 71–80.

von Horsten HH. 2006. Current strategies and techniques for theanalysis of low-abundance proteins and peptides from complexsamples. Current Analytical Chemistry 2: 129–138.

Walker JM. 2005. The Proteomics Protocols Handbook. HumanaPress, Totowa, NJ.

Walsh BJ, Herbert BR. 1999. Casting and running vertical slab-gelelectrophoresis for 2D-PAGE. In: AJ Link (ed.) 2-D ProteomeAnalysis Protocols. Humana Press, Totowa, NJ, pp. 245–253.

Wang J et al. 2006. Proteomics and its role in nutrition research.Journal of Nutrition 136: 1759–1762.

Wang M, Lakatta EG. 2002. Altered regulation of matrixmetalloproteinase-2 in aortic remodeling during aging. Hyper-tension 39: 865–873.

Westerfield M. 2000. The Zebrafish Book. A Guide for the Lab-oratory Use of Zebrafish (Danio rerio), 4th edn. University ofOregon Press, Eugene, OR.

Westermeier R, Naven T. 2002. Proteomics in Practice. Wiley-VCH, Weinheim.

Wheelock AM, Goto S. 2006. Effects of post-electrophoretic anal-ysis on variance in gel-based proteomics. Expert Review of Pro-teomics 3: 129–142.

Wilm M. 2009. Quantitative proteomics in biological research. Pro-teomics 9: 4590–4605.

Wilm M et al. 1996. Femtomole sequencing of proteins from poly-acrylamide gels by nano-electrospray mass spectrometry. Nature379: 466–469.

Woessner JF. 1991. Matrix metalloproteases and their inhibitors inconnective-tissue remodelling. FASEB Journal 5: 2145–2154.

Woods RK et al. 2002. Prevalence of food allergies in young adultsand their relationship to asthma, nasal allergies, and eczema.Annals of Allergy, Asthma & Immunology 88: 183–189.

Yates JR. 1998. Mass spectrometry and the age of the proteome.Journal of Mass Spectrometry 33: 1–19.

Yates JR. 2004. Mass spectral analysis in proteomics. An-nual Review of Biophysics and Biomolecular Structure 33:297–316.

Yates JR et al. 2009. Proteomics by mass spectrometry: approaches,advances, and applications. Annual Review of Biomedical Engi-neering 11: 49–79.

Yu C-J et al. 2003a. Proteomics and immunological analysis of anovel shrimp allergen, pen m 2. Journal of Immunology 170:445–453.

Yu YL et al. 2003b. Proteomic studies of macrophage-derived foamcell from human U937 cell line using two-dimensional gel elec-trophoresis and tandem mass spectrometry. Journal of Cardio-vascular Pharmacology 42: 782–789.

Zeece MG et al. 1989. High-resolution two-dimensional elec-trophoresis of bovine caseins. Journal Of Agricultural and FoodChemistry 37: 378–383.

Documents

Food Biochemistry and Food Processing (Simpson/Food Biochemistry and Food Processing) || Application of Proteomics to Fish Processing and Quality