|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 REPRODUCTIVE AND RENAL PATHOLOGIES IN THE GIANT TOAD ( Bufo marinus ) ARE ASSOCIATED WITH HUMAN LAND USE PRACTICES By KRISTA ANN-MARIE MCCOY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007
2 2007 Krista Ann-Marie McCoy
3 To Michael Thank you for enduring the in itial insanity and for celebrating th e successes
4 ACKNOWLEDGMENTS I acknowledge and thank Michael McCoy for bei ng a constant source of wonderful ideas, constructive feedback, and friendship. I thank m y advisors Colette M. St Mary and Louis J. Guillette Jr. for providing me with the opportunity to conduct high quality science in a safe and intellectually stimulating environm ent. I thank Dr. C. Kenneth Dodd and Craig W. Osenberg for providing comments and critical f eedback on my writing and statisti cal approaches. I thank Dr. Steven M. Phelps for all the wonderful convers ations about my work and for provoking me to think about things from a completely different perspective. I have had many undergraduate and high school research assistants th at have been instrumental in my success. Although there are too many to list here, I would like to thank Patr ick A. Barnes Jr., Vict oria Bender, Michele Fromowitz, Kristin Hopkins, Michele A. Rehr er, and Heather Wineman for help collecting animals. I also thank Eric A. DeVeaux, Steven Lroch, and Brittany Nagy for help conducting histology. I acknowledge Laurie l Bortnick, Chelsey Campbell, and Kim Hoang for help with various aspects of my work and for spear head ing various important i ndependent projects. Lauriel Bortnick characterized the color and nuptial pad differences, and Chelsey Campbell conducted the arm width measurements in toads th at are presented in chapter 2. Kim Hoang was instrumental in the development and data colle ction for the renal pathology study presented in chapter 5 of this volume. I also acknowledge the staff members of the Zoology Department and the School of Natural Resources and Environmen t for logistic, monetary, and at times (much needed) emotional support. I would like to th ank Lori Albergotti, Leslie Babonis, and Kelly Hyndman for much appreciated critical fee dback on my writing, for thought provoking (and fun) conversations about physiology, for helping me conduct molecular lab work, and for being a great people and friends. I th ank Dieldrich Bermudez and Heather Hamlin for their help conducting the hormone assay presented in chapte r three. I thank Caren Helbing, Lan Ji, and
5 Mark Gunderson for their help conducting the cDNA array study presented in chapter four. I also so the Zoology departments graduate students for a lot of fun times that were absolutely required for maintaining ones sanity. Finall y, I thank my mom and dad for finding their own ways to help me develop into a strong, fearle ss, independent person motiv ated by the desire to find the truth.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES.......................................................................................................................10 ABSTRACT...................................................................................................................................12 CHAP TER 1 ENDOCRINE DISRUPTING CHEM ICALS IN AMPHIBIANS ......................................... 14 Introduction................................................................................................................... ..........14 Brief Introduction to the Endocrine System ........................................................................... 15 Definition of Endocrine Disrupting Che micals...................................................................... 16 The Nature and Sources of E ndocrine Disrupting Che micals................................................18 Modes of ActionHow and Why?........................................................................................21 Nuclear Receptors........................................................................................................... 22 Hormone Catabolism and Carrier Protein Binding......................................................... 23 Steroid Biosynthesis........................................................................................................24 Epigenetics......................................................................................................................25 Phenotypic Responses........................................................................................................... .28 Thyroid............................................................................................................................29 DevelopmentHomeobox Gene Expression .................................................................. 34 Reproduction...................................................................................................................37 Behavior..........................................................................................................................45 Implications................................................................................................................... .........49 Conservation and Ecology...............................................................................................49 What Needs to Be Known...................................................................................................... 54 2 GROSS GONADAL ABNORMALITIES AND ALTERE D SECONDARY SEXUAL TRAITS ARE ASSOCIATED WITH AGRICULTURE....................................................... 62 Introduction................................................................................................................... ..........62 Methods..................................................................................................................................64 Study System...................................................................................................................64 Study Sites and Collection Techniques........................................................................... 66 Analysis of Primary Sexual Characteristics.................................................................... 67 Analysis of Secondary Sexual Characteristics................................................................ 69 Nuptial pads..............................................................................................................69 Color pattern.............................................................................................................69 Forelimb size............................................................................................................70 Results.....................................................................................................................................70 Analysis of Primary Sexual Characteristics.................................................................... 70
7 Secondary Sexual Characteristics.................................................................................... 71 Nuptial pads..............................................................................................................71 Mottling....................................................................................................................71 Forelimb size............................................................................................................71 Discussion...............................................................................................................................72 3 MARINE TOADS LIVING IN AGRI CULTURAL AREAS HAVE ALTERED GONADAL FUNCTION .......................................................................................................84 Introduction................................................................................................................... ..........84 Methods..................................................................................................................................85 Histological Preparations.................................................................................................86 Quantification of Spermatogenesis..................................................................................86 Hormone Assays..............................................................................................................86 Statistical Tests.............................................................................................................. ..87 Site Comparisons............................................................................................................. 88 Sex Comparisons............................................................................................................. 88 Results.....................................................................................................................................89 Spermatogenesis Site Com parisons.............................................................................. 89 Spermatogenesis Sex Comparisons.............................................................................. 90 Hormone Concentrations Site Comparisons................................................................ 90 Hormone Concentrations Sex Comparisons................................................................. 90 Discussion...............................................................................................................................91 4 GENE EXPRESSION PROFILE DI FFERE NCES AMONG OVARIAN, TESTICULAR, AND INTERSEX GONADAL TISSUE OF WILD CAUGHT ADULT MARINE TOADS................................................................................................................105 Introduction................................................................................................................... ........105 Methods................................................................................................................................107 Isolation of RNA........................................................................................................... 108 cDNA preparation and labeling.....................................................................................109 Gene Expression Profiling.............................................................................................109 Quality Control and Statistical Analyses....................................................................... 112 Identifying correlated gene expression patterns-cluster analysis........................... 113 Biomarker identification-correspondence analysis................................................ 113 Test for differential gene expression-robust analyses ............................................ 113 Results...................................................................................................................................114 General Characterization of Gene Profiles....................................................................114 Correlated Gene Expression Patterns............................................................................ 115 Biomarker Identification............................................................................................... 116 Expression Level Comparisons..................................................................................... 117 Discussion.............................................................................................................................118 Correlated Gene Expression.......................................................................................... 118 Biomarkers of Sex......................................................................................................... 119
8 5 RENAL PATHOLOGIES IN MARINE TO ADS ARE ASSOCIATED WITH LAND USE PATTERNS .................................................................................................................. 141 Introduction................................................................................................................... ........141 Methods................................................................................................................................143 Characterizatizing Land Use Type................................................................................ 143 Sample Collection, and Histology................................................................................. 144 Response Variables and Statistics................................................................................. 145 Results...................................................................................................................................146 Discussion.............................................................................................................................147 APPENDIX: CHEMICAL USAGE LIST................................................................................... 156 LIST OF REFERENCES.............................................................................................................158 BIOGRAPHICAL SKETCH.......................................................................................................183
9 LIST OF TABLES Table page 4-1 Ovary and intersexed tissue (Ovary/I ntersex), testis and intersexed tissue (Testis/Intersex), and ovary and testis (Ovary/Testis) fold change differences ...............126 4-2 Genes identified in correspondence analys is as being associated (or disassociated) with a specific reproductive organ type.. .........................................................................130 4-3 Gene expression level comparisons. Estimates of levels of expression, standard errors and pvalues. ...........................................................................................................131 A-1 A list of chemicals used at each agricultu ral s ite were provided by the farm managers at each facility. ...............................................................................................................156
10 LIST OF FIGURES Figure page 1-1 The endocrine system regulates physiologi cal processes by integrating signals from the external environment via the sensory system and internal signals that are conveyed by endogenous (or internal) hormones..............................................................58 1-2 Lipophilic hormones easily diffuse from a capillary into a targ et cell and bind to specific intracellular pro tein r eceptors. Most EDCs identified to date influence gene expression by interacting with thes e receptors found within the cell................................ 59 1-3 Hormones can cross the nuclear memb rane, bind specific receptors, and induce allos teric transformations that enable the hormone-receptor complex to bind to response elements (REs) and modulate gene expression................................................... 60 1-4 Steroid hormone synthesis involves a com plex network of chemical reactions that convert a cholesterol based substrate in to a product (e.g. ho rmones) through the activity of several subs trate-specific enzymes................................................................... 61 2-1 Male (M) and female (F) Bufo marinus have several sexually dim orphic characters....... 77 2-2 Southern Florida (USA) map showing each collection s ite. The area around Belle Glade (BG) and Canal Point (CP) is a patc hwork of agricultural fields. The area around Wellington (WT) and Lake Worth (LW) are suburban developments.................. 78 2-3 Percentage of individuals that were classified as inte rsexed, abnorm al, and male at each site. Total number of testicular, Bidders organ, and female tissue abnormalities are plotted for each collection site.............................................................. 79 2-4 Percentages of male (N=59), abnormal (N=12), and intersexed (N=25) toads with zero through three nuptia l pads per forelim b. No individuals scored as female had nuptial pads................................................................................................................... .....80 2-5 Mean mottling score across an individuals body for each sex. Mottling was evaluated by drawing a transect line from th e left eye to the vent directly onto the image of each individual and counting the number of color changes (see Figure 1)........ 81 2-6 Mean mottling score for females and males across sites. Agricultural intensity increases along the X-axis from left to right for each graph.............................................. 82 2-7 Mean forelimb width for each sex or abnormality group measured across the radioulna distal to the hu m erus (elbow), and perpendicular to the arm axis..................... 83 3-1 Summary data of primary and secondary m orphological traits of Bufo marinus collected from habitats that vary in the intensity of agriculture........................................ 97
11 3-2 Cross section of a testis and mean nu mber of sprem atocysts of each cell type per lobule across sites of varyi ng agricultural intensities........................................................ 98 3-5 Percentage of toads with a median sperm rank of 0-3 by sex group including nonagricultural males (NAgM), agricultural males (AgM), agricultural abnormal toads (A), and agricultural intersexed toads (I)......................................................................... 101 3-6 Testosterone concentration by site. The m ost agricultural site is on the right. Bars are 95% confidence intervals........................................................................................... 102 3-7 Hormone concentrations of non-agri cultural m ales (NAgM) agricultural males (AgM), agricultural abnormal toads (A), and agricultural intersexed toads (I)............... 103 3-8 Estradiol 17 to testosterone ratio concentrations am ong sex groups including nonagricultural males (NAgM), agricultural males (AgM), agricultural abnormal toads (A), and agricultural intersexed toads (I)......................................................................... 104 4-1 Representative region of three arrays depicting dif fer ences in gene expression among sex groups........................................................................................................................136 4-2 Percentage of genes that had a fold cha nge that was higher, equal, or lower relative to another sex group.. .......................................................................................................137 4-3 Gene expression correlations. a. ova ry and intersexed ti ssue gene expression, b. testis and intersexed tissue gene expression, and c. ovary and testis tissues. Solid line represents the one to one line. ................................................................................... 138 4-4 Cluster analysis of fold change expre ssion levels of each gene that met the data quality requirements and showed a fold change higher than 1.5 increases or 0.67 decreases in at least one category. ................................................................................. 139 4-5 Association among gene expression and gonadal type. This figure was generated by Dr. Mary Lesperance University of Victoria, British Columbia Canada........................ 140 5-1 Selected renal pathologies................................................................................................ 153 5-2 Mode of granuloma number across sites. The m ost suburban site is on the left and agriculture increases along the x-axis..............................................................................154 5-3 The percentage of male toad s with various renal pathologies ......................................... 155
12 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy REPRODUCTIVE AND RENAL PATHOLOGIES IN THE GIANT TOAD ( Bufo marinus ) ARE ASSOCIATED WITH HUMAN LAND USE PRACTICES By Krista Ann-Marie McCoy December 2007 Chair: Colette St. Mary Cochair: Louis J Guillette, Jr Major: Interdisciplinary Ecology Many chemicals used by humans (such as pest icides) are known to disrupt the endocrine system of wildlife and lead to reproductive abnormalities. Despite some notable laboratory studies, the potential effects of these chemicals under field conditions have not been well examined. Therefore I surveyed fi ve populations of the giant toad ( Bufo marinus ) that varied in proximity to either agricultural or suburban areas, to evaluate wh ether this gradient of land use left a signature on their reproductive and rena l biology. I demonstrat e that male toads ( Bufo marinus ) from agricultural areas in south Florida are feminized (e.g., they have female color patterns), demasculinized (e.g., they have smaller forelimbs, and fewer nuptial padsmasculine traits), and are more likely to have intersex gona dal tissues (i.e. ovaries and testes in the same animal) than toads from less agricultural (more subur ban) sites. Toads from agricultural habitats also had reduced spermatogenesis and testosterone levels relative to male toads from suburban habitats. To develop hypotheses about the mechanisms th at result in toads from agricultural areas having such dramatic reproductive abnormalities, I compared the expression levels of several genes in ovaries, testes, and intersex gonads using a cDNA array. Many ovarian genes were
13 expressed similarly between ovarian and intersex gonadal tissue, which is surprising because the intersex toads are most likely genetically ma le, and because on average intersex gonads were only 30% ovary by weight. I also developed poten tial biomarkers for the intersex condition and identified gene networks that could be impor tant for normal (and abnormal) gonadal function. Given the striking difference in gonadal abnorma lities across sites, I also determined if toads from these sites had differences in toxi n-induced renal disease. Renal pathologies associated with pollutant exposur e occurred at all of th e sites but tended to be increased at the most agricultural sites. Ther efore, although both agri cultural and suburban s ites are polluted; the types of pollutants at agricultural sites are associated with se vere reproductive abnormalities in addition to renal pathologies while those at subu rban sites are not. My work has increased our understanding of how human-induced environm ental alterations can harm wildlife.
14 CHAPTER 1 ENDOCRINE DISRUPTING CHEMICALS IN AMPHIBIANS Introduction Endocrine disrupting chem icals (EDCs) are ubi quitous pollutants and their presence in the environment has created an important and difficu lt issue in wildlife conservation (Colborn and Smolen, 1996). Any physiological, behavioural or developm ental process that is controlled or influenced by the endocrine system is suscepti ble to modification by EDCs. In fact, EDCs are known to have dramatic effects on vertebrate development, physiology, and behaviour (Crews and McLachlan, 2006; Fox, 2001; Guillette and Iguchi, 2003, 2004; Hayes et al., 2002 ; Milnes et al. 2006). Therefore, EDCs can change the re productive success of individuals, alter patterns of sexual selection, and reduce th e long-term fitness and persiste nce of affected populations. Indeed, environmental contaminants, such as ED Cs, could have further reaching and much more important effects on amphibians and other wildlife than is presently appreciated. Amphibians are thought to be particularly suscep tible to the effects of EDCs because they have permeable skin, often reproduce in areas that receive large amounts of run-off and can be in intimate contact with environmental contaminants in all stages of their lif e history. For example, recent studies have documented reproductive abnor malities in amphibians from agricultural areas where EDCs are used (Hayes et al. 2002; Hayes et al. 2003; Ouellet et al. 1997; Russell et al. 1995; see Chapter 5 this volume). In addition, b ecause EDCs are distributed globally they have been hypothesized to contribute to the global decline of amphibian populations (Carey and Bryant, 1995; Hayes et al. 2003). It is important to note, howe ver, that relatively little is known concerning the role EDCs potentially play in amphibian population declines. This is an important question to be addressed via research at all levels of or ganization, including the molecular and cellular mechanistic levels, and individual, population, a nd community levels.
15 Although, to date, relativel y little is known about the effects of EDCs on amphibians, there is substantial literature from other species, and a growing understandi ng of the endocrinology of amphibians from genes to individual physiolo gical performance. Therefore, since many generalities have been observed, literature on these topics from other vertebrates and various ecological systems are used here to better unders tand the data currently available for amphibians specifically, and examples of endocrine di sruption in amphibians are emphasized. Brief Introduction to the Endocrine System The endocrine system consists of ductless endoc rine glands that secrete hormones directly into the bloodstream or into surrounding tis sues. Endogenous hormones transmit information and regulate biological processes (Fig. 1-1). Lipophilic hormones, such as steroid hormones, are well suited for signaling because they diffuse read ily from a source cell (or capillary) to a target cell, where they bind to specific protein receptors (Fig. 1-2). The binding of a steroid hormone to its receptor induces an allo steric transformation that enable s the hormone-receptor complex to bind to high affinity sites in the DNA and modulate gene transcription (Mangelsdorf et al. 1995; Yamamoto, 1985) (Figure 3). Steroid hormone r eceptors belong to a super-family of nuclear receptors that function as transcription factors, capable of altering the gene expression of thousands of genes (Terasaka et al. 2004; Watanabe et al., 2003). Therefore, hormones are secreted from an endocrine gland, and are transported through the body to regulate biological processes in distant tissues by modulating ge ne expression. Lipophilic hormones, including the steroids, retinoids, and thyroid hormones, are im portant regulators of development, cellular differentiation and organ physiology in all vertebrates studied to date (Mangelsdorf et al. 1995). Indeed, most EDCs identified to date are known to influence gene expression by interacting with the steroid, thyroid, and re tinoid super-family of receptors (Crews and McLachlan, 2006) (Fig. 1-2). In addition to inter actions with hormone receptors, EDCs alter the
16 function of hormone degradation, carrier proteins and key steroidogenic enzymes, all of which are crucial for normal gene expression and hormonal homeostasis (Guillette and Gunderson, 2001; Sanderson, 2006) (Fig. 1-1). A ltered gene expression can l ead to a variety of negative effects-including abnormalities in the structure (via development and growth) and function of organs associated with hormonal signaling and regulation, or induction of cancers. EDCs are also known to influence gene expression epigen etically, which can i nduce transgenerational effects (Crews and McLachlan, 2006; Mukerjee, 2006). Definition of Endocrine Disrupting Chemicals Endocrine disrupting chem icals are defined differently by va rious entities ; the nuances among definitions can have important implications for legislation and can influence the way in which research on EDCs is conducted. For ex ample, the World Health Organizations (WHO) International Program on Chemical Safety (IPCS) recently organized and conducted a Global Assessment of the State-of-the -Science of Endocrine Disrupto rs (IPCS, 2002). The current definition of an endocrine disruptor according to that program is an exogenous substance or mixture that alters function(s) of the endocrine system and cons equently causes adverse health effects in an intact organism, or its progeny, or (sub)populations (IPCS, 2002; Sanderson, 2006). The European Commissions de finition is similar to this but also includes a definition of a potential endocrine disruptor which acknowledges that consideration of a chemicals safety can be appropriate if that substance could possibly disrupt the endocrine system (EuropeanCommission, 1997). However, the US EPA ( http://www.epa.gov/scipoly/oscp endo/edsparchiv e/2-3attac.htm ) defines an endocrine disruptor as an exogenous agent that interfer es with the synthesis, secreti on, transport, binding, action, or elim ination of natural hormones in the body wh ich are responsible for the maintenance of homeostasis, reproduction, development and/or beha vior. The EPA explicitly states that the
17 agency does not consider endocrine disruption to be an a dverse effect per se, but rather to be a mode or mechanism of action potentially leadi ng to other outcomes, for example carcinogenic, reproductive, or developmental effects. Thes e differences are important because the WHOs or European Commissions defi nition suggests that altered endocrine function can induce adverse health effects through any number of mechanisms, while the EPAs industry-friendly definition requires demonstration of a particular mechanism of t oxicity that leads to adverse health effects beyond disruption of the endocrine system itself. Although it is clearly extremely informative to understand the exact mechanism of toxicity of a particular environmental pollutant and to identify relevant consequences of exposure, these details can, and do, take many years, and in some cases decades, to work out. One consequence is that many endocrine disrupting chemicals can remain in use for many years. In addition, th is definition creates a situation where significant time, and research funding, is spent working out mechanisms of action, and less time is spent in the field understanding effects at the population level. Endogenous hormones control a wide variety of living processes a nd are physiologically active at extremely low concentrations (in the pg /ml or ng/ml range). Consequently, pollutants that behave like endogenous hormones can induce effects at concentrations well below traditionally perceived toxic levels. Tradit ional chemical evaluation and risk-assessment procedures were not designed to manage such an unexpected mode of action. As a result, current regulations on the use of EDCs were made ma ndatory only after significant environmental damage had already occurred (Matthiessen and Johnson, 2007). For the purposes of this chapter, EDCs ar e defined by paraphrasing Crews and McLachlan, (2006) and Tabb and Bloomberg (2006) The resulting definition implies that there are multiple
18 mechanisms through which endocrine disruption can occur and a divers ity of effects and consequences to exposure: Endocrine disrupting chemicals are a class of environmental pollutants that mimic or block transcriptional activation elicited by endogenous hormones such that they resemble (or alter) natural biologic al signals (hormones) and, thus can be misinterpreted by the organism leading to abnormal gene expression and phenotype. The Nature and Sources of Endocrine Disrupting Chemicals Many endocrine disrupting chem i cals are persistent either be cause they have a long halflife or because they are added to environmental sy stems at a rate faster than they break down or are metabolized. EDCs are also ubiquitous and many are distributed glob ally---even areas once believed to be pristine are affected by thei r run off or fallout. Once deposited, EDCs can biomagnify in food chains (Bjerregaard et al. 2001; Goksoyr, 2006; Guillette et al. 2006; Lie et al. 2003; Nations and Hallberg, 1992; Skaare et al. 2000; Thurman and Cromwell, 2000). For example, organochlorine pesticides (e.g. DDT) and industrial chemicals, such as polychlorinated biphenyls (PCBs), are found in high concentrations in arctic ecos ystems. Skaare et al. (2000) showed that polar bears ( Ursus maritimus ) and glaucous gulls ( Larus hyperboreus) from Svalbard, in the Arctic Ocean m idway between Norway and the North Pole contain high levels of PCBs. Polar bear PCB levels were greater th an eight tim es higher than those found in ringed seals ( Phoca hispida ), an important polar bear prey (Severinsen et al., 2000; Skaare et al. 2000). Similarly, glaucous gull PCBs are up to three orde rs of magnitude higher than in crustaceans and fish which make up their major food items (Borga et al., 2003; Skaare et al. 2000). Interestingly, there are se x differences in exposure to lipophilic environmental contaminants such as PCBs that are related to differences in excretory pathways. For example, egg yolk and milk are important, female-specifi c, excretory pathways for lipophilic compounds (Skaare et al. 2000). The occurrence of high levels of PCBs and DDE (a metabolite of DDT) in
19 the eggs of the green frog ( Rana clamitans) and spring peeper (Pseudacris crucifer ) in southern Ontario, Canada, suggests that sex-specific e xposure and excretory pathways occur in amphibians. Therefore, larvae are exposed to en vironmental as well as to maternally derived (via the yolk) concentrations of contaminants at critical developmental stages (Russell et al. 1997; Russell et al. 1995). These exposures could certainly be negati vely affecting amphibian populations in many places, but the ecological impli cations have not been well studied (Russell et al. 1997; Russell et al. 1995). Some EDCs enter the environment as natural sources such as hormones originating from humans, live stock, or plants (phytoestrogens). For example, some estrogenic hormones, such as 17beta-estradiol and estrone, are naturally synthesized and excreted by women. Several studies have shown that these chemicals, as well as 17 al pha ethinylestradiol, the synthetic estrogen in contraceptives, can be found in aquatic environments associated with domestic waste water treatment plants at concentrations up to 5, 12, 47 and 7.5 ng/L respectively (e.g. Belfroid et al., 1999). Although many waste water treatment sites contain concentrations below detection limits, 17 alpha -ethinylestradiol is known to i nduce the development of intersexed gonads in some gonochoristic fish at levels well be low those detected by Belfroid (Jobling et al. 2002 ; Lnge et al., 2001; Norris and Carr, 2006; Palace et al. 2002; Pawlowski et al., 2004; Woodling et al. 2006). Furthermore, in the Ruhr district of Germany, 17 alpha eth inylestradiol (used in birth control) has been detected in surface waters in concentrations betw een 14 ng/L, which can induce intersexed gonads. The naturally synthesized estrogens, 17 -estradiol and estrone, however, were below the quantification limit of 1 ng/L (Lintelmann et al., 2003). Another source of naturally derived EDC po llution involves the hormones released from the waste of animals on feedlots or from wa ste used as fertilizer (reviewed by (Johnson et al.,
20 2006). These hormones are either na turally synthesized or prescribe d as growth stimulants, or for other purposes. Intensive livestock rearing or agriculture leads to runoff of androgenic hormones at levels high enough to induce endocrine disruption in wild fathead minnows ( Pimephales promelas) (Ankley et al. 2003; Matthiessen et al. 2006; Orlando et al., 2004; Soto et al. 2004). Steroid hormones, such as 17 beta -estradiol and te stosterone, have even been found in spring water from mantled karst aq uifers in agricultur al areas (Peterson et al. 2000). Therefore, hormone contamination from agricultur al practices is not limited to surface waters and is likely distributed acro ss long distances via ground water. Phytoestrogens are endogenous compounds f ound in plants and are structurally and functionally similar to estradiol and other endogenous estrogens. They are released into the environment in a variety of ways, including sewage treatment plant effluent (Pawlowski et al. 2003; Spengler et al. 2001), runoff from agricultural ar eas treated with manure (Burnison et al. 2003) and effluent from wood pulp mills (Clotfelter et al. 2004; Mahmood-Khan and Hall, 2003). Although it is easy to unde rstand how pollution containing na tural or synthetic hormones can alter the endocrine systems of wildlife, there is an extraordinarily divers e array of man-made (synthetic) chemicals that also induce endocrine disruption. Thes e chemicals include pesticides, flame retardants, plastics or plasticizers, hea vy metals, industrial compounds (e.g. coolants or insulators), pharmaceuticals (e.g. antidepressants) and cosmetics (Goksoyr, 2006; Propper, 2005). Each of these broad categories contains many different chemicals that are known or suspected to induce endocrine disruption. Importantly, indivi dual chemicals within and among these general categories (e.g. pesticides) can have ve ry different chemical structures or be from different chemical classes, which complicates th e ability to identify or predict chemicals that cause endocrine disruption.
21 Despite the large diversity of chemicals that can disrupt the endocrine system, they often have similar effects, albeit via different mechanisms of toxicity. For example, most known EDCs are estrogenic (induce estrogen-like effects) which can lead to feminization of males or development of intersexed gonads (Crews and Mc Lachlan, 2006). However, a growing literature also has shown that aquatic pollutants potentially alter thyroid, androgen, progesterone and retinoic acid signaling (Guillette, 2006). In fact, many PCBs and polybrominated biphenyl ethers (PBDEs), used for decades as flame re tardants, alter thyroid hormone function (Zoeller et al. 2002). A growing list of fungi cides and other pesticides have anti-androgenic activity (Gray et al. 2006). Interestingly, no synthetic environmental compounds have been reported to have androgenic effects (McLachlan, 2001) except phar amaceutical agents (i.e. synthetic androgens) prescribed by, and released int o, the environment as part of ca ttle feedlot operations (Durhan et al. 2007). Plant sterols found in paper-mill efflue nt, however, can be metabolized by native aquatic bacteria to produce androgens, and have been shown to masculinize female fish (e.g. Gambusia affinis holbrooki) (Denton et al., 1985; Parks et al. 2001; Toft et al., 2004). Thus, it is important to recognize that no single endocri ne mechanism is at ris k, but rather, EDCs can alter various endocrine signaling pathways via multiple mechanisms. Modes of ActionHow and Why? The m ost commonly studied mode of action of EDCs is their action as a ligand (e.g. a hormone), thus inducing inappropriate activation or antagonism of nuclear hormone receptors (Fig. 1-2). However, hormone receptors can be influenced through several other mechanisms. EDCs can modulate receptor degr adation, or alter receptor activity, such as by changing the activity or availability of co-factors essentia l for receptor function (Tabb and Blumberg, 2006). In addition, EDCs can modulate the degradation and clearance of endogenous hormones or alter synthesis of hormones, such that inappropriate amounts or types of steroid hormones are formed
22 (Guillette and Gunderson, 2001; Sanderson, 2006; Tabb and Blumberg, 2006). Finally, and possibly most importantly, some EDCs ha ve been shown to alter genome-wide DNA methylation patterns. Methyla tion controls gene expression and if it occurs during specific developmental windows, gene expression patterns can be permanently altered, and these changes are inherited by offspring (Anway et al., 2005; Edwards and Myers, 2007). Nuclear Receptors Nuclear receptors are pro tein receptors that bind to specific hormones (also called ligands) (Figs. 2, 3). This binding changes the conformati on of the receptor such that it can bind with another receptor-ligand complex or with a coactivator-ligand complex. After this dimerization, the complex binds to specific DNA sequences in the regulatory region of target genes (called responsive elements) (Fig. 1-3), and tr iggers gene expression (Janosek et al. 2006). There are 47 known nuclear receptors and, as a group, they c onstitute the nuclear receptor superfamily. These receptors are involved in modulating a wide range of physiol ogical functions across eukaryotes including cellular growth and proliferation, deve lopment and differentiation, and maintenance of homeostasis (Blumberg and Evans, 1998; Janosek et al. 2006). EDCs can bind to and activate a receptor and induce expr ession of responsive genes (McLachlan, 2001; Sanderson, 2006). Alternatively, other EDCs bind to the receptor and block the endogenous hormone from its receptor, which results in inap propriately low levels of ligand-responsive gene expression. Such influences on sex hormone receptors have been extensively studied over the past few decades (Tabb and Blumberg, 2006); for a review see (Janosek et al. 2006). Some chemicals, such as ethanol, however, activate transcription of estrogen-controlled genes by influencing up-stream genes, not by bindi ng the estrogen receptor (ER) (Janosek et al. 2006). Many members of the nuclear receptor s uperfamily are degraded via the ubiquitinproteasome pathway. Ubiquitin labels old or da maged proteins for degradation, which is then
23 carried out by proteosomes (proteins that degrade other proteins). Receptor turnover is one mechanism whereby cells control gene expres sion and prevent over stimulation by endogenous hormones (Tabb and Blumberg, 2006). EDCs can modulate receptor degradation because they do not induce the same proteosome response th at endogenous hormones elicit. For example, (Masuyama and Hiramatsu, 2004) showed that in the presence of estradiol both estrogen receptor alpha (ERalpha) and beta (ERbeta) interacted wi th active sites within the proteosome and were degraded. However, in the presence of Bis phenol A (a constituent of many plastics) which activates ER transcription, ER beta degradation was blocked (Masuyama and Hiramatsu, 2004). Reduced degradation of ERbeta leads to a build up of these receptors in the cell, and thus is expected to lead to increased expression of ERbeta-responsive gene s (Tabb and Blumberg, 2006). Other EDCs have been shown to alter ligand receptor activity (Jansen et al. 2004) or the activity of co-activators essential for r eceptor function (Tabb and Blumberg, 2006). Hormone Catabolism and Carrier Protein Binding Altering hormone biotransfor m ation (synthesis) and catabo lism (break down) is yet another way EDCs can modulate hormone balance (Guillette and Gunderson, 2001; Tabb and Blumberg, 2006). For example, two nuclear receptors are important regula tors of the catabolism of xenobiotics (a biologically active pollutant or drug), chemicals, and steroid hormones. These receptors are: (1) human steroid and xenobiotic receptor/rodent pregnane X receptor (SXR/PXR) and (2) constitutive androsta ne receptor (CAR) (Forman et al., 1998; Kretschmer and Baldwin, 2005; Tabb and Blumberg, 2006; Xie et al., 2000). Both of these receptors are highly expressed in the liver and intestine, and control induction of several ke y enzymes that orchestrate the biotransformation, catabolism, and transport of endogenous hormones and xenobiotics. Many EDCs have been found to activate these receptors and increase expression of their target genes (Kretschmer and Baldwin, 2005; Tabb and Bl umberg, 2006). EDC-i nduced activation of
24 enzymes that degrade hormones and xenobiotics has a two-fold effect. First, endogenous hormones have a shorter half-life and thus are biologically active for less time. Second, in degrading xenobiotics, these enzymes increase le vels of EDC metabolites, some of which are more toxic than the parent compounds. Plasma-binding proteins sequest er and control the bioavailability of hormones such that they cannot pass through the plasma membra ne and bind to nuclear receptors (Nagel et al. 1997). Many xenobiotic chemicals do not bind to plasma-binding proteins and, thus, are readily available to the cell and can induce endocrine disruptive actions (Arnold et al. 1996; Crain et al. 1998). In contrast, Danzo (1997) showed that several xenobiotics indu ced disassociation of endogenous ligands from their bindi ng proteins and competed for binding proteins just as strongly as did natural ligands. Therefore, so me xenobiotics can competitively bind to steroidbinding proteins, thereby increasing the bioavaila blity of endogenous hormones because they are free to enter the cell (Danzo, 1997). In addition, several studies have sh own that receptors for these binding proteins can be present on the pl asma membranes of some cells. The hormonebinding protein complex binds to th ese receptors and initiates a si gnal transduction cascade (for a review see Danzo, 1998). Xenobiotics that bind in place of the endogenous hormones can inhibit or stimulate signal tr ansduction (Danzo, 1998). Steroid Biosynthesis Steroid hormone synthesis involv es a com plex network of chemical reactions that convert a cholesterol-based substrate into a product (Fig. 1-4) through th e activity of several substratespecific cytochrome P450 enzymes, steroi d dehydrogenases, and reductases (Guillette et al., 2007; Sanderson, 2006). The cytochrome P450 enzyme s are especially important because each one is responsible for specific chemical conversi ons that are essential for steroid biosynthesis (Miller, 1988). Importantly, the activity of many of these enzy mes can be modulated by various
25 endocrine-disrupting chemicals, resulting in impaired development, growth, sexual differentiation, reproduction, and the development of particular cancers (for a review see Sanderson, 2006). For example, aromatase (CYP19) is responsible for controlling the ratelimiting step in the conversion of androgens to estrogens (e.g. test osterone to 17betaestradiol), and its modulation by EDCs has been shown to in terfere with the homeostasis and function of sex steroid hormones, resulting in feminizati on and demasculinazation (Crain and Guillette, 1997). For example, several triazine herbicid es (atrazine, simazine, and propazine) and a number of their metabolites (atr azine desethyl and atrazine de sisopropyl) induced aromatase activity and gene expression up to 2.5-fold in human H295R adrenocortical carcinoma cells (Fan et al. 2007; Sanderson et al. 2002; Sanderson et al. 2000). Similar results were shown in alligator neonates (Crain and Guillette, 1997) This mechanism could explain why male amphibians exposed to artrazine in laborator y and field studies ar e feminized and have intersexed gonads (Hayes et al., 2002; Hayes et al. 2003). Alternatively, Ankley et al. (2005) demonstrated the fungicides prochloraz and fenarimol inhibit aromatase activity as well as bind to the androgen nuclear receptor in the fathead minnow ( Pimephales promelas). Both fungicides caused significant anti-androgeni c alterations in endocrine f unction of the fish, thereby decreasing reproductive success (Ankley et al. 2005). Epigenetics Epigenetics ref ers to changes in gene function that are heritable (mitotically or meiotically) but do not involve changes in nucleotide sequence. For example, all the cells in the body of a multi-celled organism originate from the same single-celled zygote and thus share the same DNA sequences. Processes that lead to cellular differentiation and maintenance of specific cell functions involve epigenetic ch anges in genes rather than ch ange in nucleotide sequence (e.g. mutation). In other words, gene function and expression among different cells is altered in a
26 mitotically heritable way, in this case wit hout changing the DNA sequence. One mechanism through which this occurs is by methylation of DNA. This occurs when methyl groups are added to cytosines in the DNA. Methylation is i nherited by daughter cells through DNA synthesis and cell division, and typically lead s to suppression of gene expre ssion, whereas de-methylation is associated with increased gene expression (Ballestar and Esteller, 2002; Crews and McLachlan, 2006; Edwards and Myers, 2007; Holliday and Pugh, 1975). For example, several of the inactive genes on female Barr bodies are methylated an d the active and inactive states of these chromosomes are inherited from mother cell to daughter cells. Therefore, the pattern of DNA methylation among chromosomes (and alleles) of somatic cells is stably maintained, but germ cells and pre-implantation embryos can expe rience genome-wide re-programming of the methylation pattern that generates cells with new (meiotically ) heritable developmental potential (Reik et al. 2001; Reik and Walter, 2001a). Epigenetic processes, via differential methylation, are also involved in gene imprinting which occurs when genes function differen tly depending on whether they are located on maternal or paternal chromosomes (Reik and Walter, 2001b). The DNA of eggs and sperm are methylated differently and inheritance of these differences leads to differential gene expression between the sexes (Zuccotti and Monk, 1995). In addition, during gonadal sex determination DNA in primordial germ cells is demethylated and remethylated in a sex-specific manner (Reik and Walter, 2001a). The methylation patterns that occur in the germ line can be permanent and inherited from parent to offspring. The mechanisms controlling methylati on during development are under intense investigation, and estrogens are thought to be involved in some cases (McLachlan, 2001). For example, estrogens, through interactions with thei r receptors, are known to in crease the levels of
27 c-fos and c-jun in target cells (Kamiya et al., 1996; McLachlan, 2001). C-fos and c-jun are transcription factors that are up-regulated in response to ma ny physiological signals, and in turn up-regulate transcription of ot her genes involved in a diverse array of functions. Overexpression of c-fos leads to an-upregulation of (for example) cytosine methyltransferase, the enzyme involved in increasing DNA methylatio n (McLachlan, 2001). Estrogens influence on methylation, however, is complex and can be ti ssue and gene specific. For example, 17beta estradiol exposure in male white Leghorn roosters is associated w ith removal of methyl groups at the estrogen response element (estradiol-receptor bi nding site) of a vitellogenin (yolk protein) gene, and this demethylation is associated with inappropriate transcript ion and translation of vitellogenin in males (Jost et al. 1990; Saluz et al. 1986). A similar mechanism of vitellogenin activation is known to occur in frogs (e.g. Xenopus laevis ) (Andres et al. 1984; Crews and McLachlan, 2006). Hormone exposure early in life can alter methyl ation, and thus epigene tically alter the set point for later response to the same or different hormones (Saluz et al. 1986). Given the potential role of hormones, such as estrogens, in controlling methylation, and the complexity of the mechanisms involved, one should expect EDCs that modulate plasma concentrations of estrogens, or influence their receptor expression, to induce changes in methylation patterns in ways that could be heritable. Indeed, it has r ecently been demonstrated that several EDCs can influence epigenetic programming through DNA methylation, and that when these changes occur at particular stages during de velopment they are permanent a nd inherited by offspring (Crews and McLachlan, 2006; Tabb and Blumberg, 2006). For example, Anway et al. (2005) showed that rat pups ( Rattus rattus ) exposed to the fungicide vinclo zilin (anti-androgenic) or the insecticide methoxyclor (estroge nic) during sexual differentia tion experienced inappropriate
28 genome-wide methylation patterns in the male germ line. The altered methylation patterns were associated with altered sperm development and lower reprod uctive success (Anway et al., 2005). This methylation pattern and reduction in fertility was transmitted through the male germ line to nearly all males in each generati on across the four generations that were examined. The ability for endocrine disrupting chemicals to reprogram the germ line and induce trans-generational pathologies clearly has importa nt implications for ecology, e volution, and conservation of exposed wildlife, including amphibians. Phenotypic Responses Exam ining the effects of EDCs is difficult and complex for a variety of reasons. There are multiple ways to define endocrine disruption, multiple physiological mechanisms through which disruption occurs, diverse abnormalities and pa thologies that can be induced by EDCs, and multiple levels at which to study these effects. In addition, most toxicological studies have focused on identifying effects of high concentrations of contaminants (e. g. mg/L), often focusing on measuring end points such as mortality. Fewe r studies have examined sublethal effects of chemicals, such as endocrine disruption, at ecologically, or physiologically, relevant concentrations (e.g. g/L). In particular, it is difficult to understand the effects of EDCs on amphibian populations because relatively little research on the effect of these chemicals has been conducted on this group of vertebrates, and almost all of it has been based on very few species under laboratory conditions. It is known, however, that several chemicals can m odulate timing of metamorphosis (e.g.Freeman et al., 2005), as well as drastica lly alter organ development. For example, Hayes et al (2003) collected leopard frogs ( Rana pipiens) that were intersexes (male and female gonads in the same gonochoristic individual) from sites ac ross the United States that were contaminated with the widely used herbicide at razine, as well as other agricultu ral chemicals. In addition, the
29 present authors (in preparation) have found a clear relationship between exposure to agricultural habitats where known EDCs, such as atrazine, ar e used, and intersexed primary and secondary sexual characteristics in Bufo marinus They also observed that pollutants such as nitrate, largely from fertilizer use, a ubiquitous global pollutant of fresh water and near marine ecosystems, can alter the rate of development of tadpoles, presumably through an effect on thyroid hormone function, and can alter ovar ian steroidogenesis (Barbeau et al. 2007; Edwards et al., 2006). Thyroid The thyroid system is highly conserved am ong vertebrate species, so thyroid hormone chemistry, its synthesis and delivery system, its receptors, and its regulation within the hypothalamic-pituitary-thyroid (HPT) axis are all comparable across vertebrates (Zoeller and Tan, 2007; Zoeller et al. 2007). The thyroid gland and its hormones regulate growth, development, metabolism, and reproduction, so alterations of the thyroid system caused by EDCs are of concern (Opitz et al. 2006a). The thyroid system involves many mechanisms that can be altered by EDCs (as discussed above), su ch as alterations in receptor binding, hormone synthesis, and clearance. Thes e processes and mechanisms are affected in similar ways across diverse taxa, including ma mmals and amphibians (Fort et al. 2007; Opitz et al. 2006a). Several groups of chemicals are known, or are t hought to, induce thyroid disruption (Boas et al. 2006). For example, polychlorinated biphenyls (PCBs) (such as coolant/insulating fluids and flame retardants), dioxins (e.g. agent orange, chemi cals found in some insecticides, and industrial byproducts from chlorination) and polybrominated biphenyl ethers (PBDEs) (such as some flame retardants) are known to cause insufficient synt hesis of thyroid hormone (hypothyroidism) in exposed animals. Environmentally relevant doses appear to affect hu man thyroid function as well. Although data on the effects of flame re tardants on human populations is limited, there is great concern as the concentratio ns of these chemicals have dramatically increased in mothers
30 milk over the past decade. These compounds appear to be similar in action to PCBs, which have been studied in detail for decades and which clearly influence thyroid function. Also of major concern are the phthalates, used as plasticizers and chemical stab ilization agents (e.g. in personal care products), as they also affect thyroid function, but appear to be stimulatory (Boas et al., 2006). Structural and functional changes required for successful amphibian metamorphosis are dependent on the presence and appropriate concentrations of the thyroid hormones 3,5,3triiodothyronine (T3), and 3,5,3,5-tetraiodothyronine (T4) (Crump et al. 2002; Fort et al. 2007; Sachs et al. 2000). Thyroid hormones bind to nuclear t hyroid receptors (TR) and this complex can activate, or inhibit, gene tr anscription in a tissue-specific manner. Particular tissues are proliferated after thyroid horm one exposure, whereas others u ndergo apoptosis (programmed cell death) (Crump et al., 2002; Helbing et al. 1996; Helbing et al. 1992; Helbing and Atkinson, 1994; Sachs et al. 2000). Blocking the natural synthe sis of thyroid hormone inhibits metamorphosis whereas exposing tadpoles to exogenous thyroi d hormone induces precocious metamorphosis (Sachs et al. 2000). Therefore, thyroid horm ones are essential for stimulating the complex and diverse genetic programs requi red for amphibians to su ccessfully complete metamorphosis. The essential nature of t hyroid hormones for amphibian metamorphosis, coupled with the fact that thyroid function is well conserved acro ss vertebrates, has lead to the suggestion that tadpoles are an ideal bioassay system to identi fy thyroid function disruption by environmental contaminants (Crump et al., 2002; Degitz et al. 2005; Fort et al., 2007; Opitz et al. 2005; Zhang et al., 2006). A chemical evaluation pr otocol referred to as the Xenopus metamorphosis assay (XEMA) has been develo ped and evaluated via a ring test where six laboratories participated to evaluate inte r-laboratory variation in results (Opitz et al. 2005).
31 Opitz et al (2005) demonstrated that the XEMA test can provide a sensitive, robust, and practical approach for the detec tion and evaluation of chemicals th at affect the thyroid system. The herbicide acetochlor [2-chloroN -(ethoxymethyl)N -(2-ethyl-6-methylphenyl) acetamide] has been shown to accelerate metamo rphosis in several amphibian species (Cheek et al. 1999; Veldhoen and Helbing, 2001). Acetocholor, however, only accelerates metamorphosis in animals that are primed by previous e xposure to thyroid hormone. Therefore, it was suggested that acetochlor did not bi nd to and activate the thyroid receptor directly but enhanced T3 action by different receptor-mediated mechanisms (Cheek et al., 1999). Veldhoen and Helbing (2001) showed that an environmenta lly relevant dose of acetochlor significantly enhanced TRbeta mRNA levels in T3-primed Rana catesbeiana within 24 hours. Therefore, acetochlor does not bind to TRbeta but it does increase TRbeta expression, so that endogenous thyroid hormone can bind (Vel dhoen and Helbing, 2001). Crump et al (2002) investigated acetochlors influence on gene expression in Xenoupus leavis by developing a 420-gene cDNA array from known frog genes. They found that 26 genes are modulated by thyroid hormone and acetochlor. Twenty-four genes were up-regulated whereas only two were inhibited. A detailed understanding of the physiological changes induced by altering this number of thyroid-dependent genes required for normal metamorphosis or the long-term effects on juvenile and adult metabolism and health are unknown. Given that there is a natural spike in thyroid hormones just before metamorphosis, metamorphic animals coul d be naturally primed by thyroid hormone. Under these conditions, acetochlor exposure coul d induce higher than usual TRbeta (and other) mRNA levels and cause the animals to proceed through metamorphosis faster than normal. The ecological implications of this acceleration are as of yet unknow n, but could lead to higher mortality, smaller size at metamorphosis, and altered morphology.
32 Another approach to the investigation of EDCs is to study their effects at the tissue level. Changes in thyroid histology (morphology) are indi cative of alterations in the physiology of the pituitary-thyroid axis. For example, inhibition of thyroid hormone secreti on inhibits the negative feed-back on the pituitary which would normally stop secretion of thyroid stimulating hormone (TSH). This induces increased TSH that result s in an increase of th e size of thyroid gland follicles (hypertrophy of thyrocytes) and this effect can be observed and quantified histologically. For example, perc hlorate, an oxidizer used in so lid-fuel rockets, inhibits thyroid hormone synthesis by inhibiti ng the sodium-iodide symporter that acts as an ion pump transporting iodide into thyroid epithelial cells so that they can synt hesize thyroid hormones (Opitz et al. 2006b; Tietge et al., 2005). The effects of perchlorate on cricket frogs ( Acris crepitans) from contaminated streams in central Texas were evaluate d through histological analyses. Individuals living in streams with th e greatest mean water perc hlorate concentrations (~ 25 microg/L) showed significantly greater fol licle cell hypertrophy. In addition, a significant positive correlation was found between thyroid follicle cell height and mean water perchlorate concentrations across all sites (Theodorakis et al. 2006). Similar results were found for Campostoma anomalum a species of fish, from the sa me habitats, which suggests that perchlorate affects thyroid func tion across many taxa through a similar mechanism. Perchlorate reduced thyroid hormone, and the thyroid follicle s experienced prolonged stimulation by TSH. Since the mechanism whereby perchlorate e ffects thyroid function is known, it is an appropriate chemical to use to help understand how effects induced by EDCs at the tissue level can scale up to influence processes, such as meta morphosis, at the individual level. For example, African clawed frogs ( Xenopus laevis ) exposed to ecologically relevant concentrations of ammonium perchlorate (59 and 140 microg/L) experienced significant hypertrophy of the
33 thyroid follicular epithelium, as well as inhibition of forelim b emergence, tail resorption, and hindlimb development relative to control animals (Goleman et al. 2002). All these findings suggest that perchlorate alters thyroid function (as expected), but they also demonstrate that many thyroid-hormone-dependant developmenta l programs required for metamorphosis are negatively affected. Interestingl y, only the highest concentrati on of perchlorate reduced wholebody thyroxine concentrations rela tive to the control treatment. This finding highlights the importance of measuring multiple endpoints; if t hyroid hormone concentrations were measured without consideration of other morphological en dpoints, the conclusion would have been drawn that perchlorate does not influence thyroid function in this species. In additi on to alterations in metamorphic development, Goleman et al (2002) found that perchlor ate induced a skewed sex ratio such that significantly fe wer males were recorded at metamorphosis. They concluded that ammonium perchlorate altered thyroid activ ity and gonadal differen tiation in developing X. laevis, which suggests that animals that proceed ed through seemingly normal metamorphosis could suffer lower reproductive succ ess as adults. Importantly, the concentrations of perchlorate used in this study were below c oncentrations reported in surf ace waters contaminated with ammonium perchlorate, so pe rchlorate contamination likely poses a threat to normal development in natural amphi bian populations (Goleman et al., 2002). In a similar study, Tietge et al (2005) exposed X. laevis tadpoles to sodium perchlor ate during metamorphosis, and measured thyroid histology as well as metamorp hic timing. Histological effects on the thyroid were found at levels as low as 16 microg/L a nd metamorphosis was retarded significantly by perchlorate concentrations as low as125 microg/L. Histol ogical effects occurred at concentrations below those required to induce delays in metamorphosis which suggests that thyroid histology is a more sensitive biomarke r for detecting thyroi d-disruption induced by
34 perchlorate (Tietge et al. 2005). Importantly, it is unclear ho w such alterations in the thyroid during development influences la ter life stages (e.g. adult metabolism) independent of whether metamorphosis was delayed. DevelopmentHomeobox Gene Expression A significant am ount of the ea rly work on EDCs focused on developmental endpoints as the developing embryo is sensitive to very sma ll concentrations of hormones required for normal development of many organs and organ systems (Bern, 1992). It is recognize today that the developing embryo has sensitive windows of re sponsiveness when endocrine signals, at appropriate concentrations, indu ce organizational changes that fundamentally modify cellular differentiation and endocrine respon siveness later in life (Guillette et al. 1995). Many diverse genes, including the Homeobox (Hox) genes, have been shown to be responsive to endocrine signaling, making them targets for EDCs. Hox genes are evolutionarily conserved tran scription factors that are essential for orchestrating embryonic development as well as cellular differentiation in adults. The expression of particular collections of Hox gene s in specific places within the embryo creates a collinear expression continuum th at regulates development in a segment-specific pattern and determines the anterior-posterior axis in th e developing embryo (Dafta ry and Taylor, 2006). Hox genes are also essential in the adult where they mediate cellular differentiation and control the developmental plasticity required to spec ify the function of new cells. Although Hox genes have been studied, and their role in cell fate determination is well known, the mechanisms by which their expression is controlled are not well understood. Recent work has demonstrated that nuclear receptors and their corre sponding ligands (hormones) can regulate Hox gene expression in the embryo and adult (reviewed by (Daftary and Taylor, 2006). For example, in mice ( Mus musculus ) estrogens are necessary for the normal Hox gene expression required during
35 embryonic development of the female reproductive tract (Block et al., 2000; Daftary and Taylor, 2006). Exposure to diethylstilbestrol (DES), a synthetic non-steroidal estrogen, results in altered endogenous estrogen signaling and a change in the spatial ex pression of Hoxa9, Hoxa10, and Hoxa11, such that each is expressed more caudally throughout the developing female reproductive tract leadin g to alterations in its morphology (Block et al., 2000; Daftary and Taylor, 2006). Human females exposed to DE S during development e xhibit reproductive tract abnormalities with phenotypes that can be explained by a similar shift in Hox gene expression (Block et al. 2000; Daftary and Taylor, 2006). For exam ple, glandular tissue normally present in the uterus and cervix is observed in the vagina in women exposed in-utero to DES, which is consistent with caudal displacement of HoxA10 and HoxA11 expression (Block et al. 2000; Daftary and Taylor, 2006). This work demonstr ates that endocrine regulation of Hox genes by estrogens during embryogenesis is necessary across a variety of ta xa. Estrogens are also thought to control functional differentiation in the adul t reproductive tract (Daftary and Taylor, 2006). Given the critical role of hormones such as the estrogens in modulating Hox gene expression, any perturbation to these endogenous signals by ED Cs is expected to induce developmental abnormalities (Iguchi et al. 2001). Indeed, methoxychlor (a replacement insecticide for DDT) reduces estrogen binding to the estrogen receptor (ER), and disrupts the ability of the 17beta-estradiol-ER complex to bi nd to the estrogen response element (ERE) for the HoxA10 gene in endometrial cells, resulti ng in a decrease in HoxA10 gene expression (Fei et al. 2005). Therefore, methoxychlor functions as an endocrine disruptor by affecting 17beta estradiol signaling in endometrial cells. In mi ce and rats, these changes inhibit modification of the endometrium (to create the dicidua), whic h blocks implantation of the embryo, and has deleterious effects on fertility. Importantly, in mice, this reduction of HoxA10 gene expression
36 is permanent and continues in adults even if they were only exposed as neonates (Fei et al. 2005). The mechanisms inducing continued repressi on of Hox gene expression in the absence of continued methoxychlor exposure are not known, but likely result from epigenetic modification, possibly via increase d methylation (Fei et al. 2005; McLachlan, 2001). Indeed, Wu et al (2005) clearly demonstrated that differences in methylation of HoxA10, and the subsequent decline in HoxA10 gene expression, is sign ificantly associated with in creased endometiosis in woman, and they argue that enodmetriosis is an epigenetic disease. Unfortunately, to our knowledge no work ha s explicitly evaluated the occurrence and importance of EDC exposure on Hox gene expression in amphibians. It is likely that similar effects are operating in amphibians because Hox genes, and their roles in embryonic development and cellular differentiation in adults are highly conserved. Important aspects of amphibian development and metamorphosis are c ontrolled by Hox genes. For example, the number, position, and type of limbs in amphibi ans depend on the proper expression of specific Hox genes. Retinoic acid regulat es Hox gene expression and alte ring the quantity of retinoids during development results in seri ous developmental abnormalities across various taxa (Gardiner et al. 2003; Koussoulakos, 2004). A pplication of retinoic acid to chicken limb buds alters homoeobox gene expression and induces duplica tion of limbs (Ogura and Evans, 1995). Indeed, one of the most compelling explanati ons for the skeletal dy splasias observed in frogs from Minnesota, and elsewhere, is that the severely malf ormed limbs are due to exposure to environmental contaminants that function as retinoid (e.g. Vitami n A and retinoic acid) mimics (Gardiner et al. 2003). Gardiner et al (1999) analyzed skeletal abnormalities in the Minnesota frogs and identified two major classe s of abnormalities (Gardiner and Hoppe, 1999). First, the initiation of limb development was altered, which induced absent or supernumerary
37 limbs. Second, limb growth and pattern formation were modified such that primary and supernumerary limbs both had particular sk eletal abnormalities, in cluding truncation and dwarfism. Skeletal elements were also folded back on themselves, forming characteristic "bony triangles" (Gardiner et al. 2003; Gardiner and Hoppe, 1999). Retinoid (e.g., retinoic acid) exposure during limb bud development can induce the formation of boney triangles in frogs, chickens, and mice (Gardiner et al. 2003). In addition, retinoic acid can induce supernumerary limbs, incomplete or missing limbs, complete l imb duplications, as well as duplicated pelvic girdles across diverse taxa, includ ing several frog species (Gardiner et al. 2003). Most of these abnormalities have been observed in field-collected malformed specimens. In addition, Gardiner et al (2003) extracted hydrophobic substances from water samples collected from sites where malformed frogs were routinely found, and tested them for their ability to activate the retinoic acid receptor (RAR). Biologically active retinoids were found in water samples from severely affected sites. Therefore, it is possible that a single pollutant that func tions as a retinoid mimic could be responsible for all the malformed phenotypes that have been observed (Gardiner et al. 2003). Interestingly, a recent survey of 5,264 hylid and ranid metamorphs in 42 Vermont wetlands demonstrated that proximity to agricu ltural land was associated with an increase in limb malformations (Taylor et al. 2005). No analysis of the mechanisms underlying these limb abnormalities have yet been carried out (Gardiner et al. 2003). It seems likely, however, that retinoic acid mimics could modulate Hox gene expression and induce a variety of abnormalities, depending on the developmental stage and length of time of exposure. Reproduction Reproductive organs produce a variety of hor mones including steroids such as testosterone, estrogen, and progest erone, as well as peptide horm ones such as inhibin, activin, and Mllerian inhibiting hormone, so they function as endocrine glands and they appear to be a
38 major target for EDCs. Several environmental chemicals alter the development and function of the reproductive system, thereby al tering endocrine system function. These affects have been documented across all classes of vertebra tes including humans (reviewed by (Milnes et al. 2006). Although the mechanisms that drive sex determination and gonadal development appear to vary widely among species, the underlyi ng genetic and endocrine control of gonadal development, growth, and function, as well as the roles of hormones on secondary sexual characteristics, are highly conserved among verteb rates. Most EDCs studied to date exhibit estrogenic or anti-androgenic activity and can influence sex ratio, sexual maturation, gonadal morphology and function, as well as spermatogenesis, fertility, hormone levels (steroidogenesis, metabolism), secondary sexual characterist ics and reproductive behaviour (Milnes et al. 2006). Clearly, all of these characteristics are important for mainta ining healthy and persistent populations, especially for taxa that are known to be declining globally, such as amphibians. The effects of EDCs on the development a nd functioning of the reproductive system of amphibians are not well understood re lative to other wild life su ch as fish, alligators, and mammals. Amphibians, however, could be espe cially sensitive to EDCs because gonadal differentiation in this taxon is highly sensitive to sex hormone s and occurs over an extended period of time, sometimes beginning prior to me tamorphosis and lasting for several months (Hayes, 1998; Qin et al. 2003). In genetic males, estrogen can induce a complete and permanent sex reversal or induce development of intersexed gonads (ovarian tissue and testic ular tissue in the same gonad). Intersexed gonads typically oc cur in two forms: oocytes can be scattered throughout the testes forming an ovo -testis, or the ovary can be di stinct and well delineated from the testicular tissue (Milnes et al. 2006). These different ph enotypes likely occur through
39 different molecular mechanisms and/or timing of exposure, but these details have not yet been fully investigated. Although the data on the effects of EDCs on the reproductive system of amphibians are relatively limited, it has been clearly demonstrat ed that EDCs can induce female-skewed sex ratios and alter gonadal development and con centrations of hormones. Tavera-Mendoza et al (2002a,b) exposed Xenopus laevis tadpoles to ecologica lly relevant levels (21 microg/L) of atrazine for 48 hours during sexual di fferentiation. Histological an alysis of the ovaries showed that atrazine-exposed females experienced a 20% reduction in primary germ cells compared to 2% in controls. Atrazine-exposed males had a 57% reduction in testicular volume, and primary spermatogonial cell nests were reduced by 70%. Furthermore, Sertol i (nurse) cells, which provide nutritive support for the developing ge rm cells, declined by 74% and testicular resorption occurred in 70% of the males, while 10% failed to fully develop testes. These findings are significant because primary germ cells are thought to comprise the total number of germ cells for the entire reproduc tive life of the organism, and such a remarkable reduction in only 48 hours suggests that atrazine, even in pulse d exposures, could severely negatively affect reproductive success of both females and males (Tavera-Mendoza et al. 2002a, b). As mentioned above, African clawed frogs ( Xenopus laevis ) exposed to environmentally relevant doses of ammonium perchlorate duri ng metamorphosis had female-skewed sex ratios (Goleman et al., 2002). Several other chemicals, however have been demonstrated to induce female-biased sex ratios. For example, Kloas et al (1999) exposed X. laevis tadpoles to estradiol (positive control), nonylphenol (use d in surfactants, detergents, and pesticides), bisphenol A (plastic or plasticizer), oc tylphenol (surfactant) and but ylhydroxyanisol (pharmaceutical for treating retroviruses). Each chemical indu ced a significantly higher number of females
40 compared to controls. These results demonstrat e that a diverse array of chemicals can, and do, alter sex ratios making them female-biased. There are two ways in which this type of bias can occur. Males could, in general, be less resist ant to chemicals and suffer higher mortality than females. Alternatively, the genotypic males in these experiments may undergo sex reversal so that they appear female. Many chemicals have been shown to induce feminization of male gonadal tissue. Qin et al (2003) exposed X. laevis to Aroclor 1242 and Aroclor 1254 (aroclors are mixtures of PCBs) and recorded both gross morphol ogical and histological differen ces relative to the vehicle control. Control animals had normal ovaries or testes from the standpoint of gross morphology whereas PCB-exposed animals had abnormally si zed and asymmetrical testes as well as ovotestes. Ovotestes, in that study, were characterized by ovarie s in the cranial and/or caudal portion of the gonad with testes located medially. PCB exposure did not alter the proportion of females across treatments but the portion of males with morphologically normal testes was reduced. Histological examinations revealed that testes that were classi fied as morphologically normal, as well as those classi fied as abnormal upon gross evalua tion, had oocytes interspersed throughout the tissue. In additi on, testes of PCB-exposed animal s were more loosely organized and had fewer seminiferous tubules, spermatogoni a, and spermatozoa than was the case for controls. Qin et al (2003) argued that Aroclor 1242 and Aroclor 1254 feminize gonadal differentiation in X. laevis and that X. laevis could be especially sensitive to endocrine disruption. Thus, this species is an appropriate model for studying endocrine disruption. Hayes et al (2002) showed that very low ecologi cally relevant levels of the common herbicide atrazine can induce X. laevis tadpoles to develop inte rsexed gonads and exhibit decreased larynx size (demasculin ized), and they suggested that these changes could reduce
41 fertilization success and alter breedi ng call characteristics. Indeed, X. laevis, collected from agricultural areas in South Africa where atrazine and other chemicals are used, suffer decreased testosterone levels (Hecker et al. 2004). Importantly, atrazine is a widespread compound and the concentrations of atrazine used in that st udy are realistic levels (e .g. "acceptable" for drinking water according to the EPA) to which many amphi bian species are expose d in the wild, putting them at risk of impaired sexual development. Indeed, much work has been conducted on X. laevis and this species has proven to be an important model since it is easy to obtain, breed, and rear in the laboratory. Several studies, however, have documented feminization in amphibian species other than X. laevis For example, leopard frogs (Rana pipiens) exposed to ecologically relevant concentrations of atrazine (>0.1 ppb) had retarded gonadal development (gona dal dysgenesis) and testicular oogenesis (intersexed). Males that devel oped slowly and were thus expose d to atrazine for longer periods of time, also experienced oocyte growth (oogenesis and vitellogenesis) (Hayes et al. 2003). Hayes et al. (2003) also conducted field surveys of sites that varied in atr azine exposure across the United States and observed gonadal dysgenesis and intersexuality in animals from sites contaminated with atrazine. They concluded that these integrated laborat ory and field studies on leopard frogs, coupled with similar findings in X. laevis, demonstrate the effects of atrazine on amphibians in general and suggest that atrazine and other endocrine-disr upting pesticides could be playing a major role in amphibian declines around the world. Another important group of contaminants th at is commonly found in the environment due to its importance to humans is plastic that co ntains bisphenol A or plasticizers (various phthalates) that give plastic its pliability. Bisphenol A has been shown to induce female skewed sex ratios in X. laevis (Mosconi et al. 2002). Ohtani et al (2000) exposed genetically male
42 tadpoles of Rana rugosa to dilute solutions of the plasticiz er dibutyl phthalate and three 17beta estradiol-positive controls for four days during gonadal sexual differentiation (days 19-23). At day 40, the gonads of the tadpoles were examin ed histologically. Positive control and dibutylphthalate-treated, genetically male tadpoles had complete ovar ies or were intersexed in a dose-dependent manner. Therefore, dibutyl phth alate was able to induce intersexuality, but it was approximately 1000-fold less potent than 17beta -e stradiol (Ohtani et al. 2000). Even considering the reduced potency, Ohtani et al (2000) concluded that dibutyl phthalate is an estrogen-like hormone that alters testicular differentiation, and t hus is a dangerous environmental contaminant. It is known that gonadal development in many amphibian species can be altered by exposure to estrogens or to estrogenic contamin ants (Hayes, 1998). Therefore one would expect larvae living in environments containing sewage effluent that is contaminated with human birthcontrol chemicals to have altere d gonadal development. To test this hypothesis, Park and Kidd (2005) added low concentrations of 17alpha-ethinylestradiol (EE2) to an experimental lake in northwestern Ontario, Canada. A target concen tration of 5 ng/L wa s maintained by adding 17alpha-ethinylestradiol to the treatment lake three times a week duri ng the open-water seasons (May to October) for three years (2001, 2002, and 2003) Egg masses were reared in cages at the EE2 lake and in two refere nce lakes. In the EE2 lake, hatching success was reduced significantly in green frogs ( Rana clamitans) but not in mink frogs ( Rana septentrionalis ), relative to each species in the reference lakes. Ethinylestradiol had no effect on se x ratios of either species and no intersex gonads were observed in tadpoles of e ither species at the re ference sites. Although no green frog tadpoles were intersexed at the EE2 site, approximately 6% and 13% of the caged mink frog tadpoles were intersexed in 2001 and 2002 respectively. Therefore, these two species
43 responded differently to EE2 exposure. Green frogs suffered higher mortality at hatching but survivors did not have intersexed gonads, wh ereas mink frogs did not have high hatchling mortality but did experience altered gonadal development. This hi ghlights the fact that different species respond differently to the same envi ronmental contaminant. Un-caged mink frog tadpoles were also studied at each of the sites, and EE2 had no effect on sex ratios, but by the third year 28.6% of wild EE2-exposed first-year tadpoles had intersexed gonads whereas none were intersexed in the reference lakes. These results indicate that concentrations of 17alphaethinylestradiol, comparable to those found in sewage treatment effluents and some surface waters, can affect hatching success as well as go nadal development in native amphibians (Park and Kidd, 2005). In addition to intersexed or completely femi nized gonads, alterations in circulating steroid concentrations have been reported in amphibi ans exposed to EDCs. For example, Hayes et al (2002) exposed adult African claw ed frogs to 25 ppb atrazine (lev el acceptable in drinking water according to the EPA) and found that they exhibited a 10-fold decrea se in testosterone concentrations relative to animals housed in 0 ppb. They hypothesized that atrazine increased aromatase, the enzyme that converts testoste rone to estrogen, and cau sed the decrease in testosterone. As discussed above, this mechanis m is known to occur in ot her vertebrate species (Fan et al. 2007; Sanderson, 2006; Sanderson et al. 2002; Sanderson et al. 2000). Altered gonadal development and hormone con centrations associated with EDC exposure induce other abnormal physiological responses. On e important example of this is the induction of the normally female-specific protein vitellogeni n (yolk protein) in males. Typically males do not express the vitellogenin gene but when they are exposed to estrogen or estrogenic chemicals this gene is upregulated. Ther efore, plasma vitellogenin in ma les is a biomarker for estrogenic
44 xenobiotics. For example, adult male red-eared slider turtles ( Trachemys scripta ) and African clawed frogs given injections of estradiol, the sy nthetic estrogen diethylstilbestrol (DES), or the insecticide DDT experienced an induction of vi tellogenin. In both species, estradiol and DES treatments induced more vitellogenin than did DD T, which demonstrates that DDT is estrogenic, but as expected from previous studies investigati ng its estrogenicity, it do es not elicit as strong a response as do actual estrogens (Palmer and Palmer, 1995). Since that study, and many others establishing vitellogenin as a re liable biomarker, researchers ha ve begun to use vitellogenin to detect estrogen exposure in amphibians (Palmer et al., 1998); methodology reviewed by (Wheeler et al. 2005). Mosconi et al (2002) reported a dosedependent induction of vitellogenin in male European frogs (Rana esculenta ) and crested newts (Triturus carnifex ) exposed to the surfactant 4-nonylph enol. They also demonstrated that 4-nonylphenol inhibited gonadotropin release from the hypothalamus and prolactin secretion by the pituitary, but increased plasma androgen concentrations. Impor tantly, the metabolic cost associated with production and elimination of vitellogenin and it s impacts on reproductive success of males have not been considered or investigated in detail (Milnes et al. 2006). Other physiological responses that can be al tered when EDCs are misinterpreted as hormonal signals involve seconda ry sexual traits. These include many sexually dimorphic characteristics such as differences in forelimb size, or colouration, or the presence of nuptial pads. For example, female colouration in the reed frog ( Hyperolius argus ) is estrogen dependent, so Noriega and Hayes (2000) used earl y induction of female colour pattern as a biomarker for estrogenic activity. H. argus tadpoles were exposed to the insecticide o,p-DDT and six of its congeners and in vivo colour changes were compared among treatments. Estradiol, o,p-DDT, o,p'-DDE and o,p'-DDD prematurely indu ced adult female colo uration in juvenile
45 animals, whilst p,p'-DDT, p,p'-DDE and p,p'-DDD did not (Noriega and Hayes, 2000). This concept that estrogen-dependent colouration could be used as a biomarker of estrogenic chemicals was utilized by McCoy et al (unpublished data) during fiel d surveys of giant toads ( Bufo marinus ) living in sugarcane agricultural areas in South Florida. It was found that as many as 40% of the males (defined as having testes and nuptial pads) were intersexed (also had ovaries) and many males, including individuals with morphologica lly normal testes, exhibited female colouration, whereas those from reference sites were not intersexed, and exhibited the sexually dimorphic male colour pattern. Behavior The endocrine system and the central nervous system (CNS) are integrated such that the proper functioning of one is dependent upon the pr oper functioning of the other. For example, particular regions of the brain are known to be sexually dimorphi c and these differences occur in response to different hormonal influences dur ing development. Having the proper hormonal milieu at crucial stages in brain organization is essential for future (adult) sex-specific behavioural responses (Moore et al. 2005; Palanza et al. 2002). Altering this hormonal milieu during fetal development can perman ently change adult behaviour. The hypothalamus links the nervous and endocrine systems by releasing stimulating or inhibiting hormones through blood vessels to the anterior pituitary which in turn signals endocrine glands such as the thyroid, adrenal, or gonad via tropic (releasing) hormones (Fig. 11). These endocrine glands can then synthesize and secrete specific hormones that induce various physiological effects and provide feedback that alters pituitary and hypothalamic (CNS) function. Since these systems ar e integrated, hormonal effects on behaviour (CNS) can be direct, or indirect (e.g. altering thyroi d hormones can influence metabolism that in turn influences behaviour) (Zala and Penn, 2004).
46 The endocrine and nervous (neuroendocrine) systems are so highly integrated that chemicals which alter the endocrine system al so influence the central nervous system and therefore behaviour. There are numerous diverse examples of e ndocrine disruption leading to alteration of behaviour; thus, Zala and Penn (2004) suggested that endocrine disrupting chemicals be renamed as "neuroendocrine disrupters". Indeed, behaviour is becoming a more common endpoint in standard toxico logical assays. Many researcher s suggest that behaviour is a useful bioindicator of endocri ne disruption because it is an easily measured, integrated physiological response that is rela ted to alterations of specific ne ural pathways and is dependent on environmental context (Clotfelter et al. 2004; Crews et al., 2007; Panzica et al. 2005). An animals behaviour is indicative (a surrogate or sentinel) of its health and the tools needed to evaluate behaviour are relatively inexpensive and easy to implem ent. Many of the physiological mechanisms that control behaviour, however, are not understood in detail, so it can be difficult to determine the mechanism(s) through which specif ic EDCs modulate behaviour. Thus, there is some debate about the usefulness of behaviour as a measure of endocrine disruption (Clotfelter et al. 2004). If one cannot ascribe an endocrine mechanism, how can one know that disruption of the endocrine system is leading to the altere d behaviour? Understand ing how EDCs alter the mechanisms controlling and modulating behaviour is an extremely important and open field of exploration. Although the mechanisms of endocrine or ne uroendocrine disruption leading to altered behaviour are not always clea r, there are numerous examples of known EDCs that induce alterations in behaviour. Recent review s by Zala and Penn (2005) and Clofelter et al (2005) demonstrated that locomotion, balance, f eeding, antipredator behaviour, communication, aggression, various aspects of learning, and re productive behavior (inc luding courtship, mate
47 choice, mating, nesting, and parental care) can al l be altered by chemicals that are known to disrupt the endocrine system. Many aspects of amphibian behaviour are altered by chemical pollutants including locomotion, activity, exploratory behavior, resp onse to cues emanating from predators, communication, learning, and mating. The followi ng discussion is restricted to examples describing how known EDCs alter behaviour in amphibians. PCB 126 (insulating fluid for transformers and capacitors) has been show n to reduce swimming speed in tadpoles of Rana clamitans and R. pipiens, which suggests that both feeding an d predator escape could be affected in natural systems; these eco logical consequences, however, were not tested (Rosenshield et al. 1999). Bridges (1999) explicitly tested whether res ponses by tadpoles to predators were altered in the presence of the insectic ide carbaryl. Tadpoles of Hyla versicolor were exposed to two ecologically relevant levels of carbaryl (1.25 a nd 2.50 mg/L), the vehicle control, and a negative control (acetone solvent and wa ter respectively) for only 24 hour s. After the exposure, tadpole behaviour was examined in the presence a nd absence of adult red-spotted newts ( Notophthalmus viridescens ) housed in a mesh container. The containe rs allowed visual, tactile, and chemical cues to be transmitted but did not allow pred atory attack. Tadpole activity, measured as percentage of time spent swimming, resting, and feeding and time spent in refugia, was recorded every three minutes for one hour (Bridges, 1999). Control tadpoles spent 20% of their time in refugia in the presence of a pred ator, but did not enter refugia when no predator was present; instead they actively fed. Animals in the lowcarbaryl treatment spen t roughly equal amounts of time in refugia whether a predator was present or not, whereas those in high-carbaryl treatment spent less time in refugia when the predator was present than when it was absent. In addition,
48 tadpoles exposed to high levels of carbaryl we re, on average, less activ e than were those in control treatments but they spent more time foragi ng in the presence of a predator than in its absence. Control tadpoles in Bridges (1999) experiment were able to effectively detect predators, hide in their presence, and actively fo rage in their absence. In other words, they behaved adaptively. The tadpoles exposed to ca rbaryl, however, did not show these adaptive behaviours, especially at the hi gh carbaryl concentration. This study clearly demonstrates that EDCs can alter behaviour in unexpected ways. In a similar study, salamander larvae ( Ambystoma macrodactylum ) exposed to methoxychlor (replacement insecticide for DDT ) exhibited a reduced startle response and traveled shorter distances in response to distur bance; they also experi enced increased predation by dragonflies (Eroschenko et al., 2002; Verrell, 2000). Importantly, this was not the first time that exposure to a known EDC induced behaviour that increased predation. Three decades earlier, Cooke (1970, 1971) showed that common frog tadpoles ( Rana temporaria ) exposed to DDT were uncoordinated and hyperactive and we re preferentially preyed upon by warty newts ( Triturus cristatus ). In addition to the incr eased predation that prey experience after chemical exposure, predators could suffer increased bioaccumulation if th ey preferentially eat exposed individuals (Cooke, 1970; Cooke, 1971). The occurrence of maladaptive behaviour in th e presence of EDCs suggests that learning could be altered by exposure to such chemi cals. To examine this hypothesis, Steele et al (1999) exposed tadpoles of Rana clamitans and R. catesbeiana to 0 or 750 microg Pb/L for 5-6 days; then, animals were conditioned to associate a light source (conditioned s timulus) with a mild shock (unconditioned stimulus). Tadpoles exposed to lead had a higher mean response time and
49 showed less avoidance. It was concluded that sublethal exposure to lead adversely affects acquisition learning and rete ntion in tadpoles (Steele et al. 1999). In addition to maladaptive antipredator responses, feeding, and learning behaviour, exposure to EDCs also has been associated with altered communication and reproductive behaviour in amphibians. For example, Park and Propper (2002) exposed male red spotted newts ( Notophthalmus viridescens ) to ecologically relevant levels of the insecticide endosulfan or to an acetone carrier control for four days. After th e exposure period, olfactory response tests were run in a Y-maze in which each arm contained a fe male used to attract males. Individuals exposed to the low concentra tion of endosulfan took longer to respond to the presence of females. Morphological measurements of phero mone glands demonstrat ed that endosulfan decreased the alveolar and luminal area. Park and Popper (2002) also studied the mating success of individuals from endosulfan treatments by placing males with a female and recording the presence of spermatophores after 24 hours. Th ey found that males from endosulfan treatments had dramatically lower mating success and conclude d that environmental chemicals may lead to population declines in amphibians by disrupti ng chemical communication and mating behaviour (Park and Propper, 2002). Implications Conservation and Ecology Contam inants are distributed globally; even areas once believed to be pristine are affected by run off, wind blown contaminants, or fallo ut through precipitation (Davidson and Knapp, 2007; Sparling et al., 2001; Thurman and Cromwell, 2000). Importantly, many amphibians are very sensitive to contaminants and are exposed throughout their life history. For instance, they have highly permeable skin as adults and larv ae. Many reproduce in aquatic environments affected by runoff or precipitation, so adults can be exposed thr oughout their lives, in transit to
50 breeding sites, and while breeding. Their larvae experience critic al hormone-regulated developmental stages while in these potentially polluted environments, making them particularly susceptible to endocrine disr uption. Living in, and around, cont aminated areas and undergoing critical stages of development while exposed to endocrine-disrupting chem icals is expected to negatively affect individuals as well as populations (Hayes et al 2006b; but see Glennemeier and Begnoche 2002). The link between pollutants and amphibian de clines has, in some cases, been weak, occurring as a post hoc hypothesis after declines have occurr ed. For example, the last remaining natural populations of Wyom ing toads were known to be exposed to fenthion, an organophosphate insecticide used for mosquito control (Lewis et al., 1985). Importantly, fenthion is known to block androgen recepto rs (in mammals) and androgen-dependent physiological processes (Tamura et al., 2001). Carey and Bryant ( 1995) speculated that this exposure could have been involved in the ul timate extinction of w ild Wyoming toads ( Bufo hemiophrys baxteri ). Male toads in the last populati on showed reduced clasping behaviour during breeding, which is likely an androgen-dependent behaviour, and hatchability of the few fertilized clutches was low (Carey and Bryant, 19 95). Appropriate studies to evaluate the effects of fenthion on Wyoming toad populations were not conducted because the connection between the contaminant and the populati on extinction was made too late. Although Wyoming Toads have been reintroduced, they ar e not currently self sustaining, and research on these wild populations is hampered by the scarci ty of the toads (Dreitz, 2006). Importantly, however, several strong lines of evidence suggest that pollution induces population declines and ex tinctions of amphibians. For exam ple, population declines of several amphibian species are associated with wind-borne agricultural chemicals (Davidson and Knapp,
51 2007; Sparling et al., 2001). In an important and creative study, Sparling et al (2001) used altered cholinesterase (ChE) activity to demonstr ate that exposure to wind-borne pesticide was associated with reduced population status in the Pacific Tree Frog ( Hyla regilla ). Cholinesterase (ChE) is an enzyme that rapidly clears the neurotransmitter acetylcholine from neuronal synapses. The clearance of acetylcholine m odulates neuron-to-neuron communication and is essential for proper muscle function. Many insectic ides and some herbicid es that are typically sprayed in agricultural areas contain pyrethoids and organophos phates. These chemicals are known to inhibit ChE activity and reduced ChE ac tivity is a biomarker of exposure to these pesticides. Sparling et al (2001) found that cholinesterase (ChE) activity in H. regilla tadpoles from mountainous areas downwind of the highly agricultural Central Valley, where the most severe amphibian population declines in California have occurred, was depressed compared to those from sites on the coast or north of the valley where declines are less precipitous (Sparling et al. 2001). In addition, as many as 86% of the populati ons in areas with reduced ChE had detectable levels of endosulfan, up to 50% had measurable organophosphorus levels, and 40% carried DDT residues. This study clearly establishes an association between pesticide exposure, physiological response to that exposure, and poor population st atus. Although a cause-and-effect relationship is difficult to establish with field studies that use such epidemiological approaches, these data are highly suggestive of such a relationship. An alternative hypothesis that could explai n population declines in these areas of California is that introduced non-native fish ha ve increased mortality of tadpoles and driven populations to decline. Theref ore, Davidson and Knapp (2007) st udied the relative effect of introduced fish and pesticides on the mountain yellow-legged frog ( Rana muscosa). They
52 conducted an enormous set of field surveys that included 6,831 sites stud ied over a seven-year period (1995-2002). At each site, they quantified ha bitat characteristics a nd presence or absence of R. muscosa and fish. They calculated windborne pesticide exposure for each site from pesticide application records and predomin ant wind directions. The probability of R. muscosa presence was significantly reduced by the presence of fish and pesticides. However, the effect of pesticides at the landscape scale was much stro nger than the effects induced by fish (Davidson and Knapp, 2007). Importantly, th e degree of protection from wi ndborne pesticides was also a significant predictor of R. muscosa presence. These results demonstrate that windborne pesticides are contributing to amphibian declin es in pristine locations (Davidson and Knapp, 2007). Many agricultural contaminants are known to function as endocrine disrupting chemicals that induce effects at very low concentrations Californias amphibian populations could be suffering declines not because they are exposed to high quantities or leth al concentrations of contaminants, but because they are exposed to very low concentrations of hormonally active agents that alter thyroid function, deve lopment, reproduction, and behaviour. These contaminant-associated declines are not surprising given what is known about EDCs, but it is important to note that few studies have investigated the population-level effects of chronic exposure to pollution, nor have any, to our knowledge, explicitly modeled the way in which EDCs could be involved in the global amphibian de cline. Chemical pollutants, including EDCs, are expected to play an important role in modulating population dynamics because they can directly induce mortality and, at sublethal levels, they infl uence growth and development, decrease the ability of larvae to avoid predators, and alter reproductive success (Carey and Bryant, 1995). In addition, because the immune system interacts direc tly with the endocrine
53 system, EDCs can affect susceptibility to pathogens (Carey and Bryant, 1995; Hayes et al. 2006b). Importantly, EDCs could contribute to recent outbreaks of infectious disease in amphibians, especially when they occur in comp lex mixtures as is typical of contaminated environments. For example, Hayes et al (2006) studied the effects of nine different agricultural pesticides used on cornfields in the Midw estern United States by exposing tadpoles of Rana pipiens to very low concentrations (0.1ppb) of each chemical, separately or in various combinations, and measuring differences in growth, development, gonadal differentiation, and immune function. Many of the single chemical treat ments led to differences in larval growth and development but the pesticide mixtures had more se vere effects. Pesticide mixtures damaged the thymus which was associated with immunosuppre ssion and the contracti on of flavobacteria meningitis (gram negative bacteria). This disease left the animals debilitated; they were in poor condition and unable to hold them selves upright. Interestingly, the control animals tested positive for the presence of this bacteria but did not show signs of the disease (Hayes et al. 2006). This type of interaction between contam inants and disease agents, making the disease more pathogenic, has been hypothesized for many years (Carey and Bryant, 1995). Another recent study has investigated the infl uence of EDCs on dis ease susceptibility by exposing tiger salamanders ( Ambystoma tigrinum ) to atrazine and measuring peripheral leukocyte levels and susceptibility to Ambystoma tigrinum virus (ATV), a pathogen implicated in some amphibian die-offs (Forson and Storfer, 2006). Atrazine significan tly decreased peripheral leukocyte levels and increased susceptibility of the larvae to ATV infection, illustrating that atrazine influences the immune system. Interesti ngly, there were familial differences in infection rates, making particular genotypes more (or less) sensitive. Over all, this study shows that in
54 addition to its effects on testicular development, ecologically releva nt concentrations of atrazine can have immunosuppressive effects that coul d contribute to ATV epizootics (Forson and Storfer, 2006). Evolution Another important consequence of EDC exposure is that it can a lter the evolutionary trajectory of a species. A classic example of pollutant-induced evolutionary change is the darkening of the peppered moth (Biston betularia ) in response to the darkening of trees by soot in England during the industrial revolution. There are also many taxa that have evolved pesticide resistance. Pollutants can induce strong selective forces by increasing mortality and altering genetic variability. Sublethal levels of pollutants can also influen ce evolutionary processes. For example, intergenerational transfer of contam inants (through egg yolk or mothers milk) can alter development, re productive physiology, metabolism, and behaviour in maladaptive ways (Fox, 1995). In addition, many recent experiments have demonstrated that gene imprinting and induction of epigenetic modifications through met hylation is controlled, at least is some cases, by estrogen or other hormones and can be modulat ed by EDCs such as methoxyclor. Therefore EDCs can induce persistent, heritable phenotypic ch anges, independent of mutagenesis. These phenotypic variants will be subjec t to selection and the evoluti onary trajectory of affected species is expected to be altered. Evolutionary changes are in many cases irreversible and they impart an environmental legacy that is broader than that of the polluta nt itself (Leblanc, 1994). What Needs to Be Known Hayes et al (2006) reported that fewer than 30 pub lished laboratory and field studies have addressed low concentration, endoc rine-disrupting effects on am phibians. A few studies have examined the effects of contaminant exposure on wild amphibian populations (Davidson and Knapp, 2007; Ouellet et al. 1997; Sparling et al. 2001) but establishing caus e and effect in such
55 studies is difficult, and descri bing physiological alterations in ducing effects at the population level is daunting. Regardless of these difficulti es, understanding the popula tion-level effects of contaminants on amphibians and the role of ED Cs in global amphibian decline is extremely important, albeit understudied. For example, a lthough several studies have demonstrated that some chemicals can induce feminization of males and intersexuality, little work has focused on the population-level effects of male feminization; this problem is not restricted to amphibians (Hayes et al. 2006a). Many studies on the effects of contaminants such as EDCs have been conducted in controlled laboratory settings. A lthough these studies are extrem ely important, and necessary, for determining the mechanisms of toxicity, th ey do not even attempt to replicate natural systems. For example, many studies investigate th e effects of chemicals at concentrations much higher than individuals e xperience in the wild. Reported effects from these studies do not help explain whether exposed populations will suffer similar effects because the mechanisms of toxicity of EDCs can depend on concentration. A number of studies ha ve documented inverted U-shaped dose responses to EDCs; that is, low to intermediate c oncentrations can produce larger effects than high doses due to the integrated na ture of physiological resp onses, and the presence of negative feedback mechanisms. For example, survival was significantly lower for four different amphibian species when they were expo sed to 3 ppb of atrazine compared with either 30 or 100 ppb and these mortality patterns depend ed on tadpole stage (Storrs and Kiesecker, 2004). Therefore, the deficiency in studies on the effects of lower, ecologically relevant concentrations of EDCs in amphi bians has likely resulted in an underestimation of the impacts of these pesticides (Carey and Bryant, 1995; Hayes et al. 2006a; Hayes et al. 2002). In addition, wild amphibian populations are exposed to co mplex mixtures of contaminants and other
56 stressors, such as UV radiati on, but most studies ha ve investigated the effects of single contaminants without considering other environmental variables (Blaustein et al. 2003; Hayes et al. 2002; Sullivan and Spence, 2003). Contaminant concentrations also fluctuate drastically in natural systems, and these fluctuations can i nduce unexpected and unpredic table effects, but few studies have investigated fluctuat ing contaminant exposures (Edwards et al. 2006). Furthermore, most studies focus on effects of cont aminants during, but not after, exposure. This can lead to underestimations of the effects of contaminant because survival of later life-history stages (e.g. adults) could be nega tively affected whereas earlier, exposed, stages are not (Harris et al. 2000; Rohr and Palmer, 2005; Rohr et al., 2006b). More realistic e xposures that replicate wild systems, and investigate effects across amphibi an life history will help describe the effects that contaminants have on amphibian populations. Questions that attempt to link individual-lev el effects induced by a contaminant in the laboratory to effects in the field are difficult and require integrati on across broad fields that have traditionally been isolated from one another. Although recent studies have begun to address these issues, it is not clear how the mechanisms of toxicity or the effects of chemical pollutants change relative to competition, predation, alterations in animal density, food quantity, or the presence of pathogens (Forson and Storfer, 2006; Relyea and Mills, 2001; Rohr et al. 2004; Rohr et al., 2006b; Romansic et al. 2006). Knowing how ecologi cal factors influence an organisms response to chemicals will help predic t the effects of pollutant s in the wild, and will advance an understanding of how to restore polluted habitats (Rohr et al. 2006a). The phenomenon of endocrine disruption has been focused on documenting estrogenic, androgenic, antiandrogenic, and an tithyroidal actions of specific chemicals, and these pathways of toxicity have now been established in numerous vertebrate species (Guillette, 2006). Many
57 potential mechanisms of endocrine disruption, ho wever, have not been studied. For example, little work has been conducted on the effects of EDCs on other steroid pathways, such as those including the progestins (e.g. progesterone) and glucocorticoi ds (e.g. the stress hormone corticosterone) or on the retinoids (Berube et al. 2005; Guillette, 2006; Hinson and Raven, 2006; Leiva-Presa and Jenssen, 2006). Importantly, it is known that progester one is important for oocyte maturation in amphibians and that EDCs such as methoxychlor, disrupt progesteroneinduced oocyte maturation in Xenopus leavis (Guillette, 2006; Pickford and Morris, 1999, 2000; Pickford and Morris, 2003). In addition, the adrenal is thought to be the most common toxicological target or gan for EDCs (Harvey et al. 2007) but relatively little work has focused on understanding the effects of EDCs on the stre ss axis (hypothalamus-pituitary-adrenals) (Pottinger, 2003). Evaluation of the adrenals are also neglected in EDC screening and testing for regulatory purposes (Harvey et al. 2007). Finally, retinoid imbalances are associated with multiple effects, including changes in sec ondary sexual characteristics, inhibition of spermatogenesis, and decreased embryo survival. Importantly, retinoid homeostasis is affected by several contaminants (Berube et al. 2005). Future studies on endocrine disruption must broaden current approaches. The mechanisms a nd endocrine end points examined need to be expanded, and studies that allow ev aluation of how organisms are aff ected in wild habitats need to be conducted (Guillette, 2006).
58 Figure 1-1 The endocrine system regulates physiological processes by inte grating signals from the external environment via the sensory system and internal signals that are conveyed by endogenous (or internal) hor mones (e.g. GnRH, FSH, LH, steroids, inhibin, activin). These hormones are secret ed by various organs (dark gray boxes), and regulate specific functions (light gray boxes) with in target organs. Any physiological, behavioral or developmental pro cess that is controlled or influenced by the endocrine system is susceptible to modification by EDCs. For example, EDCs alter the synthesis of steroid hormone s (e.g. within the gonad), by altering key steroidogenic enzymes. Altered steroid production affects feedback mechanisms to the brain which changes endocrine regul ation. Altered ster oid production also changes gene expression within target cel ls and organs. EDCs can also directly interfere with gene expression by binding to nuclear steroid receptors. At the liver, EDCs can interfere with hormone degradati on and clearance, and alter the function of steroid plasma binding proteins. All of thes e processes (light gray boxes) are crucial for hormonal homeostasis.
59 Figure 1-2 Lipophilic hormones (grey crosses) easily diffuse from a capillary into a target cell and bind to specific intracellular protein receptors (Black Vs). This complex (receptors and ligand) binds to the DNA and induces gene expression that modulates a wide range of physiological functions such as cellular growth and proliferation, development and differentiation, and maintena nce of homeostasis (slow responses). Most EDCs (white trapezoids) identified to date influence gene expression by interacting with the steroid, thyroid, a nd retinoid receptors found within the cell (Black Vs). Some EDCs can bind to a nd activate a ligand receptor which induces gene expression of responsive genes, whereas others block gene expression by competing for receptors without mimicki ng biological function. Recent evidence suggests that EDCs can also bind to membra ne receptors and influence rapid cellular responses such as metabolism.
60 Figure 1-3 Steroid, retinoid, and thyroid hormones are lipophilic substances that can cross the nuclear membrane (dashed circle). These hormones bind specific receptors (grey oval labeled R) that induce allosteric tran sformation and enable the hormone-receptor complex to bind to high affinity sites in the DNA, called response elements (REs). This binding modulates gene transcription. EDCs can interfere with receptor binding, modulate receptor degradation, a lter receptor activity, or th e activity and availability of co-factors (ovals shown above and between the receptor complex and RNA polymerase). All of these processes are essential for normal receptor function and gene expression. Altered gene expression can lead to a variety of negative effects including abnormalities in the structure a nd function of organs, or induction of cancers.
61 Figure 1-4 Steroid hormone synthe sis involves a complex network of chemical reactions that convert a cholesterol based substrate in to a product (e.g. ho rmones) through the activity of several substrate-specifi c cytochrome P450 enzymes, steroid dehydrogenases, and reductases. The activ ity of many of these enzymes can be modulated by various endocrine disrup ting chemicals resulting in impaired development, growth, sexual differentia tion, reproduction, and the development of particular cancers.
62 CHAPTER 2 GROSS GONADAL ABNORMALITIES AND A LTERE D SECONDARY SEXUAL TRAITS ARE ASSOCIATED WI TH AGRICULTURE Introduction Endocrine signaling regulates reproductive system developm ent and plays an im portant role in organizing embryonic tissues to respond to hormonal signals la ter in life (Guillette et al. 1995). Therefore, exposure to endocrine disr upting chemicals that alter normal hormonal signaling during embryonic development can permanently change reproductive system morphology and function of adults (Guillette et al. 1995; McLachlan, 2001). Many of studies that have documented endocrine disrupting effects and associated reproductive abnormalities in wildlife populations have focused on the effects of sewage or pulp mill effluent on fish (Jobling et al. 2002a; Jobling et al. 2006; McMaster, 1995; McMaster et al. 1991; Munkittrick et al., 1994; Tyler & Routledge, 1998; van Aerle et al. 2001; Woodling et al. 2006). However, many of the chemicals used in agricu ltural practices are known endocrine disruptors that influence the thyroid and reproductive systems of non-target wild life. For example, cattle feedlot effluent contains androgens from hormone implants that negatively affect the endocrine and reproductive systems of wild fathead minnows ( Pimephales promelas) (Ankley et al. 2005; Ankley et al., 2001; Ankley et al., 2003; Orlando et al. 2004). Many pesticides disrupt the endocrine system of non-target wildlife and induce diverse reprodu ctive abnormalities. For example, pesticides from farming practices negatively affect steroi dogenesis, circulating hormone levels, hepatic transformation of androgens, and gonadal and thyroid morphology of American alligators ( Alligator mississippiensis) (Guillette, 2001; Guillette et al. 1994; Guillette & Iguchi, 2003; Hewitt et al. 2002). Pesticides such as the fungicide vinclozolin negatively affect fish and mammal reproduction, and alter germ cell DNA me thylation patterns making these reproductive effects heritable which could lead to population d eclines (Ankley et al. 2005; Crews et al. 2007;
63 Gray, Ostby & Kelce, 1994; Gray, Ostby & Marsha ll, 1993). Recent studies have suggested that agricultural contaminants are associated with amphibian reproductive abnormalities and population declines (Davidson & Knapp, 2007; Ouellet et al. 1997; Sparling, Fellers & McConnell, 2001). Amphibians are declining on a global s cale and the role th at pollutants play in these declines is relatively unclear. Amphibians make appropriate models for studyi ng the effects of EDCs on wildlife, in general, and might serve as bio-indicators of EDC exposure. Their gonadal differentiation and secondary sexual characteristics are highly sensit ive to sex hormones, and chemicals that alter sex hormone function (Hayes, 1998; Qin et al. 2003). Alterations of normal gonadal function and hormone concentrations can generate identifiable changes in secondary sexual traits that are indicative of endocrine disrupti on and may be used as visual tags or biomarkers for such disruption (Hayes & Menendez, 199 9; Noriega & Hayes, 2000; va n Wyk, Pool & Leslie, 2003). The objectives for this study were to survey Bufo marinus across a gradient of agricultural intensities to investigate whether repr oductive abnormalities in gonadal morphology are associated with agriculture, and to determine if gonadal abnormalities are associated with altered gonadal function. One of the primary functions of gonads is to produce hormones that control the expression of secondary sexual traits. Therefore, I compared the secondary sexual characteristics of toads with morphologically normal and abnormal gonads to determine whether gonadal function was compromised in toads with abnormal gonadal morphology. I also investigated whether specific secondary sexual characteristics vary as a function of abnormal gonadal morphology in predictable ways that promote the development endocrine disruption biomarkers. I hypothesized that toads collected from sites with greate r agricultural activity, where known EDCs are routinely used (Appendi x 1), would have a higher incidence of
64 reproductive abnormalities than toads collected from suburban sites with little to no agriculture nearby. In addition I hypothesized that to ads with gonadal abnormalities would display predictable and easily measured abnormalities (bio markers) in their secondary sexual traits. The proportion of agriculture at each collection site in this stu dy is inversely related to the degree of urbanization, and its asso ciated stressors. Suburban sites are not pristine habitats; they are typically polluted with polyaromatic hydro carbons (PAHs) associated with road runoff (Crosbie & Chow-Fraser, 1999; Maltby et al. 1995). PAH concentrations have been used to estimate the proportion of urbanized land among different watersheds (Standley, Kaplan & Smith, 2000). Agricultural areas, however, are typi cally polluted with pesticides used to control weed, insect, and rodent populations (Crosbie et al. 1999). Although, many people living in suburban areas use pesticides on their lawns urbani zed and agricultural hab itats are polluted with different milieus of chemicals that are expected to induce different suites of pathologies. I find that reproductive abnormalities indicative of endoc rine disruption increase with agricultural intensity and are correlated with agricultural pollutants. Methods Study System Fe male Bufo marinus are multi-colored (with spots rangi ng from white to dark brown or black, Figure1), whereas male s are a solid toffee color (L ee, 1996; Zug, 1979). The dark coloration found on females is caused by high concentr ations of melanin, which is thought to be modulated by estrogen (Costin & Hearing, 2007; Shah & Maibach, 2001). Therefore, populations of males that have higher endogenous levels of estr ogen, or have been exposed to estrogenic chemicals are predicted to have darker skin and more mottling typical of females. Male Bufo marinus have larger forelimb bones, and hypertrophied musculature relative to similarly sized females (Lee & Correales, 2002) (Fi gure 2-1). Forelimb musc ulature in males is
65 androgen dependant, and is believed to aid males during amplexus (Lee et al. 2002). Reductions in androgens are likely to reduce forelimb size, and pot entially negatively affect male reproductive success. Male Bufo marinus also have nuptial pads on their first through third phalanges that function in mating (Figure 2-1) (Epstein & Blackburn, 1997; van Wyk et al., 2003). The presence of epidermal hooks, secretory activity, and the number of nup tial pads are androgen dependent and occur only when animal s are reproductively active (Epstein et al. 1997; van Wyk et al. 2003). Thus, nuptial pad activity and mor phology during the breeding season is an index of both androgen and repr oductive status (Epstein et al., 1997; van Wyk et al. 2003). Another key characteristic of B ufo marinus that makes them an especially useful model in studies of endocrine disruption is that both male and female bufonids possess a Bidders organ, which has been characterized as a non-functional or vestigial ovary (Far ias, Carvalho-e-Silva & do Brito-Gitirana, 2002). Although it contains ooc ytes at various stages of oogenesis, there are several clear differences between ovaries and the Bidders organ. Morphologically, an ovarian pocket does not form in the Bidders or gan (Falconi, Dalpiaz & Zaccanti, 2004; Falconi, Dalpiaz & Zaccanti, 2007; Witschi, 1933). In ad dition, the oocytes of the Bidders organ do not accumulate vitellogenins, which are the fe male-specific egg yolk protein precursors (lipophosphoglycoproteins) produced in the liver unde r estrogen stimulation, and transported to maturing oocytes (Brown, del Pino & Krohne, 2002; Falconi et al., 2004; Falconi et al. 2007; Farias et al. 2002). Although the functiona l significance of the Bidde rs organ is unknown, this organ does not undergo vitellogenesi s (oocytes are not yol ked) in normal individuals (males or females); oogenesis in the Bidde rs organ normally ends with degeneration of large previtellogenetic oocytes (Brown et al. 2002; Falconi et al. 2004; Falconi et al. 2007; Farias et
66 al. 2002). Even when high levels of gonadotropins are administered, ooge nesis of the Bidders organ remains inhibited (Pancak-Roessler & No rris, 1991). Therefore, bufonids are not hermaphroditic, and under normal conditions the Bi dders organ does not function as an ovary in males (Hayes, 1998). Experimental castration, ho wever, induces the oo cytes in the Bidders organ to become morphologically identical to ovarian oocytes, and they can then undergo vitellogenesis (Brown et al. 2002). In addition, reducing the le vel of dehydrotestosterone in developing bufonids increases development and growth of Bidders organ oocytes (Zaccanti et al. 1994). Therefore, testes inhi bit vitellogenesis in the Bidder's organ of males, and this inhibition is likely controlled, in part, by testicular androgens (Pancak-Roessler et al. 1991; Zaccanti et al. 1994) Study Sites and Collection Techniques I surveyed five sites in south Florida that va ried in habitat type from completely agricultural to suburban (Figure 2-2), to investigate the asso ciation between intensity of agriculture and reproduc tive abnormalities in B. marinus I calculated the total amount of agricultural land within a 5.6 km2 area around each collection site, to determine the intensity of agriculture. The home range of B. marinus is ~2km2 (Zug, Lindgren & Pippet, 1975), so this grid includes the area likely to be experi enced by a toad from each local site. I imported Google Earth digital satellite images and scale into Image Pro image analysis program (on August 20, 2007), and centered a 5.6 km2 grid over each collection site. Each of the 9 cells within the grid re presented approximately 622 m2. Using these quadrats I estimated the percentage of agricultural area (vegetable a nd sugarcane farming, no lives tock) within each cell, and averaged these values for each site. The five study sites ranged in percentage of agricultural land from 0 to 97%. Two sites (Wellington and Lake Worth) were completely suburban and contained zero agricultural land, but they varied in proximity to agriculture. Therefore, to
67 distinguish these two sites I calc ulated distance from each site to the closest agricultural area. Wellington (WT) was closer to fa rm land (5.18 km away) than Lake Worth (LW) (22 km away). The sites were then assigned an agricultural intensity score of one through five that was based on the percentage of agriculture in the 5.6 km2 area around the site or its proximity to agriculture when there was no agriculture nearby (Figure 2-2). Lake Worth (LW) did not contain any agriculture, was completely suburban, and it was farthest from an agricultural area; therefore it was assigned a rank of one. Wellington (WT) was assigned a rank of two. Homestead (HS) contained 34% agricultural land and was ranked as three. Canal Point (CP-51.2%) and Belle Glade (BG-97%) contained the most agricultural land and were ranked 4 and 5 respectively. At least twenty sexually mature adult indi viduals (>91mm, (Cohen & Alford, 1993) were collected from each site. Coll ections occurred at BG, CP, an d WT during the summers of 2005 and 2006, HS toads were collected in summer 2005 and JP animals were collected in summer 2006. There was no effect of year on the percentage or types of abnormalities documented at sites where collections occurred in more than one year. Toads from all locations were collected, while feeding, under streetlight s. Each individual was euthanized with an overdos e of the anesthetic MS222 (0.3% Tricaine Methanesulfonate, pH 7), photographed (Konica Minolta Dimage Z10 w ith 3.2 mega pixels), and measured for mass (Ohaus Scott 2) and snout vent length (Spi dig ital caliper). Gonads were removed and one gonad from each individual was fixed in neutral buffere d formalin for gross morphological analysis. Analysis of Primary Se xual Characteristics Individuals were classified in to four broad categories ba sed on their gonad m orphology to determine whether gonadal abnormalities increased with agriculture intens ity: 1) normal males, 2) abnormal males, 3) intersex individuals, and 4) females. Normal males had testes and nonvitellogenic Bidders organs, a nd no ovaries or oviducts. Ab normal males had testes and
68 vitellogenic Bidders organs, but no ovaries or oviducts. Intersex individuals had testes, and ovaries or oviducts (t he Bidders organ was not considered for this grouping). Females had ovaries, oviducts, non-vitellogenic Bidders orga ns, and no testes. Normal, abnormal, and intersex individuals were ranked (1,2, and, 3 respectively) and differences in the mean rank across sites were ev aluated with a Kruskal-Wallis one way analysis of variance to test whether sex group rank varied across sites. Kenda lls Tau rank correlation was used to determine if the change in the sex group rank corresponded to agricultural intensity (test of association). I created a scoring index to categorize individuals based on the type and total number of gonadal abnormalities each toad possessed. No obvious gonadal abnormalities were seen in females. Therefore, I focused on gonadal abnormalities in normal and abnormal males and intersex individuals. Scores were assigned base d on abnormalities in each of the following three tissue types: 1) testes, 2) Bidders organ, and 3) ovaries/ oviducts (in the presence of testes and a Bidders organ). Abnormalities involving testes included multiple and misshaped testes. Bidders organ abnormalities included vitellogenic Bidders, or multiple Bidders organs. Abnormalities involving ovaries/oviducts included the presence of regressed ovaries, vitellogenic ovaries, multiple ovaries, or oviducts (in the presence of testes and Bidders organs). As an example of my scoring protocol, an individual that had misshaped testes (1 abnormality in testes category), a vitellogenic Bidders organ (1 abnor mality in Bidders organ category), and ovaries that were also vitellogenic (2 abnormalities in ova ry/oviduct category) received a score of 4. The number of individuals with each abnormality type was evaluated across sites with a KruskalWallis on way analysis of variance (SPSS 15.0) to determine if the occurrence of specific abnormalities differed across sites, and Kendalls Ta u rank correlation was used to test whether
69 individual abnormalities were associated with agricultural intensity. In addition, the total number of abnormalities were summed and simila rly evaluated to determine if more gonadal abnormalities occurred at specific sites and if the number of abnormalities corresponded to increased agricultural intensity. I was intere sted in evaluating whether the frequency of abnormalities corresponded with increasing agricultu ral intensity therefore I used a one-tailed test for all Kendalls Tau rank correlations (Siegel & Castellan, 1988). Analysis of Secondary Sexual Characteristics Therefore, I used photographs to compare num ber of nuptial pads, body color pattern, and forelimb size in relation to sex group (female, normal male, abnormal male, and intersex) to determine if secondary sexual traits varied amon g normal, abnormal, and intersex individuals. Nuptial pads I calculated the m ean number of nuptial pads per forelimb for each intersexed, abnormal, and male individual. No individuals scored as female had nuptial pads. Univariate one way analysis of variance (ANOVA, SPSS 15.0) was used to determine if the number of nuptial pads was different among groups with different degrees of reproductive abnormalities (normal, abnormal and intersexed). Color pattern I characterized the am ount of mottling across an individuals body (all sexes included) by drawing a transect line from the left eye to the vent directly onto the image of each individual using Image Pro Plus analysis program, and coun ted the number of colo r changes that occurred across this transect (Figure 2-1) Mottling scores of sex groups were compared using a one way ANOVA. I analyzed male and female mottli ng scores across sites using a two-way ANOVA (SPSS 15.0) to test for differences in color di morphism (by sex) across sites (sex*site).
70 Forelimb size I m easured the width of the forelimb across th e radioulna distal to the humerus (elbow), and perpendicular to the arm axis using Image Pro Plus analysis software (F igure 2-1) to test for differences among sex groups. I used ANCOVA with log10 elbow width as the dependent factor, log10 (body length) as the covariate, and gona dal abnormality group (male, abnormal, or intersexed) as the fixed effect to account for the influence of body size. Slopes were not significantly different across sex groups (interaction not signifi cant), so the interaction was removed from the analysis, and the simplifie d ANCOVA (SPSS 15.0) was used to estimate marginal means for the elbow width of each group (+/95% confidence interval). Results Analysis of Primary Se xual Characteristics Sex group rank (m ales=1, abnormal=2, intersex=3 ) increased as a function of agriculture (KW = 25.16, df = 4; p<0.001; Kendalls Tau = 0.8, p=0.008; Figure 2-3a). More intersex individuals occurred at the two most agricultural sites relative to the other sites. The total number of gonadal abnormalities also differe d among sites and increased with increasing agriculture (KW = 21.2, df = 4; p<0.001; Figure2 -3b, Kendalls Tau =0.8, p=0.008). Testicular abnormalities were relatively common but did not vary across sites or increase as a function of agricultural intensity (KW = 5.148, df = 4; p=0.272; Kendall s Tau =0.3, 0.24>p<0.41, data not shown). Bidders organ abnormalities varied marginally across sites (KW = 9.04, df = 4; p=0.06; Figure 2-3c) and was highest at the agricultural sites (Figure 2-3c), but did not statically correspond to increasing agriculture (Kendalls Tau =0.4, p=0.24). The rank of female tissue abnormalities was significantly different among sites and increased with increasing agricultural intensity (KW = 17.3, df = 4; p=0.002; Figure 2-3d, Kendalls Tau =0.8, p=0.008).
71 Secondary Sexual Characteristics Nuptial pads The m ean number of nuptial pads was greate st for males (mean = 2.3) and was lower for abnormal males (mean = 2.1) and intersexed individuals (mean = 1.7) (F = 3.5, df = 2,93; p=0.035; Figure 2-4). Fifty percent of the normal males had three nuptial pads per forelimb (one on each digit), whereas only 33% of the abnormal and 29% of intersexed individuals had three nuptial pads (Figure 2-4). Fortysix percent of intersexed indivi duals had zero or one nuptial pad per forelimb compared with <20% for either abnormal or male individuals. Mottling Toads differed in m ottling score as a f unction of sex category (F=22.9, df= 3, 150; p<0.001). Females were more mottled (mean mottling score 9.6) than males (mean mottling score 5.4), and intersex indivi duals were intermediate (mean mottling score 7.6) between and significantly different from both females and nor mal males (Figure 2-5). Abnormal males had a mean mottling score similar to intersex individua ls (7.4), but showed such high variation that they were not distinguishable from any sex groups (Figure 2-5). There was a significant decrease in mottling sexual dimorphism when males and females were compared by site, which was indicated by th e significant interaction between sex and site (F=2.7, df= 4, 103; p<0.033). Although female mottling did not change as a function of agriculture, the sexual dimorphism in mottling became less pronounced as agricultural intensity increased (Figure 2-6). Forelimb size Forelim b width was also significantly different among sex groups (F = 65.48, df 3, 89; p<0.001) compared at the grand mean of SV L=2.08). Females had significantly smaller forelimbs than any other group (raw mean = 1.15 cm), and as with color pattern intersex
72 individuals were intermediate (raw mean = 1.37cm) to and significantly different from both males (raw mean 1.51cm) and females (Figure 2-7). Abnormal males (raw mean 1.47cm) had larger forelimbs than females, and their forelimbs were intermediate in size between intersexed or normal males. Discussion W ildlife species living in agricultural areas, or areas polluted by ag ricultural runoff, are known to have various reproductive abnormalities, and controlled laboratory experiments have demonstrated that chemicals used at these s ites can alter endocrine system function (Guillette et al. 1994; Hayes et al. 2003; Orlando et al., 2004). This study shows that reproductive abnormalities in primary sexual traits of Bufo marinus are associated with agriculture. Intersexed individuals occur more commonly at agri cultural sites than at less agricultural sites. The number and developmental stage of female gona dal tissues found in the pr esence of testes in this study is considered a measure of the severity of the intersexed c ondition. This index of intersexuality increased with agricultural intens ity in a dose dependent manner (Figure 2-3 d). This dose response relationship is suggestive of a causa l link between agricultural contaminants and development of intersexed condition. The exact cause for these abnormalities can not be identified by this study. However, nitrate (fertiliz er), and some of the pesticides (e.g.s atrazine, glyphosate) used at the sites in this study are known endocrine di sruptors, and are likely inducing the documented abnormalities. Toads from agricultural sites had a variet y of gonadal deformities that alter gonadal function and might reduce reproductive success. For example, vitellogenesis of the Bidders organ in males is only known to occur after castra tion, which has lead to the conclusion that the testes are necessary in males to suppress vitellogenesis w ithin the Bidders organ (Brown et al., 2002; Pancak-Roessler et al. 1991; Zaccanti et al. 1994) However, we found that
73 approximately 20% of the individuals (with test es) at the agricultural sites had vitellogenic oocytes within their Bidders or gans, which suggests that the testes of these toads do not function normally to suppress the Bidders organ. My data demonstrates that secondary sexual characteristics were altered in toads with abnormal and intersexed gonads, which suggests th at testicular function was compromised in toads with these abnormalities. Secondary sexua l characteristics are maintained via endocrine signaling, and are typically modulated via se x steroids produced in the gonads (Hayes et al., 1999). For example, nuptial pads are androgen de pendent breeding glands found on the digits of the forelimbs of several anuran (frog and toad) species, and function in mating (Emerson et al., 1999; Epstein et al., 1997; Lynch & Blackburn, 1995; Rastogi & Chieffi, 1971; Thomas & Licht, 1993; Thomas, Tsang & Light, 1993; van Wyk et al., 2003; Wetzel & Kelley, 1983). In this study, the proportion of individuals with maxi mal nuptial pad developm ent (e.g. three nuptial pads per forelimb) was lower in intersexed in dividuals suggesting that androgen status was compromised. Nuptial pad development is seas onal, and non-reproductive toads do not possess well developed nuptial pads. Th erefore, decreased nuptial pad development in intersexed animals suggests that they might not be as reproductively active rela tive to normal males. Forearm width, another androgen dependent sexually dimorphic secondary sexual characteristic, was altered in intersex individual s relative to males sugges ting that they did not build the hypertrophied musculature typical normal levels of androgens, and thus properly functioning testes (Lee et al. 2002). The dimorphism in nuptial pads and forearm width aids the male during amplexus, and is under sexual selection (Lee et al. 2002). Therefore, abnormalities found in this study might adversely influence reproductive fitness and alter sexual selection.
74 The occurrence of intersex individuals c ould arise through feminization of males or masculinization of females. If females were masculinized I would have found individuals with relatively complete female reproductive tracts, an d some combination of male traits (e.g., nuptial pads, decreased mottling, wider arms, small testes). Individuals did not have morphologies that appeared to be incrementally more masculine. No toads with relatively normal looking ovaries and small or partially developed testes were coll ected. Instead, intersex individuals typically had normal sized testes, but varied in ovary si ze and stage of oogenesis. In addition, no females (ovaries and oviducts, but no testes) were collect ed that had nuptial pads, or solid coloration. Females did not differ in color pa ttern across sites, but males cat egorized as normal were more mottled in agricultural areas, and the typical sexual dimorphism in mottling was not observed at the most agriculture site (Figure 2-6). The sexually dimorphic external morphological traits of B. marinus make them an ideal model species for studying the effect s of EDCs because these traits function as a visual tag for determining whether an individual is experienci ng endocrine disruption. Female coloration in the reed frog ( Hyperolius argus ) is estrogen dependent, and earl y induction of female color pattern has been used as a biomarker for estr ogenic activity (Noriega & Hayes, 1999; Noriega et al. 2000). The dark skin colo ration typical of female B. marinus is due to dense melanin concentrations that are thought to be modulated, at leas t in part, by estrogen and may, therefore, function as a biomarker of estrogen exposur e as well (McLeod, Ranson & Mason, 1994; Shah et al. 2001; Slominski et al. 2004; Thorton, 2002). The gross morphology of the Bidders organ is an ideal biomarker for androgen status and testis function because it is known to develop into a vitellogenic ovary after experimental castration. Thus, animals exposed to antiandrogeni c chemicals are expected to have vitellogenic
75 Bidders organs. However, only bilateral orchid ectomy has resulted in Bidders organ oocyte growth and increased oogenesis, so the presence of one testis is sufficient to inhibit Bidders organ development (Pancak-Roessler et al. 1991). Therefore, the presence of abnormal Bidders organs in toads investigated in this study suggests that testicular function was severely altered in agricultural toads. In addition to having diagnostic external and internal biomarkers of endocrine disruption, Bufo marinus are widely distributed and could be used to study the relationship between various agricultural practices an d the types and severiti es of reproductive abnorm alities. A global study would provide important informati on about the types of pesticides that are associated with most abnormalities. In addition, B. marinus could be used to study the global distribution of EDCs and their effects across diverse habitats. Although this study was not design ed to determine the specific cause of the reproductive effects observed, a few of the pesticides used at these agricultural sites cause endocrine disruption and have well studied mechanisms of toxicity. For example, glyphosate (Round Up) and atrazine are used at both sites and are known to disrupt steroidogenesis (Walsh et al. 2000(Oliveira et al. 2007; Soso, Barcellos & Ranzan i-Paiva, 2007). Glyphosate disrupts steroidogenic acute regulatory (StA R) protein expression which modul ates the initial step in the steroidogenic pathway (Walsh et al., 2000). Such disruption leads to reductions in steroid hormone production, including both testosterone and es trogen, across diverse taxonomic groups. Glyphosate exposure leads to reductions in andr ogen and estrogen dependent traits. Although this study suggests that androgen dependent traits are reduced (demasculinization), possibly due to reduced androgens, I also found that estrogen mediated traits such as melanin are increased (feminization). Research conducted across severa l vertebrate classes have demonstrated that
76 atrazine exposure inhibits androgen-mediated development, and could produce estrogen-like effects by increasing aromatase transcription (Crain et al. 1997; Fan et al. 2007; Sanderson et al. 2002). Therefore, the mechanisms of toxicity of atrazine are consistent with the patterns of demasculinization and feminization that we obser ved in this study, but ot her studies have not found strong relationships between agricultural us e of atrazine and reproductive abnormalities in amphibians (Murphy et al. 2006). The timing, amount, and fluc tuation of contaminant exposure are important factors that determine the phenotyp ic response to EDCs, so these factors could explain the differences among studies (Edwards et al., 2006; Guillette et al. 1995; Milnes et al. 2006). Many of the observed gonadal abnormalities associated with agriculture in this study are likely reproductive deformities that occur during development and are maintained throughout the toads adult lives. Inte rsexed gonads in this study are dis tinct ovarian and testicular tissues suggesting that reproductive development is fundamen tally altered in toads living in these areas. The morphology of the intersexed gonads of th ese toads is dramatically different from the ovotestes that have been found in many fish species around the worl d after exposure to synthetic estrogens (Jobling et al. 2002a; Jobling et al. 2002b; Jobling et al., 1998; Jobling et al. 1996; Jobling et al. 2006; Woodling et al. 2006). Ovotestes occur when oocytes are found within the testes suggesting that endocrine signals leading to the differen tiation of particular cells are disrupted. However, the abnormalities that we find in B. marinus from agricultural areas suggest that entire developmental cascades leading to the differentiation of th e gonad are altered such that both male and female reproductive organs form. Both types of intersex phenotypes (distinct organs vs. ovotestes) alter gonada l function leading to altered secondary sexua l traits, and could
77 decrease reproductive success. However, the developmental mechanisms leading to these different intersexed phenotypes are likely dist inct and induced via different pathways. Because, EDCs are distributed globally they ha ve been hypothesized to be a factor in the global decline of amphibian populations (Carey & Bryant, 1995; Davidson et al., 2007; Hayes et al. 2002; Sparling et al. 2001). However, it is important to note, that relative ly little is known about the role EDCs play in amphibian populatio n declines (McCoy and Guillette in press). I believe that B. marinus could be developed into a bioindi cator of EDC exposure to facilitate molecular, cellular, and population level studies. With some creativ ity this toad which is often detrimental to wildlife species where it has been introduced, coul d be valuable for identifying, understanding, and mitigating the effects of EDCs on wildlife in general. Figure 2-1 Male (M) and female (F) Bufo marinus have several sexually dimorphic characters. Males (M) are less mottled than females (F). Mottling scores were enumerated by counting the number of color changes th at occurred across the yellow transect running down the back of the toad from the ey e to the vent. Males have larger arms than females of equivalent size. The por tion of the forelimb that was measured and compared among the groups is depicted as a black line. Males also can have nuptial pads on the first three digits (this male was scored as having two nuptial pads).
78 Figure 2-2 Field Sites. a. Southern Florida (USA ) showing each collection site. b. Close up view of the four most northern collection site s. The area around Belle Glade (BG) and Canal Point (CP) is a patchwork of agri cultural fields. The area around Wellington (WT) and Lake Worth (LW) are suburban developments. c. Close up view of the most southern collection site. The ar ea around Homestead (HS) is a mixture of agricultural and nonagricultural land.
79 Figure 2-3 Individuals (LW=11; WT=16, HS=14, CP=34, BG=33) that had both testes and ovaries or oviducts were categorized as intersexed (I), those with vitellogenic Bidders organs but no ovaries or oviducts were abnormal males (A), and males (M) had testes, a non-vitellogenic Bidders or gan, and no ovaries or oviducts. a. Percentage of individuals that were classified as inters exed, abnormal, and male at each site. b. Total number of testic ular, Bidders organ, and female tissue abnormalities were enumerated and the percen tage of individuals having a specific frequency of abnormalities are plotted for each collection site. c. Percentage of toads with or with out Bidders organ abnormalitie s at each collection site. d. Percentage of individuals with or without a specific number of female tissues or advanced stage of reproductive development. Aall of these i ndividuals also had tes ticular testes. For example, intersexed toads that had ovaries (1 ) and oviducts (1), or that had ovaries (1) that were vitellogenic (1) were both scored as having two abnormalities. Agricultural intensity increases along the X-axis fr om left to right for each graph.
80 Figure 2-4 Percentages of male (N=59), abnorma l (N=12), and intersexed (N=25) toads with zero through three nuptia l pads per forelimb. No individuals scored as female had nuptial pads.
81 Figure 2-5 Mean mottling score across an individuals body for each sex. Mottling was evaluated by drawing a transect line from th e left eye to the vent directly onto the image of each individual and counting the number of color changes (see Figure 1). Samples sizes were: females = 55, inte rsex = 29, abnormal = 11, and male = 59.
82 Figure 2-6 Mean mottling score for females and males across sites. Agricultural intensity increases along the X-axis from left to ri ght for each graph. Sample sizes were: FemalesLW=10, WT= 5, HS=4, CP= 16, BG =20; Males LW=10, WT= 15, HS=10, CP= 10, BG =10
83 Figure 2-7 Mean forelimb width for each se x or abnormality group measured across the radioulna distal to the hume rus (elbow), and perpendicular to the arm axis. Mean values are corrected for body size and repres ent estimated marginal means from the ANCOVA with log10 elbow width as the dependent factor, log10 (body length) as the covariate, and gonadal abnormality group (male, abnormal, or intersexed) as the fixed effect. Sexes are compared at a log10 SVL of 2.08. Bars are 95% Confidence intervals. Sample sizes were: females = 28, intersex = 17, abnormal = 9, and male = 41.
84 CHAPTER 3 MARINE TOADS LIVING IN AGRICULTURAL ARE AS HAVE ALTERED GONADAL FUNCTION Introduction In adult vertebrates, gonads pe rform two primary functions. Th ey are the location of germ cell production (e.g. spermatogenesis in males), and they function as important endocrine glands producing a variety of steroid a nd peptide hormones. As a resu lt, effects of EDCs on gonadal development can affect spermatogenesis and endocrine hormone production. The gonads of animals that are exposed to e ndocrine disrupting chemicals (EDC s) during critical periods of development or sexual differentiation can be perm anently altered so that they can not function normally in adults (Guillette et al. 1995; Guillette et al., 1994). Such permanent alterations that occur during embryonic development are referr ed to as organizational abnormalities. Although the effects of endocrine disrupting chemicals (EDCs) have been documented extensively in vertebrates (Edwards, Moore & Guillette, 2005; Milnes et al. 2006), amphibians are hypothesized to be especially sensitive to th eir effects (McCoy and Guillette, 2008 in press; Chapter 1). Amphibians have highly permeable sk in and many live in habitats that receive runoff or precipitation that cont ains EDCs. Their larvae experience critical hormone-regulated developmental stages, especially during metamorphosis, making them particularly susceptible to permanent (organizational) endocrine disr uption in polluted environments (e.g., Hayes et al 2002). I previously demonstrated a relationship betw een exposure to agricu ltural habitats where endocrine disrupting chemicals (EDCs) are used and abnormal and intersex gonads in Bufo marinus (Chapter 2). Gonad abnormalities were associ ated with demasculinized male secondary sexual characteristics and reductions in sexua lly dimorphic coloration (Figure 1). This association suggested that gona dal function was compromised si nce gonadally-derived steroid
85 hormones modulate secondary sexual characteristics. Here, I inve stigate whether two additional measures of gonadal function, spermatogenisis a nd sex hormone concentrations, are altered in B. marinus that have gonadal abnormalities resulting from inhabiting areas in close proximity with agriculture. Specifically, I tested the hypotheses that 1) spermatoge nesis is altered in toads living in agricultural areas where EDCs are applied re lative to suburban areas, 2) spermatogenesis is increased in males from non-agricultural areas rela tive to males, abnormal males, and intersexed toads from agricultural areas, 3) estradiol-17 and testosterone concen trations are altered in toads living in agricultural areas where EDCs are applied relative to suburban areas, and 4) estradiol-17 and testosterone concentrations are di fferent among sex groups. Plasma estradiol17 concentrations are expected to be greater in toads (e xcluding females) from more agricultural sites than from subur ban areas, and testosterone is expected to show the opposite pattern. Females are expected to have higher estradiol-17 concentrations relative to all other sex groups, whereas males from non-agricultural ar eas are hypothesized to have lower levels than male, abnormal, and intersexed toads from agricultural areas. Testos terone concentrations are expected to be lower in females than in all other groups, whereas ma les from non-agricultural areas are expected to have highe r concentrations relative to male, abnormal, and intersexed toads from agricultural areas. Methods I surveyed five sites in south Florida that va ried in habitat type from completely agricultural to suburban to inve stigate the associati on between intensity of agriculture and reproductive abnormalities in Bufo marinus At least 20 sexually mature adult individuals (> 91mm) (Cohen & Alford, 1993) were collected from each site during the summers of 2005 and 2006. Each individual was euthanized with an overdose of the anesthet ic MS222 (0.3% Tricaine
86 Methanesulfonate, pH 7), and 3 to 4 ml of w hole blood were collected via cardiac puncture. Whole blood was centrifuged, and plasma was colle cted for hormone assays. Plasma samples were frozen at -20C immediately, transported back to the laborato ry on dry ice, and then placed at -80C until hormone assays were conducted. Gonads were removed and fixed in neutral buffered formalin for gross morphological (see Ch apter 1) and histologi cal analyses (reported here). Histological Preparations Fixed tissues were cut in ha lf, dehydrated in alcohol, em bedded in paraffin wax, and section so that the middle portion of each individu als testis was cross-sectioned and analyzed. Sections were 8 m thick, and were stained us ing standard Hemotoxylin and Eosin (Presnell & Schreibman, 1997). Quantification of Spermatogenesis Five slides per anim al were examined using an Olympus BX50 microscope,, and one section was evaluated per slide. Sections examin ed were approximately 180m apart so that the same cells were not evaluated in different sect ions. Spermatogenic cells in five lobules per section were quantified, yielding a total of 25 lo bules for each individual. For each lobule the number of spermatocysts containing either primar y spermatocytes, secondary spermatocytes, or spermatids (both round and elongate) were quantified. The number of spermatozoa (loose and organized) per lobule was categor ized as 0 (approximately 0-25% of the total area of the lobule), 1 (25-50%), 2 (50-75%), or 3 (75-100%). Hormone Assays Estrad iol-17 (E2) and testosterone (T) concentrations were analyzed with a validated radioimmunoassay using the 96-well FlashPlate PLUS system (Per kinElmer, City). Each flash plate was coated with 10 0L of antibody (1:30,000 E2, or 1:15,000 T) purchased from Fitzgerald
87 Industries International, Inc. (20-ER06 Estradiol-17 beta6 CMO lot#P410311520-TR05 Testosterone-19, Lot # P4033111), covered with aluminum foil to prev ent evaporation, and incubated at room temperature for at leas t two hours. Standard curves (either E2 or T) were prepared with the following concentrations: 1800, 600, 200, 66.7, 22.2, 7.4, and 2.5 pg/100 L for each assay plate. Samples were extracted wi th diethyl ether and reconstituted with phosphate buffered saline gelatin (PBSG; 0.1M, pH 7.0) buffe r buffer prior to being loaded into the assay plate. After antibody incubation each well was washed two times with 150 L of PBSG buffer and each well was loaded with 100 L of standard hormone or unknown sample, and the appropriate radiolabeled hormone (either E2 or T). Plates were inc ubated at room temperature for three hours and counts per minute of radi oactivity were analyzed using a Microbeta 1450 Trilux counter (PerkinElmer). Prior to perf orming hormone assays, these procedures were validated by ensuring a reconstitu ted plasma spiked standard curve was linear with the standard curve (internal standard) 2) and that dilution curves of extracted and reconstituted plasma were linear. Statistical Tests Statistical analyses were carried ou t using SPSS version 15.0. The composition of spermatogenic cell types was evaluated by site and sex group using separate analysis of variance (ANOVA) tests. Post Hoc comparisons to assess the affects of site and sex on spermatogenic cell numbers where conducted using Tukey HSD tests. Medians of ranked sperm abundance were evaluated among sites and sexes using Kruskal-Wallis one way analysis of variance tests. When 3 or more groups are evaluated in a Kruskal-Wallis test and when the number of observations within groups exceeds 5 the distribution of KW approximates the Chi square distribution, thus the chi square statistic is reported here (Siegel & Castellan, 1988).
88 Hormone concentrations were valuated us ing a linear mixed model where assay and assay by dependent factor (site or sex) interaction were identified as the random effects. Estradiol 17 to testosterone concentra tion ratios were Log 10 transformed and evaluated using a one way analysis of variance. Site Comparisons Sperm atogenesis and hormone concentrations were compared among non-agricultural males, and agricultural male, abnormal, and inters ex toads collected from different sites that varied in agricultural intensity. Agricultural inte nsity of each site was ranked from 1 to 5 based on the total % of agricult ural land within a 5.6 km2 area of each site, and distance to closest agricultural area (fu ll methodological explanation in Chapter 2). In brief, Belleglade (BG-97%, rank of 5) and Canal Point (CP-51.2%, rank of 4) were the most agricultural; Homestead (HS34%, rank of 3) was intermediate, and Welli ngton (WT),and Lake Worth (LW) were nonagricultural (suburban sites with no agriculture in the vicinity). Wellington was ranked 2 because it was close to agricultural areas (inten sive agriculture 14 km), and Lake Worth, ranked as 1, was farther from agriculture (intensive agriculture 37 km). Sex Comparisons Toads were grouped into different sex cat egories based on gross gonadal morphology and their occurrence in non-agricultural (WT, LW ) or agricultural (BG a nd CP) sites: 1) nonagricultural males (NAgM), 2) ag ricultural males (AgM), 3) abnormal agricultural males (A), 4) intersexed agricultural males (I), and 5) female s (for sex hormone concentration comparisons only). Toads defined as Males had normal Bi dders organs and lacked female tissue. Abnormal males had apparently normal testes but abnormal Bidders organs and no ovaries or oviducts, whereas intersex toads had testes al ong with ovaries or oviducts. Toads from the intermediate agricultural intensity site (HS) we re not included in sex gr oup analyses because it
89 was unclear whether they should be considered as non-agricultural ma les or as agricultural males (e.g. several HS toads were coll ected near a tree farm). The number of spermatogenic cells were compared among non-agricultural males (NAgM), 2) agricultural males (A gM), 3) abnormal agricultural males (A), 4) intersexed agricultural males (I). Females were not evaluated for spermatogenesis. Previous work found no differences among females across sites (see Chapter 2). Therefore, hormone concentrations among females we re statistically compared to determine if it was appropriate to pool females into one group that could be efficiently compared to the other sex groups. Females did not vary across sites in either hormone (Estradiol 17 : df=4,59.08; F=1.05, p=0.39; Testosterone: df=4, 69.26 F=1.67 p= 0.17). Therefore, females were pooled into a single group for among sex group hormone comparisons. The ratio of each hormone concentration (E to T) was evaluated by sex group, with females excluded, to test whether the balance between estradiol 17 and testosterone differed among non-agricultural males and agricultural toads. Results Spermatogenesis Site Comparisons Sperm atogenesis was reduced in toads living in agricultural areas. The number of primary spermatocytes (df=4, 91; F=8.59, p<0.001) was signi ficantly lower at Belle Glade and Canal Point (most agricultural sites) re lative to lake Worth (most suburban). Secondary spermatocytes (df=4,91; F=7.13, p<0.001) and spermatids (d f=4,91; F=8.90, p<0.001) were significantly reduced in toads from the Belle Glade and Canal Point (most agricultural sites) compared to the more suburban sites Lake Wort h, Wellington, and Homestead. Median sperm rank also varied across sites (df=4, Chisquare=14.978, p=0.005). Lake Worth and Wellington, non-agricultural
90 sites, had a higher percentage of individuals with median sperm ranks two and three whereas agricultural sites commonly had sp erm ranks of one (Figure 3). Spermatogenesis Sex Comparisons The num ber of primary spermatocytes (df=3,77, F=6.88, p<0.001), secondary spermatocytes (df=3, F=7.41, p<0.001), and sp ermatids (df=3, 77; F=9.28, p<0.001) were significantly higher in males from non-agricultural areas relative to abnormal, intersexed and male toads from agricultural sites (Figure 4). Median sperm rank also significantly varied by sex (df=3, Chi-square=9.739, p=0.021, Figure 5) with the lowest sperm rank (1) occurring more frequently in abnormal, intersexed and male toads from agricultural sites th an in male toads from non-agricultural sites. However, a similar per centage of non-ag ricultural males and intersex toads had the highest sperm ranks. Hormone Concentrations Site Comparisons Plasm a estradiol-17 (E2) among toads from different sites were not different (df=4, 11.58, F=1.63, p=0.233), whereas testosterone levels were significantly differe nt across sites (df=4, 9.36, F=5.50, p=0.015, Figure 6). Hormone Concentrations Sex Comparisons There were significan t differences in estradiol-17 (E2) concentrations among sex groups (df=4, 20.94; F=11.09, p<0.001), but these differences were driven entirely by the high E2 concentrations in females. Testosterone con centrations were signifi cantly different among sex groups (df=4, 160.4; F=17.59, p<0.001). Males from non-agricultural areas had the highest testosterone concentrations and females and intersexed frogs had the lowest (Figure 7B). The ratio of estradiol 17 to testosterone c oncentration was signifi cantly different among sex groups (df=3, 90; F=4.242, p<0.007). Intersex toads had the greatest estrad iol to testosterone ratio (Figure 8).
91 Discussion The underlying genetic and endocrine control of gonadal developm ent, growth, and function, and the roles of gonadally -derived hormones on secondary sexual characteristics, are highly conserved among vertebrates. Nevertheless, these physiological mechanisms can be altered by exogenous exposure to su b-lethal concentrations of endocrine disrupting chemicals (EDCs) causing impaired development, gr owth, sexual differentiation, and reproduction (Sanderson, 2006). For example, aromatase (CYP 19) is responsible for controlling the ratelimiting step in the conversi on of androgens to estrogens (e.g. testosterone to 17 estradiol), and its modulation by EDCs (e.g. tamoxifen and atrazine) interfer es with homeostasis and the function of sex steroid hormones resulting in feminization, demasculinization, and even sex reversal (Crain & Guillette, 1997). In Chapter two of this volume, I presented re sults that demonstrated that both primary (e.g. gonadal morphology) and secondary se xual traits of toads are altere d in agricultural areas where chemicals with endocrine disruptive potential are used. In addition, I observed that toads that had gonadal abnormalities also had feminized and de masculinized secondary sexual traits. One of the primary functions of gonads is to pr oduce hormones that control the expression of secondary sexual traits, thus my results suggested that testicular func tion (e.g. spermatogenesis and hormone production) was reduced in toads li ving in agricultural areas and in toads with gonadal morphological abnormalities. Estrogens were hypothesized to to increase and testosterone to decrease ( due to increased aromatase) in toads from agricultural sites and in toads with gonadal abnormalities generating the combination of feminized estrogen-dependent and demasculinized androgen-dependent traits. In this study, testicular f unction was reduced in toads liv ing in agricultural areas. Spermatogenic cell types were reduced in toads li ving in agricultural habitats relative to those
92 that did not. The reduction in testicular f unction was associated with abnormal gonadal morphology. For example, intersex toads had re duced spermatogenic cell numbers relative to normal males from non-agricultural areas. Th e physiological mechan ism(s) through which spermatogenesis is altered in to ads living in agricultural areas, or in those with intersex gonadal tissue is unknown. Production of primary and secondary spermatocytes is initiated by follicle stimulating hormone (FSH), whereas differentia tion of spermatids is controlled by androgens produced by Sertoli cells induced by lutenizing hormone (LH). Bo th FSH and LH are secreted by the anterior pituitary after stimulation from the hypothalamus (Sriraman et al 2004; Rastogi et al 2005; Minucci et al 2005). Therefore, the re duction of the number of spermatogenic cells suggests that the hypothalamus or pituitary could be altered in toads living in agricultural areas as has been reported in another study of frogs (Sower, Reed & Babbitt, 2000). Alternatively, the Sertoli cells could be altered such that they can not respond to pituitary hormones appropriately. Regardless of the mechanisms, reduced spermat ogenesis is likely asso ciated with reduced fertility and could have important implications for the long term pers istence affected populations. Each of these hypothesized mechanisms could be a result of organizational alterations that occurred because toads were exposed to EDCs during embryonic development (Guillette et al. 1995; Sower et al., 2000). Such embryonic exposures pe rmanently alters developing tissues such that they never function properly in adul ts. In addition, current body burdens and daily exposure to EDCs can exacerbate organizati onally derived abnormalities as well as induce activational pathologies. Intersexed individual s have experienced organizational alterations of gonadal morphology, and it is notworthy that the number of primary and secondary spermatocytes and sprematocytes did not differ between intersexed and normal males from agricultural areas; males and intersexed individuals from agricultural areas show the same
93 reduction in spermatogenesis relative to males fr om non-agricultural sites. This result suggests that current exposure could be reducing sper matogenesis independent of gonadal morphology. However, the developmental stage at which expos ures occur, as well as the concentration and length of time of exposure, and differences in the toxicant milieu can induce different abnormality phenotypes. Therefore, toads at agri cultural areas might have different degrees or severities of abnormalities. For example, re duced spermatogenesis could be a result of organizational abnormalities induced by larval exposures to pulsed sublethal concentrations of pesticides, whereas toads that were exposed for longer time periods, or to particular concentrations or milieus of pesticides had additional gross morphological abnormalities. The mechanisms through which these gross morphol ogical and functional abnormalities occur are unknown, and further research is required before any of these hypotheses can be explicitly evaluated. Although non-agricultural males had higher sperm numbers than individuals from agricultural sites, a greater percentage of inters exed individuals had sperm ranks of three than normal agricultural males. Although unexpected, ther e are plausible explanations for this result. First, intersexed individuals might not be reproductively active. If they do not display breeding behavior, breed as frequently, or attract as ma ny mates as normal males, then spermatozoa could accumulate relative to normal males. Second, es trogen increases the number of spermatogonia, the spermatogenic stem cells, in Rana esculenta (Minucci et al. 1997). Although circulating estradiol 17 levels were not significantly higher in intersexed individuals the ratio between circulating estradiol 17 and testosterone was significantly higher. In addition, estrogens from the ovarian tissue of intersexed individuals could act locally on th e testes and increase
94 spermatogonia there by increasing spermatoge nesis and eventually the production of spermatozoa (relative to agricultural males). Although female estrodiol 17 concentrations were significan tly greater than in all other sex groups, there were no differences across normal non-agricultural males and any of the agricultural toads. These patterns are contrary to expectations because toads from agricultural sites, and especially those with gonadal abnormalities, have altere d secondary sexual traits that are sex hormone dependent. However, the lack of significant differences between intersexed toads and the other sex groups is influenced by the inherent variation in estradiol 17 concentration with intersexed individuals. Thes e toads varied in stage of oogenesis, and thus were expected to va ry in estradiol 17 concentrations. In add ition, endogenous hormones are activated at extremely low concentrations, so small differences in estradiol 17 concentrations could induce large effects (Welshons et al., 2003). Furthermore, the ph ysiological responses that result from endogenous hormones are not simply a function of having enough of one hormone; they are also a function of the re lative quantities of different type s of hormones. Importantly, the ratio between estradiol 17 and testosterone concentrati ons was significantly greater in intersexed toads relative to all other sex groups, which c ould feminize and demasculinize secondary sexual traits. There are several ways in which secondary sexual traits could be modulated without changes in circulating hormone levels. I predic ted that circulating leve ls of estrogens would increase in agricultural toads and especially in intersexed individuals because atrazine (the herbicide used at the agricultural sites) increases aromatase expression and thus could lead to the feminized colorations patterns observed (see Figure 1). However, these color patterns could be caused by increased aromatase expression within the skin cells which would lead to the
95 feminized phenotype without requiring higher ci rculating estrogen con centrations (Thorton, 2002). Synthesis of circulating androgens occurs in the Leydig cell of the te stes, and is induced by lutenizing hormone. Circulating testosterone levels were lowe r in intersexed, abnormal, and normal male toads from agricu ltural sites relative to non-agricultural male s which suggest that androgen synthesis could be reduced generally, or that the toad s are clearing androgens more readily by liver biotransformation. Although test osterone concentrations were not different among sex groups within agricu ltural sites, these groups va ried in secondary sexual characteristics such as arm size and nuptial pad quantity (Chapter 2). However, these demasculinized traits could be induced through several mechanisms. The relative ratios between estradiol 17 and testosterone could be important in the activation of masculine traits. In addition, the expression of 5 reductase, the enzyme that convert s testosterone to the more potent androgen 5 dihydrotestosterone (DHT) could be lower. DHT influences secondary sexual traits and its concentrations could be altered across agricultural site s and in animals with gonadal abnormalities even when testosterone concentrati ons are not reduced. Binding proteins could also be modulated by EDCs, and if they become more numerous they will bind more testosterone or DHT and there will be less endogenous hormone fr ee to exert physiological effects. Therefore total hormone concentrations would be simila r among the sex groups in the agricultural areas, but the percentage that was free to exert masculin izing effects would be lower in the intersexed toads. Androgen receptor gene expression could also be reduced by EDCs which would make androgen responsive cells blind and unable to de tect circulating androgens. The particular mechanisms through which these abnormalities are occu rring need to be studied in more detail.
96 However, my results demonstrate that gonadal function is altered in animals living in agricultural areas. Living in or near contaminat ed areas and undergoing critical stages of development while exposed to endocrine-disrupting ch emicals is expected to negativ ely affect both individuals and populations (Hayes et al 2006 but see Glennemeier and Begnoc he 2002). Most EDCs studied to date have estrogenic or anti-a ndrogenic activity and can influe nce sex ratio, sexual maturation, gonadal morphology and function, and spermatogenesis, fertility, hormone levels (steroidogenesis, metabolism), secondary sexua l characteristics and reproductive behavior (Milnes et al. 2006). All of these characteristics are important for maintaining ecologically functioning populations, especially for taxa that are known to be declining globally, such as many amphibian species.
97 Figure 3-1 Examples and summary data of prim ary and secondary morphological traits of Bufo marinus collected from habitats that vary in the intensity of agriculture. a. Normal male (M) gonads have testes (T), a non-vitellogenic Bidders organ (BO), and no ovaries or oviducts. b. Abnormal male (A) gonads have testes (T) vitellogenic or oogenic Bidders organs (VBO), but no ovaries or oviducts, c. Intersexed gonads (I) have both testes (T) and ovari es (OO) or oviducts. Bidders organs and ovaries were distinguished by the density of the tissue. Bidders organs are thicker and more dense and ovaries are thin, sac-like, and convoluted. Ovaries were also completely separate from the Bidders organ. d. Photograph of a male from a non-agri cultural site; Solid coloration, large forelimb width, and dark mu ltiple nuptial pads are diagnositic. e. Percentage of individuals that were classified as male (b lack), abnormal (dark gray), and intersexed (light gray) at each site. Agricultural intensity increases along the Xaxis. f. Photograph of an intersexed toad. Similarly sized intersexed toads had more mottling, smaller forearm widths and fewer nuptial pads.
98 Figure 3-2 a. Cross section of a testis. The round lobule in the center of the photograph contains several groups of spermat ogenic cells. Each group is called a spermatocyst. Spermatocysts contain only one cell type either: primary spermatocytes, secondary spermatocytes, sprematids (round or elongate), and sperm. Sperm can also be loose which is indicative of reproductive readin ess. The largest round cells are primary
99 spermatocytes and include the more obvious groups of cells al ong the edge of the lobule (light purple cysts). The smaller r ound cells are secondary spermatocytes. No round spermatids are observed in these lobules but they are easily distinguished from secondary spermatocytes by their small size. Elongated spermatotids are light purple whereas sperm is dark purple to black. b. Mean number of sprematocysts of each cell type per lobule across sites of varyi ng agricultural intensities. Agricultural intensity increases from left to right. Letters a, b, and c represent significant differences among sites for primary sperma tocytes, e and f denote differences for secondary spermatocytes, and g and h denote differences among sites for spermatids. Bars are 95% confidence intervals. Figure 3-3 Percentage of males with a median sperm rank of 0-3 across sites. Agricultural intensity increases from left to right along the X-axis.
100 Figure 3-4 Mean number of sprematocysts of each cell type per lobule across sex groups including non-agricultural males (NAgM), agricultural males (AgM), agricultural abnormal toads (A), and agricu ltural intersexed toads (I). Letters a and b represent significant differences based on post hoc a Tukeys test among sex groups for primary spermatocytes, c and d denote differe nces for secondary spermatocytes, and e and f denote differences among sites for spermatids. Bars are 95% confidence intervals.
101 Figure 3-5 Percentage of toads with a median sperm rank of 0-3 by sex group including nonagricultural males (NAgM), agricultural males (AgM), agricultural abnormal toads (A), and agricultural intersexed toads (I).
102 Figure 3-6 Testosterone concentration by site. The mo st agricultural site is on the right. Bars are 95% confidence intervals.
103 Figure 3-7 Hormone concentr ations. a Estradiol 17 and b Testosterone concentration across sex groups. Non-agricultural males (NAgM) agricultural males (AgM), agricultural abnormal toads (A), and agricu ltural intersexed toads (I). Bars are 95% confidence intervals around the mean of the untransformed data.
104 Figure 3-8 Estradiol 17 to testosterone concentrati ons among sex groups including nonagricultural males (NAgM), agricultural males (AgM), agricultural abnormal toads (A), and agricultural intersexed toads (I). Letters represent significant differences among sex groups. Bars are 95% confidence inte rvals. Solid line indicates a ratio of one.
105 CHAPTER 4 GENE EXPRESSION PROFILE DIFFERENCES AMONG OVARIAN, TESTICULAR, AND INTERSE X GONADAL TISSUE OF WILD CAUGHT ADULT MARINE TOADS Introduction Over the last century, hum an-induced envir onmental changes have occurred on a global scale. Extensive urbanization, i ndustrial farming practices, and th e use of chemicals, such as pesticides and pharmaceuticals, have resulted in increased concentrations of environmental contaminants (Swarup & Patra, 2005). Many of these pollutants alter th e endocrine systems of wildlife and humans and induce reproductive ab normalities (Colborn & Clement, 1992). All classes of vertebrates and many invertebrates experience developmental and functional abnormalities of the reproductive system after exposure to endocrine disrupting chemicals (EDCs: (Edwards, Moore & Guillette 2006; Edwards & Myers, 2007; Milnes et al. 2006; Porte et al. 2006; Weltje & Schulte-Oehlm ann, 2007). In many cases, different chemicals induce similar reproductive abnormalities (phenotypes) across diverse vertebrate taxa (Milnes et al., 2006). Similarity in response is likely due to cons erved mechanisms of genetic and endocrine control of gonadal development, growth, and function among vertebrates. In other words, developmental and functional abnormalities observed in EDC-exposed animals could stem from disruption of conserved genetic networks. Id entification of gene networks disrupted by contaminant exposure can improve characterizati on of the physiological a nd genetic mechanisms involved in toxin-induced patholog ies. Monitoring gene expression profiles of exposed animals could permit the identification of chemical signatu res that can be used as biomarkers, and can be linked to specific mechanisms of toxicity as well as whole organism effects. In addition, this approach promotes a system wide understanding of the effects of pollutants and can be more
106 sensitive than morphological or developmenta l endpoints (Denslow, Garcia-Reyero & Barber, 2007; Helbing et al., 2007). DNA arrays are commonly used to rapidly id entify alteration in 100s to 1000s of mRNA transcripts and allow for the id entification of genes and gene networks that underlay phenotypic traits associated with contaminant exposure. Several laborat ory studies have discovered genes whose expression levels are signif icantly altered by exposure to sp ecific chemicals, leading to identification of novel mechanisms of toxicity and potential biomarkers of exposure (Denslow et al. 2007). Field studies have compared gene ex pression of organisms collected from polluted and reference environments demonstrating differe ntial expression of specific genes associated with polluted environments (Denslow et al., 2007). Field studies are important because gene expression is highly dependent on environmental context, and exposure to contaminants under controlled conditions likely induc es very different genetic re sponses relative to responses generated in a variable environm ent. Therefore, field studies provide information about the mechanisms through which natural populations are affected by toxins. Field studies are challenging, however, because many variables are not controlled (e.g., length of time of contaminant exposure), and the additional variati on can decrease statistical power and mask the effects of contaminants. Relating gene expr ession to specific phenotypes or pathologies (phenotypic anchoring) can help explain some of this inherent variation making comparisons more powerful, and help link alteration in gene expression changes to traditional toxicological endpoints (Denslow et al. 2007; Schimidt, 2003). Indeed, Dens low et al. (2007) recently argued that establishing the relationship between altered gene expression and traditional markers of toxicology (i.e., phenotypic traits ) is a critical step required to advance the study of aquatic toxicology. The emerging field of ecotoxicogenomics seeks to do this (Schimidt, 2003).
107 In previous studies, I showed that the in cidence and severity of gonadal abnormalities (indicative of endocri ne disruption) in Bufo marinus are associated with changes in land use patterns (chapter 1 and 2 of this volume). The frequency of gonadal pathologies in these toads was higher in areas characterized by intensive agricultural practices (primarily sugar cane farming) than in suburban areas. Toads in agricultural areas had a greate r incidence of intersex gonads and abnormal Bidders organs, feminized skin coloration and demasculinized forelimb morphology. Furthermore, spermatogenesis and test osterone levels were reduced in male and intersex toads living in agricultural habitats. In this study, I characteri zed gene expression profiles of gonadal tissues from normal female (only ovary tissue), normal male (only testis tissue), and in tersex (testis and ovary tissue) B. marinus using an anuran DNA array (Crump et al., 2002b). First, I characterize gene expression and identify genes that have corr elated expression patterns among the different gonadal tissues. Second, I identify potential bi omarkers of sex by determining specific genes that are most closely associated with each gonadal t ype. Third, I test for differences in the levels of gene expression of a subset of genes to identify those that are differentially expressed in each reproductive organ. This study is the first to investigate genome wide changes in gene expression associated with normal and intersex gonadal tissue in amphibians. Methods Gonadal tissues used in this study were coll ected during the summ er of 2005. Toads from 2 sites with low or intermediate agriculture (Wellington and Homestead) and 2 agricultural locations (Canal Point and Belle Glade) were collected and euthanized with an overdose of MS222 (0.3% tricaine methanesulfonate, pH 7) One gonad from each individual was placed in RNALater ( Ambion, Inc) for gene expression an alyses and the other gonad was preserved for histological description (reporte d in chapter 3). Gonadal tissues from 5 individuals of each
108 sexual condition (normal male, normal female and intersex) were hybridized to separate arrays (N = 15). Ovaries and testes were from toad s from Homestead and We llington, whereas intersex gonads were collected from Belle Glade, and Canal Point, FL, USA. Gene expression is dynamic and can respond to environmental variation by changing over a period of minutes to hours, handling and cap ture stress were minimized and similar among sites and sexes. Toads were captured, handl ed, and held (15min-2 hours) under similar conditions. In addition, gene e xpression data were carefully sc reened for high variation and genes that varied significantly within a sex condition (e.g. perhaps due to differences in handling) were excluded from further analyses (s ee statistics section in methods). This means that several potentially important genes associat ed with variation in gon adal function within a sex group were not analyzed in this study. For ex ample, each of the five females used in this study varied in stage of vitell ogenisis from nonvitellogenic (a nd regressed) to moderately vitellogenic. Therefore, a portion of the genes associated with vite llogenisis could have been left out of this study due to high, among female, varia tion in their expression levels. Although, these genes are not studied here, this data is available for future work, and should be investigated further. Isolation of RNA For intersex gonads, I separated and weighed the ovarian and testicular portions and isolated RNA separately so that RNA of each tissu e type could be quantified in future studies. For this paper, I was interteres ted in mRNA expressed in the entire gonad, so after RNA isolation (see below), I constituted the to tal RNA (prior to cDNA preparat ion) by mixing the appropriate ratio (determined by mass) of testicular and ov arian RNA extract. None of the ovaries of intersex individuals were vitellogenic. For th e non-intersex gonads, I weighed each tissue and extracted RNA as indicated below.
109 Total RNA was isolated using TRIzol reagent according to the manu facturers protocol (Invitrogen Canada Inc., Burli ngton, Ontario, Canada). Gonada l tissue was homogenized in 1 mL TRIzol reagent for each 100 mg of tissue wi th a 3 mm diameter tungsten-carbide bead in safe-lock Eppendorf 1.5 ml microcentrifuge tube s. For homogenization, a Retsch MM301 Mixer Mill (Fisher Scientific Ltd, Ottawa, ON) was used at 20 Hz for 6min intervals until all tissue was completely homogenized. Mixing chambers were rotated 180 degrees at each 6-min interval. After phase separation with chloroform and RN A precipitation with isopropanol, the isolated RNA was re-suspended in up to 40l diethyl pyr ocarbonate (DEPC)-treat ed RNase-free water and stored at -70C. cDNA preparation and labeling cDNA was prepared according to the manufacturers protocol (Invitrogen) except single stranded cDNA was purified using QIAquick PCR purification colu mns (Qiagen) prior to the double strand synthesis step. SM ART cDNA was radiolabled with [ -32P] dATP using a HexaLabel DNA labeling kit (Fermentas). The labeled cDNA was purified using QIAquick PCR purification columns (Qiagen). Radioactivity wa s quantified for each sample so that standard quantities of radioactive cDNA could be a dded to each array. Immediately prior to hybridization, the radiolabeled cDNA was heat denatured for 5 min at 95C and quickly cooled on ice for 5 min. Gene Expression Profiling The MAGEX cDNA array, which is a hete rologous frog m ulti-gene cDNA array containing gene probes origina ting from conserved protein-en coding sequence regions, was purchased from Viagen X Biotech Inc. (Victoria, BC, Canada), and is described in detail elsewhere (Crump, Lean & Trudeau, 2002a; Veldhoen et al., 2006). Briefly, a cDNA sequence for each gene target was spotted in duplicate at adjacent grid positions on a nylon membrane.
110 Approximately 90 percent of the 420 gene sequences on the array originated from Xenopus laevis, and the others were isolated from Rana catesbeiana Each membrane contained three intron controls to monitor for genomic DNA cont amination. Prehybridi zation, hybridization, and posthybridization washes were performed at 48C. Five replicate hybridizations were performed for each gonad type, and each gonad was collected from separate individua l. Testes were collected from males from Wellington (3) and Homestead (2). Ovaries were collected from Wellington (2), Homestead (2), and Belle Glade (1). Intersex tissue was collected from intersex toads from Belle Glade (2) and Canal Point (3). A single ovary was collect from Belle Glade an ag ricultural site because it was most similar to the other ovaries in its stage of oogenesis. No reproductive abnormalities have been identified in female Bufo marinus from Belle Glade. Hybridizatio ns were carried out in 20 ml of hybridization solution containing 4x SSC, 10% (w/v) dextran sulfate, 1.0% (w/v) SDS, and 0.5% (w/v) Blotto. Pre-warmed hybridization solutio n was added to each hybridization tube (35 mm internal diameter x 150 mm length; Amersham) containing an array membrane and allowed to pre-hybridize for 2 h. Radiolabeled cDNA samples we re then added to a fi nal concentration of 8 x 107 cpm/ml and allowed to hybridize overnight After hybridization, the membranes were rinsed briefly with 50 ml 2x SSC, washed twi ce with 50 ml 2x SSC/0.1% SDS for 15 min, once with 50 ml 0.1x SSC/1.0% SDS for 25 min, and rinsed with 50 ml 0.1x SSC. The arrays were placed on 3M filter paper (Rose Scientif ic Ltd, Edmonton, AB, Canada) soaked with ddH20 and wrapped with plastic wrap. Each processed membrane was exposed to a phosphorimager screen (Molecular Dynamics In c., Sunnyvale, CA, USA) for five days. Hybridization signals were collected using a Storm 820 Ge l and Blot Imaging System (Amersham) at 50 m resolution. The resulting im age data were converted to a standard 8-bit
111 TIFF file using Photoshop V5.0 (Adobe Systems In c, San Jose, CA, USA), and auto-leveled which darkens the blots increasing the signal inte nsity. Both non auto and auto-level images were prepared for analysis to account for signal saturation. Relative expression for each gene target was collected from the image data using ImaGene Version 5.6.1 (BioDiscovery Inc, El Segundo, CA, US A). Some genes were blotted more than twice so within array variation could be assess ed. Signal intensities for each gene and blank positions were determined from the median spot pixel intensities and corrected by subtracting the local median background pixel intensities. Signal intensities that were derived from areas of non-specific hybridization on the arrays were not included in the final analysis. A non-signal background was determined from the median inte nsity value plus one st andard deviation of blank positions across the auto-level data set. Signal intensities for gene positions exhibiting values below median background were adjusted to this value so that background variation was illuminated and would not be mistaken for trea tment variation. Saturated gene positions identified in auto-level data were replaced across all data sets by the corresponding values obtained in the non auto-level analysis. Both non autoand auto-leveled data for each of the 15 arrays (5 per sex group) were normalized using a geometric mean derived fr om the median signal intensities from the following genes: ribosomal pr otein L8, GAPDH, cytoplasmic beta actin, NM23/nucleoside diphosphate kinase, and ubiquitin. The choice of gene transcripts used for normalization was dictated by spot quality, consiste ncy, and passing a robust estimated standard deviation cut-off (see Statistics below).
112 Quality Control and Statistical Analyses Non-specific binding and radiation contam ination can occur on an array m aking it impossible to accurately detect the expression of genes in that area. Contamination is easily detected because it forms a splotch that does not match the grid within which the genes are spotted. A gene was also excluded from analyses if it was represented by fewer than 5 data points (at least one spot on each ar ray) or 3 arrays per gonad type (each having at least two spot intensities for that gene). Gene expression data obtained for each gonadal t ype were analyzed for consistency using a intra-class correlation wi th a 2-way random effects model (reliability analysis SPSS Version 12.0, Chicago, IL, USA). The correlation coeffici ent for each set of arrays was never less than 0.85 (Ovary -0.918, Tes tis -0.848, Intersex -0.922). As an additional quality measure, and to reduce the chance of maki ng a type I error, I excluded genes that showed high variance among arrays within a gonadal type. The variance cut off was based upon a robust estimated standard deviation for each gene among replicate arrays [(maximum value-minimum value)/2]. If the robust estimated standard de viation was greater than 2 (i.e., if the maximum expression for a gene was >4-times the minimum expression within a gonad type) the gene was excluded from the final list of genes considered in further analyses. For the remaining genes, I calculated the relati ve expression (i.e., fold-change) of ovary (or testis) to intersex (i.e., ovary/inters ex or testis /intersex) or ovary re lative to testis (ovary/testis). Only genes exhibiting positive or negative fold cha nge values greater than or equal to 1.5 for at least one comparison were used in any analysis. Traditionall y, array studies use a 2 fold threshold, but since the goal of this work was to identify as many differentially expressed genes as possible I relaxed this (to 1.5) to incr ease the total number of genes available for consideration. Others were considered to be equivalent across all sex groups and thus unimportant for this study. Pearson correlations between gene expression values in ovaries
113 versus intersex tissue, testis versus intersex tissue, and ovari es versus testis tissue were conducted in SPSS 15.0. Identifying correlated gene expres sion patterns-cluster a nalysis Fold change values were im ported into Cluster 3 (Eisen et al. 1998), log2 transformed, and average linkage clustering was used to produce a hierarchical cluster tr ee which was visualized in Treeview (Eisen et al. 1998). Genes that showed similar relative expression patterns (fold change) across the sex groups were identified. For example, genes that were similarly expressed between ovary and intersex tissue but were expressed at relatively lower levels in testis tissue clustered together. Biomarker identification-correspondence analysis Correspondence analysis was used to identify gene transcripts that had strong expression levels that were associated with a particul ar gonadal type. S ince the data sets were a combination of data generated from non auto-lev eled and auto-leveled data (which were on different scales), I could not directly perform CA on the signal intensity va lues. Instead, the fold changes of one sex relative to another were used (ovary/intersex, testis/intersex, and ovary/testis). To make increases and decrease s symmetrical around zero, the fold changes were natural log transformed. Thus, an increase (e.g., a doubling: ln (2) = 0.69) would be seen as equivalent in magnitude to a comparable decrease (e.g., a halving: ln (0.5) = -0.69). Test for differential gene expression-robust analyses All tests for differential gene expression were perform ed in the R statistical programming environment (R development core 2007). To look for significant differences in gene expression across sex groups, each of the 89 differentially expressed genes were analyzed in separate robust analyses of deviance tests. This analysis is part of a family of statistical techniques (Robust Statistics) for estimating the parameters of parame tric models while dealing with violations of
114 idealized assumptions (Huber 1981, Hampel et al 1986, Venables and Ripley 1999). For example, many types of data contain extreme valu es or outliers resulting in departures from normality or other parametric assumptions. Robust statistics emulate classical methods, but are not prone to estimation errors when data include outliers or exhibit other small departures from model assumptions. Multiple comparisons inflate the probability of type I errors. However, because the goal of this analysis was to genera te hypotheses about mechanisms and to identify candidate genes for further study, inferences ba sed on a Bonferoni corre cted alpha level of 0.0005 would rule out many potentially important genes. Thus, I present non-corrected p-values, but also note the genes whose expression patter ns are significantly di fferent after Bonferoni correction. Results General Characterization of Gene Profiles Differential gene expression across the different gonadal types is shown for a representative region of the M AGEX array (Figure 1). Of th e 420 genes on the array, 259 gene transcripts were detectable above background ex pression levels. Of the original 259 genes expressed, 89 m et all data quality requirements and displayed a fold-change in relative gene expression of greater than 1.5 or less than 0.67 in at least one se x comparison (Table 1). These genes varied in their function a nd included: 12 genes involved in cell growth control, 4 involved with chromatin structure, 11 in signal transduction, 29 with transcri ption, 5 with hormonal regulation, 5 with metabolism, 6 with transp ort/ and binding, 6 w ith apoptosis/protein processing, 7 involved with cell structure, 1 of unknown function, and 3 involved with protein translation. A high percentage of genes were down regulated in ovaries relative to intersex (50%) or testicular (46%) tissues (Figure 2). Twenty percent of the gene s in ovaries were up regulated
115 relative to testes (O/T), and 11% were up regulated relative to in tersex gonadal tissue (O/I). Thirty eight percent of the genes show similar expression levels between ovaries and intersex tissues (O/I) but were expressed at higher or lowe r levels in testis tissu e, whereas 32% of genes had similar levels of expression in testes a nd intersex gonads and were higher of lower in ovaries. In general, males seem to have roughl y equivalent percentage s of genes that are up regulated, equivalent, or down regulated relative to intersex gonads (Figure 2). Although gene expression among these tissues is strongly correl ated, overall gene expression between ovaries and intersex tissue were more hi ghly correlated (R= 0.92) than betw een testes and intersex tissue (R= 0.83) or ovaries and test es (R= 0.74, Figure 3). Correlated Gene Expression Patterns Nine gene clusters were identified, each with correlation coefficien ts (CC) > 0.90 (Figure 4). Gene cluster one (CC=0.94; Figure 4), includes m any genes th at are important in cell cycle regulation and are involved in both oogenesis and spermatogenesis (cyclin H and B2, nerve growth factor, inportan alpha 1a, suvivin). Ge ne cluster one (CC=0.94; Figure 4), includes many genes that are important in cell cycle regul ation and are involved in both oogenesis and spermatogenesis (cyclin H and B2, nerve growth factor, inportan al pha 1a, suvivin). This cluster in characterized by similar gene expression in ovaries and intersex gonads, whereas testes showed lower expression levels. Gene cluste r two (CC=0.93; Figure 4) includes genes that are typically thought to regulate development (retinoic acid converting enzyme, sonic hedgehog, and Pax-2) and function as transcription factors in ad ults. Altered regulation of these genes has been associated with development of cancers (Shang, 2006), and according to the cluster analysis both ovaries and testes had lower expr ession relative to intersex gona ds, whereas ovaries and testes had comparable expression levels. Gene cl uster three (CC=0.90; Fi gure 4) predominantly includes genes that are involved in cell growth and tran scription; ovaries and testes had lower
116 expression relative to intersex gonads, but ovaries also tended to have lower expression than testes (compared to cluster #2). Cluster four (CC=0.95; Figure 4) contains a variety of genes involved in transcription, and tr anslation. Ovaries had lower leve ls of gene expression than either intersex gonads or testes, whereas testes and intersex gonads had similar expression levels. Cluster five (CC=0.98; Figur e 4) also included genes associated with transcription and in this case ovaries had lower gene expression than in tersex gonads and testes, whereas testes had higher levels of gene expression than intersex gonads. Gene cluster si x (CC=0.95) also included several genes associated with transcription and was characterized by ovaries with variable expression levels (both higher a nd lower) relative to intersexes, but testes had higher expression levels relative to intersexes and ovaries. Clus ter seven (CC=0.96) included two genes expressed in ovaries at higher levels than intersexes or testes, whereas testes and intersex tissue were similar. Cluster eight (CC=0.98) included genes involved in cell growth and signal transduction that were expressed similarly be tween ovaries and testes and we re higher relative to intersex tissue. Cluster nine (CC=0.99) included two gene s one of which was relatively similar across all groups and the other tended to be expressed at hi gher levels in ovaries an d testes relative to intersex tissue and was lower in ovary compared with testis. Biomarker Identification The correspondence analysis identified 18 ge nes with expression pa tterns that were strongly associated with specific sexes and m ight serve as biomarkers of sex type (Figure 5, Table 2). This association coul d be due either to a specific gene being highly expressed or relatively absent in a specific gonad type. For example, if a specific gene was expressed in males and females but was very low in inters ex gonads the corres pondence analysis would identify its absence as being associated with in tersex condition. The expr ession patterns of four genes were associated with ovary tissue. Thes e included one gene involved in cell growth, two
117 controlling chromatin structure, a nd one involved in gene transcrip tion (Table 2). Five different genes showed a strong association with testic ular tissue. These in cluded three signal transduction genes, one associated with transcri ption, and another involve d in ion transport. Three genes were identified to be closely associated with intersex tissue. One of these genes is associated with cell growth, anot her with signal transduction, and the third with transcription. Three other groups were identified that were associated with two sex groups, but not the third including: one metabolic gene that was si milar in testis and intersex tissue, two signal transduction genes and one transcription gene were associated with ovary and testicular tissues, and two cell growth genes associated with ovary and intersex tissue (Table 2). Expression Level Comparisons Twenty five of the 89 genes differed significantly in their levels of gene expression (alpha = 0.05), and eight differed signifi cantly after Bonferronni corr ection (alpha=0.0005, Table 3). Of the eight genes that differed after Bonferronni correction f ive (adenosine A1 receptor a1R gene, adenosine A1 receptor a1R gene, Na K tr ansporting ATPase beta subunit, E2 ubiquitin conjugating enzyme Ubc9, and co llagenase 4 precursor) were si gnificantly different between testicular and intersex tissue (alpha=0.0005), and not between ovarian and intersex (lowest alpha=0.131). Two (mitochondrial cytochrome c oxi dase subunit 1 and elongation factor 1 alpha chain) were similar between testicular and inters ex tissue, and differed in the ovary (one higher one lower). One gene (cyclin H) was significan tly different from intersex tissue and ovarian tissue which expressed higher levels and sifferent from tes ticular tissue expressed which expressed lower quantities. None of these gene s were identified as biomarkers of gonadal type. Two genes, osteogenic protein-1 homolog pr ecursor (cell growth), and retinoic acid converting enzyme (hormonal regulation), were significantly (alpha=0.05 ) up regulated in both
118 males and females relative to intersexes (Table 3), whereas sonic hedgeh og (transcription factor) were down regulated in both males and females relative to intersexes. Twelve genes differed significantly in their relative levels of expression in females relative to intersexes but not in males relative to in tersexes. However, only two of the 12 were significantly up regulated in females relative to intersex gonads: elongation factor 1 alpha chain (protein translation) and RAG2 (chromatin structure). The remaining 9 genes were all down regulated including 4 genes involved with si gnal transduction (colorectal cancer tumor suppressor, casein kinase 1 epsilon, activated protein kinase C receptor, calcineurin A), two transcription factors (Pitx1, and liver helicase), thr ee genes involved in metabolism (ornithine decarboxylase, aldolase B, mitochondrial cytochro me c oxidase subunit 1), and one structural gene (alpha 1 collagen type II). Nine genes differed significantl y in males relative to intersex es but did not differ between females and intersexes. Two of these, E2 ubi quitin conjugating enzyme (involved in apoptosis and protein processing) and wee 1A kinase (involved in cell grow th control), were significantly down regulated in males (Table 3). The remaining 7 genes were all up regulated in males relative to females and intersexes, including tw o genes involved with signal transduction (phospholipase C gamma 1a, adenosine A1 receptor), two genes involved with transport (sodium phosphate cotransporter, Na K transporting ATPa se beta subunit), one involved in hormonal regulation (thyrotropin releasing hor mone precursor), one involved with cell growth (cyclin B2), and one in apoptosis and protein pr ocessing (collagens 4 precursor). Discussion Correlated Gene Expression The cluster analys is identified several cluste rs that contained genes associated with transcription that could be especially important since they control expression of other genes.
119 This group of genes was the most highly repres ented on the array (91 transcription related genes). The fact that these gene s did not cluster together sugges ts that expression of different transcription related genes are more (or less) important for each gonadal. For example, both clusters three and four contained genes associated with transcrip tion, but only in cluster three did intersex gonads have higher expression levels than ovaries and testes. In vestigating these genes, and the genes they control, could help iden tify mechanisms involved in the morphology and function of this intersex pathology. Comparing different clusters also helps to identify the mechanisms involved with the gonadal pathologies and normal gonadal function. Cl usters one and six were similar in that ovary and intersex tissues had similar expression levels, and testicular tissue had either lower (#1) or higher (#6) expre ssion. These genes could be important for understanding the mechanisms through which intersex tissues are feminized (act like ovary tissue), and for identifying genes and the genetic mechanisms that distinguish ovaries from testes. In clusters two and eight, ovarian and testicular tissue were relatively similar whereas intersex tissue had either higher (#2) or lowe r (#8) expression levels (re lative to both ovaries a nd testes). Therefore, these genes distinguish intersex individuals from females and males, and identify possible genetic mechanisms involved in the structure and function of the intersex condition. Cluster four is characterized by genes that ar e similar in testes and intersex tissues, but are different in ovaries; thus they represent a suite of genes th at could be important for ovary function. Biomarkers of Sex Three of the four genes associated w ith ova ries in the correspondenc e analysis are known to be expressed in vertebrate ovaries (Wee 1-A kinase, Cytosi ne 5methyltransferase, and deoxyribonuclease gamma). In Xenopus leavis Wee 1-A kinase incr eases phosphorylation of cdc2 (a cyclin-dependent kinase), and inhibits cells from undergoing mitosis (Mueller, Coleman
120 & Dunphy, 1995). In mammals, Wee 1 A kinase, cdc 2 and other cell cycle control proteins are involved in controlling the meiotic arrest of oocytes (El Touny & Banerjee, 2006; Park et al., 2004). Cytosine 5methyltransferase (down) regulat es gene expression and embryogenesis and is found in Xenopus leavis stage III oocytes. Concentrations increase as the oocytes mature and in late stage oocytes, DNA methyltransferase is translocated into the nucleus, where it down regulates gene expression and f acilitates the formation of chromatin and rapid DNA replication. This gene expression down regulation occurs on a genomic scale and is maintained through fertilization and the early stag es of embryogenesis. Reduction s in DNA methyltransferase leads to premature gene transcription and body plan defects and thus has important implications for embryonic development (Kimura, Suetake & Tajima, 1999; Stancheva & Meehan, 2000). Less is known about the function of deoxyri bonuclease gamma in gonadal tissues, but it has been localized in both ovary and tetes and is involved in degrading DNA during apoptosis (Shiokawa et al. 2006). Early response -1 (ER-1) has not previously been localized in the ovary but similarly to Cytosine 5methyltransferase it represses gene expression (Ding, Gillespie & Paterno, 2003). Indeed, ovaries tended to have a higher percentage of genes that were down regulated relative to intersex a nd testicular tissues. DNA methyltransferase and ER-1 are likely involved in this down regulation (Ding et al., 2003). All of the genes associated with ovary tissu e in the correspondence analysis clustered together in cluster #1, and intersex gonads disp layed similar gene expres sion levels to females for each of these genes. Therefore not only do intersex gonads share a common pattern of gene expression with females for a random suite of ge nes, they similarly expr ess the genes that are most closely associated with ovary tissue, or define femaleness.
121 Less is known about the genes a ssociated with testicular tiss ue in this study. There is much debate in the literature a bout the role of SERCA-1 in sper matogenesis and whether it is important for inducing the acrosome reacti on required during fertilization (Harper et al. 2005; Jimenez-Gonzalez et al. 2006; Lawson et al., 2007; Triphan et al. 2007). SERCA-1 has been localized in sperm and sp ermatozoa (Jimenez-Gonzalez et al. 2006; Lawson et al., 2007), and is known to regulate calcium concen trations and calcium signaling which are essential for cell homeostasis (Chami et al., 2001). Neuromedin is expr essed at high levels in Xenopus laevis ovaries (Wechselberger, Kreil & Richter, 1992); however, in this study it was highly expressed in Bufo marinus testes. Indeed, neuromedin is more hi ghly expressed in rat testes relative to ovaries, but there is no detailed information on its function in either of th ese reproductive tissues (Kilgore et al. 1993; Rucinski et al. 2007). Despite the lack of in formation on the role of these genes (their protein products) on gonadal function, the pattern de rived from the array data are interesting as they identify genes that are expressed in a sexually dimorp hic pattern in the gonads and could be important for testic ular function. Expression levels of four out of the five genes that were associated with maleness, or testicul ar tissue, were more similar between ovaries and intersex gonads than between te stes and intersex tissue sugges ting that intersex tissue is demasculinized. All three of the genes associated with the intersex conditi on modulate cell pr oliferation. Cyclin A2 has been localized in ovarian granulos a cells, is required for th e S phase and G2 to M phase transitions in the cell cycle, and plays an important role in regulating both meiosis and mitosis of female germ cells (Persson et al., 2005). However, cyclin A2 has also been localized in spermatogonia and regulates mitosis in male germ cells (Diederichs et al. 2005; KajiuraKobayashi, Kobayashi & Nagahama, 2005; Ravn ik & Wolgemuth, 1996). Cyclin A2 is up
122 regulated in intersex gonads relative to both ovaries and testes, and likely increases cell proliferation of male and female germ cells. Pax 2 was also up regulat ed in intersex gonads relative to ovaries and testes, and its expression is also linked to increased cell proliferation. Pax 2 is a transcription factor that is typically up regulated during development and then is switched off in adult life via increased me thylation (Shang, 2006). If methyl ation of PAX 2 is disturbed, it becomes upregulated and leads to hyperplasia and malignancies, and thus is considered an oncogene (Shang, 2006). The up regulation of PAX 2 suggests that intersex individuals could have altered DNA methylation patter ns. The third gene that is associated with intersex gonads, casein kinase 1-epsilon, phosphoryl ates many different substrates and modulates a variety of processes such as cell differen tiation, proliferation, chromoso me segregation and circadian rhythms. Alterations in casein ki nase expression has been linked to neurodegenerative diseases and cancer (Knippschild et al., 2005). The ovaries of the intersex toads used in th is study were non-vitello genic whereas two of the female ovaries were undergoi ng early stages of vitellogenisi s. Despite these differences, ovary and intersex gonadal tissue gene expressi on was similar. The strong correlation between ovarian and intersex tissue gene expression (Figure x) and the f act that intersex toads express genes associated with ovaries similarly to fe males and do not express genes associated with testes similarly to males, suggests that intersex go nads are more ovarian than testicular. Intersex ovaries were relatively small, and the average ovar y to testis mass ratio in intersex gonads was 0.33. Although the same amount of radiolabeled cDNA was added to each array, the intersex cDNA was a mixture of ovary and testicular cDNA whereas the fema le ovary was not. Thus, the similarity in gene expression betw een female and intersex gonads o ccurs in spite of the fact that the portion of intersex ovary was smaller than th e female ovary added to the array. Therefore,
123 the ovary portion of the intersex gonad could be expressing higher levels of genes relative to the female ovary, or that the testicular portion is also expressing these genes. Studying the genes that are expressed differentially in the ovarian and testicular portions of in tersex tissues can help identify underlying functional differences in these tissues relative to normal ovarian or testicular tissue. However, no study has analyzed each portion of intersex tissue separately. Endocrine signaling pathways and the mech anisms underlying reproductive development are relatively conserved among vertebrates sugges ting that specific suites of genes might be associated with normal gonadal function and fertilit y. Recent studies have identified genes or groups of genes that are important for fertility in mice and humans. Rockett et al (2004) found that 40 genes were differentially expressed between testicular tissue of fertile and infertile mouse lines. These mouse lines included animals that do not develop spermatogonia, or experience arrested spermatogenesis at the spermatocyte, or spermatid stage. In the same study, 19 genes were expressed differently between samples co llected from humans with normal and abnormal spermatogenesis (Rockett et al. 2004). Comparisons of the m ouse and human gene expression profiles lead to the hypothesis th at differential expression of ge nes associated with cell cycle control, heat shock proteins (hsp), and proteo lytic proteins are mechan istically linked and are associated with altered spermatogenesi s and infertility in mammals (Rockett et al. 2004). Indeed, similar types of genes were differentially expressed in this study and could be associated with differences in fertility be tween normal and abnormal toads. Altered expression patterns of these or similar networks of genes could under lie the mechanisms of toxicity of EDCs and explain similarities in observed pathologies across taxonomic groups. The use of cDNA arrays has advanced the field of ecotoxicology by providing a way to assess the effects of toxicants on multiple gene tr anscripts and biochemical pathways in a single
124 study (Denslow et al., 2007). Arrays contribute to th e development of a system wide understanding of toxicity by providing informa tion on known and unknown pathways of toxicity and by linking those pathways to phenotypic changes. However, arrays have limitations because gene expression varies extensively with several inherent biological processes including age, sex, and genetic polymorphisms, as well as with exte rnal phenomena such as temperature, day length, and season (Denslow et al. 2007). Arrays describe gene expre ssion profiles for organisms collected at a single point in time under a specific suite of e nvironmental conditions. Biological pathways could vary in sensitivity to contaminan ts across seasons, sexes, or genetically distinct populations. In addition, cDNA arrays do not provide a truly qua ntitative measure of gene expression, but describe relative changes (Denslow et al., 2007). To understand the effects of endocrine disrup ting chemicals, multiple endpoints must be measured. Previous work in this volume has show n that the occurrence and severity of intersex gonads is higher in toads collected from agricu ltural sites relative to those found in suburban sites. However, the mechanisms through which these abnormalities occur have not been studied and much work is needed linking gene expres sion, morphology and physiological performance. This study has documented expression profiles associated with phenotypically normal ovarian and testicular tissue and compared those profiles to those of pathol ogical intersex tissue. This approach has identified many different genes an d clusters of genes with correlated expression patterns expressed across these gonad types. I have identified specific genes associated with each gonadal type that should be studied furthe r to determine whether they are dependable biomarkers for each sex group. Future studies will continue to investigate these biomarkers of sex over a larger sample size of individuals us ing QPCR so that indivi dual variation can be
125 investigated and quantified. This work shows that intersex Bufo marinus living in agricultural areas in South Florida express gene s in a pattern that is more similar to females than to males. .
126Table 4-1 Fold change differences. Ovary and intersexed tissue (Ovary/Intersex), te stis and intersexed tissue (Ttestis/Interse x), and ovary and testis (Ovary/Tes tis). Cluster numbers refer to clusters from the hierarchical cluster analysis identified in Figure 4-4. Functional codes: 1) cell growth control, 2) chromatin st ructure, 3) signal transduction, 4) transcription, 5) hormonal re gulation, 6) metabolism, 7) transport/ and bi nding, 8) apoptosis/pr otein processing, 9) structural, 10) unknown function, 11) protein translation. Female Male Female Gene GenBank Cluster Functional Intersex Intersex` Male alpha skeletal actin X03470 1 9 0.02 -0.55 0.57 c-Jun 5650725 1 4 -0.11 -0.77 0.67 cyclin B2 AB005253 1 1 -0.57 -1.53 0.97 cyclin H U20505 1 1 -0.42 -1.34 0.91 cytosine-5-methyltransferase D78638 1 2 0.03 -0.91 0.94 deoxyribonuclease gamma AF059612 1 2 0.08 -0.74 0.82 E2 ubiquitin conjugating enzyme (Ubc9) U88561 1 8 -0.08 -0.68 0.60 ER1 AF015454 1 4 0.27 -0.95 1.23 importin alpha 1a L36339 1 7 -0.07 -0.74 0.67 nerve growth factor X55716 1 1 -0.15 -0.72 0.57 survivin AW198821 1 8 -0.30 -0.84 0.54 wee 1A kinase U13962 1 1 0.10 -0.63 0.73 adult beta-globin J00978 2 7 -0.83 -0.94 0.11 aldolase B AB030822 2 6 -0.72 -0.88 0.16 cdc2 AB005254 2 1 -0.45 -0.92 0.47 cyclin E L23857 2 1 -0.44 -0.43 -0.01 GSK-3 binding protein AF062738 2 3 -0.59 -0.65 0.07 katanin p60 AF177942 2 9 -0.32 -0.55 0.23 myelin proteolipid protein Z19522 2 9 -0.47 -0.47 0.00 Pax-2 AF179300 2 4 -1.17 -1.44 0.27 PCNA M34080 2 2 -0.28 -0.52 0.23 retinoic acid converting enzyme AF057566 2 5 -0.53 -0.78 0.25 sonic hedgehog L39213 2 4 -0.60 -1.15 0.55
127Table 4-1 (continued) Gene GenBank Cluster Functional Female Male Female Intersex Intersex` Male AP-2 M59455 3 4 -0.73 -0.15 -0.57 B-50/GAP 43 X87582 3 1 -0.57 -0.36 -0.21 casein kinase 1-epsilon AF183394 3 3 -0.99 -0.67 -0.33 Cu/Zn SOD X16585 3 6 -0.60 -0.34 -0.26 cyclin A2 X85746 3 1 -1.10 -0.97 -0.13 ets-2a M81683 3 4 -0.67 -0.46 -0.21 fos-related antigen-2 U37374 3 4 -0.79 -0.18 -0.60 iodothyronine 5'-deiodinase III L28111 3 5 -0.51 -0.16 -0.35 MAP kinase phosphatase X83742 3 3 -0.80 -0.64 -0.17 MCM2 U44047 3 1 -0.53 -0.18 -0.35 MCM5/CDC46 U44048 3 1 -0.58 -0.40 -0.17 Myf-5 X56738 3 4 -0.91 -0.65 -0.26 NEDD 8 Not Submitted 3 8 -0.77 -0.43 -0.34 ornithine decarboxylase X56316 3 6 -0.70 -0.56 -0.14 p54 mRNA binding protein M80257 3 4 -0.72 -0.33 -0.39 TBP-binding repressor Dr1 AB003582 3 4 -0.61 -0.37 -0.24 TFIID subunit D50054 3 4 -0.77 -0.47 -0.29 Xefiltin U63711 3 9 -0.86 -0.59 -0.27 activated protein kinase C receptor AF105259 4 3 -0.31 0.15 -0.46 alpha-1 collagen type II M63595 4 9 -1.05 -0.07 -0.98 ATP-dependent RNA helicase X57328 5 4 -0.20 0.24 -0.44 beta tubulin L06232 5 9 -0.22 0.22 -0.44 brain factor 1 AF101387 4 4 -0.46 0.28 -0.74 calbindin D28K U76636 4 7 -0.71 -0.08 -0.63 cellular retinoic acid bindi ng protein S74933 5 5 -0.38 0.44 -0.82 c-myc I X14806 5 4 -0.17 0.38 -0.55 colorectal cancer tumor suppressor U10986 4 3 -0.91 0.57 -1.47 DLL3 Distal-less 3 L09729 4 4 -0.41 0.02 -0.43 DLL4 Distal-less 4 L09728 4 4 -0.59 0.30 -0.89 elongation factor-1 alpha chain M25504 4 11 -0.91 0.09 -1.00
128Table 4-1 (continued) Gene GenBank Cluster Functional Female Male Female Intersex Intersex` Male hsp30B X02512 5 7 -0.47 0.59 -1.06 liver helicase AF302423 4 4 -1.65 -0.05 -1.59 MAM domain protein U37376 5 10 -0.21 0.20 -0.42 mitochondrial cytochrome c oxidase subunit 1 M10217 4 6 -3.30 0.08 -3.39 myelin basic protein AB000736 4 9 -0.71 -0.03 -0.68 nervous system-specific RNA binding protein M34894 5 4 -0.64 0.87 -1.51 neuromedin L01530 5 3 -0.32 0.92 -1.24 nuclear factor I-C1 L43149 5 4 -0.35 0.53 -0.87 Oct-1 X17190 4 4 -0.46 0.12 -0.58 Pitx-1 AF155206 4 4 -0.71 0.27 -0.98 protein S10 S57432 4 11 -1.00 0.07 -1.06 ribosomal protein L32 X55030 4 11 -0.38 0.08 -0.46 thyroid-stimulating hormone beta L07618 5 5 -0.36 0.46 -0.82 vitellogenin 65210 4 6 -0.44 0.26 -0.70 adenosine A1 receptor (a1R gene) AJ249842 6 3 0.00 0.64 -0.64 CCCH zinc finger protein C3H-3 AF061982 6 4 -0.05 0.63 -0.68 circadian CLOCK AF203107 6 4 0.22 0.84 -0.62 collagenase-4 precursor L76275 6 8 0.00 0.56 -0.56 Fez AF195021 6 4 0.16 0.84 -0.68 metalloproteinase (XMMP) U82541 6 8 0.23 1.03 -0.80 Na+/K+ transporting ATPase beta subunit M37788 6 7 0.00 0.74 -0.74 PKC-related kinase 2 AF027183 6 3 0.22 0.62 -0.40 SERCa1 fast skeletal muscle X63009 6 3 0.68 1.74 -1.05 sodium phosphate cotransporter L78835 6 7 -0.23 1.49 -1.72 thyrotropin-releasing hormone precursor X64056 6 5 0.37 1.29 -0.92 Xsal-1 L46583 6 4 0.00 0.40 -0.40
129Table 4-1 (continued) Gene GenBank Cluster Functional Female Male Female Intersex Intersex` Male recombination activation (RAG-2) L19325 7 2 0.95 0.00 0.95 Tcf-3 co-repressor CtBP AF152006 7 4 0.92 0.28 0.64 BMP1b Y09660 8 8 0.74 0.61 0.13 calcineurin A AF019569 8 3 1.77 1.22 0.55 DNA binding protein E12 X66959 8 4 0.78 0.56 0.22 L-myc L11363 9 4 0.31 0.45 -0.14 Mad1 L77888 8 1 0.45 0.44 0.01 Max2/Max4 L09738 8 4 1.62 1.12 0.50 osteogenic protein-1 homolog precursor U40034 8 1 0.75 0.69 0.06 phospholipase C-gamma-1a AF090111 9 3 1.28 1.68 -0.39
130 Table 4-2 Genes identified in correspondence analysis as being associated (or disassociated) with a specific reproductive organ type. Functi on codes: 1) cell growth control, 2) chromatin structure, 3) signal transducti on, 4) transcription, 5) hormonal regulation, 6) metabolism, 7) transp ort/ and binding, 8) apopto sis/protein processing, 9) structural, 10) unknown functi on, 11) protein translation. Gene Code Functional code Associated with ovary tissue FJ7 wee 1-A kinase 1 FM55 DNAse gamma 2 FM62 cytosine 5-methyltransferase 2 FN44 ER-1 4 Associated with teticular tissue FM24 SERCa1 fast skeletal muscle 3 FI9 neuromedin 3 FB2 colorectal cancer tumor suppressor 3 FH20 nervous system-specific RNA binding protein 4 FH15 sodium phosphate cotransporter 7 Associated with intersex tissu: FJ30 cyclin A2 1 FJ20 casein kinase 1-epsilon 3 FD9 Pax-2 4 Low in ovary but equal in test icular and intersexed tissue FM40 mitochondrial cytochrome c oxidase subunit 1 6 Low in intersex, but equal in testicular and ovary tissue FM34 phospholipase C-gamma-1a 3 FN17 calcineurin A 3 FJ8 Max2/Max4 4 Low in testis, but equal in ovary and intersexed tissue FJ36 cyclin H 1 FJ38 cyclin B2 1
131Table 4-3 Gene expression level comparis ons. Estimates of levels of expre ssion, standard errors and pvalues. Gene Female SE F vs. I Pr> t Intersex SE M vs. I Pr>t Male SE Female and different from intersex osteogenic protein 1 homolog precursor 5.2E+03 1.2E+03 0.033 + 2.7E+03 2.0E+02 0.042 + 6.0E+03 1.6E+03 *cyclin H 3.3E+03 8.9E+02 0.015 4.6E+03 4.2E+02 <0.001 1.2E+03 8.4E+02 sonic hedgehog 1.0E+04 1.2E+04 0.044 2.4E+04 5.7E+03 0.013 6.3E+03 1.2E+04 retinoic acid converting enzyme 6.8E+03 3.6E+03 0.017 + 1.2E+04 1.7E+03 0.002 + 4.9E+03 3.5E+03 Female and intersex significantly different recombination activation RAG 2 2.8E+03 NA <0.001 + 1.1E+03 NA 0.295 1.1E+03 1.3E+00 colorectal cancer tumor suppressor 1.2E+03 1.5E+03 0.045 2.9E+03 7.4E+02 0.454 4.5E+03 2.8E+03 casein kinase 1 epsilon 1.3E+04 9.8E+03 0.028 2.7E+04 4.0E+03 0.065 1.6E+04 9.4E+03 activated protein kinase C receptor 3.1E+04 1.4E+04 0.033 5.1E+04 5.7E+03 0.674 5.6E+04 1.7E+04 calcineurin A 6.6E+03 1.4E+03 0.001 + 1.4E+03 2.3E+02 0.726 2.9E+03 4.4E+03 Pitx 1 1.3E+04 1.1E+04 0.043 2.6E+04 5.6E+03 0.59 3.2E+04 1.6E+04 liver helicase 1.7E+03 3.0E+03 0.007 6.5E+03 1.5E+03 0.933 6.4E+03 3.2E+03 ornithine decarboxylase 1.5E+04 7.9E+03 0.01 2.7E+04 3.8E+03 0.441 2.0E+04 1.2E+04 aldolase B 1.1E+04 5.7E+03 0.001 2.3E+04 2.9E+03 0.142 1.3E+04 9.6E+03
132Table 4-3 (continued) Gene Female SE F vs. I Pr> t Intersex SE M vs. I Pr>t Male SE *mitochondrial cytochrome c oxidase subunit 1 1.6E+03 6.6E+03 <0.001 4.2E+04 3.3E+03 0.391 4.5E+04 7.1E+03 alpha 1 collagen type II 2.0E+03 1.9E+03 0.03 4.5E+03 8.8E+02 0.83 5.0E+03 2.8E+03 *elongation factor 1 alpha chain 7.4E+02 3.0E+02 <0.001 + 1.6E+03 1.4E+02 0.509 1.7E+03 4.2E+02 Male and intersex significantly different wee 1A kinase 9.0E+03 4.2E+03 0.954 9.1E+03 1.8E+03 0.022 3.9E+03 3.8E+03 cyclin B2 1.8E+04 1.2E+04 0.098 3.0E+04 6.0E+03 0.005 + 7.5E+03 1.2E+04 *adenosine A1 receptor a1R gene 1.1E+03 7.3E+01 0.823 1.1E+03 NA <0.001 + 2.6E+03 <0.001 phospholipase C gamma 1a 3.9E+03 1.5E+03 0.076 1.7E+03 3.2E+02 0.012 + 5.9E+03 1.7E+03 thyrotropin releasing hormone precu rsor 4.0E+03 2.0E+03 0.408 2.9E+03 7.5E+02 0.036 + 8.3E+03 3.0E+03 *sodium phosphate cotransporter 1.2E+03 8.0E+02 0.287 1.7E+03 4.0E+02 <0.001 + 3.9E+03 8.2E+02 Na K transporting ATPase beta subunit 1.1E+03 2.0E+02 0.789 1.1E+03 1.0E+02 <0.001 + 3.0E+03 2.0E+02 *E2 ubiquitin conjugating enzyme Ubc9 1.2E+03 2.2E+02 0.131 1.5E+03 8.8E+01 <0.001 7.1E+02 2.0E+02 *collagenase 4 precursor 1.1E+03 NA 0.349 1.1E+03 NA <0.001 + 4.7E+03 NA Not significantly different nerve growth factor 7.0E+03 7.0E+03 0.306 1.1E+04 3.3E+03 0.092 4.6E+03 6.8E+03 Mad1 2.4E+03 1.6E+03 0.896 2.5E+03 7.2E+02 0.45 3.4E+03 1.9E+03 cyclin A2 1.1E+03 3.4E+02 0.271 1.3E+03 1.7E+02 0.473 1.2E+03 3.5E+02
133Table 4-3 (continued) Gene (not significantly different) Female SE F vs. I Pr> t Intersex SE M vs. I Pr>t Male SE cyclin E 1.1E+03 1.2E+03 0.387 1.7E+03 6.2E+02 0.537 1.2E+03 1.3E+03 cdc2 1.3E+04 1.5E+04 0.832 1.5E+04 4.5E+03 0.102 7.0E+03 9.0E+03 MCM5 CDC46 1.0E+04 1.1E+04 0.433 1.4E+04 5.3E+03 0.526 1.1E+04 1.1E+04 MCM2 4.5E+03 1.7E+03 0.156 6.2E+03 5.8E+02 0.449 5.6E+03 1.4E+03 B 50 GAP 43 1.8E+04 1.3E+04 0.083 3.0E+04 6.1E+03 0.34 2.2E+04 1.4E+04 PCNA 2.2E+04 1.3E+04 0.739 2.5E+04 5.4E+03 0.183 1.7E+04 1.1E+04 deoxyribonuclease gamma 3.7E+03 3.0E+03 0.835 4.0E+03 1.3E+03 0.136 1.8E+03 2.7E+03 cytosine 5 methyltransferase 1.2E+04 6.4E+03 0.856 1.1E+04 2.4E+03 0.057 5.8E+03 5.1E+03 neuromedin 1.8E+03 7.5E+02 0.634 2.0E+03 3.1E+02 0.317 3.8E+03 2.1E+03 MAP kinase phosphatase 1.6E+04 1.0E+04 0.056 2.8E+04 4.5E+03 0.143 1.7E+04 1.1E+04 PKC related kinase 2 1.5E+03 8.2E+02 0.653 1.3E+03 4.4E+02 0.109 2.4E+03 1.1E+03 GSK 3 binding protein 1.5E+04 1.3E+04 0.104 2.7E+04 6.1E+03 0.152 1.6E+04 1.3E+04 SERCa1 fast skeletal muscle 3.6E+03 1.6E+03 0.188 1.9E+03 4.0E+02 0.525 5.4E+03 5.7E+03 ets 2a 2.6E+03 3.6E+03 0.337 4.5E+03 1.8E+03 0.478 3.0E+03 3.8E+03 c myc I 1.3E+04 8.4E+03 0.455 1.7E+04 4.0E+03 0.998 1.7E+04 1.1E+04 L myc 5.2E+03 2.9E+03 0.995 5.2E+03 1.3E+03 0.854 5.6E+03 3.7E+03 circadian CLOCK 2.2E+03 9.7E+02 0.79 2.0E+03 4.0E+02 0.905 2.3E+03 2.5E+03 c Jun 3.1E+04 2.4E+04 0.814 3.4E+04 1.1E+04 0.131 1.7E+04 2.1E+04 nuclear factor I C1 2.0E+04 2.1E+04 0.214 3.4E+04 1.0E+04 0.537 4.2E+04 2.3E+04 X1 Oct 1.1E+04 9.8E+03 0.271 1.7E+04 4.9E+03 0.944 1.8E+04 1.3E+04 AP 2 1.3E+03 1.5E+03 0.157 2.4E+03 7.7E+02 0.755 2.2E+03 1.6E+03 TBP binding repressor Dr1 1.7E+03 1.6E+03 0.536 2.2E+03 7.5E+02 0.364 1.5E+03 1.5E+03
134Table 4-3 (continued) Gene (not significantly different) Female SE F vs. I Pr> t Intersex SE M vs. I Pr>t Male SE TFIID subunit 1.7E+04 1.7E+04 0.223 2.8E+04 8.1E+03 0.476 2.2E+04 1.7E+04 Tcf 3 co repressor CtBP 4.2E+03 2.0E+03 0.145 1.6E+03 3.8E+02 0.779 1.8E+03 8.5E+02 brain factor 1 2.1E+04 1.5E+04 0.157 3.3E+04 6.7E+03 0.858 3.5E+04 1.8E+04 CCCH zinc finger protein C3H 3 2.0E+04 1.1E+04 0.964 2.0E+04 3.9E+03 0.208 3.1E+04 1.3E+04 DLL4 1.5E+04 1.2E+04 0.146 2.5E+04 5.6E+03 0.876 2.6E+04 1.6E+04 DLL3 4.3E+02 2.7E+02 0.402 5.5E+02 1.2E+02 0.851 5.8E+02 2.5E+02 Fez 4.9E+03 2.5E+03 0.823 4.6E+03 1.2E+03 0.213 9.4E+03 4.8E+03 DNA binding protein E12 2.5E+03 7.8E+02 0.078 1.5E+03 2.8E+02 0.28 2.3E+03 9.9E+02 Xsal 1 1.2E+03 1.3E+02 0.798 1.2E+03 5.6E+01 0.344 1.4E+03 2.6E+02 p54 mRNA binding protein 3.7E+04 4.1E+04 0.115 7.3E+04 2.0E+04 0.451 5.4E+04 4.4E+04 Myf 5 1.1E+03 3.0E+03 0.24 3.0E+03 1.5E+03 0.284 1.3E+03 3.0E+03 ATP dependent RNA helicase 1.5E+03 8.7E+02 0.819 1.4E+03 2.4E+02 0.472 1.7E+03 7.2E+02 fos related antigen 2 1.5E+03 2.0E+03 0.179 2.9E+03 1.0E+03 0.952 2.8E+03 2.2E+03 ER1 3.1E+03 1.6E+03 0.551 2.4E+03 4.7E+02 0.113 1.4E+03 1.0E+03 cellular retinoic acid binding protein 1.5E+03 1.4E+03 0.596 1.9E+03 7.4E+02 0.618 2.5E+03 1.9E+03 thyroid stimulating hormone beta 1.2E+03 3.0E+02 0.178 1.4E+03 1.5E+02 0.78 1.8E+03 1.7E+03 iodothyronine 5 deiodinase III 1.9E+04 1.6E+04 0.243 3.0E+04 7.6E+03 0.216 1.8E+04 1.6E+04 Cu Zn SOD 1.8E+04 1.9E+04 0.231 3.1E+04 9.1E+03 0.247 1.9E+04 1.9E+04 vitellogenin 1.1E+04 1.1E+04 0.162 1.9E+04 5.3E+03 0.583 2.3E+04 1.2E+04 importin alpha 1a 6.6E+03 4.1E+03 0.91 6.9E+03 1.3E+03 0.072 3.7E+03 2.9E+03 calbindin D28K 1.6E+03 1.1E+03 0.275 2.3E+03 4.6E+02 0.8 2.1E+03 1.1E+03 adult beta globin 1.1E+04 4.2E+04 0.522 2.5E+04 2.1E+04 0.5 9.8E+03 4.2E+04
135Table 4-3 (continued) hsp30B 2.1E+03 1.1E+03 0.507 2.5E+03 4.3E+02 0.352 4.1E+03 2.0E+03 NEDD 8 3.2E+04 2.6E+04 0.091 5.8E+04 1.2E+04 0.37 4.4E+04 2.7E+04 Gene (not significantly different) Female SE F vs. I Pr> t Intersex SE M vs. I Pr>t Male SE metalloproteinase XMMP 4.9E+03 1.9E+03 0.219 3.1E+03 4.6E+02 0.073 9.2E+03 3.6E+03 survivin 6.9E+04 3.4E+04 0.145 1.0E+05 1.4E+04 0.054 4.9E+04 3.8E+04 BMP1b 2.1E+03 7.2E+02 0.126 1.3E+03 2.4E+02 0.212 1.9E+03 6.8E+02 katanin p60 6.4E+03 1.1E+04 0.271 1.3E+04 5.2E+03 0.196 5.7E+03 1.1E+04 myelin proteolipid protein 1.1E+03 9.6E+01 0.382 1.2E+03 4.7E+01 0.209 1.1E+03 9.5E+01 myelin basic protein 1.4E+03 2.1E+03 0.307 2.5E+03 1.0E+03 0.984 2.4E+03 2.5E+03 beta tubulin 3.9E+02 2.6E+02 0.839 4.2E+02 1.2E+02 0.343 5.5E+02 2.6E+02 Xefiltin 5.5E+03 1.1E+04 0.162 1.4E+04 5.4E+03 0.302 6.8E+03 1.2E+04 alpha skeletal actin 8.2E+02 1.8E+02 0.923 8.3E+02 5.3E+01 0.169 6.0E+02 2.1E+02 MAM domain protein 4.6E+02 6.1E+01 0.088 5.3E+02 2.1E+01 0.101 6.3E+02 7.6E+01 protein S10 1.9E+04 1.9E+04 0.139 3.5E+04 9.0E+03 0.529 4.2E+04 1.9E+04 ribosomal protein L32 2.8E+02 1.4E+02 0.253 3.7E+02 6.6E+01 0.436 4.9E+02 2.2E+02 Pax 2 1.6E+03 NA 0.833 1.1E+03 NA 0.276 1.1E+03 NA
136 Figure 4-1 Representative region of three arrays depicting differences in gene expression among sex groups. a. Array of ovary tissue from a female toad. b. Array of testis tissue from a male toad. c. Array of an inte rsexed gonad. Genes are spotted in duplicate, and were differentially expressed across sex gr oups. Boxes indicate a single gene that was up regulated in ovaries relative to testes and intersexed tissues. Ovals indicate a gene that that expressed similarly in ov aries and testes, but down regulated in intersexed tissue. Crosses indicate a gene that was similar in ovaries and intersexed tissue, but down regulated in testes.
137 Figure 4-2 Percentage of genes that had a fold cha nge that was higher, equal, or lower relative to another sex group. Fold change equals the median expression value of one sex organ (e.g. ovary) divided by the median expre ssion level of another (e.g. intersex).
138 Figure 4-3 Gene expression correlat ions. a. ovary and intersexed tissue gene expression, b. testis and intersexed tissue gene expression, and c. ovary and testis tissues. Solid line represents the one to one line. a c b
139 Figure 4-4 Cluster analysis of fold change expression levels of each gene that met the data quality requirements and showed a fold change higher than 1.5 increases or 0.67 decreases in at least one category. Number s indicate clusters that had correlation coefficients of at least 0.90.
140 Figure 4-5 Association among gene expression and gonadal type. Black points represent a gene transcript. The further away a gene transcript is relative to the origin the more that gene deviates from an average response. If the point lies close to or along the line representing a gonad type (purple ovary, green testis, orange intersex), then it is more strongly related to that gonadal type. If gene are located in the same half of the plot, those gonadal tissue types are more related. Th is is particularly true for the halves represented by the X -axis which accounts for >65% of th e data variation. Therefore, ovarian and intersex gene expression patte rns are more similar to each other than either is to testicular gene expression. The names of th e genes identified as closely associated with each gonadal type are listed in Table 2. This figure was generated by Dr. Mary Lesperance University of Victoria, British Columbia Canada.
141 CHAPTER 5 RENAL PATHOLOGIES IN MARINE TOADS ARE ASSOCIATED WITH LAND USE PATTERNS Introduction Many pollutants are broadly distributed as a result of wide scale use, run off, and fallout in precipitation (Davidson & Knapp, 2007; Lie et al. 2003; Sparling, Fellers & McConnell, 2001; Thurm an & Cromwell, 2000). Even areas once be lieved to be pristine are now known to be affected by environmental pollutants (e.g. (Lie et al. 2003; Thurman et al., 2000). As a result, environmental contaminants are a pan global h ealth concern for humans and wildlife (Colborn & Clement, 1992; Colborn, vom Saal & Soto, 1993; Edwards, Moore & Guillette, 2006; Guillette et al. 1995; Guillette & Gunderson, 2001; Norris & Carr, 2006; Vos et al. 2000). Kidneys are the principle excretory organs for all vertebrates, filtering toxins from the blood and excreting them, so excessive or chronic exposure to exogenous toxins can lead to renal pathologies. Many endogenous me tabolic toxins and exogenous c ontaminants are removed from the body via active transport across the proximal t ubule which makes it a targ et for toxin-induced pathologies (Sweet, 2005). For example, acute tubular necrosis (a disease of the kidneys proximal tubules) occurs in humans that have inge sted medications or pesticides, or in wildlife that live in polluted habitats (Chan et al. 1998; Ortiz, De Canales & Sarasquete, 2003; Rankin, 2004), and leads to loss of tubular brush boarder s, epithelial cell thi nning, vacuolization, and degeneration (Kern, 1999; Ortiz et al. 2003). Interstitial tissue around the prox imal tubule is also an importa nt target for toxin-induced pathologies such as inte rstitial nephritis, which in volves infiltration of ly mphocytes and leads to inflammation. Chronic intersti tial nephritis typically occurs after long term exposure to pharmaceuticals or environmental toxins, and is furt her associated with interstitial fibrosis, an irreversible renal injury characterized by the accum ulation of matrix proteins in the interstitium
142 (Eddy, 1996; Kern, 1999; Ortiz et al. 2003). Interstitial fibrosis is the most important cause of chronic renal failure in humans (Sipes, McQueen & Gandolfi, 1997). Renal pathologies associated with pollutant exposure have been documented in a diverse array of wildlife. For example, the kidneys of herring gulls (Larus argentatus) from the Great Lakes (USA) region contaminated with planar ha logenated aromatic hydrocarbons (e.g., used in solvents and some pesticides) had diagnostic path ologies consistent with interstitial nephritis (Fox, Grasman & Campbell, 2007). In most inst ances of interstitial nephritis and tubular necrosis, glomeruli are unaffected. However, Balt ic grey seals (Halichoer us grypus), polar bears ( Ursus maritimus ), and sledge dogs (Canis familiaris) exposed to DDT, polychlorinated biphenyls (PCBs), or organohalogens had glomerular pathologies including large hy aline bodies, diffuse thickening of the glomer ular capillary walls, and scle rosis (Bergman, Bergstrand & Bignert, 2001; Sonne et al. 2006; Sonne et al., 2007b). Polar bears ( Ursus maritimus ) from eastern Greenland had several re nal pathologies that were co rrelated with body burdens of specific contaminants. For example, the occurrence of interstitial fibrosis in Polar bears from eastern Greenland was correlated with polybromin ated diphenyl ether (a flame retardant) concentrations in their blubber, whereas diffuse glomerular capillary wall thickening was found to be associated with chlordane (a pesticide) concentrations (Sonne et al., 2006). As evidenced by these examples, studies of the effects of pollutants on renal phys iology have focused predominantly on mammals. No studies have ch aracterized toxin-induced renal pathologies in amphibians. Although many studies have linked exposure to specific chemicals with specific renal diseases, none have compared re nal pathologies occurring in anim als living in different types of human-dominated landscapes. Urban and s uburban areas are typically polluted with
143 polyaromatic hydrocarbons (PAHs) associated with road runoff (Crosbie & Chow-Fraser, 1999; Maltby et al. 1995), whereas agricultural areas are more often polluted with several pesticides used to control weed, insect, and rodent populations (Crosbie et al., 1999; Kreuger, 1998). These associations are so strong that PAH and pesticide (i.e. meto lachlor) concentrations have been used to estimate the proportion of urba nized and agricultural land, respectively, among different watersheds ( Crosbie et al. 1999; Standley, Kaplan & Smith, 2000). Thus, both urbanized and agricultural habitats are polluted with mixtures of chemicals that likely induce renal pathologies. The goal of th is study was to investigate renal pathologies diagnostic of toxin exposure in Bufo marinus living in suburban and agricultural habitats. Methods Characterizatizing Land Use Type I collected Bufo marinus from five sites in South Flor ida with different suburban and agricultural land use. Land use type for each site was determined by importing Google Earth digital satellite images (downloaded August 20, 2007) and a distance key of each site into Image Pro-Plus (Media Cybernetics Inc) imag e analysis program. I centered a 5.6 km2 grid (with 9 ~622 m2 cells) over each collection site. The home range of B. marinus is ~2km2 (Zug, Lindgren & Pippet, 1975), so this grid incl udes the area likely to be experien ced by a toad from each local site. Land use type for the five sites was defi ned by the percentage of agricultural land, proximity to nearest ag riculture (for those sites lacking agriculture), and amount of suburban development (i.e. roadways) within the 5.6 km2 area. Agricultural landuse in this study is prodominantly sugarcane (northern sites), or mixed vegetable and silvicultu re (southern site), and not livestock ranches. I estimated the percentage of agricultural area (km2) per cell and averaged the percentages across the grid to dete rmine the mean percentage of agriculture for
144 each site. The percent agricultural coverage among the sites varied from 0 to 97% (Belle Glade 97% Canal Point 51.2%, Homestead 34%, Well ington 0%, and Lake Worth-0%). Belle Glade and Canal Point were almost entirely agri cultural land with few roads, whereas Homestead included a mixture of agricultura l fields, buildings, and roadways. Wellington and Lake Worth, had zero agriculture and were distinguished from each other by the distance to the closest agricultural area, amount of bu ildings and roadway cover. Wellington was 5.18 km from agriculture and had less suburban development, whereas Lake Worth was 22 km from agriculture and had more suburban development. Toads collected from suburban areas often were found in shopping center parking lots where they are likely exposed to oil residue s (polyaromatic hydrocarbons PAHs), antifreeze (e.g. ethylene glycol), and car exhaust, but they were also collected from lawns next to roads and parking areas. Suburban lawns are often treated with a variety of chemicals some of which are similar to those used at agricultural sites. Toad s collected from agricultural areas were collected from within agricultural fields, grassy areas on the margins of agricultural fields and from the parking lots of the field stati ons. Thus, in addition to exposur e to high levels of agricultural contaminants it is likely that these toads were also exposed to pollutants from cars, albeit to a lesser degree (the parking lots a these sites were only used by st ation employees and were not as busy as the shopping mall parking lo ts in the suburban sites). Sample Collection, and Histology After toads were captured, they were euthani zed with an overdose of the anesthetic MS222 (0.3% Tricaine Methanesulfona te, pH 7), the kidney-gonad complex was rem oved, and one kidney and gonad from each individual was fixed in neutral buffered formalin for histological analysis.
145 Kidney tissue from 82 toads was evaluated. Some of the toads had gondal abnormalties or intersex gonads. However, there was no association between gonadal abnormalities (e.g. intersex gonad or mishaped tistis) and kidney ab normalities within each site. For example, the presence of testicular abnormalities in toad s from Lake Worth and Belle Glade were not assocaited with the total number of kidney abnormalities (LW=X2 = 2.36, df=4, p=0.67; BG=X2 = 8.61, df=4, p=0.20). Therefore, all individuals within sites were pooled for site comparisons of renal pathologies, regardless of the presence of gonadal abnormalities. Kidney tissue from 68 male and 14 intersex toads was dehydrated, embe dded in paraffin, and sectioned at 7-8 m. Approximately 10 slides were made for each i ndividual; every other slide was stained with hematoxylin and eosin. To ensure that the same cells were not counted more than once, one section per stained slide (total of 5) was evaluated per individual, and sections were at least 90 m apart. At least 11 toads from each site were analyzed (LW=11 male, 0 intersex; WT=16 male, 0 intersex; HS=11 male, 3 intersex, CP=16 male, 4 intersex; BG=14 male, 7 intersex). Response Variables and Statistics I recorded the presence or abse nce of vaculization, degraded tubule cells, dilated tubules, interstitial ly mphocyte in filtration, and interstitial scarring for each toad (assessed across the five sections). Each of these responses was examined using a chi square analys is to determine if the incidence of each pathology varied among site s more than expected by chance. I counted the number of fungal elements, and granulomas in each section and averaged these counts across sections for each toad. These data were analyzed using Kruskal Wallis oneway analyses of variance to determine if th e abundances of these abnormalities (number per section per toad) differed among sites. I also combined some of the response variable s to examine pathologies indicative of toxininduced tubular necrosis and chr onic interstitial nephritis (Ker n, 1999). I determined the number
146 of pathologies present in a toad that were indicative of toxin-induced tubular necrosis and scored each toad 0-3 depending on the presence of vac uolization, degrading tubule cells, and dilated tubules. I also determined the number of patholog ies indicative of inters titial nephritis and toads received a score of 0-3 based on the presence of interstitial lymphocyte in filtration, presence of granulomas, or interstitial scarring. These res ponses were analyzed with Kruskal Wallis oneway analyses of variance to determine whether th e pathologies associated with each renal disease were common at specific sites. Results Several renal pathologies diagnos tic of exposure to toxicants we re identified in this study (Figure 1 A-E). The presence of vacuolization of tubule cells varied among sites more than expected by chance (X2 = 15.868, df=4, p=0.003). Approximately 55% of the toads at Lake Worth and Canal Point had vaculized tubule cells whereas only 20-25% of the Homestead and Belle Glade toads and 0% of the Wellington toads had this pathology. Approximately 25% of the toads at Lake Worth, the most suburban site, and Belle Glade, the most agricultural site, had dilated tubules, wh ereas very few if any toads at the other three sites had this pathology (X2 = 15.044, df=4, p=0.005; Figure 1 C and D;and Figure 3A). Degenerating (severely vacuolarized) tubule ce lls (Figure 1 B, Figure 3B) and interstitial fibrosis (epithelial scarring and hyperplasia Fi gure 1E; Figure 3C) vari ed in frequency among sites (X2 = 23.773, df=4, p<0.001 and X2 = 17.359, df=4, p=0.002 respectively). Degenerating tubule cells were most common at Belle Glade a nd Canal Point, and epit helial scarring was most common at Belle Glade. Lymphocyte infiltration was not significantly different among sites (X2 = 7.961, df=4, p=0.093 Figure 3D). Pathogenic fungal infections were found only in six toads across all sites. Lake Worth (LW) had no individuals with f ungal infections whereas West Pa lm (WP), Homestead (HS), and
147 Canal Point (CP) each had one, and Belle Glad e (BG) had three toads with fungal bodies. Therefore, statistical analysis was not conducted for this pathology. The number of granulomas was significantly different across sites (KW = 10.218, df=4, p=0.037; Figure 2). Toads obtained from the most agricultural site, Belle Glade ha d more granulomas (mode = 5; median 7.5) than toads from any other site (Figure 2). Howeve r, one toad from Wellington had 51 granulomas which is higher than any animal from BG (max 47). The number of abnormalites consistent with toxin-induced tubular necrosis were significantly different among sites (KW = 23.7, df=4, p<0.001, Figure 3E). Lake Worth (most suburban) and Canal Point and Belle Glade (agr icultural) had the hi ghest percentage of individuals with at least one pathology associated with tubul ar necrosis (73%, 80%, and 71% respecitvely), whereas 12.5% and 29% of th e individuals at Wellington and Homestead repectively had at least one. The number of pathologies consis tent with interstitial nephritis were significantly different among sites (KW = 15.3, df=12, p=0.004, Figure 3F) and oc curred more commonly than those indicative of tubular necrosis. Toads with all three interstitial pathologies investigated where found at every site except for Wellington (Figure 3F). Fifty percent of the individuals at Wellington had at least one patholo gy consistent with interstitial nephritis. However, 70%-75% of the toads at Lake worth, Homestead, and Canal Point, and 95% of the toads collected at Belle Glade had at least one pathology associated w ith interstitial nephrit is. The number of pathologies associated with interstitial nephriti s were lowest at Wellingt on and highest at Belle Glade (Figure 8B). Discussion A wide variety of drugs and environm ental po llutants induce nephrotox icity. For example, chronic renal failure in humans is associated w ith long term use or abuse of analgesics (Amdur et
148 al. 1996). The kidney is unusually su sceptible to environmental pollu tants for several reasons. It receives more blood, relative to organs of similar volume and thus, the kidney is exposed to greater quantities of toxins. Th e processes involved w ith filtering metabolic waste products and forming nitrogenous waste for excretion also concentr ate toxins in the tubular fluid. As water is reabsorbed, toxins in the tubul e fluid become highly concentrat ed, which can lead to their diffusion into the tubular cells (Amdur et al. 1996; Sipes et al. 1997). This concentrates toxins within the tubule cell. In addition, many pollutants are metabolized in the kidney, which contributes to its susceptibilit y, as chemicals can be bioactivat ed and made more toxic during their metabolism (Amdur et al. 1996; Sipes et al. 1997). The specific pathologies identified in this study have been associated with toxicant exposure in other organisms, including humans polar bear, dogs, birds, and fish (Chan et al., 1998; Eddy, 1996; Fox et al., 2007; Ortiz et al. 2003; Rankin, 2004; Sonne et al., 2006; Sonne et al. 2007a). In addition, the dominant types of pollutants in a specific area depend on local land use (Crosbie et al., 1999). For example, Crosbie and C how-Fraser (1999) demonstrated that polycyclic aromatic hydrocarbons (PAHs) were mo st concentrated in urban sites, whereas metolachlor was most concentrated in agricultural sites. These a ssociations were so strong that the authors argued that the pollutants could be used to determine land use practices or as surrogates to land use measures (Crosbie et al. 1999). However, many of the pollutants present at each of the five study sites in this study are unknown (but see appendix 1 for some known pollutants). Each site has a milieu of pollutants, and the mechanisms of toxicity leading to renal pathologies is complex. For example, the agricu ltural sites have high pesticide (e.g., atrazine and glyphosate), and fertilizer (leading to high levels of nitrates, a poten tial renal toxin) application. Urban settings likely have high PAHs (due to hi gher traffic and road density), as well as
149 residential application of fertilizers and pesticides In addition, the mixtures of chemicals at the two most agricultural sites (CP and BG; pre dominantly sugarcane) likely differ from the agricultural contaminants used at the farms ar ound Homestead which grow other types of plants (e.g., trees and vegetables). Although it is known that certain renal pathologi es are associated w ith pollution exposure and that pollutants vary with land use, this is the first study that examines th e occurrence of particular renal pathologies in toads from habitats that vary in land use. This study demonstrates that renal pathologies consistent w ith toxin exposure were present in Bufo marinus collected from all sites. Granulomas, degrading (necrotic) tubule cells, and interstiti al fibrosis occurred more frequently at the most agricultural site s (Belle Glade and Canal Point), whereas dilated tubules occurred only occurre d at the most suburban and the most agricultural sites. The more frequent occurrence of some toxin-induced path ologies at the most agricultural and suburban sites suggests that toads in thes e areas may experience greater expos ure to toxins or to a greater variety of toxins. Toxin-induced tubular necrosis is characte rized by proximal tubular epithelium necrosis where tubule cells become vacuolized, and necro tic. As the tubule cell s are destroyed, only the basement membrane around the tubule remains (Figure 1). This study demonstrates that pathologies associated with this tubule disease were generally more common in agriculture. A higher percentage of the animal s at the agricultural sites ha d two of the three (recorded) pathologies diagnostic of this dise ase. The only animals that had all three pathologies consistent with tubule necrosis were collected from the most agricultural site. However, a high percentage of toads at the most suburban site (70%) had at least one pathology associated with this debilitating disease which was mu ch higher than the sites with intermediate agricultural and
150 suburban exposures. The presence of tubular pathologies across si tes concomitant with specific pathologies occurring most often at agricultural sites (degenerati ng tubule cells) suggests that there are chemicals that have the same mechanisms of toxicity across site s, but that agricultural areas have additional chemicals that work th rough other mechanisms to induce additional pathologies. Alternatively, th e kidney could respond to toxic ch emicals of various forms in a similar manner. That is, specific renal pathol ogies are not due to specific chemicals, but represent a generalized response to perturbation by toxi c chemicals. Interstitial nephritis is characterized by the pr esence of inflamation in the renal interstitium (Kern, 1999). Inflammation is a natural biologica l response to cellular or tissue level damage and involves the infiltration of plasma and leukoc ytes from the blood into the injured tissues. Interstitial nephritis can be acute or chronic, and the diagnostic feature of chronic interstitial nephritis is the presence of in terstitial fibrosis (Kern, 1999). Fibrogenesis is characterized by excessive formation of extracellular matrix prot eins such as collagen in response to tissue damage or pathogens. In humans, progressive inte rstitial fibrosis is the strongest determinant of chronic renal failure, and the most common cause is the use of medications and thus is toxininduced. For example, in a recent study, 92% of the interstitial nephr itis cases investigated were drug induced (Clarkson et al. 2004). However, in terstitial nephritis can also be induced by bacterial or viral infections, or rarely, by hereditary diseas es. Differential diagnosis among drug/toxin or bacterial/viral induced interstitia l nephritis depends on the dominant types of leukocytes present in the interstitium (Kern, 1999). A predominance of neutrophils, which have diagnostic saddle shaped nuclei, in the inflammato ry infiltrate indicates th e presence of bacterial infection (Kern, 1999). However, lymphocytes and eosinophils predominate in drug induced interstitial nephritis. In a ddition, granuloma formation (dense clusters of lymphocytes and
151 multinucleated giant cells) occurs in approximate ly 30% of drug-induced interstitial nephritis cases, and adverse reactions to drugs is thought to be the most common cause of these formations (Kern, 1999). However, the presence of fungal infection can also induce granuloma formation. The majority of leukocytes observed infiltrati ng the renal interstitium in this study were lymphocytes, which have large nuclei and sm all amounts of cytoplasm relative to other leukocytes, and are associated w ith toxin induced interstitial nephritis. When examined specifically, lymphocyte infiltration (although not significantly), and the number of granulomas were highest at the most agricu ltural site. Some of the observe d granulomas were due to fungal infection, as they encapsulated fungal bodies. However, fungal inf ections were rare (~ 7%) and granulomas occurred when fungal bodies were not present. Tubular injury and necrosis can induce a cascade of physiologi cal events that lead to interstitial inflammation and fibrogenesis. As inte rstitial fibrosis worsens it can lead to further necrosis and reduced filtration rate. These diseas es either alone or in conjunction are associated with renal failure, which occurs when glomerular filtration rate decreases to a level at which kidneys can no longer function properly. As ki dneys fail many other systems are affected. Blood pressure can become increased due to exce ss fluids. Urea and other chemicals such as potassium accumulate in the blood, which can lead to edema and electrolyte imbalances. In addition, the kidney stops produci ng proper amounts of bicarbonate, which can lead to decreased blood pH, and death. The pathologies I document he re are expected to ha ve negative effects on the health of Bufo marinus and are likely occurring in and a ffecting other wildlife species in these or similar habitats. The similarity between the renal pathologies found across diverse taxa
152 exposed to toxicants, and those found in humans taking medications such as antibiotics is striking. The specific milieu of chemicals present at each site is expected to induce different pathologies depending on the tissue type. In previous studies I have demonstrated that morphological and functional gonadal abnormalitie s occur more often in toads living in agricultural areas where endocri ne disrupting chemicals known to alter the reproductive system are used. Agricultural toads also exhibit rena l pathologies that are diagnostic of pollution exposure. However, there is no association between gonadal and renal pathologies. Although gonadal abnormalities occur more often in toads liv ing in agricultural areas toxin-induced renal pathologies are not limited to agricultural sites. Therefore, toads across all five s ites are exposed to pollution, but the specific pollutants used in the most agricultural sites (appendix 1) induce endocrine disruption altering gonada l development whereas pollutants at the other sites do not. This suggests that the chemiacal milues at agri cultural sites work though different mechanism of toxicity relative to thos e at the suburban sites. Often, scientists think of anim als as models for understandin g human disease. In this study, a detailed understanding of human renal diseases was used to understand those of wildlife. Future research should focus on understanding the similarities and differences in the mechanisms through which environmental contaminants in fluence renal physiology and pathogenesis of kidney diseases in wildlife and humans. A more detailed understanding of the mechanisms of toxin-induced renal pathologies, as well as the chemicals involved, could help devise strategies to avoid or mitigate the impacts of toxins on humans and animals living in human-modified environments.
153 Figure 5-1 Selected renal pathologies. A. A normal kidney section. The distal t ubules (short arrow), proximal tubule with br ush borders (long arrow), and glomeruli (r ound arrow head) are distingu ishable (40X bar is 50 m). B. Tubule cell vacuolization indicated by (small arrow) has become so extensive that the cel ls are degrading. C. Dilated tubules and thinning tubule cells. Proximal tubules have no brush borders (double headed arrow). D. Tubular necrosis, dead and dying tubule cell casts are located within the tubules. Lymphocyte infiltration is exte nsive (short arrow), and tissue in this area is becoming fibrous and scarred. As the tubule cells are destroyed, only the basement membrane around the tubule remains (long arrow). The accumulation of lymphocytes in the renal tissue indicates an immune response. High infiltration of lymphocytes can cause the de gradation of the adjacent tissue structures. E. Chronic inters titial nephritis is associated with interstitial fibrosis an accumulation of matrix proteins in th e interstitium (long arrow), hyperplasia of interstitial epithelial cells, and lymphocyte infiltration (short arrow).
154 Figure 5-2 Mode of granuloma number across site s. The most suburban site is on the left and agriculture increases along the x-axis.
155 Figure 5-3 A. The percentage of male toads with the presence vs. absence of dilate d tubules across five study sites. B. Presence and absence of degrading (necrotic) tubule cells across sites. C. Presence and absence of inters titial fibrosis across sites. D. Presence and absence of lymphocyt e infiltration acr oss sites. E. Percentage of toads found at each site with zero three pathologies diagnostic of t oxin-induced tubular necrosis (maximum value is 3). F. Percentage of toads found at each site with zero three pathologies diagnostic of toxin induced interstitial nephritis (maximum value is 3). The most suburban site is on the left and the mo st suburban site is on the right. (LW=11 male, 0 intersex; WT=16 ma le, 0 intersex; HS=11 male, 3 intersex, CP=16 male, 4 intersex; BG=14 male, 7 intersex)
156APPENDIX CHEMICAL USAGE LIST Table A-1 A list of chem icals used at each agricultural site were provided by the farm managers at each facility. Information on what the chemical is used for and whether or not it is a known endocri ne disruptor was derived from the Pesticide Action Network North America ( http://www.pesticid einfo.org/Index.htm l ). Literature citations unde r sources include primary literature sources that report endocrine disrup tive effects of each known endocrine disruptor. Chemical trade name Active ingredient(s)-chemical class Use Endocrine disruptor Site applied Source(s) Atrazine Atrazine--Triazine Herbicide Known Reproductive BG and CP Danzo, 1997; Fan et al., 2007; Hayes et al., 2006b; Sanderson et al., 2002 Ametryn Ametryn--Triazin e Herbicide Unknown BG Asulox Asulam sodium salt-Carbamate Herbicide Unknown BG and CP Envoke Trifloxysulfuron-sodium-Sulfonylurea Herbicide Unknown BG and CP Mocap (active ingredients are ethroprop and PNCB) Ethroprop--Organophosphate PNCB--Substituted Benzene Nematicideinsecticide Unknown Suspected Thyroid BG Hurley et al., 1998 PROWL 3.3 EC Pendimethalin--2,6Dinitroaniline Herbicide Suspected Thyroid
157 Table A-1 (Continued) Roundup UltraMAX Glyphosate, isopropylamine salt--phosphonoglycine Herbicide Known Reproductive BG and CP Howe et al., 2004; Hurley et al., 1998; Oliveira et al., 2007; Walsh et al., 2000 Sencor DF Metribuzin--Triazine Herbicide Suspected Thyroid CP Porter et al., 1993 Thimet 20-G Phorate--Organophosphate Unknown BG and CP
158 LIST OF REFERENCES Am dur, M.O., Casarett, L.J., Doull, J., & Klaassen, C.D. (1996) Casarett and Doull's toxicology :the basic science of poisons 5 edn. McGraw-Hill, New York. Andres, A.C., Muellener, D.B., & Ryffel, G. (198 4) Persistence, methylation and expression of vitellogenin gene derivatives after injection into fertilized-eggs of Xenopus laevis Nucleic Acids Research 12, 2283-2302. Ankley, G.T., Jensen, K.M., Durhan, E.J., Makynen, E.A., Butterworth, B.C., Kahl, M.D., Villeneuve, D.L., Linnum, A., Gray, L.E., Ca rdon, M., & Wilson, V.S. (2005) Effects of two fungicides with multiple modes of action on reproductive endocrine function in the fathead minnow ( Pimephales promelas). Toxicological Sciences 86, 300-308. Ankley, G.T., Jensen, K.M., Kahl, M.D., Makyne n, E.A., Hornung, M.W., Korte, J.J., & Gray, L.E., Jr. (2001) Effects of the hormone growth supplement trenbolone on reproductive endocrinology of the fathead minnow. In Society of Toxicology Conference. Ankley, G.T., Jensen, K.M., Makynen, E.A., Ka hl, M.D., Korte, J.J., Hornung, M.W., Henry, T.R., Denney, J.S., Leino, R.L., Wilson, V.S., Cardon, M.C., Hartig, P.E., & Gray, L.E., Jr. (2003) Effects of the a ndrogenic growth promoter 17 -trenbolone on fecundity and reproductive endocrinology of the fathead minnow. Environmental Toxicology and Chemistry, 22 1350-1360. Anway, M.D., Cupp, A.S., Uzumcu, M., & Skinner, M.K. (2005) Epigene tic transgenerational actions of endocrine disruptors and mate fertility. Science 308 1466-1469. Arnold, S.F., Collins, B.M., Robinson, M.K., Guillette, L.J., Jr., & McLachlan, J.A. (1996) Differential interaction of natural and synt hetic estrogens with extracellular binding proteins in a yeast estrogen screen. Steroids, 61, 642-646. Ballestar, E. & Esteller, M. (2002) The imp act of chromatin in human cancer: linking DNA methylation to gene silencing. Carcinogenesis, 23 1103-1109. Barbeau, T.R. & Guillette, L.J., Jr. (2007) Altered ovarian steroids in Xenopus laevis exposed to environmentally relevant concentrations of nitrate. Journal of Herpetology 2007, 590596. Belfroid, A.C., Van der Horst, A., Vethaak, A.D ., Schafer, A.J., Rijs, G.B.J., Wegener, J., & Cofino, W.P. (1999) Analysis and occurre nce of estrogenic hormones and their glucuronides in surface water and waste water in the Netherlands. Science of the Total Environment 225, 101-108. Bergman, A., Bergstrand, A., & Bignert, A. (2001) Renal lesions in Baltic grey seals ( Halichoerus grypus ) and ringed seals ( Phoca hispida botnica ). Ambio, 30, 397-409.
159 Bern, H. (1992). The fragile fetus. In Chemically-Induced Alterations in Sexual and Functional Development: The Wild life/Human Connection (eds T. Colborn & C. Clement), Vol. XXI, pp. 9-15. Princeton Scientific Publishing Company, Princeton. Berube, V.E., Boily, M.H., DeBlois, C., Dassylv a, N., & Spear, P.A. (2005) Plasma retinoid profile in bullfrogs, Rana catesbeiana, in relation to agricu ltural intensity of subwatersheds in the Yamaska River drainage basin, Quebec, Canada. Aquatic Toxicology 71, 109-120. Bjerregaard, P., Dewailly, E., A yotte, P., Pars, T., Ferron, L., & Mulvad, G. (2001) Exposure of Inuit in Greenland to organochlorines through the marine diet. Journal of Toxicology and Environmental Health Part A 62, 69-81. Blaustein, A.R., Romansic, J.M., Kiesecker, J.M., & Hatch, A.C. (2003) Ultraviolet radiation, toxic chemicals and amphibi an population declines. Diversity and Distributions, 9, 123140. Block, K., Kardana, A., Igarashi, P., & Taylor, H.S. (2000) In utero di ethylstilbestrol (DES) exposure alters Hox gene expression in the developing Mullerian system. Faseb Journal 14, 1101-1108. Blumberg, B. & Evans, R.M. (1998) Orphan nuclear receptors new ligands and new possibilities. Gene Develop. 12, 3149-3155. Boas, M., Feldt-Rasmussen, U., Skakkebaek, N.E., & Main, K.M. (2006) Environmental chemicals and thyroid function. European Journal of Endocrinology 154 599-611. Borga, K., Gabrielsen, G.W., & Skaare, J.U. (2003) Comparison of organochlorine concentrations and patterns between freeranging zooplankton and zooplankton sampled from seabirds' stomachs. Chemosphere, 53, 685-689. Bridges, C.M. (1999) Effects of a pesticide on tadpole activity and predat or avoidance behavior. Journal of Herpetology 33, 303-306. Brown, F.D., del Pino, E.M., & Krohne, G. (2002) Bidder's organ in the toad Bufo marinus : Effects of orchidectomy on the morphol ogy and expression of lamina-associated polypeptide 2. Development Growth & Differentiation 44, 527-535. Burnison, B.K., Hartmann, A., Lister, A., Servos M.R., Ternes, T., & Van Der Kraak, G. (2003) A toxicity identification evaluation approach to studyi ng estrogenic substances in hog manure and agricultural runoff. Environmental Toxicology and Chemistry 22, 2243 2250. Carey, C. (2000) Infectious disease and worldw ide declines of a mphibian populations, with comments on emerging diseases in coral reef organisms and in humans. Environmental Health Perspectives 108 143-150.
160 Carey, C. & Bryant, C. (1995) Possible interrel ations among environmental toxicants, amphibian development, and decline of amphibian populations. Environmental Health Perspectives 103, 13-17. Chami, M., Gozuacik, D., Lagorce, D., Brini, M., Falson, P., Peaucellier, G., Pinton, P., Lecoeur, H., Gougeon, M.L., le Maire, M., Rizzuto, R., Brechot, C., & Paterlini-Brechot, P. (2001) SERCA1 truncated proteins unable to pump calcium reduce the endoplasmic reticulum calcium concentration and induce apoptosis. Journal of Cell Biology 153 1301-1313. Chan, B.S.H., Lazzaro, V.A., Seale, J.P., & Duggin, G.G. (1998) The renal excretory mechanisms and the role of organic cations in modulating the renal handling of paraquat. Pharmacology & Therapeutics 79, 193-203. Cheek, A.O., Ide, C.F., Bollinger, J.E., Rider, C.V., & McLachlan, J.A. (1999) Alteration of leopard frog ( Rana pipiens) metamorphosis by the herbicide acetochlor. Archives of Environmental Contamination and Toxicology 37 70-77. Clarkson, M.R., Giblin, L., O'Connell, F.P., O'Ke lly, P., Walshe, J.J., Conlon, P., O'Meara, Y., Dormon, A., Campbell, E., & Donohoe, J. (2004) Acute interstitial nephritis: clinical features and response to corticosteroid therapy. Nephrology dialysis and Transplantation 19, 2778-2783. Clotfelter, E.D., Bell, A.M., & Levering, K.R. ( 2004) The role of animal behaviour in the study of endocrine-disrupting chemicals. Animal Behaviour, 68, 665-676. Cohen, M.P. & Alford, R.A. (1993) Growth, survival and activity patterns of recently metamorphosed Bufo marinus Wildlife Research 20, 1-13. Colborn, T. & Clement, C., eds. (1992) Chemically-induced Alterations in Sexual and Functional Development: Th e Wildlife/Human Connection Vol. XXI, pp 403. Princeton Scientific Publishing Company Inc., Princeton. Colborn, T. & Smolen, M.J. (1996) Epidemiologi cal analysis of pers istent organochlorine contaminants in Cetaceans. Reviews of Environmental Contamination and Toxicology 146, 91-172. Colborn, T., vom Saal, F.S., & Soto, A.M. (1993) Developmental effects of endocrine-disrupting chemicals in wildlife and humans. Environmental Health Perspectives 101 378-384. Cooke, A.S. (1970) The eff ect of p,p'-DDT on tadpoles of the common frog ( Rana temporaria). Environmental Pollution 1, 57-71. Cooke, A.S. (1971) Selective predation by newts on frog tadpoles treated with DDT. Nature, 229, 275-276.
161 Costin, G.E. & Hearing, V.J. ( 2007) Human skin pigmentation: me lanocytes modulate skin color in response to stress. Faseb Journal 21, 976-994. Crain, D.A. & Guillette, L.J., Jr. (1997a) Endocri ne-disrupting contaminan ts and reproduction in vertebrate wildlife. Reviews in Toxicology 1, 47-70. Crain, D.A., Guillette, L.J., Jr., Rooney, A.A., & Pickford, D.B. (1997b) Alterations in steroidogenesis in alligators (Alligator mississippiensis ) exposed naturally and experimentally to environmental contaminants. Environmental Health Perspectives 105 528-533. Crain, D.A., Noriega, N., Vonier, P.M., Arnol d, S.F., McLachlan, J.A., & Guillette, L.J., Jr. (1998) Cellular bioavailability of natural hor mones and environmental contaminants as a function of serum and cyto solic binding factors. Toxicology and Industrial Health 14, 261-273. Crews, D., Gore, A.C., Hsu, T.S., Dangleben, N.L., Spinetta, M., Schallert, T., Anway, M.D., & Skinner, M.K. (2007) Transgenerational ep igenetic imprints on mate preference. Proceedings of the National Academy of Sc iences of the United States of America 104, 5942-5946. Crews, D. & McLachlan, J. (2006) Epigenetic s, evolution, endocrine disruption, health, and disease. Endocrinology 147, S4-S10. Crosbie, B. & Chow-Fraser, P. (1999) Percentage land use in the watershed determines the water and sediment quality of 22 mars hes in the Great Lakes basin. Canadian Journal of Fisheries and Aquatic Sciences 56, 1781-1791. Crump, D., Lean, D., & Trudeau, V.L. (2002a) Oc tylphenol and UV-B radiation alter larval development and hypothalamic gene expression in the leopard frog ( Rana pipiens ). Environmental Health Perspectives 110, 277-284. Crump, D., Werry, K., Veldhoen, N., van Aggelen, G., & Helbing, C.C. (2002b) Exposure to the herbicide acetochlor results in increased t hyroid hormone-dependent gene expression and accelerated metamorphosis in Xenopus laevis. Environmental Health Perspectives 110 1199-1205. Daftary, G.S. & Taylor, H.S. (2006) Endocrine regulation of Hox genes. Endocrine Reviews 27, 331-355. Danzo, B.J. (1997) Environmental xenobiotics may disrupt normal endocrine function by interfering with the binding of physiological ligands to steroid receptors and binding proteins. Environmental Health Perspectives 105 294-301. Danzo, B.J. (1998) The effects of e nvironm ental hormones on reproduction. Cellular and Molecular Life Sciences, 54, 1249-1264.
162 Davidson, C. & Knapp, R.A. (2007) Multiple stresso rs and amphibian declines: Dual impacts of pesticides and fish on yellow-legged frogs. Ecological Applications 17, 587-597. Degitz, S.J., Holcombe, G.W., Fl ynn, K.M., Kosian, P.A., Korte, J.J., & Tietge, J.E. (2005) Progress towards development of an amphibi an-based thyroid screening assay using Xenopus laevis Organismal and thyroidal res ponses to the model compounds 6propylthiouracil, methim azole, and thyroxine. Toxicological Sciences 87, 353-364. Denslow, N.D., Garcia-Reyero, N., & Barber, D.S. (2007) Fish 'n' chips: the use of microarrays for aquatic toxicology. Molecular Biosystems, 3, 172-177. Denton, T.E., Howell, W.M., Allison, J.J., McCollum, J., & Marks, B. (1985) Masculinization of female mosquitofish by expos ure to plant sterols and Mycobacterium smegmatis Bulletin of Environmental Contamination and Toxicology 35, 627-632. Diederichs, S., Baumer, N., Schultz, N., Hamra, F.K., Schrader, M.G., Sandstede, M.L., Berdel, W.E., Serve, H., & Muller-Tidow, C. (2005) Expression patterns of mitotic and meiotic cell cycle regulators in testic ular cancer and development. International Journal of Cancer 116, 207-217. Ding, Z.H., Gillespie, L.L., & Paterno, G.D. (2003) Human MI-ER1 alpha and beta function as transcriptional repressors by recruitment of histone deacetylase 1 to their conserved ELM2 domain. Molecular and Cellular Biology 23, 250-258. Dreitz, V.J. (2006) Issues in species rec overy: An example-based on the Wyoming toad. Bioscience 56, 765-771. Durhan, E.J., Lambright, C.S., Makynen, E.A., Lazorchak, J., Hartig, P.C., Wilson, V.S., Gray, L.E., & Ankley, G.T. (2007) Identification of metabolites of trenbolone acetate in androgenic runoff from a beef feedlot. Environmental Health Perspectives 114 (suppl 1) 65-68. Eddy, A.A. (1996) Molecular insights into renal interstitial fibrosis. Journal of the American Society of Nephrology 7 2495-2508. Edwards, T.M., McCoy, K.A., Barbeau, T., McCoy, M.W., Thro, J.M., & Guillette, L.J., Jr. (2006a) Environmental context determines nitrate toxicity in southern toad ( Bufo terrestris ) tadpoles. Aquatic Toxicology, 78, 50-58. Edwards, T.M., Moore, B.C., & Guillette, L.J., Jr (2005) Metamorphic sex change: Evolutionary insight for endocrine disruption studies. Integrative and Comparative Biology 45, 991991. Edwards, T.M., Moore, B.C., & Guillette, L.J., J r., (2006b) Reproductive dysgenesis in wildlife: a comparative view. International Journal of Andrology 29, 109-120.
163 Edwards, T.M. & Myers, J.P. (2007) Environmen tal exposures and gene regulation in disease etiology. Environmental Health Perspectives in press. Eisen, M.B., Spellman, P.T., Brown, P.O., & Botstei n, D. (1998) Cluster an alysis and display of genome-wide expression patterns. Proceedings of the National Academy of Science (USA) 95, 14863-14868. El Touny, L.H. & Banerjee, P.P. (2006) Identifi cation of both Myt-1 and Wee-1 as necessary mediators of the p21-independent inactivati on of the Cdc-2/cyclin B1 complex and growth inhibition of TRAMP cancer cells by genistein. Prostate 66, 1542-1555. Emerson, S.B., Greig, A., Carroll, L., & Prins, G.S. (1999) Androgen rece ptors in two androgenmediated, sexually dimorphic characters of frogs. General and Comparative Endocrinology, 114, 173. Epstein, M.S. & Blackburn, D.G. (1997) Histolo gy and histochemistry of androgen stimulated nuptial pads in the leopard frog ( Rana pipiens ). Canadian Journal of Zoology 75, 427477. Eroschenko, V.P., Amstislavsky, S.Y., Schwab el, H., & Ingermann, R.L. (2002) Altered behaviors in male mice, male quail, and sa lamander larvae following early exposures to the estrogenic pesticide methoxychlor. Neurotoxicology and Teratology 24, 29-36. European-Commission (1997) European Workshop on the Impact of Endocrine Disrupters on Human Health and the Environment. In E nvironment and Climate Research Programme, Vol. DG XII, European Commission. Falconi, R., Dalpiaz, D., & Zaccanti, F. (2004) Ul trastructural aspects of gonadal morphogenesis in Bufo bufo (Amphibia Anura) 1. sex diffe rentiation, Vol. 301A, pp. 378-388. Falconi, R., Dalpiaz, D., & Zaccanti, F. (2007) Morphological Aspects of Gonadal Morphogenesis in Bufo bufo (Amphibia anura): Bidder's Or gan Differentiation, Vol. 290, pp. 801-813. Fan, W.Q., Yanase, T., Morinaga, H., Gondo, S., Okabe, T., Nomura, M., Komatsu, T., Morohashi, K.I., Hayes, T.B., Takayanagi R., & Nawata, H. (2007) Atrazine-induced aromatase expression is SF-1 dependent: impli cations for endocrine disruption in wildlife and reproductive cancers in Humans. Environmental Health Perspective 115, 720-727. Farias, C.F., Carvalho-e-Silva, S.P., & do Brito -Gitirana, L. (2002) Bidder's organ of Bufo ictericus: a light and el ectron microscopy analysis. Micron, 33 673-679. Fei, X.L., Chung, H.J., & Taylor, H.S. (2005) Methoxychlor disrupts uterine Hoxa10 gene expression. Endocrinology, 146, 3445-3451.
164 Forman, B.M., Tzameli, I., Choi, H.S., Chen, L. Simha, D., Seol, W., Evans, R.M., & Moore, D.D. (1998) Androstane metabolites bind to and deactivate the nuclear receptor CAR. Nature, 395, 612-615. Forson, D. & Storfer, A. (2006) Effects of atrazine and iridovirus infection on survival and lifehistory traits of the long-toed salamander ( Ambystoma macrodactylum ). Environmental Toxicology and Chemistry 25, 168-173. Fort, D.J., Degitz, S., Tietge, J., & Touart, L.W. (2007) The hypothalamic-pituitary-thyroid (HPT) axis in frogs and its role in frog development and reproduction. Critical Reviews in Toxicology, 37, 117-161. Fox, G.A. (1995) Tinkering with the ti nkerer: Pollution verses evolution. Environmental Health Perspectives 103, Suppl. 4, 93-100. Fox, G.A. (2001) Effects of endoc rine disrupting chemicals on wild life in Canada: Past, present and future. Water Quality Research Journal of Canada 36, 233-251. Fox, G.A., Grasman, K.A., & Campbell, G.D. (2007) Health of herring gulls ( Larus argentatus ) in relation to breeding loca tion in the early 1990s. II. Ce llular and histopathological measures. Journal of Toxicology and Environm ental Health-Part a-Current Issues 70, 1471-1491. Freeman, J.L., Beccue, N., & Rayburn, A.L. (2005) Differential metamorphosis alters the endocrine response in anuran larvae exposed to T3 and atrazine. Aquatic Toxicology 75, 263-276. Gardiner, D., Ndayibagira, A., Grun, F., & Blumberg, B. (2003) Deformed frogs and environmental retinoids. Pure and Applied Chemistry 75, 2263-2273. Gardiner, D.M. & Hoppe, D.M. (1999) Environm entally induced limb malformations in mink frogs ( Rana septentrionalis ). Journal of Experimental Zoology 284, 207-216. Goksoyr, A. (2006) Endocrine disruptors in the ma rine environment: mechanisms of toxicity and their influence on reproductive processes in fish. Journal of Toxicology and Environmental Health Part A 69, 175-184. Goleman, W.L., Carr, J.A., & Anderson, T.A. (2002) Environmentally relevant concentrations of ammonium perchlorate inhib it thyroid function and alter sex ratios in developing Xenopus laevis Environmental Toxicology and Chemistry 21 590-597. Gray, L.E., Jr., Ostby, J.S., & Kelce, W.R. (1994 ) Developmental effects of an environmental antiandrogen: The fungicide vinclozolin alte rs sex differentiation of the male rat. Toxicol. Appl. Pharmacol. 129, 46-52.
165 Gray, L.E., Jr., Wilson, V.S., Stoker, T., Lambright, C., Furr, J., Noriega, N., Howdeshell, K., Ankley, G.T., & Guillette, L.J., Jr. (2006) Adverse effects of environmental antiandrogens and androgens on repr oductive development in mammals. International Journal of Andrology 29, 96-104. Gray, L.E.J., Ostby, J.M., & Marshall, R. (1993) The fungicide vinclozolin inhibits morphological sex differentiation in the male rat. Biol. Reprod. Suppl. 48, 97. Guillette, L.J., Jr. (2001) Developmental abnormal ities in alligators liv ing in contaminated environments. Toxicology, 164, 12-12. Guillette, L.J., Jr. (2003) Reproduction and environmental contaminants: Endocrinology, evolution, and alligators. Biology of Reproduction 68, 85-85. Guillette, L.J., Jr. (2004) Contaminants, reprodu ctive endocrinology and wildlife: The evolving field of signal disruption. Hormones and Behavior, 46, 130-131. Guillette, L.J., Jr. (2006) Endocrine disr upting contaminants Beyond the dogma. Environmental Health Perspectives 114 9-12. Guillette, L.J., Jr., Crain, D.A., Rooney, A.A., & Pickford, D.B. (1995) Organization versus activation the role of en docrine-disrupting contaminants (Edcs) during embryonicdevelopment in wildlife. Environmental Health Perspectives 103, 157-164. Guillette, L.J., Jr., Edwards, T.M., & Moore, B.C. (2007) Alligators, environmental contaminants and steroid hormones. Environmental Science in press Guillette, L.J., Jr., Gross, T.S., Masson, G.R., Matter, J.M., Percival, H.F., & Woodward, A.R. (1994) Developmental abnormalities of the gonad and abnormal sex-hormone concentrations in juvenile alligators from contaminated and control lakes in Florida. Environmental Health Perspectives 102, 680-688. Guillette, L.J., Jr. & Gunderson, M.P. (2001) Alterations in development of reproductive and endocrine systems of wildlife populations exposed to endocrine-disrupting contaminants. Reproduction 122, 857-864. Guillette, L.J., Jr. & Iguchi, T. (2003) Cont aminant-induced endocrine and reproductive alterations in reptiles. Pure and Applied Chemistry, 75, 2275-2286. Guillette, L.J., Jr., Kools, S., Gunderson, M.P., & Bermudez, D.S. (2006). DDT and its analogues: new insights into their endocr ine disruptive effects on wildlife. In Endocrine Disruption: Biological Basis for Hea lth Effects in Wildlife and Humans (eds D.O. Norris & J.A. Carr), pp. 332-355. Oxford University Press, New York.
166 Harper, C., Wootton, L., Michelangeli, F., Lefi evre, L., Barratt, C., & Publicover, S. (2005) Secretory pathway Ca2+-ATPase (SPCA1) Ca 2+ pumps, not SERCAs, regulate complex [Ca2+], signals in human spermatozoa. Journal of Cell Science 118, 1673-1685. Harris, M.L., Chora, L., Bishop, C.A., & Bogart, J.P. (2000) Speciesand age-related differences in susceptibility to pesticid e exposure for two amphibians, Rana pipiens, and Bufo americanus. Bulletin of Environmental Contamination and Toxicology 64 263-270. Harvey, P.W., Everett, D.J., & Springall, C.J. (2007) Adrenal toxicology: a strategy for assessment of functional toxicity to th e adrenal cortex and steroidogenesis. Journal of Applied Toxicology, 27, 103-115. Hayes, T.B. (1998) Sex determination and prim ary sex differentiation in amphibians: Genetic and developmental mechanisms. Journal of Experimental Zoology 281, 373-399. Hayes, T.B., Case, P., Chui, S., Chung, D., H aeffele, C., Haston, K., Lee, M., Mai, V.P., Marjuoa, Y., Parker, J., & Tsui, M. (2006a) Pe sticide mixtures, endoc rine disruption, and amphibian declines: Are we underestimating the impact? Environmental Health Perspectives 114, 40-50. Hayes, T.B., Collins, A., Lee, M., Mendoza, M ., Noriega, N., Stuart, A.A., & Vonk, A. (2002) Hermaphroditic, demasculinized frogs after ex posure to the herbicide atrazine at low ecologically relevant doses. Proceedings of the National Academy of Sciences of the United States of America 99, 5476-5480. Hayes, T.B., Haston, K., Tsui, M., Hoang, A., Haeffele, C., & Vonk, A. (2003) Atazine-induced hermaphroditism at 0.1 ppb in American leopard frogs (Rana pipiens ): Laboratory and field evidence. Environmental Health Perspectives 111, 568-575. Hayes, T.B. & Menendez, K.P. (1999) The effect of sex steroids on primary and secondary sex differentiation in the sexually dichromatic reedfrog ( Hyperolius argus : Hyperolidae) from the Arabuko Sokoke Forest of Kenya. General and Comparative Endocrinology 115, 188-199. Hayes, T.B., Stuart, A.A., Mendoza, M., Collins, A., Noriega, N., Vonk, A., Johnston, G., Liu, R., & Kpodzo, D. (2006b) Characterization of atrazine-induced gonadal malformations in African clawed frogs ( Xenopus laevis ) and comparisons with effects of an androgen antagonist (cyproterone acetate) and exogenous estrogen (17 -estradiol): Support for the demasculinization/feminization hypothesis. Environmental Health Perspectives 114, 134-141. Hecker, M., Giesy, J.P., Jones, P.D., Jooste, A. M., Carr, J.A., Solom on, K.R., Smith, E.E., Van der Kraak, G., Kendall, R.J., & Du Preez, L. (2004) Plasma sex steroid concentrations and gonadal aromatase activities in African clawed frogs ( Xenopus laevis ) from South Africa. Environmental Toxicology and Chemistry 23, 1996-2007.
167 Helbing, C., Gallimore, C., & Atkinson, B.G. (1996) Characterization of a Rana catesbeiana HSP30 gene and its expression in the liver of this amphibian duri ng both spontaneous and thyroid hormone-induced metamorphosis. Developmental Genetics 18, 223-233. Helbing, C., Gergely, G., & Atkinson, B.G. (1992) Sequential up-regulation of thyroid-hormone beta-receptor, onithine transcarbamylase, and carbamyl-phosphate synthetase messengerRNAs in the liver of Rana catesbeiana tadpoles during spontaneous and thyroid hormone-induced metamorphosis. Developmental Genetics 13, 289-301. Helbing, C.C. & Atkinson, B.G. (1994) 3,5,3'-t riiodothyronine-induced carbamyl-phosphate synthetase gene-expression is stabilized in the liver of rana catesbeiana tadpoles during heat-shock. Journal of Biological Chemistry 269 11743-11750. Helbing, C.C., Ji, L., Bailey, C.M., Veldhoen, N ., Zhang, F., Holcombe, G.W., Kosian, P.A., Tietge, J., Korte, J.J., & Degitz, S.J. (2007) Identification of gene expression indicators for thyroid axis disruption in a Xenopus laev is metamorphosis screening assay Part 2. Effects on the tail and hindlimb. Aquatic Toxicology, 82, 215-226. Hewitt, A.E., Crain, D.A., Gunderson, M.P., & Gu illette, L.J., Jr. (2002) Thyroid status in juvenile alligators (Alligator mississippiensis) from contam inated and reference sites on Lake Okeechobee, Florida, USA. Chemosphere, 47, 1129-1135. Hinson, J.P. & Raven, P.W. (2006) Effects of endocrine-disrupting chemicals on adrenal function. Best Practice and Research Clin ical Endocrinology and Metabolism 20, 111120. Holliday, R. & Pugh, J.E. (1975) DNA modificat ion mechanisms and gene activity during development. Science, 187 226-232. Iguchi, T., Watanabe, H., & Katsu, Y. (2001) Developmental effects of estrogenic agents on mice, fish and frogs: a mini-review. Hormones and Behavior, 40, 248-251. IPCS (2002) Global Assessment or the state of the science of endocrine disruptors (eds T. Damstra, S. Barlow, A. Bergman, R. Kavlock & G. Van der Kraak), Vol. WHO/PCS/EDC/02/2, pp. 1-180. Wo rld Health Organization. Janosek J., Hilscherova, K., Blaha, L., & Holoubek, I. (2006 ) Environmental xenobiotics and nuclear receptorsInteractions, e ffects and in vitro assessment. Toxicology in Vitro 20, 18-37. Jansen, M.S., Nagel, S.C., Miranda, P .J., Lobenhofer, E.K., Afshari, C.A., & McDonnell, D.P. (2004) Short-chain fatty acids enhance nuclear receptor activity through mitogenactivated protein kinase activati on and histone deacetylase inhibition. Proceedings of the National Academy of Sciences of the United States of America 101, 7199-7204.
168 Jimenez-Gonzalez, C., Michelange li, F., Harper, C.V., Barratt, C.L.R., & Publicover, S.J. (2006) Calcium signalling in human spermatozoa: a spec ialized 'toolkit' of channels, transporters and stores. Human Reproduction Update 12, 253-267. Jobling, S., Beresford, N., Nolan, M., Rodgers-Gra y, T., Brighty, G.C., Sumpter, J.P., & Tyler, C.R. (2002a) Altered sexual maturation and gamete production in wild roach ( Rutilus rutilus ) living in rivers that recei ve treated sewage effluents. Biology of Reproduction 66, 272-281. Jobling, S., Beresford, N., Nolan, M., Rodgers-Gra y, T.P., Brighty, G., Sumpter, J., & Tyler, C.R. (2002b) Wild intersex roach ( Rutilus rutilus ) have reduced fertility. Biol. Repro., 67, 515-524. Jobling, S., Nolan, M., Tyler, C.R., Brighty, G., & Sumpter, J.P. (1998) Widespread sexual disruption in wild fish. Environ. Sci. Technol. 32 2498-2506. Jobling, S., Sheahan, D., Osborne, J.A., Matthiess en, P., & Sumpter, J.P. (1996) Inhibition of testicular growth in rainbow trout ( Oncorhynchus mykiss) exposed to estrogenic alkylphenolic chemicals. Environmental Toxicology and Chemistry 15, 194-202. Jobling, S., Williams, R., Johnson, A., Taylor, A ., Gross-Sorokin, M., Nolan, M., Tyler, C.R., van Aerle, R., Santos, E., & Brighty, G. (2006) Predicted exposures to steroid estrogens in UK rivers correlate with widespread sexual disruption in wild fish populations. Environmental Health Perspectives 114, 32-39. Johnson, A.C., Williams, R.J., & Matthiessen, P. (2006) The potential steroid hormone contribution of farm animals to freshw aters, the United Ki ngdom as a case study. Science of the Total Environment 362, 166178. Jost, J.P., Saluz, H.P., McEwan, I., Feavers, I.M., Hughes, M., Reiber, S., Liang, H.M., & Vaccaro, M. (1990) Tissue specific expression of avian vitellogenin gene Is correlated with DNA hypomethylation and invivo specific protein-DNA interactions. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences 326, 231-240. Kajiura-Kobayashi, H., Kobayashi, T., & Na gahama, Y. (2005) Cloning of cDNAs and the differential expression of A-type cyclins and Dmc1 during spermatogenesis in the Japanese eel, a teleost fish. Developmental Dynamics 232, 1115-1123. Kamiya, K., Sato, T., Nishimura, N., Goto, Y ., Kano, K., & Iguchi, T. (1996) Expression of estrogen receptor and proto-oncogene messenger ribonucleic aci ds in reproductive tracts of neonatally diethylstilbestrol-exposed fema le mice with or without postpuberal estrogen administration. Experimental and Clinical Endocrinology and Diabetes 104, 111-122. Kern, W .F. (1999) Atlas of renal pathology Saunders, Philadelphia.
169 Kilgore, W.R., Mantyh, P.W., Mantyh, C.R., Mc Vey, D.C., & Vigna, S.R. (1993) Bombesin Grp-Preferring and Neuromedin B-Preferring Receptors in the Rat Urogenital System. Neuropeptides 24, 43-52. Kimura, H., Suetake, I., & Tajima, S. (1999) Xenopus maintenance-type DNA methyltransferase is accumulated and translocated into germinal vesicles of oocytes. Journal of Biochemistry 125, 1175-1182. Knippschild, U., Gocht, A., Wolff, S., Huber, N., Lohler, J., & Stoter M. (2005) The casein kinase I family: participation in multi ple cellular processes in eukaryotes. Cellular Signalling 17 675-689. Koussoulakos, S. (2004) Vertebrate limb development: from Harrison's limb disk transplantations to targeted disruption of Hox genes. Anatomy and Embryology, 209, 93105. Kretschmer, X.C. & Baldwin, W.S. (2005) CAR a nd PXR: Xenosensors of endocrine disrupters? Chemico-Biological Interactions 155 111-128. Kreuger, J. (1998) Pesticides in stream water within an agricultural catchment in southern Sweden, 1990-1996. Science of the Total Environment 216, 227-251. Lnge, R., Hutchinson, T.H., Croudace, C.P., Si egmund, F., Schweinfurth, H., Hampe, P., Panter, G.P., & Sumpter, J.P. (2001) Effects of the synthetic estrogen 17 ethinylestradiol on the life-cycle of the fathead minnow ( Pimephales promelas) Environmental Toxicology and Chemistry 20, 1216. Lawson, C., Dorval, V., Goupil, S., & Leclerc, P. (2007) Identification and localisation of SERCA 2 isoforms in mammalian sperm. Molecular Human Reproduction 13, 307-316. Leblanc, G.A. (1994) Assessing deleterious ecosyst em-level effects of e nvironmental pollutants as a means of avoiding evolutionary consequences. Environmental Health Perspectives 102, 266-267. Lee, J.C. (1996) The amphibians and reptiles of the Yucatn Peninsula Comstock, Ithica. Lee, J.C. & Correales, A.D. (2002) Sexual dimor phism in hind-limb muscle mass is associated with male reproductive success in Bufo marinus. Journal of Herpetology 36, 502-505. Leiva-Presa, A. & Jenssen, B.M. (2006) Effect s of p,p '-DDE on retinoid homeostasis and sex hormones of adult male European common frogs ( Rana temporaria). Journal of Toxicology and Environmental Health Part A Current Issues, 69, 2051-2062. Lewis, D.L., Baxter, G.T., Johnson, K.M., & St one, M.D. (1985) Possible extinction of the Wyom ing toad, Bufo hemiophrys baxteri Journal of Herpetology 19, 166-168.
170 Lie, E., Bernhoft, A., Riget, F., Belikov, S.E., Boltunov, A.N., Derocher, A.E., Garner, G.W., Wiig, ., & Skaare, J.U. (2003) Geographical distribution of organochlorine pesticides (OCPs) in polar bears ( Ursus maritimus ) in the Norwegian and Russian Arctic. Science of the Total Environment 306 159-170. Lintelmann, J., Katayama, A., Kurihara, N., S hore, L., & Wenzel, A. (2003) Endocrine disruptors in the environment. Pure and Applied Chemistry 75 631-681. Lynch, L.C. & Blackburn, D.G. (1995) Effects of testosterone administation and gonadectomy on nuptial pad morphology on overwintering male leopard frogs, Rana pipens AmphibiaReptilia 16, 113-121. Mahmood-Khan, Z. & Hall, E.R. (2003) Occurren ce and removal of plant sterols in pulp and paper mill effluents. Journal of Environmenta l Engineering Science 2, 17-26. Maltby, L., Boxall, A.B.A., Forrow, D.M., Calow, P., & Betton, C.I. (1995) The effects of motorway runoff on freshwater ecosyste ms. 2: Identifying major toxicants. Environmental Toxicology and Chemistry 14, 1093-1101. Mangelsdorf, D.J., Thummel, C., Beato, M., He rrlich, P., Schutz, G., Umesono, K., Blumberg, B., Kastner, P., Manuel, M., Chambon, P., & Evans, R.M. (1995) The nuclear receptor superfamily: the second decade. Vell 83, 835-839. Masuyama, H. & Hiramatsu, Y. (2004) Invol vement of suppressor for Gal 1 in the ubiquitin/proteasome mediated degr adation of estrogen receptors. The Journal of Biological Chemistry, 279 12020-12026. Matthiessen, P., Arnold, D., Johnson, A.C., Pepper, T.J., Pottinger, T.G., & Pulman, K.G.T. (2006) Contamination of headwater str eams in the United Kingdom by oestrogenic hormones from livestock farms. Science of the Total Environment 367, 616-630. Matthiessen, P. & Johnson, I. ( 2007) Implications of research on endocrine disruption for the environmental risk assessment, regulation a nd monitoring of chemicals in the European Union. Environmental Pollution 146 9-18. McLachlan, J.A. (2001) Environm ental signaling: what embryos and evolution teach us about endocrine disrupting chemicals. Endocrine Reviews, 22, 319-341. McLeod, S.D., Ranson, M., & Mason, R.S. (1994) Eff ects of estrogens on human melanocytes in vitro The Journal of Steroid Biochemistry and Molecular Biology. 49, 9-14. McMaster, M.E. (1995) Mechanisms of reproductive dysfunction in white sucker (Catostomus commersoni), brown bullhead (Ictalurus nebulosus), and goldfish (Carassius auratus) exposed to various pulp mill effluents and polycyclic a romatic hydrocarbon contaminants. Ph.D., University of Guelph.
171 McMaster, M.E., Van Der Kraak, G.J., Portt, C. B., Munkittrick, K.R., Sibley, P.K., Smith, I.R., & Dixon, D.G. (1991) Changes in hepatic mixed function oxygenase (MFO) activity, plasma steroid levels and age at maturity of a white sucker ( Catastomus commersoni ) population exposed to bleached kraft pulp mill effluent. Aquatic Toxicology, 21, 199-218. Miller, W.L. (1988) Molecular biolo gy of steroid hormone synthesis. Endocrine Reviews 9, 295318. Milnes, M.R., Bermudez, D.S., Bryan, T.A., Edwards, T.M., Gunderson, M.P., Larkin, I.L.V., Moore, B.C., & Guillette, L.J., Jr. (20 06) Contaminant-induced feminization and demasculinization of nonmammalian verteb rate males in aquatic environments. Environmental Research 100, 3-17. Minucci, S., DiMatteo, L., Chie ffi, P., Pierantoni, R., & Fasa no, S. (1997) 17 beta-estradiol effects on mast cell number and spermatogonial mitotic index in the testis of the frog, Rana esculenta Journal of Experimental Zoology 278, 93-100. Moore, F.L., Boyd, S.K., & Kelley, D.B. (2005) Historical perspective: hormonal regulation of behaviors in amphibians. Hormones and Behavior, 48, 373-383. Mosconi, G., Carnevali, O., Franzoni, M.F., Co ttone, E., Lutz, I., Kloas, W., Yamamoto, K., Kikuyama, S., & Polzonetti-Magni, A.M. (2002) Environmental estrogens and reproductive biology in amphibians. General and Comparative Endocrinology 126, 125129. Mueller, P.R., Coleman, T.R., & Dunphy, W. G. (1995) Cell-Cycle Regulation of a Xenopus Wee1-Like Kinase. Molecular Biology of the Cell, 6, 119-134. Mukerjee, D. (2006) Special secti on. Endocrine disruptors Introduction. Environmental Research 100 1-2. Munkittrick, K.R., Van Der Kraak, G.J., McMaster M.E., Portt, C.B., van den Heuvel, M.R., & Servos, M.R. (1994) Survey of receiving-wate r environmental impacts associated with discharges from pulp mills. 2. Gonad size, live size, hepatic EROD activity and plasma sex steroid levels in white sucker. Environmental Toxicology and Chemistry 13, 10891101. Murphy, M.B., Hecker, M., Coady, K.K., Tompsett, A.R., Higley, E.B., Jones, P.D., Du Preez, L.H., Solomon, K.R., Carr, J.A., Smith, E.E., Kendall, R.J., Van Der Kraak, G., & Giesy, J.P. (2006) Plasma steroid hor mone concentrations, aromatas e activities and GSI in ranid frogs collected from agricu ltural and non-agricultural s ites in Michigan (USA). Aquatic Toxicology, 77, 153-166. Nagel, S.C., vom Saal, F.S., Thayer, K.A., Dh ar, M.G., Boechler, M., & Welshons, W.V. (1997) Relative binding affinity-serum modified access (RBA-SMA) assay predicts the relative
172 in vivo bioactivity of the xenoestroge ns Bisphenol A and octylphenol. Environmental Health Perspectives 105 70-76. Nations, B.K. & Hallberg, G.R. (1992) Pesticides in Iowa precipitation. Journal of Environmental Quality 21, 486-492. Noriega, N.C. & Hayes, T.B. (1999) Preliminary evaluations of the Hyperolius argus estrogen screen. American Zoologist 39, 64A-64A. Noriega, N.C. & Hayes, T.B. (2000) DDT congene r effects on secondary sex coloration in the reed frog Hyperolius argus : a partial evaluation of the Hyperolius argus endocrine screen. Comparative Biochemistry and Physio logy B-Biochemistry and Molecular Biology, 126 231-237. Norris, D.O. & Carr, J.A. (2006) Endocrine Disruption: Biological Basis for Health Effects in Wildlife and Humans Oxford University Press, NY. Ogura, T. & Evans, R.M. (1995) A retinoic ac id triggered cascade of Hoxb1 gene activation. Proceedings of the National Academy of Sc iences of the United States of America 92, 387-391. Ohtani, H., Miura, I., & Ichikawa Y. (2000) Effects of dibutyl phthalate as an environmental endocrine disruptor on gonadal sex differe ntiation of genetic males of the frog Rana rugosa. Environmental Health Perspectives 108, 1189-1193. Oliveira, A.G., Telles, L.F., Hess, R.A., Mahecha, G.A.B., & Oliveira, C.A. (2007) Effects of the herbicide Roundup on the epididymal region of drakes Anas platyrhynchos Reproductive Toxicology, 23, 182-191. Opitz, R., Braunbeck, T., Bogi, C., Pickford, D.B ., Nentwig, G., Oehlmann, J., Tooi, O., Lutz, I., & Kloas, W. (2005) Description and initial evaluation of a Xenopus metamorphosis assay for detection of thyroid system-disrup ting activities of e nvironmental compounds. Environmental Toxicology and Chemistry 24, 653-664. Opitz, R., Hartmann, S., Blank, T., Braunbeck, T., Lutz, I., & Kloas, W. (2006a) Evaluation of histological and molecular e ndpoints for enhanced detection of thyroid system disruption in Xenopus laevis tadpoles. Toxicological Sciences 90, 337-348. Opitz, R., Trubiroha, A., Lorenz, C., Lutz, I., Hartmann, S., Blank, T., Braunbeck, T., & Kloas, W. (2006b) Expression of sodium-iodide sy mporter mRNA in the thyroid gland of Xenopus laevis tadpoles: developmental expression, effects of antithyroidal compounds, and regulation by TSH. Journal of E ndocrinology 190, 157-170. Orlando, E.F., Kolok, A.S., Binzcik, G.A., Gates, J.L., Horton, M.K., Lambright, C.S., Gray, L.E., Soto, A.M., & Guillette, L.J., Jr. (2004) Endocrine-disrupting effects of cattle
173 feedlot effluent on an aquatic sent inel species, the fathead minnow. Environmental Health Perspectives 112 353-358. Ortiz, J.B., De Canales, M.L.G., & Sarasquete, C. (2003) Histopathological changes induced by lindane (gamma-HCH) in various organs of fishes. Scientia Marina 67, 53-61. Ouellet, M., Bonin, J., Rodrigue, J., DesGrange s, J.L., & Lair, S. (1997) Hindlimb deformities (ectromelia, ectrodactyly) in free-livi ng anurans from agricultural habitats. Journal of Wildlife Diseases 33, 95-104. Palace, V.P., Evans, R.E., Wautier, K., Bar on, C., Vandenbyllardt, L., Vandersteen, W., & Kidd, K. (2002) Induction of vitellogenin and histol ogical effects in w ild fathead minnows from a lake experimentally treated with th e synthetic estrogen, ethynylestradiol. Water Quality Research Journal of Canada 37. Palanza, P., Morellini, F., Parmigiani, S., & vom Saal, F.S. (2002) Ethological methods to study the effects of maternal exposure to estr ogenic endocrine disrupters A study with methoxychlor. Neurotoxicology and Teratology 24, 55-69. Palmer, B.D., Huth, L.K., Pieto, D.L., & Selcer, K.W. (1998) Vitellogenin as a biomarker for xenobiotic estrogens in an amphibian model system. Environmental Toxicology and Chemistry 17 30-36. Palmer, B.D. & Palmer, S.K. (1995) Vitellogeni n Induction by xenobiotic estrogens in the redeared turtle and African clawed frog. Environmental Health Perspectives 103, 19-25. Pancak-Roessler, M.K. & Norris, D.O. (1991) The Effects of Orchiectomy and Gonadotropins on Steroidogenesis and Oogenesis in Bidde rs Organs of the Toad Bufo-Woodhousii. Journal of Experimental Zoology 260, 323-336. Panzica, G.C., Viglietti-Panzica, C., & Ottin ger, M.A. (2005) Introduction: neurobiological impact of environmental estrogens. Brain Research Bulletin 65, 187-191. Park, B.J. & Kidd, K. (2005) Effects of the synthe tic estrogen ethinylestradi ol on early life stages of mink frogs and green frogs in the wild and in situ. Environmental Toxicology and Chemistry, 24 2027-2036. Park, C.E., Kim, Y.H., Jeon, E.H., Cha, K.Y., Lee, S.H., & Lee, K.A. (2004) Expression of Wee1 and its related cell cycle com ponents in mouse early stage follicles. Cells Tissues Organs 177, 221-228. Park, D. & Propper, C.R. (2002) Endosulfan affect s pheromonal detection and glands in the male red-spotted newt, Notophthalmus viridescens Bulletin of Environm ental Contamination and Toxicology 69, 609-616.
174 Parks, L.G., Lambright, C.S., Orlando, E.F., Guillette, L.J., Jr., Ankley, G.T., & Gray, L.E., Jr. (2001) Masculinization of female mosquito fish in kraft mill effluent-contaminated Fenholloway river water is associated w ith androgen receptor agonist activity. Toxicological Sciences 62, 257-267. Pawlowski, S., Ternes, T., Bonerz, M., Kluczka, T., van der Burg, B., Nau, H., Erdinger, L., & Braunbeck, T. (2003) Combined in situ and in vitro assessment of the estrogenicactivity of sewage and surface water samples. Toxicological Sciences 75, 57-65. Pawlowski, S., van Aerle, R., Tyler, C.R., & T ., B. (2004) Effects of 17a-ethinylestradiol in a fathead minnow( Pimephales promelas) gonadal recrudescence assay. Ecotoxicology and Environmental Safety 57 330. Persson, J.L., Zhang, Q., Wang, X.Y., Ravnik, S. E., Muhlrad, S., & Wolgemuth, D.J. (2005) Distinct roles for the mammalian A-type cyclins during oogenesis. Reproduction 130 411-422. Peterson, F.W., Davis, R.K., & Orndorff, H.A. (2000) 17 -estradiol as an indicator of animal waste contamination in mantled karst aquifers. Journal of Environmental Quality 29, 826-834 Pickford, D.B., Guillette, L.J., Jr., Crain, D.A ., Rooney, A.A., & Woodward, A.R. (2000) Phallus size and plasma dihydrotestosterone concentrat ions in juvenile American alligators ( A. mississippiensis) from contaminated and reference populations. J. Herpetol., 34, 233-239. Pickford, D.B. & Morris, I.D. (1999) Effects of endocrine-disrupting contaminants on amphibian oogenesis: methoxychlor inhibits progesterone-induced maturation of Xenopus laevis oocytes in vitro Environmental Health Perspectives 107, 285-292. Pickford, D.B. & Morris, I.D. (2003) Inhibiti on of gonadotropin-induced oviposition and ovarian steroidogenesis in the African clawed frog ( Xenopus laevis ) by the pesticide methoxychlor. Aquatic Toxicology 62, 179-194. Porte, C., Janer, G., Lorusso, L.C., Ortiz-Zarra goitia, M., Cajaraville, M.P., Fossi, M.C., & Canesi, L. (2006) Endocrine disruptors in marine organisms: Approaches and perspectives. Comparative Biochemistry and Physio logy C-Toxicology & Pharmacology 143, 303-315. Pottinger, T.G. (2003) Interactions of endocrine disrupting chemi cals with stress responses in wildlife. Pure and Applied Chemistry 75, 2321-2333. Presnell, J.K. & Schreibm an, M.P. (1997) Humasons Animal Tissue Techniques. The Johns Hopkins University Press, Baltimore, MD. Propper, C.R. (2005) The study of endocrine di srupting compounds: past approaches and new directions. Integrative and Comparative Biology 45, 194.
175 Qin, Z.F., Zhou, J.M., Chu, S.G., & Xu, X.B. (2003) Effects of Chinese domestic polychlorinated biphenyls (PCBs) on gonadal differentiation in Xenopus laevis. Environmental Health Perspectives 111, 553-556. Rankin, G.O. (2004) Nephrotoxicity i nduced by Cand N-arylsuccinimides. Journal of Toxicology and Environmental Health-Part B-Critical Reviews 7, 399-416. Rastogi, R.K. & Chieffi, G. (1971) Effect of an antiandrogen, cyproterone acetate, on the pars distalis of pituitary and thumb pad of the male green frog, Rana esculenta L. Steroidologia 2, 276-282. Ravnik, S.E. & Wolgemuth, D.J. (1996) The devel opmentally restricted pa ttern of expression in the male germ line of a murine cyclin A, cyclin A2, suggests role s in both mitotic and meiotic cell cycles. Developmental Biology 173, 69-78. Reik, W., Dean, W., & Walter, J. (2001a) Epigenetic reprogramming in mammalian development. Science, 293 1089-1093. Reik, W. & Walter, J. (2001b) Evolution of im printing mechanisms: the battle of the sexes begins in the zygote. Nature Genetics 27, 255-256. Reik, W. & Walter, J. (2001c) Genomic imprinting: Parental influence on the genome. Nature Reviews Genetics 2, 21-32. Relyea, R.A. & Mills, N. (2001) Predator-induced stress makes the pesticide carbaryl more deadly to gray treefrog tadpoles (Hyla versicolor ). Proceedings of the National Academy of Sciences of the Unit ed States of America 98, 2491-2496. Rockett, J.C., Patrizio, P., Schmid, J.E., H echt, N.B., & Dix, D.J. (2004) Gene expression patterns associated with infertilit y in humans and rodent models. Mutation ResearchFundamental and Molecular Mechanisms of Mutagenesis 549 225-240. Rohr, J.R., Elskus, A.A., Shepherd, B.S., Crowley, P.H., McCarthy, T.M., Niedzwiecki, J.H., Sager, T., Sih, A., & Palmer, B.D. (2004) Mult iple stressors and sa lamanders: Effects of an herbicide, food limitation, and hydroperiod. Ecological Applications 14 1028-1040. Rohr, J.R., Kerby, J.L., & Sih, A. (2006a) Comm unity ecology as a framework for predicting contaminant effects. Trends in Ecology and Evolution 21, 606-613. Rohr, J.R. & Palmer, B.D. (2005) Aquatic herb icide exposure increases salamander desiccation risk eight months later in a terrestrial environment. Environmental Toxicology and Chemistry, 24 1253-1258.
176 Rohr, J.R., Sager, T., Sesterhenn, T.M., & Palmer, B.D. (2006b) Exposure, postexposure, and density-mediated effects of atrazine on amphi bians: Breaking down net effects into their parts. Environmental Health Perspectives 114, 46-50. Romansic, J.M., Diez, K.A., Higashi, E.M., & Bl austein, A.R. (2006) Effects of nitrate and the pathogenic water mold Saprolegnia on survival of amphibian larvae. Diseases of Aquatic Organisms 68, 235-243. Rosenshield, M.L., Jofre, M.B., & Karasov, W.H. (1999) Effects of polychlorinated biphenyl 126 on green frog ( Rana clamitans ) and leopard frog ( Rana pipiens ) hatching success, development, and metamorphosis. Environmental Toxicology and Chemistry 18, 24782486. Rucinski, M., Ziolkowska, A., Neri, G., Trejter, M., Zemleduch, T., Tyczewska, M., Nussdorfer, G.G., & Malendowicz, L.K. (2007) Expression of neuromedins S and U and their receptors in the hypothalamus and endocrine glands of the rat. International Journal of Molecular Medicine 20 255-259. Russell, R.W., Gillan, K.A., & Haffner, G.D. ( 1997) Polychlorinated biphenyls and chlorinated pesticides in southern On tario, Canada, green frogs. Environmental Toxicology and Chemistry, 16 2258-2263. Russell, R.W., Hecnar, S.J., & Haffner, G.D. (1995) Organochlorine pesticide residues in southern Ontario spring peepers. Environmental Toxicology and Chemistry 14, 815-817. Sachs, L.M., Damjanovski, S., Jones, P.L., Li, Q., Amano, T., Ueda, S., Shi, Y.B., & IshizuyaOka, A. (2000) Dual functions of thyroid hormone receptors during Xenopus development. Comparative Biochemistry and Physio logy B-Biochemistry and Molecular Biology, 126 199-211. Saluz, H.P., Jiricny, J., & Jost, J.P. (1986) Genomic sequencing reveals a positive correlation between the kinetics of strand-specific DNA demethylation of the overlapping estradiol glucocorticoid receptor-binding sites and th e rate of avian vitellogenin messenger RNA synthesis. Proceedings of the National Academy of Sciences of the United States of America 83, 7167-7171. Sanderson, J.T. (2006) The steroid hormone bios ynthesis pathway as a target for endocrinedisrupting chemicals. Toxicological Sciences, 94 3-21. Sanderson, J.T., Boerma, J., Lansbergen, G.W.A ., & van den Berg, M. (2002) Induction and inhibition of aromatase (CYP19) activity by various classes of pesticides in H295R human adrenocortical carcinoma cells. Toxicology and Applied Pharmacology, 182, 4454.
177 Sanderson, J.T., Seinen, W., Giesy, J.P., & va n den Berg, M. (2000) 2-chloro-s-triazine herbicides induce aromatase (CYP19) activit y in H295R human adrenocortical carcinoma cells: A novel mechanism for estrogenicity? Toxicological Sciences 54, 121-127. Schimidt, C.W. (2003) Toxicogenom ics: An Emerging Discipline. Environmental Health Perspectives Toxicogenomics 111, A20-A25. Severinsen, T., Skaare, J.U., & Lydersen, C. (2000) Spatial distribution of persistent organochlorines in ringed seal (Phoca hispida) blubber. Marine Environmental Research 49, 291-302. Shah, M.G. & Maibach, H.I. (2001) Estrogen and Skin An Overview. American Journal of Clinical Dermatology, 2, 143-150. Shang, Y.F. (2006) Molecular mechanisms of oestrogen and SERMs in endometrial carcinogenesis. Nature Reviews Cancer, 6, 360-368. Shiokawa, D., Hatanaka, T., Hatanaka, M., Shik a, Y., Nishikawa, A., & Tanuma, S. (2006) CDNA cloning and functiona l characterization of Xenopus laevis DNase gamma. Apoptosis 11 555-562. Siegel, S. & Castellan, N.J., Jr. (1988) Nonparametric statistics for the behavioral sciences second edn. McGraw-Hill, New York. Sipes, G., McQueen, C.A., & Gandolfi, A.J. (1997) Comprehensive Toxicology (Renal Toxicology) 1 edn. Pergamon, New York. Skaare, J.U., Bernhoft, A., Derocher, A., Gabrie lsen, G.W., Goksyr, A., Henriksen, E., Larsen, H.J., Lie, E., & Wiig, (2000) Organochlorines in top predators at Svalbard-occurrence, levels and effects. Toxicology letters 112-113, 103-109. Slominski, A., Tobin, D.J., Shibahara, S., & Wortsman, J. (2004) Melanin Pigmentation in Mammalian Skin and Its Hormonal Regulation. Physiological Reviews 4, 1155-1228. Sonne, C., Dietz, R., Larsen, H.J.S., Loft, K.E., Kirkegaard, M., Letcher, R.J., Shahmiri, S., & Moller, P. (2006a) Impairment of cellular immunity in West Greenland sledge dogs ( Canis familiaris ) dietary exposed to polluted Minke whale ( Balaenoptera acutorostrata ) blubber. Environmental Science & Technology 40 2056-2062. Sonne, C., Dietz, R., Leifsson, P.S., Asmund, G., Born, E.W., & Kirkegaard, M. (2007a) Are liver and renal lesions in East Greenland polar bears (Ursus maritimus ) associated with high mercury levels? Environmental Health 6. Sonne, C., Dietz, R., Leifsson, P.S., Born, E.W ., Kirkegaard, M., Letcher, R.J., Muir, D.C.G., Riget, F.F., & Hyldstrup, L. (2006b) Are or ganohalogen contaminants a cofactor in the
178 development of renal lesions in East Greenland polar bears (Ursus maritimus)? Environmental Toxicology and Chemistry 25, 1551-1557. Sonne, C., Fonfara, S., Dietz, R., Kirkegaard, M. Letcher, R.J., Shahmiri, S., Andersen, S., & Moller, P. (2007b) Multiple cytokine and acute -phase protein gene tr anscription in west Greenland sledge dogs ( Canis familiaris ) dietary exposed to organic environmental pollutants. Archives of Environmental Contamination and Toxicology 53, 110-118. Sonne, C., Leifsson, P.S., Dietz, R., Kirkegaard, M., Mller, P., Jensen, A.L., Letcher, R.J., & Shahmiri, S. (2007c) Renal lesions in Greenland sledge dogs ( Canis familiaris ) exposed to a natural dietary cocktail of persistent organic pollutants. Toxicological and Environmental Chemistry, 89, 563-576. Soso, A.B., Barcellos, L.J.G., & Ranzani-Paiva, M.J. (2007) Chronic exposure to sub-lethal concentration of a glyphosate-based herbic ide alters hormone profiles and affects reproduction of female Jundia ( Rhamdia quelen ) Environmental Toxicology and Pharmacology 23, 308-313. Soto, A.M., Calabro, J.M., Prechtl, N.V., Ya u, A.Y., Orlando, E.F., Daxenberger, A., Kolok, A.S., Guillette, L.J., Jr., le Bizec, B., Lange, I.G., & Sonnenschein, C. (2004) Androgenic and estrogenic activity in water bodies receiving cattle f eedlot effluent in eastern Nebraska, USA. Environmental Health Perspectives 112, 346-352. Sower, S.A., Reed, K.L., & Babbitt, K.J. (2000) Limb malformations and abnormal sex hormone concentrations in frogs. Environmental Health Perspectives 108, 1085-1090. Sparling, D.W., Fellers, G.M., & McConnell, L.L. (2001) Pestic ides and amphibian population declines in California, USA. Environmental Toxicology and Chemistry 20, 1591-1595. Spengler, P., Krner, W., & Metzger, J.W. ( 2001) Substances with estrogenic activity in effluents of sewage treatment plants in southwestern Germany. 1. Chemical analysis. Environmental Toxicology and Chemistry 20, 2133-2141. Stancheva, I. & Meehan, R.R. (2000) Transient depletion of xDnmt1 leads to premature gene activation in Xenopus embryos. Genes & Development 14, 313-327. Standley, L.J., Kaplan, L.A., & Smith, D. (2000) Molecular tracer s of organic matter sources to surface water resources. Environmental Science & Technology 34, 3124-3130. Steele, C.W., Strickler-Shaw, S., & Taylor, D.H. (1999) Effects of sublethal lead exposure on the behaviours of green frog ( Rana clamitans), bullfrog ( Rana catesbiana ) and Am erican toad ( Bufo americanus) tadpoles. Marine and Freshwater Behaviour and Physiology 32, 1-16. Storrs, S.I. & Kiesecker, J.M. ( 2004) Survivorship patterns of la rval amphibians exposed to low concentrations of atrazine. Environmental Health Perspectives 112, 1054-1057.
179 Sullivan, K.B. & Spence, K.M. (2003) Effects of sublethal concentrations of atrazine and nitrate on metamorphosis of the African clawed frog. Environmental Toxicology and Chemistry 22, 627-635. Swarup, D. & Patra, R.C. (2005) Environmenta l pollution and its impact on domestic animals and wildlife. Indian Journal of Animal Sciences 75, 231-240. Sweet, D.H. (2005) Organic anion transporter (Slc 22a) family members as mediators of toxicity. Toxicology and Applied Pharmacology 204, 198-215. Tabb, M.M. & Blumberg, B. (2006) New modes of action for endocrine-disrupting chemicals. Molecular Endocrinology 20, 475. Tamura, H., Maness, S.C., Reischmann, K., Dorm an, D.C., Gray, L.E., & Gaido, K.W. (2001) Androgen receptor antagonism by the or ganophosphate insecticide fenitrothion. Toxicological Sciences 60, 56-62. Tavera-Mendoza, L., Ruby, S., Brousseau, P., Fournier, M., Cyr, D., & Marcogliese, D. (2002a) Response of the amphibian tadpole Xenopus laevis to atrazine during sexual differentiation of the ovary. Environmental Toxicology and Chemistry 21 1264-1267. Tavera-Mendoza, L., Ruby, S., Brousseau, P., Fournier, M., Cyr, D., & Marcogliese, D. (2002b) Response of the amphibian tadpole Xenopus laevis to atrazine during sexual differentiation of the testis. Environmental Toxicology and Chemistry 21 527-531. Taylor, B., Skelly, D., Demarchis, L.K., Slade, M.D., Galusha, D., & Rabinowitz, P.M. (2005) Proximity to pollution sources and ri sk of amphibian limb malformation. Environmental Health Perspectives 113 1497-1501. Terasaka, S., Aita, Y., Inoue, A., Hayashi, S., Nishigaki, M., Aoyagi, K., Sasaki, H., WadaKiyama, Y., Sakuma, Y., Akaba, S., Tanaka, J., Sone, H., Yonemoto, J., Tanji, M., & Kiyama, R. (2004) Using a customized DNA mi croarray for expre ssion profiling of the estrogen-responsive genes to evaluate estr ogen activity among natural estrogens and industrial chemicals. Environmental Health Perspectives 112 773. Theodorakis, C.W., Rinchard, J., Carr, J.A., Park, J.W., McDaniel, L., Liu, F.J., & Wages, M. (2006) Thyroid endocrine disrup tion in stonerollers and cric ket frogs from perchloratecontaminated streams in east-central Texas. Ecotoxicology, 15 31-50. Thomas, E.O. & Licht, P. (1993) Testicular and androgen dependence of skin gland morphology in the anurans, Xenopus laevis and Rana pipiens. Journal of Morphology, 215, 195-200. Thomas, E.O., Tsang, L., & Light, P. (1993) Comperative histochem istry of the sexually dimorphic skin glands of anuran amphibians. Copeia 1993.
180 Thorton, M.J. (2002) The biological actions of estrogens on skin. Experimental Dermatology, 11 487-502. Thurman, E.M. & Cromwell, A. (2000) Atmospheri c transport, deposition, and fate of triazine herbicides and their metabolites in pristi ne areas at Isle Royale National Park. Environmental science and technology 34, 3079-3085. Tietge, J.E., Holcombe, G.W., Flynn, K.M., Kosian, P.A., Korte, J.J., Anderson, L.E., Wolf, D.C., & Degitz, S.J. (2005) Metamorphic inhibition of Xenopus laevis by sodium perchlorate: Effects on devel opment and thyroid histology. Environmental Toxicology and Chemistry, 24, 926-933. Toft, G., Baatrup, E., & Guillette, L.J., Jr. (2004) Altered social behavior and sexual characteristics in mosquitofish ( Gambusia holbrooki ) living downstream of a paper mill. Aquatic Toxicology, 70, 213-222. Triphan, J., Aumuller, G., Brandenburger, T., & Wilhelm, B. (2007) Localization and regulation of plasma membrane Ca2+-ATPase, in bovine spermatozoa. European Journal of Cell Biology, 86, 265-273. Tyler, C.R. & Routledge, E.J. (1998) Oestrogenic effects in fish in English rivers with evidence of their causation. Pure and Applied Chemistry 70 1795-1804. van Aerle, R., Nolan, M., Jobling, S., Christians en, L.B., Sumpter, J.P., & Tyler, C.R. (2001) Sexual disruption in a second sp ecies of wild cyprinid fish (the Gudgeon, gobio gobio) in United Kingdom freshwaters. Environmental Toxicology and Chemistry 20 2841-2847. van Wyk, J.H., Pool, E.J., & Leslie, A.J. (2003) The Effects of Anti-Androgenic and Estrogenic Disrupting Contaminants on Breeding Gl and (Nuptial Pad) Morphology, Plasma Testosterone Levels, and Plasma Vitellogenin Levels in Ma le Xenopus laevis (African Clawed Frog). Archives of Environmental Conatamination and Toxicology, 44, 247-256. Veldhoen, N. & Helbing, C.C. (2001) Detection of environmental endocrine-disruptor effects on gene expression in live Rana catesbeiana tadpoles using a tail fin biopsy technique. Environmental Toxicology and Chemistry 20, 2704-2708. Veldhoen, N., Skirrow, R.C., Ji, L., Domanski D., Bonfield, E.R., Bailey, C.M., & Helbing, C.C. (2006) Use of heterologous cDNA arrays and organ culture in the detection of thyroid hormone-dependent re sponses in a sentinel frog, Rana catesbeiana Comparative Biochemistry and Physiology D-Genomics & Proteomics 1, 187-199. Verrell, P. (2000) Methoxychlor increases susceptibility to predation in the salam ander Ambystoma macrodactylum Bulletin of Environmental C ontamination and Toxicology, 64, 85-92.
181 Vos, J.G., Dybing, E., Greim, H.A., Ladefoged, O., Lambre, C., Tarazona, J.V., Brandt, I., & Vethaak, A.D. (2000) Health effects of endoc rine-disrupting chemi cals on wildlife, with special reference to the European situation. Critical Reviews in Toxicology 30, 71-133. Walsh, L.P., McCormick, C., Martin, C., & Stocco, D.M. (2000) Roundup inhibits steroidogenesis by disrupting st eroidogenic acute regulatory (StAR) protein expression. Environ Health Perspect, 108, 769-776. Watanabe, H., Suzuki, A., Kobayashi, M., Taka hashi, E., Itamoto, M., Lubahn, D.B., Handa, H., & Iguchi, T. (2003) Analysis of temporal ch anges in the expression of estrogen-regulated genes in the uterus. Journal of Molecular Endocrinology 30, 347-358. Wechselberger, C., Kreil, G., & Richter, K. (1992) Isolation and Se quence of a Cdna-Encoding the Precursor of a Bombesin-Like Pep tide from Brain and Early Embryos of Xenopus Laevis. Proceedings of the National Academy of Sciences of the United States of America 89, 9819-9822. Welshons, W.V., Thayer, K.A., Judy, B.M., Taylor J.A., Curran, E.M., & vom Saal, F.S. (2003) Large effects from small exposures. I. Mech anisms for endocrine-d isrupting chemicals with estrogenic activity. Environmental Health Perspectives 111, 994-1006. Weltje, L. & Schulte-Oehlmann, U. (2007) The se ven year itch progre ss in research on endocrine disruption in aquatic invertebrates since 1999. Ecotoxicology, 16 1-3. Wetzel, D.M. & Kelley, D.B. (1983) Androgen a nd gonadotropin effects on male mate calls in South African clawed frogs, Xenopus laevis Hormones and Behavior 17 388-404. Wheeler, J.R., Gimeno, S., Crane, M., Lopez-Ju ez, E., & Morritt, D. (2005) Vitellogenin: A review of analytical methods to dete ct (anti) estrogenic activity in fish. Toxicology Mechanisms and Methods 15, 293-306. Witschi, E. (1933) Studies in sex differentiati on and sex determination in amphibians VI The nature of bidder's organ in the toad. American Journal of Anatomy 52, 461-515. Woodling, J.D., Lopez, E.M., Maldonado, T.A., Norris, D.O., & Vajda, A.M. (2006) Intersex and other reproductive disrupti on of fish in wastewater effluent dominated Colorado streams. Comparative Biochemistry and Physiology, Part C 144 10-15. Xie, W., Barwick, J.L., Simon, C.M., Pierce, A. M., Safe, S., Blumberg, B., Guzelian, P.S., & Evans, R.M. (2000) Reciprocal activation of Xenobiotic response genes by nuclear receptors SXR/PXR and CAR. Genes and Development, 14, 3014-3023. Ya mamoto, K.R. (1985) Steroid r eceptor regulated tr anscription of specific genes and gene networks. Annual Review of Genetics, 19, 209-252.
182 Zaccanti, F., Petrini, S., Rubatta, M.L., Stagni, A.M., & Giorgi, P.P. (1994) Accelerated Female Differentiation of the Gonad by Inhib ition of Steroidogenesis in Amphibia. Comparative Biochemistry and Physiology a-Physiology 107 171-179. Zala, S.M. & Penn, D.J. (2004) Abnormal behaviour s induced by chemical pollution: a review of the evidence and new challenges. Animal Behaviour, 68, 649-664. Zhang, F., Degitz, S.J., Holcombe, G.W., Kosi an, P.A., Tietge, J., Veldhoen, N., & Helbing, C.C. (2006) Evaluation of gene expr ession endpoints in the context of a Xenopus laevis metamorphosis-based bioassay to de tect thyroid horm one disruptors. Aquatic Toxicology, 76, 24-36. Zoeller, R.T., Dowling, A.L.S., Herzig, C.T.A., Iannacone, E.A., Gauger, K.J., & Bansal, R. (2002) Thyroid hormone, brain development, and the environment. Environ. Health Perspec. 110 (Suppl. 3) 355-361. Zoeller, R.T. & Tan, S.W. (2007) Implications of research on assays to characterize thyroid toxicants. Critical Reviews in Toxicology 37, 195-210. Zoeller, R.T., Tan, S.W., & Tyl, R.W. ( 2007) General background on the hypothalamicpituitary-thyroid (HPT) axis. Critical Reviews in Toxicology 37 11-53. Zuccotti, M. & Monk, M. (1995) Methylation of the Mouse Xist Gene in Sperm and Eggs Correlates with Imprinted Xist Expression and Paternal X-Inactivation. Nature Genetics, 9, 316-320. Zug, G.R. (1979) The marine toad, Bufo marinus : a natural history re sume of native populations Smithsonian Institut ion Press, Washington. Zug, G.R., Lindgren, E., & Pippet, J.R. (1975) Di stribution and ecology of marine toad, Bufo marinus, in Papua New-Guinea. Pacific Science 29, 31-50.
183 BIOGRAPHICAL SKETCH Krista McCoy has spent much of her time as a student being a teacher and a mentor. As an undergraduate, she volunteered for two years at the Virginia Living Museum in Newport News Va. where she interpreted exhibits for th e public and conducted li ve animal programs. After she graduated from Old Dominion Univers ity she worked for two years at Three Lakes Nature Center in Richmond Va. where she conduct ed nature hikes, hands-on animal programs, and nature programs that followed Virginias sta ndards of learning for students from grades K12. After receiving her masters degree with Re id Harris at James Madison University, she coinstructed a course for the Smithsonian Nationa l Museum of Natural Historys Monitoring and Assessment of Biodiversity Progr am (SI/MAB). This program was designed to teach wildlife managers from diverse countries and cultures (e.g. Nepal, India) wildlif e monitoring techniques to enhance their ability to manage and cons erve biodiversity in their own countries. As a PhD student at the Univer sity of Florida, Krista was selected to be an NSF GK-12 fellow in the Science Partners in Inquiry Ba sed Collaborative Education (SPICE) program. While in this program she taught middle school students the value of ecosystem health and biodiversity while instilling an appreciation of learning through inquiry based projects. Through SPICE she also had the opportunity to mentor students from diverse backgrounds typically underrepresented in science. During summer 2004, she invited two of her middle school students to come to the University of Florida and help conduct her disse rtation research. She also mentored a high school student through an experiment that was presented at the Science Fa ir. The student won first prize at the regional show and was invited to the State where she won third prize, and to the International Competition where she earned honorabl e mention. Krista has also mentored two other high school students thr ough the University of Floridas Student Scientist Training
184 Program (SSTP). These students spent 35 hrs per week in the laboratory with Krista, and she taught them how to care for experimental animals, and collect and analyze data to be presented in both poster and lecture formats. Krista has also mentored many undergra duate students by providing them with undergraduate research o pportunities related to he r research, or by helping them develop their own independent projects. Four of her students have completed independent honors projects and graduated with highest honors. She has also co-authored two papers with undergraduates, and she is currently working on two additional papers with undergraduates that will be submitted for publication before the end of the year. In additi on, Krista has participated in the Strategies for Ecology, Education, Development, and Sustainability program (SEEDS: http://www.esa.org/seeds/) at the 89th annual meeting of the Ecological Society of America. This program is designed to help students from unde rrepresented groups navigate a large scientific meeting and to network with other scientists without becoming overwhelmed and intimidated. Krista will continue to mentor, teach, and learn as she conducts postdoctoral research, through Boston University, at the Smithsonian Tropi cal Research Institute in Gamboa, Panama. She looks forward to continuing her disserta tion research and begi nning a variety of new projects.