Landscape of snake’ sex chromosomes evolution spanning 85 MYR reveals ancestry of sequences despite distinct evolutionary trajectories

Most of snakes exhibit a ZZ/ZW sex chromosome system, with different stages of degeneration. However, undifferentiated sex chromosomes and unique Y sex-linked markers, suggest that an XY system has also evolved in ancestral lineages. Comparative cytogenetic mappings revealed that several genes share ancestry among X, Y and Z chromosomes, implying that XY and ZW may have undergone transitions during serpent’s evolution. In this study, we performed a comparative cytogenetic analysis to identify homologies of sex chromosomes across ancestral (Henophidia) and more recent (Caenophidia) snakes. Our analysis suggests that, despite ~ 85 myr of independent evolution, henophidians and caenophidians retained conserved synteny over much of their genomes. However, our findings allowed us to discover that ancestral and recent lineages of snakes do not share the same sex chromosome and followed distinct pathways for sex chromosomes evolution.

www.nature.com/scientificreports/ least two transitions involving XY and ZW and independent turnovers of these sex chromosome systems may have occurred in Pythonoidea and Booidea superfamilies. The gene content of both the Z and the W chromosomes are thought to be relatively conserved in snakes 24,27,28 . Still, the high variability of W chromosome (regarding morphology and/or gene content) in major clades suggests a remarkable role of repetitive sequences accumulation in their architecture and evolution 24,25,29 . Unlike the W, the Z chromosomes are thought to be more stable among Serpentes lineages, and non-drastic shifts in morphology have been reported in different snake clades [23][24][25]30 .
Although suggested as having an independent origin, the homomorphic XX/XY chromosomes in Boa and Python are yet to be characterized by molecular cytogenetic techniques. Thus, their relationships and transitions among both homomorphic and heteromorphic ZZ/ZW in Pythonoidea and Booidea superfamilies [31][32][33] still remain unanswered. In the amazonian red-tailed Boa constrictor (formally Boa constrictor constrictor), which has 2n = 36 chromosomes, the fourth chromosomal pair is thought to represent the putative sex pair, which would be a typical feature for henophidians (i.e. a former superfamily of the suborder Serpentes, which harbors boas, pythons and other old lineages of snakes, usually referred as "Primitive Snakes") 31 . In addition, two different classes of sequences, the PBI-MspI and EQU-BamHI-4 (EQU-BamHI-4 being putatively reported as sex-linked 27,35 ) have been identified on the 4th homomorphic pair of Boa constrictor females, therefore, being identified as the sex pair in these studies 34,35 . Similarly, in most of the pit vipers, rattlesnakes and colubrids, the 4th chromosomal pair also represent the sex chromosomes 24,25,[36][37][38] , even when homomorphic, as already identified through accumulation of Bkm repeats in some species 39,40 , suggesting a conserved trend for sex chromosome evolution in Snakes.
Homomorphic sex chromosomes are frequently observed among non-avian reptiles. In the Serpentes suborder, for instance, they are found in major clades of henophidian and caenophidian species (i.e. Caenophidia is a monophyletic group that contains over 80% of all the extant species of snakes, commonly referred as "Advanced Snakes") 30,31,39,41 . On the other hand, well-differentiated sex chromosomes are more common in the more recently diversified groups of snakes, the advanced lineages 25 , but also present in the former groups of snakes as Typhlopoidea and Booidea 32,42 . The molecular and cytogenetic mechanisms of evolution of homomorphic sex chromosomes in snakes have not been the subject of rigorous studies compared to well-differentiated ones. Therefore, many homomorphic or micro W or Y sex chromosomes in the ancestral lineages of snakes remained undetected. Serpentes was thought to have a well-stable sex chromosome system and, despite different levels of W degeneration, as well the occurrence of multiple sex chromosomes (e.g. Z1Z2/Z1Z2W and Z/W1W2), only ZW system had been described until recently 43,44 . Indeed, the arise of a putative and independently evolved XY sex chromosome system in Boidae and Phytonidae raised questions regarding the cytogenetic and molecular mechanisms involving the evolution of sex chromosomes in snakes. Why did snakes independently evolve a new and homomorphic sex chromosome system solely in ancestral lineages? Did snakes retain homology of XY and ZW chromosomes along the Serpentes' evolution owing to some evolutionary advantage conferred by the shift and transitions between these systems? Why has the XY sex chromosome system been lost in the more advanced lineages?
Application of molecular cytogenetic tools, such as chromosomal mapping using Bacterial Artificial Chromosomes (BAC-FISH) and comparative genomic hybridization (CGH) have been instrumental in overcoming limitations in identification of undifferentiated or cryptic sex chromosomes in several vertebrate groups such as fishes [45][46][47] , amphibians 48,49 , and reptiles 20,50,51 . In this study, we aimed toward understanding the relationship between the homomorphic XY and heteromorphic ZW chromosomes found in some ancestral (henophidians) and more recent (caenophidians) snakes. For that, we performed an extensive comparative analysis of the amazonian red-tailed boa (Boa constrictor constrictor) chromosomes (homomorphic XY) through cross-species comparisons using whole genomic DNA (gDNA) from several caenophidian species with varying degrees of the ZW sex chromosomes differentiation. We also performed WCP (Whole Chromosomal Painting) of a highly degenerated W sex chromosome and mapped BACs specific for several genes on Boa constrictor chromosomes. We identified chromosome homologies of sequences for all analyzed species, however, with different patterns of accumulation, which enabled us to infer the relationship and landscape of snake' sex chromosomes evolution spanning 85MYR.

Results
Chromosome painting with amazonian pit-viper (Bothrops atrox) W paints and cross-species chromosome painting to Boa constrictor chromosomes. The isolated W chromosome probe of B. atrox (BaW) was amplified and the homology was tested onto metaphase spreads of the same species (Fig. 1). We carried out cross-species chromosome painting using BaW probe to metaphase spreads of male and female of B. constrictor in order to test for the homology of the sex chromosomes between Henophidia and Caenophidia. The BaW probe hybridized completely on a small metacentric (W chromosome/pair 4) of Bothrops atrox (Fig. 1). However, in male and female B. constrictor metaphase spreads the BaW probe showed faint hybridization signals on microchromosomes and on the centromeric position of the 7th pair ( Fig. 1), with no differences between males and females.
Comparative genomic hybridization. Comparison between male and female gDNAs of Boa constrictor (Fig. 2), produced intense and faint hybridization signals on macrochromosomes and microchromosomes, respectively (Fig. 3), co-located with C-banded regions as previously reported 31 . Small-shared signals were observed on the pericentric regions of the p arms of chromosome pair 1, whereas strong bright signals were observed on the centromere of chromosome pairs 2 and 4 and on the telomeric position of chromosome pair 7. www.nature.com/scientificreports/ Male or female-specific hybridization signals were neither detected in B. constrictor male and female nor on the homomorphic sex chromosomes (putatively the chromosome pair 4) (Fig. 3). Interspecific hybridization among henophidian (Boa constrictor-red-tailed boa) (ancestral lineage XY) and caenophidian (advanced lineages ZW) (pitvipers-Bothrops, bushmaster-Lachesis, rattle snakes-Crotalus and puffer snake-Spilotes) species (Fig. 4a, b), revealed that all species share conserved sequences to that of Boa constrictor chromosomes, particularly with the macrochromosomes pairs 1, 4, 7 (Figs. 5a-c, 6a-d). The crossspecies hybridization (Boa constrictor/Bothrops bilineatus) revealed shared sequences on the centromeric position of the 4th and 7th pairs (Fig. 5a). However, the Boa constrictor/Bothrops taeniatus pitviper comparisons showed hybridization signals only on the centromeric position of the 4th pair (Fig. 5b). The Boa constrictor/Bothrops atrox-amazonian pitviper comparisons, on the other hand, showed hybridization signals only on the centromeric position of the 7th pair (Fig. 5c).
Comparisons between Boa constrictor/Lachesis muta revealed shared hybridization signals on the 1st and 7th chromosomal pairs. However, Lachesis hybridization signals on the 1st pair were more intense than Boa constrictor signals (Fig. 6a). The Boa constrictor/Crotalus terrificus showed shared sequences on the 4th and 7th pairs, similar to that of the Boa constrictor/Bothrops bilineatus (Fig. 6b). The Boa constrictor/Crotalus ruruima comparisons showed the same hybridization pattern to that of Boa / Lachesis, with signals only on the 1st and 7th pairs, likewise with more intense signals on the 1st pair (Fig. 6c). Unlike most patterns, Boa constrictor/Spilotes pullatus showed hybridization signals on 3 chromosomal pairs: near the pericentromeric region of the 1st pair (similar to that of the Boa / Lachesis and Boa/C. ruruima) and on the centromere of both 4th and 7th pairs (similar to those of the Boa/B. bilineatus and Boa/C. terrificus) (Fig. 6d). The 2nd pair was the sole representative with strong Boa constrictor specific hybridization signals, but with no shared regions of gDNA with all caenophidian snakes.

Discussion
In reptiles, independent turnovers and transitions among sex chromosomes systems (XY and ZW) and sexdetermining mechanisms (TSD and GSD) within closely related species are more common than previously thought, being thus, a widespread feature among non-avian reptiles 1,3,4,54,55 . Unlike most snakes, Boa imperator was reported to have XY homomorphic sex chromosomes 26 , and its sister species Boa constrictor, shares ancestry presenting also an XY system. Therefore, it is plausible that transitions between homomorphic ZW and XY have occurred in the Boidae family without much substantial genotypic innovation (e.g. considering the 4th pair of boas as the putative sex chromosomes, the XY and ZW are morphologically similar), as reported in the Japanese frog Glandirana rugosa 56 . In our study, we did not detect any sex-specific pattern using intra-and interspecific CGH experiments (Figs. 3, 5, 6), suggesting that only minute sequence differences exist between sex chromosomes (putatively the pair 4). A similar pattern was observed in the Sanziniidae family, a sister group to Boidae (Fig. 8) 32 , but the Z and W chromosomes of Acrantophis sp. cf. dumerili are morphologically well-differentiated nevertheless.
Whilst CGH has been applied for the identification of undifferentiated or cryptic sex chromosomes across a range of vertebrates ranging from fish to reptiles 45,46,50,51,57,58 , this technique, in some cases, may not be efficient in detecting specific sex domains (even in the heteromorphic sex chromosomes) as already seen in amphibians 59 and in the well-differentiated ZW present in Acrantophis sp. cf. dumerili (Booidea) 32 . Perhaps some ancestral lineages still need more time to achieve sex-specific signatures (e.g. morphological changes, accumulation of sequences, heterochromatinization), or simply use alternative mechanisms for sex chromosome evolution, which makes it difficult to detect, especially when they retain huge traits of homology, as here observed in Boa constrictor.
Our comparative cytogenetic analysis suggests that henophidian and caenophidians indeed followed different evolutionary pathways regarding the origin of their sex chromosomes. Several genes share ancestry between putative homomorphic X and Y chromosomes of Python (at that time considered to be Z and W) and the Z chromosomes of caenophidians 27 , suggesting that X, Y and Z chromosomes can easily undergo transitions in ancestral lineages conferred by the similarity of morphology and gene content. For instance, even though located in different positions regarding other snakes' lineages, the genes linked to the putative sex pair of Boa constrictor male and female points homology with the independently evolved putative pair of burmese python (Python bivitattus XY), with the sex pair of habu pit viper (Protobothrops flavoriridis ZW) and the four-lined ratsnake www.nature.com/scientificreports/ (Elaphe quadrivirgata ZW) (Fig. 9). Interestingly but not surprisingly, once sex chromosomes evolve fastly and independently across lineages 4 , the XY of Python bivittatus seems to share more similarities with caenophidians than with the other sole representative XY system existing in Serpentes, the XY present in Boa 26 . Regardless, this shared ancestry, in spite of some fine adjustments in the gene position on the sex pair, indicates that henophidian and caenophidian snakes do not share the same set of sex-determining genes, since other genes located on the putative XY of Boa also share homology with the second pair of Elaphe quadrivirgata (Caenophidia), which partially correspond to the Z chromosome of chicken. Furthermore, the mapping of BaW chromosome probe also provided strong evidence that caenophidian and henophidian snakes do not share the same sex chromosomes, because the W of Bothrops atrox (Caenophidia) has homology with the 7th autosomal pair of Boa constrictor and not with the putative homomorphic sex chromosomes (4th pair) (Fig. 1), which correspond to a well-differentiated ZW system in the sister group (Sanziniidae) 32 . CGH also revealed that, among all the 7 caenophidians www.nature.com/scientificreports/ snakes involved in our comparative study, six of them shared ancestry with the 7th pair of Boa constrictor, that somehow share some degree of homology with the W sex chromosome of B. atrox. Perhaps this 7th pair represent a large conserved segment of the henophidians and caenophidians ancestor. To fully understand the real status of ZW-XY-ZW transitions and homology of sequences, combined whole genome sequencing and refined cytogenetic approaches will be required, especially in representatives from the four major clades of Serpentes suborder (Typhlopoidea ZW, Pythonoidea XY, Booidea ZW/XY, and Colubroidea ZW) (Fig. 8), where the XY sex chromosome system arose only twice and remained morphologically undifferentiated. Notably, our study revealed that the 4th pair of Boa also shares homology with the Z, W and 2nd chromosome pair of chicken (Fig. 9). While chicken's Z partially correspond to the Squamata chromosome 2 2,29,60 , however, at least 2 genes (CHD1, APTX) located on the 4th pair of Boa constrictor also share homology with the second pair of Elaphe quadrivirgata and the bearded dragon (Pogona vitticeps). As hypothesized by Ezaz and colleagues, this synteny among different squamate clades and chicken Z chromosome could represent part of an ancestral super-sex chromosome for Aminiotes 2 . Interestingly, two sex-linked genes in Pogona vitticeps also share homology with the Boa constrictor 2 (BAC containing genes OPRD1 and RCC1 ) and 7 (NR5A2) chromosome pairs (Fig. 10). These genes correspond to the chicken chromosomes 17 and 23 19 . Concordantly, OPRD1 / RCC1 and NR5A2 genes, also mapped in the yellow and green anacondas (Eunectes notaeus and Eunectes murinus), cerrado rainbow boa (Epicrates crassus) and in the amazonian puffer snake (Spilotes pullatus), showed a similar scenario (Viana personal communication). Although the homology of Squamates 2 and chicken Z is considered a conserved trait across lineages 2,12,29, 61-63,76 , the Boa constrictor 2 shares homology to the chicken chromosome 17 and 23, whereas the chicken Z, W and 2 with the putative XY of amazonian red-tailed boa (4th pair) (Figs. 9, 10), which highlights the homology and ancestry of sequences among close and distantly related lineages, possibly remnants of a common evolutionary history among avian and non-avian reptiles.
It is intriguing that after the divergence of Henophidia and Caenophidia in the Upper Cretaceous (~ 85 MYR) 64,65 snakes still share conserved sequences across lineages even after such long period of independent evolution (Fig. 11). Even more puzzling is that some closely related lineages (e.g. C. terrificus and C. ruruima) show a divergent pattern of gDNA hybridization on the Boa constrictor chromosomes (Figs. 6b,c, 11), perhaps  (Figs. 5a-c, 11). This evolutionary landscape might be product of the mechanisms that shape the processes of chromosomal differentiation during the evolution, as for example the association with TEs (Transposable Elements) and SSRs (Simple Short Repeats) sequences, that triggers an important role on the genome architecture leading to independent evolution processes (e.g. silencing, deleting, or increasing genomic regions) 63,[66][67][68] . This seems to be also the case for the snakes here analyzed. Nevertheless, all caenophidians used in our study shared sequences with B. constrictor chromosomes, representing a possible inheritance of ancestry, being the assortment of hybridization patterns due to the tempo of sequence divergence and transient evolutionary mechanisms linked to their evolution spanning ~ 85 my of independent evolution. However, we were not able to identify any sex-specific sequence from all caenophidians gDNA derived probes and W chromosome probes (BaW), that somehow showed the same hybridization pattern in both male and females of Boa constrictor. In fact, this is not surprising because hybridization has not even detected within Boa comparisons, such sex-specific patterns. The shared sequences and different patterns could simply be the result of the convergent accumulation of repetitive sequences during Snakes' evolution. However, all caenophidian species used here share the same W sex chromosome (Viana personal communication).
This lack of sex-specific signals in Boa (XY) from caenophidian (ZW) gDNA derived probe is likely that the sex-linked sequences in advanced snakes are different and, therefore, do not share any similarity with those sexlinked sequences in Boa. This suggests an independent evolution of sex chromosome sequences in snakes, but in caveats, given the similarity of morphologies and gene content of putative sex pair of henophidian and the www.nature.com/scientificreports/ sex chromosomes of caenophidian snakes we cannot conclusively infer which homomorphic system, XY or ZW really occurs in ancestral lineages (Boidae and Pythonidae). Regardless, our study provides first evidence that caenophidian and henophidian snakes have a common evolutionary history but likely evolving a different set of sex-determining sequences, where the sex chromosomes followed divergent evolutionary pathways. However, in henophidians, the real status of homology with the cryptic sex chromosomes of Boa and the heteromorphic ZW present in the sister group (Sanziniidae) is yet to be investigated, which will require developing probes from  www.nature.com/scientificreports/ Y sex-linked markers of Boa imperator for cross species chromosome mapping. Such combined methods of genomics and cytogenetics will enable us to unreveal the dynamic evolutionary history and transitions between XY and ZW sex chromosomes system in the major clades of Serpentes. This study is part of a series of further cytogenetic and genomic studies, focusing on Neotropical reptiles and their hidden evolutionary diversity.

Material and methods
Sampling, mitotic chromosomes preparation, and DnA extraction. Snakes  in the cross-species mapping. Chromosomal preparations were obtained following 69 . The gDNA of males and females for all species were extracted from blood using the Wizard Genomic Purification Kit (Promega), according to the manufacturer's recommendations. We also highlight that in our present study, no animal needed to be euthanized.
Microdissection of W sex chromosome of Bothrops atrox and preparation of the BaW chromosome paints. We performed microdissection using an inverted phase-contrast microscope Zeiss Axiovert.
A1 (Zeiss, Oberkochen, Germany) equipped with Eppendorf TransferMan NK 2 micromanipulator (Eppendorf, Hamburg, Germany). We prepared glass needles from 1.0 mm diameter capillary glass using a glass capillary puller, Sutter P-30 Micropipette Puller (Sutter Instrument, Novato, Calif., USA) and sterilized using ultraviolet irradiation. We microdissected a W chromosome from freshly prepared slides of a female B. atrox using a glass needle and the micromanipulation system, subsequently transferring the W chromosome into 0.2 ml PCR tubes. The W chromosome DNA (BaW) was amplified using GenomePlex Single Cell Whole Genome Amplification Kit (Sigma-Aldrich, St. Louis, Mo., USA) according to the manufacturer's protocol with slight modifications according to 70 . The volume of the reactions was scaled down to half, and the PCR amplification step was increased to 30 cycles. The W chromosome paint of B. atrox was labeled by nick translation means incorporating SpectrumGreen-dUTP (Abbott, North Chicago, Ill., USA). The hybridization was carried out for 1 day in the B. atrox chromosomes (control) and 3 days in cross-species chromosome painting (Boa constrictor male and female).
preparation of probes for cGH. The gDNA of males and females of all species was used for comparative approaches focused on an intraspecific comparison between males and females of Bc (Boa constrictor), with special emphasis on the homomorphic sex chromosomes in this species and in an interspecific genomic comparison among henophidian and caenophidian species. For intraspecific comparisons, male and female-derived gDNA of Boa constrictor were hybridized against male and female metaphase chromosomes of the species (Fig. 2). The female-derived gDNA was labeled with biotin-16-dUTP and male gDNAs with digoxigenin-11-dUTP by Nick translation means (Roche, Mannheim, Germany). Interspecific comparisons gDNA of male and female of all caenophidian species were hybridized against metaphase chromosomes and gDNA of male and female of Boa constrictor (Bc) (Fig. 4a,b). For this purpose, the gDNA of caenophidians male and female (green pit viper / Bothrops bilineatus = Bb; forest pit viper Bothrops taeniatus = Bt; amazonian pit viper / Bothrops atrox = Ba; bushmaster/Lachesis muta = Lm; common rattle snake/Crotalus terrificus = Ct; north rattle snake/Crotalus ruruima = Cr and puffer snake/Spilotes pullatus = Sp) were labeled with digoxigenin-11-dUTP (red), whereas   www.nature.com/scientificreports/ from Pogona vitticeps genomic BAC library as previously described in Ezaz et al. 12,52 , Young et al. 53 and Deakin et al. 19 . All 8 BACs were anchored to P. vitticeps metaphase chromosomes as control (data not shown). BAC DNA was extracted using the Promega Wizard Plus SV Minipreps DNA Purification System following the manufacturer's protocol, with volumes scaled up for 15 ml cultures. The BACs were labeled with SpectrumOrange-dUTP or SpectrumGreen-dUTP (Abbott, North Chicago, Ill., USA) and hybridized for 2 days. The slides were then washed twice in 0.4 × SSC, 0.3% IGEPAL (Sigma-Aldrich) at 55 °C for 5 min each and after air-dried, counterstained using DAPI (1.2 µg/ml) and mounted in an antifade solution (Vector, Burlingame, CA, USA).
Microscopy and image analyses. Images were captured using an Olympus BX51 microscope (Olympus Corporation, Ishikawa, Japan) with CoolSNAP. For W painting and BAC-FISH, images were captured using a Zeiss Axioplan epifluorescence microscope equipped with a CCD camera (Zeiss). ISIS software was used for microphotography and analyzing images.
ethics statement. We declare that all procedures and experimental protocols were approved and performed under the rules of the Ethics Committee of the National Institute of Amazonian Research (Permission number: 018/2017).