1 THE ROLE OF ZINC AND COPPER ACCUMULATION IN QUEEN CONCH Strombus gigas REPRODUCTIVE DEFICIENCY AT NEARSHORE SITES IN THE FLORIDA KEYS By DANIEL JAMES SPADE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011
2 2011 Daniel James Spade
3 To my parents, for your love and support
4 ACKNOWLEDGMENTS I thank my chair Dr. Nancy Denslo w and supervisory committee Dr. David Barber, Dr. Gustav Paulay, and Dr. Madan Oli, for their excellent mentorship and support I thank current and former members of the Denslow and Barber laboratories for h elp and guidance. In p articul ar, I thank Dr. Christopher Martyniuk and Dr. R. Joseph Griffitt for training and guidance in microarray processing and data analysis and Dr. Martyniuk for training in mRNA cloning and real time RT PCR analysis I thank Kevin Kroll for assistance in setting up and carry ing out conch exposure experiments. I thank April Feswick for training in ICP MS sample processing and analysis. Additionally, I thank everyone in the laboratory who helped care for conchs or process samples, or who discussed my project with me improvi n g my thought process, my methodological approach and my understanding of the project I thank collaborators who were essential to th e work performed for this dissertation particularly Robert Glazer and Gabriel Delgado at the Florida Fish and Wildlife Con servation Commission, for their expert ise and guidance in working with queen conch for generously providing conch tissue samples, and for allowing me to participate in the 2009 field sampling effort I thank Megan Davis and Amber Shawl at the Harbor Bran ch Oceanographic Institute, for generously providing Florida fighting conchs and for guidance in setting up recirculating saltwater holding tanks and caring for conchs. I thank Nancy Brown Peterson at the University of Southern Mississippi for guidance an d training in interpretation of conch gonadal histology. I thank the staff of the UF Interdisciplinary Center for Biotechnology Research, particularly Dr. Li Liu, Dr. D Gigi Ostrow, and Dr. Yanping Zhang, all of whom provided guidance and assistance in w orking with the queen conch microarray. I thank Irvy Quitmyer from the Florida
5 Museum of Natural History for guidance in developing methods and processing samples for conch shell metal analysis. I thank the University of Florida for awarding me the Alumni Fellowship, as well as sources of funding for the work presented here, including the UF College of Veterinary Medicine for a Consolidated Faculty Re search Development Award, Mote Marine Laboratory ( grant no. POR 2007 36 ), The United States Environmental P rotection Agency (grant no. #X7974799 03), and the Florida Fish and Wildlife Conservation Commission (grants no. #F2410 and #NG06 106). I thank the UF Graduate Student Council, the Veterinary Graduate Student Association, the Department of Physiological S ciences, the Army Corps of Engineers, and the Society of Toxicology for providing travel funds that allowed me to attend scholarly meetings, present my research, and participate in a Pellston workshop on ecotoxicology in the 21 st century. I thank my family : Alison for always listening, Nathan for always challenging me, Mom and Dad for putting us ahead of everything else It was only because of their loving support, their guidance, and the example of their hard work and enjoyment of learning that I pursued this opportunity. Finally, I thank Michelle for her love and encouragement, for getting me through the toughest parts of this project and above all for put ting up with me while I wrote this dissertation.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 11 LIST OF FIGURES ................................ ................................ ................................ ........ 13 LIST OF OBJECTS ................................ ................................ ................................ ....... 15 LIST OF ABBREVIATIONS ................................ ................................ ........................... 16 CHAPTER 1 LITERATURE REVIEW ................................ ................................ .......................... 20 Background ................................ ................................ ................................ ............. 20 Significance of Queen Conch in the Florida Keys ................................ ................... 22 Economic Significance of Queen Conch in the Florida Keys ............................ 22 Ecological Significance of Queen Conch in the Florida Keys ........................... 22 Queen Conch Life History ................................ ................................ ....................... 23 Queen Conch Demographics and Reproductive Stressors in the Florida Keys ...... 27 Copper and Zinc Exposures Lead to Reduced Reproductive Success in Gastropods ................................ ................................ ................................ .......... 30 Copper and Zinc in South Florida ................................ ................................ ........... 37 Related Work with Queen Conch ................................ ................................ ............ 39 Queen Conch Genetics ................................ ................................ .................... 39 Production of an Oligonucleotide Microarray for Queen Conch ........................ 39 Metals in Queen Conch ................................ ................................ .................... 40 Overall Hypotheses ................................ ................................ ................................ 41 Significance of this Dissertation ................................ ................................ .............. 41 2 METHODS ................................ ................................ ................................ .............. 43 Strombus gigas Sample Collection ................................ ................................ ......... 43 Algae Sample Collection ................................ ................................ ......................... 45 In Viv o Exposures of Strombus alatus to Copper and Zinc ................................ ..... 45 Sources, Care, and Maintenance of Conchs ................................ .................... 45 Preliminary In Vivo Exposures ................................ ................................ .......... 47 Production of an Alternative Diet ................................ ................................ ...... 48 50 Day In Vivo Time Course Exposure of Fighting Conchs to High Oral Doses of Copper and Z inc ................................ ................................ ............. 49 Source of Montastraea faveolata Tissue Samples ................................ .................. 52 Histological Analysis ................................ ................................ ............................... 52 Preparation and Purification of RNA from Tissue Samples ................................ ..... 54 Reagents and Solutions for RNA Extraction ................................ ..................... 54
7 RNA Pr eparation Method A: Preparation of RNA Samples Using Guanidinium/Phenol/Chloroform Extraction and Isopropanol Precipitation ... 54 RNA Preparation Method B: Preparation of RNA Samples Using a Cs Cl Gradient Centrifugation Step ................................ ................................ ......... 55 Further Processing and Purification of RNA ................................ ..................... 56 RNA Quantification and Quality Analysis ................................ .......................... 56 Microarray Experimental Processing ................................ ................................ ...... 57 Cloning of Strombus gigas and Strombus alatus Partial Transcripts ...................... 59 Real Time RT PCR ................................ ................................ ................................ 59 18S Real Time RT PCR to Validate Performance of RNA Samples Prepared by Method B ................................ ................................ .................. 63 Validation of 18S rRNA as an Internal Reference Gene for Real Time RT PCR in Testis Samples ................................ ................................ ................. 64 Collection of Samples for Conch Shell Metal Testing ................................ ............. 64 Validation Study ................................ ................................ ............................... 64 Historical Shell Metal Study ................................ ................................ .............. 65 Metal Analysis by Inductively Coupled P lasma Mass Spectrometry (ICP MS) ....... 65 Determination of Blood Total Protein ................................ ................................ ...... 67 Statistical Analyses ................................ ................................ ................................ 67 3 CESIUM CHLORIDE GRADIENT CENTRIFUGATION IMPROVES THE QUALITY OF TOTAL RNA PREPARATIONS FROM THE GASTROPOD Strombus gigas AND THE CORAL Montastraea faveolata ................................ ..... 75 Background ................................ ................................ ................................ ............. 75 Results ................................ ................................ ................................ .................... 77 Quality of RNA Prepared by Method A ................................ ............................. 77 Quality of RNA Prepared by Method B ................................ ............................. 78 Cy3 Labeling of RNA Samples ................................ ................................ ......... 78 Real Time RT PCR for Strombus gigas 18S rRNA ................................ .......... 79 Discussion ................................ ................................ ................................ .............. 80 RNA Quality was Significantly Improved by CsCl Gradient Centrifugation ....... 80 CsCl Method May Provide Benefits for Downstream Applications ................... 81 Summary ................................ ................................ ................................ ................ 82 4 QUEEN CONCH ( Strombus gigas ) TE STIS REGRESSES DURING THE REPRODUCTIVE SEASON AT NEARSHORE SITES IN THE FLORIDA KEYS ... 87 Background ................................ ................................ ................................ ............. 87 Results ................................ ................................ ................................ .................... 89 Microarray Analysis of Testicular Transcription ................................ ................ 89 Real Time RT PCR ................................ ................................ .......................... 90 Tissue M etal Burdens ................................ ................................ ....................... 92 Correlations among Microarray, Histology, and Metal Data ............................. 92 Discussion ................................ ................................ ................................ .............. 93 Conch Testis Gene Expression ................................ ................................ ........ 94 Potential Role of Metals as a Reproductive Stressor ................................ ....... 97
8 High Throughput Sequencing for Gastropod Transcriptomics .......................... 99 Summary ................................ ................................ ................................ .............. 100 5 DIFFERENCES IN OVARIAN APOPTOSIS, TRANSLATION, AND LIPID METABOLISM I N THE DIGESTIVE GLAND OF NS FEMALE QUEEN CONCHS ................................ ................................ ................................ .............. 108 Background ................................ ................................ ................................ ........... 108 Results ................................ ................................ ................................ .................. 110 Ovarian Histology of Conchs Used in this Study ................................ ............ 1 10 Morphometric Data ................................ ................................ ......................... 111 Tissue Metal Concentrations ................................ ................................ .......... 111 Microarray Experiments: Quality Control ................................ ....................... 111 Differentially Expressed Transcripts per the Microarray Experiments ............ 112 Enriched Biological Processes ................................ ................................ ....... 113 Real Time RT PCR Quality Control ................................ ................................ 114 Real Time RT PCR Validation of Microarray Results ................................ ..... 115 Discussion ................................ ................................ ................................ ............ 116 Understanding Morphometric Differences between OS and NS ..................... 116 Metal Accumulation in NS Female Conchs ................................ .................... 118 Presence of a Masculinized Female Nearshore ................................ ............. 121 Real time RT PCR Validation of Microarray Results ................................ ...... 123 Transcriptional Status of Genes Known to Contribute to Reproductive Development ................................ ................................ ............................... 125 Reduced Translational Processes in Nearshore Conch Digestive Gland and Ovary ................................ ................................ ................................ .......... 128 NS Ovarian Development and Apoptosis ................................ ....................... 130 Lipid Metabolism in the Digestive Gland Might Play a Major Role in Ovarian Development ................................ ................................ ............................... 134 What Gene Expression Says about Metal Homeostasis ................................ 135 Summary: Implications for Reproduction of Nearshore Conchs .......................... 136 6 ZINC AND COPPER ACCUMULATION IN CONCH DIGESTIVE GLAND IN THE FIELD AND THE LABORATORY, W ITH POSSIBLE IMPLICATIONS FOR REPRODUCTION ................................ ................................ ................................ 142 Background ................................ ................................ ................................ ........... 142 Results ................................ ................................ ................................ .................. 145 March 2009 Queen Conch Field Study ................................ .......................... 145 Copper and Zinc Concentrations in the Tissues of Wild Queen Conchs ........ 146 Copper, Zinc, Cadmium, and Lead Concentrations in Algal Samples Collected from Sites of Conch Aggregations ................................ ............... 147 Real Time RT PCR Quality Control ................................ ................................ 148 Vitellogenin Expression in Wild Queen Conchs ................................ .............. 148 50 day In Vivo Time Course Exposure Study ................................ ................. 148 Histological Analysis of In Vivo Gonad Development ............................... 150 Copper and Zinc in Florida Fighting Conchs Exposed In Vivo ........................ 151
9 Cloning of Strombus alatus Vite llogenin Partial mRNA ................................ .. 151 Real time RT PCR Analysis of Vitellogenin mRNA Expression In Vivo .......... 152 Discussion ................................ ................................ ................................ ............ 153 Field Collected Queen Conchs: Persistent Trends in Metals and Reproductive Status ................................ ................................ .................... 153 Algal Metal Concentrations ................................ ................................ ............. 156 Feeding Study Temperature Control and Water Quality ................................ 159 Metal Accumulation in In Vivo Exposed Conchs ................................ ............ 161 Indicators of Reproductive Status in In vivo Exposed Conchs ....................... 164 Summary of Findings ................................ ................................ ............................ 165 7 DEVELOPMENT OF METHODS T O MEASURE TRACE METAL CONCENTRATIONS IN CONCH SHELL ................................ ............................. 177 Background ................................ ................................ ................................ ........... 177 Results ................................ ................................ ................................ .................. 178 Consistency of Data from Bur Type Tests ................................ ...................... 178 Trends in Metal Concentrations from Monroe County, FL, Strombid Shells ... 179 Discussion ................................ ................................ ................................ ............ 180 8 CONCLUSIONS ................................ ................................ ................................ ... 185 Evidence for an Exogenous Zinc Source ................................ .............................. 188 In Vivo Zinc Exposure: Effects on the Ovary ................................ ........................ 191 Potential Sources and Effects of Cu and Zn in the Nearshore Environment ......... 193 Multiple Stressors Could Inhibit Conch Reproduction ................................ ........... 195 How Clear is the Nearshore Offshore Distinction? ................................ ............... 198 H ow Likely is Nearshore Reproduction to Affect Conch Population Growth? ....... 199 Methods Developed in this Dissertation and Associated Projects ........................ 200 Refinements and Future Directions ................................ ................................ ...... 201 Further Work with Zn and Cu in the NS Florida Keys ................................ ..... 201 Investigating Other P ossible Stressors in the NS Environment ...................... 202 Improving upon Study Design for Conch Trace Metal Exposures .................. 202 Implications for Con ch Management: Translocation Efforts .......................... 203 Closing Statements ................................ ................................ ............................... 204 APPENDIX A LIST OF DIFFERENTIALLY REGULATED PROBES IN THE TESTI S MICROARRAY EXPERIMENT ................................ ................................ ............. 206 B CONCENTRATIONS OF ALL METALS MEASURED IN 2007 WILD CAUGHT MALE QUEEN CONCHS ................................ ................................ ...................... 207 C CONCENTRATIONS O F ALL METALS MEASURED IN 2007 WILD CAUGHT FEMALE QUEEN CONCHS ................................ ................................ ................. 209
10 D LIST OF DIFFERENTIALLY REGULATED GENES IN THE DIGESTIVE GLAND MICROARRAY EXPERIMENT ................................ ................................ 211 E LIST OF DIFFERENTIALLY REGULATED GENES IN THE OVARY MICROARRAY EXPERIMENT ................................ ................................ ............. 212 F ADDITIONAL ANALYSES OF OVARY AND DIGESTIVE GLAND MICROARRAY DATA TO DETERMINE THE APPROPRIATENESS OF INCLUDING THE MASCULINIZED FEMALE ................................ ....................... 213 G CROSS HYBRIDIZATION OF Strombus alatus RNA ONTO THE Strombus gigas MICROARRAY ................................ ................................ ............................ 215 LIST OF REFERENCES ................................ ................................ ............................. 217 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 240
11 LIST OF TABLES Table page 2 1 Histological classification of stages of gonad development in Strombus gigas and Strombus alatus ................................ ................................ ........................... 71 2 2 Primers used for Strombus gigas partial mRNA and rRNA cloning and real time RT PCR ................................ ................................ ................................ ...... 72 3 1 Comparison of RNA quality data for several projects prepared by each of two methods. ................................ ................................ ................................ ............. 83 3 2 Cy3 labeling of Strombus gig as and Strombus alatus RNA samples prepared by different methods ................................ ................................ ........................... 84 4 1 Enriched Gene Ontology (GO) biological processes in the testis microarray experiment. ................................ ................................ ................................ ....... 102 4 2 Comparison of testis gene expression results by microarray and real time RT PCR. ................................ ................................ ................................ ........... 102 4 3 Validation of 18S rRNA as an internal reference gene for February 2007 conch testis gene expression assays. ................................ .............................. 103 4 4 Non parametric correlations among histological indices of testis development, metal concentrations, and gene expression data. ...................... 104 5 1 Morphometric data for female queen conchs collected in the February and June 2007 sampling efforts. ................................ ................................ ............. 137 5 2 Enriched Gene Ontology (GO) biological processes in the digestive gland microarray experiment. ................................ ................................ ..................... 137 5 3 Enriched Gene Ontology (GO) biological processes in the ovary microarray experiment. ................................ ................................ ................................ ....... 138 5 4 Apoptosis associated genes differentially regulated in the ovary microarray experiment. ................................ ................................ ................................ ....... 138 5 5 Comparison of digestive gland gene expression results b y microarray and real time RT PCR. ................................ ................................ ............................ 139 5 6 Comparison of ovary gene expression results by microarray and real time RT PCR ................................ ................................ ................................ ............ 139 6 1 Copper and zinc concentrations in the tissues of wild queen conchs collected in March 2009. ................................ ................................ ................................ .. 169
12 6 2 Metal concentrations in algal samples collected from sites of conch aggregations in the Fl orida Keys, March, 2009. ................................ ............... 170 6 3 Water chemistry data from the 50 day in vivo Cu and Zn feeding study ........... 170 6 4 Histological score s for ovarian development of female fighting conchs from the in vivo Cu and Zn feeding study. ................................ ................................ 170 B 1 Concentrations of metals measured in 2007 wild caught male queen conchs. 207 C 1 Concentrations of all metals measured in 2007 wild caught female queen conchs. ................................ ................................ ................................ ............. 209
13 LIST OF FIGURES Figure page 2 1 Queen conch sampl ing sites in the Florida Keys ................................ ................ 73 2 2 Typical phenol chloroform RNA extraction followed by precipitation or an additional CsCl gradient centrifugatio n step ................................ ....................... 74 3 1 High quality S. gigas and M. faveolata total RNA profiles in capillary electrophoresis ................................ ................................ ................................ ... 85 3 2 Real time RT PCR a ssay for queen conch 18S rRNA ................................ ........ 86 4 1 Hierarchical clustering of significantly differentially regulated probes in the conch testis microarray experiment ................................ ................................ .. 105 4 2 Pathway analysis of differentially regulated probes from the conch testis microarray experiment ................................ ................................ ...................... 106 4 3 Concentrations of Zn and Cu in tissues of male c onc hs collected in February 2007 ................................ ................................ ................................ ................. 107 5 1 Copper and zinc concentrations measured in female queen conch tissues coll ected in February and June 2007 ................................ ............................... 139 5 2 Differentially regulated genes in the digestive gland and ovary microarray experiments ................................ ................................ ................................ ...... 140 5 3 Overlap in differentially expressed genes between the ovary and digestive gland microarray experiments ................................ ................................ .......... 141 6 1 Concentrations of Zn and Cu in the tissues of female and male queen c onchs collected in March 2009 ................................ ................................ .................... 171 6 2 Vitellogenin mRNA expression in ovary of queen female queen conchs collected in March 2009 ................................ ................................ .................... 172 6 3 Tank water temperatures during pre exposure and exposure per iods for the 50 day in vivo Cu and Zn feeding study ................................ ........................... 172 6 4 Representative images of ovarian histology in fighting conchs ........................ 173 6 5 Concentrations of Cu and Zn in female fighting conch digestive gland and ovary from the in vivo feeding study ................................ ................................ 174 6 6 Alignment of Strombus alatus vitellogenin partial mRNA sequence wi th the Strombus gigas sequence ................................ ................................ ................ 175
14 6 7 Vitellogenin mRNA expression in the ovary of S. alatus from the 50 day in vivo feeding study ................................ ................................ ............................. 176 7 1 Cu and Zn concentrations from calcareous materials used for bur tests .......... 183 7 2 Cu and Zn concentrations in archived conch shell samples from the Fl orida Museum of Natural Hist ory ................................ ................................ ............... 184 8 1 Comparison of digestive gland Zn concentrations in conchs fro m 2007 and 2009 field studies ................................ ................................ .............................. 205 F 1 Volcano plot representing analysis of ovary microarray data on arbitrarily determined groups ................................ ................................ ............................ 214 F 2 Dendrograms showing the results of hierarchical clustering analysis performed on all genes for the dig estive gland and ovary microarray studies .. 214 G 1 Percent present calls for Strombus gigas and Strombus alatus cRNA samples hybridized to the Strombus gigas microarray ................................ ..... 216
15 LIST OF OBJECTS Object page A 1 List of differentially regulated probes in the testis microarray experiment (.csv file 50 KB) ................................ ................................ ................................ ......... 206 D 1 List of differentially regulated genes in the digestive gland microarray experiment (.csv file 23 KB) ................................ ................................ .............. 211 E 1 List of differentially regulated genes in the ovary microarray experiment (.csv file 33 KB) ................................ ................................ ................................ ......... 212
16 LIST OF ABBREVIATION S 1 ATL Aquatic Toxicology Laboratory at the University of Florida CA cortical alveolar oocyte Ctr1c copper transporter 1c DG digestive gland DS Delta Shoal, a nearshore site in the Florida Keys E IF5A eukaryotic translation initiation factor 5A ES Eastern Sambo, an offshore site in the Florida Keys ESR EV early vitellogenic oocyte FW RI Fish and Wildlife Research Institute of the Florida Fish and Wildlife Conservation Commission GO Gene Ontology GST glutathione S transferase HBOI Harbor Branch Oceanographic Institution of Florida Atlantic University ICBR Interdisciplinary Center for Bi otechnology Research at the University of Florida ICP MS inductively coupled plasma mass spectrometry LV late vitellogenic oocyte NG neural ganglia, specifically the buccal ganglia for all mentions in this dissertation NS nearshore: conch aggregations loc ated between the Florida Keys and the Hawk Channel OS offshore: conch aggregations located south of the Hawk Channel, extending to the reef tract 1 Several transcript abbreviations are listed here; this list only includes the transcripts for which cloning and real time RT PCR assays were designed, as these are mentioned several times throughout the text. While many other g enes were included in the microarray analysis, these are identified in the text.
17 PG primary growth oocyte PS Pelican Shoal, an offshore site in the Florida Keys RPL32 ribosomal protein L32 ( large subunit) RT PCR reverse transcription polymerase chain reaction SR Sombrero Reef, an offshore site in the Florida Keys Stard7 StAR related lipid transfer ( START ) domain containing 7 TepII thiolester containing protein II TI Tingler Island, a nearshor e site in the Florida Keys UF University of Florida VTG vitellogenin
18 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE ROLE OF ZIN C AND COPPER ACCUMULATION IN QUEEN CONCH Strombus gigas REPRODUCTIVE DEFICIENCY AT NEARSHORE SITES IN THE FLORIDA KEYS By Daniel James Spade August 2011 Chair: Nancy Denslow Major: Veterinary Medical Science s The queen conch ( Strombus gigas ) populati on in the Florida Keys has been depleted since at least the 1980s and is recovering slowly. C onchs found in ne arshore (NS) aggregations lack gonad development compared to offshore (OS) conchs and are not known to reproduce T he nearshore environment app arently interfere s with conch reproduction, as translocation of NS conchs to OS aggre gations leads to increased ga metogenesis In 2007 field studies, m ean zinc concentration in the digestive gland of male and female NS conchs was higher than OS and there was a non significant trend toward elevated copper NS. Both Zn and Cu can inhibit gastropod reproduction and so were hypothesized to be causative agents in NS conch reproductive dysfunction A cust om oligonucleotide microarray identified decreased expre ssion of genes related to spermatogenesis and small GTPase mediated signal transduction in testis In ovary, differences were observed in apoptosis and translation related genes suggesting atresia of NS ovaries. In the digestive gland expression of li pid metabolism genes wa s
19 altered NS, possibly limiting oogenesis Vitellogenin (VTG) mRNA expression was much higher OS than NS by real time RT PCR. A 2009 study confirmed that trends in digestive gland Zn values and VTG mRNA expression are pers istent A lgal (food) metal concentrations overall were low but Cu was higher NS, while Zn did not differ significantly In vivo exposures of the surrogate Strombus alatus to Cu and Zn were used to test the hypothesis that exogenous Zn and Cu exposure causes conch reproductive dysfunction. Conchs exposed to metals accumulated the metal in the digestive gland and showed ovarian atresia more often than control; Zn exposed conchs exhibited an early peak in vitellogenesis, though these trends were not significant Ove rall, there is an association between Zn accumulation and reproductive dysfunction in NS queen conchs. However, in vivo exposures indicated that metals alone may not cause the reproductive dysfunction observed NS. This may indicate that Zn and/or Cu are part of a suite of stressors that interfere with NS conch reproductive development. These results will help Florida wildlife managers to identify the healthiest conch aggregations for broodstock or translocation in efforts to restore the conch population in Florida.
20 CHAPTER 1 LITERATURE REVIEW Background The q ueen conch ( Strombus gigas ) is a benthic marine gastropod of ecological and economic importance to the Florida Keys and throughout the greater Caribbean region While Florida once supported a thri ving queen conch population and both industrial and recreational conch fishery activities, the conch population is now heavily depleted, presumably due in part to past overfishing  Florida queen conch harvest peaked in 1966, and r estrictions on the harvest began in 1977  A complet e moratorium on the fishery was enacted in 1985  followed by listing under the Co nvention on International Trade in Endangered Species (CITES) in 1992  Despite harvest protections, the queen conch population in the Florida Keys has failed to recover to the desired level [3,5] though it is believed that the rate of recovery has increased since 2000  While human fishing pressure is the greatest cause of adult queen conch mortality  and therefore has contributed to depletion of conch stocks, queen conchs in aggregations at near shore (NS) sites in the Florida Keys also suffer from reduced seasonal development of gonad tissue, and it is believed that no reproduction occurs in the NS aggregations while conch s residing in offshore (OS) aggregations reproduce successfully [1,3,8,9] The NS and OS habitat patches are separated by the deep w ater Hawk Channel, which has a silt substrate that is believed to act as a barrier to adult and juvenile passage; the result is a geographic separation between NS and OS sites [1,3,10] and a significant disparity i n reproductive output between conch aggregations in the two areas.
21 Several major observations led to this hypothesis. First, Glazer and Quintero  monitoring efforts, NS conchs had not been observed reproducing, and that a small (n = 7 conchs total) sample of OS and NS conchs collected in May and Ju ne 1996 showed a uniformly greater degree of gonad development, assessed histologically, OS than NS. Second ly a large scale reciprocal translocation study (n = 576 conchs total) perfo rmed in 1999 found no observations of copulation or egg laying in NS co nchs, but both behaviors were observed in OS conchs. Translocated OS conchs reproduced at lower rates at NS sites, relative to OS conchs at OS sites, and translocated NS conchs were observed laying egg masses at OS sites, indicating that the OS environmen t supports reproduction, while the NS environment does not, and further indicating that the factor inhibiting reproduction in NS conchs is transient and reversible  Finally based on the same translocated conchs it was reported that NS conch gonad tissue is often immature or regressed, and is made up of a smaller percentage of gametogenic follicle tissue, during spring, summer, and fall se asons when OS conchs have mature gonads capable of supporting reproduction, and that conchs translocated from NS to OS show greater rates of gonadal maturity and a greater proportion of gametogenic tissue in the gonad than NS conchs remaining at NS sites, after only three months  This indicates that the NS environment in some way prevents conch reproduction by inhibiting development of the gonad in both male and female conchs. Therefore, it is critical to identify stressors present in the NS area tha t could interfere with conch reproduction To that end, t he aim of this dissertation project was to determine the degree to which queen conch reproduction in NS aggregations within the Florida Keys
22 might be inhibited by elevated l evels of copper (Cu) and zinc (Zn) two trace metals that are known to interfere with reproduction in several gastropods. Significance of Queen Conch in the Florida Keys Economic Significance of Queen Conch in the Florida Keys Queen conch is valued for both its shell and its meat throughout the Caribbean  and conch f isheries are worth millions of dollars to nations that support a sustainable harvest. However, overharvested conch fisheries are a common problem [2,11] The value of queen conch fisheries throughout the Caribbean was estimated at $40 million USD in 1994  and this value is believed to be second only to spiny lobster in Caribbean fisheries  I n the Bahamas alone the conch fishery has an estimated annual economic impact of $4.457 million and 9, 800 seasonal jobs  However, fishing pressure on conch populations increased with increasing conch ex ports in the 1970s and 1980s [2,7] leading to depletion of stocks even in countries with very large conch harvests, such as Belize  The commercial conch fishery in Florida was never as productive as in some other Caribbean l ocations  Still, a s illustrated by the impact of the conch fishery on Caribbean economies, a queen conch fishery could have a positive impact on the economy of the Florida K eys, in Monroe County, Florida. Ecol ogical Significance of Queen Conch in the Florida Keys Queen conch is also an important species from an ecological standpoint. Conchs are herbivores consuming algae preferentially  including Batophora [14,15] but also consuming seagrass and detritus [2,16 18] While algae are believed to be the major carbon source at least for juvenile conchs  and conch and Aplysia grazing significantly reduced Laurencia algal biomass in an experimentally nutrient enriched environment  conchs can also significantly impact on detrital biomass  A lso, a s
23 a large, benthic invertebrate associated with coastal and back ree f environments, queen conch has the capacity to serve as a sentinel organism for effect s that could be contributing to the health or decline of the ecosystem. The health of c oral reefs in Florida has been investigated, and the corals are known to be subje ct to temperature related bleaching  Moreover, human impacts on corals are suspected to be related to the reef decline, as evide nced by the discovery that Florida corals are contaminated with bacteria and viruses found in human fecal matter  Additional information about the stressors contributing to reproductive failure in queen conch could potentially bolster a whole ecosystem understanding of Florid coastal ecosystems. Queen Conch Life History In order to understand the effects that a toxicant might have on the conch population from a whole life cycle perspective, it is necessary to understand the life history of queen conch. Adult con chs copulate and lay egg masses that contain an approximate 400,000 [18,23] to 700,000  eggs ; average number of egg masses per female per season has been estimated at 9.4 based on an enclosure study  or up to 25 in a separate field study  Therefore the literature suggest s that queen conchs are highly fecund Known cues for reproduction are the seasonal parameters of temperature and photoperiod, which corre late with egg laying according to studies performed in Venezuela  and the Bahamas  Temperature also appeared to drive reproduction in an aquaculture experiment  The length of the reproductive season varies, and was reviewed by Stoner et al.; the season runs at least from June to September, but in most locations is longer  Veliger development and torsion b egin in the conch egg capsule which hatch after  After hatching, conchs undergo
24 metamorphosis through three pelagic veliger stages (two lobed, four lobed, and six lobed) over the course of approximately three weeks [4,29] beco ming competent for metamorphosis around 18 to 26 da ys after hatch when the larvae can be induced by algal or detrital cues, according to Davis and Stoner  However, the duration of the pelagic veliger stage may be variable, as up to the swim crawl stage lasting approximately 52 days  The veliger settles to the benthic substrate as it undergoes metamorphosis all of the larval source and sink populations are located with respect to the a dult population in the Florida Keys  There is evidence to suggest spora dic recruitment of larvae from the western Caribbean Sea [31,32] but this has recently been disputed by Delgado et al.  who argue that the Florida queen conch population is lar gely dependent on recruitment of larvae from Florida Specifically, they point to higher larval densities in the Florida Keys than Dry Tortugas or Florida Straits during the peak of the reproductive season (June), at causes drift vials released upstream of the Florida Keys to bypass the Keys and be recovered upstream. Drift vials are floating scintillation vials meant to mimic the planktonic drift of larvae. An allozyme study of Caribbean conch populations from ni ne sites not including Florida indicated that there is a high degree of gene flow among the populations, but they remain genetically distinct  A second allozyme study conducted by Campton et al.  compared queen conchs from Florida and the Bahamas using allozyme frequencies at ten loci, including several of those assayed in the previous study. T heir results again indicated that genetic similarity among populations was high despite significant differences at
25 some loci. Therefore, queen conchs in Florida and the rest of the Caribbean are genetically similar but not identical, indicating some gene flow caused by larval dispersal b ut also some reliance on local larval sources. After larvae settle, they develop through the juvenile stage to the sexually mature adult stage, characterized by a flaring shell lip  Reproductive development begins as the conch reaches adult length, at or before the time that the lip develops, and ends after the lip is completely formed and has thickened [35,36] Age at sexual maturity ha s been estimat ed at 3. 6 to 4 years in a wild population in Puerto Rico  ba sed on an assumed minimum lip thickness of 4 mm, which was reported to be the minimum size at in a p revious study  in Belize; n ote that this was the minimum for males, while the minimum for females was 6 mm. In a Colombian population, the lip thickness at whic h 50 % of samples showed mature gonadal histology was 13 mm for males and 17.5 mm for females  The gonad histology of the different stages of queen c onch reproductive development has been described in detail by Egan  and Avila Poveda and Baquiero Cardenas  The two authors use slightly diff erent terminology to describe the appearance of resting stage (sexually immature) conchs, maturing gonads marked by increasing size of follicular tissues and increasing presence of mature gametes ng mature gametes, and the normal post mating appearance of male and female gonads. Egan also describes the connective tissue present in all gonad sections, which makes up the majority of tissue in im mature gonads, as signet tissue; signet tissue contains albumen and is likely critical for the energy requirements of gonad development  In addition to normal
26 histology, t he influence of the nearshore environment of the Florida Keys on gonad histology was described by Glazer and Quintero  and Delgado et al.  who added information about the appearance of abnormal regression in some NS gonads. Glazer and Quintero  also argued that NS signet tissue had reduced polysaccharide levels, based on p eriodic acid Schiff (PAS) stain ing which identifies carbohydrates in tissue sec tions  Mortality is assumed to be much higher for larval and juvenile conchs t han adults. Stoner and Ray  demonstrated that predation is higher for juvenile conchs outside of aggregations than within aggregations. Randall  lists various species of gastropods, cephalopods, crustaceans, and fishes that have been reported to predate upon both adult and juvenile conchs. Weil and Laughlin also state that high rates of mortality in juveniles can be caused by both predation and in shallow waters desic cation  However, Berg and Olsen  review several studies of conch mortality, and determine that natural juvenile and adult mortality can be estimated at M = 0.1, or an annual survival rate of slightly greater than 90%. It follows logically, then, given the high rate of survival of adult conchs and the likelihood that juveniles will survive to reproductive maturity, that reproduction and/o r recruitment (survival and settling of larvae) would be major influences on the population growth rate for conchs. However, fishing mortality is often significantly greater than natural mortality  and as previously mentioned, many conch populations are overfished [2,11] C onchs exhibit considerable movement and migrati on throughout their life cycle Juveniles generally undergo a migration from shallow to deeper water  and this could be to take advantage of different habitat type s or food sources as burying and
27 feeding behavior change with maturation  However, it does not appear that an OS area is necessary for the act of reproduction, as there have also been reports that conchs migrate t o shallower water close to shore in the summer months in order to reproduce [23,24,41] or migrate between patches of differing substrate types for feeding and breeding  In the Florida Keys, Glazer et al. report that conchs seek out coarse sand substrate during reproduction  It is possible that in many locations, conchs migrate twice annually to shallow, sandy NS habitats in t he summer for breeding and to deeper OS seagrass beds in the winter and that this may be related to differences in optimal feeding and breeding habitats. Queen Conch Demographics and Reproductive Stressors in the Florida Keys The queen conch population ha s been monitored regularly since closure of the fishery. From 1987 to 1990 in early monitoring studies both juvenile and adult populations appeared to decrease  or to vary by substrate type  though the large monitoring area and sparse, aggregated distribution were cited as problems for making population size estimates. Juvenile data were more variable than adult data [5,10] which is reasonable giv en the large size of juvenile aggregations that have been found in some conch habitats  as well as the burying behavior and migration exhibited by juvenile conchs  Recent estimates suggest that adult queen conch population size in the Florida Keys has increased to an estimated 27,000 individuals by 2001, from a lowest observed estimate of 5,750 in 1992, according to transect data collected by the Florida Fish and Wildlife Conservation Commission (FWC)  and may be increasing at a greater rate since 2000  Stoner and Ray briefly review population density estimates from various sites in the Caribbean, and Florida is close to the lowest reported densities, with 0.50 adult conchs/ha reported in 1987 88; this is much lower
28 than the 130 1582 adult conchs/ha reported in 1972 1974 in Nearby Cuba but similar to the 0.52 adult conch s /ha reported in Bermuda in 1988  Given all of this information, i t appears that the Florida Keys conch population is st ill significantly smaller than what would be necessary to support a fi shery but has begun to recover slowly  Queen conch populations are known to be subject to the Allee effect, or inverse density dependence, which means that population growth rate i ncreases with increasing population density, at least to a certain point  probably due to reduced reproduction at low density  Given this information, and given the observed reproductive impairment in the near shore dispers e across Hawk Channel and onto the reef tract, managers are interested in supplementing the density of mating aggregations OS by translocating individuals from NS to OS [1,3] Translocation experiments have led to successful reproductive development, assessed histologically, and to observations of but not observations of copulation [3,9] ; further, translocation of OS conchs to NS sites results in re duced rates of spawning, relative to OS conchs that are captured and re released at OS sites  In addition to translocation, mariculture has been attempted as a strategy for wild queen conch stock enhancement, but was ultimately unsuccessful. While conchs have been cultured from egg masses in recirculating saltwater [4,47] nave hatchery reared ind ividuals are subject to greater predator induced mortality  Though this could be mitigated to some degree by predator exposure in culture, the cost of a ma riculture program is expensive, estimated at approximately nine dollars per indi vid ual 
29 While translocation of NS conchs to the OS environment results in development of normal gonad tissue and reproductive activity within three months  this says little about the reas on for the failure of near shore conchs to develop gonad tissue. I t appears that some factor in the NS environment currently precludes reproduction ; this is supported by the impact of the NS environment on spawning in the reciprocal transplant study by McC arthy et al.  Therefore, a goal of this project wa s to identify possible stressors that could contribute to the failure of near shore con chs to develop gonad tissue and subsequently to successfully reproduce. This information could be vital to the restoration of a thriving S. gigas population in south Florida. A number of chemical and physical stressors that might contribute to the lack of NS conch gonadal development have been considered. T he UF Analytical Toxicology Core Laboratory quantified common organic contaminants including pharmaceuticals and pesticides in queen conch tissues and sediments from the sites of interest in the Florida Keys While organic compounds such as ethinylestradiol were detected in samples, there were no significant gradients from near shore to offshore t o indic ate that those contaminants might contribute to the problem as outlined in a final report to the US E PA by Glazer et al.  Water chemistry data were of obvious interest because of the influ ence of temperature and photoperiod on reproduction. Also hypoxia has recently been shown to cause effects similar to endocrine disrupting chemicals in fish such as the Atlantic croaker Micropogonius undulatus  and carp Cyprinus carpio  However, there is little evidence that temperature, dissolved oxygen, or other water quality parameters are significantly correlated with the observed effect on conch reproduction. In my own analysis of data collected by the Southeast Environmental
30 Research Center at F lorida International University, it appears that there is a slight decrease in dissolved oxygen over the years studied  However, I could find no obvious trend in temperature or dissolved ox ygen from nearshore to offshore. I do presume, t hough, that bottom water temperatures are lower and probably less variable at deeper OS sites. Copper and Zinc Exposures Lead to Reduced Reproductive Success in Gastropods Cu and Zn effects in gastropods have been studied relatively extensively. A number of studies have demonstrated adverse sublethal health effects of excess Cu and Zn on gastropods, including effects on growth and development, and also on reproduction. A study on two subspecies of Helix aspersa demonstrated that Cu and Zn bioaccumulate re adily, with the largest values measured in the foot and the viscera, respectively  suggesting that gastropods in general are likely to take up excess Cu and Zn, which might subsequently produce toxic effects. Similarly, Gimbert et al.  describe Zn uptake from contaminated soil by Helix aspersa noting that Zn accumulates in the granular and cell debris fractions of the viscera (digestive gland and gonad) over an 84 day exposure with 17 fold higher concentrations than in the foot and other soft tissues; further, in this experiment, very little Zn was eliminated during an 84 day depuration phase Gomot DeVa u fleury and Pihan also describe the uptake of Zn and other metals in Helix aspersa from contaminated soil, with accumulati on in the visceral mass being greater than the foot  Not only do metals ac cumulate in gastropods, but there is evidence that metal exposures cause a generalized stress response, as evidenced by effects on enzymes related to xenobiotic biotransformation and neurological function. Cu treatment affect s
31 Ex vivo activities of biomar kers including catalase (stress response increased ) glutathione S transferase (xenobiotic biotransformation increased ) and acetylcholinesterase (neurological function decreased ) in Hexaplex trunculus  Sim ilarly Helix aspersa nucleotidase (nucleotide catabolism decreased ) but not ATPase (cellular energetics) or c holinesterase (neuromuscular function) activity was a ltered in vitro by Zn while Cu only decreased ATPase activity in the same experiment  The sum of sublethal stresses can have effects on energy budgets and in fact Moolman et al. described a negative effect of Zn on energy stores in Melanoides tuberculata and Helisoma duryi  Development is of interest for life cycle toxicity studies, especially in gastropods, which have rather complex life cycles involving metamorphosis and multiple life stages which can be interrupted by exposure to Cu and Zn. Cu and Zn are both toxic to early life stage s of abalone Haliotis rubra causing both death and effects on morphological development  B oth Cu and Zn caused mortality, an d also affected growth in a feeding study with juvenile Helix engaddensis ; the authors of this study found the growth inhibition of Zn, unlike Cu, to be irreversible  In a similar feeding study, Cu and Zn both reduced growth rates in juvenile Helix aspersa with EC 50 values of 1200  Finally, one study of Cu has been performed with juvenile queen conch s; the findings included decreased feeding and excretion rates as well as increased righting time, an effect on behavior that could stem from the energetic impact of metal exposure  Given these sublethal effects of Cu and Zn exposure it is logical that Cu and Zn might a lso affect fecundity and population growth rate. Metal effects on the abundance
32 of marine organisms, which could be influenced by effects on reproduction, have been given some attention in the literature. Stark  determined that 56 day exposure to Cu EDTA in water (20 and 60 ppm treatments) or in a plaster containing 50 g Cu EDTA inserted into sedi ment reduced numbers of naturally occurring gastropods in sediment samples and observed a similar effect on bivalves and polychaetes The use of Cu EDTA plaster was meant to mimic the gradual contamination of soils with Cu in the field. In a related fie ld study, the author found that gastropods were more abundant in bays with l ower concentrations of Cu, Pb, and Zn  These findings are consistent with a field study conducted by Rygg  that indicated that sediment metal concentrations are negatively correlated with abundance of sediment animals with the strength of association in the order: Cu>Pb>Zn. In all of these studies the observed effect on gastropod abundance could have been caused by lethality, reduced reprod uction, or both, or in the field by some mechanism of contaminant avoidance. Ravera  briefly reviewed impacts of heavy metals, including Cu and Zn, on freshwater pulmonate gastropod reproduction in 1991 A notable theme of this review was the variability of e ffect doses across species and metals. In spite of this variability, fertility, fecundity, or maturation of pulmonates was often impacted at Zn concentrations Since the time of that review additiona l studies have been identified impacts of Cu and Zn on gastropod reproduction. Laskowski and Hopkin  demonstrated that exposures to Cu and Zn, as well as Pb and Cd, reduced fecundity of Helix aspersa a sexually reproducing hermaphrodite (EC 20 for Cu and Zn = ; the treatments also reduced feeding rates, which may have affected reproductive output L ow dose aqueous Zn exposures were
33 shown to reduce both fecundity (egg laying reduced by 500 ppb Zn 2+ ) and fertility (egg hatching eliminated by 1000 ppb Zn 2+ ) in Biomphalaria glabrata another sexually reproducing hermaphrodite, by Munzinger and Guarducci  In another reproductive study, Ducrot et al.  model ed, using the dynamic energy budget, the negative effects of Zn on reproduction in Valvata piscinalis yet another se xually reproducing hermaphrodite, finding a reduction in fecundity at 628 to 993 mg Zn/kg sediment, depending on the duration of the study and the method of determining effect concentrations and an ultimate decrease in population growth rate However, th is effect might not be universal; in one study by Co e urdassier et al.  exposure to an effluent containing Fe, Cr and Zn resulted in bioaccumulation of Cr and Zn, but with unclear effects on fecundity in Lymnaea pal ustris another sexually reproducing hermaphrodite, as fecundity was reduced by som e effluent treatments, but not those with the highest metal concentrations. This highlights the complex ity of mixture toxicity studies and the difficulty of attributing an effect of a mixture to a single stressor. ing with respect to metal exposures in Lymnaea was that Cr and Zn were bioaccumulated, while Fe was apparently regulated. Several studies showing effects of copper or copper compounds on gastropod repr oduction have been conducted in addition to the aforeme ntioned study by Laskowski and Hopkin  Copper oxychloride (Cu 2 Cl(OH) 3 ), a fungicide, decreased oocyte production in Helix aspersa in a dose dependent manner at two sublethal concentra 2 Cl(OH) 3 /g food), and this was coupled with a dose dependent accumulation of Cu in the ovotestis, in an experiment by Snyman et al.  Dorgelo et al.  showed that e xposure to 30 Cu in lake water resulted in
34 decreased offspring production in Potamopyrgus jenkinsi a n ovoviviparous and parthenogenic gastropod Real et al.  and hatching i n Stagnicola vulnerata Finally, in a study by Rogevich et al.  a 9 month exposure to waterborne Cu reduce d fecundity in the Florida apple snail Pomacea pal udosa which is dioecious, at low environmentally relevant concentrations of 8 and ; Cu also accumulated in the soft tissues As with Zn the effects of Cu on reproduction might not be universal across the gastro pods in all exposure conditions In a 36 day experiment reported by Pea and Poscidio  sublethal Cu concentrations of 20 to 67.5 did not affect Pomacea canaliculata r eproduction T he weight of the evidence suggests that Cu and Zn are important stressors that affect growth, development, energy availability, and reproduction in gastropods. Further, reproductive effects have been documented in gastropods across several different reproductive strategies. It should be noted that all of the studies of Cu or Zn effects on reproduction rep orted in the literature involve aquatic or terrestrial gastropods, while marine gastropods have n ot been a major focus, despite reports that metals in marine sediments are associated with reduced numbers of gastropods and bivalves [63 65] However, the inclusion of both feeding studies and aquatic exposure stu dies indicates that reproductive effects of Zn and Cu can result from oral or aqueous routes of exposure. The oral route of exposure, which is more likely in the marine environment, is believed to be the major route of exposure for many animals in nature according to Croteau and Luoma  and is the primary route Zn uptake for the marine gastropods Nassarius teretiusculus and Babylonia formasae habei 
35 Several other reports cover accumulation and e ffects of Cu and Zn on marine gastropods, regarding endpoints other than reproduction. Gay and Maher  argue that the marine gastropod Bembicium nanum can be used as a bioindicator for contamination of marine ecosystems with metals including Cu and Zn, because the snail accumulates metals according to environmental concentrat ions, though factors such as size and season of collection also influence metal concentrations. Taylor and Maher  argue that Bembic i um auratum and Austrochochlea constricta Cu and Zn concentrations are consiste ntly higher in gonad and digestive gland, with some variation over time that cannot be explained by reproductive state, gender, salinity, or temperature, and so these marine gastropods would make appropriate biomonitoring species. Nicolaidou and Nott  found species differences in the accumulation of metals from a smelting plant in se veral marine gastropods, with Cerithium vulgatum accumulating all metals in the study except for Cu, including Zn; notably, in several species most metals, including Zn and Cu, accumulated to a greater concentration in viscera than foot. Ying et al.  however, found that Cu and Zn were apparently regulated not accumulated by Polyni ces sordidus and so this species was not an appropriate bioindicator of environmental Cu and Zn concentrations. Tsai et al.  found that Zn reduced growth of abalone Haliotis diversicolor with a NOEC of 65.2 Again, in Strombus gigas Cu exposure has been shown to affect feeding and behavior  Liao and Lin  found that Zn was toxic to abalo ne Haliotis diversicolor supertexta with 96 h LC 50 = 1.2 mg Zn/L. It should be noted that Zn and Cu have numerous physiological roles most of which have been best characterized in mammals, but probably many of which are
36 conserved across species. Well est ablished roles of Zn from the mammalian literature include function in metalloenzymes [84,85] in cell membranes  as an antioxidant  and as an inhibitor of apoptosis  Cu is also an important element in the function of some metalloenzymes [84,89] with a particular importance in respiration  I n some gastropods, including Strombus gigas the blood pigment hemocyanin represents another important physiological need for Cu [91 93] Because of the important p hysiological roles of Cu and Zn and the potential for toxicity of excess metals they are heavily regulated by a suite of transport proteins, many of which are inducible by transcription factors that respond to metal concentrations in tissues  While many of the transport mechanisms for trace metals are also likely to be conserved across species, they are not as well characterized for gastropods as for mammals. Still, gastropods have systems for r egulation of trace metals, but those systems differ from mammals. While some of the aforementioned papers argue that a marine gastropod species does not accumulate Zn or Cu because of regulatory mechanisms [80,81] still others argue that high levels of exposure to one or both of the metals will lead to accumulation [53 56,70,74,76 78,80,83] possibly because of the normal function or overwhelming of storage mechanisms relate d to de toxification. When accumulated. Zn and Cu tend to accumulate to a greater degree in viscera than foot or other tissue s [53 56,74,79,80] with the exception that Cu accumulates to a greater degree in the foo t than viscera of two Helix aspersa subspecies  It should be noted that some studies of small gastropods include visceral tissues such as parts of lysis, [53
37 55] Elevated levels of metals in the viscera can be explained by the roles of the kidney and digestive gland as sinks fo r storage and detoxification of trace metals, a phenomenon that has been thoroughly described in Lymnaea stagnalis [94,95] Therefore, exposure to high exogenous doses of Zn or Cu could cause toxicity or sublethal effects such as effects on reproduction or growth, and also accumulation of the metal, probably in the digestive gl and, in a generalized marine gastropod, with differences likely to exist among species. Copper and Zinc in South Florida Sources of copper an d zinc exist close to shore in the south Florida though no complete survey of metal concentrations in Florida Keys media has been published A report prepared for the Florida Keys National Marine Sanctuary Water Quality Protection Progra m notes elevated concentrations of zinc and copper, as well as lead and mercury, in canals in the Florida Keys, which the author associated with boat traffic  A doctoral dissertation submitted to Rice University in 1 97 5 provides prob ably the most in dept h survey of metal concentrations in u pper Florida Keys sediments; findings included elevated concentrations of several metals, including Zn, around Key Largo, Largo Sound and in Biscayne Bay areas under the influence of high human population density, can al systems and wastewater. The author cites a background Zn level of approximately 2 ppm throughout the study area, with some sediment samples reaching as high as 100 ppm  It should be cautioned that sources and fluxes of metals can change considerably over time, and could be different today than they were in 1975; however, the association between human activity and release of metals, including Zn into the environment is likely persistent.
38 In more recent studies, Carna han et al. found that sediment concentrations of Cu and Zn, as well as Ni, Pb, Hg, and Cr, in Biscayne Bay, north and east of the F lorida Keys, are greate st near the city of Miami and where canals empty into the Bay  A survey of some south ern and western Florida water and marine sediment samples found that some contaminants including Zn, were more concentrated in seagrass sediments than elsewhere  which is interesting because of the p reference of queen conchs for seagrasses such as Thalassia testinudum and Cymodocea manatorum over other substrate types  at least during the reproductive season [24,41] Finally, Caccia et al. reported elevated concen t rations of copper and zinc in marina sediments relative to levels throughout the Florida Bay  While this last study covers the Florida Bay, north of the Florida Keys, rather than the area inhabited by queen conchs, it again associates boat traffic with elevated metal levels. Additionally, copper is currently a st ressor of interest for a number of south Florida terrestrial species, most prominently the Florida apple snail Pomacea pa ludosa [74,101] High levels of copper contamina tion in the Florida Everglades have resulted in accumulation of copper in apple snails, potentially threatening its predator, the Florida snail kite Rostrhamu s sociabilis plumbeus  While Cu and Zn effects on conch reproduction have not been studied, th e aforementioned juvenile S. gigas study demonstrated negative effects of Cu on feeding and excretion rates in juvenile conchs  No published stu dies have addressed effects of zinc in species of interest to south Florida, but it is clear that both copper and zinc can cause reproductive to xicity in gastropods in general, and that sources of both metals including canals, boats, agriculture, and muni cipal wastewater, are present in south Florida.
39 Related Work with Queen Conch It should be noted that some previous work has been conducted with queen conch with respect to m etal effects and especially with reproduction two area s of emphasis in this di sse rtation. Gene expression is the major methodological foundation for the work described in this dissertation. While no genomic or transcriptomic work has previously been conducted with conchs except for the development of the microarray used for this pro ject several genetic studies have been published and should be mentioned. Background on reproductive development, another area of emphasis for this dissertation, has already been discussed. Queen Conch Genetics Little sequence data for queen conch exist ed prior to undertaking of the sequencing project that laid the foundation for this dissertation. However, conch genetics had been studied to some degree. A molecular phylogeny of strombids was published by Latiolais et al., based on sequence distances i n cytochrome oxidase and histone H3 genes  Ad ditionally, m icrosatellite sequences have been identified to address the queen conch population ecology  but no extensive microsatellite analyses have been published. The two allozym e studies previously mentioned are the only queen conch population genetics studies in the literature [33,34] Production of an Oligonucleotide Microarray for Queen Conch A n oligonucleotide microarray for Strombus g igas was construct ed as described in Spade et al.  Briefly a normalized cDNA library was produced from pooled queen conch tissues by Robert Griffitt (UF). Tissue samples included gonad, digestive gland, foot muscle, and neural ganglia and were collected from Sombrero Reef on 9 June ock and E astern Sambo on 15 March 2005. The library
40 was sequenced on the 454 GS FLX pyrosequencer at the University of Florida Interdisciplinary Center for Biotechnology Research ( ICBR ) assembled and annotat ed by Li Liu at ICBR. Using the resulting sequ ence data, Nancy Denslow (UF) and Li Liu designed an oligonucleotide microarray This platform (GPL8934) was used for all queen conch microarray experiments in this dissertation. Metals in Queen Conch As previously mentioned, one study performed in 1984 d etermined that exposure of juvenile queen conchs to aqueous sub lethal concentrations of c opper, as low as 30 resulted in righting time, showing an effect of Cu concentrations on behavior, possibly mediated by energetic effects, with no effects on growth  Aside from this study, I do not know of any work with trace metal effects that has been published in queen conch. However, a recent study of copper, zinc, and le ad concentrations in sediments and edible queen conch muscle tissue was performed in Cuba : the authors concluded that, at least in muscle, bio accumulation of the metals was low, but that some lead levels exceeded Cuban public health limits, while some cop per levels exceeded the levels set by the former British Ministry of Agriculture, Fisheries, and Food  The reader should be cautioned, however, that in most studies gastropods accumulate metals to a greater degree in the viscera than in the foot or other soft tissues [53 55,74,79,80] Therefore, considerable questions remain with regard to both the potential for bioaccumulation and the possible effects of exposure to trace metals in nearshore conchs in th e Florida Keys.
41 Overall Hypotheses The goal of the work performed in this dissertation was to determine what relationship exists between met al accumulation, particularly Zn and Cu and development of gonads in NS conchs. These goals were achieved using ge ne expression and histological approaches toward understanding the differences between developing gonads and associated tissues such as digestive gland, coupled with analysis of metal concentrations in conchs and algal samples by inductively coupled plasma mass spectrometry (ICP MS). The overall hypotheses of this dissertation were: 1. That gene expression in nearshore conch testis, ovary, and digestive gland would explain variations in reproductive development between NS and OS, providing insight into the re sponsible stressors; 2. That metal concentrations, particularly Zn and Cu, in NS conch tissues would be higher than in OS tissues on a consistent basis; 3. That In vivo exposure of conchs to Zn and Cu would lead to reduced gonad development and patterns of gene expression similar to NS conchs. Significance of this Dissertation Studying the biology of reproduction, differences in gene expression, and the effects of metal contamination on reproduction in queen conch s has provided relevant data for the management of the species, as well as for the understanding of overall ecosystem health in the Florida Keys. Mollusks are commonly used as bioindicator species when studying the effects of xenobiotics  and meta ls in particular  This is the case because mollusks typically take up metals at rates that correspond with environmental concentrations  and also because exogenous contaminants often negatively impact molluscan reproduction  Because of the experiments performed in this dissertation we have a better understanding of the relationship between trace metal concentrations Cu and especially Zn and conch reproduction, and the
42 mechanisms that are important for conch reproductive development We have gained some understanding of the effects of Cu and Zn on fecundity in wild queen conchs, which will aid in the development of management strategies. Of equal importance, this work has provided some understanding o f conch reproductive biology in terms of physiological control, important signaling pathways and biological processes and this has helped to understand how stressors may interfere with these processes Therefore, t his information will be useful for manag ers who are attempting to promote the recovery of the queen conch population in Florida, and also interesting in the context of basic conch biology
4 3 CHAPTER 2 METHODS 2 S t rombus gigas Sample Collection Queen conch samples were collecte d and tissues proce ssed in 2007 sampling efforts by Robert Glazer and Gabriel Delgado of the Florida Fish and Wildlife analyses Sampling was conducted by free diving or SCUBA diving. Shel l length, shell lip thickness, total mass, and soft tissue mass were recorded at the time of collection. Shell mass was estimated by subtracting soft tissue mass from total mass. Only sexually mature adult queen conchs, identified by a fully flared lip [4,35 37] were collected. Tissue samples used for the comparison of reproductive (OS) versus non reproductive (NS) queen conchs by histology, real time reverse transcription polymerase chain reaction (real time RT PCR), and metals analysis were from conchs collected from Pelican Shoal (OS 4 male and 4 female conchs ) and Tingler Island (NS 4 male and 4 female conchs ) on 15 February 2007 and from East Sambo Island (OS 2 male and 5 female conchs ock (NS 4 male and 2 female conchs ; 1 female later determined to be male based on gonad histology ) on 7 June 9 June 2007. Testis, ovary, and digestive gland microarray analyses were conducted using 15 February 2007 samples from Pelican Shoal and Tingle r Island ( Figure 2 1 ) 2 Some of the methods reported in this chapter were previously published in manuscripts corresponding to Chapters 3 and 4. The references, as listed in the bibliography, are: 105. Spade DJ, Griffitt RJ, Liu L, Brown Peterson NJ, Kroll KJ, e t al. (2010) Queen conch ( Strombus gigas ) testis regresses during the reproductive season at nearshore sites in the Florida Keys. PLoS One 5: e12737. 108. Spade DJ, Knoebl I, Denslow ND (2011) Cesium chloride gradient centrifugation improves the quality of total RN A preparations from the gastropod Strombus gigas and the coral Montastraea faveolata J Exp Mar Biol Ecol 402: 43 48.
44 Along with FWRI, I designed and participated in an additional sampling effort in March, 2009. In 2009 a total of 22 adult conchs were collected from Delta Shoal (OS) s Rock (NS). Four male and 4 female conchs were c ollected from Delta Shoal (OS) and processed immediately on 11 March 2009. An additional 3 male and 4 female conchs were collected from Delta Shoal (OS) on 11 March 2009, and processed on 12 March 2009, after a 16 hour hold in NS flow through water. Four male and 3 processed on 11 March 2009, after a 28 hour hold in NS flow through water. No rainfall and no significant changes in temperature occurred over the span during which sampling and processing was performed. In all cases, a dult male and female conchs were collected live and transported immediately to the FWRI laboratory in Marathon, FL. Conchs were then euthanized and tissues were harvested. For molecular assays and de termination of tissue metal burdens, gonad, digestive gland, neural ganglia (buccal ganglia) blood, and foot muscle samples were frozen immediately in liquid nitrogen. For descriptions of the adult conch anatomy and precise locations of organs, see Littl e  Frozen tissue samples were maintained at 80 o C until further analysis. For histology, an approximately 1 cm 3 piece of gonad tissue was cut from the center of the gonad, fixed in 10% neutral buffered formalin for a minimum of 7 d ays and sent to Nancy Brown Peterson (University of Southern Mississippi) for processing and analysis The results of the histology analysis will be commented on as necessary to aid in interpretation of the work reported in this dissertation.
45 Algae Sample Collection During the queen conch collection conducted on 10 11 March 2009 15 algae samples were collected at each site (ESR, NS; DS, OS) using a randomized grid design. A grid was developed in ArcView GIS software, mapping the entire area of interest at each site with a 5m x 5m square grid. A random number generator (Excel function =RANDBETWEEN()) was used to select 15 sites from each grid by row number and column number in the grid. Latitude and longitude coordinates for each site were retrieved in ArcView, and sampling locations were found by GPS in the field. Divers collected algae in an acid rinsed ( 1 % HCl in de ionized water for 2 h, de ionized water rinse, air dry) 50 mL Corning polystyrene conical vial. All algae samples were immediately plac ed on ice and were frozen at 20 o C on return to the FWC lab in Marathon, then kept on ice until placing them in the 20 o C freezer on return to UF. Algae samples were later identified from photographs by Gabriel Delgado of FWRI. In V ivo Exposures of Stromb us alatus to Copper and Zinc Sources Care, and Maintenance of Conchs Adult Florida fighting conchs, Strombus alatus were donated from the conch aquaculture program at the Harbor Branch Oceanographic Institution of Florida Atlantic University (HBOI) or w ere collected from Cedar Key, FL, by Ben Olaivar (UF). In all cases, o nly adult conchs were collected, and all conch s had a flared shell lip, indicative of sexual maturity  Fighting conchs were housed in the Aquatic Toxicology Laboratory at the University of Florida (ATL). Prior to introduction of conchs, a system of three recirculating artificial seawater holding tank s w ere designed based on the design used at HBOI [27,110] with modifications The three approximately 380 L round fiberglass tanks were connected to a single sump of approximately 114 L and in line
46 pump. Water flowed into the tanks through spray bars located both above the w ater surface and below the undergravel support, and flowed out of the tanks through a central drain below the undergravel support, into an external standpipe that determined the height of the water column (approximately 44.5 cm) A layer of approximately 2 5 cm Caribbean crushed coral aragonite sand (CaribSea, Inc, Fort Pierce, FL) was used as su bstrate and biological filter. As opposed to the HBOI system, no external fluidized sand filter was used. Artificial seawater was produced using carbon filtered, dechlorinated Gainesville city water, treated at ATL, mixed initially with Red Sea Salt (Red Sea USA, Houston, TX), and later with ProLine Super Salt Concentrate (Aquatic EcoSystems, Inc., Apopka, FL) and fine solar salt (Morto n Salt, Chicago, IL), accordi ng recommendations, to produce salinity of ca. 33 Air was cons tantly bubbled into each tank using an airstone. To encourage biological filtration, the system was initially inoculated with FritzZyme TurboStart 900 nitrifying bacteria and dosed with approximately 1 g NH 4 + and 50 mg NO 2 as an aqueous solution of NH 4 Cl and NaNO 2 Total ammonia nitrogen (NH 3 N) and nitrite nitrogen (NO 2 N) were monitored using Instant Ocean test kits, and conchs were added to the tank only after no ammo nia or nitrite nitrogen were detectable. Salinity, pH, temperature dissolved oxygen nitrite and ammonia were monitored regularly when conchs were present in the holding tanks. Water was amended with fresh water to adjust salinity as necessary the surf ace was regularly skimmed, the substrate regularly cleaned using a net or siphon and when significant algae built up on the sides of the tank the system was cleaned thoroughly and water was replaced.
47 The initial diet used for holding and as control for exposures from 7 January 2009 until 17 September 2010 was a gelatin based diet developed and used at HBOI. This consisted of koi pellet food (Mazuri, PMI Nutrition International, St. Paul, MN), sea lettuce ( Ulva sp. ) and seawater, bound with gelatin  During the holding period, conchs were fed to satiation every 2 3 days, and uneaten food was removed the following day by net or siphon to prevent a rise in ammonia, bacterial growth, or other water quality issues. Preliminary In V ivo Exposures Two preliminary exposure experiments with Florida fighting conchs, Strombus alatus were used to determine the exposure regime for an i n vivo time course feeding experiment with Cu and Zn. All tissues from these studies were collected and stored as indicated for S. gigas field collections. First, a 7 day experiment was conducte d in which three female conchs were fed the control HBOI diet, and three female conchs were fed a diet supplemented with Zn as ZnCl 2 at the intermediate nominal concentration of 200 ng/mg using smaller, app roximately 75 L, exposure tanks The exposure ta nks each had an individual electric pump, sand substrate/biofilter, undergravel filter, and sponge filter. The sand in each tank was inoculated with sand from the holding tank prior to introduction of conchs, in order to promote biofilter activity. T issu es collected from this experiment were eventually used to test the efficacy of hybridizing labeled Strombus alatus cRNA to the Strombus gigas oligonucleotide microarray. Second, a 60 day In vivo oral dose response study was conducted, in which four male an d four female conchs per g roup were fed either control HBOI diet or diets supplemented with Zn, as ZnCl 2 at nominal concentrations of 20, 200, or 2000 ng
48 Zn /mg food or Cu as CuCl 2 2H 2 O at nominal concentrations of 2, 20, or 200 ng Cu /mg food. Zn level s in the control HBOI diet were determined to be too high for this study, measuring 87.95 ng Zn /mg, and were apparently quite variable, as a sample of the medium dose Cu treatment food, which was not supplemented with Zn measured 305.71 ng Zn /mg by ICP MS Further, following the study, conchs had uniformly high concentration s of Zn in the digestive gland, averaging 4268.75 ng Zn /mg for controls and 3991.65 ng Zn /mg in the high dose Zn group. Note that no baseline measurement was made prior to the beginni ng of this study, and therefore it was impossible to know whether this full burden was accumulated during the experiment. Also, these values may be under estimated, as the Zn concentrations in digested samples were sometimes >100 ppb, which may exceed the linear range of the ICP MS. Ultimately, the data from this experiment could not be interpreted with respect to Zn effects on reproductive development, and an alternative diet was required. Production of an Alternative Diet A second control diet was desig ned based on the metal exposure experiment with Helix aspersa that w as reported by Laskowski and Hopkin  with modifications. This diet contained (by dry weight) 28.4% dried, chopped spinach, 28.4% bran, 14.2% dried, ground carrot, 14.2% skim milk powder, 14.2% gelatin, and 0.01 % CaCO 3 The food was mixed in milli Q water, to an ap proximate dry/wet ratio of 26.0 %. The major modifications were addition of spinach to the recipe, and intentionally using less water to achieve a higher dry/wet weight ratio. Zn and Cu concentrations in this food were lower than the HBOI diet. An initial test batch measured 53.76 ng Zn/mg food and 4.06 ng Cu/mg food, and a batch of experimental control food used for the subsequent 50
49 day time course study measured 36.5 0 ng Zn/mg food and 4. 7 7 ng Cu/m g food. This diet was used for further maintenance and e xperimental feeding of conchs. 50 D ay In Vivo Time Course Exposure of Fighting Conchs to High Oral Doses of Copper and Zinc Fighting conchs used for the 50 day in vivo study were collected by dredging from Cedar Key, FL, by Ben Olai var in the UF Department of Biology and arrived to ATL on 12 July (7 conchs) 9 September (12 conchs) and 27 September 2010 (45 conchs) Conchs were not numbered until the arrival of the third group. Therefore, the conch s from 12 July and 9 September had been combined prior to numbering. The seven conchs from 12 July were fed with the HBOI diet for approximately two months, and could not be faithfully separated from the 12 conchs from 9 September. This will be discussed in the context of metal accumulation. A baseline sample was processed on 8 October 2010 ( 96 d relative to the start of the exposure experiment) consisting of 3 female conchs, which were euthanized and tissues were collected for histological and ICP MS analysis. Histological analysis indicated that the gonads two of these conchs were Reg enerat ing (see Table 2 1) meaning that they had likely ceased reproductive activity for the season while the third was not analyzed due to a lack of gonad tissue in th e histological section Therefore, in an effort to encourage complete regression, the ambient temperature was adjusted to give a tank water concentration in holding tanks of approximately 17 o C to approximate winter water temperatures at Cedar Key and held at that temperature from 25 October to 11 October 2010 to allow complete regression of gonads. The temperature was then increased incrementally until 11 January 2011 to reach an approximate water temperature of 25 o C, similar to mid spring water temperat ures at Cedar Key that would encourage reproductive development. Water
50 temperatures were chosen based on NOAA Coastal Water Temperature Guide data for Cedar Key, FL  The strategy was chosen based on a report that conchs at HBOI resumed reproduction in late winter with an artificial increase in tank water temperature [2 7] This is discussed further in Chapter 6. Conchs were moved to experimental tanks on 10 January 2011. For the 50 day time course in vivo feeding study, Florida fighting conchs S. alatus were hous ed in 75 L exposure tanks outfitted as previously des cribed, but with an additional airstone in each tank Beginning on 12 January 2011, c onchs were fed an average of 4.83 g (wet) food/tank*day, all of which was consumed on all but eight days of the study (days 1, 2, 4, 5, 6, 29, 34, and 35) Nominal food concentrations were: control (0 ng/mg added Zn or Cu), Cu 200 (200 ng added Cu /mg food as CuCl 2 2H 2 O), and Zn 2000 (2000 ng added Zn /mg food, as ZnCl 2 ). Each exposure and control tank contained 4 female and 2 male fighting conchs except for the 14 d Cu treated group, which contained 3 female and 3 male fighting conchs The males were present to encourage reproductive development and activity, but were not included in the analyses of metal accumulation or gonad histology, due to low sample size. Sampli ng days were 0 d (control group used as 0 d time point for control as well as treatments), 7 d, 14 d, and 50 d. Sampling was conducted prior to feeding or cleaning on sampling days, and so ca. 24 hours passed between feeding and sampling at each time poin t Conchs were euthanized and tissues to be used for gene expression assays, ICP MS, and histology were collected and maintained according to the methods described fo r the queen conch field studies, with one exception: from 14 d and 50 d conchs, the medi al ca. 1 cm of the complete digestive gland/gonad mass was removed, two thin transverse
51 sections were cut and fixed for histological analysis, and the remainder was used to obtain samples of digestive gland and gonad for ICP MS and RNA extraction, to impro ve confidence in obtaining the correct tissues. Note that S. alatus gonad is a much smaller tissue than S. gigas gonad, and some histological samples prior to the 14 d time point contained parts of visceral mass other than gonad. Therefore, if gonad was n ot present in the histological section, the sample was not included in further analyses of the gonad. Additionally, if the ICP MS analysis indicated unusual characteristics, such as a lower concentration of Zn than Cu in the r observed for a gonad sample confirmed histologically, it was removed from further analysis. No such problems were encountered for sampling of digestive gland, which is a large and easily identifiable tissue. During the exposure period, prior to feeding or cleaning, salinity, temperature, and dissolv ed oxygen were measured in each tank daily, and pH was measured in each tank on 38 of 51 study days (including day 0) Sediment was cleaned daily, using a net to remove residual food, feces, and detritus, aft er which conchs were fed. A thorough cleaning was first performed on day 9, after water appeared cloudy in one tank, indicating possible bacterial growth, and ammonia nitrogen in that tank measured 0.2 ppm on day 7, using an Instant Ocean test kit There after, a ll tanks were cleaned thoroughly 12 times during the study (once every 4.17 days, on average for the entire duration of the study ): all tank surfaces were scrubbed, sediment was cleaned by siphon, and approximately 75 % of the tank water volume was replaced. Water samples from each tank were collected in 50 mL polypropylene vials and frozen on an
52 approximately weekly basis and were later used to measure ammonia nitrogen and nitrite nitrogen, using LaMotte test kits 3315 and 3352, respectively. Th e samples that were tested were collected on days 0, 7, 14, 21, 28, 38, 43, and 49, and were collected 0, 7, 1, 4, 0, 1, 0, and 0 days, respectively, after a partial water change and cleaning. Nitrite (ppm NO 2 ) was calculated according to the manufacture unionized ammonia (ppm NH 3 ) was calculated according to the US EPA report number EPA 600/3 79 091  Freezing of samples for later ammonia analysis has been previously validated [113,114] and has been used in at least one recent large scale nutrient study  However, it should be noted that measurements of frozen samples could be variable, and that glass containers appear to give more consiste nt results than plastic. Source of Montastraea faveolata Tissue Samples All Montastraea faveolata samples were obtained under Florida Keys National Marine Sanctuary Research Permit no. FKNMS 2009 095 from experiments performed jointly at the United States Environmental Protection Agency (USEPA) Gulf Ecology Division and the University of Florida. Tissue from M. faveolata colonies cultured and used in laboratory experiments was snap frozen in liquid nitrogen and stored at 80 o C until being processed. Histol ogical Analysis For the queen conch field collections, gonadal histology was analyze d and scored by Nancy Brown Peterson, University of Southern Mississippi. For the fighting conch in vivo exposure studies, formalin fixed tissue samples were sent to Histo logy Tech Services, Inc., Gainesville, FL, where they were embedded in paraffin, mounted, and stained with haematoxylin and eosin. I analyzed histological sections of ovary and
53 scored them based on the apparent stage of development (scores discussed below ) as well as the percentage of gametogenic tissue in the section (0 25% = 1, 25 50% = 2, 50 75% = 3, 75 100% = 4) and presence or absence of atresia The histology of the queen c onch reproductive cycle has been described in detail by Avila Poveda and Ba quiero Cardenas  and Egan  using slightly different terminology. T he influence of the nearshore environment of the Florida Keys on gonad histology was described by Glazer and Quintero  Delgado et al.  and Spade et al.  Delgado et al.  and Spade et al. (Chapter 4)  use a terminology system based on Egan  but with additional stages to describe the processes of development a nd regression in greater detail Because of the variation in terminology, Table 2 1 presents a compar ison of terms used in the various studies. The terminology used in Spade et al.  is maintained here for Chapter 4. However, Chapte rs 5 and 6 use a simpler terminology adapted from the standardized terminology for fish gonadal staging proposed by Brown Peterson et al.  This seems to be the best terminology, and applies to the progression of conch gonads, despite having been developed for fish. This nomenclature system includes stages for D eveloping, studies) and R egressing animals, as well as Regenerating animals, which have concluded reproduction for the season. It does not, however, attempt to use multiple terms to describe the progressing stages of development, as these are difficult to different iate. In fact Egan  asserts that
54 Preparation and Purification of RNA from Tissue Samples Reagents and Solutions for RNA Extraction RNA STAT 60 reagent was obtained from Tel Test, Friendswood, TX, US A. Cesium chloride, diethyl pyrocarbonate, chloroform, isopropanol, Tris HCl, sodium acetate, glacial acetic acid, and SDS were obtained from Fisher Scientific, Pittsburgh, PA, USA. 0.5 M EDTA, pH 8.0, was obtained from Applied Biosystems/Ambion, Austin, TX, USA. Molecular biology grade ethanol was obtained from Sigma Aldrich, St. Louis, MO, USA. Ultrapure nuclease free water (Gibco) was obtained from Invitrogen, Carlsbad, CA, USA. 5.7 M CsCl; 3 M sodium acetate, pH 5.2; and TES (10 mM tris HCl/5 mM ED TA/1% w/v SDS) solutions were prepared according to the method of Kingston et al.  RNA Preparation Method A: Preparation of RNA Samples Using Guanidinium/Phenol/Chloroform Extraction and Isopropanol Precipitation Frozen tissue samples were handled in liquid nitrogen and immediately homog 60, using an IKA T10 Basic S1 handheld tissue homogenizer (IKA Works, Wilmington, NC, USA), except in the case of Montastraea faveolata samples, which were ground under liquid nitrogen using a Dremel tool with a grinding bit or by hand with a mortar and pestle and mixed in RNA STAT 60 with a vortex mixer. Samples were incubated at room temperature for 5 min, Vortex mixer and incu bated for an additional 2 min at room temperature. Samples were then centrifuged for 15 min at 4 o C and 20,817 g (14,000 rpm) in an Eppendorf 5417R microcentrifuge. The aqueous layer was removed by pipette to another tube, and extracted a second time with
55 centrifuged as before. The aqueous layer from the second extraction was removed to another tube, and RNA was precipitated in one volume of isopropanol (approx. 600 750 in the microcentrifuge by centrifuging for 20 min at 4 o C and 20,817 g (14,000 rpm). The resulting RN A pellet was washed twice in 75 % ethanol, dried for 5 min at room temperature, and reconstituted in ultrapure nuclease free water or RNA secure Resuspension Solution (Ambion/Applied Biosystems) (Figure 2 2 method A). This method will hereafter be referred to as RNA Preparation Method B: Preparation of RNA Samples Using a CsCl Gradient Centrifugation Step Frozen tissue samples were handled under liquid nitrogen, homogenized in RNA STAT 60, and extracted twice using RNA STAT 60 and chloroform as described for Method A top of the aqueous layer by pipette and was subsequently l ayered over 1.4 mL 5.7 M CsCl solution in a 2.0 mL polyallomer bell top Quick Seal ultracentrifuge tube (Beckman Coulter, Brea, CA, USA), using a 3 mL syringe and 19 ga needle. Sample tubes were sealed and centrifuged at 228,000 g (80,000 rpm) for 3 hours at 23 o C in a Beckman Coulter Optima TLX benchtop ultracentrifuge, using the TLA 120.2 rotor. M sodium acetate, pH 5.2, overnight at 80 o C. The RNA was pelleted in the microcentrifuge by spinning for 20 min at 20,817 g (14,000 rpm) and 4 o C. The pellet free water, a second precipitation was performed for 30 min at 80 o C, and RNA was
56 pelleted as before. The pellet was finally dried for 10 min at room temperature and free water (Figure 2 2 method B). This method will hereafter be Further Processing and Purification of RNA In the case of Strombus gigas diges tive gland RNA used for microarrays and all Strombus gigas and Strombus alatus samples used for real time RT PCR, RNA was l for or S. gigas testis samples used for microarrays and for all RNA samples used for real time RT PCR, an aliquot of each RNA sample was DNase treated Column cleanup was performed prior to DNase treatment for testi s, ovary and digestive gland RNA from 2007 field samples used for QPCR. Later, this process was reversed for the 2009 field samples and samples from the in vivo study, in order to ensure that no DNase was carried over from DNase treatment; samples perform ed well in real time RT PCR regardless of the order of these processes. RNA Quantification and Quality Analysis All RNA samples were quantified using the NanoDrop ND 1000 spectrophotometer (NanoDrop Tec hnologies, Wilmington, DE, USA) based on the A 260 mea surement Also, the spectrophotometric ratios A 260 /A 280 and A 260 /A 230 were measured on the NanoDrop ND 1000 to determine the purity of the RNA sample. A quality check was performed using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA) for all samples used to compare RNA preparation methods, and for all samples used in microarray analysis ; o nly samples that appeared to be high purity and intact (un degraded) were used for microarray analysis For RNA samples
57 used only for real tim e RT PCR, a random sampling of at least 8 RNA samples per experiment was assessed using the Agilent 2100 Bioanalyzer, and in all cases these samples showed high integ rity, with the possible exception of the NS ovary samples from February 2007; this is disc ussed further in Chapter 5. When applicable, intactness of RNA was assessed by the RNA Integrity Number (RIN) calculated by the Agilent 2100 software. RIN is a measure of RNA integrity ranging from a low value of 1 (degraded) to a high of 10 (intact), ba sed on a Bayesian model that incorporates characterist ics of multiple features of mammalian total RNA electropherogram s : 18S and 28S rRNA peaks, the small RNA area, the RNA marker peak, and regions in between  Because total RNA electropherograms differ significantly between invertebrates a nd mammals, this calculation could not always be made In particular, I have observed that undegraded conch total RNA samples show no 28S rRNA band; therefore, no RIN can be calculated to quantify RNA integrity for these samples. This is likely due to th Ishikawa  and that is present in gastropods including Hal iotis rufescens  For some Montastraea faveolata samples, RIN calculation was precluded apparently by the high ratio of 18S rRNA:28S rRNA, which is sometimes greater than 4.0. Despite the impossibility of cal culating RIN in some cases only high quality samples that appeared un degraded in Bioanalyzer profiles were used for further analysis. Microarray Experimental Processing Testis (n = 4 NS and 4 OS one of each group later removed due to outlying probes ) o vary (N = 3 NS and 4 OS) and female digestive gland (n = 3 + 1 technical replicate NS and 4 OS) microarray experiments were performed using Strombus gigas RNA from conchs collected during the 2007 field collections. An additional 2 S. gigas
58 mantle sample s, 2 S. alatus mantle samples, and 6 S. alatus ovary samples were labeled and hybridized along with the S. gigas ovary samples as preliminary experiments, and are reported in Chapter 3, as well as Appendix G S. gigas and S. alatus RNA sample us ed for microarray analysis was labeled with Cyanine 3 fluorescent dye labeled cytosine triphosphate ( Cy3 CTP ) following the Agilent protocol One Color Microarray Based Gene Expression Analysis (publica tion no. G4140 90040 ) which in the case of the ovary and digestive gland microarrays was modified for half volume reactions Montastraea faveolata RNA was labeled with Cy3 in two experiments. In the first conducted by Iris Knoebl (UF) four M. faveolata samples were labeled with Cy3 CTP, following the Agi lent protocol. A Pimephales promelas (fathead minnow, a teleost fish that has been used for numerous microarray projects) sample was included as a positive control, and a sample of M. faveolata RNA combined with P. promelas RNA was used to detect inhibiti on of labeling by components in the M. faveolata sample. In a separate experiment, I prepared M. faveolata RNA by Method B; this RNA was labeled with Cy3 by Yanping Zhang (ICBR), using the Ambion Amino Allyl MessageAmp II aRNA Amplification Kit, per the m samples, RNA quantity and Cy3 concentration were determined using the NanoDrop ND 1000 spectrophotometer. Microarray scanning and feature extraction was performed at ICBR using an Agilent G2505 B Microarray Scanner and Agi lent Feature Extraction Software (v9.5) All microarray data here reported are MIAME compliant; raw and normalized microarray data have been submitted to the GEO database (testis series GSE17379 ; digestive gland and ovary super series GSE29758 ), accor ding to MIAME standards 
59 Cloning of Strombus gigas and Strombus alatus Partial Transcripts DNase treated conch RNA sample s from tissues of interest (testis, ovary, or digestive gland) were pooled and reverse t ranscribed to produce cDNA, using Invitrogen SuperScript II Reverse Transcriptase and random primers, per the Primers used for PCR amplification were designed in the program Primer3  In the case of 18S r RNA, primers were designed based on alignment of 18S rRNA from the gastropod Bursa rana (NCBI accession # X94269.1) and the bivalve Nucula sulcata (NCBI accession AF207642.1) (Table 2 2 ). All other primers were designed from the 454 GS FLX derived S. giga s cDNA library sequences  Strombus alatus p utative vitello genin (VTG ) was amplified using primers designed for the homologous Strombus gigas sequence. Sequences of interest were amplified in a PCR reaction with Invitrogen Taq polymerase, according to the pGEM T Easy vector (Sigma Aldrich, St. Louis, MO, USA) and Invitrogen One shot Top10 chemically competent E. coli cells or One shot Mach1 T1 phage resistant chemically competent E. coli cells Three transcripts, EIF5A, RP L32, and VTG, were cloned with Ali Agha, an undergraduate whom I oversaw. All clone sequences were confirmed by Sanger sequencing at ICBR and sequences (Sanger sequence for 18S clone, 454 sequences for all others) were submitted to NCBI (accession s: Tabl e 2 2 ). Real Time RT PCR Real time RT PCR assays were performed on RNA samples from conchs used for microarray analysis, in order to validate parallel forms reliability between the microarray and real time RT PCR. All primers used for real time RT PCR wer e identical to pri mers
60 used for cloning (Table 2 2 ), with the exception of the 18S rRNA real time RT PCR assay. The 638 bp clone of 18S rRNA was used to design a second set of primers that was more optimal for real time RT PCR. Plasmids containing each c loned sequence were used to create a standard curve consisting of eight points in a serial dilution from 1x10 2 through 1x10 9 copies/reaction. Real time RT PCR was performed as a two s tep process. In step one, DNAse treated RNA was reverse transcribed usi SuperScript II reverse transcriptase and random primers protocol In step two, real time PCR reactions were run using SYBR Green Supermix (Bio Rad iCycle r real time PCR thermal cycler. All reactions were run using a two step protocol with an initial denaturation at 95 o C for 3.5 min followed by 40 cycles of denaturation at 95 o C for 10 s and annealing and extension at 58 o C for 1 min during which re al time quantification was enabled. Following the reactions, a dissociation curve was run beginning at 55 o C and increasing to 95 o C at 0.5 o C intervals every 10 s. Standards and experimental samples were run at least in duplicate, along with two negative c ontrols for treated RNA samples were pooled and water was used in place of reverse transcriptase during the ater was used in place of template cDNA during the real time PCR reaction. The specific parameters of reactions for different studies are given below. In testis, real time RT PCR assays were performed for 18S rRNA, Ctr1c, GST, Stard7, and TepII (Chapter 4 ) 18S rRNA was run on samples collected in February and June, 2007, and March, 2009 for the 18S validation experiment described below
61 All target genes ( all other than 18S ) were quantified only in samples collected in February, 2007. treated total RNA isolated from each testis sample was concentrated in a SpeedVac to achieve the necessary concentration, without drying, reverse transcribed, and the real time PCR reaction was run with cDNA equivalent to 100 ng total RNA for al l target genes, while the equivalent of 100 pg RNA was run for 18S rRNA, in order to keep values within the range of the standard curve. A pool RT control. 18S rRNA expression wa s consistent across samples, and so was used as the internal reference gene for target genes. In the digestive gland and ovary of conchs collected in February, 2007, real time RT PCR assays were performed for 18S rRNA, EIF5A, RPL32, and VTG (Chapter 5). 2 DNase treated total RNA isolated from each sample was concentrated in a SpeedVac to achieve the necessary concentration, without drying, and reverse transcribed with the exception of two NS ovary samples that RNA These two sam ples were diluted to the same concentration as other samples prior to the real time PCR step. was used to create a common RT sample. The reverse transcription protocol was The real time PCR step was run with cDNA equivalent to 100 ng total RNA for EIF5A and VTG, 10 ng total RNA for RPL32, and 1 ng total RNA for 18S rRNA. In the cases of digestive gland and ovary, 18S rRNA was very different between OS and NS groups, and so was
62 considered an inappropriate reference gene for this sample set and those run afterward. Data were normalized to ng total RNA; th is is discussed further in Chapter 5. For March 2009 S. gigas ovary and digestive gland rea l time RT PCR was performed for VTG (Chapter 6) 600 ng of DNase treated RNA was reverse transcribed and cDNA equivalent to 100 ng total RNA used for each sample i n the real time PCR step i n all samples except for one ovary sample with low yield, for which 100 ng DNase treated RNA was reverse transcribed and cDNA equivalent to 10 ng total RNA was used in the real time PCR step. All values were expressed as copy num ber/ng total RNA. A pool totaling 600 ng RNA from 2 ovary and 3 digestive gland samples was used to produce the RT control. In order to determine whether the use of different RNA and cDNA amounts would affect the quantification of the mRNA, several samp les were replicated at both concentrations described above. For the two ovary samples that were replicated, there was very little difference. Mean SEM of log 10 (copy number/ng RNA) for 2 technical replicates at each concentration gave values of 6.03 0.01 and 6.78 0.14 for one ovary sample and 3.61 0.04 and 3.60 0.04 for a second ovary sample, which supports the accuracy of these data. However, digestive gland samples were markedly different, only amplifying in the dilute replicates, which indic ates that multiple dilutions might not give the same values for very low expression genes. Note that digestive gland values were not reported in Chapter 6 due to the high frequency of samples that did not amplify ; however, 4 of the 5 samples used for the RT pool amplified, indicating that it was an appropriate control For S. alatus ovary and digestive gland samples from the 50 day in vivo feeding study, real time RT PCR was run for VTG. For ovary DNase treated
63 total RNA was reverse transcribed was used in lieu of the optional RNaseOUT enzyme due to low RNA yield for many samples. 10 different ovar y samples was used to produce the RT control. All but two real time PCR reactions were run with cDNA equivalent to 100 ng total RNA ; the remaining two (159 24, from the 14 day Cu treatment and 159 06, from the 50 day Zn treatment) were run with cDNA equ ivalent to 10 ng total RNA due to low RNA yield See the above validation results from S. gigas ovary for justification. For all S. alatus digestive gland samples, 1 transcribed and real time PCR was run with cDNA equivalent to 100 ng total RNA. digestive gland samples was used to produce the RT control. 18S Real Time RT PCR to Validate Performance of RNA Samples Prepared by Method B To assess performance in real time RT PCR of RNA prepared by Method B in a real time RT PCR assay with colu mn cleanup and DNase treatment the following assay was performed, bas ed on the Strombus gigas 18S rRNA assay described by Spade et al. [105,108] Two queen conch testis RNA samples were chosen at random, in addition to a RT control protocol. The resultant cDNA samples were diluted to a concentration equivalent to 20 :10 dilutions were performed in series to give cDNA sample was run in an RT Rad iQ SYBR Green supermix, forward and reverse primer at a final
64 reaction cDNA amounts equivalent to 10 ng, 1 ng, 100 pg, 10 pg, and 1 pg RNA, respectively, for each sample. These samples were run against a seven point standard curve consisting of plasmid containing queen conch partial 18S rRNA (NCBI accession no. GU198749) in a series of dilutions ranging from 1x10 9 copies through 1x10 3 copies per reaction. Finally, a NTC sample was run All samples, standards, and controls were ru n in duplicate, and the mean values are reported. Validation of 18S rRNA as an Internal Reference Gene for Real Time RT PCR in Testis Samples The use of 18S rRNA as a reference gene for RT PCR was validated by measuring 18s rRNA by the method described ab ove for 23 conch testis samples, collected in February, 2007, June, 2007, and March, 2009. Initial quantity for each sample was calculated as 18S rRNA copy number/ng total RNA. Data were analyzed in JMP v8 using a two Collection of Samples for Conch Shell Metal Testing Validation Study In order to determine whether shell samples could be sampled using a dental drill without causing Cu and Zn contamination, a preliminary sampling stud y was conducted using shell like material primarily composed of calcium carbonate at varying densities. F our technical replicate samples each of chalk, limestone, and marble were collected using three dif ferent dental burs composed of different materials: carbide, diamond or tit anium nitride coated carbide versus a control crude sample collected by breaking
65 sample material without use of a drill/bur. The bur was acid rinsed in a 2 % (v/v) solution of Optima grade H NO 3 (Fisher) between sample collections Historical Shell Metal Study Shells of 41 conchs of the species Strombus gigas Strombus alatus and Strombus costatus were obtained from the Florida Museum of Natural History at UF (FLMNH). These samples were all from Monroe County, FL (i.e. the Florid a Keys and surrounding areas), and ranged in date of live collection from 1936 to 1989. Shell material was sampled from these shells at an identical location just anterior of the suture on the body whorl (outermost shell whorl ) In one case, the sample w as a whole juvenile shell, and so the entire shell was collected and processed. For larger shells, the exterior of the sampling site was first superficially ground off with the titanium nitride dental bur. One S. costatus shell (coded 123708 ) was sampled repeatedly, four times with a new (previously unused) titanium nitride bur, and four times with an old (previously used for the validation study) titanium nitride bur, to determine whether the old bur contaminated the sample. In all cases, t he bur was ac id rinsed in a 2 % (v/v) solution of Optima grade H NO 3 (Fisher Scientific ) between sample collections. The powdered shell samples were kept in a 1.5 mL microcentrifuge tube until being weighed and digested for ICP MS analysis. Metal Analysis by Inductively Coupled Plasma Mass Spectrometry (ICP MS) ICP MS was used to determine levels of 58 Ni, 65 Cu, 66 Zn, 88 Sr, 107 Ag, 111 Cd, 118 Sn, 202 Hg, and 238 U in blood, digestive gland, foot, neural ganglia, and testis of male and female queen conchs collected in February and June 2007 and March, 2009, field collections and 65 Cu and 66 Zn in shell samples and male and female queen conch digestive gland and gonad of fighting conchs from the feeding studies and for 65 Cu,
66 66 Zn 111 Cd, and 208 Pb levels in algae collected in 2009 and food used in fighting conch feeding studies For 2007 field samples, ICP MS sample digestion and data collection was carried out by April Feswick (UF), and I performed data analysis. For subsequent analyses, I performed digestion, data collecti on, and data analysis. Tissue samples were weighed and digested to completion in 0.5 1.0 mL HNO 3 (Optima, Fisher Scientific) at 140 C A second HNO 3 digestion was performed if necessary. The final digestion step was in 0.5 1.0 mL 30% ultrapure H 2 O 2 at 110 C. Samples were dried to near completion (approximately 100 L) and diluted with milli Q water to a volume of 5 mL and a final concentration of 2% HNO 3 and finally passed syringe filter. The reconstituted samples were analyzed for total metal content using an XSeries 2 ICP MS (Thermo Elect ron Corporation, Winsford, Cheshire, UK) with 115 In as an internal standard. Samples were quantified against analyte specific standards with concentrations of 1, 5, 10, 50, 100, 500, and 1000 ppb each analyte. The lower limit of detection for this assay was set at 0.5 ppb analyte in the digested sample. Some Cu and Zn concentrations in blood, gonad, or digestiv e gland samples were >1000 ppb, particularly Zn in digestive gland. Therefore, i n a separate experiment, additional standards with concentration s of 5000, 10,000, and 50,000 ppb 66 Zn and 65 Cu were quantified to confirm that the instrume nt is linear through this range and to account for digested samples with high concentrations of 66 Zn and 65 Cu. This range covered all samples reported in tables or figures except for three digestive gland samples from the feeding study with Zn concentrations up to 61,320 ppb, thus confirming that the instrument was linear through the range of concentrations corresponding to all or virtually all of the samples quanti fied for all studies reported.
67 Determination of Blood Total Protein Total protein in blood samples used for metal measurement was determined using the Thermo Scientific Coomassie Plus Assay Kit, based on the method of Bradford  All samples were run in triplicate and read on a 96 well plate reader. Blank absorbance was subtracted from measured values and sam ples were quantified against the quadratic fit of a seven point standard curve consisting of 0.05, 0.1, 0.2, 0.4, 0.6, 0.8, and 1.0 mg/mL protein. 2007 male conch blood ICP MS concentrations were determined as ng analyte/mg total blood weight, and were re ported as such in Spade et al.  Therefore, those values we re not changed in Chapter 4. Subsequent blood samples from 2007 female conchs and 2009 male and female conchs are reported as ng analyte/mg blood total protein. Statistical Analyses RNA quality and labeling data and raw microarray data were analyzed in JM P Genomics. Spectrophotometric ratio and RIN data for RNA samples that were prepared by two different methods (Chapter 3) were compared using paired t tests. Means for strombid RNA Cy3 labeling data were compared using one way ANOVA, followed by Tukey Kr amer HSD. Successes and failures in Cy3 labeling were compared using a chi square test. Gene expression microarray data were analyzed as follows For the testis microarray, n on uniform spots were flagged and removed from the dataset; r ows not containing at least two data points for both groups (OS and NS) were deleted Array data were median centered prior to performing one way ANOVA on the factor location (OS/NS) to identify differentially regulated transcripts (p<0.01 or p<0.05 for further analyses, F DR=5%). For the digestive gland and ovary microarrays, data were Loess normalized (smoothing parameter = 0.2, one iteration) and analyzed using one way
68 ANOVA on the factor location ( OS/NS ) at which point control probes were evaluated. A second analysis was then performed to emphasize only annotated data: for any plus/minus sequence probe pair, the probe with the lesser signal was removed from the dataset; any probes designed from multiple contigs annotated with the same Gene Title were combined by takin g the mean of raw signal values. Condensed signal values were then Loess normalized (smoothing parameter = 0.2, one iteration); ANOVA was performed as before. For all microarray data, h ierarchical clustering analysis was performed using the program Clust er  and visualized in the program Java TreeView  Data were median centered by gene and complete linkage clustering was based on centered correlation (standard Pearson correlation) For microarray data, further functional analysis was performed: functional the FatiGO tool within the Babelomics suite  or directly in JMP Genomics, using the gene set enrichment function (previously called column enrichment). All terms with a nominal p value of p<0.05 (no post hoc correction) were considered to be enriched. Finally, important connections within enriche d biological processes in testis were illustrated using Pathway Studio 7 (Ariadne Genomics, Rockville, MD, USA ), operating on the ResNet 7.0 mammalian database, updated with zeb rafish, Danio rerio annotation (this pathway analysis is not a statistical test but an additional data analysis step). Real time RT PCR data were analyzed for differences between OS and NS groups using the Kruskall Wallis non parametric test calculator available at http://elegans.wsmed.edu/~leon/stats/utest.html (Chapter 4) or in JM P (Chapter 5) ; p values are reported. Because of larger sample size for real time RT PCR in Chapter 6,
69 data were checked for normality (Shapiro Wilk) and constant variance in SigmaPlot and analyzed by ANOVA, followed by Tukey Kramer HSD if they passed th e assumptions. If not, they were analyzed by the Kruskall Wallis test. Field samples were analyzed on the factor OS vs. NS, while in vivo data were analyzed by histological stage, percentage of oogenic tissue in histological sections, and by time point w ithin treatment for the time course study. ICP M S data were imported into JMP, and in the case of male queen conch data (Chapter 4) analyzed for difference of means using two way ANOVA, with the two factors being tissue and location; this analysis was fol lowed by the post hoc Tukey Kramer HSD test for multiple comparisons (p<0.05). For subsequent metal data collected from both conch tissues and algal samples in the field it was hypothesized that NS metal values would be larger, and there was no desire to compare concentrations across tissues. Therefore, groupwise comparisons were made using the Kruskall Wallis test or Mann Whitney test Water chemistry data from the 50 day in vivo feeding study, was analyzed for differences across treatment using the Kr uskall Wallis test. Morphometric data from the 2007 field collected female conchs were also checked for normality by the Shapiro Wilk test in JMP, then analyzed for differences in NS versus OS conchs using two tests, assuming unequal va riance. N on was used to compare gene expression, metal concentra tion, and histological data in Chapter 4, and also to ascertain trends in shell metal data over time (Chapter 7), due to low sample size for the former da taset and inconstant variance in the latter dataset. This analysis was perform ed in JMP. Finally, for digestive gland metal concentrations from the in vivo
70 feeding study, simple linear regression of Cu and Zn concentrations onto time were performed in Si gmaPlot, after testing for normality of errors and constant variance.
71 Table 2 1. Histological classification of stages of gonad development in Strombus gigas and Strombus alatus Stage Brown Peterson  Characteristics Egan  Avila Poveda  Delgado  Spade  0 Immature signet tissue with scattered undeveloped follicles "0" resting 1 Developing PG, CA, EV oocytes beginning gametogenic early development early developing developing mid development developing mature late development 2 Spawning Capable LV oocytes, ooducts possible ripe spawning ripe spawning capable 3 Regressing PG oocytes, collapsed follicles spent post spawning spent regressing atresia* atretic* 4 Regenerating signet ti ssue with scattered undeveloped follicles regressed regenerating no tissue # no tissue # Stages from Brown Peterson et al. were adapted for Chapters 5 and 6, and are described in the subject of Chapters 5 and 6. References describing conch histology are identified by first author. Spade et al. terminology was maintained for Chapter 4. CA cortical alveolar, EV early vitellogenic, LV late vitellogenic, PG primary growth. Imm ature and Regenerating stages look very similar, but a conch with a flared lip is assumed to be sexually mature, and therefore not to fall into the Immature stage. *Atresia/atretic oocytes are a condition that may or may not accompany regression, and so w ere scored separately for Chapters 5 and 6. # No Tissue is an abnormal condition characterized by a total lack of follicles in the histological section
72 Table 2 2 Primers used for Strombus gigas partial mRNA and rRNA cloning and real time RT PCR Sequenc e Purpose Direction Amplicon Size (bp) Accession Number 18S rRNA Cloning Forward GTTTCCCATCCTACGCTTCC 636 GU198749 Reverse AGACAAATCGCTCCACCAAC 18S rRNA Real time Forward TCGGTCTTATTTTGCTGGTTT 226 GU198749 Reverse ATCGCT AGTTGGCATCGTTT Ctr1c Real time Forward ACAAGGGCGGAAGAAGAAGT 158 JN105870 Reverse GGCTTTCAGTACCCAAACGA TepII Real time Forward GTCACGGCTGACTCCTTCTC 151 JN105871 Reverse TAAAGAACACGCCGATCTCC GST Real time Forward TATGGCAAGACCAACATGGA 174 JN105 873 Reverse ATTCGCGTAAAAGCCAAAGA Stard7 Real time Forward GCGCTGTTGCTGAACATAAA 183 JN105872 Reverse CTTCTTGCACACCATCTCGTT EIF5A Real time Forward CACGCATAGAGCCCATATCA 138 JN105868 Reverse TCACTGGCAAGAAGATGGAA RPL32 Real time Forward TGGGC TTAACCCGAAGGTAT 117 JN105867 Reverse GGTCGTGGGAACAACAAATC VTG Real time Forward GAGGGACAAAACAAGGGACA 213 JN105869 Reverse CACGTGGATTACACCGTCTG tr ansporter 1c; TepII, thiolester containing protein II; GSTA1, glutathione S transferase alpha 1; Stard7, StAR related lipid transfer (START) domain containing 7 ; EIF5A, eukaryotic translation initiation factor 5A; RPL32, ribosomal protein large subunit 32; VTG, vitellogenin. VTG Primers also used for Strombus alatus VTG clon ing and real time RT PCR. [Part of table reprinted with permission from Spade DJ et al. 2010. Queen conch ( Strombus gigas ) testis regresses during the reproductive season at nearshore sites in the Florida Keys (Page 4, Table 2). PLoS One 5:e12737 ]
73 Fi gure 2 1. Queen conch sampling sites in the Florida Keys. NS: Tingler Island, TI; Sombrero Reef, SR; Delta Shoal, DS. Image credit: Robert Glazer, FWRI. [Reprinted with permission from Spade DJ et al. 2010. Queen conch ( Strombus gigas ) testis regresses during the reproductive season at nearshore sites in the Florida Keys (Page 2, Figure 1). PLoS One 5:e12737 ]
74 F igure 2 2 Typical phenol chloroform RNA extraction can be followed by precipitation (Method A) or an additional CsCl gradient centrifugation step (Method B). The dashed lines represent the optional use of CsCl gradient centrifugation or column based met [Reprinted with permission from Spade DJ et al. 2011. Cesium chloride gradient centrifugation improves the quality of total RNA preparations from the gastropod Strombus gigas and the coral Montastraea faveolata (Page 45 Figure 1). J Exp Mar Biol Ecol 402:43 48.]
75 CHAPTER 3 CESIUM CHLORIDE GRAD IENT CENTRIFUGATION IMPROVES THE QUALITY OF TOTAL RNA PREPARATIO NS FROM THE GASTROPO D Strombus gigas AND THE CORAL Montastraea faveolata 1 Background Obtaining high quality RNA is essential for a number of downstream applications, including real time RT PCR and microarray analysis. However, obtaining RNA from marine invertebrates can be very difficult for various reasons. For instance, Groppe and Morse  indicate that preparing RNA from the gastropod Haliotis rufescens is problematic, because endogenous contaminants cause contamination and degradation of RNA samples. The authors suggest the contaminants might be proteoglycans, which ca n interfere with enzymes that bind RNA. Indeed, other samples high in proteoglycans, such as mammalian cartilage, also present challenges for RNA preparation  Cnidarian tissue samples might also pose problems for RNA extraction. Some cnidarian species have been of interest to researchers looking for natural inhibitors of rev erse transcriptase for antiretroviral applications [128,129] ; logically, if an RNA sample becomes contaminated with endogenous inhibitors of reverse transcriptase during the extraction process, then it will present significant difficulty for use in applications requiring reverse transcription. In spite of these issues, microarray studies have been published in the last five years involving gastropod species such as Biomphalaria glabrata  Strombus gigas  Haliotis asinina  Aplysia kurodai  and Lymnaea stagnalis  1 The contents of this chapter, including all tables and figures, have been published, and are reproduced with permission from Elsevier. Only the analyses for which D. Spade was directly responsible are reported here. Reference: 108. Spade DJ, Knoebl I, Denslow ND (2011) Cesium chloride gradient centrifugation improves the quality of total RNA preparations from the gastropod Strombus gigas and the coral Montastraea faveolata J Exp Mar Biol Ecol 402: 43 48.
76 Additionally, partial transcriptome sequencing has been successfully completed with the corals Acropora palmata and Montastraea faveolata  Clearly, therefore, RNA can be successfully extracted from gastropods and cnidarians, and can be labeled for microarrays. All of the aforementioned authors used a phenol/chloroform based extraction kit that utilizes guanidinium as a denaturant, based on the method of Chomczynski and Sacchi  However, for Strombus gigas Linnaeus, 1758, and Montastraea faveolata Ellis, 1786, samples, we have found this prep aratory procedure to be ineffective, and to yield RNA samples prone to failure in downstream applications, especially when compared to RNA samples from fish. If co extraction of endogenous contaminants such as proteoglycans is the major reason for the poo r quality of an RNA preparation, then a method that isolates RNA based on physical, in addition to chemical, properties should produce higher purity preparations than chemical extraction alone. One such physical method of RNA isolation is CsCl gradient ce ntrifugation, which separates biomolecules based on density, and which has been used successfully in the gastropod Haliotis rufescens  A similar cesium trifluoroacetate method was used by Smale and Sasse  to prepare human cartilage RNA. Therefore, we hypothesized th at CsCl gradient centrifugation would improve the quality of our Strombus gigas and Montastraea faveolata RNA preparations. We modified the methods presented in Groppe and Morse  and Kingston et al.  to develop a CsCl supplemented RNA extracti on method. This method succeeded in improving A 260 /A 280 and A 260 /A 230 ratios for Strombus gigas testis samples, as well as the RNA Integrity Number (RIN) for Montastraea faveolata samples, relative to a method that relies solely on chemical extraction. C sCl
77 centrifugation increased success in RNA labeling with cyanine 3 (Cy3) slightly, but not significantly, for S. gigas and M. faveolata but not Strombus alatus Gmelin, 1791, RNA samples. Finally, Strombus gigas testis RNA prepared using the CsCl centrif ugation step performed well in a real time RT PCR assay. Results Quality of RNA P repared by Method A Strombus gigas testis RNA samples had very low A 260 /A 280 and A 260 /A 230 ratios, averaging 1.57 and 0.36, respectively (Table 3 1). RIN could not be calcula ted due to the total lack or severe reduction of the 28S rRNA peak (Figure 3 1 A), which was previously observed in S. gigas  rRNA that has also been observed in the gastropod Haliotis rufescens  in several other mollu scs, and generally in protostomes  Montastraea faveolata RNA prepared by this method showed fewer signs of contamination tha n S. gigas with a mean A 260 /A 280 ratio of 2.11 and A 260 /A 230 ratio of 1.93. For ten of 13 M. faveolata samples, RIN could be calculated, and the mean RIN value was 7.63. For the remaining three samples, unexpected signals or signal ratios interfered wit h RIN calculation; for instance, the 28S rRNA:18S rRNA ratio was often as high as 4.0 (Figure 3 1 precluding the calculation of RIN values, which is based on normal signal features in mammalian samples  Ho wever, the electropherograms show no evidence of degradation (e.g. no increased baseline signal between ribosomal bands or below the 18S rRNA band) in the cases for which RIN could not be calculated. Therefore, the high 18S:28S ratio appears to be normal for intact RNA from this species.
78 Quality of RNA P repared by Method B A 260 /A 280 ratios were near optimal values for all groups, with group means ranging from 1.97 to 2.07 (Table 3 1). Method B followed by additional RNeasy column cleanup gave group means of 2.01 for testis and 2.13 for digestive gland in Strombus gigas samples. A 260 /A 230 ratios were similarly adequate, ranging from 1.83 to 2.14 by Method B alone, and from 1.98 to 2.26 with Method B followed by RNeasy column cleanup. Only seven of 16 Mont astraea faveolata preparations gave RIN values the remainder again were confounded by signals or signal ratios considered anomalous by the RIN algorithm but these values averaged 9.03. For samples that can be directly compared between the two methods, M. faveolata samples had slightly lower A 260 /A 280 ratio (p<0.0001), but higher RIN (p=0.0015) when prepared by the CsCl supplemented method. S. gigas testis samples had much higher A 260 /A 280 and A 260 /A 230 ratios (p<0.0001) when prepared by the CsCl metho d. When S. gigas digestive gland samples were subjected to column cleanup using the RNeasy Mini Kit, this increased the mean A 260 /A 280 ratio (p<0.0001) from 2.00 to 2.13 but had no effect on A 260 /A 230 ratio. Cy3 Labeling of RNA S amples Strombus gigas RNA samples subjected to Cy3 labeling for microarrays failed at some rate in all groups. Success rates (Table 3 2) are based on the number of samples cRNA on the first attempt. Testis samples prepare d by Method A succeeded in 62.5 % of reactions (five out of eight). Ovary and digestive gland samples prepared by Met hod B had success rates of 71.4 % (f ive of seven) and 85.7 % (six of seven), respectively. However, S. alatus ovary s amples prepared by Method B succeeded only 66.7 % of the
79 time (4 of 6), and S. alatus or S. gigas mantle samples prepare d by Method B succeeded only 50 % of the time (2 of 4). Given small sample size and tissue and species differences, these trends are tent ative, and no differences in success rate were confirmed statistically. There was no significant difference in mean Cy3 specific activity across groups, despite a trend toward higher activity with Method A. Mean yield of Cy3 labeled cRNA was significantl y higher, however, for the S. gigas digestive gland samples prepared by Method B with column purification than for any other sample group. Montastraea faveolata RNA samples prepared by Method A and labeled with Cy3 using the Cy3 CTP protocol by Iris Knoeb l (Agilent) succeeded 25 % of the time (1 of 4). In the same experiment designed by I. Knoebl a Pimephales promelas sample prepared by Method A labeled adequately, while a combination of M. faveolata and P. promelas RNA prepared by method A failed to lab el  M faveolata RNA samples prepared by Method B and labeled by Yanping Zhang using the Amino Allyl (Amb ion) protocol succeeded in 93.8 % of reactions (14 of 15), a considerably higher rate than for other M. faveolata groups. Real T ime RT PCR for Strombus gi gas 18S rRNA RNA prepared using Method B followed by RNeasy column cleanup and DNase treatment worked very well in a QPCR assay for 18S rRNA (Figure 3 2 ). The efficiency of the assay was 92.5 % 2 18S rRNA copy number calculations were very similar for the two randomly chosen testis cDNA samples; log 10 copy number (mean SEM) was 2 This assay was originally run against an eight point standard curve. The efficiency of the curve was 115.3%, due to unexpectedly early amplification of the 1e2 copy standard, which was likel y caused by a dilution error, as there was no non specific signal in any well. Therefore, that point was removed to achieve the 92.5% efficiency reported here.
80 8.69 0.07 and 8.70 0.09 for the two samples. RT controls did not amplify within the range of the standards for two of the five dilutions; for the remaining three dilutions, the RT control amplified a minimum of 12.64 cycles after the samples. NTC amplified after 30.16 cycles, 9.46 cycles after the last sample to amplify, and after all standards had amplified. Discussion RNA Q uality was S ignificantly I mproved by CsCl G radien t C entrifugation The addition of CsCl centrifugation in the preparation procedure improved A 260 /A 280 and A 260 /A 230 ratios for Strombus gigas testis samples (Table 3 1). Montastraea faveolata samples prepared by Method B showed a modest but significant dec rease in A 260 /A 280 ratio. However, it should be noted that an A 260 /A 280 value of 1.9 to 2.0 generally indicates high purity RNA  Therefore, a value above 2.0, as seen with the M. faveolata samples prepared without CsCl, might not be superior. Regardless, both group means are close to 2.0. The significantly higher RIN values of M. faveolata RNA prepared by Method B indicates that this method is at least as likely as Method A to produce intact RNA. Note also that in a separate experiment a S. gigas ovary RNA sample that had be en prepared by Method A was re purified by diluting in water, layering over CsCl, and proceeding through the CsCl centrifugation and precipitation steps, as indicated by the dashed line in Figure 2 2 This resulted in improved A 260 /A 280 and A 260 /A 230 rat ios, the former increasing from 1.92 to 1.96 and the latter from 0.70 to 2.36. Significant absorbance at 230 nm, resulting in the low A 260 /A 230 ratios seen with some S. gigas samples prepared by Method A, may indicate contamination with proteins or organi cs  Proteins containing amino acids with aromatic side chains tend to absorb at 280 nm and the peptide bond absorbs from
81 very low wavel engths up to at least 235 nm [138,139] The CsCl centrifugation step is likely responsible for separating RNA from endog enous contaminants in our samples, either protein or other small molecules, resulting in the improved A 260 /A 280 and A 260 /A 230 ratios. I recognize that the RNA solvent affects spectrophotometric readings  ; however, this should have little effect on our conclusions. RNA prepared by Method B was dissolved in ultrapure water, which may actually give lower A 260 /A 280 readings than a slightly alkaline buffer, according to Wilfinger et al.  but our samples consistently gave A 260 /A 280 readings of 1.97 to 2.07. CsCl M ethod M ay P rovide B enefits for D ownstream A pplications There were no statistically sig nificant differences in success rate for strombid samples in Cy3 labeling of RNA among RNA samples prepared by Method A or Method B with or without column cleanup (Table 3 2). However, there was a slight trend toward increased success with Method B, and a n even larger jump in success rate with Method B followed by column cleanup. Making an inference on these differences is confounded by the fact that RNA samples derived from each different tissue were only prepared by one method. It seems that there stil l may be some tissue or species specific properties that result in labeling failure for certain samples. Montastraea faveolata RNA samples prepared by Method A failed in most Cy3 labeling attempts using the Cy3 CTP (Ag ilent) labeling protocol  The fact tha t a Pimephales promelas RNA sample alone could be labeled using this series of methods, but a combined P. promelas / M. faveolata RNA sample failed, suggests that there might be some inhibitory factor co extracting with M. faveolata RNA when using Method A. The Agilent labeling procedure requires both reverse transcription and in vitro transcription/amplification; M. faveolata RNA samples gave lower cRNA yield and Cy3
82 specific activity than the P. promelas sample, indicating that either or both process(es) m ight be inhibited by some endogenous contaminant. Further, M. faveolata RNA prepared by Method B and labeled by the Ambion Amino Allyl protocol succeeded at a very high rate  While the sample size is too small to make statistical inference, and the two labe ling methods were not both attempted with RNA prepared by each method, this supports the notion that RNA prepared by Method B might be free of the endogenous co extracting contaminants that interfere with the amplification and labeling reactions. Performa nce in real time RT PCR was very good when using Strombus gigas testis RNA prepared by Method B followed by column purification and DNase treatment (Figure 3 2 ). The acceptably high efficiency (92.5 %) and high degree of consistency among replicates indica tes that these samples were free of most contaminants that can interfere with enzymatic assays. Both samples that were assayed for 18S rRNA were diluted to five different concentrations, and SEM for starting quantity values were very small for both sample s. This indicates that the assay gave reproducible results regardless of the input amount of cDNA. Summary This CsCl supplemented RNA preparation method (Method B) led to improved indicators of quality for Strombus gigas and to some degree for Montastra ea faveolata RNA samples, consistent with our hypothesis. While performance in microarray labeling procedures is difficult to compare with small sample sizes, and differences were not confirmed statistically, it appears that there may be some benefit to u sing the CsCl supplemented method, especially given the subsequent success of M. faveolata samples with amino allyl labeling. The excellent performance in real time RT PCR of S.
83 gigas testis RNA samples prepared using CsCl (Method B) and column purificati on further indicates that they are relatively free of contaminants that interfere with enzymatic assays. Given the difficulty of preparing and using RNA from some marine invertebrates, this method, adapted from Groppe and Morse  may improve the performance of RNA dependent assays for researchers working with S. gigas and M. faveolata as well as other gastropod and cnidarian species. Table 3 1. Comparison of RNA quality data for several projects prepar ed by e ach of two method s Species Tissue Method A260/A280 A260/A230 RIN Montastraea faveolata whole polyp B 1.97 1.83 9.03 Montastraea faveolata whole polyp A 2.11 1.93 7.63 Strombus alatus ovary B 2.06 1.98 Strombus gigas DG B 2.00 2.14 S trombus gigas DG B + column 2.13 1.98 Strombus gigas ovary B 1.97 1.87 Strombus gigas testis B + column 2.01 2.26 Strombus gigas testis A 1.57 0.36 Strombus spp. mantle B 2.07 2.13 Abbreviations: DG digestive gland, R IN RNA Integrity Number. For Method A and B details refer to Figure 2 2 RIN could not be calculated for Strombus samples. Further, RIN for some Montastraea faveolata samples was not reported due to presence of signals considered anomalous by the RIN algorithm; see Results for further details. *Indicates significance in paired t placed next to the group with the greater mean. [Reprinted with permission from Spade DJ et al. 2011. Cesium chloride gradient centrifugation improves the quality o f total RNA preparations from the gastropod Strombus gigas and the coral Montastraea faveolata (Page 45, Table 1). J Exp Mar Biol Ecol 402:43 48.]
84 Table 3 2 Cy3 labeling of Strombus gigas and Strombus alatus RNA samples prepared by different methods Spe cies Tissue Method Cy3 SA Succes s Fail P Strombus alatus ovary B 1.22 9.24 4 2 0.667 Strombus gigas DG B + column 13.02 13.02 6 1 0.857 Strombus gigas ovary B 2.70 16.12 5 2 0.714 Strombus gigas testis A 2.42 20.84 5 3 0.625 Strombus spp. mantle B 2.86 11.78 2 2 0.500 All All A 2.42 b 20.84 5 3 0.625 All All B 2.21 b 12.67 11 6 0.647 All All B + column 13.02 a 13.02 6 1 0.857 Abbreviations: DG digestive gland, Cy3 Cyanine 3. For Method A a nd B details refer to Figure 2 2 ves the amount of cRNA produced in an average labeling reaction. Cy3 SA indicates the specific activity of Cy3 dye in the labeled c RNA sample. Success and Fail refer to the number of successful and unsuccessful labeling reactions, counting only the first attempt for any sample. P is the corresponding proportion of successes. *Indicates that cRNA is significantly different across groups based on one way ANOVA (Tukey); means not connected by the same letter differ significantly. Cy3 SA did not differ sig nificantly across groups (one way ANOVA), nor did Proportion (chi square). [Reprinted with permission from Spade DJ et al. 2011. Cesium chloride gradient centrifugation improves the quality of total RNA preparations from the gastropod Strombus gigas and t he coral Montastraea faveolata (Page 46, Table 2). J Exp Mar Biol Ecol 402:43 48.]
85 Figure 3 1 High quality S. gigas and M. faveolata t otal RNA profiles in capillary electrophoresis. Abbreviation: RIN RNA integrity number. Electropherograms obtai ned using the Agilent 2100 Bioanalyzer. (A) Strombus gigas ovary RNA sample no. 22 18 (RIN not applicable). (B) Montastraea faveolata RNA sample no. C 12 (RIN = 9.90). [Reprinted with permission from Spade DJ et al. 2011. Cesium chloride gradient centri fugation improves the quality of total RNA preparations from the gastropod Strombus gigas and the coral Montastraea faveolata (Page 46, Figure 2). J Exp Mar Biol Ecol 402:43 48.]
86 Figure 3 2 Real time RT PCR assay for queen conch 18S rRNA. Abbreviation : SQ starting quantity. The assay was efficient (A), and 18S rRNA SQ determined by the assay was very consistent regardless of the input amount of total RNA (B). SQ is expressed as the mean SEM of log 10 (18S rRNA copies/ng total RNA) of five technic al replicate dilutions for each sample, after taking the mean of duplicate wells. See footnote 2 for a comment on the efficiency. [Reprinted with permission from Spade DJ et al. 2011. Cesium chloride gradient centrifugation improves the quality of total R NA preparations from the gastropod Strombus gigas and the coral Montastraea faveolata (Page 47, Figure 3). J Exp Mar Biol Ecol 402:43 48.]
87 CHAPTER 4 QUEEN CONCH ( Strombus gigas ) TESTIS REGRESSES D URING THE REPRODUCTIVE SEASON AT NEARSHORE SITES I N THE FL ORIDA KEYS 1 Background Queen conch ( Strombus gigas ) is a species of significant ecological and economic importance throughout its range. For example, the estimated economic value of the annual conch fishery in the Bahamas is approximately $4.457 million, representing 9,800 seasonal jobs  The queen conch is also a large benthic invertebrate associated with coral reef ecosystems, and therefore could serve as an indicator species for toxic effects contributing to the decline of the Florida coral reef ecosystem. As a r esult of the queen conch population decline in Florida, a complete moratorium on the Florida conch fishery was declared in 1986 [1,3] The queen conch was listed under the Convention on International Trade in Endan (CITES) Appendix II in 1992  However, recovery of adult conchs in spawning aggregations within the Florida Keys has been modest. In 2001, the number of adult conchs in offshore spawning aggregations was estimated at 27,000, up from a lowest observed estimate of 5,750 in 1992, according to transect data collected by the Florida Fish and Wildlife Conservation Commission (FWC)  It is believed that little or no r eproduction occurs in near shore (NS) aggregations, and that this might contribute to the slow recovery of the population [1,3] A study of conch reproduction found that NS conchs failed to develop adequate gonad t issue for reproduction, but that translocation of NS conchs to the offshore (OS) environment resulted in development of normal 1 The contents of this chapter, including all tables and figures, have been published, and are reproduced under the terms of the Public Library of Science Creative Commons License. Only the analyses for which D. Spade is responsible are reported here. Citation: 105. Spade DJ, Griffitt RJ, Liu L, Brown Peterson NJ, Kroll KJ, et al. (2010) Queen conch ( Strombus gigas ) testis regresses during the reproductive seas on at nearshore sites in the Florida Keys. PLoS One 5: e12737.
88 gonad tissue and reproductive activity within three months  However, the causes of reproductive failure of NS conchs remain unknown. Human impacts on coastal marine ecosystems are ever increasing, and threats include inputs of nutrients, organic contaminants, and metals, as well as changes in temperature, decreasing ocean pH, and deoxygenation  While many of these factors can theoretically affect reproduction, one plausible cause for reproductive failure in a NS marine gastropod is heavy meta l exposure. A number of gastropod studies have related heavy metal exposure, in particular exposure to Cu [67,71,74] and Zn [67,69,70] to reduced fecundity reproductive o utput usually measured as egg laying. Despite the link between exposure to Cu and Zn and decreased reproductive output in gastropods, past studies consider mostly female mediated effects at the individual level. In the Florida Keys, both male and female reproductive development is inhibited near shore [1,3] Given that heavy metals are known to inhibit gastropod egg laying, and that general and point sources for metal contamination exist close to shore in the Flor ida Keys [96,98,99] our general hypothesis is that heavy metals are likely to contribute to the reproductive failu re observed near shore. For the present study, our specific hypotheses were: 1. that testis transcript ional data would identify candidate gene expression pathways affected by near shore environmental stressors, and 2. that tissue concentrations of heavy metals in NS conchs would exceed those of OS conchs. For this study we developed and used a microarray to i dentify gene expression differences between the testes of NS and OS conchs. Gene expression data was anchored in histopathology to provide a more complete understanding of the dysfunction in testis development in NS conchs. Additionally, we used inductiv ely
89 coupled plasma mass spectrometry to quantify nine analytes, including Cu and Zn, in conch tissues and to determine whether their concentrations correlate with histological and gene expression evidence of NS reproductive dysfunction Results H istopathol ogy for conchs used in this study was analyzed by Nancy Brown Peterson, who reported that OS conchs apable testis phenotypes, while NS conch testis had less spermatogenic tissue in section and less advanced lobules in most cases This was true in both February and June, but perhaps most importantly, the difference between OS and NS development was greater in June than in February June, they were evenly distri by a spermatogenic index, which was based on both the stage score and the percent of gametogenic tissue in the s ection  These values were used for the correlation analysi s discussed below. For further description of histological stages, see Table 2 1. Microarray A nalysis of T esticular T ranscription 255 differentially regu lated probes (58 up and 197 down in NS with respect to OS conchs) were identified by ANOVA ( n=3, p<0.0 1, FDR=5%) ( F igure 4 1, Appendix A ). At a less stringent p value, 1147 differentially regul ated probes (341 up and 806 down) were identified (p<0.05, FDR=5%) (Appendix A ) B ased on differentially regulated probes all OS and NS individuals cluster ed sepa rately from one another, indicating that the identified set of transcripts show a clear diffe rence between these two presumably outbred groups of wild conchs (Figure 4 1 ) Differentially regulated genes we re predominantly down regulated in this experiment ; a t a cutoff of p<0.01, the proportion of
90 differentially regulated transcripts that are down regulated was 77.4 percent. The two most up regulated probes with annotation (p<0.01) were Similar to Glutathione S transferase (GST, 15.24 fold up regulated NS) and Collagen 1, Alpha 1 (COL1A1, 10.26 fold up regulated NS). The two most down regulated probes with annotation (p<0.01) were RIKEN CDNA F730014I05 Gene a mouse genome sequence (13.83 fold down regulated NS), and Dolichyl phosphate Ma nnosyltransferase Polypeptide 2 Regulatory Subunit (Dpm2, 4.92 fold down regulated NS). Functional enrichment analysis based on GO terms for biological process identified 11 significantly enriched terms in the differentially regulated gene list (Table 4 1 ). The most signif of seven differentially regulated genes was down regulated. Another notable term was on the list (p= 0.052). Pathway analysis ( Pathway Studio) further illustrated the results of the enrichment analysis ( F igure 4 2 ): m ost affected transcripts were down regulated ( blue color), many of these transcripts are found in the mitochondria, and there we re a large number of associations with the cell processes proliferation and Real T ime RT PCR Efficiencies of the real time RT PCR assays here reported ranged from 92 .5 percent to 1 08.6 percent (Table 4 2 ) and their correlation coefficients ranged from 0.98 8 to 0.999 The difference between threshold cycles of the last experimental sample to amplify and the first negative control well to amplify in any reaction was at least 6.49 cycles and 9.83 cycles for RT and NTC controls, respectively. For each assay, the dissociation curve indicated that a single amplicon was produced. By real
91 time RT PCR (n=4), two of the four genes, Stard7 and TepII, were significantly differentially regulated (p=0.029 and 0.014, respectively); the direction of regulation was the same as determined by microarray. The fold change was similar to that determined by microarray for Stard7(1.87 by real time RT PCR compared to 2.31 by microarray), but smaller for TepII (5.66 by real tim e RT PCR compared to 29.66 by microarray). For GST, the fold change was smaller, but the direction of regulation (4.71 fold up regulated NS) was similar to that determined by microarray (15.24 fold up regulated NS). This difference in real time RT PCR wa s not significant according to the Kruskall Wallis test (p=0.100). For one gene, Ctr1c, the direction of regulation determined by real time RT PCR was opposite that determined by microarray, though the difference was essentially zero (1.03 fold down NS by RT PCR compared to 1.74 fold up NS by microarray). This change was not significant by Kruskall Wallis (p=0.443). Therefore, RT PCR results significance was reduced when mea sured by RT PCR, compared to microarray. The use of 18S rRNA as a reference gene for RT PCR was validated by measuring 18S rRNA by the method described above for 23 conch testis samples, collected in February, 2007, June, 2007, and March, 2009 (Figure 2 1 ) Initial quantity for each sample was calculated as 18S rRNA copy number/ng total RNA. Data were analyzed in JMP v8 using a two ls between OS and NS, with no statistically significant differ ence according to ANOVA (Table 4 3 ). While no reference gene is perfect [142,143] 18S rRNA appears to be the best internal reference for this experimen t. 18S rRNA expression does vary across some sites and
92 collection times, but is remarkably consistent for the samples here presented, collected actin, was differentially regulated between NS and OS in the microarray study ( Appendix A ProbeName UF_Sgi_AF_101275 ), making it a poor candidate for our internal reference in real time RT PCR. Tissue M etal B urdens ICP MS results for all nin e analytes are given in Appendix B (sample size va ries: n=2 8/group, specifically enumerated in Table B 1 ). 66 Zn was present at a significantly higher level in the digestive gland of NS conchs (831.85 ng/mg) than OS conch digestive gland (84.53 ng/mg), or any other tissue at either site (Figure 4 3 A). In addition, although not statistically significant, the concentration of Zn in the NS testis (83.96 ng/mg) was approximately 15 fold higher than in the OS testis (5.43 ng/mg) (Figure 4 3 A). 65 Cu, conversely, was not significantly higher in any of the N S tissue means compared to the corresponding OS means. However, there was a non significant (p=0.65), approximately five fold difference between 65 Cu levels in NS (34.77 ng/mg) and OS (6.60 ng/mg) gonad (Figure 4 3 B, Appendix B ). In the tissue term of t he two way ANOVA, concentrations of 58 Ni, 66 Zn, 111 Cd, and 238 U were significantly higher in digestive gland than any other tissue. 1 18 S n, despite being detected only at very low concentrations in these samples, was found at its highest concentrations in the neural ganglia. 65 Cu levels were highest in the blood, which in molluscs contains a copper based hemocyanin pigment  Correlations among M icroarray, Histology, and M etal D ata Correlation an alysis was based on testis histological conditions (n=7 8), metal concentrations in testis and digestive gland (n=3 8), and expression levels of
93 differentially regulated transcripts under the GO biological processes spermatogenesis and small GTPase mediate d signal transduction as determi ned by microarray (n=3) (Table 4 4 0.655), 0.509), though this was not statistically significant ( p=0.110). Digestive gland Zn was also significantly and inversely correlated with four of the 11 transcripts included in the analysis; gonad Zn was correlated with two of the 11. SI was significantly correlated with six of the 11 genes in the analysis. D iscussion Histological analysis performed by Nancy Brown Peterson provides a physiological anchor for gene expression at two sites in the Florida Keys  While I acknowledge that site specific effects may play a large role in testis development, these observations mirror results from conchs collected at similar NS and OS areas of the Florida Keys in 1999  and 1996  suggesting that NS conchs show a persistent, long term reduction in reprod uctive capability. Moreover, the histology analysis of Nancy Brown Peterson for 2007 conchs showed a more dramatic reduction than that reported previously The histological data suggested that NS conchs begin to undergo spermatogenesis early in the repro ductive season (February), but may regress by mid season (June) [1 05] These results complement the results of our microarray and ICP MS experiments. NS conch testis transcription differed from OS in the GO biological processes proton transport (GO:0015992), spermatogenesis (GO:0007283), small GTPase mediated signal tr ansduction (GO:0007264), and others (Table 4 1, Figure 4 2 ). This supports specific hypothesis (1), and also suggests that inhibition of small GTPase (Ras)
94 mediated signaling in NS testis contributes to NS reproductive failure. ICP MS analysis indicated that Cu and Zn were elevated in some NS conch tissues, providing preliminary support for specific hypothesis (2), and creating the hypothesis that Cu and Zn may be a causative factor in reproductive failure of NS conchs in the Florida Keys. It is importan t to note that site specific differences in metal concentrations and gene expression surely exist. Future studies will incorporate metal and gene expression data from additional sites to determine whether differences in these parameters are as consistent as the histological differences observed throughout the NS and OS Florida Keys. Conch T estis Gene E xpression The gene expression analysis in the conch testis reveals, logically, that spermatogenesis associated transcripts are down regulated NS. Correspond ingly, mitochondrial transcripts are significantly down regulated in NS testes. The effects on proton transport identified by the GO enrichment analysis could be either a cause or a result of the observed reduction in spermatogenesis in NS testes, given t he important role of mitochondria in spermatozoa and in sperm maturation [144 146] Our finding is likely the result of the reduction in mature spermatozoa, and consequent numeric reduction in mitochondria, in NS t estes as opposed to OS. Under the Biological Process GO:0007283, spermatogenesis, we identified differentially regulated transcripts with major roles in spermatogenesis in species ranging from Drosophila to humans, including d egenerative spermatocyte homol og 1 (DEGS1)  ; Similar to Kiser (homologous to slowmo)  ; p roteasome activator subunit 4 (PSME4/PA200)  ; DnaJ related, subfamily B, member 13 ( D NA JB13 ) [150 ,151] which is also related to the TSARG genes in rats  and mice  ; and
95 nuclear autoantigenic sperm protein (histone binding) (NASP)  These genes, important for the process of spermatogenesis in a wide range of species, appear to be conserved in queen conch, and were all down regulated NS in the present study. A surprising result of the GO enrichment analysis was the e nrichment of the term related to R as GTPase s, proto oncogene s involved in mammalian tumor formation and developmental disorders [155 ] Seven genes that fall under this GO term were differentially regulated in our experiment, including related Ras viral oncogene homolog (Rras) ; Ras related protein 1b (Rap1b) ; RAB1A member of Ras oncogene family ; T cell lymphoma invasion and metastasis 1 (TIAM1); RAB member of ras oncogene family 4 like (RABL4); ADP ribosylation factor like 1 (ARL1); and 4R79.2, a hypothetical GTP binding protein identified in Caenorhabditis elegans All of these genes are down regulated with the exception of TIAM1 (Ap pendix A ). Ras function has been described in invertebrates including ascidians, for which Ras signaling is involved in embryonic tissue development  and Drosophila for which Rap1 is involved in cell adhesion and polarity during epidermal growth factor receptor mediated tissue growth  Ras genes are also known to be involved in vertebrate and invertebrate testis development. The Ras cyclin D2 pathway is involved in mouse spermatogonial stem cell development in vitro  MAPK and Rap GEF signaling pathways are also involved in testis development and renewal in Drosophila  Therefore, Ras GTPase signaling may play a major role in conch tes tis tissue growth and differentiation. Histological SI was correlated with six of the 11 differentially regulated transcripts involved in spermatogenesis or small GTPase mediated signaling (Table 4 4 ). This s uggest s that
96 transcription of these genes is i ndicative of the overall maturation of the testis tissue in queen conchs, and that perturbation of normal transcription of these genes is d etrimental to spermatogenesis. Transcripts evaluated by real time RT PCR were selected based on their differential re gulation between NS and OS, according to the microarray study ( Appendix A ) and their varied and interesting biological functions. GO biological processes of these gene products include : Ctr1c, copper transmembrane transport; TepII, antib acterial humoral response; GST glutathione metabolic process; Stard7, no biological process (but related to steroidogenic acute regulatory ( StAR ) protein) The results of our real time RT PCR assays were largely successful in validating the changes observed in the microa rray study. TepII, GST, and Stard7 were confirmed by real time RT PCR, though the GST result was not statistically significant. Ctr1c, however, was essentially unchanged between NS and OS samples in real time RT PCR, with a 1.03 fold change in the direct ion opposite that determined by microarray. The difference in results between platforms is possibly due in part to the small sample size (n=4) used for both assays; increased sample size would lend power to the analyses. Unfortunately, permitting regulat ions limit sample size for a protected species such as S. gigas It is also possible that for Ctr1c our probe was designed to a region with homology to other proteins or isoforms in the SLC31 family of copper transporters, causing the lack of consistency between microarray and real time RT PCR. Changes in TepII, GST, and Stard7 may indicate that stressors affecting NS conchs cause changes in immune response, xenobiotic metabolism/redox balance, and steroidogenesis,
97 respectively. However, these are single gene changes, and so should be interpreted carefully. Potential R ole of M etals as a R eproductive S tressor Our ICP MS data indicated that Cu and Zn, two known reproductive toxicants in gastropods, were elevated in some NS conch tissues. Our study also inc luded other analytes with known toxic effects, including Ni, Ag, Cd, Sn, Hg, and U. Sr was included due to its role in shell building; it is known to be physiologically beneficial in gastropods at low doses, but toxic at high levels  However, few differences were observed for the latter seven analytes. The effects of Cu and Zn on gastropod reproductive output have been well documented, although most examples relate to females. In laboratory exposures, Cu has resulted in reduced fecundity in Helix aspersa  reduc ed egg laying and a dose dependent reduction in hatching in Pomacea pal udosa  and, as copper oxychloride, reduced oocy te number in the ovotestis of Helix aspersa  Zn exposures, likewise have impacted reproduction in numerous studies, resulting in reduced fecundity and population growth rate in Valvata piscinalis  reduced fecundity in Helix aspersa  and, as an effluent containing Zn, Cd, and Fe, mortality and reduced egg laying in Lymnaea palustris  General and point sources of heavy metals in south Florida includ e storm water runoff, roadway contaminants, septic system leachate, and boats, which may be responsible for high levels of Hg, Pb, Zn, and Cu in waterways  Elevated Cu, Zn, Cr, Hg, Pb, and Ni levels have been identified in Biscayne Bay, adjacent to the city of Miami, as well as at the outflow of canals  Additionally heavy metals including Cu and Zn have been detected in sediments and seagrass beds, both habitats occupied by conchs, as well as in surface waters at multiple sites throughout south Florida with Cu
98 sometimes exceeding guidelines for aquatic life and sediment quality  Taken together, this information suggests that potential sources of Cu and Zn contamination exist in the Florida Keys and are likely to be primarily on land or close to shore, further supporting the plausibili ty of these metals interfering with NS testis development. In the present study, Zn was elevated in the digestive gland, and possibly in t he gonad, of NS conchs (Figure 4 3 A). Coupled with the knowledge that Zn causes reduced fecundity in other gastropod species [67,70] this finding suggests that Zn may contribute to the reproductive failure of NS conchs. The observed NS digestive gland mean concentration of 831.85 ng Zn/mg tissue is similar to the body burden ob served (approx. 200 reduced fecundity in Lymnaea palustris  While available data in the literature focus on female mediated reproductive inhibition measured as reduced fecundity, studies of fecundity may miss mechanistic effec ts in both males and females. Further, while the gonad is the apparent site of action for any potential toxicant, accumulation of Zn in the digestive gland in the present study is also likely to be a significant finding. The digestive gland is adjacent t o the gonad and is believed to be a site of metal accumulation and detoxification in gastropods [54,95,161,162] While a recent study indicates that Zn concentrations in the testis of the Japanese eel Anguilla japo nica track the progression of spermatogenesis  it is important to note that an excess of Zn from external sources could still have a deleterious effect, as is possible in the present study. The relationship b etween Zn and spermatogenesis is likely complex, and should be the subject of further study.
99 No significant differences in Cu concentrations within any tissue were found between NS and OS. The mean concentration of 34.77 ng Cu/mg tissue observed in NS co nch testis in this study is still only a fraction of the toxic levels accumulated in studies by Rogevich et al.  (396.6 0 ng Cu/mg tissue) and Snyman et al.  (260.47 ng Cu/mg tissue), but is approximately five times the OS mean of 6.60 ng Cu/mg tissue Further, the aforementioned studies measured whole body Cu rather than tissue specific accumulation. Blood levels of Cu in our study (40.18 ng Cu/mg tissue NS, 58.90 ng Cu/mg tissue OS) were the highest of any tissue, and it would be difficult to separ ate the Cu contribution of hemocyanin in a tissue to the amount actually bound up in cells. In other words, blood Cu bound in hemocyanin might obscure differences between tissues. Therefore, Cu might still be a factor in NS reproductive failure, and futu re studies will attempt to test this possibility. It should also be noted that many environmental factors could be considered stressors in a complex environmental mixture, and as with all real world situations, multiple stressors are likely involved at ou r NS sites. The inverse correlations between Cu and Zn concentrations in the digestive gland and SI (Table 4 4 ) provide support for the argument that accumulation of metals, including Zn and possibly Cu, in the conch digestive gland affects development of the conch testis. These hypotheses will be examined in future studies. High T hroughput S equencing for Gastropod T ranscriptomics The approximately 60,000 extant gastropods make up the largest class within the 100,000 member phylum Mollusca, the second l argest animal phylum  However, very little work has been done in the area of gastropod geno mics. A PubMed search for 16 July 2010 yielded only 14 results, one of which was non germane. Two of the remaining 13 papers discussed toxicogenomics as a tool for
100 understanding endocrine disruption in invertebrates [165,166] The remaining 11 papers applied to only five genera of gastropods: Helix  Lymnaea  Haliotis  Aplysia  and Biomphalaria [169,170] or to schistosomes that use b oth humans and gastropods as hosts [171 175] A fielded search for only 13 results, consisting of the two submissions here reported, in addition to two platforms (GPL3635 and G PL3636) and two gene expression datasets (GSE4628 and GSE18783) for Aplysia californica one platform (GPL7421) and one gene expression dataset (GSE13039) for Haliotis asinina and two platforms (GPL9129 and GPL9483) and two gene expression datasets (GSE16 596, GSE18705, and GSE22037) for Biomphalaria glabrata The use of high throughput sequencing allowed us to make a significant contribution to this growing field. Still, aside from several heavily studied genera, one of which ( Biomphalaria ) has direct im portance for human health, the entire realm of gastropod genomics remains to be developed. Summary This study has provided new information regarding the reproductive failure of NS conchs in the Florida Keys. The major findings of this study include the fo llowing: first, that failure of NS conchs to reproduce is coupled with a reduction in NS testis development, as previously reported  and premature regression of NS testis. Second, the microarray results indicate that reduced testis tissue in NS male conchs is concurrent with a decrease in the expression of many genes related to spe rmatogenesis and mitochondrial function. Transcription of small GTPase related signaling genes is clearly affected, and this may contribute to the lack of testis tissue development, but this requires further study. Finally, this study supports the hypoth esis
101 that heavy metals may contribute to the reproductive failure of NS conchs. Zn and possibly Cu are elevated in the NS conch digestive gland, and Zn may be elevated in the testis. Given that Zn and Cu are known to reduce gastropod fecundity, the possi bility that these same metals may also inhibit gametogenesis in both males and females merits further consideration. Note that this study characterized effects of the NS environment on reproductive tissue of male conchs. While many gastropod reproduction studies rely on egg laying (i.e. female mediated effects) as the measure of average reproductive success [67,69 71,74] the phenomenon observed in the NS Florida Keys affects both males and females [1,3] Conchs rely on mate pairing and copulation  rather than broadcast spawning or other mating strategies that would require fewe r reproductive males Logically, this lack of male reproductive maturity could have a significant impact on the conch population. Future studies will aim to assess transcriptional effects on the ovaries of affected NS females, in addition to males. Alth ough the testicular regression in NS conchs appears to be a persistent problem in the Florida Keys, it is apparently reversible at the level of the individual, as many NS conchs transplanted to OS areas become Spawning Capable  This suggests that transcriptional effects, which can immediately and transiently respond to environmental factors, can play an important role in understanding the disparity in conch reproduction from NS to OS, as well as identifying responsible factors. Therefore, the combination of microarray studies with more traditional approaches will yield useful inform ation for managers as they work to facilitate the recovery of NS queen conch populations in the Florida Keys.
102 Table 4 1 Enriched Gene Ontology (GO) biological process es in the testis microarray experiment Biological Process GO Term ID % of DR % of ot her p value proton transport GO:0015992 2.90 0.75 0.005 membrane fusion GO:0006944 0.83 0.00 0.005 virus induced gene silencing GO:0009616 0.83 0.00 0.005 receptor clustering GO:0043113 0.83 0.00 0.005 aromatic compound metabolic process GO:0006725 1. 24 0.13 0.011 seryl tRNA aminoacylation GO:0006434 0.83 0.03 0.015 cilium biogenesis GO:0042384 0.83 0.03 0.015 small GTPase mediated signal transduction GO:0007264 2.90 1.07 0.023 prostaglandin biosynthetic process GO:0001516 0.83 0.07 0.029 protein kinase C activation GO:0007205 0.83 0.07 0.029 neuron differentiation GO:0030182 0.83 0.10 0.045 spermatogenesis GO:0007283 1.66 0.52 0.052 refe rs to the percent of all other transcripts with GO annotation that fall under the term. P value is the raw (nominal) p [Reprinted with permission from Spade DJ et al. 2010. Queen conch ( Strombus gigas ) testis regresses dur ing the reproductive season at nearshore sites in the Florida Keys (Page 7, Table 5). PLoS One 5:e12737] Table 4 2 Comparison of testis gene expression r esults by microarray and real time RT PCR. Microarray Real Time RT PCR Gene Fold Change Direction p Fold Change Direction p Efficiency 2 Ctr1c 1.75 up 0.029 1.03 down 0.443 108.2 % TepII 29.66 up 0.020 5.66 up 0.014 92.5 % GSTA1 15.24 up 0.009 4.71 up 0.100 95.1 % Stard7 2.32 down 0.024 1.89 down 0.029 100.7 % Real time RT PCR values are normalized to 18S rRNA (18S rRNA e fficiency = 108.6 %). Fold change is the ratio of NS mean to OS mean Direction of regulation is in NS samples, with respect to OS. P value determined by ANOVA (FDR) for microarray and by Mann Whitney test for real time RT PCR. [Repr inted with permission from Spade DJ et al. 2010. Queen conch ( Strombus gigas ) testis regresses during the reproductive season at nearshore sites in the Florida Keys (Page 9, Table 6). PLoS One 5:e12737] 2 All assays were run against eight point standard curves, but the efficiencies of the 18S rRNA and Ctr1c assays were greater than 110%, likely due to dilution errors. Therefore, several poin ts were removed from each curve to obtain the efficiencies here reported. With all of the points included, the efficiencies were 142.7% and 117.1% for the two assays, respectively. The remaining assays were reported with all eight points included in the standard curve, though the 1e2 standard for TepII did not amplify.
103 Table 4 3 Validation of 18S rRNA as an intern al ref erence gene for February 2007 conch testis gene expression assays. Collection Mean 18S copies/ng (NS) SEM (NS) N (NS) T K Mean 18S copies/ng (OS) SEM (OS) N (OS) T K 02/2007 249250.0 48436.2 4 ab 252000.0 26498.4 4 ab 06/2007 111882.5 55357.8 4 bc* 344000. 0 35000.0 2 a 03/2009 33624.5 17324.5 2 c 210785.4 18636.2 7 abc followed by Tukey Kramer HSD for multiple comparisons. Within each analyte, values no t connected by the same letter are significantly different. *NS samples for 06/2007 were contaminated with digestive gland. Microarray and real time RT PCR reported in the present study was conducted with 02/2007 samples. For a note on efficiency, see foo tnote 2. [Reprinted with permission from Spade DJ et al. 2010. Queen conch ( Strombus gigas ) testis regresses during the reproductive season at nearshore sites in the Florida Keys (Supporting Information, Table S1). PLoS One 5:e12737]
104 Table 4 4 Non par ametric correlations among histological indices of testis development metal concentrations, and gene expression data SI DG Zn Testis Zn DG Cu Testis Cu PSME4 KISER DNAJB13 DEGS1 RRAS RAB1B RAB1A TIAM1 RABL4 ARL1 4R79.2 SI DG Zn 0.655 Testis Zn 0.382 0.733 DG Cu 0.509 0.836 0.661 Testis Cu 0.345 0.333 0.164 0.491 PSME4 0.771 0.600 0.800 0.429 0.100 KISER 0.771 0.771 0.700 0.543 0.100 0.714 DN AJB13 0.886 0.543 0.600 0.314 0.314 0.943 0.771 DEGS1 0.829 0.829 0.600 0.600 0.600 0.600 0.943 0.714 RRAS 0.829 0.657 0.900 0.371 0.200 0.943 0.771 0.771 0.657 RAB1B 0.886 0.886 0.700 0.600 0.600 0.600 0 .771 0.771 0.886 0.714 RAB1A 0.714 0.886 1.000 0.657 0.100 0.771 0.771 0.657 0.714 0.886 0.829 TIAM1 0.943 0.771 0.600 0.429 0.500 0.657 0.886 0.771 0.943 0.771 0.943 0.771 RABL4 0.829 0.829 0.600 0.600 0.500 0.6 00 0.943 0.714 1.000 0.657 0.886 0.714 0.943 ARL1 0.771 0.771 0.700 0.543 0.100 0.714 1.000 0.771 0.943 0.771 0.771 0.771 0.886 0.943 4R79.2 0.657 0.657 0.800 0.600 0.100 0.943 0.771 0.886 0.657 0.829 0.543 0.714 0.600 0.657 0.7 71 SER, similar to kiser; DNAJB13, DnaJ related subfamily B member 13; DEGS1, degenerative spermatocyte homolog 1 lipid desatura se; RRAS, related Ras viral oncogene homolog; RAB1B, Ras related protein 1B; RAB1A, RAB1A member Ras oncogene family; TIAM1, T cell lymphoma invasion and metastasis 1; RABL4, Rab member of Ras oncogene family like 4; ARL1, ADP ribosylation factor like 1; 4 R79.2, 4R79.2 hypothetical protein. *Indicates significance (p<0.05). [Reprinted with permission from Spade DJ et al. 2010. Queen conch ( Strombus gigas ) testis regresses during the reproductive season at nearshore sites in the Florida Keys (Page 10, Tabl e 7). PLoS One 5:e12737]
105 Figure 4 1. Hierarchical clustering of significantly differentially regulated probes in the conch testis microarray experiment Red color represents expression of a gene at a level greater than the row (gene) average, and blu e color represents expression lower than the row average. The map shows a clear distinction between NS and OS testis samples based on the 256 differentially regulated transcripts. Approximately one fourth of the regulated transcripts are up regulated in NS relative to OS; the majority are down regulated. [Reprinted with permission from Spade DJ et al. 2010. Queen conch ( Strombus gigas ) testis regresses during the reproductive season at nearshore sites in the Florida Keys (Page 7, Figure 3). PLoS One 5:e1 2737]
106 Figure 4 2. Pathway analysis of differentially regulated probes from the c onch testis microarray experiment. Pathway Studio (Ariadne Genomics) was used to find all shortest paths between genes falling under significantly enriched GO Biological Pr ocesses in the testis. Red color represents up regulation; blue color represents down regulation. Genes: ARL1, zgc:92883 (ADP ribosylation factor like 1); ATP5B, ATP synthase, H+ transporting, mitochondrial F1 complex, beta polypeptide; ATP5C1, ATP synt hase, H+ transporting, mitochondrial F1 complex, gamma polypeptide 1; ATP5G2, ATP synthase, H+ transporting, mitochondrial F0 complex, subunit C2 (subunit 9);
107 ATP5H, ATP synthase, H+ transporting, mitochondrial F0 complex, subunit d ; CRP, si:ch211 234p6.13 ( Danio rerio hypothetical protein) ; DEGS1, im:6909319 (degenerative spermatocyte homolog, lipid desaturase); DGKZ, hypothetical LOC571856 ( similar to diacylglycerol kinase, iota); DNAI2, dynein, axonemal, intermediate chain 2; DNAJB13, DnaJ (Hsp40) relate d, subfamily B, member 13; GSTM4, glutathione S transferase mu 4; MOV10, si:dkeyp 38g6.3 (Moloney leukemia virus 10); NAPA, N ethylmaleimide sensitive fusion protein attachment protein alpha; PGDS, prostaglandin D2 synthase, hematopoietic; PICK1, hypotheti cal protein LOC791503; PPME1, zgc:56239 (protein phosphatase methylesterase 1); PSME4, hypothetical LOC561538 ( proteasome (prosome, macropain) activator subunit 4); RAB1A, RAB1A member RAS oncogene family; RABL4, RAB, member of RAS oncogene family like 4; RAP1A, RAP1A, member of RAS oncogene family; SARS2, seryl tRNA synthetase 2, mitochondrial; SLMO2, slowmo homolog 2 (Drosophila) ( similar to kiser); VAPA (VAMP (vesicle associated membrane protein) associated protein A, 33kDa Organelles, clockwise from t op center: mitochondrion, endoplasmic reticulum, Golgi complex, nucleus. [Reprinted with permission from Spade DJ et al. 2010. Queen conch ( Strombus gigas ) testis regresses during the reproductive season at nearshore sites in the Florida Keys (Page 8, Fi gure 4). PLoS One 5:e12737] Figure 4 3. Concentrations of Zn (A) and Cu (B) in tissues of male conchs collected in February 2007 Letters indicate significant difference in 2 way ANOVA, with the two factors tissue and location followed by Tukey Krame r HSD (p < 0.05). Note different y axis for Cu and Zn. Break in Zn data (A) omits 150 800 ng/mg. DG = digestive gland; NG = neural ganglia. [Reprinted with permission from Spade DJ et al. 2010. Queen conch ( Strombus gigas ) testis regresses during the r eproductive season at nearshore sites in the Florida Keys (Page 9, Figure 5). PLoS One 5:e12737]
108 CHAPTER 5 DIFFERENCES IN OVARIAN APOPTOSIS, TRANSLATI ON, AND LIPID METABOLISM IN THE DIGESTIVE GLA ND OF NS FEMALE QUEEN C ONCHS Background Queen conch reprodu ction at nearshore (NS) sites in the Florida Keys is known to be limited by the reduced egg laying and gonadal development of conchs living in NS aggregations [3,8,9] an effect that is reversible by translocating N S conchs to offshore (OS) aggregations  NS to OS translocation also results in spawning (e.g. egg laying), while OS to NS translocation reduces rates of spawning, relative to conchs remaining OS  Recent work with male queen conchs from several of these sites suggested that NS males begin to develop spermatogenic tissue in the testis early in the reproductive season, but regress prema turely, while OS male conchs progress and become Spawning Capable ; this is coupled with gene expression effects on such biological processes as spermatogenesis and small GTPase mediated signal transduction as well as significant accumulation of Zn in the NS male conch digestive gland (Chapter 4)  Zn and Cu both trace metals are known to detrimentally impact reproduction in gastropods, and this is gen erally measured in terms of egg laying or egg hatching [67 69,71,72,74] Therefore, while it is interesting to understand r eproductive effects in male conchs, and male conchs are important for reproduction in conchs, which rely on copulation  effects of the NS environment on o vari an development and oogenesis can be more thoroughly interpreted within the context of the literature on other gastropods. Moreover, e gg laying is directly related to the fecundity p arameter in many matrix or life table population models, which demonstrate s the importance of ovarian development for understanding population effects.
109 Several authors have made excellent use of fecundity in population models to determine the potential influence of xenobiotics on population growth rate projections. For instance Miller et al. were able to project the impact of androgenic and anti estrogenic compounds on egg laying in fathead minnow, Pimephales promelas with significant implications for projections of population growth  Population growth rate is not always most elasti c or most sensitive to reproduction ; for instance, Salice et al. determined juvenile survival to have a greater influence than fecundity on population growth rate for the snail Biomphalaria glabrata when exposed to Cd  However, the use of fecundity as a major element of population growth models, and th e fact that reproductive effects of contaminants can alter population growth at sublethal concentrations indicates that understanding the reproductive status of females conchs in NS aggregations could prove critical to understanding the overall influence o f the NS environment on the ecology of queen conchs in the Florida Keys Further, as previously noted, natural mortality of adult [14,18,24] and perhaps even juvenile  queen conchs is presumed to be low, and so effects on reproduction and recruitment could be v ery important for queen conch population dynamics. In this study, I aimed to determine whether the accumulation of Zn in NS conch digestive gland occurs in females, as it does in males, and also to determine how gene expression differs between NS and OS fe male conchs with relation to reproductive status determined histologically in order to better understand the influence of the NS environment on the ability of female conchs to reproduce. It is a goal of ecotoxicologists to develop stronger links between individual responses and population outcomes, and so this work is critical to the understanding of the adverse outcome pathway (AOP) for
110 queen conchs in the NS Florida Keys environment  As the demographics of queen conchs in this environment become better understood, any information about potential biomarkers obtained from these microarray studies could be used to aid modeling and population projections for queen conch, as has been the case in several other species. For this study, the specif ic hypotheses were : 1. that Zn concentration s would be higher in NS than OS female queen conch digestive gland, and 2. that g ene expression differences in both ovary and digestive gland would show disparities in important processes for reproductive development between NS and OS. Results Ovarian Histology of Conchs Used in this Study G onadal sex and histological stage of ovarian development was analyz ed by Nancy Brown Peterson at the University of Southern Mississippi and findings are summarized here to provide context for gene expression data There were two major findings: first, there was a marked disparity between NS and OS ovarian stage, with all OS conchs in the Spawning Capable stage and all NS conchs in either the Reg enerating or Developing stage (Tabl e 2 1) ; second, one individual classified as a male based on observations of external genitalia during the collection was clearly fem ale based on gonad histology, containing oocytes as advanced as the cortical alveolar stage ( Nancy Brown Peterson, personal communication). This individual was considered to be a masculinized female, and was included in the gene expression analysis of female conchs. See Discussion for further details.
111 Morphometric Data Mean shell lengths and soft tissue mass of conchs collec ted for this study were nearly identical (Table 5 1). However, overall mass and shell mass were significantly greater in OS conchs. Further, lip thickness was significantly greater in OS conchs. The difference in overall mass was presumably a result of the difference in shell lip thickness and the consequent difference in shell mass between locations. Tissue Metal Concentrations ICP MS analysis indicated that mean Zn concentration in NS digestive gland was significantly higher than in OS digestive gland (Figure 5 1 and Appendix C) as hypothesized based on data from the 2007 male queen conchs. There were no other significant differences between NS and OS samples in Zn or Cu measured concentrations within tissue. Relatively high levels of Cu, Zn, and Sr were found in most tissues, including blood, while most other analytes were in the low detectable range or below the limit of detection for many samples. While t he biological significance of other metal differences is questionable due to low concentration s two way Kruskall Wallis tests found significant differences with OS higher than NS in all of the following comp arisons: Ni in digestive gland, Ag in digestive gland and foot muscle, Cd in ovary, and Hg in digestive gland (Appendix C). Microarray Experi ments : Quality Control All digestive gland and ovary microarray quality control reports from Agilent Feature Extraction software showed an acceptable linear dynamic range for spike in controls (r 2 = 0.95 1.00), and none showed problems with background median or variance, large numbers of outlying spots, or other common issues. There were slightly more outliers on ovary arrays than in the digestive gland arrays, but there were no
112 warning flags, as no array contained one percent or more non uniform spots 21 control probes were found in the differentially regulated probe list for digestive gland These included 3 of 200 (1.5 %) spike in probes, 3 of 102 (2.9 %) hybridization negative controls, and 16 of 250 (6.4%) control probes designed from distant spec ies 57 control probes were found in the differentially regulated probe list in the ovary analysis These included 29 of 200 (14.5 %) spike in probes, 8 of 102 (7.8%) hybridization negative controls, 13 of 34 Agilent internal experimental controls (38.2%) and 6 of 250 (2.4%) control probes designed from distant species. No differential signal was detected for dark or bright corner controls, reserved probe controls, or hairpin negative controls. This indicates an overall low rate of differential signal i n control probes. Note especially that control probes designed from distant species include some highly conserved sequences, e.g. a cytochrome P450 sequence, and so these are not perfect negative controls. Differentially E xpressed T ranscripts per the Micr oarray Experiments ANOVA analysis of microarray data returned a total of 1 273 experimental probes in digestive gland and 1564 probes in ovary with differential signal (p<0.01, FDR = 5 %) Of these probes, 470 in digestive gland and 635 in ovary we re annota ted with a Gene Title. After removing the probe with lesser mean signal from each sense/antisense probe pair combining expression data for multiple probes annotated with the same gene title, and repeating Loess normalization for gene data (see Methods) a second ANOVA in digestive gland found 387 differentially regulated genes, 191 up regulated and 196 down regulated NS (Figure 5 2 A and C ; Appendix D ) The final ANOVA analysis of ovary data found 558 differentially regulated genes, 241 up regulated and 317 down regulated NS (Figure 5 2 B and D; Appendix E) In the NS digestive gland,
113 t he most up regulated gene was T cell Lymphoma Invasion and Metastasis 2 (TIAM2 ; 10.26 fold ), and the most down regulated gene was Vitellogenin 2 (VTG2 ; 445.63 fold ) In o vary, t he most up regulated gene was T cell lymphoma invasion and metastasis ( the same as in the digestive gland ; 44.93 fold ), and the most down regulated gene was Sideroflexin 1 (SFXN1; 23.60 fold) In the ovary dataset ANOVA was not performed for 53 of 4328 genes due to missing spots resulting from removal of non uniform outliers. Overlap in differentially regulated genes was low: 54 genes were differentially expressed in both studies, making up 13.95% of the total for the digestive gland and 9.68% of the total for the ovary. Enriche d B iological P rocesses Nine enriched biological processes in the digestive gland (Table 5 2) and six in the ovary (Table 5 3 ) were identified at a nominal p value cutoff of p<0.05 with no correction for multiple comparison s The most enriched biological process in digestive gland was regulation of protein metabolic process, with five of the ten genes under that term being differentially regulated, followed by lipid metabolic process, translational elongation, regulation of translation, iron ion transport, cellular iron ion homeostasis, small GTPase mediated signal transduction, cation transport, and transport. In ovary, the most enriched biological process was translation, which accounted for 38 differentially regulated tra nscripts. The second most enriched process was apoptosis, with 9 differ entially regulated transcripts, followed by protein amino acid deposphorylation, regulation of cell shape, phosphate transport, and nuclear mRNA splicing via spliceosome. Of the 38 di fferentially regulated transcripts under the process annotated as mitochondrial ribosomal proteins (10 of 11 down regulated NS), and 18 were annotated as ribosomal proteins (17 of 18 down regulated
114 NS). Of 9 differentially regulated genes falling under the GO biological process regulated NS, and two up regulated (Table 5 4 ). Real T ime RT PCR Quality Control R eal time RT PCR standard curves for genes of interest had efficiency values ranging from 97.6 to 110.5 % and correlation coefficients between 0.99 7 and 0.999 18S rRNA had an efficiency of 122.5% and correlation coefficient of 0.964, but was not re run due to the considerable difference between NS and OS groups, which rendered it a poor reference gene for these tissues Amplification detected in negative control (NTC or RT) wells occurred after the lowest concentration sample well amplified for 18S rRNA (at least 1.45 cycles), EIF5A (at least 3.37 cycles), and RPL32 (at least 2.92 cycles), and was nonspec ific (likely primer dimer ) and occurred at least 1.54 cycles after amplification of the lowest concentration standard for VT G Any sample amplifying after the lowest concentration standard was considered to be below the limit of detection for the assay; i ts copy number value was set to the value of the lowest concentration standard to avoid a bias toward significant groupwise differences. In all cases, melt curve s indicated a single amplicon for any sample above the limit of detection. Note that after per forming microarray experiments, additional RNA had to be prepared for real time RT PCR for all ovary samples and one digestive gland sample. The Agilent 2100 Bioanalyzer electropherograms showed intact RNA for all samples used for microarray and all used for real time RT PCR with the possible exception of the NS ovary samples. These samples yielded less RNA than the others, and were analyzed using the Bioanalyzer pico chip. They appeared to be possibly more degraded than the OS samples, with some signal no worse than that in the illustration of a RIN = 7 sample in Schroeder et al.  However, no RIN can be
115 calculated for conch RNA samples [105,108] The NS samples also had slightly lower average A 260 /A 280 (1.92 NS, 2.14 OS) and lower A 260 /A 230 (1.04 NS, 2.14 OS) ratios than OS samples, prior to DNase treatment. Therefore, these samples may not have been ideal. Real time RT PCR data from the ovary should therefore be interpreted cautiously, as opposed to the digestive gland, fo r which all RNA samples were high quality, and confidence in th e real time RT PCR data is high. Real Time RT PCR Validation of M icroarray R esults 18S rRNA was differentially expressed between NS and OS samples for both ovary and digestive gland, and was th us determined to be an in adequate internal reference gene Therefore, r eal time RT PCR results for target mRNAs are express ed as copy number/ng total RNA (Tables 5 5 and 5 6). Results showed a high degree o f similarity with the microarray In most cases fold differences determined by real time RT PCR were greater than those determined by microarray. EIF5A and RPL32 were both down regulated in both tissues though the RPL32 result was not significant in digestive gland The VTG mRNA that was 445.6 fold lower expressed in the NS than in the OS digestive gland according to microarray was in fact 222 fold lower NS according to real time RT PCR. VTG mRNA was not significantly different between NS and OS according to microarray, but was a remarkable 165,692 fold lower in NS ovary than OS ovary according to real time RT PCR. The difference between these two estimates appeared to be caused by saturation of the microarray signal in OS samples, which is reasonable given the very high expr ession of VTG in mature oocytes. However, ovary real time RT PCR data should be interpreted cautiously due to questions regarding RNA quality.
116 Discussion Understanding Morphometric Differences between OS and NS This study was not intended to determine whether differences in siz e parameters existed between NS and OS conchs; however, these data were collected in order to perform a thorough analysis. OS and NS queen conch shell lengths for individuals used in this study were nearly identical, indicating that the animals had reache d a maximum adult shell length that is common to both sites. This falls into line with the idea that only mature adult queen conchs evidenced by reaching the maximum shell length and showing a flared lip which are therefore capable of reaching sexual ma turity [4,35,36] were collected in this study. However, the significant differences in lip thickness, shell mass, and overall mass (the latter attributable to shell mass) are interesting. Because this was not bas ed on a large sample size it is possible that these differences occurred by chance. However, it is also possible that this is another effect assoc iated with the NS environment, indicating that conditions NS are not optimal for shell production. Given th e association between NS sites and Zn accumulation, it is worth noting that Jordaens et al.  found a negative correlation between Zn concentration and shell strength and thickness in a field study of Cepea nemo ralis but the authors cautioned that the association with polluted sites was tenuous. Still, it seems Zn exposure could possibly affect shell thickness in a gastropod. Differences in shell lip thickness are likely not attributable to the age of the conch s or directly attributable to distance from shore ; while a study in the Bahamas observed thicker shell lips moving away from shore, this was coupled with increasing shell lengths  and shell lip formation does not occur prior to reaching maximal shell length [35,36] which has apparently happened in both OS and NS conchs in the present
117 study. Further, minimum shell lip thickness for maturity was reported as 4 mm based on a population in Belize  thinner than all of the animals in this study. A study in Colombia reports a much thicker 17.5 mm lip thickness for maturity in females  but that number is based on the thickn ess for which 50 % of samples showed what they called mature ov aries (similar to Spawning Capable in Brown Peterson et al.  Poveda and Baquei ro Cardenas  see Table 2 1) This i s a more conservative number, but one that might be skewed b y asynchrony or seasonality of the reproductive cycle as the proportion of mature conchs varies seasonally in the same population  Appeldoorn notes that the lip takes less than three months to form after reaching adult size, but that growth rates vary between different populations studied  Therefore, based on comparisons with the literature, it seems likely that all of the conchs collected in the NS group are old enough to have reached sexual maturity. This is further supported by the finding of Delgado et al.  that NS conchs t ranslocated to OS proceeded to develop gonads, and could reach maturity within three months. Appeldoorn  also states that lip t hickness is not a good indicator of weight, but that length is which runs counter to the data presented cu rrently, in which shell weight was largely responsible for the difference in total weight between NS and OS Stoner  des cribes the use of a condition factor for juvenile queen conch, which is calculated as tissue weight/shell length and which is suppressed by artificially increased densities in enclosures. Applying this to the adult queen conchs used in the present study, we find condition factors of 2.52 for NS and 2.44 for OS females in the study. While this cannot be compared directly to the values calculated for juvenile conchs in the Stoner
118 paper it suggests that NS and OS conchs are of similar physical condition. Coupled with the observation that they have reached a size normally assumed to be greater than the minimum for reproduction, it is safe to assume that some other factor must be at possibly overa ll shell thickness or density is abnormally reduced. Metal Accumulation in NS Female Conchs The major finding in the metal analysis conducted in this field study, that Zn accumulates in NS conch digestive gland (Figure 5 1), mirrors what was observed in males collected during the same sampling effort (Chapter 4)  The trend in blood and gonad is also toward higher Zn NS, but those trends were not confirmed stati stically. Therefore, in female conchs as in males, greater accumulation of Zn in digestive gland occurs along with the failure to devel op mature gonadal tissue in NS aggregations. A link between excess Zn exposure and decline in female reproductive capacity has been demonstrated in several gastropods [67 69] and so it is logically possible, but n ot proven, that in the current study an exogenous exposure to Zn could lead to both the accumulation of Zn in NS conch digestive gland and the reduction in reproductive development. It was discussed in Spade et al. (Chapter 4)  that Zn is required for spermatogenesis in the Japanese eel  which is also the case in mice  Similarly, Zn concentration increases with progressing stage s of Xenopus oocytes  and is required for meiosis during mouse oogenesis  It might, then, be presumed that accumulation of Zn in non reproductive conch digestive gland could be beneficial or be related to mobilization of Zn in anticipation of imminent ovarian development. However, that seems unli kely to be the case, given that the trend in ovary is also toward higher Zn concentrations in non reproductive NS conchs.
119 A ccumulation in digestive gland is more likely to be related to accumulation of excess metals which has been reported in numerous ga stropod studies [53 56,70,76 78,80,83] and which is logically consistent with the reported effects of exogenous Zn on reproduction in several gastropods [66 69] Note also that in the marine gastropods Bembic i um auratum and Austrocochlea constricta Taylor and Maher  report that Zn concentration is not explained by reproductive state, indicating that there may or may not be such a n association between gonad development and Zn concentrations in gastropods or at least not in all gastropods While no significant differences in Cu levels between NS and OS were detected in the current study or in male queen conchs (Chapter 4)  Cu remains an analyte of interest becaus e of the known ability of C u compounds to reduce gastropod fecundity [67,71 74] despite at least one report to the contrary  and also the presence of general sources of Cu c lose to shore [98,102] in marinas  and on seagrass beds  in south Florida It might be possible for Cu to exert an effect at low doses, with no discernable accumulation, but it seems unlikely. One problem, however, with determin ing whether Cu has accumulated in NS conchs at a greater rate than in OS conchs is identify ing the contribution of blood h emocyanin to measured levels in tissues. While this may obscure differences in tissue levels, the trends in Cu concentrations in conc h tissues are not significant and do not obviously favor NS or OS conchs. In fact the measured mean value for Cu in digestive gland was slightly higher NS, but in gonad was slightly higher OS. Note that mean Cu levels appear higher in females (Figure 5 1 A) than in males (Chapter 4); this is because Cu in the present chapter is reported as ng Cu/mg total protein, rather than ng Cu/mg blood. The concentration of proteins in blood
120  statement that the composition of conch blood is similar to seawater, and this is the reason that blood Cu concentrations in this chapter appear higher and less variable than the values reported for ma les. It bears mentioning that Cu might not have the affinity for accumulation in digestive gland that Zn cle arly has. I n the snails Helix aspersa asper s a and Helix aspersa maxima Cu accumulation occurs to a high degree in the viscera, but unlike Zn, Cu a ctually accumulates to higher concentrations in the foot  A recent study in Cuba confirmed that a small amount of Cu, Pb, and Zn can accumulate in conch muscle tissue  The authors note that accumulation was low, but sometimes exceeded human health limits for Cu and Pb. However, they do not measure the concentrations of these metals in digestive gland or gonad, and so do not comment on the relative accumulation rates of Cu and Zn in those tissues. In the present stud y, mean Cu concentrations in foot muscle were 2.80 ng Cu/mg NS and 4.61 ng Cu/mg OS, significantly lower than the 6.4 32.6 mg Cu/kg measured by Diaz Rizo et al. in the Cuban study  Altogether, our data do not implicate Cu accumulation as being as likely an issue for reproduction of NS conchs as Zn accumulation. S ignificances detected for metals other than Zn do not suggest that metals in general accumulate to a greater degree in NS conchs; rather, several analytes were found at significantly hi gher levels OS (Appendix C). However, many of these differences are between metals of very low concentrations at both sites, so the biological significance of dif ferences might be minimal. One possible exception is Cd, a toxic metal that is found at 3.58 ng/mg in OS ovary, an d only 0.81 ng/mg in NS ovary. This is a very low concentration that apparently does not affect reproduction OS, but
121 even low levels of non essential metals can be problematic for normal biological function, and so it would be pruden t to monitor Cd in the future. Presence of a Masculinized Female Nearshore While it was not the major focus of our study, the acc idental sampling of an apparently masculinized female at a NS site in the Florida Keys is an interesting finding that could hav e significant implications for understanding conch reproductive health in the NS environment. However, this finding is not easily interpreted in the context of contaminants known to be present in this environment. Tributyltin exposure has long been consi dered the major causative influence in development of imposex  the imposition of a penis and/or vas defere ns on the normal genital systems of females in dioecious species of gastropod, as reported by Gibbs and Bryan [18 4] after earlier reports on the phenomenon by Smith  and Blaber  The description of imposex is similar to the conch observed in the present study, and the association between female penis development and TBT accumulation has in fact been observed in queen conchs in the British Virgin Is lands  In the present sampling effort, however, tissue concentrations of t in were all below the limit of detection, and so we did not make an effort to speciate tins or attempt to quantify butyltins in conch tissues. Similar observatio ns of conchs with ovarian tissue in histological sections of the ovary that also possess a verge have been made by Reed in Strombus pug i lis [188,189] and Strombus luhuanus  and by Avila Poveda  and Reed  in Strombus gigas which is the term chos en for the present report The masculinization of Strombus luhuanus was so common in one  High imposex rates, but no measurement of butyltins, have also been observed in the
122 dog conch, Strombus canari um in Malaysia  Therefore, this phenomenon of masculinization has been previously reported in numer ous strombids. It is important to note that while organotins were long thought to be the exclusive cause of imposex, recent evidence suggests that other compounds could cause the phenomenon. Castro et al. determined that imposex was mediated by the Retin oid X Receptor (RXR) in the dogwhelk Nucella lapillus and that imposex could be induced by 9 cis retinoic acid  ; this has been observed for the rock shell Thais clavigera as well  Therefore, any compound with the ability to activate the RXR could theoretically induce an imposex like masculinized phenotype. Through this or other mechanisms, stressors other than butyltins could likely cause the masculinization of the female queen conch o bserved in the present study. as long as the NS males in the study, which averaged 45.75 5.83 cm, and slightly shorter than the OS males, which avera ged 68.07 5.13 cm. While in many reports of imposex, the penis that develops is referred to as a small and non functional Strombus gigas Titley  reported a mean female penis length of 20.6 cm and mean male penis length of 15.86 cm for one highly contaminated site. While the mean lengths were shorter overall in the Titley masculinized females can have peni s lengths similar to, or exceeding, male penis length, as was observed in the present study. Because one of the individuals used in this study wa s a masculinized female, and because of the small sample size available for analysis and inherent in microarray
123 methods, there was some concern that the imposex individual could bias the gene expression analysis toward detecting spurious differences between groups. Therefore, a second analysis of ovary data was conducted, in which the conchs were placed into rando mized groups without respect to location (A ppendix F ). This analysis found very few differences between the arbitrary groups only 40 probes indicating that differences in gene expression in the masculinized female individual were not sufficient to dri ve the analysis To further satisfy the question of the influence of the masculinized female conch on gene expression, unsupervised hierarchical clustering analysis was performed using the data for all of the genes on both the ovary and digestive gland mi croarrays rather than clustering only differentially regulated genes (Figure F 2) This analysis indicated that the imposex individual (NM3) clustered closely with the other nearshore females, in no way indicating a gene expression state that was interm ediate to or separate fro m the two locations Taken together, these two analyses demonstrate that gene expression in the ovary and digestive gland of the masculinized female are quite similar to the normal females in the study, and that differences detect ed in the study are reliable differences between OS and NS gene expression in ovary and digestive gland. This follows logically with the knowledge that normal ovarian development can proceed even when penis development occludes the genital openings of imp osex female dog whelks, preventing egg laying  and also the observation that masculinized strombid females can copulate and lay eggs  Real time RT PCR Validation of Microarray Results Real time RT PCR results in this study were normalized to total RNA which appears to be the appropriate approach for the current study as opposed to norm alization to an internal reference gene For conch testis data (Chapter 4) 18S
124 rRNA was used as an internal reference for real time RT PCR  In that experiment I demonstrated that for the queen conch samples in question, but not for all sample groups in testis, 18S rRNA was consistently expressed across groups. In the current experiment, 18S rRNA is clearly not stably expressed across groups (Table s 5 5 and 5 6 ). Variable 18S expression has been observed in recent studies with other systems, as well [195 198] N ormaliza tion to total RNA and quantification against an external standard curve was successful for this study similar to Tricarico et al.  who demonstrated that this method can be more reliable than many reference ge nes. Bustin has advocated normalization to total RNA even calling it [142,143] though recently supporting the use of multiple reference genes when possible  A similar procedure of normalization based on total RNA was recently reported in an environmental toxicology study by Connon et al.  While it is cautioned that normalization to total RNA does not account for the presence of inhibito rs or reverse transcriptase or Taq polymerase [142,143,199] I am confident that the CsCl based RNA preparation method used for these experiments is likely to remove most inhibitors, as was apparently the case with the coral samples analyzed in Spade et al. (Chapter 3)  Another concern for using total RNA normalization or, I would suspect any normalization method, is RNA integrity  While it is impossible to determine RIN for queen conch RNA samples [105,108] I am confident that the digestive gland RNA samples were of high integrity, based on visual inspection of the electropherograms from the Agilent 2100 Bioanalyzer. Ovary RN A appeared to be possibly subject to moderate degradation, which could have contributed to the increased fold differences between NS and OS, particularly for VTG. However,
125 the overall similarity between microa rray and real time RT PCR data support the val idity of our microarray results. Transcriptional Status of Genes Known to Contribute to Reproductive D evelopment VTG mRNA was not significantly differentially regulated in ovary according to the microarray, but was an impressive 160,085 fold different betw een sites according to real time RT PCR. This seems unreasonable at first but saturation of microarray probes may be responsible Clea rly, there is some level of VTG mRNA present in NS samples, as the log 2 intensity for those spots on the microarray was high (mean = 15.36 ) However, the OS probes were clearly saturated, with a log 2 intensity mean of 17.07 ; t he 99 th percentile of all genes for mean log 2 intensity was 1 5.54 This resulted in a small and non significant difference between groups when meas ured by microarray. However, the real time RT PCR assay has a much larger dynamic range than the microarray spanning eight orders of magnitude and returned a larger fold difference. A similar difference between microarray and real time RT PCR for a v itellogenin mRNA was observed recently by Skillman et al.  who measured induction of VTG mRNA by 17 ethinylestradiol at 10.42 fold by microarray and 65 ,000 fold by real time RT PCR. While NS values for VTG we re lower than expected when measured by real time RT PCR (one sample below detection and one that did not amplify), OS VTG values determined by real time RT PCR were also slightly lower than expected; they had similar copy numbers to OS values for EIF5A an d RPL32 while VTG probe intensity was higher than either gene for OS samples on the microarray This likely indicates a difference in the ability of the primers to bind the VTG sequence, relative to the
126 microarray probe, and could be explained by the exa ct position of the corresponding sequence segments transcript) Again, NS VTG values could have been further depressed if RNA integrity was reduced relative to OS samples. Therefore, despite the difference between values measured by microarray and real time RT PCR, the present data clearly indicate a considerable impact on vitellogenin production in ovary. Th e significant difference in VTG mRNA levels between NS and OS digestive gland was unexpec ted. In gastropods, vitellogenin is generally believed to be produced, presumably at both the transcript and protein level, directly in the ovary. In fact, Matsumoto et al. confirmed using both real time PCR and in situ hybridization that the ovarian fol licle cells are the site of vitellogenin mRNA production in the Pacific abalone, Haliotis discus hannai and that the vitellogenin mRNA is not detected in hepatopancreas  There are notable differences in reproductive strategy across the gastropods, and this might mean there are considerable differences in vitellogenin production. Gagnai re et al. recently studied the biochemical characteristics of vitellogenin proteins in three diverse gastropods, including a parthenogenic asexual reproducer, Potam o pyrgu s antipodarum a hermaphrodite, Valvata piscinalis and a gonochoristic/dioecious sexual reproducer, Lithoglyphus naticoides Interestingly, they were able to detect vitellogenin proteins in all three specie s using electrophoresis and silver staining, but vitellogenin  This is of interest from the perspective of the current study, because it speaks to the degree to which we can interpret information from other spec ies. While the Matsumoto et al. study very clearly confirmed that vitellogenin mRNA is found only
127 in the ovary of Haliotis discus hannai it is not impossible that there is some contribution of the digestive gland to VTG mRNA production in conch Thus, w e interpret the finding of differential regulation of VTG in the digestive gland with caution While it appears plausible that the result seen in digestive gland could indicate contamination with ovary during dissection, we are confident that there was li ttle contamination between the two tissues, despite their close physical association, and this is borne out by the microarray data, in which we found only a small overlap in significant results (Figure 5 3). There is no question that vitellogenin has been successfully used as a biomarker of endocrine disruption in fish  A perhaps more overlooked function of vitellogenin in fish is its use as a marker of reproductive s tatus of females  which fits into the greater context of an adverse outcome pathway (AOP) for the effects of anti estrogenic chemical s in female fish  In molluscs, including some gastropods but mostly in bivalves, which have a more define d role of estrogens in reproductive development than gastropods vitellogenin protein levels have been investigated as potential biomarkers of estrogenic endocrine disrupting compounds  At least one study has also sought to measure vitellog enin mRNA levels in the mussel Mytilus edulis after estradiol injection, but found variable results  However, a possibility that many of these studies overlook is the potential use of vitellogenin gene expres sion as a marker of ovarian stage and/or reproductive status Given the clear di fference in VTG expression between NS and OS conchs in this study, and the impressive dynamic range of the QPCR assay, it seems that this might have the potential to quantitat ively assess more subtle changes in ovarian health and development for queen conchs. However, this would require additional characterization. For the purposes of the current study, it is
128 sufficient to note that VTG mRNA is expressed much more highly OS t han NS during the early reproductive season (February) and that this appears to also apply to the digestive gland, though ovary is clearly the major site of VTG transcription. Aside from VTG there are numerous differences in processes related to ovarian development in this microarray study. However, due to incomplete coverage of the conch transcriptome, as well as a reliance on interspecies comparisons for annotation of sequence data, the queen conch microarray does not include any probes for genes annot ated with the GO biological process results of Spade et al.  (Chapter 4), no such inference can be made in the present study. One interesting comparison between the present study an d the testis study (Chapter 4) is the appearance in digestive gland of the enriched GO biological process small GTPase mediated signal transduction. It was postulated that ras related GTPases may have been responsible for signaling in testis development. In fact they may be important signaling molecules in several conch tissues, including digestive gland. In addition the bulk of the microarray evidence points to effects on translation in both digestive gland and ovary, apoptosis in ovary, and lipid met abolism in the digestive gland in the overall development of the ovary. These processes are discussed further. Reduc ed Translational Processes in Nearshore Conch D igestive Gland and O vary The most enriched GO biological process in the ovary microarray ana lysis was translation (Table 5 3); the digestive gland analysis included the terms regulation of protein metabolic process, translational elongation, and regulation of translation (Table 5 2). Clearly, there is a marked impact on translation in NS conchs As mentioned in the results, a majority of ribosomal proteins and mitochondrial ribosomal proteins are
129 expressed at lower levels in NS ovary. Still other down regulated translation related genes NS were translation factors, including E IF3 subunits 2 and 12 and E IF5A, which was confirmed by real time RT PCR for ovary, but not for digestive gland. In digestive gland, the regulation of protein metabolism process encompassed five differentially regulated ubiquitin conjugating enzymes. The translational elo ngation process included several elongation factors, including Tu elongation factor and mitochondrial elongation factor Tu, both of which were up regulated NS. Finally, regulation of translation included down regulation of Cytoplasmic Polyadenylation Elem ent Binding Protein 2 and MAP Kinase Interacting Serine/Threonine Protein Kinase 1. Therefore, while protein synthesis overall appears to be more heavily affected in ovary, with reduced transcription for many ribosomal proteins, there is also evidence of d own regulation in transcripts related to regulation of protein synthesis and up regulation of messages required for ubiquitin related protein degradation, suggesting an overall decrease in protein production in NS digestive gland. This was an unexpected r esult, given that digestive gland histology does not change throughout the reproductive season in the same marked fashion as the ovary. However, Gros, et al. determined based on histological data that digestive gland contains ribosome rich crypt cells, an d suggested that they have a role in replacing sloughed digestive cells  The differences in metabolic activity between NS and OS coul d lead to different rates of digestive cell turnover, and therefore explain differences in translation related gene expression and expression of 18S rRNA between NS and OS digestive gland. A striking comparison to the transcriptional state observed in this study can be found in mussels, Mytilus galloprovincialis who show inactivation of ribosomes during
130 times of cellular stress, especially metabolic stress; this can be a biomarker of general stress or stresses induced by metals such as Cu, Zn, Pb, Cd, and Hg, which can be correlated with metallothionein levels  At the mRNA level, transcription of ribosomal proteins has been shown in many studies to be coordinately regulated and to be inhibited when cellular en ergy pools are reduced; this inhibition can happen through a number of pathways in model organisms, including mammalian target of rapamycin (mTOR) and several other protein kinase signaling pathways  Therefor e, in the present study ovarian transcription suggests a high level of cellular stress, resulting in down regulation of transcription for ribosomal proteins. While this does not directly implicate metals as a stressor in the present study, it is worth me ntioning that a putative metallothionein probe on the microarray ( Probe no. UF_Sgi_AM_107631 ; not currently annotated with a Gene Title because of short sequence match to Littorina littorea metallothionein mRNA) was expressed 5.74 fold higher in NS than in OS ovary. NS Ovarian Development and Apoptosis Apoptosis was the second most enriched GO biological process in the ovary enrichment analysis (Tables 5 3 and 5 4) Connections between translational status and apoptosis are extensive. For instance, in hum ans the mitochondrial ribosomal protein S29 is also called Death Associated Protein 3, and may promote cell growth in tumors, while being associated with low levels of apoptosis  While the queen conch microar ray does not contain a sequence annotated as mitochondrial ribosomal protein S29, the current study saw differential regulation of 11 mitochondrial ribosomal proteins. Perhaps more importantly for this study, E IF5A promotes apoptosis by a mitochondrial me chanism in human cells 
131 Enrichment of the apoptosis term is not only consistent with differential regulation of transcripts related to translational processes, but might be of particular importance in the cont ext of the ovary Indeed, there is a link between apoptosis and follicular atresia at least in humans and probably in fish  In fish, ovarian atresia i s often associated with stress  but this has not to my knowledge, been heavily studied in gastropods. Nonetheless, it would seem that with down regulat ion of translational processes NS, apoptosis would be more prevalent in this tissue. I t appears that mRNAs associated with apoptosis are largely down regulated in the NS ovary (Table 5 4); these have varying functions, some of which are anti apoptotic and some of which are pro apoptotic There were two up regulated genes. The up regulated transcript sphingosine 1 phosphate phosphatase 1 (SPP1, also: sphingosine 1 phosphate phosphohydrolase 1) is directly related with apoptosis, converting sphingosine pho sphate to sphingosine and ceramide which in humans is associated with apoptosis and autophagy through complex mechanisms involving tumor necrosis factor and RNA like ER kinase [216 218] The only other NS up regul ated apoptosis related transcript was NADH dehydrogenase (ubiquitinone) Fe S protein 1, 75 kDa (NDUFS1, also: NADH coenzyme Q reductase ). NDUFS1 is a target of caspase s in the mitochondria and is cleaved before ATP levels fall preceding apop tosis  U p regulation of these two genes in NS ovary provides contradictory information regarding the idea that levels of apoptosis are higher in NS than OS ovary. T he remaining seven transcripts under this term are all d own regulated. Three of these are known to protect cells from apoptosis in other species which may in fact indicate increased apoptosis NS. These include MAP kinase Activating Death Domain (MADD)
132 [220,221] Tumor Necrosis Factor Receptor Superfamily, member 11b ( tnfrsf11b also: osteoprotegerin, OPG)  and TNF Receptor Associated Factor 1 (TRAF1)  T he remaining four NS down regulated transcripts are promoters of apoptosis in other species including Htra Serine Peptidase 2 (Htra2)  DNA Fragmentation Factor, Alpha Subunit (DFFA/DFF 45) [225,226] Programmed Cell Death 2 (PDCD2)  and Eukaryoti c Translation Initiation Factor 5A (E IF5A)  The reduced expression of EIF5A NS was confirme d by real time RT PCR (Table 5 6 ). Down regulation of this latter set of genes runs counter to the concept of increa sed apoptosis in NS ovary while down regulation of the former set supports the idea Notably, many of these gene products have been heavily studied with respect to their roles in human cancer, for example tnfrsf11b /OPG can reduce rates of apoptosis of bre ast cancer cells, which would otherwise mitigate proliferation of the cancer cells  E IF5A, however, is somewhat of an enigma: while E IF5A promotes apoptosis through a p53 driven pathway  it is also implicated in the formation of hepatocellular carcinoma  Therefore, even assuming that the roles of these gene products are conserved between humans, in which they have been heavily studied, and queen conch, it is difficult to understand the delicate balance that might exis t between cell death and cell proliferative properties conferred by a gene product such as E IF5A and so differential regulation of the transcript is not absolute evidence of either a pro or anti apoptotic role in the present study. It should be mentione d that, w hile only one gene on the microarray was annotated with the title Eukaryotic Translation Initiation Factor 5A, there was also a Eukaryotic Translation Initiation Factor 5A1 (2.67 fold down NS, p = 0.036), a Eukaryotic Translation Initiation Factor 5 (1.35 fold down NS,
133 p=0.435), and a Eukaryotic Translation Initiation Factor 5A2 (1.19 fold down NS, p = 0.664). This is the result of annotation against multiple species, and it appears that the data from this study consistently suggest NS down regula tion of EIF5A and related mRNAs. Generally, it would have been expected that a NS environment not supportive of reproductive development would lead to increased expression of pro apoptotic genes However, little is known about the relationship between atr esia and normal ovarian development in conchs. It should be noted that Zn may have an anti apoptotic role, and is inversely related with the activation of caspases in some cells  Thi s is likely not directly related to the present situation, as Zn accumulates in digestive gland, not ovary, of the conchs, while the differences in apoptosis have been identified in ovary. In order to understand the apoptotic differences between NS and OS ovary, one can use fish for comparison: it has long been known that follicular atresia occurs in fish naturally, and can also be induced by exogenous progestins  S ites of apoptosis induced by xenobioti c s in other studies have included not only oocytes and follicles, but also theca and granulosa cells in the fish ova ry  It is likely that the mechanism of atresia also involves apoptosis in conch, especially given the clear differential expre ssion of apoptosis related transcripts in our dataset. However, the weight of the gene expression data in the current study do es not clearly suggest that the transcriptional state in NS ovary is more or less pro apoptotic than O S ovary This could be due to the presence of larger numbers of mature oocytes OS which was observed histologically (Nancy Brown Peterson, personal communication), and could indicate that some rate of follicular atresia is normal in conch ovary It should also be mentioned that
134 t ranscriptional data might not match the steady state levels of proteins. Therefore, it is possible that negative feedback inhibition or other factors could lead to decreased transcription of highly expressed proteins. This obscures the ability to make a definitive case about the regulation of apoptotic processes in this study, without a clear direction of regulation for apoptosis related transcripts. In any case, apoptosis is an important process in a developing ovary, and there is a clear disparity in e xpression of apoptosis related genes in ovary between NS and OS conchs. Lipid Metabolism in the Digestive Gland Might Play a Major Role in Ovarian Development It is a notable finding that the GO biological process lipid metabolism is enriched in the digest ive gland study. All eight of the transcripts falling under this classification are lower expressed in NS digestive gland than OS digestive gland, suggesting a decrease in the rate of metabolism of lipids NS. This could reflect the overall energetic stat e of the organisms nearshore, and could also hint that NS digestive gland is responsible for producing lipids bound for developing and/or maturing oocytes. This is especially likely if, in fact, some VTG2 mRNA is produced in the digestive gland. It is kn own that ovarian signet tissue contains albumen  and that differences in carbohydrate levels in signet cells might contribute to the difference in reproductive development between NS and OS  However, the close physical association of digestive gland with ovary, the apparent differences between OS and NS in digestive gland lipid metabolism, and the high energetic cost of reproduction suggest t hat the digestive gland might have a role in reproductive development through production of lipids necessary for ovarian maturation. This could be an interesting hypothesis for future studies.
135 What Gene Expression Says about Metal Homeostasis While enrich ment analysis doe s not provide direct evidence for effects of Cu or Zn on NS reproduction an other interesting finding in the digestive gland is the enrichment of the terms iron ion transport and cellular iron ion homeostasis. Three differentially regulat ed transcripts are responsible for the appearance of both of these terms in the enrichment analysis: ferritin, catalase, and cytochrome P450, family 1, subfamily A, polypeptide 1. Perhaps the most interesting of these is the ferritin gene, which is annot ated based on a match with Arabadopsis thaliana ferritin 4 (Fet4), and is the only transcript of the three that is up regulated NS (2.36 fold up regulated, p = 0.003). Fet4 not only transports Fe, but can also be induced by Zn limitation and transport bot h Zn and Cu in humans  While catalase and CYP1A1 may be related to Fe transport, they are better known for their roles in stress response and p hase one xenobiotic metabolism, respectively. There are also several other known Cu and Zn related transcripts on the microarray, including the aforementioned putative metallothionein probe, which is up regulated in NS ovary, Copper transporter 1c ( Ctr1c 7.26 fold up in NS ovary, p<0.001 ), copper ion transporter (COPT5, not differentially regulated in either comparison), and sol ute carrier family 30, member 5 (SLC30A5, 1.46 fold down in NS digestive gland, p = 0.04 4; a Zn transporter). Despite a small num ber of Zn and Cu related transcripts on the microarray, this is a relatively high frequency of differential regulation, which supports, if not exposure to Cu and/or Zn, then differences in the metabolism of these metals between OS and NS conchs. The exac t implications of regulation of these transporters in conch is difficult to interpret given the large families of metal transporters that exist and the potential for species differences. However, supports the idea that
136 metal homeostasis differs between NS and OS digestive gland to the degree that metal transporters are being differentially regulated. Summary: Implications for Reproduction of Nearshore Conchs As in male conchs, the mean Zn concentration in the digestive gland of NS female conchs is greater than in OS conchs, which supports hypothesis (1). F urther, f emale conch reproductive status appears to be as disparate between NS and OS as that of male conchs, reported previously (Chapter 4)  This is in accord with the marked differences between NS and OS sites in female histology reported in Delgado et al. [ 3] and Glazer and Quintero  It is clear that NS reproductive development is impaired evidenced by major decreases in the transc ription of ribosomal proteins, translation initiation factors, and other translation related genes, coupled with evidence for a difference in the rate of apoptosis in NS ovary ; this supports hypothesis (2) Digestive gland gene expression provides evidenc e of effects on NS lipid and protein metabolism, GTPase mediated signaling pathways, and even metal ion homeostasis, which may suggest differences in metabolism of Fe, Zn, or possibly Cu, in the NS environment. However, ICP MS analysis only indicates sign ificant accumulation of Zn in NS conchs, consistent with the previous report. This study confirms that the NS environment has marked effects on conch ovarian development and reproductive success ; morphometric data even indicate a disparity in shell thickn ess between OS and NS conchs that may indicate an overall state of significant environmental stress, possibly from multiple sources, and negative consequences for general health of female conchs in the NS aggregation at Tingler Island
137 Table 5 1. Morphom etric data for female queen conchs col lected in the February and June 2007 sampling efforts. Location n shell length (cm) lip thickness (mm)* total mass (g)* shell mass (g)* soft tissue mass (g) NS 5 22.96 1.38 11.5 0 4.71 2105.06 318.95 1527.58 23 7.78 577.48 85.25 OS 9 23.38 0.78 21.78 7.13 2817.21 363.47 2245.14 308.77 572.07 91.29 *Indicates significance in two test comparing NS and OS values assuming unequal variance (p<0.01) ; normally distributed (Shapiro Wilk p>0.05 ) Data are expressed as mean SD Table 5 2 Enriched Gene Ontology (GO) biological processes in the digestive gland microarray experiment. GO ID Biological Process Raw p value Not DR Not DR Total DR DR Total Type GO:0051246 regulation of prot ein metabolic process 0.001 5 2531 5 248 enriched GO:0006629 lipid metabolic process 0.012 28 2531 8 248 enriched GO:0006414 translational elongation 0.018 13 2531 5 248 enriched GO:0006417 regulation of translation 0.028 5 2531 3 248 enriched GO:00068 26 iron ion transport 0.028 5 2531 3 248 enriched GO:0006879 cellular iron ion homeostasis 0.028 5 2531 3 248 enriched GO:0007264 small GTPase mediated signal transduction 0.032 28 2531 7 248 enriched GO:0006812 cation transport 0.039 6 2531 3 248 enric hed GO:0006810 transport 0.041 114 2531 19 248 enriched DR (differentially regulated) and not DR describe the number of genes annotated with each GO biological process term that are or are not differentially regulated. Totals refer to the total number o f genes in the analysis with a t least one GO biological process annotation. Type indicates whether the biological process is enriched or under represented, Raw p value determined by Fisher exact test with no correction for multiple comparisons.
138 Table 5 3 Enriched Gene Ontology (GO) biological processes in the ovary microarray experiment. GO ID Biological Process Raw p value Not DR Not DR Total DR DR Total Type GO:0006412 translation 0.001 120 2357 38 384 enriched GO:0006915 apoptosis 0.005 16 2357 9 384 enriched GO:0006470 protein amino acid dephosphorylation 0.012 9 2357 6 384 enriched GO:0008360 regulation of cell shape 0.027 5 2357 4 384 enriched GO:0006817 phosphate transport 0.036 9 2357 5 384 enriched GO:0000398 nuclear mRNA splicing, via spliceosome 0.040 13 2357 6 384 enriched DR (differentially regulated) and not DR describe the number of genes annotated with each GO biological process term that are or are not differentially regulated. Totals refer to the total number of genes in the analysis with at least one GO biological process annotation. Type indicates whether the biological process is enriched or under represented, Raw p value determined by Fisher exact test with no correction for multiple comparisons. Table 5 4. Apoptosi s associated genes differentially regulated in the ovary microarray experiment Gene Title Fold Direction (NS) p value Sphingosine 1 phosphate Phosphatase 1 4.03 up 3.24E 05 NADH Dehydrogenase Fe S Protein 1, 75kDa 3.65 up 3.80E 04 MAP kinase Activating Death Domain 2.28 down 6.74E 03 Tumor Necrosis Factor Receptor Superfamily, Member 11b 2.37 down 6.53E 03 HtrA Serine Peptidase 2 2.79 down 3.51E 03 DNA Fragmentation Factor, Alpha Subunit 2.83 down 3.74E 03 Programmed Cell Death 2 2.97 down 5.06E 03 Eukaryotic Translation Initiation Factor 5A 3.34 down 1.82E 03 Tnf Receptor associated Factor 1 6.31 down 9.58E 04 Fold fold difference between OS and NS, Direction (NS) direction of regulation in NS conchs, with respect to OS. P val ue determined b y ANOVA (FDR = 5 %).
139 Table 5 5 Comparison of digestive gland gene expression results by microarray and real time RT PCR Microarray Real time RT PCR Gene Title Fold Direction p value Fold Direction p value 18S rRNA 17.0 down 0.021 VTG 4 45.6 do wn <0.001 224 1 down 0.018 EIF5A 1.9 down 0.0 57 5.6 down 0.043 RPL32 2.1 down 0.0 18 2.0 down 0.248 Fold fold difference between OS and NS, Direction direction of regulation NS, with respect to OS. P value for microarr ay determined by ANOVA (FDR = 5 %) n = 3 NS, 4 OS P value for real time RT PCR determined by Mann Whitney test, n = 4 NS, 4 OS. Table 5 6 Comparison of ovary gene expression results by microarray and real time RT PCR Microarray Real time RT PCR Gene Title Fold Direction p value F old Direction p value 18S rRNA 164.0 down 0.034 VTG 3.3 down 0.148 16 5692.3 down 0.032 EIF5A 3.3 down 0.002 81.1 down 0.034 RPL32 3.8 down <0.001 11.1 down 0.034 Fold fold difference between OS and NS, Direction direction of regulation NS, w ith respect to OS. P value for microarr ay determined by ANOVA (FDR = 5 %) n = 3 NS, 4 OS P value for real time RT PCR determin ed by Mann Whitney test, n = 3 NS, 4 OS. Fi gure 5 1 C opper (A) and zinc (B) concentrations measured in female queen conch tissues colle cted in February and June 2007, reported as mean + SEM. All values are reported as ng analyte/mg tissue mass, except for blood, which is ng analyte/mg blood total protein. *Indicates significance in Mann Whitney nonparametric test for differ ence of means (p=0.0133). n = 2 9 per group, details and f ull report of measured analytes can be found in Appendix C.
140 Figure 5 2. Differentially regulated genes in the d igestive gland (A and C) and ovary (B and D) microarray experiments Volcano plots comparing log 2 (fold difference) to significance (as log 10 p value) for each gene in the digestive gland (A) and ovary (B) microarray studies. The horizontal dotted line in each volcano plot is a significance cutoff of log 10 p value = 2, or p<0.01; all genes above this line are significant. Arrow in (A) identifies vitellogenin, with a fold difference of 445.6, determined by microarray. Heat maps for digestive gland (C) and ovary (D) significant genes demonstrate that gene expression is consistent withi n groups (NS versus OS) for all differentially regulated genes. Note tight clustering of the supposed imposex female NM3 with the other NS individuals.
141 Figure 5 3 O verlap in differentially expressed genes between the ovary (left oval) and digestive g land (right oval) microarray experiment s. Only 54 genes were differentially expressed in both studies, making up 13.95% of the total f or the digestive gland and 9.68 % of the total for the ovary
142 CHAPTER 6 ZINC AND COPPER ACCU MULATION IN CONCH DIGESTIVE GLAND IN THE FIELD AND THE LABORATORY WITH POSSIBLE IMPLICATIONS FOR REPRODUCTION Background A consistent disparity in histological indicators of gonadal development has been observed between queen conchs ( Strombus gigas ) from offshore (OS) and nearshore (NS) aggregations in the Florida Keys. This was first reported in male and female conchs from A lligator Reef (OS) and Tingler Island (NS) collected April, 1998  later in male and female conchs collected from Tingler Island and Duck Key (NS) and Alligator Reef and Pelican Shoal (OS) in March November, 1999  and most recently from conchs collected from Pelican Shoal (OS) and Ti ngler Island (NS) in February, 2007, males (Chapter 4)  and females (Chapter 5). In Chapters 4 and 5, I demonstrated that the mean Zn concentration in the digestive gland of NS queen conchs is significantly higher than that of OS queen conchs collected from the two pairs of sites sampled in 2007 While there are several other significant differences in tissue metal concentrations between NS and OS queen conchs in the field, this was the most striking difference, in terms of the si z e, significance, and consistency across sex. While the histological disparity in gonad development has been consistent over a series of sampling efforts, it is necessary to reproduce the metal data at another time point. An additional sampling effort cou ld also aid in determining whether sources of Zn are present at NS sites to explain the accumulation of Zn in conchs collected from those sites. There is relatively little information in the literature with regard to possible sources of trace metals in th e nea rshore Florida Keys environment. However, it has been reported that Zn Cu, and other heavy metals have been found at higher concentrations
143 in sediment samples from areas close to shore in south Florida, including canals, canal outflows and the area surrounding the city of Miami [96,98] Cu has been found to exceed guideline levels for water at Florida Keys sites, and Zn in sediment has been found at higher levels on seagrass beds, which are conch habitat, tha n on non seagrass sites  Caccia et al. note that most trace metals, including Zn and Cu, are found at higher concentrations in the northern and western portions than the southern portion of Florida Bay, but do not measure concentrations on the Atlantic Ocean side of the Florida Keys, where conch aggregations are found  The authors do find measured concentrations of trace metals in the Florida Bay to be lower than in other Flo rida estuaries, but higher than concentrations in the Bahamas, and suggest agricultural runoff, roadway runoff, and boat traffic as sources of metals  While it is logical that the influence of roadway runoff and boat traffic would be greater at sites closer to shore in the Florida Keys, in order to get a clear picture of the potential for conchs to be exposed to Cu and Zn, it was necessary to sample macroalga e conch food sources and determine the levels of metals in those samples. A n in vivo exposure with a surrogate conch species, Strombus alatus the Florida fighting conch, was performed in order to assess whether it i s possible for exogenous Cu or Zn ex posures to cause the reproductive effects that have been observed in nearshore queen conchs. While it is known that Cu and Zn can decrease reproduction in gastropods [66 69,71 74] that Cu has sub lethal effects on juvenile queen conch feeding and behavior (righting time)  that Cu and Zn accumulate in digestive gland of NS conchs in the Florida Keys  and that Cu and Pb concentrations exceed public health standards in edible muscle samples of some conchs in Cuba  it has
144 not been definitively demonstrated that exogenous Zn or Cu exposures can result in decreases in reproductive development of conchs. Strombus alatus was chosen as the surrogate for in vivo e xperiments because its adult size (mean = 8.78 cm for all conchs in the present study) is smaller and therefore more amenable to laboratory experiments than the much larger queen conch (mean = 22.80 cm for all adults collected from the Florida Keys in 2007 (Chapters 4 and 5). It also was deemed to be a suitable surrogate because its reproductive biology is similar to the other strombids, including queen conch [232,233] Additionally, S. alatus has reproduced in aquaculture under similar conditions to S. gigas  However, I was cautioned that S. alatus may be less sensitive to environmental stressors such as pollutants than S. gigas (Robert Glazer, personal communication). While no toxicology data exist in the literature to make a direct comparison, this idea was mostly predicated on habitat differences between queen conch and fighting conchs, the latter of which are known to bury themselves in mud dy sand or silt sediments, as has been reported in the literature for the closely related West Indian fighting conch, Strombus pugilis [234,235] For the present study, m ale and female queen conchs and algal samples were collected from OS (Delta Shoal) an March, 2009 (Figure 2 1) A dditionally, a 50 day time course feeding exposure of Strombus alatus to either Cu (219.51 ng/mg food) or Zn (1818.72 ng/mg food) was performed in the Aquatic Toxicology La boratory (ATL) at UF, as described in the Methods. Doses for the in vivo exposures were chosen based on several factors, namely that they were similar to doses that repressed reproduction without significantly affecting feeding rates or causing toxicity i n a feeding exposure with Helix aspersa 
145 the same concentrations appeared to lead to significant accumulation of metals in gonad or digestive gland in a preliminary 60 day in vivo dose finding study, and they reflected the fact that Zn concentrations in tissues of wild conchs te nd to be approximately ten fold higher than Cu concentrations. For the experiments presented in this chapter, it was hypothesized that : 1. Zn accumulation observed in NS conchs in 2007 would persi st into a 2009 collection, 2. NS algal samples would have higher le vels of Zn and possibly Cu, 3. the impaired vitellogenesis observed in 2007 would also persist, q uantifiable by real time RT PCR, and 4. an exogenous exposure to a high dose of Zn or Cu would lead to Zn accumulation and inhibit reproductive development in Str ombus alatus Results March 2009 Queen Conch Field Study As outlined in the Methods (Chapter 2), conchs were collected from Delta Shoal 1). Two groups of OS conchs were collected, one of which was held in NS flow through water for 16 h; this is referred to as through water prior to being processed, for 28 h. Histological analysis for the 2009 field study was performed by Nancy Brown Peterson at the University of Southern Mississippi. The analysis showed marked differences in reproductive capacity between OS and NS, similar to previous analyses. T he t wo NS females for which histological analysis of ovary was performed showed very little gametogenic tissue, and i n both cases showed some early spermatogenesis. These were considered female for further analysis due to sex determination based on external genitalia, but this is evidence of intersex, a severe
146 condition, coupled with reduced oogenesis in NS females. Fe males from the OS and OS held groups were all Spawning Capable. Males collected NS were all in the stages Developing or Regenerating, or no spermatogenesis was observed (the latter an abnormal condition) In contrast, OS and OS held males were all Spawni ng Capable (Nancy Brown Peterson, personal communication). Copper and Z inc C oncentrations in the T issues of Wild Queen C onchs Cu and Zn values differed significantly across treatment groups for several comparisons, according to the Kruskall Wall is nonparam etric test (Table 6 1 and Figure 6 1 ). Most notably, Zn in the digestive gland of both male and female conchs significantly differed across groups. Individual Mann Whitney tests comparing NS conchs to either OS (held or not held) groups were not signific ant for either comparison in females (p = 0.052) or for the NS vs. OS held comparison in males (p = 0.052), but was significant for the NS vs. OS not held comparison in males (p = 0.030). Male and female Zn mean concentrations in NS digestive gland in thi s study (771.14 and 1285.56 ng/mg, respectively) were comparable to those from males (831.85 ng/mg) and females (1181.76 ng/mg) from 2007 (Chapters 4 and 5). However, OS values in the current study were higher than those measured at OS sites in 2007, for both females (585.14 and 440.16 ng/mg in 2009 held and not held, respectively ; 108.45 ng/mg in 2007) and males (217.10 and 184.12 ng/mg in 2009 held and not held, respectively ; 84.53 ng/mg in 2007). These values were not compared statistically across st udies from different years. Mean DG Zn concentrations for both males and females in the present study were higher in the OS held group than the OS not held group, but these differences were not statistically verified.
14 7 In the present study, Zn also varied significantly across ovary sample groups (p = 0.022, Kruskall Wallis), with the greatest mean value NS. However, as with female digestive gland, groupwise differences between the NS and the two OS groups were barely non significant (p = 0.052, Mann Whitne y). Taken together, these trends in Zn indicate that Zn is present at higher levels in digestive gland of NS males conchs from the 2009 sampling group. The trend of elevated Zn NS also exists in female digestive gland and ovary though not significant. Two other significant differences across groups were detected by the Kruskall Wallis test. Cu was significantly different across groups in the foot muscle, with the OS not held group having a significantly greater mean than NS (p = 0.030, Mann Whitney). Also, Zn differed significantly across groups in the neural ganglia. The trend was toward the greatest mean in the OS held group, but individual comparisons were not confirmed (Kruskall Wallis) Copper, Z inc, C admium and L ead C oncentrations in A lgal S amp les C ollected from S ites of C onch A ggregations In all, 17 algae samples were collected from each site (OS and NS) that could be identified to genus level by Gabriel Delgado of FWRI. Batophora was the dominant algal genu s, totaling 1 0 samples NS and 5 OS ( Table 6 2 ), and this is a known conch food [14,15,19] Dictyota  and Laurencia  are also known to be eaten by the queen conch However, this study was not intended to identify distribution of algal species, and thi s is not a comprehensive report on distributions. The analysis identified no significant differences in Zn between NS and OS in eithe r all samples or within genus. T he OS mean Zn concentration was greater for all samples. However, mean Zn concentration in Batophora samples was greater NS, though neither trend was significant Cu was present at significantly higher concentrations in NS Batophora Pb
148 mean concentrations were significantly higher NS, both for all samples and for Batophora Cd, on the oth er hand, was significantly higher OS, both for all samples and for Batophora Real T ime RT PCR Quality C ontrol These results apply to both the field and in vivo studies. Efficiency for all real time PCR assays ranged from 96.5 to 98.8, and correlation ran ged from 0.998 to 0.999. Any amplification observed in negative control wells was non specific, or occurred after amplification of the last standard. Any sample amplifying after the last standard (1e2 copies for the in vivo study and 1e3 copies for the f ield study) was set to the value of the lowest standard, in order to avoid a bias toward detecting a difference between samples. Melt curve indicated a single amplicon, except for apparent primer dimer in very low expression samples. Vitellogenin E xpressi on in W ild Q ueen C onchs Vitellogenin mRNA expression was higher in the ovary of both OS groups of wild female conc hs than the NS group (Figure 6 2 ), with a fold difference of 93.37 for the OS not held vs. NS comparison, and 94.27 for the OS held vs. NS com parison This was significant in the overall ANOVA (p = 0.008), and both OS groups were significantly higher than NS by Tukey Kramer HSD In digestive gland, VTG mRNA expression was sporadic, with amplification in only six of 11 conchs; therefore, no mean values or analyse s are reported for digestive gland. 50 day In Vivo Time Course Exposure Study Conchs were fed an average of 4.83 g food/tank*day, with an overall mean dry/wet ratio of 0.277. Due to slight differences in the measured dry/wet ratio for ea ch treatment food, estimated dry food weight was 1.23, 1.38, and 1.41 g food/tank*day for control,
149 Cu, and Zn treatments, respectively. This difference was statistically significant (p<0.001, Kruskall Wallis test) when comparing dry food weights, but not when comparing wet food weights. Due to the error inherent in determining the dry/wet ratio experimentally, this may not have been biologically significant. Concentrations of metals in one of three batches of food used in the study were confirmed using I CP MS: Cu values were 4.17, 219.51, and 5.92 ng/mg food, and Zn values were 36.51, 36.79, and 1818.72 ng/mg food for the control, Cu 200, and Zn 2000 treatment food preparations, respectively. Therefore, treatment levels were close to nominal concentrati ons. Tempe rature was held at a mean value of 18.15 o C during the cooling period from 10/25/2010 to 11/28/2010 and was raised to an average value of 25.09 o C during the experimental period from 01/11/2011 to 03/03/2011. The average value during the interim h eating period was 21.40 o C (Figure 6 3 ). Water chemistry data indicated no significant differences in temperature, salinity, pH, or dissolved oxygen (Table 6 3 ). Nitrite levels in archived water samples were below the lowest detectable limit of 0.05 ppm NO 2 N (0.165 ppm NO 2 ) for all tested samples Ammonia levels in archived samples averaged 0.01875 ppm NH 3 N (0.0009 ppm NH 3 ) with all samples either measuring 0.0 or 0.05 ppm NH 3 N in the full analysis However, prior to measuring all samples, a prelimin ary measurement of two archived samples collected 7 and 4 days after water changes gave concentrations >2 ppm NH 3 N. A dilution of the former sample gave an estimate of 3.2 ppm NH 3 N (0.1772 ppm NH 3 ), after taking i nto account the dilution factor, though that same tank had measured 0.2 ppm NH 3 N by an Instant Ocean test kit used during the study. The samples were re frozen and all samples were
150 measured 11 days later, along with ammonia spike ins in artificial seawater and in experimental samples which ga ve a positive response during the analysis. No sample gave a measured concentration greater than 0.05 ppm during the full analysis. However, it is possible that the additional freeze thaw affected quantification of ammonia (further analysis in Discussion ). Histological A nalysis of In Vivo G onad D evelopment Histological data for the in vivo Cu and Zn feeding study, as well as baseline measures collected on conchs from 96 days prior to the beginning of the study are reported in Table 6 4. The proximity of ovary to digestive gland is illustrated by the cross section in Figure 6 4 panel A. Examples of a well developed lobule showing vitellogenesis and a regressed lobule with evidence of atresia are given in Figure 6 4 panels B and C, respectively. The his tological data from the feeding study were not analyzed statistically, due to the small sample size and qualitative nature of these observations. The two conch gonads observed in the baseline sample were both Regenerating but it should be noted that the appearance of the egg groove on these conchs was faint. Therefore, I was concerned that they may have been in their first year of sexual maturation, and in fact Immature. However, they showed signs of having previously developed, with primary oocytes, re sidual vitellogenin in one case, and signs of oocyte resorption The 0 d conchs showed some increase in gametogenesis, with no atresia in either group. Atresia was present in all groups for the remainder of the study, but in fewer control than Cu or Zn exposed conchs at 14 and 50 days. Several control individuals at days 7, 14, and 50 did, however, show significant atresia. The greatest numbers of Spawning Capable conchs were observed in Zn exposed conchs at 7 d and
151 control conchs at 14 d. Greater num bers of Regressing ovaries were present in conchs from the 14 d and 50 d samples for all treatment groups than for earlier samples. Copper and Z inc in Florida F ighting C onchs E xposed I n V ivo Cu and Zn levels in the ovary of conchs from the feeding study we re highly Figure 6 5 B and D, respectively). However, digestive gland concentrations of both Cu and Zn showed apparently higher levels for treatment groups than control at 14 and 50 days (Figure 6 5 A and C, respectively); these trends could not be statistically verified. When metal concentrations were plotted against the time point for each treatment group, some apparent trends emerged. Digestive gland Zn concentration increased signi ficantly a cross time points in Zn treated animals ( Figure 6 5 E ), and digestive gland Cu concentration likewise increased significantly across time points in Cu treated animals (Figure 6 5 F ), while neither trend was present in t he corresponding controls Cloning of Strombus alatus V itellogenin P artial mRNA Ideally, the S. gigas microarray could be used to compare gene expression in fighting conchs exposed to Cu and Zn in vivo with queen conchs collected from the field. However, in a separate analysis, this was deem ed to be an unreliable approach (Appendix G). However, results of Chapter 5 indicated that real time RT PCR analysis of VTG mRNA could be used as a quantitative indicator of ovarian development, and this could be used as a quantitative comparison between the field and laboratory settings. Therefore, a 173 base pair (not including primers) segment of S. alatus VTG mRNA was cloned using the primers designed for S. gigas VTG, reported in Table 2 2 and discussed in Chapter 5. This segment aligned very closel y between the two
152 species, with 90.75 % identity ( Figure 6 6 ). This clone was subsequently used as a standard only for S. alatus real time RT PCR assays. Real time RT PCR A nalysis of V itellogenin mRNA Expression In Vivo Quality control was described above. Ovarian VTG mRNA expression in fighting conchs from the feeding study was variable. The highest expression values were observed in conchs deemed to be Spawning Capable based on histology ( Figure 6 7 A) and this value was found to be significantly great er than the mean value for Regenerating conchs, though not significantly greater than Developing or Regressing conchs To a lesser extent, greater levels of VTG mRNA were associated with conchs showing a large percentage of oogenic tissue in the histologi cal section (Figure 6 7 B) though no groupwise significance was found by the Kruskall Wallis test Groupwise mean values for VTG mRNA expression ( Figure 6 7 C ) did not differ significantly across treatment groups within the 14 or 50 day time point s; the 7 day time point was not tested statistically due to missing values This was likely partially due to the high variability in expression. However, one possible trend appears: VTG appears to reach its highest level at 14 d in the Zn exposed group, and th en to be lower expressed at 50 d, while the other groups of animals have their highest mean VTG expression at 50 d. This trend was not significant according to the Kruskall Wallis test. Digestive gland VTG mRNA expression is not reported due to the large number of samples that did not amplify (ca. 45%) as was the case with the queen conch field samples ; this is further evidence that ovary is the major site of VTG expression in strombids.
153 Discussion Field Collected Queen Conchs: Persistent Trends in Meta ls and Reproductive Status The concentrations of Zn in NS digestive gland for both males and females in the present study were remarkably similar to those reported from a previous sampling effort both in males (Chapter 4)  and females (Chapter 5). Moreover, the trend toward higher concentrations of Zn in NS than O S digestive gland that was observed in 2007 was observed again in 2009. This consistency is an important observation. In 2007, two NS and two OS sites were compared; in 2009, one NS site was repeated, while a separate OS site was used for comparison. It was in the OS site that differences became evident: the measured average Zn concentrations in both male and female digestive glands in the current study were considerably higher than in the previous sampling effort. Though statistical comparisons were no t made across studies, this could be an important finding. Delta Shoal the OS site in this study (Figure 2 1), has for some time been considered an intermediate site with respect to reproduction (R obert Glazer, personal communication) FWRI researchers have not collected long term rep roductive data from Delta Shoal but have only observed reproduction occurring at that site in recent years. Another OS site, Grecian Rocks, which is not addressed in this study, is similar to Delta Shoal in that it is also the site of a large, apparently healthy conch aggregation, and is in the geographic area defined as OS (south of the Hawk Channel), but no reproduction has been observed at this site over the course of ten years of monitoring (Robert Glazer, personal comm unication). This indicates that the disparity in reproduction in the Florida Keys queen conch population may be influenced by factors other than the OS vs. NS location
154 of the habitat patch, such as exact distance from shore, human population density on th e island adjacent to the site, boating and other activities, as well as the mixing of Atlantic Ocean, Gulf of Mexico, and Florida Bay waters at the particular site. While a large scale analysis of these factors is beyond the scope of this study, our resul t indicates that reproductive success at a particular habitat patch might correlate inversely with digestive gland Zn, making this a valuable parameter for assessing the ability of the habitat patch to support reproduction. At the time of the present study however, all histological evidence indicates that Delta Shoal (OS) conchs are reproductively mature and capable of reproducing successfully (Nancy Brown Peterson, personal communication). While the current data do not allow comment on the past status of this site, it appears that it may be recovering, despite moderate accumulation of Zn and past evidence of little to no reproduction. The results of real time RT PCR for VTG mRNA in ovary (Figure 6 2 ) show a clear difference between NS and OS, with VTG b eing about 90 fold higher OS than NS. This is a considerable differenc e, but does not approach the 160,085 fold difference observed between OS and NS ovary in 2007 (Chapter 5). Again, this might indicate that while Delta Shoal conchs are much more reprod uctively competent than intermediate site, with reproductive capacity more limited than Pelican Shoal, the OS si te from the 2007 study. However, comparing gene expression acr oss years and in samples collected at different months of the year (February, 2007, vs. March, 2009) should be done cautiously, especially given the concerns about integrity of RNA from NS ovary samples in the 2007 study (Chapter 5).
155 The most no table findi ng in this study with respect to reproductive status was the presence of two female conchs with an intersex phenotype. These animals were determined to be female based on external genitalia, and so this was not an imposex condition, differing from the mas culinized female observed in the 2007 collection (Chapter 5). Rather, these animals were externally female, but showed ver y small amounts of spermatogenic tissue in the gonad section. The extent of spermatogenesis was limited: only spermatogonia were pr esent in one case, several spermatozoa in another case, with many empty lobules, and no evidence of oogenesis in the section. While this would be easier to interpret if oogenesis had been observed in the section, there is no known condition that results i n the imposition of female genitalia on males (i.e. the reverse of imposex). The only similar phenomenon reported in the literature is penis shedding, which has been observed in Littorina littorea  but this h as not been documented in queen conch. Moreover, an egg groove was observed on females during the present queen conch collectio n. Therefore, it seems more likely that these animals were females displaying ovarian spermatogenesis, which has been reported in Babylonia japonica often in imposex females, but also in normal females  and in Haliotis roei  As a related note, one NS individual from the 2007 study showed a similar crude spermatogenic phenotype in the gonad, d espite originally being labeled male. At the time, that animal was considered to have been miscoded as female, despite the improbability of making that mistake based on external genitalia. At this point, it seems likely that this was also an intersex fem ale. Intersex conditions have been extensively studied in fish exposed to endocrine disrupting compounds (EDCs)  C onsiderable effort has been made to
156 characterize intersexual phenotypes in fish a nd amphibians and these can include mixed gonadal tissue, various malformations of testes, ovaries, oviducts, ovarian features in the testis, testis like features in the ovary, and numerous others  In gastropods, understanding of endocrine disruption and reproductive toxicology has been limited by the differences between species gastropods, which rely heavily on n europeptide hormones, and other test sp ecies, which rely on steroid hormones for reproductive development, and presumably by the vastly different reproductive strategies of various gastropods. However, examples of endocrine disruption in gastropods have been published, with a particular em phasis on imposex induction by butyltins [241,242] Therefore, it is difficult to know whether intersex in NS conchs could be induced by Zn, Cu, or another compound present at those sites. Regardless queen conchs are dioecious  and so the presence of an intersex condition is likely to have a significant negative impact on reproduction. Algal M etal C oncentrations No significant trends in Zn concentrations were observed between OS and NS samples ( Table 6 2 ) though Zn was the highest concentration and most variable metal in most samples. The highest Zn concentration in any sample was found in a Laurencia spp. sample offshore, and this skewed the mean for OS samples. Zn was slightly, but not significantly, higher in NS than OS Batophora samples Pb (in all samples) and Cu (in Batophora sample s) concentrations were higher on average in NS samples while the mean of all Cd values was higher OS. While no differences were significant in genera other than Batophora that was surely influenced by the small sample sizes in those other genera. It is interesting that Cd is clearly higher OS, though low in both samples (0.33 ng/mg OS and 0.07 ng/mg NS) because Cd was also higher in the
157 digestive glands of OS than NS conchs in the 2007 field study (Chapter 5). This indicates that sources and distribut ion of different metals may vary significantly within this coastal marine environment. Placing the values measured in the present study into the context of metal concentrations measured at other marine locations will help to determine wh ether differences i n OS and NS are meaningful. An excellent study of algal metal concentrations in the Great Barrier Reef, Australia, was conducted by Denton and Burdon Jones  and can serve as a guide in answering this question. At the time of the study, in 1980 1981, the authors claim that the Great Barrier Reef is a rela tively low contamination environment, and this is borne out by comparisons of algal metal concentrations between Great Barrier Reef samples and more contaminated environments reviewed from other studies. Reported mean values for Zn ranged from 0.87 12.4 ng/mg in Phaeophyta, 1.2 22.8 ng/mg in Rhodophyta, 3.0 21.2 in Chlorophyta, and 2.0 38.6 in Cyanophyta (note: reported as ng/mg here, rather than the equivalent g/g values reported in the paper, for the sake of comparison) In comparison, Batophora samples from the present study would be in the low to medium range of values reported for the Great Barrier Reef, but samples of Dictyota Halimeda Pennicillus and S argassum would be in the high end of the Great Barrier Reef Valu es, or exceed them (Table 6 2 ). Cu values reported for the Great Barrier Reef ranged as high as 11.2 ng/mg, but were generally lower (<5 ng/mg), with values similar to those reported presentl y. Cd and Pb concentrations measured in the present study were both low (<0.5 ng/mg and <2 ng/mg in all sample means, respectively), making them comparable to Great Barrier Reef samples; they do not approach some heavily
158 polluted sites noted in the Denton and Bu rdon Jones study  Therefore, our study does not indicate that metal concentrations in Florida Keys algal samples are especially high. Trends in Cu, Pb, and Cd might be meaningful, but the higher mean is still a relatively low value. Zn concentrations, while not significantly different between gro ups, are higher and more variable for many genera than the values reported in Denton and Bardon Jones  A maximum value of only 50.68 ng/mg Zn was observed NS (in a Pennicillus pyroformis sample), but a high value of 561.19 ng/mg Zn was observed in a Laurencia spp. sample OS. While this does not support t he notion that algae are a source of elevated Zn in the NS environment, it does indicate that algal food sources of conchs can accumulate high levels of Zn, and that they can vary considerably across samples. Macroalgae, in particular Batophora appear to be a major food source for queen conch; Stoner et al.  determined that Batophora density and conch density are related, and that Batophora grows when plates are left in conch aggregations in locations where grazing is impossible. Stoner and Ray  have also suggested that Batophora is a preferred conch food, and might be respons ible for migratory behavior. In addition to algae, c onchs consume detritus sediments, seagrasses ( Thalassia testinudum ), epiphytes, and probably accidentally small sediment associated organisms including other gastropods, crustaceans and polychaete worm s, none of which were included in this study [2,16,17] Randall  suggested that conchs are indiscriminate herbivores, feeding on Halophila algae and sea grasses, but Robertson  claimed that conchs never intentionally eat seagrasses, but only epiphytes including Dictyota in addition to Spyridia and Acan thophora and that conchs
159 unintentionally consume sand when on sandy rather than grassy substrates It has been suggested that conchs may consume different food sources in different age classes, as juvenile gut contents contain more detritus and less alga e than adult gut contents  However, macroalgae such as Laurencia and Batophora are the primary carbon source at least for juvenile conchs  Therefore, other food sources could be of interest for future studies, and the timing of exposures to high metal foods during conch maturation or at specific times during the annual conch repr oductive cycle might be important. However, our sampling effort included several important macroalgae, and the results were interesting: while only five Laurencia samples were collected, one showed a very high concentration of Zn. Batophora was present both NS and OS, and the trend was toward elevated Cu concentrations NS. Therefore, both of these algae could be sources of metals for queen conchs, and could be interesting subjects for further study. Note that I assume a foodborne exposure is more likel y than a waterborne exposure, given the low solubility of Cu and Zn as free ions in seawater  and the low concentrations (in the low pmol/kg range) of Cu and Zn in Atlantic open ocean water samples  Feeding Study Temperature C ontrol and Water Quality Shawl and Davis found that for Strombus alatus Strombus gigas Strombus costatus and Strombus raninus breeding behavior stopped after a drop in water temperature in aquaculture systems. However, by heating tanks during winter m onths, they were able to induce egg laying in all four species  Based on this idea, and on the ambient water temperatures in Cedar Key  from where our S. alatus conchs were collected, I attempted to stimulate simultaneous development of conch gonads for the in vivo feeding study, which began during the winter (Fig ure 6 3 ) According to
160 basel ine histological data (Table 6 4 ), this appeared to work; at the baseline time point of 96 d, both conchs analyzed were in the Regenerating stage whereas conchs analyzed at 0 d were Developing or Spawning Capable Water chemi stry data from the feeding study showed no significant differences among groups Feeding rates did not differ when compared as wet weight, but did differ significantly when food was analyzed as dry weight. This was a slight difference, and if anything, f avored the treatment groups over the control. It should be noted that Shawl and Davis  report feeding a much larger quantity to S. alatus in aquaculture, a total of 150 g/day to eight conchs, or 18.75 g/conch*day, compared to the 4.83 g/tank*day (0.805 g/conch*day) used in the present study. However, the amount fed d uring this study was determined based on the amount that conchs would eat daily without leaving food behind during the holding period. It appeared to be an appropriate amount during the feeding study, because it was usually consumed by the following morni ng, but not immediately, and it also was sufficient to support some degree of ovarian development in most of the conchs in the study. Despite the high ammonia measurements in the preliminary samples average ammonia levels in the full analysis were quite l ow (0.0009 ppm NH 3 ) ; unionized ammonia may be slightly overestimated due to high ionic strength in the saline water used in this study  The preliminary samples were chosen because of a belief that they would be higher in ammonia than other samples, as the y were taken on day 7 of the study, prior to the first partial water change on day 10, and on day 21 of the study, four days after a partial water change on day 17 The day 7 sample had measured a much lower 0.2 ppm with another test kit during the study, and so the high
161 measurements may not have been accurate However, if we consider the high measured values to represent a maximum concentration, levels of unionized ammonia in these measurements were lowe r than most of the toxic effect concentrations repo rted in acute studies with other gastropods [247 249] Several studies indicate that gastropods have lower sensitivities to ammonia than other genera [247,250] However, a concentration of 0 .07 ppm NH 3 affected behavior (activity) in one 40 day exposure study with a freshwater snail  In the present study ther e were no mortalities in any group, feeding rates were consistent on almost all days of the study, and a normal level of activity persisted throughout the study indicating that the conchs were healthy Conchs sometimes position ed themselves over the airs tones, despite normal oxygenation in the water. While this may have indicated effects of ammonia on the gill, this was an intermittent behavior and did not seem to follow a pattern with increasing study time or prior to cleaning events. Further, with dai ly cleaning of sediments and partial water changes occurring every 4.17 days on average throughout the study, coupled with an active biofilter, there should have been little opportunity for ammonia levels to rise. However, due to questions about measureme nt accuracy, this cannot be determined for certain. Metal Accumulation in In Vivo Exposed Conchs Accumulation of Cu and Zn in the in vivo feeding study showed a large d egree of variability (Figure 6 5 A D ), and therefore, no g roupwise means were significan t. Variability in kinetics of uptake and elimination should be expected with a trace metal, and a good example of this is in Gimbert et al.  who model kinetics of Zn uptake and elimination; Zn body burden data reported by Gimbert et al. during the upt ake phase is quite variable, despite the obvious trend. Variability in the present study could
162 be caused by a number of factors, of course including the different rates of physiological use of the metals in each conch, but also by slight differences in fe eding rate or other behaviors. It is interesting that conchs also eat sand and detritus as a result of their pattern of  This was observed occasionally in the present study, particularly immediately after feeding, when conchs were approaching food. Based on my own observations, it does not appear that conchs find food visually, but sense the additi on of food to the tank and approach it while sifting through sand with their proboscis. The eyestalks remain pointed upward in an apparent ly defensive position, as reported by Robertson  Therefore, in addition to the normal variability in Zn kinetics, it seems it is impossible to have complete control over what conchs will eat during an experiment, despite the fact that they fed readily on the diet u sed for this study. This could have contributed to variations in Zn uptake if some conchs ate more algae, detritus, and sand than others, though the sand in the tanks was kept as clean as possible of algae and detritus. Further, seven of the conchs in th e study, including 5 to 6 females, were brought into the lab in July and fed the HBOI control diet for two months. This diet is believed to be higher in Zn than the control diet used for the study, and could have contributed to variability. However, when all 22 conchs that arrived in the first two groups were removed from the analysis, the significant trend in Zn in the digestive gland was still present. The trend in Cu was not significant, but this was likely influenced by the reduction in sample size. When digestive gland metal concentrations were a nalyzed using correlation, which takes into account the spacing between sampling time points Zn and Cu appeared to
163 accumulate significantly in the digestive gland of Zn treated and Cu treated conchs, respect ively, while there was no significant trend in control con chs for either metal (Figure 6 5 E and F ). The separation between mean Zn values in control and treated conchs was greater by 50 days than for Cu, possibly because of the higher treatment concentra tion for Zn than Cu and possibly because of differences in Zn and Cu regulation mechanisms in conchs. Overall, t his indicates that an exposure to exogenous Zn or Cu at high concentrations can lead to accumulation of the metals in the digestive gland of S. alatus as hypothesized and as has been observed in other gastropod species [53 55,79,80] While this does not definitively prove an adverse effect of Cu or Zn in S. alatus it confirms that the accumulation phen omenon observed in NS wild conchs could be explained by exposures t o elevated Cu or Zn. I t should be noted that we have confirmed most algal samples in NS and OS sites contain far lower levels of Cu or Zn than those used in the laboratory study with S. ala tus (Table 6 2 ). T herefore, t his study might not accurately model a long ter m exposure to lower dose metals, which is a more likely situation in the field. Note also that the control food levels of Cu ( 4. 7 7 ng Cu/mg food) and Zn (36.5 0 ng Zn/mg food) wer e still higher than some algal concentrations which averaged 2.14 ng Cu/mg and 33.40 ng Zn/mg, though those values were right skewed, and Zn concentrations in Batophora for instance, were lower on average. This may have led to some accumulation prior to the start of the study, based on the difference between baseline ( 9 6 d) and 0 d control values for Cu and Zn (Figure 6 5 ). This accumulation prior to day 0 may also have been affected by the lower water temperature or fluctuations in temperature during the pr e exposure period (Figure 6 1). There is precedent in the
164 literature for these kinds of effects. In the mussel Mytilus edulis Zn is accumulated more efficiently from diet at 2 o C than 12 o C.  In the arthropod Folsomia candida fluctuating temperature appears to increase Zn assimilation but this could also be driven by increased Zn accumulation at the lower of the two temperature extremes, as a second experiment determined that Zn accumulation is higher at lower temperatures  Indicators of Reproductive Status in In v ivo Exposed Conchs Histological indications of development did not conclusively show an effect of Cu or Zn on development. However histological stage data indicated that Zn exposure may initially speed the development of the conch ovary, with a number of conchs in the Zn group showing m aturity by 14 d with the mean dropping at 50 d though this tren d was not significant (Table 6 4 ). If this is a real phenomenon, inappropriate timing of development and regression could have significant implications for r eproductive success. Atresia was apparent in a number of sections including controls as evidenced by the phagocy tic bodies visible in Figure 6 4 panel C. These features appear to be the same as the phagocytes identified in Avila Poveda et al.  indicating resorption of oocytes categories in Delgado et al.  While some atresia was observed in controls, it may be meaningful that fewer control conchs than Cu or Zn f ed conchs showed evidence of atresia at 14 d and 50 d. If the NS environment is causing inappropriate atresia of NS conch oocytes in the wild, as evidenced by differences in expression of apoptosis related genes in Chapter 5, then this result could sugges t that Cu or Zn could be the causative factor in atretic regression. However, with small sample size and high variability, this is difficult to confirm. Note that it is not likely that Zn directly causes
165 atresia via an apoptotic mechanism in cells with h igh Zn concentrations as Zn has roles, at least in some species, in anti apoptotic mechanisms  and also oogenesis [182,183] However, Zn is no t accumulating in the tissue where atresia is being observed; rather, accumulation of metals in digestive gland could have a negative impact on the gonad through ener getic or other mechanisms. Note also that if, in fact, Strombus alatus is less sensitive to contaminants than S. gigas as discussed in the Background, this could have contributed to the lack of significant differences between groups. Real time RT PCR results of the in vivo study were likewise not conclusive, indicating increased vitellogenesi s with time in all treatment groups ( Figure 6 7 C ), but again show that Zn fed conchs might peak early and begin to decline by 50 d (not significant). Clearly, the similarity between S. alatus and S. gigas VTG mRNA is quite high (Figure 6 6 ), and this ass ay app eared to work very well, with an association evident between VTG mRNA levels and reproductive stage of development (Figure 6 7 A ). Therefore, this provides further support for the possibility of using VTG mRNA as a quantitative indicator of reproduc tive status in conchs, though the levels were more variable in this study than in present or previous evaluations of S. gigas VTG mRNA. Summary of Findings The c onsistency of Zn accumulation in NS conchs observed in 2009, and corroborating previous data fr om 2007 (Chapters 4 and 5)  indicate that this is a relativ ely persistent phenomenon, which has now been documented at two NS sites and in two separate years confirming hypothesis (1) for this chapter Further, low expression of VTG mRNA in the ovaries of NS conchs, coupled with histological observations of redu ced gonad development (Nancy Brown Peterson, personal
166 communication), confirm hypothesis (3), suggesting that the reproductive deficiency observed NS is, like Zn accumulation, persistent over time. The major difference in the current study was the slightl y higher OS concentrations of Zn compared to 2007 which are still lower than NS concentrations from either year This may be interesting due to the lack of historical data suggesting that the particular OS site in this study, Delta Shoal, has long suppor ted conch reproduction. Identifying a source of Zn and/or Cu that might lead to a high or prolonged exogenous exposure for conchs in the NS environment is no small task. In the current study, algal Cu and Pb levels were higher NS, while Cd was higher OS. Zn trended higher NS in Batophora samples, but this was not significant; moreover, some very high levels of Zn were detected OS, particularly in a single Laurencia sample that measured 561.19 ng Zn/mg. Therefore, hypothesis (2) was partially supported. Accumulation of a contaminant is a classic biomarker of exposure in ecotoxicology and measuring accumulation of toxic metals in aquatic organisms is a popular approach to biomonitoring that has been very successful in some applications; for example, Bened icto et al. recently conducted a successful toxic metal biomonitoring survey of over 100 sites around the Mediterranean S ea using the mussel Mytilus galloprovincialis  W ith a trace metal however this is confounded by the physiological need for the metal and the possibility of metabolic differences or other physiological processes that could alter distribution within the organism The factor that s uggests Zn accumulation in NS conchs is a response to exogenous exposures is the general knowledge that such accumulation of Zn in gastropod viscera is a common response to exposure to excess es of Zn [53 56,80] Th erefore, while Zn
167 kinetics could be affected by other factors influencing metabolism and mobilization of metals, it seems unlikely that concentrations of Zn would reach such high levels in NS digestive gland without an elevated exposure from some environme ntal medium. Identifying that medium is difficult, and may be confounded by ontogeny and variation in feeding behaviors, long duration of low concentration exposures windows of exposure and other factors. In fact, the knowledge that Zn elimination from Helix aspersa viscera during a depuration experiment happened at a much lower rate than accumulation  could indicate that some gastropods could accumulate a lifetime burden of Zn in the digestive gland and that the elevated Zn in adult S. gigas digest ive glands could have resulted from early or long term exposures. It should also be briefly mentioned that the results of the preliminary 60 day in vivo experiment showed a correlation between Zn and Cu concentrations in the gonad that indicated the metal s could be co transported or otherwise study, this was not observed, and it may have been related to the uniformly high concentrations of Zn in the digestive glands of conchs from t he 60 day study. In vivo Cu and Zn exposures led to significant t rends toward accumulation of the metals in digestive gland, partially confirming hypothesis (4) but the effect on gonad al development was not conclusive In order to improve the separation in metal accumulation between control and treated conchs, t he present study used a food recipe based on Laskowski and Hopkin  which gave lower measured Cu and Zn concentrations than the food used by Shawl et al. [27,110] ( 36.51 ng/mg and 87.95 ng/ mg, respectively ). It should be noted however, that Shawl et al. have achieved reproduction in aquaculture when using the latter food recipe. This may be an
168 indication that timing of exposure in the present study is preventing detection of an effect. F uture studies might be able to address some of these questions by using a depuration experiment, which would more closely model the translocation study performed by Delgado et al.  Still, in the present study, there was some evidence that Cu and Zn t reatments led to more frequent observations of atretic regression in the ovary, and also that Zn may have accelerated development and regression of the ovaries, based on VTG mRNA expression. While neither of these comparisons was significant, these are in teresting trends, as inappropriate timing of maturation might be detrimental to reproductive success. The failure of queen conchs in NS aggregations to develop mature gametogenic gonads during the reproductive season is a consistent phenomenon [3,8,105] In the present study, we demonstrate that it is also consistently associated with elevated digestive gland Zn concentrations that were observed by Spade et al.  and reported in Chapters 4 and 5. While algal metal levels do not absolutely identify a source of Zn NS, in vivo dietborne exposures resulted in significant accumulation of both Zn and Cu in fighting conch digestive gland. A 50 d Cu or Zn exposure in adult fighting conchs that were beginning to undergo seasonal reproductive development was not sufficient to complet ely inhibit or reverse ovarian development. However, results of the feeding studies provided clues as to mechanisms by which Cu and Zn might exert negative effects on ovarian development, including inappropriate acceleration of development and atretic reg ression.
169 Table 6 1. Copper and zinc concentrations in the tissues of wild queen conch s collected in March 2009 Sex Tissue Site/hold n Mean 65 Cu (ng/mg) 65 Cu SEM p Mean 66 Zn (ng/mg) 66 Zn SEM p F Blood NS held 3 1850.59 91.798 150.99 33.425 OS held 4 1657.31 92.038 58.59 10.222 OS 4 1773.04 107.244 63.67 23.928 DG NS held 3 17.73 1.389 1285.56 234.681 0.04 1 OS held 4 15.11 3.261 585.14 81.195 OS 4 18.90 7.549 440.16 130.038 Mantle OS held 3 18.76 5.600 16.82 3.742 Muscle NS held 3 2.75 0.717 15.90 3.471 OS held 4 3.08 0.223 13.82 1.924 OS 4 3.76 0.331 17.70 5.418 NG OS 1 3.33 16.22 Ovary NS held 3 18.57 5.811 465.38 137.019 0.022 OS held 4 20.66 11.388 40.13 10.023 OS 4 24.28 7.862 71.79 8.740 M Blood NS held 4 1715.17 93.575 96.56 23.850 OS held 3 1841.77 125.399 59.19 15.216 OS 4 1808.54 89.981 43.18 2.725 DG NS held 4 8.85 0.123 771.14 42.989 0.030 OS held 3 31.88 8.777 217.10 100.049 OS 4 15.40 3.202 184.12 26.887 Mantle OS held 2 21.00 5.710 14.69 3.135 Muscle NS held 4 1.77 0.209 0.029 15.81 2.089 OS held 3 4.17 0.871 22.79 6.078 OS 4 5.77 1.837 13.55 2.168 NG NS held 4 5.45 0.968 20.38 9.337 0.04 6 OS held 3 5.38 0.845 111.24 4.927 OS 4 8.84 2.726 11.08 1.094 Testis NS held 4 14.32 1.477 109.73 75.810 OS held 3 20.28 5.194 42.07 29.663 OS 4 11.86 3.772 13.02 1.161 value for a Kruskall Wallis nonparametric test Site/hold Site/hold held overnight in NS flow through water.
170 Table 6 2. Metal concentrations in algal samples collected from sites of conch aggregations in the Florida Keys, March, 2009. Genus Location n 65 Cu (ng/mg) 66 Zn (ng/mg) 111 Cd (ng/mg) 208 Pb (ng/mg) All NS 17 2.65 0.48 15.18 3.38 0.07 0.02 1.27 0.1 6 ** All OS 17 1.63 0.18 51.61 32.20 0.33 0.08 # 0.65 0.11 Batophora NS 10 2.40 0.60 ** 9.40 1.85 0.03 0.00 1.28 0.10 ** Batophora OS 5 1.00 0.07 6.74 2.76 0.05 0.01 0.62 0.05 Dictyota NS 1 2.86 13.46 0.07 0.99 Dictyota OS 3 1.41 0.22 25.26 12.73 0.35 0.13 1.16 0.46 Halimeda NS 1 1.04 13.54 0.04 0.48 Laurencia NS 1 0.78 2.51 0.03 0.10 Laurencia OS 5 2.17 0.45 137.03 106.79 0.21 0.08 0.51 0.18 Penicillus NS 2 2.44 0.58 34.40 16.28 0.14 0.06 1.66 0.22 Sar gassum OS 3 2.00 0.30 22.06 7.96 0.75 0.16 0.43 0.11 Values reported as mean SEM. Mann Whitney test was used to compare OS and NS values for each metal, based on the chi square approximation. Symbol placed next to the greater mean: *p<0.05; *p<0.01; # p<0.001. Table 6 3 Water chemistry data from the 50 day in vivo Cu and Zn feeding study n CTRL n Cu 200 n Zn 2000 76 34.0 0 0.07 75 33.9 0 0.06 75 33.8 0 0.06 pH 58 7.76 0.02 58 7.75 0.03 58 7.76 0.02 Temperature ( o C) 76 25.0 0 0.14 75 25.0 0 0.14 75 24.9 0 0.13 DO (mg/L) 76 5.38 0.03 75 5.30 0.03 75 5.34 0.03 No significant differences in any parameter were detected by Kruskall Wallis nonparametric test. Table 6 4 Histological scores for ovarian developm ent of female fighting conchs from the in vivo Cu and Zn feeding study. Day Treatment n Developing Spawning Capable Regressing Regenerating Atresia No Atresia 96 Control 2 0 0 0 2 0 2 0 Control 3 1 1 0 1 0 3 7 Control 3 0 1 1 1 2 1 Cu 1 1 0 0 0 1 0 Zn 4 0 3 1 0 2 2 14 Control 4 0 3 0 1 1 3 Cu 3 0 1 2 0 3 0 Zn 4 0 2 2 0 3 1 50 Control 4 1 1 2 0 2 2 Cu 4 0 2 2 0 4 0 Zn 4 2 1 1 0 3 1 Developing, Spawning Capable, R egressing and Regenerating give the number of conchs in each set of stages, a dapted from the terminology of Brown Peterson et al.  Atresia and No Atresia columns give a binary indication of whether any atresia was evident in the section.
171 Figure 6 1. Concentr ations of Zn and Cu in the tissues of female and male queen conchs collected in March 2009. Zn (A and C) and Cu (B and D) concentrations in the tissues of female (A and B) and male (C and D) wild queen conchs captured March, 2009. *Indicates groupwise si gnificance in Kruskall Wallis test (p<0.05). All values expressed as ng analyte/mg tissue, except for blood (ng analyte/mg total protein). n = 3 4 per group, enumerated in Table 6 1.
172 Figure 6 2. Vitellogenin mRNA expression in ovary of queen female q ueen conchs collected in March 2009 reported as mean + standard error of copy number/ng total RNA. Signific ant by one way ANOVA (p = 0.008 ); groups not connected by the same letter are significantly different according to Tukey Kramer HSD. NS nearshor e; OS offshore; held conchs held overnight in NS flow through water prior to processing. Sample sizes in parentheses. Digestive gland values are not reported because seven of 11 samples showed no amplification or only non specific amplification. Fi gure 6 3 Tank water temperatures during pre exposure and exposure periods for the 50 day in vivo Cu and Zn feeding study Tank water measurements represent the mean of all tanks, measured each day. Key West and Cedar Key, FL, ambient water temperature data were obtained from NOAA as the mean of each month  Cooling, heating, and experimental (exposure) periods are separated by vertical black lines.
173 Figure 6 4. Representative images of ovarian histology in fighting conchs A digestive gland (to r ight of panel) and ovary (to left, with large eosin patches staining vitellogenin) are closely associated in digestive gland. Ovary is primarily comprised of follicles and signet tissue. Total axial length is ca. 10 mm. B a follicle from a Spawning Ca pable ovary showing several stages of oocyte devel opment. C several collapsed follicles from a Regressing ovary, also showing evidence of atresia Note early vitellogenic oocyte is small and irregular, and could be in the process of resorption. Symbol s: ca, cortical alveolar oocyte; DG, digestive gland; ev, early vitellogenic oocyte; f, follicle; lv, late vitellogenic oocyte; nu, nucleus of a late vitellogenic oocyte; pg, primary growth oocyte; ph, phagocytosis indicative of atresia ; st, signet tissue
174 Figure 6 5 Concentrations of Cu (A and B) and Zn (C and D) in female fighting conch digestive gland (A and C) and ovary (B and D) from the in vivo feeding study. No differences were found within time point by Kruskall Wallis test. Sample sizes giv en in parentheses. Cu (E) and Zn (F) trends over time in digestive gland of In vivo Cu and Zn fed fighting conchs; treatment and control regression: no significant trend in control for either analyte (p = 0.673 for Zn 0.159 for Cu); significant trends for Cu in Cu 200 (R = 0.553, p = 0.032) and Zn in Zn 2000 (R = 0.715, p = 0.002) treatment groups. Errors were normally distributed (Shapiro Wilk) and variance was constant Note differe nt y axes. DG digestive gland.
175 Figure 6 6 Alignment of Strombus alatus vitellogenin partial mRNA sequence with the Strombus gigas sequence. Strombus alatus clone (Sa VTG F2R1 C1rc) and Strombus gigas clone (Q QC VTG F2R1 C2rc) used to construct rea l time RT PCR standard curves. Strombus gigas cDNA library sequence is labeled
176 Figure 6 7 V itellogenin mRNA expression in the ovary of S. alatus from the 50 day in vivo feed ing study. A VTG mRNA versus histological stage (p=0.020 ANOVA, Tukey Kramer HSD); groups not connected by the same letter are significantly different. B VTG mRNA versus percentage class of section showing oogenesis (not significant, Kruskall Wallis). C VTG mRNA steady state levels in Strombus alatus ovary throughout 50 day feeding study (n o significant differences in ovary were found within 14 or 50 d time point s, Kruskall Wallis). Sample sizes in parentheses. Digestive gland values were not stat istically analyzed due to the large number of samples with no amplification. Abbreviation: SC Spawning Capable.
177 CHAPTER 7 DEVELOPMENT OF METHO DS TO MEASURE TRACE METAL CONCENTRATIONS IN CONCH SHELL Background In the previous chapters, it has been dem onstrated that zinc accumulate s in the digestive glands of queen conchs at NS sites in the Florida Keys, and that this is associated with a lack of reproductive development Presumably, if metal deposition associated with human activity was a causative fa ctor in the reproductive effects observed at nearshore sites, this would imply that zinc deposition and possibly the deposition of other metals has increased at those sites over time. This approach has been used to analyze environmental contamination of lead, copper, zinc, cadmium and other metal s in the whole shells of live collected My t ilus edulis  Richardson et al.  used laser abl ation ICP MS to study the trends in Cu, Pb, and Zn across shell growth lines in the horse mussel Modiolus modiolus On the other hand, Puente et al.  argue that only lead and nickel concentrations were correlated between Mytilus galloprovincialis nacreous shell and sediment concentrations, due to the detoxification of trace metals thr ough vesicle storage in the mussel. In the gastropod Cepea nemoralis Jordaens et al. were able to identify differences in shell Zn concentrations that correlated with shell strength and thickness between polluted and unpolluted sites  In another gastropod, Haliotis diversicolor supertexta Lin and Liao  found that aqueous Zn exposures led to accumulation, and that concentrations of Zn in shell decreased with depur ation. Therefore, metal accumulation has been successfully identified in shells of both bivalves and gastropods. Given the success that researchers have had identifying metal concentrations and even trends across growth lines in mussel shells as well as measuring accumulation in
178 gastropods studying metal concentrations in the shells of conch s might be a useful approach not for biomonitoring, but for discerning historical trends. In order to determine whether temporal trends exist in conch shell metal concentrations a method was first developed to determine concentrations of metals in the shell accurately. After developing that method, strombid shell samples from the Florida Museum of Natural History (FLMNH) collections, serve d as an ideal sample set to test for historical patterns. It was hypothesized that: 1. a diamond or titanium nitride coated dental bur would serve to collect calcareous material from shell or other shell like materials without imparting Zn or Cu contamination, and 2. Zn, Cu, and possib ly Pb concentrations would show an increasing trend over time in conch shells from the Florida Museum of Natural History. Results Consistency of Data from Bur Type Tests Note that this preliminary analysis was somewhat problematic, as the standard curve me asurements indicated unusually high percent error for low concentration standards of both 65 Cu (334.34% at 5 ppb and 112.65% at 10 ppb, but only 11.56 % at 50 ppb in the digested sample ) and 66 Zn (449.28 % at 5 ppb and 81.44% at 10 ppb but only 6.89 % at 50 ppb in the digested sample ) and samples also showed 115 In internal standard recovery lower than 80 % for all but six samples. These problems were likely due to the large amount of calcareous material used for the analysis, and the fact that high concentra tions of alkaline earth metals can be a problem for measurement of trace metals in environmental samples [259,260] Mean SEM measured Cu and Zn values for four technical replicates are given in Figure 7 1, with te st materials arranged in order of increasing hardness. The Kruskall
179 Wallis test detected significant differences across materials for both Cu and Zn in the limestone and marble test groups (p<0.05). While Steel nonparametric test against control did not identify significance for any individual groups, it was clear that carbide and diamond bur values were more different from control samples (crude samples) than titanium nitride values for both chalk and limestone. Marble samples collected by all three bur types showed considerably greater mean metal values than control. These data supported the use of the titanium nitride bur for collecti ons, assuming that shell samples should not be a s hard as marble. Trends in Metal Concentrations from Monroe County, FL Strombid Shells No significant differences were detected in any mean metal concentrations across species, and so all three species ( S. gigas S. alatus and S. costatus ) were analyzed as one group. Note that six of 41 samples sho wed 115 In internal stand ard recovery values below 80 %. While this was not a large number and was an improvement over the validation study it could potentially have an impact on the accuracy of measurements. However, error estimates based on the standard curve were much lower f or this set of samples. No significant trend in zinc or copper was observed in the shell samples used for this study (Figure 7 2 A and C, respectively). Lead however, decreased significantly 2 B). Lead was also significantly correlated with zinc ( = 0.545, p = 0.001) and Cu ( = 0.549, p = 0.001), and zinc and copper were significantly correlated ( = 0.542, p = 0.001). While all of these Rho values are significant, they do not appear to be very strong associations. The most apparent trend in the data is the negative correlation between lead and year of collection.
180 Repeated samples from t he S. costatus s hell coded 123708 showed a greater degree of variability than expected, particularly for measured zinc values (Figure 7 2 D). However, these did not appear to indicate contamination from the bur or the collection process in general, as the measured concentrations actually decreased with repeated sampling. Therefore, these likely indicate di fferences in metal concentration with depth of sample, an important consideration for future sampling. Discussion Collection of shell material using a dental drill with a titanium nitride bur appears to be an appropriate method for gathering raw material f rom which to analyze conch shell metal concentrations and titanium nitride, but not diamond, burs accomplish this task without imparting contamination which supports our hypothesis The large size of conch shell can make this process more difficult than for mussels, in which the entire shell can be processed  ; in this analysis the average mass of shell sample was 36.61 mg, a relatively small mass The titanium nitride bur showed no signs of contributing to measured metal concentrations either in the validation study (Figure 7 1 ) or in the shell sampling from FLMNH samples (Figure 7 2 D). However, when available, o ne possible improvement to this method would be to use laser ablation  which would not require direct contact between a metallic sample apparatus and the sample. The method presented here is likely to have a considerably reduced cost per sample relative to laser ablation, thoug h. Further, sample processing using the dental drill is rapid, taking only about 3 5 minutes per sample. Our second hypothesis was that increasing trends in metal contamination would be observed over time in the historical shell sample set. This was not supported, as Cu and Zn did not show trends, while Pb showed a decreasing trend. While few trends
181 were observed in the current study, the inverse correlation of lead concentrations with time is interesting. As previously mentioned, Puente et al.  found lead and nickel to be the only metal s for which Mytilus galloprovincialis nacreous shell concentration correlated with sediment concentrations at the site of collection, and attributed this to the fact that the kinetics of metals such as Zn and Cu are regulated such that large amounts should not be transported into shell. This may in f act mean that queen conch shell is not a likely candidate for identifying historical trends in Cu or Zn ; however, Zn and Cu were detectable in conch shell, and might vary based on environmental factors in other sample sets. Notably, anthropogenic Zn and C u might not be as widely distributed in the environment as anthropogenic Pb. Further, most of the shells used for Tortugas, rather than NS sites in the middle and uppe r Keys. The decrease in Pb concentrations across the study could be related to increasing regulation of products that historically contributed to a great deal of the environmental lead burden, including gasoline, solder, and paint  ; however, one would expect the peak lead levels in the Florida Keys to have occurred after the 1930s. This trend might be clarified with additiona l samples. Further, it is possible to use isotopic ratios to assess sources of lead contamination, and this might be an interesting improvement  While this method might be better suited for detecting trends in lead concentrations than for zinc or copper, several improveme nts might be made before ruling out the use of this approach for metals other than zinc. Principally, it would be worth measuring additional samples, taken from a larger range of times and locations. According to the algal metal concentrations reported i n Chapter 6, I do not have reason
182 to believe that the Florida Keys are currently heavily contaminated with metals such as Cu and Zn. Therefore, a heavily contaminated site would make for an interesting comparison. Further, it could be interesting to atte mpt sampling from different shell locations, including the nacreous shell that was analyzed in Mytilus galloprovincialis by Puente et al.  Different locations on the shell could have different metal accumulation potentials, and some locations might be more or less variable than others, providing considerable room for refinement of th is technique. Ultimately, while this technique was successful in accurately measuring shell metal concentrations, and may have identified a meaningful trend in lead concentrations in shells of conchs collected from the Florida Keys, the present results do not suggest significant temporal trends in Zn or Cu concentrations. Interpretation of this result should be tempered, however, by the uncertainty concerning the likelihood of Cu or Zn to be found in shell, let alone to show correlation between shell conce ntrations and environmental concentrations. Assuming that Zn and Cu levels are likely to have increased over time in the area of interest, a more recent sample group would be a benefit to this study. It is possible that conchs collected live after 1989 w ould show increased Cu or Zn concentrations in shell relative to older samples. Therefore, this method is capable of discerning historical trends, an d more effort should be made, in order to determine whether environmental sources of Zn and/or Cu have inc reased since the reproductive abnormalities in NS conchs have become apparent.
183 Figure 7 1. Cu (A) and Zn (B) concentrations from calcareous materials used for bur tests. Mean SEM for each group, based on four technical replicate measures. Abbreviati ons for bur types: none crude sample used as control, C carbide, D diamond, T titanium nitride coated. *Above the control (none) group indicates significance within that material according to the Kruskall Wallis nonparametric test (p<0.05)
184 Fig ure 7 2. Cu and Zn concentrations in archived conch shell samples from the Florida Museum of Natural History 65 Cu (A) and 66 Zn (C) showed no significant trend over time for the sampling period of 1936 1989 (n = 35) ; however, the highest value of Zn was found in a sample from 1936. 208 Pb decreased significantly throug hout the sampling period (B). The r epeated S. costatus shell sample (D) showed a variable decrease across measures, especially for 66 Zn; arranged by order of collection. O1 O4 samples; N1 N4 bur samples.
185 CHAPTER 8 CONCLUSIONS This purpose of this dissertation was to better understand the biological processes involved with reproductive development in conchs, and how those processes might be inhibited by factors in the NS environment of the Florida Keys, in particular by exposure to Zn and/or Cu. The initial report that conchs at NS sites fail to reproduce was published by Glazer and Quintero  who state that no spawning (i.e. egg laying) behavior was observed at NS sites throughout the course of greater than 400 surveys conducted between 1996 and 1998. Subsequent publications confirmed that in we ekly surveys of 49 tagged NS females from March to November of 1999, no spawning was observed [3,9] Conversely, NS conchs translocated to OS sites, reproductive behavior was observed in the season following transl ocation; spawning was observed for 12.2% of 41 females observed during summer and for 18.5% of 27 females observed during fall. These percentages lagged behind those for resident OS conchs, which peaked at 46.2% (n = 39 females) spawning during spring  In the same experiment, the rate of spawning for OS conchs translocated to NS sites decreased from 8.9% (n = 56 females) in spring to 4.7% ( n = 215 females) in summer, and none of the OS to NS transplanted females were observed spawning during fall (n = 21)  Histolog ical observations of the gonads of conchs at NS sites indicated that all OS conchs (male and female) had more advanced go nad development than all NS conchs (male and female) in the 1998 report, based on a small sample size of seven total conchs  Subs equently, a larger sample size (n = 21 to 35 resident conchs of each sex from each of three seasons; n = between 10 and 26 translocated conchs of each sex during summer and fall) found that the histological disparity between NS and OS
186 was severe, with no m ore than summer and about 40% of Spawning Capable males in the fall. Among translocated conchs, about 60% of females were Spawning Capable by fall, and about 75% of males were Spawning Capable by summer. Therefore, NS conchs always showed reduced gonadal development, relative to OS, in resident populations, while translocated conchs showed significant capacity to develop gonad tissue after approximately three months of translocation  All of these ob servations helped to form the basis for the hypotheses presented in the current study. The major hypotheses presented in the Introduction of this dissertation were either fully or partially supported by the work presented in this dissertation. First, micr oarray analysis of gene expression did reveal considerable insight into the disparity in reproductive development between NS and OS conchs. In the testis of male conchs, effects on spermatogenesis related genes were associated with effects on genes relate d to small GTPase mediated signaling, which might be important for the progression of testis development or spermatogenesis (Chapter 4). The major processes effected in NS ovary included apoptosis, which may play a role in atretic processes of regression, and translation, as it appears that the production of ribosomes and associated proteins is halted at the transcript level. In the digestive gland of NS females, protein and lipid metabolism appear to be significantly altered, suggesting that limitations in the ability of the digestive gland to support ovarian development might contribute to reproductive dysfunction NS (Chapter 5). The failure of NS conchs to reproduce is clearly the result of a failure to develop sufficient gametogenic tissue in the gona d to support reproduction, as has also been repeatedly observed in conchs
187 collected in 1996  1998  2007  and 2009 (Nancy Brown Peterson, personal communication). Secondly, it was hyp othesized that NS conchs would have higher overall concentrations of metals. This was borne out in the fact that the disparity in reproduction was coupled with a consistently higher concentration of Zn in the digestive gland of NS conchs (Chapters 4, 5, a nd 6). There were also trends in Zn in the gonad, as well as trends in Cu, but these were not confirmed statistically. The relationship between accumulation of Zn in the digestive gland of NS conchs and NS conch reproductive dysfunction has been observed in male (Chapter 4)  and female (Chapter 5) queen conchs in 2007, as well as both male and female conchs in 2009 (Chapter 6). While this correlation has been thoroughly established, understanding the relationship between Zn accumulation and reproductive failure requires deeper consideration Zn is a trace metal with many physiological roles [84 87] including as an inhibitor of apoptosis  and also as a likely cofactor for spermatogenesis and oogenesis in several vertebrate species [180 183] However, excess Zn exposures are al so known to reduce fecundity, fertility or time at sexual maturity in several gastropod species [ 66 69] This phenomenon will be examined further. Finally, it was hypothesized that in vivo exposures to Cu or Zn would lead to accumulation of the metal and reduced reproduction in a strombid. This hypothesis was partially supported, as the trends in m etal accumulation over time in the digestive glands of Florida fighting conchs, Strombus alatus were significant. However, the accumulated values were quite variable both in treated and control conchs, likely because of differences in feeding rates and p hysiological requirements for the metals.
188 Effects on reproduction after 50 days were less certain. Zn treated conchs showed evidence of atresia in the ovary more often than control conchs, but sample size was low, and significant atresia was observed in some controls. Zn treated conchs also showed a higher mean VTG mRNA expression level at 14d than 50d, unlike control or Cu treated conchs. However, there were no significances in groupwise VTG mRNA levels, and this may have been by chance. Ultimately, i t appeared that conchs exposed to Cu and Zn in vivo were more likely to accumulate the metals in the digestive gland, and effects on reproduction were possible, but not confirmed absolutely after 50 days. Evidence for an Exogenous Zinc Source Accumulation of metals in the viscera of gastropod s is generally associated with an exposure to an excess of the metal from an exogenous source, as the digestive gland and kidney appear to be sinks for accumulation of metals such as Cu and Zn in gastropods [53 56,74,80,95] which some authors argue is related to detoxification or elimination of excess metals [53,54,95] However, zinc and copper differ from nonessential metals due to their prominent physiological roles in all organisms, their ubiquity, and their carefully controlled absorption, distribution, metabolism, and elimination in vertebrates  It seems, then, entirely possible that zinc accumulation in conch digestive gland nearshore could be an effect rather than a cause of the reproductive failure phenomenon both could result from some upstream process, or the two could be unrelated The zinc could be mobilized to the digestive gland due to changes in metabolic processes or reproductive development Further, it is entirely possib le that the entire zinc content of the digestive gland is effectively non toxic, due to the detoxification of metals in digestive gland. Ultimately, however, it seems likely that zinc accumulation is a sign of an exogenous exposure. This trend is occurring
189 nearshore, where metals are more likely to be found at higher concentrations [96 98] and accumulating in the tissue responsible for detoxifying metals. Gay and Maher  indicate that numerous factors can affect the concentrations mass and size, gender, reproductive state, accumulation/regulation strategies, and diet and extrinsic factors such as temperatur e, salinity, supply of metals, food availability and metal metal relationships influencing the Zn concentration in conch digestive glands in the Florida Keys. The conchs studied in this dissertation were of similar size, in terms of length and soft tissue mass, despite the differences in shell thickness, and are believed to be of similar age, again despite differences in shell thickness. Reproductive status is a consideration, but elevated Zn was obs erved in NS conch digestive gland in three months: February, March, and June, covering most of the reproductive season. While a difference was clearly observed between NS and OS digestive gland in terms of reproductive status, this was not, in other word s, a seasonal effect. Further, if Zn was being mobilized to digestive gland for reasons related to Zn need in the developing ovary or testis, elevated Zn would likely be observed in OS (developed) ovary and testis, but this was never the case. As far as e xtrinsic factors, the overall view taken in this dissertation was that exogenous supply of Zn or Cu could lead to the accumulation and effects on reproduction. Therefore, differences in supply of metals remain possible, despite the lack of significant dif ferences in algal metal concentrations measured in the field. Again, concentrations of Cu and Zn appear to be greater close to shore in several south
190 Florida media [96 98] ; Manker  reported a baseline Zn concentration of approximately 2 ppm in Florida Keys sediments, with concentrations near heavily populated areas ranging as high as 100 ppm. Metal metal relationships could occur in accumulation, especially with Zn, Cu, or Fe  as discussed in Chapter 5, when microarray results indicated effects on Fe transport mechanisms. Results of the preliminary 60 day in vivo experime nt showed a correlation between Cu and Zn concentrations in gonad, which may have indicated co transport (data not shown). The 50 day in vivo dataset and the field datasets, however, do not show such a correlation, and so the preliminary data may have bee n inf luenced by the high concentrations of Zn in the digestive glands of conchs from the preliminary study. It was previously determined that water chemistry factors are not likely to differ on average between sites, but the NS site may undergo greater ex tremes, which could also contribute to Zn accumulation. Finally, differences in diet could contribute to accumulation of metals. However, in the admittedly small sample reported in Chapter 6, both sites had numerous samples of Batophora a preferred food of conchs that is apparently generally low in metals. An algal genus of interest might be Laurencia which displayed higher and more variable Zn concentrations than the others in the sample. Overall, reasons for Zn accumulation other than elevated NS Zn do exist. However, Zn concentrations are likely to be higher closer to shore in the Florida Keys [96,97] Therefore, despite variability of Zn concentrations in wild samples, and the lack of an absolutely apparent source, it is likely that Zn accu mulation in NS conchs is related to an exogenous exposure.
191 In Vivo Zinc Exposure: Effects on the Ovary F ield data suggest a likely causal association between Zn exposure accumulation and subsequent failure to reproduce in NS queen conchs, while the effe ct of an exogenous Zn or Cu exposure on ovarian development in a controlled setting was not conclusive (Chapter 6). Zn and Cu accumulation in the digestive gland of conchs was quite variable, as should be expected for a trace metal in a feeding study, whe re rates of uptake and utilization of the metals are likely to vary from individual to individual. As shells  which could contribute to differences in feeding rate on the foo d used in the study. T he conchs did, however, appear to feed on the artificial diet readily throughout the course of the study. While Zn and Cu accumulate d over time in the digestive gland of exposed Florida fighting conchs, the development or regression of the gonad was not consistent across treatments Ultimately, due to the small sample size, the experiment may not have had the predictive p ower to determine subtle differences with significance. It seemed that Zn may have led to accelerated development of the ovary, followed by a premature regression, although this trend was not significant statistically. The connection between apoptosis an d atresia  makes this more compelling, especially as apoptosis was implicated in the NS ovary microarray analysis (Chapter 5). I f this pattern of acceler ated development and regression is real it would match rather closely with the pattern observed in NS male queen conchs in which development of the testis begins in the early reproductive season, but is followed by an untimely regression process (Chapter 4)  In the in vivo study, atresia was observed in some con trol ovaries Still, fewer atretic ovaries were observed in 14 d and 50 d controls than in Zn or Cu exposed
192 conchs, and this does support to some degree the idea that Zn and Cu have deleteriou s effects on conch oogenesis through an atretic process of reg ression. It should be mentioned that Zn has a known role in anti apoptotic processes in humans, and might be central to the control of caspases  The fact that in conchs, Zn accumulat es primarily in digestive gland, rather than directly in gonadal follicles, indicates that its likely impact on gonad development is indirect. In other words, exposure of ovarian follicles to Zn directly may not be the cause for the observed atresia or ch anges in apoptotic processes. Rather, Zn accumulation in digestive gland could indicate that detoxification of the metal results in an indirect effect, such as an effect on the overal l energy budget available for the costly process of reproduction. In fa ct, the basic concept of dynamic energy budget models in toxicology is that animals require significant energy for growth, developme nt, and reproduction, and the available energy pools can be affected by toxicants  Ducrot et al.  demonstrated that a dynamic energy budget model can be used to describe the effect of Zn exposure on reproduction in the gastropod Valvata piscinalis which provides further support for the concept that accumulation and detoxification of Zn may reduce available energy for reproduction or delay age at reproductive maturity in queen conchs. A major ques tion still remains to be asked, though: if Cu and Zn accumulation from an exogenous food source interferes with reproduction, and if the HBOI food was deemed too high in zinc to act as a control in the feeding studies (after causing accumulation of very h igh zinc concentrations in digestive gland), how is it possible that conchs have reproduced in aquaculture at HBOI [27,47,110] ? This is certainly a question worth asking, and it may have implications for the design of our study. It would
193 be interesting to know whether reproductive output remains consistent over time at HBOI or decreases with additional time in the system the latter condition would support the idea that Zn in the food could reduce reproduction. I t was suggested that the Florida fighting conch might not be a good choice for this proxy study, because it is ue, this could be a reasonable assumptio n. Fighting conchs in the wild have a tendency to bury themselves in muddy sediment s [234,235] where they would be in contact with any contaminants associated with soil particles Further, to my knowledge, queen conchs have only reproduced in aquaculture one time at HBOI (or anywhere in the world) while fighting conchs reproduce more readily in culture  Therefore, it is possible that the aquaculture conditions are more stressful for queen conchs than for fighting conch s supporting the physical or chemical environmental stressors. Potential Sources and Effects of Cu and Zn in the Nearshore Environment As previously noted, there is reason to believe that zinc, copper, and other metal concentrations would be hig h er close to s hore, where there is more human activity including boat traffic, sewage systems, plumbing, and fossil fuel use, all of which might contribute to Zn in the NS environment of the Florida Keys [96 98] This influence is further supported by the observation of human fecal contamination in samples of coral collected close to shore  If Zn or Cu from any of these sources is enriched in sediments, detritus, or algal sources close to shore, all of these could be potential routes of exposure fo r conchs, as conchs are known to consume all of these materials [2,13,14,16 19] While this dissertation did not address concentrations of metals in
194 sediments or detritus, concentrations in algal sources were actua lly relatively low, with the exception of a few outlying samples, including especially one sample of Laurencia that was collected OS (Chapter 6). It is likely that Cu and Zn concentrations are variable and are lower than those used in the in vivo study. Therefore, considerable effort might be required to characterize potential routes of exposure for conchs. The deleterious effects of zinc [67 70] and copper [67,71,72,74] on egg laying have been documented in several gastropods. However, little effort has been made to describe mechanistically why this might be the case. For copper at least, one author argues that the mechanism may be tied to lysosomal processes and affect ei ther the production of new oocytes or apoptosis of existing oocytes  In the queen conch, we can make a strong case for enhanced a poptosis at NS aggregations, based both on microarray data (Chapter 5) and the persistence of atretic oocytes in NS conch ovarian histology. Further, as previously mentioned, it is possible that Zn could affect gonadal development, and subsequently reprod uction, through an indirect mechanism related to the energy budget of the conchs [58,69,263] While this dissertation has not defined the role of Zn or Cu as reproductive stressors in conchs absolutely this knowle dge may aid in determining which stressor or stressors cause the reproductive effects of NS environment in future controlled exposure studies. It should also be noted that not all toxicant exposures are equivalent, and that not only the dose, but the sourc e and duration, can determine the poison. While it is assumed that boating activity might be a source of Zn in the NS environment, i nterestingly, one study concluded that the toxicity of zinc dissolved from sacrificial anodes in a ca thodic protection syst em may not be as severe as the toxicity of a Zn salt
195 in sea urchin  Therefore, the various possible sources of Zn in the Florida Keys may not be equivalent in their likelihood to result in accumulation or toxic effects in conch s. Moreover, t oxicity including sub lethal effects can even vary significantly across similar exposures in closely related species. Take for instance the example of copper exposures in the Florida apple snail Pomacea paludosa and the closely related Pomacea canaliculata : Rogevich et al. determined that copper exposure reduced the number of eggs laid and the proportion of eggs that hatched in Pomacea paludosa  Pe a and Poscidio, however, found that copper influenced growth rates but not reproduction for Pomacea canaliculata  It is possible that this dif ference was due entirely to differences between the two species but this is unlikely. When comparing these two studies, th e major difference was in the exposure regime; Rogevich et al. exposed the snails to 8 16 g/L copper for nine months, while Pe a an d Poscidio used a higher concentration (30 g/L) of copper, but failed to see reproducti ve effects after only 20 days. This could have implications for the current situation with conchs, as well. The in vivo exposures only appeared to result in considera ble accumulation of Zn or Cu at the 50 day time point, indicating that a longer duration of exposure may have been required to elicit effects on reproduction. Alternatively, the timing of exposure with relation to development toward reproduction could be altered in future studies. Multiple Stressors Could Inhibit Conch Reproduction Despite the necessity of choosing a controllable set of stressors to investigate, it would be foolish to assume that given the current state of coastal ecosystems an animal in a natural environment could be subject to only one a single anthropogenic stressor. It is clear tha t human activity has altered ocean water s in numerous ways, affecting chemical factors such as oxygen content, loads of nutrients and contaminant s, and
196 other parameter s  Therefore, an effect like the one seen in NS queen conch aggregations could conceivably be caused by a number of different stressors, or even by a combination of several. In fact, this is discussed at length by Delgado et al.  and to some degree by Glazer and Quintero  who first reported the disparity in reproduction between OS and NS conchs, and who suggested that it was likely to be associated with excess nutrient loads in the NS environment. Given a situation such as this, in which it is unclear whether the stressors of interest in this case zinc and copper can cause th e kind of complete inhibition of reproduction observed in NS conch aggregations, this seems all the more likely. Clearly, d etoxifying metals as in the case of accumulating th em in the digestive gland for permanent sequestration or elimination, will surely have an energetic cost  As dynamic energy budget models dictate [69,263] exposure to a stressor would limit energy available for reproduction. It may be the case that NS conchs are exposed to several stressors in addition to zinc, and that the sum of mechanisms of coping with these stressors results in an energetic cost that prohibits exp ense of energy on reproductive development altogether In fact, reproductive development in queen conchs, especially females, is likely to be a major energetic investment, with hundreds of thousands of eggs being prepared for each spawning/egg laying even t [18,23,24] ultimately derived from an organ that develops throughout the reproductive season possibly depending on the albumen  and sugars  that are found in signet tissue, and perhaps requiring the mobilization of lipid from digestive gland and the production of vitellogenin (Chapters 5 and 6), all of which are likely to carry high costs
197 One possib le general issue is that NS conch habitat might be s uboptimal in terms of available food sources, water depth, and the inability to make the migration from NS to OS during development that has been reported by Sandt and Stoner  and Stoner and Ray  In fact, Weil and Laughlin  indicate that conchs may migrat e to deeper water to take advantage of the greater amounts of epiphyte food available at those depths. However, the impact of the inability to make this migration is questionable, given that many adult conchs remain on the bank in some studies of populatio ns not exploited by a fishery  and also by the knowledge that conchs migrate to shallower water during the reproductive se ason [23,24,41] clearly suggesting that deep water is not an absolute necessity for reproductio n. Further, in the small scale survey of algal metal concentrations in this dissertation, Batophora algae, a known source of food for conchs, was collected at NS si tes more often than OS sites. Still, it would require a more extensive survey to determine whether the amount or diversity of food sources at different sites is sufficient to support conchs. It should be mentioned that historical studies of conch density generally indicate lower densities with increasing latitude  which might indicate that the Fl orida Keys are generally suboptimal habitat at the edge of the habitat range of the queen conch. Even if this is the case, there is a clear disparity in the ability of conchs to reproduce at OS versus NS sites, and this is related to a factor in the NS en vironment that differs from OS on a much smaller s patial s cale than the greater Caribbean region Another possible stressor that could cause a disparity in reproduction between habitat patches is the rate of infection with parasites. Aside from obvious ef fects on the overall health and energy available for reproduction, in some extreme cases, parasites
198 can even directly affect the sex of an organism. For example, a clear association between TBT contamination, parasite infection, and decreased reproduction has been established in amphipods in which contaminant load affects the immune system, increases parasite load, which alters the production of ecdysteroids and thereby influences reproductive development  P arasites were beyond the scope o f this dissertation, but several published reports describe p arasites affecting organs that might affect reproduction in conchs. In the conch digestive gland, which I believe to be important for supporting the development of the ovary and possibly the testis, Gros et al.  identified numerous parasites in the digestive glands of conchs which were described by Aldana Aranda et al. as being Aplicomplexa parasites  Additionally, Berg and Olsen  report finding nematodes in contact with the subesophageal ganglia of queen conchs. Effect s on the ganglia could also impact reproduction, as it is known that at least in whelks of the genus Busycon injecting extracts of the nerve tissue will result in egg laying  Therefore, parasites could be an important parameter in conch reproductive success or failure. Overall, some combination of general water quality, parasites, food availability, and chemical stressors such as Zn or Cu could theoretically result in the total failure of NS conchs to develop fun ctional gonad tissue. While a consistent association with Zn accumulation has been established in the NS population, the existence of other potential stressors leaves many more questions to be answered. How Clear is the Nearshore Offshore Distinctio n? While the current view of the queen conch reproductive problem in the Florida Keys is that there exists a clear delineation b etween OS and NS defined by the Hawk Channel, my data support the idea that Delta Shoal (2009 OS ) is an intermediate site. I
199 saw greater digestive gland zinc burdens in Delta Shoal males and females than in OS conchs from Pelican Shoal and Eastern Sambo (2007) (Figure 8 1) This observation was coupled with a smaller fold difference in VTG mRNA expression in 2009 than 2007, suggesting that Delta Shoal conchs might not be as reproductively developed as Pelican Shoal or Eastern Sambo conchs (However, NS c onchs in 2007 also expressed VTG mRNA at a very low level, which influenced the large fold difference) This falls in recent years. The idea that Delta Shoal might be impacted by contaminants to a greater degree than other OS sites including Pelican Shoal and Eastern Sambo is also logical, given that Delta Shoal is located closer to Marathon, adjacent to the same part of the Middle Florida Keys as the NS sites included in this study (Figure 2 1). Rock or Tingler Island, but at lower concentrations due to being farther from shore. Moreover, this supports the idea that development of a quantitative measure of reproductive development, such as VTG mRNA expression, could lead to a more definitive link between zinc accumulation and impairment of reproductive development. How Likely is Nearshore Reproduction to Affect Conch Population Growt h? The queen conch population in the Caribbean is not uniform genetically but there is significant larval drift through the Caribbean  resulting in populations that are genetically similar If there are numerous sources of larvae recruiting to the Florida Keys, then reproduction in the NS aggregations within the Florida Keys is likely to be rel atively unimportant for the local population. However, i f Florida relies mostly on recruitment from Florida, then reproduction in NS aggregations could be very important. This falls in line with the hypothesis of Roberts  who uses ocean current data to
200 determine which sites in the greater Caribbean have the most upstream and downstream area acting as larval source s and sink s Roberts determined th at Florida has a large upstream source for coral larvae, based on a 1 to 2 month window for recruitment. This window likely overestimates the timeframe for recruitment in conch, and the author specifically sites the conch allozyme study by Campton et al. as an example of a population that likely relies on local retention of larvae  Collection of queen conch veligers during plankton surveys con ducted by Stoner et al.  suggested recruitment of conchs from the weste rn Caribbean sea on the Florida Current, although this wa s not believed to be a constant source This was confirmed in a separate survey by Hawtof et al.  However, a more recent study by Delgado et al. argues that lower larval densities in the Tortugas and Florida Straits ( to the west, or upstream in the Florida Current) than the Florida Keys, coupled with the retention of drift vials in Mexican waters and the bypassing of the Florida Keys by drift vials released in the Florida Straits, indicate that it is more likely that r ecruitment to the Florida Keys relies on local sources  Therefore, establishing healthy breeding aggregations in the Florida Keys is likely a crucial step toward recovery of the population. Methods Developed in this D issertation and Associated Projects Prior to the execution of the work presented in this dissertation, little genetic and no genomic work had been performed with queen conch, a species whose biology and life history have been well characterized. While cer tain aspects of this project required considerable optimization, including such seemingly trivial tasks as RNA preparation, progress has been made. The cDNA library and conch microarray produced and designed by Robert Griffitt, Li Liu, and Nancy Denslow w ill be available as a tool for other conch researchers, as will the RNA preparation method and real time RT PCR
201 assays here reported. The shell metal sampling method also appears to be effective and could serve as a useful tool for future studies of spati al and temporal trends in metals. These methods have the potential to be useful moving forward for researchers interested in the basic biology, species health, and management of the queen conch. Refinements and Future Directions Further Work with Zn and C u in the NS Florida Keys The results of this dissertation showed a consistent trend in Zn accumulation in the digestive glands of NS conchs, where reproductive development was also consistently reduced, relative to OS. However, as previously discussed, th e results at the OS site Delta Shoal appeared to be intermediate to NS sites and other OS sites, with higher Zn concentrations than other OS sites. It would be an improvement on the long term understanding of this situation to include conchs from addition al sites, with larger sample sizes, and focus on Cu and Zn trends in digestive gland and ovary only. Larger sample sizes would allow researchers to determine whether the trends in metal concentrations in the gonad or Cu concentration in the digestive glan d are persistent and significant, or are just occurring by chance. Further, there may be some gradient in Zn accumulation in the digestive gland that could be strongly correlated with reproductive status. Furthermore, expanded sampling could include food sources other than macroalgae, including sediments, detritus, seagrasses, and epiphytes. The analysis of algal metal concentrations in Chapter 6 presented some interesting possible trends, but an expanded analysis would be required to identify sources of metals. Notably, in my own attempts to quantify sediment metal concentrations using ICP MS, high dissolved solids caused problems with recovery of the internal standard and so method development may require a considerable investment in time.
202 Investigati ng Other Possible Stressors in the NS Environment It has been discussed that stressors in addition to metal contaminants may exist in the NS environment, and that these may have effects on conch reproduction. A study of temperature and water chemistry dat a at the sites of adult conch aggregations would aid in understanding whether significant differences exist in mean or extreme temperature, dissolved oxygen, or other parameters that could affect reproduction, despite the lack of trends in publicly availab le data from other nearby locations. Perhaps most importantly, a larger scale study of food quantity and quality at NS and OS sites would allow for a deeper understanding of the potential impact of nutrition on reproduction at each site. If survey data i ndicate a likely role for food sources, a food supplementation study would be an appropriate approach to determine experimentally whether enhanced food quantity or quality could improve reproductive development in NS conchs. Further, future studies could expand on the understanding of NS versus OS conch energy available for reproductive development by expanding the periodic acid Schiff staining study in Hawtof et al.  or using other methods that quantify stores of energy molecules such as lipids and carbohydrates in digestive gland and the signet tissue of the gonad. Understanding the impacts of these parameters would also allow for multiple stressor studies with Cu and Zn exposures in vivo Improvin g upon Study Design for Conch Trace Metal Exposures Clearly, based on the pre vious discussion of the difference between a no effect determination in a copper reproductive study with Pomacea canaliculata  and a study that determined copper caused a reduction in egg lay ing in Pomacea paludosa  exposure conditions, duration, and species differences can have significan t effects on toxicolog y studies, and these consideration s are sometimes unpredictable. The
203 results of the in vivo copper and zinc time course feeding study were somewhat ambiguous, future studies could be refined in several ways. First, it is possible that Strombus alatus is not an adequate model for this comparison. While this study could not be conducted with queen conchs, a more suitable surrogate could be another large, primarily sand and seagrass bed dwelling conch such as the milk conch or hawkwing conch. Further, the dosing approach could be altered to more likely produce an effect. Rather than beginning the dosing regimen after development has begun for all groups, starting to feed copper and zinc during the very early part of development could model the natural situ ation more closely. Alternatively, or perhaps additionally, a depuration phase could model the translocation studies performed by Delgado et al.  while also gland. I mplications for Conch Management: Translocation Efforts Whether Zn, Cu, or a combination of stressors is responsible for the lack of reproductive development in NS conchs, the NS environment appears to be responsible. NS conchs translocated to OS develop mature gonads capable of supporting reproduction  and, unlike resident NS conchs, have been observed laying eggs [3,9] while OS conchs translocated to NS show decreased rates of egg laying relative to OS conch s that remain at OS sites  Delgado et al.  discussed the fact that, despite the successful translocation and subsequent observations of reproduction in both male and female conchs questions still re main for the long term implications of a translocation pr ogram. These questions include the selection of optimal habitats for translocation where conchs will reach a density of 200 conch s ha 1 which is optimal for reproduction the effect of removing conchs from the NS environment, and the
204 likelihood that increased larval production OS would result in increased larval retention in the Florida Keys. While the work presented in this diss ertation cannot address the latter two questions, it seems to provide important considerations for the first. With respect to predicting locations where reproduction will be successful, I would recommend measuring the concentration of zinc in the digestiv e glands of conchs resident at that site, as well as using the VTG real time RT PCR assay during the reproductive season in an effort to quantify reproductive development. This information, paired with further study of factors including food as described in the previous section, as well as substrate, size of the habitat patch, and others, could provide solid guidance for the selection of destin ation sites for translocation. Closing Statements This dissertation has led to an improve d o verall understanding of the conch, its reproductive biology, and its reproductive status with in the Florida Keys While I have not definitively answered the multifaceted question of what is causing NS conch reproductive failure, the work presented here has added to the body o f knowledge in a way that I believe is important Ultimately re establishing a thriving conch population will require the expertise of individuals in numerous disciplines. I have used molecular biology and analytical approaches to strengthen the knowled ge of the association between zinc accumulation and gonadal development in NS aggregations. This will ai d in regulatory decisions and future monitoring, and will improve the understanding of the basic biology of the queen conch. In this way, I hope that it contributes to the well being of the conch population and the health of the Florida Keys ecoregion
205 Figure 8 1 Comparison of digestive gland Zn concentrations in conchs from 2007 and 2009 field studies. Sample groups identified by sex, M or F; site ESR East Tingler Island (NS), DS Delta Shoal (OS), ES Eastern Sambo (OS), PS Pelican Shoal (OS); whether or not they were held overnight in NS flow through water, and the date of the sample. No statistical comparisons we re made as these data were collected during different ICP MS sampling runs and are reported separately in Chapters 4, 5, and 6.
206 APPENDIX A LIST OF DIFFERENTIAL LY REGULATED PROBES IN THE TESTIS MICROA RRAY EXPERIMENT Object A 1 is a comma delimited Excel f ile contai ning a list of all differentially regulated probes from the microarray experiment performed with testis from male queen conchs collected February, 2007 ( reported in Chapter 4). The table includes the Probe Name of each probe, Gene Title for any probe anno tated with a Gene Title, difference between log 2 NS mean and log 2 OS mean for each probe (Diff of Treatment = (NM) (OM)), and the p v alue determined by ANOVA with 5 % FDR (False Discovery Rate) to control for multiple comparisons. Object A 1. List of diffe rentially regulated probes in the testis microarray experiment (. c sv file 50 KB)
207 APPENDIX B CONCENTRATIONS OF AL L METALS MEASURED IN 2007 WILD CAUGHT MALE QUEEN CONCHS Table B 1. Concentrations of metals measured in 2007 wild caught male queen conchs. An alyte Organ n (NS) Mean (NS) (ng/mg) SEM (NS) TK n (OS) Mean (OS) (ng/mg) SEM (OS) TK 58 Ni Blood 8 0.11 0.028 5 0.06 0.027 DG 4 24.52 7.382 6 30.24 4.486 Foot 5 0.45 0.425 6 0.15 0.103 NG 7 2.47 2.075 6 0.53 0.120 Testis 6 3.24 1.386 5 1. 41 0.248 65 Cu Blood 8 40.18 5.514 5 58.90 10.008 DG 4 13.28 1.395 6 7.72 0.874 Foot 5 2.00 0.264 6 3.72 0.468 NG 7 24.70 18.745 6 10.97 4.919 Testis 6 34.77 14.431 5 6.60 1.064 66 Zn Blood 8 1.66 0.304 b 5 0.80 0.101 b DG 4 831.85 138 .771 a 6 84.53 31.689 b Foot 5 6.17 0.450 b 6 11.05 6.250 b NG 7 24.42 18.350 b 6 7.69 2.168 b Testis 6 83.96 49.359 b 5 5.43 0.698 b 88 Sr Blood 8 7.44 0.548 5 7.18 0.395 DG 4 21.55 5.557 6 20.61 1.695 Foot 5 8.71 0.632 6 7.59 1.260 NG 7 23.99 15.192 6 8.75 3.578 Testis 6 18.34 9.785 5 11.91 2.284 107 Ag Blood 8 0.08 0.019 b 5 0.80 0.091 b DG 4 0.04 0.010 b 6 3.90 1.585 a Foot 5 0.04 0.012 b 6 0.14 0.063 ab NG 7 0.72 0.639 b 6 0.21 0.115 b Testis 6 0.18 0.038 b 5 0.26 0.053 b 111 Cd Blood 8 0.05 0.011 c 5 0.03 0.010 c DG 4 4.76 0.746 b 6 9.58 1.084 a Foot 5 0.44 0.053 c 6 0.51 0.039 c NG 7 0.98 0.739 c 6 0.30 0.050 c Testis 6 0.67 0.300 c 5 1.35 0.939 c 118 Sn Blood 8 0.01 0.000 5 0.01 0.000 DG 4 0.02 0.004 6 0.0 3 0.010 Foot 5 0.02 0.001 6 0.02 0.003 NG 7 0.12 0.044 6 0.09 0.027
208 Table B 1. Continued. Analyte Organ n (NS) Mean (NS) (ng/mg) SEM (NS) TK n (OS) Mean (OS) (ng/mg) SEM (OS) TK Testis 6 0.04 0.009 5 0.02 0.005 202 Hg Blood 8 0.02 0.003 b 5 0.02 0.005 ab DG 4 0.03 0.016 ab 6 0.28 0.074 ab Foot 5 0.43 0.166 a 2 0.05 0.024 ab NG 7 0.25 0.124 ab 6 0.18 0.079 ab Testis 6 0.05 0.008 ab 5 0.25 0.085 ab 238 U Blood 8 0.03 0.006 5 0.01 0.000 DG 4 2.03 0.535 6 1.10 0.247 Foot 5 0.02 0 .001 6 0.02 0.003 NG 7 0.23 0.151 6 0.12 0.035 Testis 6 0.32 0.172 5 0.03 0.005 followed by Tukey Kramer HSD for multiple comparisons. Within each analyte, values not connected by the same letter are significantly different.
209 APPENDIX C CONCENTRATIONS OF AL L METALS MEASURED IN 2007 WILD CAUGHT FEMALE QUEEN CONCHS Table C 1. Concentrations of all metals measured in 2007 wild caught female quee n conchs. Analyte Organ n (NS) mean (NS) (ng/mg) SEM (NS) n (OS) mean (OS) (ng/mg) SEM (OS) p (K W) 58 Ni BL 5 4.43 0.833 2 4.74 3.959 DG 5 17.12 3.208 9 34.57 6.224 0.0388 G 4 1.57 0.422 7 3.46 1.365 M 4 0.06 0.020 9 3.95 3.893 NG 5 1.15 0.694 9 1.12 0.490 65 Cu BL 5 1943.01 348.704 2 1505.36 633.831 DG 5 40.29 14.656 9 24.13 4.095 G 4 15.50 6.934 7 20.78 9.528 M 4 3.10 1.321 9 4.61 0.793 NG 5 22.36 18.647 9 6.36 1.805 66 Zn BL 5 69.92 8.934 2 39.44 1.042 DG 5 1181.76 314.182 9 108.45 29.728 0.0278 G 4 106.90 37.048 7 29.90 4.293 M 4 4.28 0.318 9 4.89 0.612 NG 5 9.17 2.466 9 4.68 0.720 88 Sr BL 5 339.33 27.609 2 364.24 89.676 DG 5 18.36 2.774 9 21.87 2.702 G 4 12.28 6.872 7 12.10 4.550 M 4 6.28 1.499 9 7.44 0.61 8 NG 5 5.35 2.233 9 4.33 0.841 107 Ag BL 5 3.65 1.066 2 33.07 28.683 DG 5 0.20 0.071 9 1.19 0.453 0.0136 G 4 0.07 0.022 7 0.27 0.085 M 4 0.02 0.006 9 0.58 0.199 0.0055 NG 5 0.55 0.439 9 0.19 0.048 111 Cd BL 5 2.83 0.648 2 1.21 0.778 DG 5 8.17 2.241 9 6.43 2.442 G 4 0.91 0.230 7 3.58 0.486 0.0082 M 4 0.36 0.039 9 0.49 0.033 NG 5 0.40 0.118 9 0.18 0.046 118 Sn BL 5 BLD 2 BLD DG 5 BLD 9 BLD G 4 BLD 7 BLD
210 Table C 1. Continued. Analyte Organ n (NS) mean (NS) (ng/m g) SEM (NS) n (OS) mean (OS) (ng/mg) SEM (OS) p (K W) 118 Sn M 4 BLD 9 BLD NG 5 BLD 9 BLD 202 Hg BL 5 0.86 0.315 2 2.11 1.327 DG 5 0.06 0.027 9 0.32 0.094 0.0388 G 4 0.05 0.018 7 0.16 0.044 M 4 0.07 0.023 9 0.09 0.019 NG 5 0.41 0.293 9 0.34 0.197 238 U BL 5 1.97 0.539 2 BLD DG 5 2.08 0.692 9 0.96 0.178 G 4 0.09 0.020 7 0.06 0.032 M 4 BLD 9 BLD NG 5 0.10 0.023 9 0.12 0.046 l Wallis nonparametric test for difference of means (p<0.05) based on chi square approximation ; p value is reported. If no p value is reported, OS and NS values are not significantly different. BL blood, DG digestive gland, G gonad, M foot (muscl e), NG neural ganglia. BLD below limit of detection, indicated only if all samples within the group were below the limit of detection.
211 APPENDIX D LIST OF DIFFERENTIAL LY REGULATED GENES I N THE DIGESTIVE GLAND MICROARRAY EXPERIMENT Object D 1 is a comma delimited Exc el f ile containing a list of all differentially regulated gen es from the microarray experiment performed with digestive gland from female queen conchs collected February, 2007 (reported in Chapter 5 ). The table includes the Gene Title for each gene difference between log 2 NS mean and log 2 OS mean for each probe (Diff of Treatment = (NS) (OS )), which was used to determine the fold difference (Fold) and direction of regulation in NS conchs, relative to OS (Direction (NS)), and the p v alue determined by ANOVA with 5 % FDR (False Discovery Rate) to control for multiple comparisons. Object D 1. List of differentially regulated genes in the digestive gland microarray experiment (. c s v file 2 3 KB)
212 APPENDIX E LIST OF DIFFERENTIALLY REGULATED GENES I N THE OV ARY MICROARRAY EXPERIMENT Object E-1 is a comma delimited Excel f ile containing a list of all differentiallyregulated genes from the microarray experiment performed with ovary from female queen conchs collected February, 2007 (reported in Chapter 5). The table includes the Gene Title for each gene, difference between log 2 NS mean and log 2 OS mean for each probe (Diff of Treatment = (NS)-(OS)), which was used to determine the fold difference (Fold) and direction of regulation in NS conchs, relative to OS (Direction (NS), and the p-value determined by ANOVA with 5% FDR (False Discovery Rate) to control for multiple comparisons. Object E-1. List of differentially regulated genes in the ovary microarray experiment (.csv file 33 KB)
213 APPENDIX F ADDITIONAL ANALYS E S OF OVARY AND DIGESTIVE GLA ND MICROARRAY DATA TO DETERMINE THE APP ROPRIATENESS OF INCL UDING THE MASCULINIZED FEMALE Ovary microarray data were Loess normalized exactly as was performed in Chapter 5. However, for this analysis, they were randomly placed without regard to site of collection : in group A, individuals NM3, NF3, OF1, and OF2; in group B, individuals NF1, OF3, and OF4. A one way ANOVA analysis was performed exactly as previously 1). The ANOVA found significant differential sign al on 40 of 15,744 probes (0.25 %), compared to 1564 probes (9.93 %) with differential signal when the analysis wa s performed based on e 5 2). This is a strong argument that the data are not heavily biased by the inclusion of the one imposex female sample (i.e. the group containing the imposex female will not be different by default). Additionally, the data from the ovary and digestive g land microarray studies were subjected to unsupervised hierarchical clustering analysis, as previously described, but using data from all genes on the microarray (results of the second ANOVA analysis on gene, as reported in the Methods) In this way, the analysis would be equally influenced by both differentially regulated genes and those that were not differentially regulated. The analysis resulted in a complete separation of OS and NS individuals for both tissues, with NM3, the imposex female, clusterin NS females (Figure F 2). This indicates that NM3 is very similar to the normal NS females in terms of ovarian and digestive gland gene expression, and is not an outlier.
214 Figure F 1. Volcano plot representing ana lysis of ovary microarray data on arbitrarily determined groups, comparing log 2 (fold difference) to significance (as log 10 p value) for each probe in the ovary microarray dataset The horizontal dotted line in each volcano plot is a significance cutoff of log 10 p value = 2, or p<0.01; all probes ab ove this line differ significantly between group. Figure F 2. Dendrograms showing the results of hierarchical clustering analysis performed on all genes for the digestive gland (A) a nd ovary (B) microarray studies, indicating that the imposex female (NM3) clusters closely with the nearshore females in both tissues, and that the major factor in the analysis is location. NS nearshore, OS offshore.
215 APPENDIX G CROSS HYBRIDIZATION OF Strombus alatus RNA ONTO THE Strombus gigas MICROARRAY Some reports exist in the literature of microarray species cross hybridization hybridizing microarrays with genetic material obtained species closely related to the species used for probe design Examples in clude an analysis of the Senegalese sole, Solea senegalensis using an Atlantic f lounder, Platichthys flesus cDNA microarray  the use of a salmonid cDNA microarray designed with probes for sequences from Atla ntic salmon and rainbow trout for the analysis of Chinook salmon and rainbow smelt, as described by Gahr et al.  and the use of an oligonucleotide microarray designed from Xenopus tropicalis sequences for anal ysis of Xenopus laevis  In order to determine whether cross hybridization of Strombus alatus genetic material onto the Strombus gigas microarray was possible several S. alatus ovary and mantle RNA samples fro m the 7 day preliminary in vivo feeding study were labeled and hybridized along with S. gigas samples, following identical procedures, which are described in Chapter 2. S. alatus RNA was labeled successfully, with Cy3 specific activity > 6 pmol Cy3/ g RNA as recommended by Agilent at the time of the experiment (note: several samples were labeled with much higher specific activity, ca. 20 pmol/ g) microarray, defined as a si gnal exceeding the 95 th percentile of negative co ntrol signals (Figure G 1) Despite successful labeling of S. alatus samples, it appeared that only samples labeling with specific activity > 20 pmol/ g had a similar proportion of present calls to S. gigas microarray samples Note that some S. alatus ovary samples used for this analysis were not confirmed by histology (see Methods for further details); however, at least one of the samples with low percent present calls was confirmed by histology
216 and ICP MS as Strombus alatus gonad, while two of the high percent present call samples were S. alatus mantle. Given that mantle was not included in the tissue pool used to design the microarray, the tissue of origin of the genetic material did not likely have a ma jor influence on the percent of present calls. Rather, this was likely a function of label incorporation and binding efficiency across species. The reduced proportion of present calls in S. alatus samples appeared to affect background subtraction and the overall distribution of signals Therefore, it was like ly that misidentification of differential signal would occur based on the inconsistency of present calls for samples with different dye specific activities. As a result it was determined that micro array analysis of S. alatus gene expression using the S. gigas microarray is inappropriate. Figure G 1. Percent present calls for Strombus gigas and Strombus alatus cRNA samples hybridized to the Strombus gigas microarray. S. alatus samples (blue) show ed reduced signal even for samples above the 6 pmol Cy3/ g cRNA cutoff for hybridization. S. gigas samples (red) showed consistent percent present calls between 70 and 90 % for all samples over 6 pmol/ g.
217 LIST OF REFERENCES 1. Glazer RA, Delgado GA (2003) Towards a holistic strategy to man aging Florida's queen conch ( Strombus gigas ) population. In: Aldana Aranda D, editor. El caracol Strombus gigas : conocimiento integral para su manejo sustenable en el Caribe. Ycatan, Mxico: CYTED. Programa Iberoamerican de Ciencia y Tecnologia para Desar olllo. pp. 73 80. 2. Brownell WN, Stevely JM (1981) The biology, fisheries, and management of the queen conch, Strombus gigas Mar Fish Rev 43: 1 12. 3. Delgado GA, Bartels CT, Glazer RA, Brown Peterson NJ, McCarthy KJ (2004) Translocation as a strategy to rehabilitate the queen conch ( Strombus gigas ) population in the Florida Keys. Fish Bull 102: 278 288. 4. Davis M (2000) Queen conch ( Strombus gigas ) culture techniques for research, stock enhancement and growout markets. In: Fingerman M, Nagabhushanam R, editors. Recent Advances in Marine Biotechnology, Volume 4, Aquaculture, Part A, Seaweeds and Invertebrates. Enfield, NH: Science Publishers, Inc. pp. 27 59. 5. Berg CJ, Glazer RA (1995) Stock assessment of a large marine gastropod ( Strombus gigas ) using r andomized and stratified towed diver censusing. ICES Mar Sci Symp 199: 247 258. 6. Delgado GA, Glazer RA, Hawtof DB, Aldana Aranda D, Rodriguez Gil LA, et al. (2008) Do queen conch ( Strombus gigas ) larvae recruiting to the Florida Keys originate from upstr eam sources? Evidence from plankton and drifter studies. In: Grober Dunsmore R, Keller BD, editors. Marine Sanctuaries Conservation Series ONMS 08 07; Belize City, Belize. U.S. Department of Commerce, National Oceanic and Atmospheric Administration, Office of National Marine Sanctuaries, Silver Spring, MD. 7. Berg CJ, Olsen DA (1989) Conservation and management of queen conch ( Strombus gigas ) fisheries in the Caribbean. In: Caddy JF, editor. Marine Invertebrate Fisheries: Their Assessment and Management. Ne w York: John Wiley & Sons. 8. Glazer RA, Quintero I (1998) Observations on the sensitivity of queen conch to water quality: Implications for coastal development. In: 50th Proceedings of the Gulf and Caribbean Fisheries Institute; November, 1997; Merida, Me xico. pp. 78 93. 9. McCarthy KJ, Bartels CT, Darcy MC, Delgado GA, Glazer RA (2002) Preliminary observation of reproductive failure in nearshore queen conch ( Strombus gigas ) in the Florida Keys. In: 53rd Proceedings of the Gulf and Caribbean Fisheries Inst itute; Fort Pierce, FL. pp. 674 680. 10. Glazer RA, Berg CJ (1994) Queen conch research in Florida: An overview. In: Appeldoorn RS, Rodriguez Q B, editors. Queen Conch Biology, Fisheries, and Mariculture. Caracas, Venezuela: Fundacion Cientifica Los Roques
218 11. Appeldoorn RS (1994) Queen conch management and research: Status, needs and priorities. In: Appeldoorn RS, Rodriguez Q. B, editors. Queen Conch Biology, Fisheries, and Mariculture. Caracas, Venezuela: Fundacion Cientifica Los Roques. pp. 301 319. 12. Chakalall B, Crispoldi A, Garibaldi L, Lupin H, Mateo J, et al. (2007) World markets and industry of selected commercially exploited aquatic species: Caribbean queen conch ( Strombus gigas ). Available: http://www.fao.org/docrep/006/Y5261E/y5261e07.htm Accessed: 2010 Jan 4. 13. Robertson R (1961) The feeding of Strombus and related herbivorous marine gastropods. Not Nat Acad Sci Philadelphia 343: 1 9. 14. Stoner AW, Ray M (1993) Aggregation dynamics in juvenile queen conch ( Strombus gigas ): Population structure, mortality, growth, and migration. Mar Biol 116: 571 582. 15. Stoner AW, Hanisak MD, Smith NP, Armstrong RA (1994) Large scale distribution of queen conch nursery habitats: Implication s for stock enhancement. In: Appeldoorn RS, Rodriguez Q B, editors. Queen Conch Biology, Fisheries, and Mariculture. Caracas, Venezuela: Fundacion Cientifica Los Roques. 16. Stoner AW (1989) Density dependent growth and grazing effects of juvenile queen co nch Strombus gigas L. in a tropical seagrass meadow. J Exp Mar Biol Ecol 130: 119 133. 17. Stoner AW, Sandt VJ (1992) Population structure, seasonal movements and feeding of queen conch, Strombus gigas in deep water habitats of the Bahamas. Bull Mar Sci 5 1: 287 300. 18. Randall JE (1964) Contributions to the biology of the queen conch, Strombus gigas Bull Mar Sci Gulf Caribb 14: 246 295. 19. Stoner AW, Waite JM (1991) Trophic biology of Strombus gigas in nursery habitats: Diets and food sources in seagras s meadows. J Moll Stud 57: 451 460. 20. Lapointe BE, Barile PJ, Yentsch CS, Littler MM, Littler DS, et al. (2004) The relative importance of nutrient enrichment and herbivory on macroalgal communities near Norman's Pond Cay, Exumas Cays, Bahamas: a "natura l" enrichment experiment. J Exp Mar Biol Ecol 298: 275 301. 21. Manzello DP, Berkelmans R, Hendee JC (2007) Coral bleaching indices and thresholds for the Florida Reef Tract, Bahamas, and St. Croix, US Virgin Islands. Mar Pollut Bull 54: 1923 1931. 22. Lip p EK, Jarrell JL, Griffin DW, Lukasik J, Jacukiewicz J, et al. (2002) Preliminary evidence for human fecal contamination in corals of the Florida Keys, USA. Mar Pollut Bull 44: 666 670.
219 23. Robertson R (1959) Observations on the spawn and veligers of conch s ( Strombus ) in the Bahamas. Proc Malac Soc 33: 164 171. 24. Weil E, Laughlin R (1984) Biology, population dynamics, and reproduction of the queen conch Strombus gigas Linn in the Archapilago de los Roques National Park. J Shellfish Res 4: 45 62. 25. Dav is M, Mitchell BA, Brown JL (1984) Breeding behavior of the queen conch Strombus gigas Linne held in a natural enclosed habitat. J Shellfish Res 4: 17 21. 26. Stoner AW, Sandt VJ, Boidron Metairon IF (1992) Seasonality in reproductive activity and larval a bundance of queen conch Strombus gigas Fish Bull 90: 161 170. 27. Shawl AL, Davis M (2004) Captive breeding behavior of four strombidae conch. J Shellfish Res 23: 157 164. 28. D'Asaro CN (1965) Organogenesis, development, and metamorphosis in the queen co nch, Strombus gigas with notes on breeding habits. Bull Mar Sci 15: 359 416. 29. Davis M, Stoner AW (1994) Trophic cues induce metamorphosis of queen conch larvae ( Strombus gigas Linnaeus). J Exp Mar Biol Ecol 180: 83 102. 30. Stoner AW (1994) The role of larval supply in the population dynamics of queen conch and the need for metapopulation analysis. In: 47th Proceedings of the Gulf and Caribbean Fisheries Institute; Fort Pierce, FL. pp. 835 841. 31. Stoner AW, Mehta N, Lee TN (1997) Recruitment of Stromb us veligers to the Florida Keys reef tract: Relation to hydrographic events. J Shellfish Res 16: 1 6. 32. Hawtof DB, McCarthy KJ, Glazer RA (1998) Distribution and abundance of queen conch, Strombus gigas larvae in the Florida Current: Implications for re cruitment to the Florida Keys. In: 50th Proceedings of the Gulf and Caribbean Fisheries Institute; November, 1997; Merida, Mexico. pp. 94 103. 33. Mitton JB, Berg CJ, Orr KS (1989) Population structure, larval dispersal, and gene flow in the queen conch, S trombus gigas of the Caribbean. Biol Bull 177: 356 362. 34. Campton DE, Berg CJ, Robison LM, Glazer RA (1992) Genetic patchiness among populations of queen conch Strombus gigas in the Florida Keys and Bimini. Fish Bull 90: 250 259. 35. Appeldoorn RS (1988 ) Age determination, growth, mortality and age of first reproduction in adult queen conch, Strombus gigas L., off Puerto Rico. Fish Res 6: 363 378.
220 36. Egan BD (1985) Aspects of the reproductive biology of Strombus gigas [M.Sc. Thesis]. Vancouver, Canada: University of British Columbia. 147 p. 37. Avila Poveda OH, Baqueiro Cardenas ER (2006) Size at sexual maturity in the queen conch Strombus gigas from Colombia. Bol Invest Mar Cost 35: 223 233. 38. Avila Poveda OH, Baqueiro Cardenas ER (2009) Reproductive cycle of Strombus gigas Linnaeus 1758 (Caenogastropoda: Strombidae) from Archipelago of San Andres, Providencia and Santa Catalina, Colombia. Invert Reprod Dev 53: 1 12. 39. McManus JFA (1948) Histological and histochemical uses of periodic acid. Stain Tec hnol 23: 99 108. 40. Sandt VJ, Stoner AW (1993) Ontogenic shift in habitat by early juvenile queen conch, Strombus gigas patterns and potential mechanisms. Fish Bull 91: 516 525. 41. Hesse KO (1979) Movement and migration of the queen conch, Strombus gig as in the Turks and Caicos Islands. Bull Mar Sci 29: 303 311. 42. Glazer RA, Kidney JA (2004) Habitat associations of adult queen conch ( Strombus gigas L.) in an unfished Florida Keys back reef: Applications to essential fish habitat. Bull Mar Sci 75: 205 224. 43. Stoner AW, Lipcius RN, Marshall LS, Bardales AT (1988) Synchronous emergence and mass migration in juvenile queen conch. Mar Ecol Prog Ser 49: 51 55. 44. Stoner AW, Ray M (1996) Queen conch, Strombus gigas in fished and unfished locations of the Bahamas: Effects of a marine fishery reserve on adults, juveniles, and larval production. Fish Bull 94: 551 565. 45. Appeldoorn RS, Sanders IM (1984) Quanitification of the density growth relationship in hatchery reared juvenile conchs ( Strombus gigas Lin n and S. costatus Gmelin). J Shellfish Res 4: 63 66. 46. Stoner AW, Ray Culp M (2000) Evidence for Allee effects in an over harvested marine gastropod: Density dependent mating and egg production. Mar Ecol Prog Ser 202: 297 302. 47. Davis M, Shawl A (2005 ) Fighting conch, Strombus alatus and Strombus pugilis : new food candidates for aquaculture. In: 56th Proceedings of the Gulf and Caribbean Fisheries Institute; Fort Pierce, FL. pp. 769 772. 48. Stoner AW, Davis M (1994) Experimental outplanting of juvenil e queen conch, Strombus gigas comparison of wild and hatchery reared stocks. Fish Bull 92: 390 411.
221 49. Glazer R, Denslow N, Brown Peterson N, McClellan Green P, Barber D, et al. (2008) Anthropogenic effects on queen conch reproductive development in sou th Florida. US EPA. X7974799 03. 73 p. 50. Thomas P, Rahman MS, Khan IA, Kummer JA (2007) Widespread endocrine disruption and reproductive impairment in an estuarine fish population exposed to seasonal hypoxia. Proc R Soc B 274: 2693 2701. 51. Wu RS, Zhou BS, Randall DJ, Woo NY, Lam PK (2003) Aquatic hypoxia is an endocrine disrupter and impairs fish reproduction. Environ Sci Technol 37: 1137 1141. 52. Boyer JN, Briceo HO (2007) Annual report of the Water Quality Monitoring Project for the Water Quality Pr otection Program of the Florida Keys National Marine Sanctuary. Technical Report #T 407 of the Southeast Environmental Research Center for US EPA Agreement #X7 96410604 2. 76 p. 53. Gomot A, Pihan F (1997) Comparison of the bioaccumulation capacities of co pper and zinc in two snail subspecies ( Helix ). Ecotoxicol Environ Saf 38: 85 94. 54. Gimbert F, Vijver MG, Coeurdassier M, Scheifler R, Peijnenburg WJ, et al. (2008) How subcellular partitioning can help to understand heavy metal accumulation and eliminati on kinetics in snails. Environ Toxicol Chem 27: 1284 1292. 55. Gomot De Vaufleury A, Pihan F (2002) Methods for toxicity assessment of contaminated soil by oral or dermal uptake in land snails: Metal bioavailability and bioaccumulation. Environ Toxicol Che m 21: 820 827. 56. Romeo M, Gharbi Bouraoui S, Gnassia Barelli M, Dellali M, Aissa P (2006) Responses of Hexaplex ( Murex ) trunculus to selected pollutants. Sci Total Environ 359: 135 144. 57. de Souza Dahm KC, Ruckert C, Tonial EM, Bonan CD (2006) In vitro exposure of heavy metals on nucleotidase and cholinesterase activities from the digestive gland of Helix aspersa Comp Biochem Physiol C Toxicol Pharmacol 143: 316 320. 58. Moolman L, Van Vuren JH, Wepener V (2007) Comparative studies on the uptake and ef fects of cadmium and zinc on the cellular energy allocation of two freshwater gastropods. Ecotoxicol Environ Saf 68: 443 450. 59. Gorski J, Nugegoda D (2006) Sublethal toxicity of trace metals to larvae of the blacklip abalone, Haliotis rubra Environ Toxi col Chem 25: 1360 1367. 60. Swaileh K, Hussein R, Halaweh N (2002) Metal accumulation from contaminated food and its effect on growth of juvenile landsnails Helix engaddensis J Environ Sci Health B 37: 151 159.
222 61. Gomot De Vaufleury A (2000) Standardized growth toxicity testing (Cu, Zn, Pb, and pentachlorophenol) with Helix aspersa Ecotoxicol Environ Saf 46: 41 50. 62. Sanders IM (1984) Sublethal effects of copper on juveniles of the queen conch Strombus gigas Linne. J Shellfish Res 4: 31 35. 63. Stark J S (1998) Effects of copper on macrobenthic assemblages in soft sediments: a laboratory experimental study. Ecotoxicology 7: 161 173. 64. Stark JS (1998) Heavy metal pollution and macrobenthic assemblages in soft sediments in two Sydney estuaries, Australia Mar Freshwater Res 49: 533 540. 65. Rygg B (1985) Effect of Sediment Copper on Benthic Fauna. Mar Ecol Prog Ser 25: 83 89. 66. Ravera O (1991) Influence of heavy metals on the reproduction and embryonic development of freshwater pulmonates (Gastropoda, M ollusca) and cladocerans (Crustacea, Arthropoda). Comp Biochem Physiol C Pharmacol Toxicol Endocrinol 100: 215 219. 67. Laskowski R, Hopkin SP (1996) Effect of Zn, Cu, Pb, and Cd on fitness in snails ( Helix aspersa ). Ecotoxicol Environ Saf 34: 59 69. 68. M unzinger A, Guarducci ML (1988) The effect of low zinc concentrations on some demographic parameters of Biomphalaria glabrata (Say), Mollusca, Gastropoda. Aquat Toxicol 12: 51 61. 69. Ducrot V, Pery AR, Mons R, Queau H, Charles S, et al. (2007) Dynamic ene rgy budget as a basis to model population level effects of zinc spiked sediments in the gastropod Valvata piscinalis Environ Toxicol Chem 26: 1774 1783. 70. Coeurdassier M, de Vaufleury A, Crini N, Scheifler R, Badot PM (2005) Assessment of whole effluent toxicity on aquatic snails: bioaccumulation of Cr, Zn, and Fe, and individual effects in bioassays. Environ Toxicol Chem 24: 198 204. 71. Snyman RG, Reinecke AJ, Reinecke SA (2004) Changes in oocyte numbers in the ovotestis of Helix aspersa after experim ental exposure to the fungicide copper oxychloride. Bull Environ Contam Toxicol 73: 398 403. 72. Dorgelo J, Meester H, van Velzen C (1995) Effects of diet and heavy metals on growth rate and fertility in the deposit feeding snail Potamopyrgus jenkinsi (Smi th) (Gastropoda: Hydrobiidae). Hydrobiologia 316: 199 210. 73. Real M, Muoz I, Guasch H, Navarro E, Sabater S (2003) The effect of copper exposure on a simple aquatic food chain. Aquat Toxicol 63: 283 291.
223 74. Rogevich EC, Hoang TC, Rand GM (2009) Effects of sublethal chronic copper exposure on the growth and reproductive success of the Florida apple snail ( Pomacea paludosa ). Arch Environ Contam Toxicol 56: 450 458. 75. Pea SC, Pocsidio GN (2007) Influence of copper on the feeding rate, growth and reprodu ction of the golden apple snail, Pomacea canaliculata Lamarck. Bull Environ Contam Toxicol 79: 606 608. 76. Croteau MN, Luoma SN (2009) Predicting dietborne metal toxicity from metal influxes. Environ Sci Technol 43: 4915 4921. 77. Wang WX, Ke C (2002) Dom inance of dietary intake of cadmium and zinc by two marine predatory gastropods. Aquat Toxicol 56: 153 165. 78. Gay D, Maher W (2003) Natural variation of copper, zinc, cadmium and selenium concentrations in Bembicium nanum and their potential use as a bio monitor of trace metals. Water Res 37: 2173 2185. 79. Taylor A, Maher W (2006) The use of two marine gastropods, Austrocochlea constricta and Bembicium auratum as biomonitors of zinc, cadmium, and copper exposure: Effect of tissue distribution, gender, re productive state, and temporal variation. J Coastal Res 22: 298 306. 80. Nicolaidou A, Nott JA (1990) Mediterranean pollution from a ferro nickel smelter: Differential uptake of metals by some gastropods. Mar Pollut Bull 21: 137 143. 81. Ying WM, Ahsanulla h M, Batley GE (1993) Accumulation and regulation of heavy metals by the intertidal snail Polinices sordidus Mar Biol 116: 417 422. 82. Tsai J W, Chou Y H, Chen B C, Liang H M, Liao C M (2004) Growth toxicity bioassays of abalone Haliotis diversicolor sup ertexta exposed to waterborne zinc. Bull Environ Contam Toxicol 72: 70 77. 83. Liao CM, Lin MC (2001) Toxicokinetics and acute toxicity of waterborne zinc in abalone ( Haliotis diversicolor supertexta Lischke). Bull Environ Contam Toxicol 66: 597 602. 84. D iSilvestro RA, Cousins RJ (1983) Physiological ligands for copper and zinc. Annu Rev Nutr 3: 261 288. 85. Berg JM, Shi Y (1996) The galvanization of biology: A growing appreciation for the roles of zinc. Science 271: 1081 1085. 86. Bettger WJ, O'Dell BL (1 981) A critical physiological role of zinc in the structure and function of biomembranes. Life Sci 28: 1425 1438. 87. Bray TM, Bettger WJ (1990) The physiological role of zinc as an antioxidant. Free Radic Biol Med 8: 281 291.
224 88. Truong Tran AQ, Carter J, Ruffin RE, Zalewski PD (2001) The role of zinc in caspase activation and apoptotic cell death. Biometals 14: 315 330. 89. Danks DM (1988) Copper deficiency in humans. Annu Rev Nutr 8: 235 257. 90. Rutherford JC, Bird AJ (2004) Metal responsive transcripti on factors that regulate iron, zinc, and copper homeostasis in eukaryotic cells. Eukaryot Cell 3: 1 13. 91. Little C (1967) Ionic regulation in the queen conch, Strombus gigas (Gastropoda, Prosobranchia). J Exp Biol 46: 459 474. 92. van Holde KE, Miller KI Decker H (2001) Hemocyanins and invertebrate evolution. J Biol Chem 276: 15563 15566. 93. Van Holde KE, Miller KI (1995) Hemocyanins. In: Richards FM, Eisenberg DS, Edsall JT, editors. Advances in Protein Chemistry, vol 47. San Diego, CA: Academic Press, Inc. pp. 1 81. 94. Desouky MM, McCrohan CR, Jugdaohsingh R, Powell JJ, White KN (2003) Effect of orthosilicic acid on the accumulation of trace metals by the pond snail Lymnaea stagnalis. Aquat Toxicol 64: 63 71. 95. Desouky MM (2006) Tissue distribution and subcellular localization of trace metals in the pond snail Lymnaea stagnalis with special reference to the role of lysosomal granules in metal sequestration. Aquat Toxicol 77: 143 152. 96. Kruczynski WL (1999) Water quality concerns in the Florida Keys : Sources, effects, and solutions. Key West, FL: Florida Keys National Marine Sanctuary Water Quality Protection Program. 75 p. 97. Manker JP (1975) Distribution and concentration of mercury, lead, cobalt, zinc, and chromium in suspended particulates and b ottom sediments upper Florida Keys, Florida Bay, and Biscayne Bay. [Ph.D. Dissertation]. Houston, TX: Rice University. 126 p. 98. Carnahan EA, Hoare AA, Hallock P, Lidz BH, Reich CD (2008) Distribution of heavy metals and foraminiferal assemblages in sed iments of Biscayne Bay, Florida, USA. J Coastal Res 24: 159 169. 99. Lewis MA, Dantin DD, Chancy CA, Abel KC, Lewis CG (2007) Florida seagrass habitat evaluation: a comparative survey for chemical quality. Environ Pollut 146: 206 218. 100. Caccia VG, Mille ro FJ, Palanques A (2003) The distribution of trace metals in Florida Bay sediments. Mar Pollut Bull 46: 1420 1433.
225 101. Rogevich EC, Hoang TC, Rand GM (2008) The effects of water quality and age on the acute toxicity of copper to the Florida apple snail, Pomacea paludosa Arch Environ Contam Toxicol 54: 690 696. 102. Frakes RA, Bargar TA, Bauer EA (2008) Sediment copper bioavailability to freshwater snails in south Florida: Risk implications for the Everglade snail kite ( Rostrhamus sociabilis plumbeus ). Ec otoxicology 17: 598 604. 103. Latiolais JM, Taylor MS, Roy K, Hellberg ME (2006) A molecular phylogenetic analysis of strombid gastropod morphological diversity. Mol Phylogenet Evol 41: 436 444. 104. Zamora Bustillos R, Rodriguez Canul R, De Leon FJG (2007 ) Isolation and characterization of eight polymorphic microsatellite markers from pink conch ( Strombus gigas ). Mol Ecol Notes 7: 597 599. 105. Spade DJ, Griffitt RJ, Liu L, Brown Peterson NJ, Kroll KJ, et al. (2010) Queen conch ( Strombus gigas ) testis regr esses during the reproductive season at nearshore sites in the Florida Keys. PLoS One 5: e12737. 106. Diaz Rizo O, Olivares Reumont S, Viguri Fuente J, Diaz Arado O, Lopez Pino N, et al. (2010) Copper, zinc and lead bioaccumulation in marine snail, Strombu s gigas from Guacanayabo Gulf, Cuba. Bull Environ Contam Toxicol 85: 330 333. 107. O'Connor TP, Lauenstein GG (2005) Status and trends of copper concentrations in mussels and oysters in the USA. Mar Chem 97: 49 59. 108. Spade DJ, Knoebl I, Denslow ND (201 1) Cesium chloride gradient centrifugation improves the quality of total RNA preparations from the gastropod Strombus gigas and the coral Montastraea faveolata J Exp Mar Biol Ecol 402: 43 48. 109. Little C (1965) Notes on the anatomy of the queen conch, S trombus gigas Bull Mar Sci 15: 338 358. 110. Shawl AL, Jenkins D, Davis M, Main K (2005) Growth of Florida fighting conch, Strombus alatus in recirculating systems. In: 56th Proceedings of the Gulf and Caribbean Fisheries Institute; Fort Pierce, FL. pp. 773 780. 111. NOAA (2010) Coastal Water Temperature Guide, Gulf of Mexico Coast: Eastern. Available: http://www.nodc.noaa.gov/dsdt/cwtg/egof.html Accessed: 2010 October 25. 112. Thurston RV, Ru sso RC, Emerson K (1979) Aqueous ammonia equilibrium tabulation of percent un ionized ammonia. Duluth, MN: US EPA. EPA 600/3 79 091. 177 p. 113. Marvin KT, Proctor RR (1965) Stabilizing the ammonia nitrogen content of estuarine and coastal waters by freezi ng. Limnol Oceanogr 10: 288 289.
226 114. Degobbis D (1973) On the storage of seawater samples for ammonia determination. Limnol Oceanogr 18: 146 150. 115. Hamersley MR, Lavik G, Woebken D, Rattray JE, Lam P, et al. (2007) Anaerobic ammonium oxidation in the P eruvian oxygen minimum zone. Limnol Oceanogr 52: 923 933. 116. Brown Peterson NJ, Wyanski DM, Saborido Rey F, Macewicz BJ, Lowerre Barbieri SK (2011) A Standardized Terminology for Describing Reproductive Development in Fishes. Mar Coast Fish 3: 52 70. 117 Kingston RE, Chomczynski P, Sacchi N (2001) Guanidine methods for total RNA preparation. Curr Protoc Mol Biol 36: 4.2.1 4.2.9. 118. Schroeder A, Mueller O, Stocker S, Salowsky R, Leiber M, et al. (2006) The RIN: an RNA integrity number for assigning inte grity values to RNA measurements. BMC Mol Biol 7: 3. 119. Ishikawa H (1977) Evolution of ribosomal RNA. Comp Biochem Physiol B 58: 1 7. 120. Groppe JC, Morse DE (1993) Isolation of full length RNA templates for reverse transcription from tissues rich in RN ase and proteoglycans. Anal Biochem 210: 337 343. 121. Brazma A, Hingamp P, Quackenbush J, Sherlock G, Spellman P, et al. (2001) Minimum information about a microarray experiment (MIAME) toward standards for microarray data. Nat Genet 29: 365 371. 122. Roz en S, Skaletsky H (2000) Primer3 on the WWW for general users and for biologist programmers. In: Misener S, Krawetz SA, editors. Methods in Molecular Biology. Totowa, NJ: Humana Press, Inc. pp. 365 386. 123. Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Anal Biochem 72: 248 254. 124. Eisen MB, Spellman PT, Brown PO, Botstein D (1998) Cluster analysis and display of genome wide expression patterns. Proc Natl Acad Sci U S A 95: 14863 14868. 125. Saldanha AJ (2004) Java Treeview -extensible visualization of microarray data. Bioinformatics 20: 3246 3248. 126. Al Shahrour F, Minguez P, Vaquerizas JM, Conde L, Dopazo J (2005) BABELOMICS: A suite of web tools f or functional annotation and analysis of groups of genes in high throughput experiments. Nucleic Acids Res 33: W460 464.
227 127. Smale G, Sasse J (1992) RNA isolation from cartilage using density gradient centrifugation in cesium trifluoroacetate: an RNA prep aration technique effective in the presence of high proteoglycan content. Anal Biochem 203: 352 356. 128. Matthee GF, Konig GM, Wright AD (1998) Three new diterpenes from the marine soft coral Lobophytum crassum J Nat Prod 61: 237 240. 129. Rashid MA, Gus tafson KR, Boyd MR (2000) HIV inhibitory cembrane derivatives from a Philippines collection of the soft coral Lobophytum species. J Nat Prod 63: 531 533. 130. Hanington PC, Lun CM, Adema CM, Loker ES (2010) Time series analysis of the transcriptional respo nses of Biomphalaria glabrata throughout the course of intramolluscan development of Schistosoma mansoni and Echinostoma paraensei Int J Parasitol 40: 819 831. 131. Williams EA, Degnan BM, Gunter H, Jackson DJ, Woodcroft BJ, et al. (2009) Widespread trans criptional changes pre empt the critical pelagic benthic transition in the vetigastropod Haliotis asinina Mol Ecol 18: 1006 1025. 132. Lee YS, Choi SL, Kim TH, Lee JA, Kim HK, et al. (2008) Transcriptome analysis and identification of regulators for long term plasticity in Aplysia kurodai Proc Natl Acad Sci U S A 105: 18602 18607. 133. Azami S, Wagatsuma A, Sadamoto H, Hatakeyama D, Usami T, et al. (2006) Altered gene activity correlated with long term memory formation of conditioned taste aversion in Lym naea J Neurosci Res 84: 1610 1620. 134. Schwarz JA, Brokstein PB, Voolstra C, Terry AY, Manohar CF, et al. (2008) Coral life history and symbiosis: Functional genomic resources for two reef building Caribbean corals, Acropora palmata and Montastraea faveo lata BMC Genomics 9: 97. 135. Chomczynski P, Sacchi N (1987) Single step method of RNA isolation by acid guanidinium thiocyanate phenol chloroform extraction. Anal Biochem 162: 156 159. 136. Gallagher SR, Desjardins PR (2006) Quantitation of DNA and RNA w ith absorption and fluorescence spectroscopy. Curr Protoc Mol Biol 76: A.3D.1 A.3D.21. 137. Glazer AN, Smith EL (1961) Studies on the ultraviolet difference spectra of proteins and polypeptides. J Biol Chem 236: 2942 2947. 138. Mant CT, Hodges RS (1991) Hi gh Performance Liquid Chromatography of Peptides and Proteins: Separation, Analysis, and Conformation. Boca Raton, FL: CRC Press. 938 p.
228 139. Whitaker JR, Granum PE (1980) An absolute method for protein determination based on difference in absorbance at 23 5 and 280 nm. Anal Biochem 109: 156 159. 140. Wilfinger WW, Mackey K, Chomczynski P (1997) Effect of pH and ionic strength on the spectrophotometric assessment of nucleic acid purity. Biotechniques 22: 474 481. 141. Doney SC (2010) The growing human footpr int on coastal and open ocean biogeochemistry. Science 328: 1512 1516. 142. Bustin SA (2000) Absolute quantification of mRNA using real time reverse transcription polymerase chain reaction assays. J Mol Endocrinol 25: 169 193. 143. Bustin SA (2002) Quantif ication of mRNA using real time reverse transcription PCR (RT PCR): trends and problems. J Mol Endocrinol 29: 23 39. 144. Pena FJ, Rodriguez Martinez H, Tapia JA, Ortega Ferrusola C, Gonzalez Fernandez L, et al. (2009) Mitochondria in mammalian sperm physi ology and pathology: a review. Reprod Domest Anim 44: 345 349. 145. Erkan M, Sousa M (2002) Fine structural study of the spermatogenic cycle in Pitar rudis and Chamelea gallina (Mollusca, Bivalvia, Veneridae). Tissue Cell 34: 262 272. 146. Yaffe MP (1997) Mitochondrial morphogenesis: fusion factor for fly fertility. Curr Biol 7: R782 783. 147. Ternes P, Franke S, Zahringer U, Sperling P, Heinz E (2002) Identification and characterization of a sphingolipid delta 4 desaturase family. J Biol Chem 277: 25512 25 518. 148. Reeve S, Carhan A, Dee CT, Moffat KG (2007) Slowmo is required for Drosophila germline proliferation. Genesis 45: 66 75. 149. Khor B, Bredemeyer AL, Huang CY, Turnbull IR, Evans R, et al. (2006) Proteasome activator PA200 is required for normal s permatogenesis. Mol Cell Biol 26: 2999 3007. 150. Guan J, Kinoshita M, Yuan L (2009) Spatiotemporal association of DNAJB13 with the annulus during mouse sperm flagellum development. BMC Dev Biol 9: 23. 151. Guan J, Yuan L (2008) A heat shock protein 40, DN AJB13, is an axoneme associated component in mouse spermatozoa. Mol Reprod Dev 75: 1379 1386. 152. Yang HM, Liu G, Nie ZY, Nie DS, Deng Y, et al. (2005) Molecular cloning of a novel rat gene Tsarg1, a member of the DnaJ/HSP40 protein family. DNA Seq 16: 16 6 172.
229 153. Li L, Liu G, Fu JJ, Li LY, Tan XJ, et al. (2009) Molecular cloning and characterization of a novel transcript variant of Mtsarg1 gene. Mol Biol Rep 36: 1023 1032. 154. Alekseev OM, Widgren EE, Richardson RT, O'Rand MG (2005) Association of NASP with HSP90 in mouse spermatogenic cells: stimulation of ATPase activity and transport of linker histones into nuclei. J Biol Chem 280: 2904 2911. 155. Karnoub AE, Weinberg RA (2008) Ras oncogenes: split personalities. Nat Rev Mol Cell Biol 9: 517 531. 156 Nishida H (2002) Patterning the marginal zone of early ascidian embryos: localized maternal mRNA and inductive interactions. Bioessays 24: 613 624. 157. O'Keefe DD, Gonzalez Nino E, Burnett M, Dylla L, Lambeth SM, et al. (2009) Rap1 maintains adhesion be tween cells to affect Egfr signaling and planar cell polarity in Drosophila Dev Biol 333: 143 160. 158. Lee J, Kanatsu Shinohara M, Morimoto H, Kazuki Y, Takashima S, et al. (2009) Genetic reconstruction of mouse spermatogonial stem cell self renewal in v itro by Ras cyclin D2 activation. Cell Stem Cell 5: 76 86. 159. Singh SR, Hou SX (2008) Immunohistological techniques for studying the Drosophila male germline stem cell. Methods Mol Biol 450: 45 59. 160. Buchardt B, Fritz P (1978) Strontium uptake in shel l aragonite from the freshwater gastropod Limnaea stagnalis Science 199: 291 292. 161. Gros O, Frenkiel L, Aranda DA (2009) Structural analysis of the digestive gland of the queen conch Strombus gigas Linnaeus, 1758 and its intracellular parasites. J Moll usc Stud 75: 59 68. 162. Nott JA, Nicolaidou A (1990) Transfer of metal detoxification along marine food chains. J Mar Biol Assoc UK 70: 905 912. 163. Yamaguchi S, Miura C, Kikuchi K, Celino FT, Agusa T, et al. (2009) Zinc is an essential trace element for spermatogenesis. Proc Natl Acad Sci U S A 106: 10859 10864. 164. Ruppert EE, Fox RS, Barnes RD (2004) Invertebrate Ecology: A Functional Evolutionary Approach. Belmont, CA: Brooks/Cole. 963 p. 165. Iguchi T, Watanabe H, Katsu Y (2007) Toxicogenomics and e cotoxicogenomics for studying endocrine disruption and basic biology. Gen Comp Endocrinol 153: 25 29.
230 166. Iguchi T, Watanabe H, Katsu Y (2006) Application of ecotoxicogenomics for studying endocrine disruption in vertebrates and invertebrates. Environ Hea lth Perspect 114 Suppl 1: 101 105. 167. Guiller A, Bellido A, Coutelle A, Madec L (2006) Spatial genetic pattern in the land mollusc Helix aspersa inferred from a 'centre based clustering' procedure. Genet Res 88: 27 44. 168. van Rensburg MJ, Coyne VE (200 9) The role of electron transport in the defence response of the South African abalone, Haliotis midae Fish Shellfish Immunol 26: 171 176. 169. Adema CM, Hanington PC, Lun CM, Rosenberg GH, Aragon AD, et al. (2009) Differential transcriptomic responses of Biomphalaria glabrata (Gastropoda, Mollusca) to bacteria and metazoan parasites, Schistosoma mansoni and Echinostoma paraensei (Digenea, Platyhelminthes). Mol Immunol 47: 849 860. 170. Lockyer AE, Spinks J, Kane RA, Hoffmann KF, Fitzpatrick JM, et al. (20 08) Biomphalaria glabrata transcriptome: cDNA microarray profiling identifies resistant and susceptible specific gene expression in haemocytes from snail strains exposed to Schistosoma mansoni BMC Genomics 9: 634. 171. Gobert GN, Moertel L, Brindley PJ, McManus DP (2009) Developmental gene expression profiles of the human pathogen Schistosoma japonicum BMC Genomics 10: 128. 172. Waisberg M, Lobo FP, Cerqueira GC, Passos LK, Carvalho OS, et al. (2008) Schistosoma mansoni : Microarray analysis of gene expre ssion induced by host sex. Exp Parasitol 120: 357 363. 173. Gobert GN, McInnes R, Moertel L, Nelson C, Jones MK, et al. (2006) Transcriptomics tool for the human Schistosoma blood flukes using microarray gene expression profiling. Exp Parasitol 114: 160 17 2. 174. Fitzpatrick JM, Protasio AV, McArdle AJ, Williams GA, Johnston DA, et al. (2008) Use of genomic DNA as an indirect reference for identifying gender associated transcripts in morphologically identical, but chromosomally distinct, Schistosoma mansoni cercariae. PLoS Negl Trop Dis 2: e323. 175. Dillon GP, Feltwell T, Skelton J, Coulson PS, Wilson RA, et al. (2008) Altered patterns of gene expression underlying the enhanced immunogenicity of radiation attenuated schistosomes. PLoS Negl Trop Dis 2: e240. 176. Miller DH, Jensen KM, Villeneuve DL, Kahl MD, Makynen EA, et al. (2007) Linkage of biochemical responses to population level effects: a case study with vitellogenin in the fathead minnow ( Pimephales promelas ). Environ Toxicol Chem 26: 521 527.
231 177. S alice CJ, Miller TJ (2003) Population level responses to long term cadmium exposure in two strains of the freshwater gastropod Biomphalaria glabrata : results from a life table response experiment. Environ Toxicol Chem 22: 678 688. 178. Kramer VJ, Etterson MA, Hecker M, Murphy CA, Roesijadi G, et al. (2011) Adverse outcome pathways and ecological risk assessment: Bridging to population level effects. Environ Toxicol Chem 30: 64 76. 179. Jordaens K, De Wolf H, Vandecasteele B, Blust R, Backeljau T (2006) Asso ciations between shell strength, shell morphology and heavy metals in the land snail Cepaea nemoralis (Gastropoda, Helicidae). Sci Total Environ 363: 285 293. 180. Yamaguchi S, Miura C, Kikuchi K, Celino FT, Agusa T, et al. (2009) Zinc is an essential trac e element for spermatogenesis. Proc Natl Acad Sci U S A 106: 10859 10864. 181. Croxford TP, McCormick NH, Kelleher SL (2011) Moderate zinc deficiency reduces testicular Zip6 and Zip10 abundance and impairs spermatogenesis in mice. J Nutr 141: 359 365. 182. Falchuk KH, Montorzi M (2001) Zinc physiology and biochemistry in oocytes and embryos. Biometals 14: 385 395. 183. Bernhardt ML, Kim AM, O'Halloran TV, Woodruff TK (2011) Zinc requirement during meiosis I meiosis II transition in mouse oocytes is independ ent of the MOS MAPK pathway. Biol Reprod 84: 526 536. 184. Gibbs PE, Bryan GW (1986) Reproductive failure in populations of the dog whelk, Nucella lapillus caused by imposex induced by tributyltin from antifouling paints. J Mar Biol Assoc UK 66: 767 777. 185. Smith B (1980) The estuarine mud snail, Nassarius obsoletus : Abnormalities in the reproductive system. J Moll Stud 46: 247 256. 186. Blaber SJM (1970) The occurrence of a penis like out growth behind the right tentacle in spent females of Nucella lapi llus (L.). Proc Malac Soc Lond 39: 231 233. 187. Titley O'Neal CP, MacDonald BA, Pelletier E, Saint Louis R, Phillip OS (2011) The relationship between imposex and tributyltin (TBT) concentration in Strombus gigas from the British Virgin Islands. Bull Mar Sci: DOI: 10.5343/bms.2010.1093. 188. Reed SE (1994) Masculinized females in the genus Strombus : aspects of their biology and possible advantages for mariculture of conch. In: Appeldoorn RS, Rodriguez Q. B, editors. Queen Conch Biology, Fisheries, and Mari culture. Caracas, Venezuela: Fundacion Cientifica Los Roques. pp. 213 221.
232 189. Reed SE (1993) Gonadal Comparison of Masculinized Females and Androgynous Males to Normal Males and Females in Strombus (Mesogastropoda, Strombidae). J Shellfish Res 12: 71 75. 190. Reed SE (1995) Sexual Trimorphism in Strombus luhuanus Linne, 1758 (Mollusca, Gastropoda) at Shirahama, Japan. J Shellfish Res 14: 159 160. 191. Kuwamura T, Fukao R, Nishida M, Wada K, Yanagisawa Y (1983) Reproductive biology of the gastropod Strombu s luhuanus (Strombidae). Publ Seto Mar Biol Lab 5/6: 433 443. 192. Cob ZC, Arshad A, Bujang JS, Abd Ghaffar M (2010) Description and evaluation of imposex in Strombus canarium Linnaeus, 1758 (Gastropoda, Strombidae): A potential bio indicator of tributylti n pollution. Environ Monit Assess: DOI: 10.1007/s10661 10010 11698 10667. 193. Castro LF, Lima D, Machado A, Melo C, Hiromori Y, et al. (2007) Imposex induction is mediated through the Retinoid X Receptor signalling pathway in the neogastropod Nucella lapi llus Aquat Toxicol 85: 57 66. 194. Horiguchi T, Ohta Y, Nishikawa T, Shiraishi F, Shiraishi H, et al. (2008) Exposure to 9 cis retinoic acid induces penis and vas deferens development in the female rock shell, Thais clavigera Cell Biol Toxicol 24: 553 56 2. 195. Chen D, Pan X, Xiao P, Farwell MA, Zhang B (2011) Evaluation and identification of reliable reference genes for pharmogenomics, toxicogenomics and small RNA expression analysis. J Cell Physiol: DOI: 10.1002/jcp.22725. 196. Obrero A, Die JV, Roman B Gomez P, Nadal S, et al. (2011) Selection of reference genes for gene expression studies in zucchini ( Cucurbita pepo ) using qPCR. J Agric Food Chem 59: 5402 5411. 197. Ponton F, Chapuis MP, Pernice M, Sword GA, Simpson SJ (2011) Evaluation of potential r eference genes for reverse transcription qPCR studies of physiological responses in Drosophila melanogaster J Insect Physiol 57: 840 850. 198. Rho HW, Lee BC, Choi ES, Choi IJ, Lee YS, et al. (2010) Identification of valid reference genes for gene express ion studies of human stomach cancer by reverse transcription qPCR. BMC Cancer 10: 240. 199. Tricarico C, Pinzani P, Bianchi S, Paglierani M, Distante V, et al. (2002) Quantitative real time reverse transcription polymerase chain reaction: Normalization to rRNA or single housekeeping genes is inappropriate for human tissue biopsies. Anal Biochem 309: 293 300. 200. Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, et al. (2009) The MIQE guidelines: Minimum information for publication of quantitative real time PCR experiments. Clin Chem 55: 611 622.
233 201. Connon RE, Geist J, Pfeiff J, Loguinov AV, D'Abronzo LS, et al. (2009) Linking mechanistic and behavioral responses to sublethal esfenvalerate exposure in the endangered delta smelt; Hypomesus transpacific us (Fam. Osmeridae). BMC Genomics 10: 608. 202. Ho Pun Cheung A, Bascoul Mollevi C, Assenat E, Boissiere Michot F, Bibeau F, et al. (2009) Reverse transcription quantitative polymerase chain reaction: Description of a RIN based algorithm for accurate data normalization. BMC Mol Biol 10: 31. 203. Skillman AD, Nagler JJ, Hook SE, Small JA, Schultz IR (2006) Dynamics of 17alpha ethynylestradiol exposure in rainbow trout ( Oncorhynchus mykiss ): Absorption, tissue distribution, and hepatic gene expression pattern Environ Toxicol Chem 25: 2997 3005. 204. Matsumoto T, Yamano K, Kitamura M, Hara A (2008) Ovarian follicle cells are the site of vitellogenin synthesis in the Pacific abalone Haliotis discus hannai Comp Biochem Physiol A Mol Integr Physiol 149: 293 298. 205. Gagnaire B, Gagne F, Andre C, Blaise C, Abbaci K, et al. (2009) Development of biomarkers of stress related to endocrine disruption in gastropods: Alkali labile phosphates, protein bound lipids and vitellogenin like proteins. Aquat Toxicol 92: 155 16 7. 206. Ankley GT, Bencic DC, Breen MS, Collette TW, Conolly RB, et al. (2009) Endocrine disrupting chemicals in fish: Developing exposure indicators and predictive models of effects based on mechanism of action. Aquat Toxicol 92: 168 178. 207. Ankley GT, Kahl MD, Jensen KM, Hornung MW, Korte JJ, et al. (2002) Evaluation of the aromatase inhibitor fadrozole in a short term reproduction assay with the fathead minnow ( Pimephales promelas ). Toxicol Sci 67: 121 130. 208. Matozzo V, Gagne F, Marin MG, Ricciardi F, Blaise C (2008) Vitellogenin as a biomarker of exposure to estrogenic compounds in aquatic invertebrates: A review. Environ Int 34: 531 545. 209. Puinean AM, Rotchell JM (2006) Vitellogenin gene expression as a biomarker of endocrine disruption in the i nvertebrate, Mytilus edulis Mar Environ Res 62 Suppl: S211 S214. 210. Gros O, Frenkiel L, Aranda DA (2009) Structural analysis of the digestive gland of the queen conch Strombus gigas Linnaeus, 1758 and its intracellular parasites. J Moll Stud 75: 59 68.
234 211. Pytharopoulou S, Kouvela EC, Sazakli E, Leotsinidis M, Kalpaxis DL (2006) Evaluation of the global protein synthesis in Mytilus galloprovincialis in marine pollution monitoring: Seasonal variability and correlations with other biomarkers. Aquat Toxic ol 80: 33 41. 212. Joo YJ, Kim JH, Kang UB, Yu MH, Kim J (2011) Gcn4p mediated transcriptional repression of ribosomal protein genes under amino acid starvation. EMBO J 30: 859 872. 213. Jacques C, Fontaine JF, Franc B, Mirebeau Prunier D, Triau S, et al. (2009) Death associated protein 3 is overexpressed in human thyroid oncocytic tumours. Br J Cancer 101: 132 138. 214. Sun Z, Cheng Z, Taylor CA, McConkey BJ, Thompson JE (2010) Apoptosis induction by eIF5A1 involves activation of the intrinsic mitochondria l pathway. J Cell Physiol 223: 798 809. 215. Habibi HR, Andreu Vieyra CV (2007) Hormonal regulation of follicular atresia in teleost fish. In: Babin PJ, Cerda J, Lubzens E, editors. The Fish Oocyte: From Basic Studies to Biotechnological Applications. New York: Springer Verlag. pp. 235 253. 216. Lepine S, Allegood JC, Park M, Dent P, Milstien S, et al. (2011) Sphingosine 1 phosphate phosphohydrolase 1 regulates ER stress induced autophagy. Cell Death Differ 18: 350 361. 217. Le Stunff H, Galve Roperh I, Pet erson C, Milstien S, Spiegel S (2002) Sphingosine 1 phosphate phosphohydrolase in regulation of sphingolipid metabolism and apoptosis. J Cell Biol 158: 1039 1049. 218. Johnson KR, Johnson KY, Becker KP, Bielawski J, Mao C, et al. (2003) Role of human sphin gosine 1 phosphate phosphatase 1 in the regulation of intra and extracellular sphingosine 1 phosphate levels and cell viability. J Biol Chem 278: 34541 34547. 219. Ricci JE, Munoz Pinedo C, Fitzgerald P, Bailly Maitre B, Perkins GA, et al. (2004) Disrupti on of mitochondrial function during apoptosis is mediated by caspase cleavage of the p75 subunit of complex I of the electron transport chain. Cell 117: 773 786. 220. Li P, Jayarama S, Ganesh L, Mordi D, Carr R, et al. (2010) Akt phosphorylated mitogen act ivated kinase activating death domain protein (MADD) inhibits TRAIL induced apoptosis by blocking Fas associated death domain (FADD) association with death receptor 4. J Biol Chem 285: 22713 22722.
235 221. Kurada BR, Li LC, Mulherkar N, Subramanian M, Prasad KV, et al. (2009) MADD, a splice variant of IG20, is indispensable for MAPK activation and protection against apoptosis upon tumor necrosis factor alpha treatment. J Biol Chem 284: 13533 13541. 222. Rachner TD, Benad P, Rauner M, Goettsch C, Singh SK, et al. (2009) Osteoprotegerin production by breast cancer cells is suppressed by dexamethasone and confers resistance against TRAIL induced apoptosis. J Cell Biochem 108: 106 116. 223. Irmler M, Steiner V, Ruegg C, Wajant H, Tschopp J (2000) Caspase induced i nactivation of the anti apoptotic TRAF1 during Fas ligand mediated apoptosis. FEBS Lett 468: 129 133. 224. Vande Walle L, Wirawan E, Lamkanfi M, Festjens N, Verspurten J, et al. (2010) The mitochondrial serine protease HtrA2/Omi cleaves RIP1 during apoptos is of Ba/F3 cells induced by growth factor withdrawal. Cell Res 20: 421 433. 225. Zhang J, Wang X, Bove KE, Xu M (1999) DNA fragmentation factor 45 deficient cells are more resistant to apoptosis and exhibit different dying morphology than wild type contro l cells. J Biol Chem 274: 37450 37454. 226. Boulares AH, Zoltoski AJ, Yakovlev A, Xu M, Smulson ME (2001) Roles of DNA fragmentation factor and poly(ADP ribose) polymerase in an amplification phase of tumor necrosis factor induced apoptosis. J Biol Chem 27 6: 38185 38192. 227. Baron BW, Hyjek E, Gladstone B, Thirman MJ, Baron JM (2010) PDCD2, a protein whose expression is repressed by BCL6, induces apoptosis in human cells by activation of the caspase cascade. Blood Cells Mol Dis 45: 169 175. 228. Li AL, Li HY, Jin BF, Ye QN, Zhou T, et al. (2004) A novel eIF5A complex functions as a regulator of p53 and p53 dependent apoptosis. J Biol Chem 279: 49251 49258. 229. Lee NP, Tsang FH, Shek FH, Mao M, Dai H, et al. (2010) Prognostic significance and therapeutic po tential of eukaryotic translation initiation factor 5A (eIF5A) in hepatocellular carcinoma. Int J Cancer 127: 968 976. 230. Babin PJ (1987) Apolipoproteins and the association of egg yolk proteins with plasma high density lipoproteins after ovulation and f ollicular atresia in the rainbow trout ( Salmo gairdneri ). J Biol Chem 262: 4290 4296. 231. Kaptaner B, Unal G (2010) Effects of 17alpha ethynylestradiol and nonylphenol on liver and gonadal apoptosis and histopathology in Chalcalburnus tarichi Environ Tox icol: DOI: 10.1002/tox.20585. 232. Reed SE (1995) Reproductive anatomy and biology of the genus Strombus in the Caribbean: II. Females. J Shellfish Res 14: 331 336.
236 233. Reed SE (1995) Reproductive anatomy and biology of the genus Strombus in the Caribbean : I. Males. J Shellfish Res 14: 325 330. 234. Percharde PL (1968) Notes on distribution and underwater observations on the molluscan genus Strombus as found in the waters of Trinidad and Tobago. Carib J Sci 8: 47 56. 235. Reed SE (1995) Reproductive season ality, periodicity, and associated behavior in a colony of Strombus pugilis (Mollusca, Gastropoda) in Puerto Rico. Amer Malac Bull 11: 117 121. 236. Van den Broeck H, De Wolf H, Backeljau T, Blust R (2009) Comparative assessment of reproductive impairment in the gastropod mollusc Littorina littorea along the Belgian North Sea coast. Sci Total Environ 407: 3063 3069. 237. Horiguchi T, Kojima M, Hamada F, Kajikawa A, Shiraishi H, et al. (2006) Impact of tributyltin and triphenyltin on ivory shell ( Babylonia j aponica ) populations. Environ Health Perspect 114 Suppl 1: 13 19. 238. Sloan S, Gagnon MM (2004) Intersex in Roe's abalone ( Haliotis roei ) in Western Australia. Mar Pollut Bull 49: 1123 1126. 239. Sumpter JP, Johnson AC (2008) 10th Anniversary Perspective: Reflections on endocrine disruption in the aquatic environment: From known knowns to unknown unknowns (and many things in between). J Environ Monit 10: 1476 1485. 240. Hecker M, Murphy MB, Coady KK, Villeneuve DL, Jones PD, et al. (2006) Terminology of go nadal anomalies in fish and amphibians resulting from chemical exposures. Rev Environ Contam Toxicol 187: 103 131. 241. Matthiessen P (2008) An assessment of endocrine disruption in mollusks and the potential for developing internationally standardized mol lusk life cycle test guidelines. Integr Environ Assess Manag 4: 274 284. 242. Oberdorster E, Cheek AO (2001) Gender benders at the beach: Endocrine disruption in marine and estuarine organisms. Environ Toxicol Chem 20: 23 36. 243. Denton GRW, Burdon Jones C (1986) Trace metals in algae from the Great Barrier Reef. Mar Pollut Bull 17: 98 107. 244. Zirino A, Yamamoto S (1972) A pH dependent model for the chemical speciation of copper, zinc, cadmium, and lead in seawater. Limnol Oceanogr 17: 661 671. 245. Pohl C, Croot PL, Hennings U, Daberkow T, Budeus G, et al. (2011) Synoptic transects on the distribution of trace elements (Hg, Pb, Cd, Cu, Ni, Zn, Co, Mn, Fe, and Al) in surface waters of the Northern and Southern East Atlantic. J Mar Syst 84: 28 41.
237 246. Sod erberg RW, Meade JW (1991) The effects of ionic strength on un ionized ammonia concentration. Prog Fish Cult 53: 118 120. 247. Arthur JW, West CW, Allen KN, Hedtke SF (1987) Seasonal toxicity of ammonia to five fish and nine invertebrate species. Bull Envi ron Contam Toxicol 38: 324 331. 248. Watton AJ, Hawkes HA (1984) The acute toxicity of ammonia and copper to the gastropod Potamopyrgus jenkinsi (Smith). Environ Pollut A Ecol Biol 36: 17 29. 249. Alonso A, Camargo JA (2003) Short term toxicity of ammonia, nitrite, and nitrate to the aquatic snail Potamopyrgus antipodarum (Hydrobiidae, Mollusca). Bull Environ Contam Toxicol 70: 1006 1012. 250. Berenzen N, Schulz R, Liess M (2001) Effects of chronic ammonium and nitrite contamination on the macroinvertebrate community in running water microcosms. Water Res 35: 3478 3482. 251. Alonso A, Camargo JA (2009) Long term effects of ammonia on the behavioral activity of the aquatic snail Potamopyrgus antipodarum (Hydrobiidae, Mollusca). Arch Environ Contam Toxicol 56: 796 802. 252. Baines SB, Fisher NS (2008) Modeling the effect of temperature on bioaccumulation of metals by a marine bioindicator organism, Mytilus edulis Environ Sci Technol 42: 3277 3282. 253. Smit CE, Van Gestel CA (1997) Influence of temperature on the regulation and toxicity of zinc in Folsomia candida (Collembola). Ecotoxicol Environ Saf 37: 213 222. 254. Benedicto J, Andral B, Martinez Gomez C, Guitart C, Deudero S, et al. (2011) A large scale survey of trace metal levels in coastal waters of the Western Mediterranean basin using caged mussels ( Mytilus galloprovincialis ). J Environ Monit 13: 1495 1505. 255. Protasowicki M, Dural M, Jaremek J (2008) Trace metals in the shells of blue mussels ( Mytilus edulis ) from the Poland coast of Baltic Sea. Envi ron Monit Assess 141: 329 337. 256. Richardson CA, Chenery SRN, Cook JM (2001) Assessing the history of trace metal (Cu, Zn,Pb) contamination in the North Sea through laser ablation ICP MS of horse mussel Modiolus modiolus shells. Mar Ecol Prog Ser 211: 15 7 167. 257. Puente X, Villares R, Carral E, Carballeira A (1996) Nacreous shell of Mytilus galloprovincialis as a biomonitor of heavy metal pollution in Galiza (NW Spain). Sci Total Environ 183: 205 211.
238 258. Lin MC, Liao CM (1999) 65 Zn(II) accumulation in the soft tissue and shell of abalone Haliotis diversicolor supertexta via the alga Gracilaria tenuistipitata var. liui and the ambient water. Aquaculture 178: 89 101. 259. Willie SN, Tekgul H, Sturgeon RE (1998) Immobilization of 8 hydroxyquinoline onto s ilicone tubing for the determination of trace elements in seawater using flow injection ICP MS. Talanta 47: 439 445. 260. Sohrin Y, Urushihara S, Nakatsuka S, Kono T, Higo E, et al. (2008) Multielemental determination of GEOTRACES key trace metals in seawa ter by ICPMS after preconcentration using an ethylenediaminetriacetic acid chelating resin. Anal Chem 80: 6267 6273. 261. Smith D, Flegal AR (1995) Lead in the biosphere: Recent trends. Ambio 24: 21 23. 262. Rabinowitz MB, Wetherill GW (1972) Identifying s ources of lead contamination by stable isotope techniques. Environ Sci Technol 6: 705 709. 263. Nisbet RM, Muller EB, Lika K, Kooijman SALM (2000) From molecules to ecosystems through dynamic energy budget models. J Anim Ecol 69: 913 926. 264. Caplat C, Or al R, Mahaut ML, Mao A, Barillier D, et al. (2010) Comparative toxicities of aluminum and zinc from sacrificial anodes or from sulfate salt in sea urchin embryos and sperm. Ecotoxicol Environ Saf 73: 1138 1143. 265. Jacobson T, Sundelin B, Yang G, Ford AT (2011) Low dose TBT exposure decreases amphipod immunocompetence and reproductive fitness. Aquat Toxicol 101: 72 77. 266. Aranda DA, Frenkiel L, Brule T, Montero J, Cardenas EB (2011) Occurrence of Apicomplexa like structures in the digestive gland of Stro mbus gigas throughout the Caribbean. J Invert Pathol 106: 174 178. 267. Ram JL, Ram ML, Davis JP (1982) Hormonal control of reproduction in Busycon : II. Laying of egg containing capsules caused by nervous system extracts and further characterization of the substance causing egg capsule laying. Biol Bull 162: 360 370. 268. Roberts CM (1997) Connectivity and management of caribbean coral reefs. Science 278: 1454 1457. 269. Osuna Jimenez I, Williams TD, Prieto Alamo MJ, Abril N, Chipman JK, et al. (2009) Immun e and stress related transcriptomic responses of Solea senegalensis stimulated with lipopolysaccharide and copper sulphate using heterologous cDNA microarrays. Fish Shellfish Immunol 26: 699 706.
239 270. Gahr SA, Vallejo RL, Weber GM, Shepherd BS, Silverstei n JT, et al. (2008) Effects of short term growth hormone treatment on liver and muscle transcriptomes in rainbow trout ( Oncorhynchus mykiss ). Physiol Genomics 32: 380 392. 271. Chalmers AD, Goldstone K, Smith JC, Gilchrist M, Amaya E, et al. (2005) A Xenop us tropicalis oligonucleotide microarray works across species using RNA from Xenopus laevis Mech Dev 122: 355 363.
240 BIOGRAPHICAL SKETCH Daniel Spade was born in 1985 in central Pennsylvania the son of Ann and James Sp ade, and younger brother of Nathan and Alison. He grew up in Hampden Township, Pennsylvania, which will always be home, and graduated from Cumberland Valley High School in 2003. Throughout his youth, most of his recreational time was spent in the outdoors, where he found a respect for wi ld places and an interest in nature that ultimately led to his decision to study biology. Daniel graduated from the Pennsylvania State University with distinction in 2007, earning a Bachelor of Science in biology, ecology option and a minor in philosophy Dan Ph.D. study, and gained experience in molecular biology and toxicology. While working on his Ph.D., Daniel presented research at national and international conferences, incl uding a presentation in Seville, Spain and he won the award for the Best Student Presentation in Toxicogenomics at the Society of Environmental Toxicology and Chemistry meeting in Tampa, Florida, in 2008 Dan completed his Ph.D. in August 2011 having ac cepted laboratory at Brown University to study the mechanisms of testis toxicity of anti androgenic compounds in humans. Dan looks forward to furthering his career as a toxicologist and to moving back to a place where it snow s He hopes that the Phillies will continue their recent success, that he will somed ay soon visit Ireland, and that another S teinbeck is out there somewhere He remains grateful for hav ing such wonderful family and friends, and excited for what will come next.