Nonavian Reptile Reproduction Functions With a Reduced Gonadotropin System

Abstract

Vertebrate reproduction is controlled by 2 pituitary gonadotropin hormones (GtHs), FSH and LH, binding to gonadotropin hormone receptors (GtHRs) in gonadal tissues. All gnathostome vertebrates have been confirmed to possess at least 1 receptor for each GtH [LH receptor (LHR) and FSH receptor (FSHR)], except for species of the reptilian (nonavian sauropsidan) orders, such as lepidosauria, testudines, and crocodylia, which showed inexplicable reactions to heterologous amphibian, avian, and mammalian GtHs in early endocrinological studies. This study investigated the number and function of reptilian GtHRs. Genomic and transcriptomic analyses of selected tetrapod species now strongly suggest the inactivation of the LHR in all nonavian sauropsidans. This gene inactivation likely occurred independently in 3 branches of the sauropdisan clade, sparing only the avian class. Bioassays served to investigate the binding specificity of squamate, chelonian, crocodilian, avian, and mammalian GtHRs with their homologous and heterologous GtHs. The FSHR of a squamate lizard proved completely promiscuous to both its homologous GtHs, while the chelonian FSHR responded slightly stronger to the homologous LH than FSH, and the crocodylian FSHR was only stimulated by the homologous LH but not FSH. We therefore propose a modified paradigm with a neuroendocrine control of nonavian reptilian reproduction by a single GtHR and either 1 GtH in crocodylians or 2 GtHs in chelonians and squamate reptiles. Finally, we discuss hypotheses of tightly regulated temporal and spatial expression of the remaining FSHR in different gonadal somatic cells and temperature-dependent functions of the single nonavian reptilian GtHR.

Vertebrate reproduction relies on a hierarchical endocrine system along the hypothalamus-pituitary-gonadal axis. Gonadotropic hormones (GtHs; FSH and LH) are secreted from the pituitary, and the corresponding gonadotropin hormone receptors (GtHRs) are expressed in target tissues, primarily the gonads. In a complex mechanism of feedback regulations with gonadal steroid hormones (estrogen, progesterone, testosterone), GtHs trigger a series of morphological and biochemical changes in the gametogenetic tissues, resulting in gamete recruitment, selection, and maturation (1). In mammals, a dual GtH/GtHR system is present in males and females: FSH and its receptor (FSHR) are responsible for the recruitment and early maturation of gametes, whereas LH and its receptor (LHR) are responsible for the final maturation and release or ovulation of gametes (1).

This basic hierarchical mechanism applies in essence to all vertebrate species, and it was assumed that the nature and function of the regulating hormones in any vertebrate species conformed to the typical mammalian pattern (2, 3). Early comparative endocrinological studies relied on conclusions drawn from the effects of mammalian hormones administered to nonmammalian species in vivo (measuring biological effects, such as gonadal growth, gamete development, and ovulation) or gonadal tissues (radioligand binding studies). While the amphibian and avian systems responded to both mammalian hormones, FSH and LH (4-7), actinopterygian fish appeared to be only stimulated by mammalian LH but not FSH (8, 9). Lepidosaurian and chelonians reacted primarily to mammalian FSH but with more variable effects than would have been expected in a mammalian system (10-17). Therefore, significant differences in hormone and receptor structures and functions between mammalian and nonmammalian systems were assumed (18, 19).

Basic molecular structure and function of both GtHs and GtHRs appear to be conserved across all higher vertebrate taxa. FSH and LH are produced in the adenohypophysis of the pituitary gland and belong, together with the TSH, to the glycoprotein hormone family. The 3 members bear carbohydrate moieties that influence their biosynthesis, secretion, half-life, and biological potency (20). Each of these hormones is comprised of a common α subunit noncovalently bound to a variable β subunit that confers the receptor-binding specificity for each dimeric ligand (20). FSHR and LHR are transmembrane glycoproteins that belong, together with the TSH receptor , to the superfamily of G protein-coupled receptors (21). They are composed of a N-terminal extracellular domain, containing a high-affinity ligand binding domain, 7 transmembrane domains, and a cytoplasmic C-terminal domain (22, 23). All of the pituitary glycoprotein hormone receptors are encoded by a single gene and have several exons (10 in FSHR and 11 in LHR), with all the introns located in the gene section coding for the N-terminal extracellular domain region of the receptor (22). In all vertebrate groups, binding of GtHs to their cognate receptors results in the activation of adenylyl cyclase and subsequently in an increase of cAMP levels in the target cell (24-28). Gonadotropin receptors are mainly expressed in the gonads of vertebrates. In mammalian males, FSHR and LHR are expressed in testicular Sertoli and Leydig cells, respectively. In mammalian and avian females, FSHR is expressed in ovarian granulosa cells only, while LHR is expressed in theca and granulosa cells of maturing preovulatory follicles. In acinopterygian fish species, Sertoli and Leydig cells (in males) as well as theca and granulosa cells (in females) can express both gonadotropin receptors (29, 30).

The number of GtHs and GtHRs is not uniform across vertebrate taxa. The oldest extant lineage of vertebrates, Agnatha (lampreys and hagfishes), has only 1 GtH and 1 GtHR (24). Genome duplication events in gnathostomes—actinopterygian and sarcoptgerygian fish, amphibian, sauropsida (reptiles and birds), and mammals—resulted in the dual GtH-GtHR system with FSH/FSHR and LH/LHR in presumably all higher vertebrate taxa (31, 32). As GtHs and GtHRs are encoded by paralogous genes in gnathostomes, they constitute an example of ligand-receptor coevolution (33). While all gnathostomes have 2 GtHs, the presence of 2 GtHRs in actinopterygian fish, amphibian, and sauropsidan classes had long been controversial (34-37) until studies in teleost fish (38), amphibians (39, 40), and birds (41) confirmed both GtHRs. In lepidosaurians and chelonians, endocrinological research primarily focused on the biological effects of FSH and did not touch upon the presence of the LH system (12, 42-46). Similarly, molecular analyses revealed the sole presence of the FSHR gene in several chelonian, squamate, and crocodylian species (22, 47-50). Gene fragments corresponding to LHR genes of other vertebrates were detected in chelonians and crocodylians, but they contained several premature stop codons and frameshifts, rendering their function questionable; in squamate species, no evidence for LHR genes was found at all (51).

GtHRs show strict specificity for their respective GtH ligands in elasmobranch fish and mammals (23, 52-54). In these taxa, cross-activation of noncognate receptors only happens at very high, nonphysiological hormone concentrations, the so-called “specificity spillover syndrome” (27, 55). While in sturgeons, a nonteleost actinopterygiian fish, permissive activation of both FSHR and LHR by noncognate GtHs was demonstrated (56), the specificity of GtHR for their cognate ligands is variable in teleost actinopterygiian fish. Indeed, the majority of examined teleost fish show FSHR promiscuity to both GtHs combined with high LHR specificity (25, 57-60). However, in at least 1 teleost fish species, LHR cross-reacts with both GtHs (61). Amphibians appear to have the same situation: FSHR promiscuity with high LHR specificity (35, 39). In stark contrast to the wealth of research performed on GtHR specificity across fish species, there are no molecular studies available on GtHR activation specificity in avian and reptilian species. Thus, the presence of 2 GtHRs and their interactions with the 2 GtHs is unresolved for lepidosauria, testudines, and crocodylia.

Here, we aim to elucidate the composition and function of the gonadotropin-receptor system in the major nonavian sauropsidan lineages. Although the term “reptilia” has become obsolete in current phylogenetic nomenclature of sauropsidan amniota (62), for ease of reading, the vernacular zoological term “reptiles” was adopted for the group of modern nonavian sauropsidan classes comprising crocodylia, lepidosauria, and testudines. To shed light on the evolution and binding properties of these receptors, we (1) performed synteny analyses across vertebrate genomes; (2) produced novel transcriptomic data from reptilian gonads (squamate lizard, squamate snake, chelonian, crocodilian); (3) developed specific bioassays for representative species of each reptilian lineage (squamate lizard, chelonian, crocodilian) to assess the specificity of GtHs to GtHRs within the species/clade; and (4) applied cross-species bioassays to evaluate the evolutionary adaptation of the reptilian ligands and receptors. Based on the outcome of these analyses, we propose that a unique reptilian FSHR takes over the role of both GtHRs as an adaptation to the loss of a vital receptor in vertebrate endocrine regulation.

Material and Methods

Syntenic Analysis

Initial synteny analysis of the region harboring the LHR and FSHR genes in human (Homo sapiens), mouse (Mus musculus), and frog (Xenopus laevis) were done using the synteny database (http://syntenydb.uoregon.edu/synteny_db/, last accessed December 26, 2024) with a gene search for FSHR and a sliding window size of 25 genes (63). Corresponding syntenic regions of 15 additional species (for details on the species investigated, see figure legend of Fig. 1) were analyzed using the Ensembl chromosomal viewer. Regions with genes orthologous to the human LHR and FSHR as well as genes flanking these receptors were subjected to a microsynteny analyses and orthologous/paralogous genes of all the species of interest were analyzed in Ensembl (release 108, last accessed December 20, 2022) and confirmed by reciprocal tBLASTx searches. Figures were edited using Coral draw (Coral Corporation).

Figure 1.

Synteny analysis of the genomic region harboring LHR and FSHR in major vertebrate clades. The human (Homo sapiens; Hs) alignment, given in the center, was used as template and compared to the following genomes: Oa (Ornithorhynchus anatinus, platypus), Md (Monodelphis domestica, opossum), Bt (Bos Taurus, cattle), Mm (Mus musculus, mouse), Cf (Canis familiaris, dog), Xt (Xenopus tropicalis, western clawed frog), Ll (Leptobrachium leishanense, Leishan spiny toad), Gg (Gallus gallus, chicken), Tg (Taeniopygia guttata, zebra finch), Cpo (Crocodylus porosus, saltwater crocodile), Nn (Naja naja, Indian cobra), Ge (Gopherus evgoodei, Goodes thornscrub tortoise), Cpi (Chrysemys picta bellii, painted turtle), Ps (Pelodiscus sinensis, Chinese softshell turtle), Ac (Anolis carolinensis, green anole), Vk (Varanus komodoensis, Komodo dragon), and Sp (Sphenodon punctatus, tuatara). Gonadotropin hormone receptor genes (LHR and FSHR) are highlighted in red and inactivated or lost LHR are indicated by an X. Boxes with gene names represent conserved syntenic genes, whereas boxes without gene names are either species-specific genes or genes resulting from a genomic rearrangement (insertion). Broken lines represent genomic insertions. Note that due to the fragmented nature of the Chinese softshell turtle (Ps) and tuatara (Sp) genomes, we could only evaluate the region directly flanking FSHR.

De Novo Transcriptome Assembly and Analysis for GtH (FSH and LH) and GtHR (FSHR and LHR) in Reptiles

Reproductive and pituitary tissues were harvested from 1 female and 1 male of 1 representative reptile species per nonavian sauropsidan clade: testicular, ovarian, oviductal, and female pituitary tissues from the veiled chameleon (Chamaeleo calyptratus), testicular and ovarian tissues of the corn snake (Panterophis guttatus), the Greek tortoise (Testudo hermanni), and the Nile crocodile (Crocodylus niloticus) [Supplemental Table S1 (64)]. All animals were sexually mature, except for the Nile crocodile, which was represented by 70-day-old embryos, already expected to express GtHR (65). Tissues were harvested from animals either maintained or bred at the LANE animal facility (University of Geneva, Switzerland, experimentation permits GE24A/33145 and GE10619C/31521, approved by the Geneva veterinary cantonal authorities) or opportunistically from patients to be castrated surgically at the Clinic for Zoo Animals, Exotic Pets and Wildlife Zurich, University of Zurich, Switzerland.

RNA was extracted using the RNeasy kit (74104, Qiagen) according to the manufacturer's instructions. RNA quality was assessed using the Agilent Technologies 2100 Bioanalyzer, and RNA integrity number values were greater than 6.7 for all samples. The sequencing libraries were prepared using the TruSeq stranded mRNA Illumina kit, and 100 bp pair-end reads were generated using Illumina HiSeq technology.

Reads were preprocessed for trimming adapters and quality filtering (reads below Q30 Phred score were discarded) using Cutadapt software v3.4 (66). RNA sequencing (RNA-Seq) read quality was assessed before and after adapter/quality trimming using FastQC software v0.11.9 (67). De novo transcriptome assembly was performed with Trinity v.2.12.0 (68), using default parameters for each sample and individual. Finally, reads generated from all samples and for each individual were pooled to perform a global transcriptome using the same methodology. RNA-Seq data used for de novo transcriptome assemblies are available in the Sequence Read Archive database, PRJNA120771.

The quality of the assembled transcriptomes was assessed using different approaches. First, the RNA-Seq read representation of the assembly was evaluated by computing the percentage of aligning reads to the assembled transcriptome using bowtie2 software v2.4.2 (69). Second, we examined the representation of full-length reconstructed protein-coding genes by comparing the assembled transcripts against the publicly available reference proteomes of Green Anole (Anolis carolinensis) (AnoCar2.0; GCA_000090745.2), American alligator (Alligator mississippiensis) (ASM28112v4; GCA_000281125.4), Goodes thornscrub tortoise (Gopherus evgoodei) (rGopEvg1_v1.P; GCA_007399415.1), P. guttatus (UNIGE_PanGut_3.0; GCA_001185365.2), using local BLAST + software 2.12.0, with an e-value threshold of 1e-20 and -max_target_seqs = 1 (70). Additionally, transcriptome assembly validation was performed using the Benchmarking Universal Single-Copy Orthologs (BUSCO) v.5.2.2. sauropsida_db10 database to assess the completeness of the assembled transcriptome. Finally, basic summary metrics (eg, N50) were computed using Trinity and rnaQUAST software v2.2.0 (71).

Blast search (BLAST + local software 2.12.0) using the FSHR protein sequences of A. carolinensis (XP_003216196.1), G. evgoodei (XP_030410167.1), P. guttatus (XP_034283311.1), and Crocodylus porosus (XP_019404883.1) against the assembled transcriptomes of C. calyptratus, T. hermanni, P. guttatus, and C. niloticus, respectively, was performed to identify and characterize the FSHR transcripts and protein sequences. We used the same method to identify the transcripts of other genes involved in testis and ovary development and function.

For the LHR search in the assembled transcriptomes, LHR protein sequences of 7 different species of phylogenetic diversity were compared to the assembled transcriptomes of C. calyptratus, T. hermanni, P. guttatus, and C. niloticus for significant alignments using BLAST 2.12.0 + local software tBLASTn tool: mammals (Homo sapiens: AAB19917.2; Bos taurus NP_776806.1), birds (Struthio camelus: XP_009688469.1; Gallus gallus: BAA23736.1; Aptenodytes forsteri: XP_019325855.1), amphibia (Xenopus laevis: NP_001243190.1), and teleost fish (Danio rerio: AAR84281.1).

Identical methodology was used to identify and characterize the common α subunit and ß-specific subunits of C. calyptratus FSH and LH gonadotropins sequences (ccalyFSH, ccalyLH), using the ortholog sequences of α glycoprotein subunit (XP_003215586.1), FSH-specific ß-subunit (XP_003214730.1), and LH-specific ß-subunit (XP_003222770.2) of A. carolinensis. The resulting sequences served to confirm the ccalyFSH and ccalyLH sequences obtained by cloning.

Cloning of GtH (FSH and LH) Genes From the Veiled Chameleon

Pituitary tissue was collected from a female veiled chameleon in RNAlater and stored at 4 °C [Supplemental Table S1 (64)]. Total RNA was extracted from the the pituary gland using the RNeasy kit (74104, Qiagen). The RNA was reverse transcribed by the SuperScript III First-Strand Synthesis System for RT-PCR (18080051; Thermo Fisher Scientific). LH and FSH sequences were amplified by PCR on pituitary cDNA using different sets of degenerate primers designed on conserved, lowly degenerated amino acid codons. PCR products of the predicted length were ligated into the pCRII vector (Thermo Fisher Scientific) and sequenced. Based on these results, gene-specific primers were designed to carry out 3′ rapid amplification of cDNA ends. 3′ rapid amplification of cDNA ends products were sequenced and overlapping sequences assembled in the SeqMan module of Lasergene (DNAStar). At least 3 independently amplified cDNA fragments were sequenced for each nucleotide position. Primers used for amplification are shown in the Supplemental Material [Supplemental Table S2 (64)]. Sequences of the cloned ccalyGtHs were confirmed by comparison with the C. calyptratus transciptome sequence, revealing >99% identity.

Recombinant GtH Production

Single-chain recombinant FSH and LH were produced by Rara Avis Biotec S.L. as descibed previously (72-74) for representative species of 3 nonavian sauropsidan suborders (C. calyptratus, Caretta caretta, Alligator mississippiensis), the avian (Gallus gallus), and the mammalian order (Bos taurus). The cDNA sequences for respective FSH and LH were retrieved from GenBank (B. taurus glycoprotein α subunit NP_776326.1, FSH-β subunit NP_776485.1, LH-β subunit AAB59267.1; G. gallus glycoprotein α subunit NP_001264950.1, FSH-β subunit NP_989588.1, LH-β subunit ADY03193.1; C. caretta glycoprotein α subunit XP_048701022.1, FSH-β subunit XP_048710929.1, LH-β subunit XP_048689061.1; A. mississippiensis glycoprotein α subunit XP_019352338.1, FSH-β subunit NP_001274534.1, LH-β subunit BAJ14507.1). For C. calpytratrus, sequences for the glycoprotein α subunit, FSH-β subunit, and LH-β subunit were obtained from cloning pituitary tissue [Supplemental Table S1 (64)].

Briefly, Chinese hamster ovary cells in suspension were cultured in serum-free medium and transiently transfected with expression constructs encoding fusion proteins containing the entire coding sequences of FSH-β or LH-β subunits, 6 histidine residues, the 28 carboxy terminal amino acids of the human chorionic gonadotropin β subunit as a linker, and the mature sequence of the glycoprotein hormone α subunit.

After 120 hours, the cell culture was centrifuged and the supernatant was collected. Gonadotropins were quantified using an indirect ELISA developed by Rara Avis Biotec, utilizing a polyclonal antibody against the β-subunit of hCG (Thermo Fisher Scientific cat# PA5116189, RRID:AB_2900823) as the primary antibody and a polyclonal anti-rabbit antibody as the secondary (Sigma Aldrich cat# A0545, RRID:AB_257896). Briefly, plates were coated with serial dilutions (from 1:80 to 1:1280) of supernatants in PBS (pH 7.4) and incubated overnight at 4 °C (100 µL/well). The plate was then blocked with 2% BSA (200 µL/well) in PBS with 0.05% Tween-20. The primary antibody was diluted to 1:3000 in ELISA buffer (0.5% BSA in PBS with 0.05% Tween-20) and incubated overnight with shaking at 4 °C. The secondary antibody was incubated for 2 hours at 37 °C at a final 1:2000 dilution. A recombinant bovine FSH purified by metal affinity chromatography, quantified by Bradford assay and assessed for purity (>90%) by SDS-PAGE, was used as standard for this indirect ELISA. The yields of gonadotropin expression ranged from 2 ug/mL to 30 ug/mL.

In parallel, the presence of recombinant GtHs in supernatants was demonstrated by Western blot using an anti β-hCG 1:1000 (RRID:AB_2900823) as the primary antibody and anti-rabbit IgG-horseradish peroxidase (RRID:AB_257896) conjugated 1:2000 as secondary antibody [Supplemental Fig. S1 (64)].

In Vitro Bioactivity of Reptilian, Avian, and Mammalian GtHRs to Homo- and Heterologous GtHs

In vitro cultured mammalian cells expressing the GtHRs of different vertebrate species show an increase in intracellular cAMP after incubation with GtHs (25, 75, 76). The HEK293T cell line was used for in vitro bioassays based on a luciferase reporter gene assay. Cells were grown in DMEM supplemented with penicillin/streptomycin, 2 mM glutamine, and 10% fetal bovine serum and were incubated at 37 °C in 5% CO₂ and 75% relative humidity.

GtHRs of 3 nonavian sauropsidan species (C. calyptratus, C. caretta, A. mississippiensis), 1 bird (G. gallus), and 1 mammal (B. taurus) were individually subcloned in the pCDNA3.1 vector. The cDNA sequences for respective GtHRs were retrieved from GenBank (B. taurus FSHR NP_776486.1, LHR AAC24012.1; G. gallus FSHR NP_990410.2, LHR BAA23736.1; C. caretta GtHR XP_048698888.1; A. mississippiensis GtHR XP_006258435.1), except for C. calyptratus, for which the GtHR sequence was retrieved from the transcriptome analysis of gonadal tissue [Supplemental Table S1 (64)].

Gonadotropin receptor in vitro bioassay methodology is based on previously described methods (25) with minor modifications. Briefly, HEK293T cells were transiently transfected with expression constructs encoding for 1 of the gonadotropin receptors, cAMP-responsive reporter gene vector (pCRE-Luc; Agilent Technologies), and β-gal vector (Promega Corp.) for luminescence normalization. Transfected cells were seeded in 48-well plates and incubated at 37 °C under 5% CO₂ and 75% relative humidity. At 48 hours posttransfection, cells were incubated with different concentrations of GtHs for 5 hours. All treatments were done in duplicates, including negative controls. Cells were lysed and tested for luciferase and β-galactosidase activities by luminescence or colorimetric detection, respectively. Luminiscence data was recorded in duplicate for each biological replicate and then normalized by β-gal results, obtaining relative luminiscence units. The mean and SD of each relative luminiscence unit were computed for each condition using the technical and biological replicates. Dose-response data were normalized to the maximal activity of each receptor with its own specific GtH, which was set as 100%. EC₅₀ values were determined from concentration-response curves for each GtH with its own specific GtHR. At least 3 independent experiments were performed for each condition (GtHs-GtHRs).

Results

Genomic and Transcriptomic Analyses Support the Absence of a Functional LHR Gene in Reptiles

Genomic sequences of FSHR have been identified and annotated in all nonavian sauropsidan suborders (47-50), but despite the availability of high-quality reptilian genomes assembled at the chromosome level, no lepidosaurian, chelonian, or crocodylian LHR genes have been annotated so far (51). Indeed, in the Ensembl database [release 113; October 2024; (77)], as well as in the InParanoiDB database [release 9 (78)], FSHR orthologs to the human receptor (ENSG00000170820) were identified with high confidence in the genomes of chelonia, crocodiles, lizards, snakes, and the tuatara (Sphenodon punctatus), the single extant species of the sauropsidan order Rhynchocephalia. On the other hand, no lepidosaurian, chelonian, or crocodylian orthologs of the human LHR (LHCGR; ENSG00000138039) were found [Supplemental Table S3 (64)]. Similarly, in the FSHR gene, orthologs were identified in 1 chelonia, 1 crocodile, 2 lizards, and 1 snake, all with a high confidence score but none for LHR [Supplemental Table S4 (64)].

As in mammals LHR and FSHR are always found in tandem arrangement, we performed a synteny analysis of the vicinity of the FSHR and LHR genes in selected vertebrate species to investigate whether chromosomal rearrangements could be the reason for the absence of LHR in reptiles. In the human genome, FSHR and LHR are successively situated on chromosome 2 and are flanked by: a general transcription factor (GTF2A1L), stonin (STON1), the phosphoprotein phosphatase (PPP1R21) gene, and 2 forkhead box protein genes (FOXN2, FBXO11) at the 5′ and RNXN1 at the 3′ (Fig. 1). This gene arrangement is hereafter referred to as the “main block.” Among placentals (Eutheria), the synteny of an extended genomic region around the main block is conserved, whereas marsupials and egg-laying mammals share the inversion of this extended region, leaving the gene's relative position within unchanged. We also found a perfect syntenic conservation of the extended region in the clawed frog Xenopus tropicalis, but the main block is inverted in the Leishan spiny toad Leptobrachium leishanense. Genomic inversions and/or deletions of the extended region are observed in sauropsidans (birds and nonavian reptiles), yet the main block is conserved in most species. Exceptions are (1) the chicken, where several genes usually not present in the main block have been inserted between STON1 and FBXO11, and (2) the anole lizard, where we find a large genomic insertion separating STON1 and GTF2A1L.

All examined lepidosaurian, chelonian, or crocodylian species lacked a functional LHR gene at the expected location despite the conservation of the main block across vertebrates. We screened the genomic sequences between FSHR and GTF2A1L in publicly available genomes for possible traces of LHR using tBLASTx and the chicken LHR (NM_204936.2) as a template. While no traces were detected in lepidosauria, several hits were found in the crocodile Crocodylus porosus and the chelonian Caretta caretta genomes (Fig. 2). However, the identified sequences do not cover all exons, and several frame shift and nonsense mutations are present [Supplemental Table S5 (64)].

Figure 2.

Residual exons of LHR in a crocodilian and a chelonian species. Using tBLASTx and the chicken LHR, we identified residual LHR exons in the genomic sequence between FSHR and GTF2A1L in (A) the crocodile Crocodylus porosus and (B) the chelonian Caretta caretta genomes. The interrogated genomic region is highlighted in light red on the y-axis and the exons of the chicken LHR are depicted in different shaded boxes along the x-axis. The BLAST hits are represented as horizontal blue bars in the graph. In the crocodile genome, exons 1, 3, and 9 are missing (Xs) and a sequence inversion occurs between exons 10 and 11 (blue area of the y-axis). The chelonian genome lacks homologous sequences to exons 1, 3, 4, 5, 6, and 10 (Xs). A protein sequence alignment of the corresponding hits is provided. Note that several frame shift (not indicated) and nonsense mutations (red asterisks in the protein sequences) are present in the chelonian and crocodylian sequence.

To further substantiate our claim that squamate reptiles (lizards, snakes), chelonians, and crocodiles indeed lack a functional LHR gene, we performed transcriptomic analyses on gonadal tissues of the veiled chameleon (Chamaeleo calyptratus), the corn snake (Pantherophis guttatus), the Greek tortoise (Testudo hermanni), and the Nile crocodile (Crocodylus niloticus). We produced de novo transcriptome assemblies by combining the datasets of each species [Supplemental Tables S6 and S7 (64)]. These comprised a high percentage of the sauropsida BUSCO genes (78-88%), either complete or fragmented [Supplemental Fig. S2 (64)], and high protein sequence coverage when compared to publicly available protein datasets from the same or closely related species [Supplemental Fig. S3 (64)]. This comparison confirmed the presence of FSHR sequences in all species analyzed with an identity greater than 88% [Supplemental Table S8 (64)]. All pairwise alignments were significant (e-value <1e-20) with a protein coverage percentage higher than 99%. Furthermore, we successfully identified key genes involved in testis and ovary development and function such as DAZL, SPATA16, CLGN, ZMYND15, GDF9, AMH, and STAG3 and various aromatases, including CYP27A1, CYP46A1, or CYP19A1, further supporting the completeness of these transcriptomes [Supplemental Table S9 (64)]. As the veiled chameleon pituitary transcriptome was included in the assembly, we also annotated the veiled chameleon's FSH and LH, of which the deduced protein sequences were used in subsequent bioassays.

Due to the lack of published lepidosaurian, chelonian, or crocodylian LHR sequences, we queried our transcriptomes using LHR transcripts and proteins from mammals (Homo sapiens, Mus musculus), birds (Gallus gallus, Struthio camelus, Aptenodytes forsteri), amphibians (Xenopus laevis), and teleost fish (Danio rerio). In all cases, the sequence showing the highest similarity was the previously identified FSHR from each of the species under study. The second best hit was TSHR for the veiled chameleon and the Greek tortoise and a leucine-rich G protein-coupled receptor with a sequence identity of less than 26% for the corn snake and the Nile crocodile. Therefore, no sequences or transcripts compatible with an LHR were identified within the gonad-specific transcriptomes of the veiled chameleon, the corn snake, the Greek tortoise, and the Nile crocodile.

Overall, our results strongly support that sauropsids, with the exception of avians, constitute the only vertebrate group to show functional reproduction with 2 GtHs but only 1 GtHR. The deletion of the LHR gene probably occurred independently 3 times: in testudines, in lepidosauria affecting the squamate reptiles and tuataras, and in the archosauria clade, affecting crocodylia but sparing the avian class (79) (Fig. 3).

Figure 3.

Evolutionary scenario of gonadotropin hormones and receptors in tetrapoda. Two gonadotropin receptors, FSHR and LHR, are present in the amphibian and mammalian class. In the sauropsidans, inactivation of the LHR gene happened in 3 different lineages: the lepidosauria, testudines, and crocodylia but not in the aves. The ligand duplicity, FSH and LH, was nevertheless retained in all sauropsidans, yet with comparatively decreasing ability of the FSH to stimulate the remaining FSHR in the testudines and crocodylia. Blue thermometer = poikilothermy, red thermometer = homeothermy.

Ligand Specificity and Promiscuity of Reptilian GtHRs: Not All Reptiles Are Alike

In order to elucidate the response of the remaining reptilian FSHR to their homologous GtHs and to evaluate the evolutionary adaptation of the reptilian ligands and receptors, we performed functional bioassays. First, expression constructs for the single functional GtHR and both hormones of the veiled chameleon, loggerhead sea turtle (Caretta caretta), and American alligator (Alligator mississippiensis) were generated, and species-specific receptor hormone binding was tested. Then, to evaluate the adapation of the reptilian ligands and receptors during the divergence of the taxonomic groups, we used cross-over heterologous bioassays between the squamate, chelonian, and crocodilian species and tested the reptilian GtHs and GtHRs against the chicken (Gallus gallus) and cow (Bos taurus) double duality 2GtHR/2GtH systems. We selected these species because of the public availability of curated coding sequences for GtHs and GtHRs, except for the veiled chameleon, of which the sequences stemmed from our own research. In all assays, relative effects were measured based on the hormone concentrations necessary to achieve 50% (EC₅₀) of the maximum signal triggered by the receptor's cognate ligand [Supplemental Table S10 (64)].

In vitro bioactivity of reptilian, avian, and mammalian GtHs with their homologous receptor

In in vitro bioassays, the chameleon FSHR showed the highest degree of promiscuity with almost identical activation by both chameleon FSH and LH (Fig. 4A). The turtle FSHR also showed promiscuity for both its ligands but reacted more strongly with LH than with FSH (Fig. 4B). Interestingly, the single alligator GtHR, identified as FSHR by sequence, did not recognize the cognate FSH but strongly reacted with LH (Fig. 4C). In chicken, our avian representative, FSH showed strong stimulation of FSHR, but the receptor also showed slight promiscuity for LH (Fig. 4D1), whereas chicken LHR was stricly specific for LH (Fig. 4D2), as suggested in previous studies (80) and conforming with the pattern seen in amphibians (35, 39) and teleost fish (57, 59). For the bovine GtHRs, our results confirm previous reports about the high specificity of mammalian GtHRs for their respective ligands (Fig. 4E1 and 4E2).

Figure 4.

In vitro bioassays, activation of gonadotropin hormone receptors of selected reptilian (A, B, C), avian (D1, D2), and mammalian (E1, E2) species with homologous gonadotropin hormones. Grey shaded areas represent unphysiological hormone concentrations (>200 ng/mL), and the horizontal red line equals 50% receptor activation (EC₅₀). Abbreviations: am, crocodilian (A. mississippiensis); bt, bovine (B. taurus); ccaly, squamate (C. calyptratus); ccaret, chelonian (C. caretta); FSHR, FSH receptor; gg, avian (G. gallus); LHR, LH receptor.

In vitro bioactivity of reptilian, avian, and mammalian GtHRs with heterologous GtHs

To gain further insight into the evolutionary adaptations of reptilian GtHs and GtHRs, comparative heterologous bioactivity assays with squamate, chelonian, crocodylian, avian, and mammalian GtHRs and GtHs were performed. Prediction of the biological effect based on in vitro activation of receptors in a heterologous system is not possible. We therefore refrained from defining “activity” of heterologous hormones but focused our interpretation on the degree of receptor stimulation within a presumably physiological range of hormone concentration below 200 ng/mL and indicated the EC₅₀ concentration of the cognate ligand as a reference concentration.

The squamate chameleon FSHR was the most receptive receptor tested, being stimulated by all heterologous hormones sometimes at even lower concentrations than by its own species-specific GtHs (Figs. 5A1-3). Notably, the chameleon FSH and LH only interacted with the chameleon FSHR but were not able to stimulate receptors of other species (Figs. 5B1 and 5C1 and 5D1-2 and 5E1-2). The turtle FSHR appeared to be generally receptive to heterologous stimulation by both FSHs and LHs (Figs. 5B1-3), except for bovine LH and the 2 chameleon GtHs. Matching the homologous bioassay, the alligator FSHR only reacted to LHs of all taxons, except the chameleon LH (Figs. 5C1-3). In parallel, the loss of function of the alligator FSH appears to have challenged alligator LH to become the most pluripotent GtH investigated in this study, acting on almost all FSHRs and LHRs (Figs. 5A1, 5B1, 5D1-2, and 5E1), except for the bovine LHR (Fig. 5E2). Despite its lack of reactivity in its own species-specific system, alligator FSH was still recognized as FSH by heterologous FSHRs of the chameleon, turtle, chicken, and bovine (Figs. 5A1, 5B1, 5D1, and 5E1).

Figure 5.

In vitro bioassays, activation of gonadotropin hormone receptors of a selected reptilian (A1-3, B1-3, C1-3), avian (D1-4), and mammalian (E1–E4) species with heterologous gonadotropin hormones. Grey shaded areas represent unphysiological hormone concentrations (>200 ng/mL), and the horizontal red line equals 50% receptor activation (EC₅₀). Abbreviations: am, crocodilian (A. mississippiensis); bt, bovine (B. taurus); ccaly, squamate (C. calyptratus); ccaret, chelonian (C. caretta); FSHR, FSH receptor; gg, avian (G. gallus); LHR, LH receptor.

Chicken FSHR was activated by all heterologous hormones, except for chameleon GtHs and bovine LH, with the strongest reaction to the bovine FSH and both alligator GtHs (Figs. 5D1 and 5D3). On the other hand, chicken LHR strongly reacted with all LH of other species (except chameleon) but less so with FSHs (Figs. 5D2 and 5D4). The bovine FSHR was stimulated by both turtle GtHs and both alligator GtHs and to a lower degree by both chicken GtHs (Figs. 5E1 and 5E3). Thus, turtle and crocodilian LH show a considerable mammalian and avian FSHR activation. The bovine LHR is activated only by its cognate species-specific ligand LH and is the most specific receptor of all GtHRs investigated in this study (Figs. 5B2 and 5E4).

Discussion

Early research on the evolution and reactivity of nonavian sauropsidan GtHs and GtHRs was inconclusive (3). In our study, genomic and transcriptomic analyses of selected tetrapod species now strongly suggest the inactivation of the LHR in all nonavian sauropsidans. Species-specific homologous in vitro bioassays indicate that the remaining FSHR is equally sensitive to both GtHs in squamates, more sensitive to LH than FSH in testudinians, and only sensitive to LH in crocodylians, suggesting that the alligator FSHR underwent a functional change to act like an LHR-like receptor. We also systematically tested the reactivity of the GtHs and GtHRs of nonavian and avian sauropsidan and mammalian species in heterologous in vitro bioassays. Comparison to the early research is limited, as those studies did not work with the exact same species and, due to the lack of molecular technology, measured biological effects (such as steroid production in gonadal cell cultures and gonadal growth and development) in vivo or in gonadal tissue homogenates are prone to influence from many other factors (eg, season, age, hormone concentration and administration, and endogenous hormone production). Yet, the obtained molecular results are helpful to scrutinize the biological observations of the earlier studies and might give hints to potential evolutionary changes in the protein structures. The fact that early studies found a strong biological effect of mammalian FSH (including ovulation) in squamate reptiles is comprehensible based on our findings; however, the comparatively minimal biological effect of mammalian LH in squamate reptiles (10, 11, 13, 14, 17, 37, 81, 82) stands in disagreement with our results of a squamate FSHR equally sensitive to bovine FSH and LH. In contrast, the stronger effect of mammalian FSH than LH in in vivo studies or in vitro in gonadal tissue homogenates in chelonians (15, 16, 37, 82-86) was confirmed by our findings with a turtle FSHR highly sensitive to bovine FSH but not to bovine LH. Similarly, the early observations that chelonians were reactive to avian FSH (82, 85, 87) and in higher concentrations also to avian LH (82, 84, 85, 88) can be explained by our results. The purely LH-sensitive alligator FSHR explains the observations of an early study whereby mammalian LH had a stronger effect on crocodilian gonadal tissue homogenates than FSH (83) but contradicts another in vitro study that found stronger testosterone production in alligator testicular cell cultures in response to ovine FSH than to ovine LH (82).

The unexpected findings of the loss of the LHR in reptiles and the variable sensitivities and promiscuity of the remaining reptilian FSHR raise new, intriguing questions about the endocrinologic regulation of reptilian reproduction. Based on our observations, we propose a modified paradigm with a neuroendocrine control of reproduction in reptilian species by only 1 GtHR and either 1 GtH in crocodylians or 2 GtHs in chelonians and squamate reptiles. We hypothesize that in reptiles, the loss of a functional LHR was compensated by the expression of the remaining FSHR in both Sertoli and Leydig cells in males and granulosa and theca cells in females (49). Under the tested conditions, in chelonians and squamate reptilians, this receptor is activated by both FSH and LH, whereas in crocodylians, only LH is capable of stimulating the receptor; nevertheless, full gonadotropic function is granted (Fig. 6).

Figure 6.

Proposed models for the gonadotropin regulation in reptiles as compared to fish, birds, and mammals (styled after 89). Abbreviations: FSHR, FSH receptor; LHR, LH receptor.

From an evolutionary viewpoint, the increasing numbers of GtHs and GtHRs during vertebrate evolution reflect the traditional notion of increasing complexity during evolution (90). Redundant paralog GtHR genes from gene duplication events were occasionally lost, as happened in the fish clade with the third and fourth LHR, but the loss of the single reptilian LHR appears to constitute an example of reductive evolution (91) that is limited to nonavian sauropsidans but happened independently 3 times. It is noteworthy that the avian clade has retained its double duality 2GtH/2GtHR system and developed differently from its sauropsian sister branches. Allocating these losses to parallel abrupt mutational events (eg, by a transposable element) appears unlikely, which leads to the process of elimination of a gene that proved dispensible in this specific group of vertebrates. Functional redundancy of paralog promiscuous genes serves as the most likely reason for the dispensibility in reductive evolution (91). Reflections about the evolutionary drivers for this development remain purely hypothetical at this point. While the maintenance of nonfunctionalized paralogs over a long period of time after gene duplication events could have been in response to initial dosage-balance constraints (92), it may have eventually allowed nonavian sauripsidans to use gene loss as a tool to adapt to selective pressures (“less is more” hypothesis) at a later point in evolution, owing to massive shifts in environmental conditions (93). Such a massive shift could have been presented by the transfer from aquatic to terrestrial reproduction of the amniota clade, which the nonavian reptiles had to manage without the benefits of heterothermic thermoregulation, a common characteristic of the avian and mammalian clades (Fig. 3). Intuitively, the promiscuity of the reptilian FSHR might be interpreted as an ancestral trait of the receptor, and the benefits of simplifying the endocrinological regulation of reproduction to adapt to environmental pressure remain elusive; however, phylogeny of reptilian FSH and LH has demonstrated a clear differentiation into 2 different hormones. The promiscuity of the reptilian FSHR must have developed in mutual adaptation to the differentiation of the 2 ligands (94), and the reptilian patterns of GtH/GtHR systems are likely evidence for evolution in progress with the crocodylians most advanced in the process of simplifying the endocrine regulation of nonavian reptilian reproduction.

From a functional standpoint, the question remains how 2 (or in the case of crocodylians only 1) hormones and 1 receptor can control different functions in temporally concerted action (recruitment, maturation, and release of gametes). We propose the 3 following scenarios: (1) additive effects of the hormones and dose-dependent biological processes, (2) tightly regulated temporal and spatial expression of the remaining FSHR in different gonadal somatic cells, and (3) temperature-dependent functions of the only GtHR. In the case of 2 functional GtHs, as in chelonians and squamates, a continuous secretion of FSH may support maturation of gametes at lower levels of GtHR stimulation, analogous to the mammalian system, whereas ovulation may be triggered by the pulsatile secretion of LH, resulting in a peak activation of the GtHR. Alternatively, the different functions may be governed by distinct temporal expression patterns of GtHR in the different gonadal somatic cells (granulosa/Sertoli vs theca/Leydig cells). The latter appears to be the more likely scenario in the case of crocrodylians with only 1 functional GtH but leaves the question about the potential function of the apparently nonrecognized FSH in crocodiles unanswered. The third potential scenario proposes a temperature-dependent function of receptors: all vertebrate taxa with promiscuous GtHR (actinopterygian fish, amphibia, lepidosauria, testudines, crocodylia) show poikilothermy with higher dependency on ambient temperatures (Fig. 3). Since GtHRs were expressed in mammalian cells in this study, binding assays were performed at optimal mammalian temperatures of 37 °C. It is theoretically possible that promiscuity of the reptilian FSHRs changes with temperature to the point that even crocodilian FSH may be recognized by its cognate receptor within a certain temperature range. In comparative studies with mammalian, avian, amphibian, and nonavian sauropsidan gonadal tissue homogenates, it was shown that the temperature of optimal binding capacities of FSHR coincided well with the temperatures at which GtHs act in the species under natural conditions. For example, FSHR of amphibians showed higher binding capacities for FSH than that of homeotherm avian and mammalian species at low temperatures, and the FSHR of reptiles showed high binding capacities across a wide range of temperatures (34, 95-98). This must clearly be considered an impediment to the interpretation of our results. However, no studies investigated the impact of temperature on the degree of promiscuity or relative sensitivity for different GtH of GtHRs, which may have to be considered as a potential way of regulating endocrine control of reproduction in poikilotherm species. In theory, this could explain why differences in environmental temperature magnitudes and fluctuations might be linked to reproductive failure in nonnatural environments (99).

This study has raised many questions for future research, such as the evolution of the reptilian genomes, protein conformation and binding site interactions in nonavian sauropsida compared to other taxa, the expression patterns and function of the reptilian GtHR receptor in various gonadal and nongonadal tissues, the role of the apparently nonfunctional FSH in crocodylians, and the temperature dependency of hormone-receptor interactions in nonavian sauropsidan species.

Funding

This project was funded by the Foundation for Research in Science and the Humanities at the University of Zurich (STWF-20-010) and by the Wolfermann-Nägeli-Stiftung.

Author Contributions

Contributions were as follows: experimental design: M.K., J.R.G., I.G., A.T., M.C., M.G.; data collection: J.R.G., I.G., A.T., M.G.; data analysis: M.K., J.R.G., I.G., M.G.; manuscript draft: M.K., J.R.G., I.G., A.T., M.C., S.N., J.M.H., M.G. All authors provided intellectual content and editing for the final manuscript and agree on its publication.

Disclosures

The authors have nothing to disclose and no conflicts of interest.

Data Availability

Original data generated and analyzed during this study are included in this published article or in the data repositories listed in References.