Orientador: Luciana Bolsoni Lourenço Morandini / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Biologia / Made available in DSpace on 2018-08-13T04:27:41Z (GMT). No. of bitstreams: 1 Targueta_CintiaPelegrineti_M.pdf: 5224061 bytes, checksum: 8bdf78a0b206d3bdf666436bc2d077b5 (MD5) Previous issue date: 2009 / Resumo: O anuro Engystomops petersi está amplamente distribuído na bacia Amazônia, compreendendo países como o Brasil, Equador, Peru e Bolívia. Estudos prévios já descreveram variações no canto e também variações genéticas nesse anuro. Nós apresentamos aqui uma análise citogenética da população de Puyo (Equador), um sítio que inclui a região descrita como sendo a localidade-tipo dessa espécie, duas outras populações Equatorianas que apresentam isolamento pré-zigótico (Yasuní e La Selva), e uma população Brasileira do Estado do Acre. Todos os espécimes analisados citogeneticamente foram também incluídos em estudos filogenéticos utilizando seqüências de DNA mitocondrial. A sequência de parte de um gene nuclear (rodopsina) também foi utilizada neste trabalho. Uma enorme variação citogenética foi encontrada e muitos padrões cariotípicos puderam ser observados, mesmo entre populações que não apresentam isolamento pré-zigótico. Este resultado nos permite sugerir que dados cariotípicos podem ter participado no processo de especiação de Engystomops. Populações de Puyo e do Acre apresentaram cromossomos sexuais heteromórficos compondo o sistema XX/XY, o que não foi identificado nas outras populações Equatorianas. Já que as populações Equatorianas estão localizadas em um clado separado das populações do Acre, nós sugerimos a hipótese de que a presença dos cromossomos sexuais X e Y seja um caráter plesiomórfico neste grupo. A análise do gene nuclear permitiu ainda identificar seis SNPs (Polimorfismos de simples nucleotídeos) e alguns espécimes heterozigotos para esses sítios. As variações cariotípicos e moleculares aqui apresentados, em conjunto com os dados previamente descritos, corroboram a hipótese de que Engystomops petersi seja um complexo de espécies e sugerem que eventos de hibridação e introgressão possam ter ocorrido entre as populações desse grupo. / Abstract: The frog Engystomops petersi is widely distributed in the Amazon basin. Genetic divergence and speciation mechanisms of E. petersi populations have been inferred mainly from mitochondria DNA (mtDNA) and microsatellite polymorphisms, male advertisement call and sexual selection. In spite of the significant progress resulting from these studies, many cytogenetic, taxonomic and evolutionary aspects of E. petersi populations remain unclear. In the present study, cytogenetic analysis and nucleotide divergence in mtDNA and in the rhodopsin nuclear gene of E. petersi populations are described in an attempt to contribute to the understanding of this anuran. High cytogenetic variation distinguished karyotypic patterns, which included two populations with no prezygotic isolation. The Puyo and Acre populations showed heteromorphic sexual chromosomes (XX/XY), which were not identified in the other Equatorial populations. Since all the Equatorial populations were placed in a clade separated from the Acre populations, we hypothesize that the sex chromosomes X and Y is a plesiomorphic condition in this group. The results from mtDNA found here were very similar to those found in previous studies. In the rhodopsine nucleotide sequences, 6 SNPs (Single Nucleotide Polymorphism) were identified and various specimens were heterozygous for these nucleotide sites. The karyotypic and molecular data presented herein, in conjunction with previously reported data, corroborate the hypothesis that Engystomops petersi is a complex of distinct species and suggested hybridization and introgression events between populations. Moreover, tha data indicated that karyotypic evolution patterns may have played a substantial role in the Engystomops speciation. / Mestrado / Biologia Celular / Mestre em Biologia Celular e Estrutural

Orientador: Luciana Bolsoni Lourenço Morandini / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Biologia / Made available in DSpace on 2018-08-23T15:06:51Z (GMT). No. of bitstreams: 1 Targueta_CintiaPelegrineti_D.pdf: 14058240 bytes, checksum: 609267a5b0419113c3e1035f998113c6 (MD5) Previous issue date: 2013 / Resumo: Dentre as espécies do gênero Engystomops, somente as três espécies do clado Edentulus (E. freibergi, E. petersi e E. pustulosus) e duas das seis espécies do clado Duovox (E. pustulatus e E. puyango) tinham o número diploide descrito, e a localização cariotípica das NORs (regiões organizadores de nucléolo) e das regiões heterocromáticas era conhecida apenas para E. freibergi, E. petersi e E. puyango. Os dados disponíveis mostravam que as espécies do clado Duovox já cariotipadas apresentavam 2n=20, enquanto o número cromossômico diploide de todas as espécies do clado Edentulus era 2n=22, assim como o das espécies de Physalaemus e Edalorhina, gêneros mais proximamente relacionados a Engystomops. Dessa forma, uma redução do número diploide de 2n=22 para 2n=20 era considerada na história de Engystomops, mas não era possível inferir se essa era uma sinapomorfia de E. puyango e E. pustulatus ou de todo o clado Duovox. Em relação ao clado Edentulus, algumas dúvidas também persistiam. Recentes estudos propunham a presença de um complexo de espécies amazônicas sendo confundidas com E. petersi. Citogeneticamente puderam ser reconhecidos três grupos nesse complexo, que corresponderam a populações de Puyo (Equador), Yasuní (Equador) e La Selva (Equador), diferentes ainda daquele de E. freibergi (Brasil). Nem sempre é possível a identificação de homeologias cromossômicas entre esses grupos, o que dificulta o reconhecimento dos possíveis rearranjos cromossômicos envolvidos na divergência cariotípica no gênero em questão. Também permanecia pouco explorada a diferenciação entre os cromossomos sexuais heteromórficos reconhecidos em populações de E. petersi e em E. freibergi. Os objetivos do presente trabalho foram estudar citogeneticamente as espécies de Engystomops do clado Duovox, melhor caracterizar os cromossomos sexuais de E. freibergi por meio de pintura cromossômica e buscar novos marcadores citogenéticos para auxiliar na identificação de homeologias cromossômicas nesse gênero. A análise cariotípica mostrou que as espécies E. randi, E. guayaco, E. montubio, E. pustulatus e E. coloradorum apresentaram 2n=20, assim como as espécies E. puyango e E. pustulatus, previamente cariotipadas, corroborando a hipótese de essa característica provavelmente seja uma sinapomorfia do grupo Duovox. Em E. coloradorum, foi encontrada uma fêmea triploide, que consistiu no primeiro relato de poliploidia para Engystomops. Em todas e somente nas fêmeas dessa espécie foi encontrado um heteromorfismo referente à presença de NOR dentre os cromossomos 10 homólogos o que sugere que esses sejam cromossomos sexuais e que o sistema de determinação do sexo em E. coloradorum seja ZZ/ZW. Quanto à análise dos cromossomos sexuais de E. freibergi, foi possível isolar por microdissecção os cromossomos X e Y e produzir sondas a partir da amplificação de segmentos desses cromossomos. A utilização da técnica de pintura cromossômica com tais sondas permitiu inferir que as regiões proximais do cromossomo X e do cromossomo Y de E. freibergi apresentam similaridade molecular, o que sugere a possibilidade de se tratarem de regiões pseudo-autossômicas. As sondas dos cromossomos sexuais de E. freibergi não detectaram os cromossomos X e Y de E. petersi de Puyo nem os cromossomos do par 11 homomórfico de E. petersi de Yasuní, não evidenciando, portanto, grande similaridade entre eles. O isolamento, a caracterização e a localização cariotípica do DNAr 5S das espécies pertencentes ao grupo Duovox permitiram inferir a possível homeologia do par cromossômico 6 dessas espécies entre si e com o par 6 das espécies do grupo Edentulus. Ainda, o uso de sondas PcP190EcoRI, referentes a uma família de DNA repetitivo derivada de DNAr 5S, permitiu identificar os cromossomos 5 de E. randi, E. guayaco e E. coloradorum e diferenciá-los dos cromossomos 6 dessas espécies, portadores do sítio de DNAr 5S do tipo II. Além disso, essa mesma sonda identificou a região pericentromérica do braço curto do cromossomo 3 (e do cromossomo 5 de E. petersi de Yasuní) de todas as espécies do gênero, com excessão de E. coloradorum, sugerindo outra possível homeologia entre esses cromossomos. Dessa forma, com o emprego de técnicas citogenéticas clássicas e moleculares foi possível fornecer importante contribuição para o estudo da evolução do gênero Engystomops / Abstract: Of the species in the genus Engystomops, only three species of the Edentulus clade (E. freibergi, E. petersi and E. pustulosus) and two species of the Duovox clade (E. pustulatus and E. puyango) have had their diploid number described, while the NOR (nucleolus organizer region) and the heterochromatic sites were known only for E. freibergi, E. petersi and E. puyango karyotypes. The species of Duovox clade had 2n=20, while the diploid chromosome number of all species of Edentulus clade was 2n=22, which is the same diploid number of species of Physalaemus and Edalohina, genera closely related to Engystomops. Thus, the diploid number reduction from 2n=22 to 2n=20 was supposed, but it was not possible to infer if this was a synapomorphy of E. puyango and E. pustulatus or a synapomorphy of the entire Duovox clade. With regards to the Edentulus clade, some doubts also persisted. Recent studies proposed a complex of Amazonian species misidentified as E. petersi and it was possible to cytogenetically recognize three groups in this complex that corresponded to populations from Puyo (Ecuador), Yasuní (Ecuador) and La Selva (Ecuador). All of these karyotypes also differed from the E. freibergi (Brazil) karyotype. Chromosome homologies are not easily recognized between these groups, a fact that makes difficult the identification of chromosome rearrangements involved in karyotypic divergence in this genus. Additionally, the differentiation of the heteromorphic sex chromosomes found in the E. petersi and E. freibergi karyotypes remained unexplored. The goals of the present work are (1) to cytogenetically study the species of Duovox clade, (2) to better characterize the sex chromosomes of E. freibergi by means of chromosome painting and (3) to find new cytogenetic markers to identify potential chromosome homologies in this genus. The karyotypic analysis showed that the species E. randi, E. guayaco, E. montubio, E. pustulatus and E. coloradorum had 2n=20, as previously described in E. pustulatus and E. puyango. This corroborates the hypothesis that the diploid number 2n=20 is likely a synapomorphy of the Duovox clade. Among the specimens of E. coloradorum, a triploid female was found, which is the first report of polyploidy for the genus Engystomops. In all the females of this species, there was a NOR heteromorphism in the homologous chromosomes 10, suggesting that these may be ZZ/ZW sex chromosomes. Regarding the sex chromosomes of E. freibergi, it was possible to isolate the X and Y chromosomes by microdissection and to produce probes by amplifying segments of the microdissected chromosomes. The hybridization of these probes in E. freibergi metaphases showed that the proximal regions of the X and Y chromosomes of this species had molecular similarity, thus suggesting that they may be pseudoautosomal regions. The sex chromosome probes from E. freibergi detected neither the X nor the Y chromosome of E. petersi from Puyo nor the chromosomes of homomorphic pair 11 of E. petersi from Yasuní. This implies no significant similarity between them and the E. freibergi sex chromosomes. The isolation, characterization and karyotypic localization of the rDNA 5S from the species of the Duovox clade made it possible to infer the homeology of chromosome pair 6 of those species, as well as their homeology with chromosome pair 6 of the species of the Edentulus clade. Further, the probes PcP190EcoRI, which are related to a family of satellite DNA derived from rDNA 5S, made it possible to identify chromosome 5 of E. randi, E. guayaco and E. coloradorum and to differentiate them from chromosome 6, which carry type II rDNA 5S. In conclusion, both classical and molecular cytogenetic techniques provided important contributions for the study of the evolution of the genus Engystomops / Doutorado / Biologia Celular / Doutora em Biologia Celular e Estrutural

Thesis (Ph. D.)--University of Texas at Austin, 2007. / Vita. Includes bibliographical references.

Sexually selected traits are among the most costly, complex, and conspicuous elements of the phenotype. In polygynous reproductive systems, they evolve under strong selection by females. Why females favor those traits, however, is an on-going debate. Here, I use túngara frogs as a model system to study the evolution of communication under sexual selection. The wealth of available information on the behavior, neurophysiology, and reproductive biology of túngara frogs make them an ideal system to understand the patterns of signal evolution and explore the processes that have shaped them. In chapter 1 and 2, I review the taxonomy of túngara frogs (Engystomops) from western Ecuador. I describe three new species including their external morphology and advertisement calls. In chapter 3, I explore the phylogenetic relationships of túngara frogs, testing the support for alternative relationships previously reported for this group. The new phylogeny provides the framework for the comparative analysis carried out in chapters 4 and 5. In chapter 4, I present new female preference and male advertisement call data to test the sensory exploitation hypothesis of sexual selection. Using ancestral character reconstruction, I found that female preferences for complex calls did not originated before the appearance of complex calls, as predicted by sensory exploitation. Instead, my results suggest that the origin of complex calls and their preference originated at similar times. Finally, in chapter 5, I analyze the macroevolutionary patterns of call variation in male túngara frogs. A generalized least squares model demonstrates that advertisement calls have a strong phylogenetic signal. Although most species in Engystomops have distinctive calls, they share a common acoustic structure with two components that evolve at different rates. I did not found evidence of greater call differentiation among sympatric species relative to allopatric species.

Orientador: Luciana Bolsoni Lourenço Morandini / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Biologia / Made available in DSpace on 2018-08-20T20:53:49Z (GMT). No. of bitstreams: 1 Rodrigues_DeboraSilva_M.pdf: 3296545 bytes, checksum: 2a2ccc454f9a1df9f09173c2127af330 (MD5) Previous issue date: 2012 / Resumo: O gênero Engystomops apresenta ampla distribuição geográfica e constitui um interessante grupo de anuros para estudos cariotípicos. As populações de Engystomops encontradas na Amazônia têm sua identificação taxonômica ainda controversa. Análises genéticas e citogenéticas apoiam hipóteses que sugerem a existência de um complexo de espécies crípticas e especiação incipiente. Muitas vezes a variação citogenética observada entre diferentes populações estudadas dificultou o reconhecimento de homeologias cromossômicas entre os cariótipos. Uma caracterização cromossômica mais detalhada poderia auxiliar no possível reconhecimento de homeologias cromossômicas e, dessa forma, contribuir para o estudo dos processos envolvidos na divergência desses anuros. Já que o gene do DNAr 5S tem sido importante marcador genético e citogenético para estudos evolutivos e para a identificação e comparação de espécies em diversos grupos, no presente trabalho o DNAr 5S de Engystomops freibergi e de exemplares de Engystomops petersi de duas localidades Equatorianas (Puyo e Yasuní) foi estudado. Em todos os casos, dois tipos de DNAr 5S, facilmente diferenciados pelo tamanho e composição da sequência do seu espaçador não transcrito, foram isolados. A provável região promotora do gene do RNAr 5S (ICR) foi localizada nos dois tipos de sequências de DNAr 5S e a presença de possíveis sequências regulatórias adicionais foi discutida. No cariótipo de E. freibergi, sonda contendo a unidade repetitiva do DNAr 5S tipo I hibridou na região pericentromérica do braço curto dos cromossomos do par 3, e o DNAr 5S tipo II foi mapeado na região distal do braço longo dos cromossomos do par 6. A sonda formada somente pela região de NTS do DNAr 5S tipo I claramente detectou a região pericentromérica de 3p nos cariótipos de E. freibergi e E. petersi (Puyo) e de 5p no cariótipo de E. petersi (Yasuní), porém nenhum sinal distal ou intersticial foi observado. A sonda formada pela região de NTS do DNAr 5S tipo II detectou apenas a região distal de 6q nos três cariótipos estudados, corroborando a distribuição diferencial dos dois tipos de DNAr 5S nesses cariótipos. Tais sítios de DNAr 5S constituem novos marcadores cromossômicos, os quais permitem sugerir a homeologia entre o cromossomo 6 dos cariótipos de E. freibergi e de E. petersi, e entre o cromossomo 5 do cariótipo de E. petersi de Yasuní e o cromossomo 3 dos cariótipos de E freibergi e de E. petersi de Puyo. Já que os dois tipos de DNAr 5S encontrados em Engystomops são relacionados àqueles de Physalaemus tanto quanto à composição nucleotídica quanto à localização cromossômica, é ainda possível inferir que a origem desses dois tipos de sequências tenha antecedido a divergência evolutiva desses gêneros / Abstract: The Engystomops genus is widely distributed geographically and constitutes an interesting group for karyotypic studies. The taxonomic identifications of Amazonian populations of Engystomops is still controversial. Genetic and cytogenetic analyses suggest the existence of a complex of cryptic species and incipient speciation. The cytogenetic variations found among some populations prevent the recognition of chromosomal homeologies between the described karyotypes. A more detailed chromosomal characterization could help in the recognition of chromosome homeologies and, therefore, could contribute to the study of the processes involved in the divergence of these anurans. Since the 5S rDNA gene has been an important genetic and cytogenetic marker for evolutionary studies and even for the identification and comparison of species in diverse groups, in the present work the 5S rDNA of Engystomops freibergi and exemplars of Engystomops petersi from two Ecuadorian locations (Puyo and Yasuní) was studied. In all cases, two types of 5S rDNA, easily differed by size and compositon of their non-transcribed spacer, were isolated. A putative promoting region of the 5S rRNA gene (ICR) was recognized in the two types of 5S rDNA sequences and the presence of possible additional regulatory sequences was discussed. In the E. freibergi karyotypes, the entire type I 5S rDNA repeating unit hybridized to the pericentromeric region of 3p, whereas the entire type II 5S rDNA repeating unit mapped to the distal region of 6q, suggesting a differential localization of these sequences. The type I NTS probe clearly detected the 3p pericentromeric region in the karyotypes of E. freibergi and E. petersi from Puyo and the 5p pericentromeric region in the karyotype of E. petersi from Yasuní, but no distal or interstitial signals were observed. Interestingly, this probe also detected many centromeric regions in the three karyotypes, suggesting the presence of a satellite DNA family derived from 5S rDNA. The type II NTS probe detected only distal 6q regions in the three karyotypes, corroborating the differential distribution of the two types of 5S rDNA. Because the 5S rDNA types found in Engystomops are related to those of Physalaemus with respect to their nucleotide sequences and chromosomal locations, their origin likely preceded the evolutionary divergence of these genera. In addition, our data indicated homeology between chromosome 5 in E. petersi from Yasuní and chromosomes 3 in E. freibergi and E. petersi from Puyo. In addition, the chromosomal location of the type II 5S rDNA corroborates the hypothesis that the chromosomes 6 of E. petersi and E. freibergi are homeologous despite the great differences observed between the karyotypes of the Yasuní specimens and the others / Mestrado / Biologia Celular / Mestre em Biologia Celular e Estrutural

It is generally thought that in amphibians, thyroid hormones (THs) regulate metamorphosis, while sex steroids (estrogens and androgens) regulate gonadal differentiation. However, inhibition of TH synthesis in frogs alters gonadal differentiation, suggesting instead that these two endocrine axes interact during development. Specifically, THs may be involved in male development, while estrogens may inhibit tadpole metamorphosis. However, we do not currently know the mechanisms that account for these interactions, let alone how such mechanisms may differ between species. To develop and test new hypotheses on the roles of sex steroids and THs, I first examined transcriptional profiles (mRNA) of enzymes and receptors related to sex steroids and THs during embryogenesis and metamorphosis in Silurana tropicalis. Tadpoles were exposed to either an estrogen synthesis inhibitor (fadrozole) or TH (triiodothyronine, T3) during early larval or tadpole development. Acute exposures of S. tropicalis to fadrozole or T3 during early development resulted in increased expression of androgen- and TH-related genes in whole body larvae, while chronic exposure to fadrozole during metamorphosis affected gonadal differentiation but did not affect tadpole development. On the other hand, acute exposure to T3 during metamorphosis increased the expression of androgen-related transcripts both in the brain and gonad. In S. tropicalis, the results suggested that cross-talk is primarily in one direction (i.e., effect of THs on the reproductive axis) with a strong relationship between TH and androgen status. Lastly, I established developmental transcript profiles and investigated T3 regulation of brain and gonad transcripts in Engystomops pustulosus. I then compared these results with S. tropicalis and an earlier study in Lithobates pipiens. While each species developed with similar profiles, they differed in their response to T3. Exposure to T3 resulted in either an increase in androgen-related genes (S. tropicalis) or a decrease in estrogen-related genes (E. pustulosus and L. pipiens). In conclusion, these data demonstrated that cross-talk mechanisms differ among these three evolutionary separate species, but in all cases, T3 appears to affect the balance of sex steroids, stimulating the androgen system and providing potential mechanisms of the masculinising effects of THs. These results will contribute to understanding the mechanisms of hormone interactions and their evolutionary basis in frogs.

It is generally thought that in amphibians, thyroid hormones (THs) regulate metamorphosis, while sex steroids (estrogens and androgens) regulate gonadal differentiation. However, inhibition of TH synthesis in frogs alters gonadal differentiation, suggesting instead that these two endocrine axes interact during development. Specifically, THs may be involved in male development, while estrogens may inhibit tadpole metamorphosis. However, we do not currently know the mechanisms that account for these interactions, let alone how such mechanisms may differ between species. To develop and test new hypotheses on the roles of sex steroids and THs, I first examined transcriptional profiles (mRNA) of enzymes and receptors related to sex steroids and THs during embryogenesis and metamorphosis in Silurana tropicalis. Tadpoles were exposed to either an estrogen synthesis inhibitor (fadrozole) or TH (triiodothyronine, T3) during early larval or tadpole development. Acute exposures of S. tropicalis to fadrozole or T3 during early development resulted in increased expression of androgen- and TH-related genes in whole body larvae, while chronic exposure to fadrozole during metamorphosis affected gonadal differentiation but did not affect tadpole development. On the other hand, acute exposure to T3 during metamorphosis increased the expression of androgen-related transcripts both in the brain and gonad. In S. tropicalis, the results suggested that cross-talk is primarily in one direction (i.e., effect of THs on the reproductive axis) with a strong relationship between TH and androgen status. Lastly, I established developmental transcript profiles and investigated T3 regulation of brain and gonad transcripts in Engystomops pustulosus. I then compared these results with S. tropicalis and an earlier study in Lithobates pipiens. While each species developed with similar profiles, they differed in their response to T3. Exposure to T3 resulted in either an increase in androgen-related genes (S. tropicalis) or a decrease in estrogen-related genes (E. pustulosus and L. pipiens). In conclusion, these data demonstrated that cross-talk mechanisms differ among these three evolutionary separate species, but in all cases, T3 appears to affect the balance of sex steroids, stimulating the androgen system and providing potential mechanisms of the masculinising effects of THs. These results will contribute to understanding the mechanisms of hormone interactions and their evolutionary basis in frogs.

It is generally thought that in amphibians, thyroid hormones (THs) regulate metamorphosis, while sex steroids (estrogens and androgens) regulate gonadal differentiation. However, inhibition of TH synthesis in frogs alters gonadal differentiation, suggesting instead that these two endocrine axes interact during development. Specifically, THs may be involved in male development, while estrogens may inhibit tadpole metamorphosis. However, we do not currently know the mechanisms that account for these interactions, let alone how such mechanisms may differ between species. To develop and test new hypotheses on the roles of sex steroids and THs, I first examined transcriptional profiles (mRNA) of enzymes and receptors related to sex steroids and THs during embryogenesis and metamorphosis in Silurana tropicalis. Tadpoles were exposed to either an estrogen synthesis inhibitor (fadrozole) or TH (triiodothyronine, T3) during early larval or tadpole development. Acute exposures of S. tropicalis to fadrozole or T3 during early development resulted in increased expression of androgen- and TH-related genes in whole body larvae, while chronic exposure to fadrozole during metamorphosis affected gonadal differentiation but did not affect tadpole development. On the other hand, acute exposure to T3 during metamorphosis increased the expression of androgen-related transcripts both in the brain and gonad. In S. tropicalis, the results suggested that cross-talk is primarily in one direction (i.e., effect of THs on the reproductive axis) with a strong relationship between TH and androgen status. Lastly, I established developmental transcript profiles and investigated T3 regulation of brain and gonad transcripts in Engystomops pustulosus. I then compared these results with S. tropicalis and an earlier study in Lithobates pipiens. While each species developed with similar profiles, they differed in their response to T3. Exposure to T3 resulted in either an increase in androgen-related genes (S. tropicalis) or a decrease in estrogen-related genes (E. pustulosus and L. pipiens). In conclusion, these data demonstrated that cross-talk mechanisms differ among these three evolutionary separate species, but in all cases, T3 appears to affect the balance of sex steroids, stimulating the androgen system and providing potential mechanisms of the masculinising effects of THs. These results will contribute to understanding the mechanisms of hormone interactions and their evolutionary basis in frogs.

It is generally thought that in amphibians, thyroid hormones (THs) regulate metamorphosis, while sex steroids (estrogens and androgens) regulate gonadal differentiation. However, inhibition of TH synthesis in frogs alters gonadal differentiation, suggesting instead that these two endocrine axes interact during development. Specifically, THs may be involved in male development, while estrogens may inhibit tadpole metamorphosis. However, we do not currently know the mechanisms that account for these interactions, let alone how such mechanisms may differ between species. To develop and test new hypotheses on the roles of sex steroids and THs, I first examined transcriptional profiles (mRNA) of enzymes and receptors related to sex steroids and THs during embryogenesis and metamorphosis in Silurana tropicalis. Tadpoles were exposed to either an estrogen synthesis inhibitor (fadrozole) or TH (triiodothyronine, T3) during early larval or tadpole development. Acute exposures of S. tropicalis to fadrozole or T3 during early development resulted in increased expression of androgen- and TH-related genes in whole body larvae, while chronic exposure to fadrozole during metamorphosis affected gonadal differentiation but did not affect tadpole development. On the other hand, acute exposure to T3 during metamorphosis increased the expression of androgen-related transcripts both in the brain and gonad. In S. tropicalis, the results suggested that cross-talk is primarily in one direction (i.e., effect of THs on the reproductive axis) with a strong relationship between TH and androgen status. Lastly, I established developmental transcript profiles and investigated T3 regulation of brain and gonad transcripts in Engystomops pustulosus. I then compared these results with S. tropicalis and an earlier study in Lithobates pipiens. While each species developed with similar profiles, they differed in their response to T3. Exposure to T3 resulted in either an increase in androgen-related genes (S. tropicalis) or a decrease in estrogen-related genes (E. pustulosus and L. pipiens). In conclusion, these data demonstrated that cross-talk mechanisms differ among these three evolutionary separate species, but in all cases, T3 appears to affect the balance of sex steroids, stimulating the androgen system and providing potential mechanisms of the masculinising effects of THs. These results will contribute to understanding the mechanisms of hormone interactions and their evolutionary basis in frogs.