Protein expression signatures: an application of proteomics

Data processing and classification analysis of proteomic changes: a case study of oil pollution in the mussel, Mytilus edulis

BACKGROUND

Proteomics may help to detect subtle pollution-related changes, such as responses to mixture pollution at low concentrations, where clear signs of toxicity are absent. The challenges associated with the analysis of large-scale multivariate proteomic datasets have been widely discussed in medical research and biomarker discovery. This concept has been introduced to ecotoxicology only recently, so data processing and classification analysis need to be refined before they can be readily applied in biomarker discovery and monitoring studies.

RESULTS

Data sets obtained from a case study of oil pollution in the Blue mussel were investigated for differential protein expression by retentate chromatography-mass spectrometry and decision tree classification. Different tissues and different settings were used to evaluate classifiers towards their discriminatory power. It was found that, due the intrinsic variability of the data sets, reliable classification of unknown samples could only be achieved on a broad statistical basis (n > 60) with the observed expression changes comprising high statistical significance and sufficient amplitude. The application of stringent criteria to guard against overfitting of the models eventually allowed satisfactory classification for only one of the investigated data sets and settings.

CONCLUSION

Machine learning techniques provide a promising approach to process and extract informative expression signatures from high-dimensional mass-spectrometry data. Even though characterisation of the proteins forming the expression signatures would be ideal, knowledge of the specific proteins is not mandatory for effective class discrimination. This may constitute a new biomarker approach in ecotoxicology, where working with organisms, which do not have sequenced genomes render protein identification by database searching problematic. However, data processing has to be critically evaluated and statistical constraints have to be considered before supervised classification algorithms are employed.

Proteomic analysis of protein profiles during early development of the zebrafish, Danio rerio

In the present study, profiles of protein expression were examined during early development of zebrafish, an increasingly popular experimental model in vertebrate development and human diseases. By 2-DE, an initial increase in protein spots from 6 h post-fertilization (hpf) to 8-10 hpf was observed. There was no dramatic change in protein profiles up to 18 hpf, but significant changes occurred in subsequent stages. Interestingly, 49% of the proteins detected at 6 hpf remained detectable by 1 week of age. To map the protein expression patterns in 2-D gels, MALDI-TOF/TOF MS was employed to identify selected protein spots from early embryos. 108 protein spots were found to match known proteins and they were derived from 55 distinct genes. Interestingly, 11 (20%) of them produced multiple protein isoforms or distinct cleavage products. Although deyolked embryos were used in the analysis, a large number of vitellogenin derivatives remained prominently present in the embryos. Other than these, most of the identified proteins are cytosolic, cytoskeletal and nuclear proteins, which are involved in diversified functions such as metabolism, cytoskeleton, translation, protein degradation, etc. Some of the proteins with interesting temporal expression profiles during development are further discussed.

Identification of Proteomic Signatures of Exposure to Marine Pollutants in Mussels (Mytilus edulis)

Bivalves and especially mussels are very good indicators of marine and estuarine pollution, and so they have been widely used in biomonitoring programs all around the world. However, traditional single parameter biomarkers face the problem of high sensitivity to biotic and abiotic factors. In our study, digestive gland peroxisome-enriched fractions of Mytilus edulis (L., 1758) were analyzed by DIGE and MS. We identified several proteomic signatures associated with the exposure to several marine pollutants (diallyl phthalate, PBDE-47, and bisphenol-A). Animals collected from North Atlantic Sea were exposed to the contaminants independently under controlled laboratory conditions. One hundred and eleven spots showed a significant increase or decrease in protein abundance in the two-dimensional electrophoresis maps from the groups exposed to pollutants. We obtained a unique protein expression signature of exposure to each of those chemical compounds. Moreover a set of proteins composed a proteomic signature in common to the three independent exposures. It is remarkable that the principal component analysis of these spots showed a discernible separation between groups, and so did the hierarchical clustering into four classes. The 14 proteins identified by MS participate in alpha- and beta-oxidation pathways, xenobiotic and amino acid metabolism, cell signaling, oxyradical metabolism, peroxisomal assembly, respiration, and the cytoskeleton. Our results suggest that proteomic signatures could become a valuable tool to monitor the presence of pollutants in field experiments where a mixture of pollutants is often present. Further studies on the identified proteins could provide crucial information to understand possible mechanisms of toxicity of single xenobiotics or mixtures of them in marine ecosystems.

Comparison of protein expression in plasma from nonylphenol and bisphenol A-exposed Atlantic cod (Gadus morhua) and turbot (Scophthalmus maximus) by use of SELDI-TOF

The overall objective of this study was to compare the expression of plasma proteins in juvenile cod and turbot after a 3 week exposure to two different chemicals known to be estrogenic: 4-nonylphenol (NP, 29 microg/L) and bisphenol A (BPA, 59 microg/L). ProteinChip) array technology in combination with surfaced enhanced laser desorption ionisation-time of flight (SELDI-TOF) was used to investigate general responses in plasma proteins. In addition, an indirect enzyme-linked immunosorbent assay (ELISA) was used to analyse two specific biomarkers of estrogenic exposure, vitellogenin (Vtg) and zona radiata protein (Zrp) in plasma. Both methods revealed clear species specific responses. In cod, 67% of significantly altered proteins showed the same response (up or down regulated) in NP and BPA exposed animals (males and females combined). The rest were either specific to NP (10%), BPA (19%) or they showed opposite responses to the two chemicals (4%). In contrast, only 20% of significantly altered proteins were common for NP and BPA exposed turbot: 60% were altered only in NP and 17% only in BPA. Furthermore, in BPA exposed cod, 77% of the responses were common for male and females, whereas turbot showed only 21% similarity for the two genders. However, NP exposed male and female turbot showed 88% similarity in responses. As gender was not determined in NP exposed cod, gender specific responses could not be determined. ELISA results supported that cod responded clearly to both chemicals as a large increase was observed in Vtg and Zrp levels. Turbot responded strongly to NP, but seemed only slightly affected by BPA. Overall, the results indicated that cod are more sensitive or respond with less specificity to estrogenic chemicals than turbot. The relatively large degree of common responses in NP and BPA exposed cod may indicate that in cod BPA have similar mode of action as NP. Generally, the results show the potential of SELDI-TOF as a tool for comparing multiple responses, and for identifying exposure as well as gender specific responses.

An ecotoxicoproteomic approach (SELDI-TOF mass spectrometry) to biomarker discovery in crab exposed to pollutants under laboratory conditions

Ciphergen ProteinChip Technology is a proteomic tool, used for the discovery of new and sensitive biomarkers. This approach was used to evaluate the protein profile of crabs exposed to various pollutants. Two different exposure experiments were performed: spider crabs (Hyas araneus) were exposed for 3 weeks to diallyl phatalate (DAP), bisphenol A (BisA) and polybrominated diphenyl ether (PBDE-47), while shore crabs (Carcinus maeanas) were exposed to crude oil, crude oil spiked with alkylphenols (APs) and 4-nonylphenol (NP). Gender and species-related protein pattern alterations were observed and compared to controls. Results showed different responses to pollutants by the two species. Major disruption in protein peak expression was observed in samples exposed to mixtures of pollutants, i.e. oil spiked with APs. Compared to shore crab, spider crab species showed a lower degree of response in terms of number of altered protein peaks following exposure. In general, female individuals of both species showed a larger number of significantly altered proteins compared to males. Data analysis by non-metric multi-dimensional scaling (MDS) was performed. Bi-dimesional-MDS plots revealed a good separation of groups for both spider and shore crabs. In some cases, a good discrimination can also be observed between the two genders within each treatment. Results highlight the potential of crabs as sentinel organisms for the aquatic environment. The results indicate that SELDI-ToF technology is a powerful tool to discover protein expression signatures for different pollutants and sex dependent responses.

Protein carbonylation and heat shock response in Ruditapes decussatus following p,p'-dichlorodiphenyldichloroethylene (DDE) exposure: a proteomic approach reveals that DDE causes oxidative stress

Protein carbonylation and levels of heat shock proteins (hsp; 60, 70 and 90 kDa) were measured in gill, mantle and digestive gland of Ruditapes decussatus following exposure to p,p'-dichlorodiphenyldichloroethylene (DDE). Heat shock response was measured by immunoblotting using antibodies specific to heat shock proteins (hsps). Densitometry analysis of individual bands revealed no difference between control and treated samples except appearance of hsp90 in DDE-treated mantle. Carbonylated protein content was determined following 2,4-dinitrophenylhydrazine derivatization and two-dimensional electrophoresis coupled with western blotting. Immunoblotting with dinitrophenol-specific antibody revealed extensive differences in both extent and number of carbonylated proteins in mantle and digestive gland in response to DDE while gill was unaffected. These results demonstrate for the first time that DDE causes tissue-specific formation of reactive oxygen species in clams.

Utility of proteomics to assess pollutant response of clams from the Doñana bank of Guadalquivir Estuary (SW Spain)

The utility of proteomics to assess pollutant response of Scrobicularia plana clams from three sites of Guadalquivir Estuary at the southern end of the National Park of Doñana (SW Spain) has been studied. Protein expression profiles were analyzed by 2-DE in soluble fractions of S. plana gills. Nearly 2000 well-resolved spots were detected in silver-stained gels, with focused areas in the 4-6.5 pH range. Different protein expression signatures were found at each site, with the highest number of more intense spots in animals with the highest metal content. Nineteen more intense protein spots were analyzed out by nanospray-ion trap tandem mass spectrometry, de novo sequencing and a bioinformatics search for their possible identification. While sequence tags of 16 more intense protein spots were obtained, including several proteins induced by pollutant exposure of model organisms, only 2 proteins were unambiguously identified: hypoxanthine-guanine phosphoribosyltransferase (HPRT) and glyceraldehyde-3-phosphate dehydrogenase (G3PDH). Both enzymes were significantly higher in animals with the highest metal contents. Thus, we propose these two proteins, HPRT and G3PDH, as novel pollution biomarkers.

Differential protein expression profiles in anterior gills of Eriocheir sinensis during acclimation to cadmium

Using a proteomic approach, we characterized different protein expression profiles in anterior gills of the Chinese mitten crab, Eriocheir sinensis, after cadmium (Cd) exposure. Two experimental conditions were tested: (i) an acute exposure (i.e. 500 microg Cd l(-1) for 3 days) for which physiological, biochemical and ultrastructural damage have been observed previously; (ii) a chronic exposure (i.e. 50 microg Cd l(-1) for 30 days) resulting in physiological acclimation, i.e. increased resistance to a subsequent acute exposure. Two-dimensional gel electrophoresis (2-DE) revealed six protein spots differentially expressed after acute, and 31 after chronic Cd exposure. From these spots, 15 protein species were identified using MS/MS micro-sequencing and MS BLAST database searches. Alpha tubulin, glutathione S-transferase and crustacean calcium-binding protein 23 were down-regulated after an acute exposure, whereas another glutathione S-transferase isoform was up-regulated. Furthermore, analyses revealed the over-expression of protein disulfide isomerase, thioredoxin peroxidase, glutathione S-transferase, a proteasome subunit and cathepsin D after chronic exposure. Under the same condition, ATP synthase beta, alpha tubulin, arginine kinase, glyceraldehyde-3-phosphate dehydrogenase and malate dehydrogenase were down-regulated. These results demonstrate that acute and chronic exposure to waterborne Cd induced different responses at the protein expression level. Protein identification supports the idea that Cd mainly exerts its toxicity through oxidative stress induction and sulfhydryl-group binding. As a result, analyses showed the up-regulation of several antioxidant enzymes and chaperonins during acclimation process. The gill proteolytic capacity seems also to be increased. On the other hand, the clearly decreased abundance of several enzymes involved in energy transfer suggests that chronic metal exposure induced an important metabolic reshuffling.

A proteomics approach to identifying fish cell lines

Fish cell lines are relatively easy to culture and most have simple growth requirements that make cross contamination a potential problem. Cell line contamination is not an uncommon incident in laboratories handling more than one cell line and many reports have been made on cross contamination of mammalian cell lines. Although problems of misidentification and cross-contamination of fish cell lines have rarely been reported, these are issues of concern for cell culturists that can make scientific results and their reproducibility unreliable. Proper identification of cell lines is thus crucial and protocols for routine and rapid screening are preferred. Cytogenetic evaluation, DNA fingerprinting, microsatellite analysis and PCR methods have been published for inter-species identification of many cell lines, but discerning intra-species contamination has been challenging. More complex DNA fingerprinting and hybridization techniques coupled with isoenzyme analysis have been developed to discriminate intra-species contamination, however, these require complex and time consuming procedures to enable cell identification thus are difficult to apply for routine use. A simple proteomic approach has been made to identify several fish cell lines derived from tissues of the same or differing species. Protein expression signatures (PES) of the evaluated fish cell lines have been developed using 2-DE and image analysis. A higher degree of concordance was seen among cell lines derived from rainbow trout, than from other fish species. Similar concordance was seen in cells derived from the same tissues than from other tissues within the same species. These profiles have been saved in an electronic databank and could be made available to be used for discerning the origins of the various cell lines evaluated. This proteomic approach could thus serve as an additional, valuable and reliable technique for the identification of fish cell lines.

Peroxisomal proteomics, a new tool for risk assessment of peroxisome proliferating pollutants in the marine environment

In an attempt to improve the detection of peroxisome proliferation as a biomarker in environmental pollution assessment, we have applied a novel approach based on peroxisomal proteomics. Peroxisomal proteins from digestive glands of mussels Mytilus galloprovincialis were analyzed using 2-DE and MS. We have generated a reference 2-DE map from samples obtained in a well-studied reference area and compared this with peroxisomal proteomes from other sequenced genomes. In addition, by comparing 2-DE maps from control samples with samples obtained in a polluted area, we have characterized the peroxisome proliferation expression pattern associated with exposure to a polluted environment. Over 100 spots were reproducibly resolved per 2-DE map; 55 differentially expressed spots were quantitatively detected and analyzed, and 14 of these showed an increase in protein expression of more than fourfold. Epoxide hydrolase, peroxisomal antioxidant enzyme, and sarcosine oxidase (SOX) have been identified by ESI MS/MS, and acyl-CoA oxidase, multifunctional protein, and Cu,Zn-superoxide dismutase were immunolocalized by Western blotting. Our results indicate that a peroxisomal protein pattern associated to marine pollutant exposure can be generated, and this approach may have a greater potential as biomarker than traditional, single-protein markers.

Proteomic analysis of cadmium-induced protein profile alterations from marine alga Nannochloropsis oculata

Protein profile alterations following exposure to cadmium were examined in marine alga Nannochloropsis oculata through proteomic analysis. Alterations of the protein expression patterns following 10 muM cadmium treatment were analyzed on 2-dimensional gels. Out of 380 protein spots detected on 2-D gel using Coomassie staining, 11 spots were changed significantly following cadmium treatment. Because of the non-availability of molecular background information on this non-sequenced algal species, cross-species protein identification through ESI-Q-TOF MS/MS was used to identify altered proteins. Two newly induced proteins were identified as malate dehydrogenase orthologue and NADH dehydrogenase orthologue. One suppressed protein was identified to be glyceraldehydes 3-phosphate dehydrogenase A. Protein spot showing a 3-fold increase was identified as mitochondrial NADH: ubiquinone oxidoreductase orthologue. However, we could not find any matches in the database from ESI-Q-TOF MS/MS for the remaining seven proteins, thus only partial peptide sequences of these proteins were found.

Protein expression patterns in zebrafish skeletal muscle: initial characterization and the effects of hypoxic exposure

Patterns of protein expression were examined in white skeletal muscle from adult zebrafish (Danio rerio). High resolution two-dimensional gel electrophoresis resolved between 300 and 400 spots with molecular masses between 20 and 120 kDa and isoelectric points between about 5 and 8. Forty spots, representing a range of protein size, charge, and abundance were excised, digested with trypsin, and subjected to matrix-assisted laser-desorption/ionisation-time of flight mass spectrometry for protein identification. Twenty-nine spots were identified, including enzymes of energy metabolism, contractile proteins, an iron transport protein, and a heat shock protein. In addition, several spots matched theoretical proteins predicted from genome sequencing. These theoretical proteins were tentatively identified by similarity to known proteins. Patterns of muscle protein expression were then measured after zebrafish were exposed to low oxygen (16 torr) for 48 h, an exposure previously shown to increase the survival of zebrafish at more severe reductions in oxygen. Exposure to low oxygen (hypoxia) did not change the general pattern of protein expression but did affect the amounts of six low abundance proteins. The relatively subtle effects of hypoxia on patterns of muscle protein expression contrasts the widespread changes previously documented in mRNA levels in this and other species of fish during hypoxic stress. The difference between protein and mRNA expression illustrates the need to integrate both measures for a more complete understanding of gene expression in fish during hypoxic exposure.

Ecotoxicogenomics: the challenge of integrating genomics into aquatic and terrestrial ecotoxicology

Rapid progress in the field of genomics (the study of how an individual's entire genetic make-up, the genome, translates into biological functions) is beginning to provide tools that may assist our understanding of how chemicals can impact on human and ecosystem health. In many ways, if scientific and regulatory efforts in the 20th century have sought to establish which chemicals cause damage to ecosystems, then the challenge in ecotoxicology for the 21st century is to understand the mechanisms of toxicity to different wildlife species. In the human context, 'toxicogenomics' is the study of expression of genes important in adaptive responses to toxic exposures and a reflection of the toxic processes per se. Given the parallel implications for ecological (environmental) risk assessment, we propose the term 'ecotoxicogenomics' to describe the integration of genomics (transcriptomics, proteomics and metabolomics) into ecotoxicology. Ecotoxicogenomics is defined as the study of gene and protein expression in non-target organisms that is important in responses to environmental toxicant exposures. The potential of ecotoxicogenomic tools in ecological risk assessment seems great. Many of the standardized methods used to assess potential impact of chemicals on aquatic organisms rely on measuring whole-organism responses (e.g. mortality, growth, reproduction) of generally sensitive indicator species at maintained concentrations, and deriving 'endpoints' based on these phenomena (e.g. median lethal concentrations, no observed effect concentrations, etc.). Whilst such phenomenological approaches are useful for identifying chemicals of potential concern they provide little understanding of the mechanism of chemical toxicity. Without this understanding, it will be difficult to address some of the key challenges that currently face aquatic ecotoxicology, e.g. predicting toxicant responses across the very broad diversity of the phylogenetic groups present in aquatic ecosystems; estimating how changes at one ecological level or organisation will affect other levels (e.g. predicting population-level effects); predicting the influence of time-varying exposure on toxicant responses. Ecotoxicogenomic tools may provide us with a better mechanistic understanding of aquatic ecotoxicology. For ecotoxicogenomics to fulfil its potential, collaborative efforts are necessary through the parallel use of model microorganisms (e.g. Saccharomyces cerevisiae) together with aquatic (e.g. Danio rerio, Daphnia magna, Lemna minor and Xenopus tropicalis) and terrestrial (e.g. Arabidopsis thailiana, Caenorhabdites elegans and Eisenia foetida) plants, animals and microorganisms.

NMR-based metabolomics: a powerful approach for characterizing the effects of environmental stressors on organism health

It is important to assess the chronic effects of chemical, physical, and biological stressors on organisms in the environment. Appropriate methods must enable rapid, inexpensive, and multibiomarker analyses of organism health. Here we investigate withering syndrome in red abalone (Haliotis rufescens), an important wild and farmed shellfish species along the Pacific coast, using a metabolomic approach that combines the metabolic profiling capabilities of nuclear magnetic resonance spectroscopy (NMR) with pattern recognition methods. Foot muscle, digestive gland, and hemolymph samples were collected from healthy, stunted, and diseased abalone, and the extracts were analyzed by NMR. Following spectral preprocessing, principal components analyses of the metabolite profiles were conducted. Our results confirm that NMR-based metabolomics can successfully distinguish the biochemical profiles of the three groups of animals, in every type of tissue or biofluid studied. Furthermore, this discovery-based approach successfully identified novel metabolic biomarker profiles associated with withering syndrome. The application of these methods for investigating other environmental stressors is discussed, as are the advantages of NMR-based metabolomics for biomonitoring, particularly in conjunction with gene and protein expression profiling.