Print-friendly report

Coastal Ecosystem
Water Quality
Biological Contaminants

Technical Details


Photos and Video

Close window
to return to list


[Close window to return to list]

Proteomics To Develop Relevant Phenotypic Biomarkers Of Environmental Impacts In Wild Marine Fishes Of Southern California Website

Principal Investigator(s):
Kevin M. Kelley - California State University, Long Beach
Andrew Mason - California State University, Long Beach

Associate Investigator(s):
Ashraf Elamin - California State University, Long Beach
Jeffrey L. Armstrong - Orange County Sanitation District

Period: 1/1/2008 - 12/1/2010

Prior collaborative work between our research groups has uncovered several striking instances of endocrine-disrupted states in marine fish species exposed to impacted environments of the Southern California Bight, such as near wastewater treatment plant (WWTP) outfalls or urban river outlets. The findings have made it clear that endocrine and physiological disruption is evident in a variety of different species, and that different endocrine systems can be significantly and simultaneously impacted dependent upon the study location and types of contaminants present.

Most studies to date, including those conducted in our laboratories, have developed and used single or a limited set of biomarkers (e.g., measurement of hormone level, expression of a steroidogenic enzyme gene, etc.) to identify an impact in fish. However, the choice of which biomarker to use can lead to differences in data and interpretation, and such single endpoint bioassay approaches typically provide little or no insight on the larger physiological status of an impacted animal, leaving unanswered many important questions. It is the integrative perspective of an animal's biological response to a given environment that will be most directly relevant to estimating potential adverse effects.

Consequently, there is a significant need to develop more powerful diagnostic methodologies that will allow an integrated, relevant assessment of environmental effects in wildlife. While single endpoint biomarkers will continue to have value in environmental studies and assessment, the absence of a more comprehensive, “integrative” biomarker approach constrains our ability to establish cause-effect relationships, predict adverse environmental impacts on animals and to identify mitigation strategies.

Recent technological advances in the areas of genomics and proteomics have provided researchers with new tools for developing biomarkers. Most impressively, these technologies allow simultaneous measurement of multiple biomarkers that, collectively, can reflect the types of chemical (and/or other factors) exposure and the systemic (integrated) biological effects in animals. While genomics is proving useful in developing single and multiple biomarkers, it is important to recognize changes in gene transcription do not necessarily correlate with protein expression or protein activity, and that proteins are the primarily effectors for adaptive changes and cellular and physiological functions in response to environmental stimuli. Thus, proteomics provides for an assessment of relevant phenotypic alterations, in which changes in the expression of hundreds or more of proteins can be measured simultaneously. While the individual proteins of interest can be developed as specific biomarkers themselves, the larger protein expression profile (PEP), or “fingerprint”, may also be used as an integrative biomarker of the “phenotypic” condition. Thus, a comparison of a control vs. altered condition in an animal reveals a set of biomarkers specifically indicative of the altered state and environmental factors at play. Therefore, with the advent of this emergent technology, we are now presented with the exciting opportunity to develop PEPs and new biomarkers as powerful diagnostic tools for classifying chemical exposures, predicting mode(s) of action, and for assessment of biological impacts.

Preliminary data generated from 2-dimensional gel electrophoretic (2-D GE) analyses of hepatic tissues of English sole (Pleuronectes vetulus) sampled from an EPA reference site (off of Dana Point) and a WWTP location indicate substantial differences in their PEP, pointing to a strong potential for developing biomarkers reflecting environmental effects. From these animals, we have been able to identify a set of new biomarker proteins in preliminary work, whose changes in impacted English sole reflect important toxicological responses and physiological alterations.

The overall research objective of this proposal is to discover new biomarkers and develop PEPs that can eventually be used in environmental assessment as a screening tool to identify/classify chemical exposures and their impacts in wild marine fish in the urban ocean setting. As part of this effort, specific biomarkers and PEPs will be correlated with environmental and bioaccumulated tissue contaminant levels, as well as with physiological and endocrine measures of impact, allowing a highly integrative (“systems” approach) interpretation of the data.

In prior collaborative work between CSULB and OCSD, English sole collected at the OCSD outfall and other impacted sites on the San Pedro Shelf have been found to consistently exhibit significant endocrine and physiological disruption. In this project, we plan to utilize this well-documented, impacted population and compare them to English sole an EPA reference site near Dana Point, CA. Fish will be caught by otter trawl coinciding with the biannual Ocean Monitoring Program of OCSD and sampled on board for blood and tissue (liver). During these monitoring efforts, OCSD takes sediment samples for the analysis of contaminant levels in the environment (metals and organics), in addition to collecting a host of other oceanographic, environmental data. Blood plasma will be prepared and saved at -80 C for subsequent endocrine analyses (cortisol, IGF-I, sex steroids, thyroid hormones) to document that the fish are physiologically impacted. Liver samples from these fish will be prepared for proteomic analysis, and metal and organic contaminant measurements. The CSU system-wide core facilities for proteomics (see below) and trace chemical analysis (ICP-MS and GC/MS will be utilized) are housed on the Long Beach campus and will be utilized in these studies. For the proteomic analysis, liver samples will be subjected to comparative 2D GE to identify differences in hepatic protein expression in fish collected from control and outfall locations. Consistently altered proteins will be then be analyzed by MALDI-TOF/TOF mass spectrometry and identified using NCBI databases and de novo sequencing approaches. Our preliminary data demonstrate the use of both techniques to identify a subset of novel biomarkers in English sole. Through the studies proposed here, a highly integrative approach will be taken. Thus, the same individuals will be measured for specific novel biomarkers, PEPs, specific contaminants (hepatic and from field site), and endocrine parameters of growth, thyroid, stress response, and reproductive status. This will enable correlative analyses to be performed, identifying relationships among environmental contaminants, measured factors and protein phenotypic changes, particularly specific protein biomarkers.

Worldwide, adverse health effects to wildlife and humans resulting from environmental exposure to persistent industrial, pharmaceutical, and natural chemicals have been increasingly documented, raising significant international concerns. The effects of such chemicals are often transduced through endocrine systems, with other important effects including actions on detoxification enzyme systems, intracellular signaling pathways or proteins, and the like. Singly, and collectively, these perturbations can significantly alter physiological homeostasis and impair health. Surprisingly, these issues have been significantly understudied on the Pacific Coast, particularly in the Southern California Bight region, which is a marine environment affected by one of the largest human populations in the U.S.A. Studies to understand impacts specifically in marine environments are also lacking worldwide.

The proposed studies will take a first step toward developing specific phenotypic biomarkers of environmental effects in the Southern California Bight and are positioned to point to potential causative factors (contaminants) in the environment. This project brings together academic research laboratories offering expertise in fish endocrine physiology and toxicology, proteomics, and trace chemical analysis, with an important regional WWTP (OCSD), which has direct interest in the information to be generated and is providing both expertise and the cost-matching to support a successful outcome.

It is the intent of this project to take advantage of this defined environment-based experimental model to develop the proteomic technology on the English sole. We have already seen the power of the technique in our preliminary work, and with the results of the proposed study as a “proof of technique”, we plan to expand its use among our collaborations and colleagues locally, and also to those interested from outside of our region (a wide distribution of the English sole populations, combined with close pleuronectid relatives in similar environments worldwide, support this possibility). The P.I.s maintain ongoing, important collaborations with key regional governmental agencies (e.g., OCSD, CLAEMD, LACSD, SCCWRP, SWQCB) with whom there is a vigorous degree of cooperation and communication, and the outcome of this work (development of effective diagnostic and screening tools) will be directly disseminated to them and hopefully incorporated into ongoing environmental monitoring efforts regionally and beyond. Funding the development of new technologies is an essential role of the Sea Grant Program, given the very limited capability of these regional agencies to do so.

We will continue in our ongoing substantial efforts to actively communicate our scientific findings directly to local agencies, to publish the data in the professional peer reviewed literature and other publications, to establish informational web-pages under www.csulb.edu/depts/endo and www.csulb.edu/programs/iirmes with links to agencies, organizations and universities, and we will present the results at professional conferences and other meetings to provide outreach to stakeholders. In addition, a graduate student is planned as a SG trainee in this project, providing a direct educational benefit to a training marine scientist, and the involvement of three other students is planned, two of whom are underrepresented minority undergraduates.

Summary and Update, April 2010

Proteomics to Develop Phenotype Biomarkers of Environmental Impacts in Wild Marine Fishes of Southern California
(Kevin Kelley, Cal State University Long Beach)

Previous research has shown that anthropogenic chemicals in the marine environment have caused endocrine and physiological disruption, such as changes to reproductive, metabolic, and growth status, that has strong potential to threaten survival in a variety ofþdifferent species residing in coastal California, including flatfish, surfperch and sculpin. Most all studies to date that have monitored these types of impacts to fish have used single or a limited set of biomarkers (e.g., measurement of hormone or mRNA level). These researchers are developing more powerful diagnostic tools using protein expression profiles (PEP), or "fingerprints," that will enhance assessment and understanding of environmental effects in marine organisms. This information will help lead to better identification of pollutants and the mechanisms by which they impact marine wildlife and ecosystems, which in turn can help state resource managers and regulators reduce the sources of contamination that negatively affect wildlife.

In collaboration with the Orange County Sanitation District (OCSD) Ocean Monitoring Program, we captured and sampled wild English sole residing at the OCSD outfall location offshore of Huntington Beach, at the OCSD reference ("far-field") location offshore of Bolsa Chica, and from an EPA reference location offshore of Dana Point. These samples were collected in August of two consecutive years (2008, 2009), and included blood plasma and liver tissues from each individual. Plasma samples were analyzed by radioimmunoassay procedures to determine concentrations of four different hormones, including the stress hormone, cortisol, the thyroid hormone, thyroxine (T4), the fish-specific androgen, 11-ketotestosterone (11-KT), and the estrogen, 17ß-estradiol (E2). Livers were measured for organic contaminant concentrations (e.g., chlorinated pesticides, PCBs, PAHs) using gas chromatography/mass spectrometry (GC/MS) techniques. In addition, proteome analyses of subsamples of the same livers are nearly completed, which determine PEP differences among fish and between study sites. The latter process is leading to the elucidation of new protein biomarkers of environmental impact. In addition, analysis of the relationships between contaminant exposures and phenotype changes in the animals are beginning to identify specific contaminant chemicals (and classes of contaminants) as candidate causative agents in the southern California marine environment that may underlie the phenotypic changes.

Animals from the different study sites exhibit distinct patterns of tissue-accumulated contaminants. In general, fish from Dana Point have the lowest overall hepatic concentrations in each class of contaminant, with the rare exception that a few PAH congeners (e.g., phenanthrene, 2,3,5-trimethylnapthlene) are relatively elevated in the Dana Point fish. In agreement with continuing OCSD ocean monitoring efforts (e.g., 2007-8, 2008-9 OCSD Ocean Monitoring Annual Reports), total PCBs are highest in fish sampled from the OCSD outfall location (564.5 +- 109 ng/g ww) as compared with the far-field site (287.4 +- 68.9 ng/g ww) and Dana Point (123.8 +- 32.1 ng/g ww). The individual PCB congeners also exhibit this spatial pattern (Figure 1). Also consistent with prior OCSD monitoring data, PAHs (notwithstanding the congeners mentioned above) and DDTs did not exhibit any outfall-associated differences, while chlordanes were higher in fish from the outfall location.

The successful measurements of contaminant concentrations within individual fish livers, and the range of concentrations among the different contaminants observed, enables us to statistically evaluate the degree to which each contaminant measured in the fish may be related to endocrine alterations and other phenotypic changes (e.g., PEPs). Our results to date clearly indicate that different classes of contaminants are significantly correlated with different kinds of endocrine disruption. For example, while PCB congeners showed almost no relationships with cortisol response in the fish, there were numerous, highly significant correlations between the thyroid hormone T4 and several PCB congeners. Such data are in agreement with our recently reported findings in San Francisco Bay on two other fish species (sculpin and surfperch; Brar et al., 2010). In both studies, several of the same PCB congeners were significantly correlated with thyroid disruptions (PCBs 095, 097, 099, 105, 110), and there was also substantial concordance in the PCB congeners that were not related to thyroid disruptions. In the present study, Pearson correlation coefficients were R=0.6-0.75 (p<0.05) for 16 congeners (a representative relationship is shown in Figure 2). Both studies also point to PCB congeners that are not associated with dioxin-like effects (i.e., those activating arylhydrocarbon receptors and down-stream biotransformation activities) but rather which may direct their effects on fish through thyroid-disrupting activities. T4 levels were also related to chlordanes (R=0.67-0.75, p<0.05), but not to DDT or DDT metabolites. Cortisol responses, on the other hand, were significantly correlated with DDT and DDT metabolite concentrations (R=0.7-0.85, p<0.05), but not with chlordanes. Additionally, while the PAH, anthracene, showed some relationship with cortisol response (R=0.5, p<0.05), T4 showed some relationships with napthalenes (e.g., for 2-methylnaphthalene, R=0.71, p<0.01) that were not evident for cortisol. In contrast to these endocrine systems, there are no discernable relationships between the sex steroids, E2 and 11-KT, and organochlorines (chlorinated pesticides) or PAH compounds. Correlations with PCBs, however, are beginning to show some significant relationships with both sex steroids in ongoing analyses. Overall, the findings to date are providing novel insights into the potential endocrine disruptive effects of existing environmental contaminants to which resident fish are exposed. The concordance with the study of San Francisco Bay fish, and the differentiation of types of endocrine system impacts with distinct chemical classes, provide some confidence that important candidate EDCs in the southern California Bight may be emerging from this study.

The hepatic proteome of each of the above experimental fish have been analyzed using 2-D gel electrophoresis and computer-assisted analytical techniques to determine differences in the expression of specific proteins. Results to date indicate consistent, and sometimes substantial, alterations in different hepatic proteins across the study sites. At present, we are in an extensive process to identifying each protein of interest, using MALDI TOF/TOF mass spectrometry approaches. We have thus far identified a number of proteins representing distinct, important physiological and cellular roles in the animals. These include induced proteins involved in detoxification and stress responses (e.g., glutathione transferase, Cu/Zn superoxide dismutase, homogentisate 1,2-dioxygenase, heat shock protein GP96), metabolic adaptation or gluconeogenesis enzymes typically altered in catabolic responses (aldolase B, enolase-alpha, fatty acid binding proteins, fumarylacetoacetate hydrolase), intracellular signal transduction proteins (calmodulin, eukaryotic translation initiation factor-3), cell structural proteins (intermediate filament, actin, keratin), and endocrine response proteins (vitellogenin). This work is ongoing, and the plan is to identify more than 50 different expressed proteins. The second phase of the identification process is to use de novo sequencing approaches to identify proteins that are not confidently identifiable using the primary MS spectra. The findings to date already provide insight into the physiological status of fish residing in the outfall location: they are undergoing toxicological responses to organic contaminant exposures and possibly to oxidants, they appear to be in a catabolic condition in which storage tissues must be used to provide metabolic fuel to the animals (typical under conditions of physiological stress), their liver tissues are exhibiting intracellular signaling and structural alterations, and they have associated disruptions in their endocrine systems. When the protein identification phase of the study is finalized, we will then complete the correlative analyses (as described above for contaminants vs. endocrine parameters) to elucidate relationships between the different expressed proteins (and classes of proteins), hepatic concentrations of contaminants, and the endocrine system parameters. Upon completion of the last phase of the study through summer 2010, we expect to provide detailed, novel insight into the physiological condition of the fish, in addition to identifying specific environmental contaminants as candidate causative agents in endocrine and phenotypic impacts in the fish.

Two graduate students, two undergraduate students, and a technician have been directly involved in this research and received training in the techniques involved. Sea Grant Trainee, Ms. Claire Waggoner, has made excellent progress, completing her coursework and generating significant databases that form the basis of her Thesis and which contribute significantly to this project. Co-P.I. Dr. Jeffrey Armstrong of the OCSD Ocean Monitoring Division has been communicating directly with managers at his institution on the project activities and its findings. He is also providing expertise statistically to evaluate our database using multivariate analyses.

A variety of scientific conference presentations by the P.I. and students have served to disseminate findings generated from this work, including (1-3 presentations/conference; *invited):

International Congress on the Biology of Fish (Portland, OR, July 2008)*

Society for Environmental Toxicology and Chemistry (SETAC) North America (Tampa, FL, November 2008)*

Southern California SETAC Annual Meeting (Dana Point, CA, May 2008)*

20th (Oakland, CA, January 2008) and 21st (Los Angeles, CA, January 2009) Annual CSU Biotechnology Symposium

CALFED Science Conference (Sacramento, CA, October, 2008)

California Water Environment Association Annual Meeting (Monterey, CA, March, 2009)*

SETAC North America (New Orleans, LA, November 2009)

Society for Integrative and Comparative Biology (SICB) Annual Meeting (Seattle, WA, January 2010)*

World Aquaculture 2010 Annual Meeting (San Diego, CA, March 2010)*

We are also outreaching to the public, at the annual University by the Sea celebration in Long Beach, CA (October 2008 lecture called "Pollution in Our Seas--What Happens to the Fish?"), at the Fellows Colloquium at CSULB (April 2010 lecture called "Sex, Drugs, and Radical Chemistry"), in the Beach Review (article called "New Views of Water Quality"), and in the State of the Estuary 2008 (SF Estuary Project, article called "Environmental Endocrine Disruption"). Additional outreach activities have included proposing and discussing the use of these technologies in environmental studies to the San Francisco Estuary Institute (SFEI) and San Francisco Bay resource managers present at the Annual Meeting of the Regional Monitoring Program (Oakland, CA, October 2, 2007), and RMP meetings at SFEI (2008, 2009). A new website is under construction to enhance outreach to both professional and lay audiences (www.ISC-BIO.org/EEL).

We expect a minimum of two manuscripts to be produced as a result of this work, which will be submitted for publication in the peer-reviewed scientific literature.

Figure 1. Hepatic concentrations of polychlorinated biphenyl (PCB) congeners (indicated across x-axis) in English sole from the OCSD outfall location (T1) and far-field (T11) monitoring stations and Dana Point (DP) sampled in August 2009. The outfall location (T1) is associated with higher PCB exposures in fish as compared with sites T11 and DP (p<0.05, most congeners).

Figure 2. Relationship between hepatic concentrations of PCB 186 (ng/g wet weight) and plasma concentrations of the thyroid hormone, thyroxine (T4; ng/ml). Comparable to results for 15 other PCB congeners as well as for total PCBs, significant correlations have been observed between exposure to PCBs and T4 (R=0.60--0.75, p<0.05).


Publications & other print media:
Evidence for thyroid endocrine disruption in wild fish in San Francisco Bay, California, USA. Relationships to contaminant exposures
Aquatic Toxicology, Volume 96, Issue 3, 18 February 2010, Pages 203-215
Navdeep K. Brar, Claire Waggoner, Jesus A. Reyes, Russell Fairey, Kevin M. Kelley



Video, electronic, and computer products:

Photo4 Photo1
FIGURE 1: Study site within the OCSD Ocean Monitoring Program area off shore of Orange County, CA. The EPA reference site is located south of OCSD, off shore of Dana Point, CA (Lat 33Ý 25.000', Long 117Ý 41.000'), while the T-1 outfall site is located off shore of Huntington & Newport Beaches, CA (Lat 33Ý 34.641', Long 118Ý 00.567'). Both trawl sites have a depth of 60 m. FIGURE 2: Endocrine disruption in English sole residing at the OCSD outfall site (T1), in comparison with fish sampled from OCSD far-field reference sites. Hormone concentrations were measured by specific radioimmunoassay and are expressed as
mean ± SEM of ng/ml values (n is shown in parentheses; asterisk indicates significant difference, p<0.01). Panel [A] illustrates that stress-induced cortisol responses* are significantly reduced by >50% in fish from the outfall location (*response after identical 40 min catching and handling protocol, previously established to determine peak stress-induced cortisol levels). Similarly in panel [B], plasma concentrations of the growth regulator IGF-I (panel [B]) and of the thyroid hormone, thyroxin (panel [C]), are significantly depressed in fish sampled from the outfall location.
Photo2 Photo3
FIGURE 3: Results of 2-D gel electrophoretic analysis of hepatic tissues from English sole residing at the EPA reference site (left) and the OCSD T-I outfall site (right). Five proteins exhibiting >2-fold changes between fish were excised and analyzed by MALDI TOF/TOF mass spectrometry (see next figure). FIGURE 4: MS spectrum of tryptic peptides from protein “c” (calmoduliin), as identified by 2D GE (shown in Fig. 3). Individual peptide mass/charge (m/z) values shown at the top of each peptide peak.