Chromosomal gene movements reflect the recent origin and biology of therian sex chromosomes

Mammalian sex chromosomes stem from ancestral autosomes and have substantially differentiated. It was shown that X-linked genes have generated duplicate intronless gene copies (retrogenes) on autosomes due to this differentiation. However, the precise driving forces for this out-of-X gene "movement" and its evolutionary onset are not known. Based on expression analyses of male germ-cell populations, we here substantiate and extend the hypothesis that autosomal retrogenes functionally compensate for the silencing of their X-linked housekeeping parental genes during, but also after, male meiotic sex chromosome inactivation (MSCI). Thus, sexually antagonistic forces have not played a major role for the selective fixation of X-derived gene copies in mammals. Our dating analyses reveal that although retrogenes were produced ever since the common mammalian ancestor, selectively driven retrogene export from the X only started later, on the placental mammal (eutherian) and marsupial (metatherian) lineages, respectively. Together, these observations suggest that chromosome-wide MSCI emerged close to the eutherian-marsupial split approximately 180 million years ago. Given that MSCI probably reflects the spread of the recombination barrier between the X and Y, crucial for their differentiation, our data imply that these chromosomes became more widely differentiated only late in the therian ancestor, well after the divergence of the monotreme lineage. Thus, our study also provides strong independent support for the recent notion that our sex chromosomes emerged, not in the common ancestor of all mammals, but rather in the therian ancestor, and therefore are much younger than previously thought

The role of CTCFL/BORIS in Meiosis

The complexity of mammalian genome organization demands a complex interplay of DNA and proteins to orchestrate proper gene regulation. CTCF, a highly conserved, ubiquitously expressed protein has been postulated as a primary organizer of genome architecture because of its roles in transcriptional activation/repression, insulation and imprinting. Diverse regulatory functions are exerted through genome wide binding via a central eleven zinc finger DNA binding domain and an array of diverse protein-protein interactions through N- and C- terminal domains. CTCFL has been identified as a paralog of CTCF expressed only in spermatogenic cells of the testis. CTCF and CTCFL have a highly homologous DNA-binding domain, while the flanking amino acid sequences exhibit no significant similarity. Genome- wide mapping of CTCF binding sites has been carried out in many cell types, but no data exist for CTCFL apart from a few identified loci. The lack of high quality antibodies prompted us to generate an endogenously flag-tagged CTCFL mouse model using BAC recombination. IHC staining using anti-flag antibodies confirmed CTCFL localization to type Β spermatogonia and preleptotene spermatocytes and a mutually exclusive pattern of expression with CTCF. ChIP followed by high-throughput sequencing identified 10,382 binding sites showing 70% overlap but representing only 20% of CTCF sites. Consensus sequence analysis identified a significantly longer binding motif with prominently less ambiguity of base calling at every position. The significant difference between CTCF and CTCFL genomic binding patterns proposes that their binding to DNA is differentially regulated. Analysis of CTCFL binding to methylated regions on a genome wide scale identified approximately 1,000 loci. Methylation-independent binding of CTCFL might be at least one of the mechanisms that ensures distinct binding patterns of CTCF and CTCFL since CTCF binding is methylation- sensitive. Co-localization of CTCF with cohesin has been well established and analysis of CTCFL and SMC3 overlap identified around 3,300 binding sites from which two related but distinct consensus sequence motifs were derived. Because virtually all data for cohesin binding originate from mitotically proliferating cells, the anticipated overlap is expected to be considerably higher in meiotic cells. Meiosis-specific cohesin subunit Rec8 is specific for spermatocytes and 6 out of the 12 identified binding sites are also bound by CTCFL. In conclusion, this was the first genome-wide mapping of CTCFL binding sites in spermatocytes, the only cell type where CTCF is not expressed. CTCFL has a unique binding site repertoire distinct from CTCF, binds to methylated sequences and shows a significant overlap with cohesin binding sites. Future efforts will be oriented towards deciphering the role CTCFL plays in conversion of chromatin structure and function from mitotic to meiotic chromosomes. - La complexité de l'organisation du génome des mammifères exige une interaction particulière entre ADN et protéines pour orchestrer une régulation appropriée de l'expression des gènes. CTCFL, une protéine ubiquitaire très conservée, serait le principal organisateur de l'architecture du génome de par son rôle dans l'activation / la répression de la transcription, la protection et la localisation des gènes. Diverses régulations sont opérées, d'une part au travers d'interactions à différents endroits du génome par le biais d'un domaine protéique central de liaison à l'ADN à onze doigts de zinc, et d'autre part par des interactions protéine-protéine variées au niveau de leur domaine N- et C-terminal. CTCFL a été identifié comme un paralogue de CTCF exprimé uniquement dans les cellules spermatiques du testicule. CTCFL et CTCF ont un domaine de liaison à l'ADN très homologue, tandis que les séquences d'acides aminés situées de part et d'autre de ce domaine ne présentent aucune similitude. Une cartographie générale des sites de liaison au CTCF a été réalisée pour de nombreux types cellulaires, mais il n'existe aucune donnée pour CTCFL à l'exception de l'identification de quelques loci. L'absence d'anticorps de bonne qualité nous a conduit à générer un modèle murin portant un CTCFL endogène taggué grâce à un procédé de recombinaison BAC. Une coloration IHC à l'aide d'anticorps anti-FLAG a confirmé la présence de CTCFL au niveau des spermatogonies de type Β et des spermatocytes au stade préleptotène, et une distribution mutuellement exclusive avec CTCF. Une méthode de Chromatine Immunoprecipitation (ChIP) suivie d'un séquençage à haut débit a permis d'identifier 10.382 sites de liaison montrant 70% d'homologie mais ne représentant que 20% des sites CTCF. L'analyse de la séquence consensus révèle un motif de fixation à l'ADN nettement plus long et qui comporte bien moins de bases aléatoires à chaque position nucléotidique. La différence significative entre les séquences génomiques des sites de liaison au CTCF et CTCFL suggère que leur fixation à l'ADN est régulée différemment. Appliquée à l'échelle du génome, l'étude de l'interaction de CTCFL avec des régions méthylées de l'ADN a permis d'identifier environ 1.000 loci. Contrairement à CTCFL, la liaison de CTCF dépend de l'état de méthylation de l'ADN ; cette modification épigénétique constitue donc au moins un des mécanismes de régulation expliquant une localisation de CTCF et CTCFL à des sites distincts du génome. La co- localisation de CTCF avec la cohésine étant établie, l'analyse de la superposition des séquences de CTCFL avec la sous-unité SMC3 identifie environ 3.300 sites de liaison parmi lesquels deux mêmes motifs consensus distincts par leur séquence sont mis en évidence. La presque quasi-totalité des données sur la cohésine ayant été établie à partir de cellules en prolifération mitotique, il est probable que la similitude au sein des séquences consensus soit encore plus grande dans le cas des cellules en méiose. La sous-unité Rec8 de la cohésine propre à l'état de méiose est spécifiquement exprimée dans les spermatocytes. Or 6 des 12 sites de liaison identifiés sont également utilisés par CTCFL. Pour conclure, ce travail constitue la première cartographie à l'échelle du génome des sites de liaison de CTCFL dans les spermatocytes, seul type cellulaire où CTCFL n'est pas exprimé. CTCFL possède un répertoire unique de sites de fixation à l'ADN distinct de CTCF, se lie à des séquences méthylées et présente un nombre important de sites de liaison communs avec la cohésine. Les perspectives futures sont d'élucider le rôle de CTCFL dans le remodelage de la structure de la chromatine et de définir sa fonction dans le processus de méiose.

Less mortality but more relapses in experimental allergic encephalomyelitis in CD8-/- mice.

Mice lacking in CD8 were generated from homologous recombination in embryonal stem cells at the CD8 locus and bred with the experimental allergic encephalomyelitis (EAE)-susceptible PL/JH-2u through four backcross generations to investigate the role of CD8+ T cells in this model of multiple sclerosis. The disease onset and susceptibility were similar to those of wild-type mice. However, the mutant mice had a milder acute EAE, reflected by fewer deaths, but more chronic EAE, reflected by a higher frequency of relapse. This suggests that CD8+ T lymphocytes may participate as both effectors and regulators in this animal model.

The balanced lethal system of crested newts: a ghost of sex chromosomes past?

Balanced lethal systems are more than biological curiosities: as theory predicts, they should quickly be eliminated through the joint forces of recombination and selection. That such systems might become fixed in natural populations poses a challenge to evolutionary theory. Here we address the case of a balanced lethal system fixed in crested newts and related species, which makes 50% of offspring die early in development. All adults are heteromorphic for chromosome pair 1. The two homologues (1A and 1B) have different recessive deleterious alleles fixed on a nonrecombining segment, so that heterozygotes are viable, while homozygotes are lethal. Given such a strong segregation load, how could autosomes stop recombining? We propose a role for a sex-chromosome turnover from pair 1 (putative ancestral sex chromosome) to pair 4 (currently active sex chromosome). Accordingly, 1A and 1B represent two variants (Y(A) and Y(B)) of the Y chromosome from an ancestral male-heterogametic system. We formalize a scenario in which turnovers are driven by sex ratio selection stemming from gene-environment interactions on sex determination. Individual-based simulations show that a balanced lethal system can be fixed with significant likelihood, provided the masculinizing allele on chromosome 4 appears after the elimination of the feminizing allele on chromosome 1. Our study illustrates how strikingly maladaptive traits might evolve through natural selection.

Strong reciprocity or strong ferocity? A population genetic view of the evolution of altruistic punishment.

Strong reciprocity, defined as a predisposition to help others and to punish those that are not helping, has been proposed as a potent force leading to the evolution of cooperation and altruism. However, the conditions under which strong reciprocity might be favored are not clear. Here we investigate the selective pressure on strong reciprocity by letting both limited dispersal (i.e., spatial structure) and recombination between helping and punishment jointly determine the evolutionary dynamics of strong reciprocity. Our analytical model suggests that when helping and punishment are perfectly linked traits (no recombination occurring between them), strong reciprocity can spread even when the initial frequency of strong reciprocators is close to 0 in the population (i.e., a rare mutant can invade). By contrast, our results indicate that when recombination can occur between helping and punishment (i.e., both traits coevolve) and is stronger than selection, punishment is likely to invade a population of defectors only when it gives a direct fitness benefit to the actor. Overall, our results delineate the conditions under which strong reciprocity is selected for in a spatially structured population and highlight that the forces behind its evolution involves kinship (be it genetic or cultural).

Stimulation of fission yeast and mouse Hop2-Mnd1 of the Dmc1 and Rad51 recombinases.

Genetic analysis of fission yeast suggests a role for the spHop2-Mnd1 proteins in the Rad51 and Dmc1-dependent meiotic recombination pathways. In order to gain biochemical insights into this process, we purified Schizosaccharomyces pombe Hop2-Mnd1 to homogeneity. spHop2 and spMnd1 interact by co-immunoprecipitation and two-hybrid analysis. Electron microscopy reveals that S. pombe Hop2-Mnd1 binds single-strand DNA ends of 3'-tailed DNA. Interestingly, spHop2-Mnd1 promotes the renaturation of complementary single-strand DNA and catalyses strand exchange reactions with short oligonucleotides. Importantly, we show that spHop2-Mnd1 stimulates spDmc1-dependent strand exchange and strand invasion. Ca(2+) alleviate the requirement for the order of addition of the proteins on DNA. We also demonstrate that while spHop2-Mnd1 affects spDmc1 specifically, mHop2 or mHop2-Mnd1 stimulates both the hRad51 and hDmc1 recombinases in strand exchange assays. Thus, our results suggest a crucial role for S. pombe and mouse Hop2-Mnd1 in homologous pairing and strand exchange and reveal evolutionary divergence in their specificity for the Dmc1 and Rad51 recombinases.

Intracellular excision and reintegration dynamics of the ICEclc genomic island of Pseudomonas knackmussii sp. strain B13.

Genomic islands are DNA elements acquired by horizontal gene transfer that are common to a large number of bacterial genomes, which can contribute specific adaptive functions, e.g. virulence, metabolic capacities or antibiotic resistances. Some genomic islands are still self-transferable and display an intricate life-style, reminiscent of both bacteriophages and conjugative plasmids. Here we studied the dynamical process of genomic island excision and intracellular reintegration using the integrative and conjugative element ICEclc from Pseudomonas knackmussii B13 as model. By using self-transfer of ICEclc from strain B13 to Pseudomonas putida and Cupriavidus necator as recipients, we show that ICEclc can target a number of different tRNA(Gly) genes in a bacterial genome, but only those which carry the GCC anticodon. Two conditional traps were designed for ICEclc based on the attR sequence, and we could show that ICEclc will insert with different frequencies in such traps producing brightly fluorescent cells. Starting from clonal primary transconjugants we demonstrate that ICEclc is excising and reintegrating at detectable frequencies, even in the absence of recipient. Recombination site analysis provided evidence to explain the characteristics of a larger number of genomic island insertions observed in a variety of strains, including Bordetella petri, Pseudomonas aeruginosa and Burkholderia.

Control of hair follicle cell fate by underlying mesenchyme through a CSL-Wnt5a-FoxN1 regulatory axis.

Epithelial-mesenchymal interactions are key to skin morphogenesis and homeostasis. We report that maintenance of the hair follicle keratinocyte cell fate is defective in mice with mesenchymal deletion of the CSL/RBP-Jkappa gene, the effector of "canonical" Notch signaling. Hair follicle reconstitution assays demonstrate that this can be attributed to an intrinsic defect of dermal papilla cells. Similar consequences on hair follicle differentiation result from deletion of Wnt5a, a specific dermal papilla signature gene that we found to be under direct Notch/CSL control in these cells. Functional rescue experiments establish Wnt5a as an essential downstream mediator of Notch-CSL signaling, impinging on expression in the keratinocyte compartment of FoxN1, a gene with a key hair follicle regulatory function. Thus, Notch/CSL signaling plays a unique function in control of hair follicle differentiation by the underlying mesenchyme, with Wnt5a signaling and FoxN1 as mediators.

Analysis of c-Myc function in mouse epidermis

Abstract The c-myc gene is one of the most frequently mutated oncogenes found in human tumors. c-Myc has been implicated in the regulation of various biological processes including cell cycle progression, cellular growth, differentiation, angiogenesis, immortalization and apoptosis. To assess the normal role of c-Myc in epithelial cell types in vitro and in vivo we have deleted the c-myc gene in keratinocytes and in the adult skin epidermis by conditional Cre/loxP mediated recombination. Similar to what we have previously shown in mouse embryonic fibroblasts acute elimination of c-Myc activity in cultured keratinocytes causes cells to cease proliferation and adapt a flat cell morphology. Mutant cells accumulate in a diploid Ki67neg stage, indicative of a quiescent Go stage. This demonstrates that c-Myc activity is essential to maintain keratinocytes in a productive cell cycle. In addition, mutant keratinocytes showed a defect in Ca2+ induced induction of the differentiation marker Keratin 1 suggesting a role for c-Myc during differentiation. To assess the in vivo role of c-Myc we used a tamoxifen inducible K5::CreERT transgene to delete the c-myc gene in the adult skin epidermis. Unexpectedly, despite strong c-Myc expression in the basal compartment it is not required for maintenance of the skin epidermis in the adult mouse. The epidermis appeared normal with respect to both proliferation and differentiation. In addition, no selection against c-Myc deficient epidermal cells occurred over many months, further confirming that c-Myc is dispensable for normal skin homeostasis. Even more surprising, TPA induced hyperproliferation also occurred in a c-Myc independent manner. Treatment of the skin with the mutagen DMBA prior to TPA is a classical way to induce papillomas by selecting for mutations that lead to dominant activation of the oncogene Ha-Ras. Most interestingly tumor formation was severely inhibited suggesting that tumor progression requires endogenous c-Myc. Further studies are required to address whether the role of c-Myc in the activation of telomerase or the Werner protein, or its role to induce angiogenesis is required for skin tumor progression, In conclusion, this work shows that while c-Myc is not required for maintenance or hyperplasia of mouse epidermis, it is essential for skin tumor progression in collaboration with Ras. Résumé Le gène c-myc est un des oncogènes les plus fréquemment mutés dans les tumeurs humaines. c-Myc est impliqué dans la régulation de processus biologiques variés, comme la progression du cycle cellulaire, la croissance cellulaire, la différenciation, l'angiogenèse, l'immortalisation et l'apoptose. Pour caractériser le rôle physiologique de c-Myc dans les cellules de type épithélial in vitro et in vivo, le gène c-myc a été délété dans des kératinocytes primaires et dans l'épiderme de peau de souris adultes par des recombinaisons conditionnelles (système Cre/loxP). De la même façon que dans les fibroblastes d'embryon de souris, l'élimination aiguë de l'activité de c-Myc dans les kératinocytes en culture primaire provoque l'arrêt de la prolifération des cellules et leur applatissement morphologique. Les cellules mutantes restent dans un stade diploïde Ki67neg, indiquant un stade quiescent Go. Cela démontre que l'activité de c-Myc est essentielle pour maintenir les kératinocytes dans le cycle cellulaire. De plus, les kératinocytes mutants montrent une déficience pour le marqueur de différenciation Kératine 1 au cours de la différenciation induite par le calcium, suggérant un rôle de c-Myc dans la différenciation cellulaire. Pour comprendre le rôle de c-Myc in vivo, le transgène K5::CreERT inductible par le tamoxifen a été utilisé pour déléter le gène c-inyc dans l'épiderme de souris adultes. Etonnemment, malgré une forte expression de c-Myc dans le compartiment basal de l'épiderme, ce gène n'est pas nécessaire pour la maintenance de l'épiderme de la peau chez la souris adulte. L'épiderme apparait normal avec une prolifération et une différenciation physiologique des cellules. De plus, il n'y a pas de sélection contre les cellules épidennales c-Myc déficientes après plusieurs mois, ce qui confirme que c-Myc n'est pas nécessaire pour l'homéostasie normale de la peau. Encore plus surprenant, une hyperprolifération est également induite par du TPA chez les souris mutantes, impliquant une voie de prolifération indépendante de c-Myc. Le traitement de la peau par le mutagène DMBA avant le traitement au TPA est une voie classique d'induction de papillomes, par sélection de mutations conduisant à l'activation de l'oncogène Ha-Ras. La formation des tumeurs est fortement inhibée chez les souris mutantes, suggérant que la progression des tumeurs nécessite la présence endogène de c-Myc. De nouvelles études sont nécessaires pour savoir si c-Myc a un rôle dans l'activation de la télomérase ou de la protéine de Werner, ou encore dans l'angiogénèse, qui sont nécessaires pour la progression tumorale. En conclusion, ce travail montre que même si c-Myc n'est pas nécessaire pour la maintenance ou l'hyperplasie de la peau de souris, il est essentiel pour la progression des tumeurs de la peau en collaboration avec Ras.

High level expression of human neuropeptide Y receptors in mammalian cells infected with a recombinant vaccinia virus.

Neuropeptide Y (NPY) is a 36 amino acid peptide present in the central and peripheral nervous system. Numerous studies point to a role of NPY in cardiovascular regulation. NPY effects are mediated through stimulation of specific cell surface G protein-coupled receptors. To allow biochemical studies of the receptor and of its interaction with the ligand, we have developed a potent expression system for NPY receptors using a recombinant vaccinia virus. A human NPY receptor cDNA was fused to a strong vaccinia virus promoter and inserted into the viral genome by homologous recombination. Recombinant viruses were isolated and tested for their ability to induce NPY binding site expression following infection of mammalian cell lines. Using saturation and competition binding experiments we measured a Bmax of 5-10 x 10(6) NPY binding sites per cell. The Kd for the binding of NPY is about 20 nM. Labelling of infected cells with a fluorochrome-labelled NPY indicated that the recombinant protein integrates into the cell membrane.

The ATPase activity of RecA is needed to push the DNA strand exchange through heterologous regions.

The role of ATP hydrolysis during the RecA-mediated recombination reaction is addressed in this paper. Recent studies indicated that the RecA-promoted DNA strand exchange between completely homologous double- and single-stranded DNA can be very efficient in the absence of ATP hydrolysis. In this work we demonstrate that the energy derived from the ATP hydrolysis is strictly needed to drive the DNA strand exchange through the regions where the interacting DNA molecules are not in a homologous register. Therefore, in addition to the role of the ATP hydrolysis in promoting the dissociation of RecA from the products of the recombination reaction, as described earlier, ATP hydrolysis also plays a crucial role in the actual process of strand exchange, provided that the lack of homologous register obstructs the process of branch migration.

Disentangling reasons for low Y chromosome variation in the greater white-toothed shrew (Crocidura russula).

Y chromosome variation is determined by several confounding factors including mutation rate, effective population size, demography, and selection. Disentangling these factors is essential to better understand the evolutionary properties of the Y chromosome. We analyzed genetic variation on the Y chromosome, X chromosome, and mtDNA of the greater white-toothed shrew, a species with low variance in male reproductive success and limited sex-biased dispersal, which enables us to control to some extent for life-history effects. We also compared ancestral (Moroccan) to derived (European) populations to investigate the role of demographic history in determining Y variation. Recent colonization of Europe by a small number of founders (combined with low mutation rates) is largely responsible for low diversity observed on the European Y and X chromosomes compared to mtDNA. After accounting for mutation rate, copy number, and demography, the Y chromosome still displays a deficit in variation relative to the X in both populations. This is possibly influenced by directional selection, but the slightly higher variance in male reproductive success is also likely to play a role, even though the difference is small compared to that in highly polygynous species. This study illustrates that demography and life-history effects should be scrutinized before inferring strong selective pressure as a reason for low diversity on the Y chromosome.

Mutation of yeast Ku genes disrupts the subnuclear organization of telomeres.

The mammalian Ku70 and Ku86 proteins form a heterodimer that binds to the ends of double-stranded DNA in vitro and is required for repair of radiation-induced strand breaks and V(D)J recombination [1,2]. Deletion of the Saccharomyces cerevisiae genes HDF1 and HDF2--encoding yKu70p and yKu80p, respectively--enhances radiation sensitivity in a rad52 background [3,4]. In addition to repair defects, the length of the TG-rich repeat on yeast telomere ends shortens dramatically [5,6]. We have shown previously that in yeast interphase nuclei, telomeres are clustered in a limited number of foci near the nuclear periphery [7], but the elements that mediate this localization remained unknown. We report here that deletion of the genes encoding yKu70p or its partner yKu80p altered the positioning of telomeric DNA in the yeast nucleus. These are the first mutants shown to affect the subnuclear localization of telomeres. Strains deficient for either yKu70p or yKu80p lost telomeric silencing, although they maintained repression at the silent mating-type loci. In addition, the telomere-associated silencing factors Sir3p and Sir4p and the TG-repeat-binding protein Rap1p lost their punctate pattern of staining and became dispersed throughout the nucleoplasm. Our results implicate the yeast Ku proteins directly in aspects of telomere organization, which in turn affects the repression of telomere-proximal genes.

Within-population polymorphism of sex-determination systems in the common frog (Rana temporaria).

In sharp contrast with birds and mammals, the sex chromosomes of ectothermic vertebrates are often undifferentiated, for reasons that remain debated. A linkage map was recently published for Rana temporaria (Linnaeus, 1758) from Fennoscandia (Eastern European lineage), with a proposed sex-determining role for linkage group 2 (LG2). We analysed linkage patterns in lowland and highland populations from Switzerland (Western European lineage), with special focus on LG2. Sibship analyses showed large differences from the Fennoscandian map in terms of recombination rates and loci order, pointing to large-scale inversions or translocations. All linkage groups displayed extreme heterochiasmy (total map length was 12.2 cM in males, versus 869.8 cM in females). Sex determination was polymorphic within populations: a majority of families (with equal sex ratios) showed a strong correlation between offspring phenotypic sex and LG2 paternal haplotypes, whereas other families (some of which with female-biased sex ratios) did not show any correlation. The factors determining sex in the latter could not be identified. This coexistence of several sex-determination systems should induce frequent recombination of X and Y haplotypes, even in the absence of male recombination. Accordingly, we found no sex differences in allelic frequencies on LG2 markers among wild-caught male and female adults, except in one high-altitude population, where nonrecombinant Y haplotypes suggest sex to be entirely determined by LG2. Multifactorial sex determination certainly contributes to the lack of sex-chromosome differentiation in amphibians.

Ever-young sex chromosomes in European tree frogs.

Non-recombining sex chromosomes are expected to undergo evolutionary decay, ending up genetically degenerated, as has happened in birds and mammals. Why are then sex chromosomes so often homomorphic in cold-blooded vertebrates? One possible explanation is a high rate of turnover events, replacing master sex-determining genes by new ones on other chromosomes. An alternative is that X-Y similarity is maintained by occasional recombination events, occurring in sex-reversed XY females. Based on mitochondrial and nuclear gene sequences, we estimated the divergence times between European tree frogs (Hyla arborea, H. intermedia, and H. molleri) to the upper Miocene, about 5.4-7.1 million years ago. Sibship analyses of microsatellite polymorphisms revealed that all three species have the same pair of sex chromosomes, with complete absence of X-Y recombination in males. Despite this, sequences of sex-linked loci show no divergence between the X and Y chromosomes. In the phylogeny, the X and Y alleles cluster according to species, not in groups of gametologs. We conclude that sex-chromosome homomorphy in these tree frogs does not result from a recent turnover but is maintained over evolutionary timescales by occasional X-Y recombination. Seemingly young sex chromosomes may thus carry old-established sex-determining genes, a result at odds with the view that sex chromosomes necessarily decay until they are replaced. This raises intriguing perspectives regarding the evolutionary dynamics of sexually antagonistic genes and the mechanisms that control X-Y recombination.