The multiple sex chromosomes of platypus and echidna are not completely identical and several share homology with the avian Z

A comparative study of the karyotype of the short-beaked echidna shows that monotremes appear to have a unique XY sex chromosome system that shares some homology with the avian Z.


Background
Monotreme mammals are receiving increasing attention in genomic research, with interests varying from karyotype evolution and gene mapping, to comparative sequencing. This should not come as a surprise, as monotremes (mammalian Subclass Prototheria) occupy a unique branch at the base of the mammalian phylogenetic tree, and serve as an evolutionary outgroup for marsupial and eutherian species (that together comprise Subclass Theria). The time of divergence of Prototheria and Theria is estimated to be in the Early Jurassic (166 million years ago (MYA)), while marsupials and eutherians diverged in the Late Jurassic (148 MYA) [1]. Five extant monotreme species are recognized; platypus (Ornithorhynchus anatinus), short-beaked echidna (Tachyglossus aculeatus) and three long-beaked echidnas (Zaglossus bruneiji, Zaglossus attenboroughi, Zaglossus bartoni). Zaglossus bartoni is divided into three subspecies Z. b. smeenki, Z. b. diamondi, and Z. b. clunius [2].
A full karyotype characterization is essential for genomic research in any species. It is particularly important for monotremes because of their exceptional sex chromosome complement. The inclusion of a set of tiny chromosomes was recognized early and thought to be a reptilian feature [3], but this suggestion was later refuted [4]. A surprise was the discovery of several unpaired chromosomes [5]. A final identification and description of the platypus unpaired chromosomes was achieved only recently by our chromosome painting studies [6,7]. The 21 autosome pairs were assigned by chromosome paints. Ten paints identified ten unpaired mitotic chromosomes as well as the ten members of the meiotic chain and the homologous regions between them. Five paints identified X chromosomes present in single copy in males but two copies in females, and five paints identified Y chromosomes that were present only in males. It was, therefore, concluded that the ten male unpaired chromosomes consisted of five X and five Y sex chromosomes. The ten sex chromosomes form a multivalent chain at meiosis held together by chiasmata within homologous pairing regions. Alternate segregation of these chromosomes into X 1 X 2 X 3 X 4 X 5 and Y 1 Y 2 Y 3 Y 4 Y 5 sperm was proposed and must be very efficient as shown by meiotic analysis of spermatids and sperm using the paint probes [6]. Remarkably, X 5 shows some homology with the chicken Z, as demonstrated by its inclusion of the DMRT-1, DMRT-2 and DMRT-3 orthologues [6,8]. Chicken Z is largely homologous to parts of human chromosomes 5 and 9, with some genes represented on 8 and 18 [9]. A region containing ATRX, RBMX and genes flanking XIST, present on Xq in human and other therians, maps to chromosome 6 in platypus [10], as does SOX3, the gene from which the sex-determining SRY gene evolved (M Wallis, personal communication), and this is consistent with the absence of a platypus homologue of the Y-linked SRY. Other genes involved in the eutherian sex determining pathway have recently been mapped to platypus autosomes, so do not qualify as candidate primary sex determining genes [11]. There is no platypus homologue of the human X-borne XIST in platypus [12] and marsupials [13]. In addition, platypus Ensembl release 44 and separate mapping work (F Veyrunes, personal communication) show an absence of human X-linked orthologues from platypus Xchromosomes, contradicting original localizations using radioactive fluorescent in situ hybridization (FISH) with heterologous cDNA probes [14][15][16][17][18]. It follows that SRY and the therian XY sex determining system have evolved between 166 and 148 MYA after the divergence of monotremes and before the divergence of marsupials, which is being explored further (F Veyrunes, personal communication).
To provide new clues to the organization, function and evolution of the platypus multiple sex chromosomes, we defined the sex chromosomes of the distantly related short-beaked echidna, T. aculeatus, and established the sex chromosome order in the echidna multivalent chain. Our genome-wide comparison using chromosome painting between echidna and platypus (called Tac (for T. aculeatus) and Oan (for O. anatinus) in this report) showed, surprisingly, that one member of the Oan chain is replaced by an autosome in Tac, and the X homologous to Oan X 5 occupies a central position in the Tac chain rather than a position at the end as seen in Oan. To investigate the participation of the ancestral avian Z in the evolution of the monotreme sex chromosome system and to map genes to the members of the sex chromosomes, we also localized the platypus homologues of genes on chicken autosomes and Z. We conclude that the ancestral monotreme sex chromosome system bears considerable homology to the sex chromosomes of birds.

Characterization of the short-beaked echidna karyotype
The male short-beaked echidna has 63 chromosomes and the female 64 [5]. We characterized and classified the male karyotype by flow cytometric analysis, flow sorting, and chromosome painting. The chromosomes produced 36 peaks in the flow karyotype ( Figure 1); 17 peaks represent single chromosome pairs and 4 peaks represent 2 chromosome pairs each. The homologues of chromosomes 1, 6, 16, and 27 (which are frequently heteromorphic) each sort in two different peaks. The nine remaining peaks represent single chromosomes, which we show to be the nine unpaired sex-chromosomes that constitute the meiotic chain in the male echidna. Thus, the male echidna has 27 autosome pairs and 9 sex chromosomes.
The chromosome paints were used to identify autosome pairs and sex chromosomes in male and female echidnas. Figure 2 shows a G-banded karyotype of male (upper) and female (lower) echidna chromosomes arranged in size and identified by the chromosome painting results (see below). The upper part of both the male and female karyotypes show 27 pairs of autosomes that should form bivalents at meiosis, and the lower parts show the nine unpaired sex-chromosomes in male and five paired sex-chromosomes in female. The paired autosomes can be divided into a group that contains submetacentric chromosomes (chromosomes 1 to 8) and a group of smaller metacentric and submetacentric chromosomes. Chromosomes 3, 6 and X 5 contain nucleolus organizer regions ( Figure 2) determined by Ag-NOR (nucleolar organizing region) staining and by FISH using a probe specific for 28S rDNA (Figure 3).
Special attention was given to chromosome pair 27, as it shows complete homology with platypus X 4 and partial homology with platypus Y 3 and Y 4 (see below), and might, therefore, be an unrecognized sex chromosome. However, both the paints, produced from the two peaks representing the heteromorphic chromosome pair 27, painted both chromosomes 27 totally in both male and female metaphases (Figure 4a). Based on this analysis, the pair was defined as an autosome pair and this was confirmed by analysis of meiotic preparations, which revealed that 27 is not part of the chain.
Chromosome painting in echidna reveals non-specific signals that are present on more than one chromosome pair, indicated by coloured bars next to chromosomes in Figure 2. For instance, when the paint specific for chromosome 1 was hybridized to metaphases, stronger signals were visible on the region indicated on chromosome 1 and a region on chromosome 2. A hybridized chromosome 4 paint gave strong signals on chromosomes 4, 7, and 16 (and others as indicated), weaker signals on X 1 and Y 1 , and even weaker signals on chromosomes 1 and 2.
The chromosome specificity of these regions was apparent using paints produced by microdissection of these regions. Hybridization to these regions (which hamper identification of chromosomes, and especially pairing regions of the sex chromosomes) was not blocked by pre-hybridization with echidna Cot-1 DNA. To facilitate chromosome identification, an image enhancement procedure was developed to remove these non-specific signals from the image [19]. Molecular characterization of these regions is not considered in this paper.

Sex chromosomes in the echidna male
The nine chromosome paints that identified unpaired chromosomes (denoted by X 1 , Y 1 , X 2 , Y 2 , and so on) were used to identify the sex chromosomes of the echidna and predict their order in the meiotic chain. Paint X 1 hybridized to the whole of chromosome X 1 and the long arm of Y 1 (Figure 4b), whereas paint Y 1 hybridizes to the single chromosome Y 1 and the short arm of X 1 (Figure 4c). X 1 is known to be the first chromosome in the chain [20], and its short arm pairs with the acrocentricY 1 q (second chromosome). The paint X 2 shows a complete coverage of a single chromosome X 2 as well as signals on the pairing regions on Y 1 p and Y 2 p (Figure 4d). This result shows that X 2 is the third and Y 2 the fourth in the chain. Paint Y 2 covers Y 2 and paints the pairing regions on X 2 and X 3 (Figure 4e). Paint X 3 paints the entire chromosome X 3 as well as the long arm of Y 2 (Figure 4f), so is the fifth chromosome    Short-beaked echidna karyotype female in the chain. The order of the last four elements is less certain at this stage. Paint Y 3 covers the tiny chromosome Y 3 with no signal denoting pairing regions on X 3 and X 4 ( Figure 4g). Paint X 4 hybridized to chromosome X 4 and to Y 3 and the heterochromatic centromeric regions of chromosome 22 or 23 ( Figure 4h). Y 3 is also painted by chromosome paints 25 and 26, suggesting that it contains shared large non-specific sequences. Paint Y 4 hybridized to chromosome Y 4 and to X 5 p ( Figure 4i). Paint X 5 paints the whole X 5 and the long arm of Y 4 ( Figure 4j); it also showed hybridization to a heterochromatic centromeric region on an autosome.

Sex chromosomes in the echidna female
The same set of nine sex chromosome paints was hybridized to female metaphases to verify which element is an X-chromosome (defined as having one copy in the male and two copies in the female) and which element is a Y-chromosome (one copy in the male and absent in the female). Figure 5 shows some examples of these hybridizations. For instance, chromosome paint Y 2 hybridized to the pairing regions of X 2 and X 3 , but identified no copy of the male-specificY 2 ( Figure 5b).
The results show that indeed X 1 -X 5 are X chromosomes and Y 1 -Y 4 are Y chromosomes. These results also clarified the order of the alternating X and Y chromosomes.
Homologous regions between adjacent X and Y chromosomes are, therefore, demonstrated for all members of the chain of nine except between the small X 3 , Y 3 and X 4 . These pairing regions all include the distal end of one chromosome arm and do not cross the centromere of the unpaired chromosome. The order of the first five chromosomes of the chain can be deduced from the homology relationships as X 1 Y 1 X 2 Y 2 X 3 . However, the order of the last two X and two Y chromosomes is uncertain by chromosome painting on echidna mitoses, as the pairing regions are too small to detect. However, the results of cross-species painting (see below) and our analyses of chromosome painting of meiotic chains (Figure 5e-g) revealed that X 4 is the seventh element in the chain ( Figure  5g), confirming the order shown in Figure 2.

Genome wide comparison between echidna and platypus
Cross-species chromosome painting was used to define chromosome regions conserved between Oan and Tac and to identify rearrangements that differentiate the karyotypes of the two species. Figures 2a and 6 show the Oan and Tac homology maps, with the homologous regions indicated on the right of each chromosome.

Comparison of platypus-echidna autosomes
Cross-species painting showed that entire Oan chromosomes 1, 4, 5, 9, 11, 14, 16, and 19  G-banded karyotype of T. aculeatus Figure 2 (see previous page) G-banded karyotype of T. aculeatus. Top: the male has 27 chromosome pairs and 9 unpaired chromosomes. Three kinds of information are given next to the chromosomes. Chromosomes 3, 6, and X 5 contain the NOR regions. Certain chromosomes have specific regions represented by colored bars on the left of the chromosomes, 'w' means that the region is relatively under-represented (see text). The numbers on the right refer to platypus chromosome paints that hybridized to the indicated regions. Middle: the pairing regions of the nine sex chromosomes determined by chromosome painting on mitotic preparations. Those of Y 3 with X 3 could not be determined in mitotic metaphases. Bottom: G-banded female karyotype of T. aculeatus. The female has 32 chromosome pairs and no unpaired chromosomes.
The NOR regions of T. aculeatus Figure 3 The NOR regions of T. aculeatus. FISH with a 28S specific probe was used for identification.
Tac homologous regions for Oan 18, 17p and 21 could not be determined, probably because these regions are homologous to Tac chromosomes with large amounts of 'non-chromosome-specific' repetitive DNA (see first section of results). Similarly, homologous regions were not determined for short regions on Tac 3p, 4p, 5p, 6p, 8p, 13p, 15q, 16p, 17q, 18, 20, 21, 23, 24p and the large regions on 1q and 2q. These blocks correspond to specific Tac regions identified by paints made from the equivalent regions by microdissection (depicted in different colors at the left in Figure 2 top).
The NOR-bearing regions are not on homologous chromosomes in the platypus and echidna. In platypus, Oan 6 contains the NOR region; this chromosome is homologous to Tac 16. In echidna, Tac chromosomes 3, 6, and X 5 are the NOR bearing chromosomes ( Figure 3); these chromosomes are homologous to Oan 5, 7, and 12p.

Comparison of platypus and echidna sex chromosomes
Cross-species painting with Oan and Tac X-Y probes shows that Oan X 1 , Y 1 , X 2 , Y 2 , X 3 are homologous to Tac X 1 , Y 1 , X 2 , Y 2 , and X 3 , respectively (Figures 8a-d and 9a-c), but one X chromosome in each and homologous regions of the flanking Y chromosomes are non-homologous. Reciprocal chromosome painting shows that Oan X 5 and Tac X 4 are homologous.
The Oan X 5 paint hybridized to Tac X 4 (Figure 8h), the Tac X 4 paint hybridized to Oan X 5 ( Figure 9f); neither hybridization detected pairing regions in adjacent Y chromosomes. Confirmation that the large Tac X 4 chromosome is the genetic homologue of the Oan X 5 is provided by the assignment of the DMRT1 gene complex to Tac X 4 q ( Figure 10a).
An important result was that Oan Y 3 and X 4 paints hybridized to a Tac autosome, and that Tac X 5 paint hybridized to an Oan autosome. Oan paint X 4 hybridized to the whole Tac 27 (Figure 8f). Oan paint Y 3 hybridized to a small region of Tac chromosome 27 (Figure 8e), confirmed as an autosome ( Figure  4a). Oan paint Y 4 hybridized to Tac Y 4 p and the distal end of Tac X 4 p ( Figure 8g). Reciprocally, Tac X 5 paint hybridized to platypus chromosome 12p (Figure 9d). Tac paint Y 4 hybridized to four regions: the two homologues Oan 12p, Oan X 5 p and Oan Y 4 p and Tac 27 identified Oan Y 3 , Oan X 4 , and Oan Y 4 q ( Figure 9e). Many metaphases were observed to confirm these signals. We conclude that Tac Y 4 q and X 5 represent autosomes in platypus and parts of Oan Y 3 , X 4 and Y 4 q are homologous to autosomal regions in echidna.
In the echidna, there are only four Y chromosomes, compared to five in the platypus, suggesting that the small Y 5 in platypus has been completely lost in the echidna lineage [21]. A surprising result, therefore, was the hybridization of the Oan Y 5 paint to Tac Y 3 , a strong indication that Oan Y 5 is represented in the echidna chromosome chain of nine ( Figure  8i). No reliable signals on chromosomes other than Tac Y 3 were observed. Attempts to hybridize the chromosome paint of Tac Y 3 paint onto Oan male metaphases produced no reliable signal.

Mapping chicken-human homologous genes
In order to test the hypothesis that the bird Z chromosome is represented in the monotreme sex chromosome chain, platypus homologues of nineteen chicken-Z genes (together with chicken 2, 3, and 13 genes) were mapped to platypus chromosomes by two independent methods, PCR amplification of chromosome specific DNA and FISH localization of platypus bacterial artificial chromosome (BACs). Gene mapping results are summarized in Figures 11 and 12, and see Additional data file 1. This file also shows platypus contigs that contain the mapped and predicted genes extending the regions of homology.
Gene loci in chicken are designated in the following section according to their established conserved synteny with human chromosomes. For instance, chicken Z is homologous with regions of human chromosomes 5, 9, 8 and 18 and chicken chromosome 2, 3 and 13 also share homology with human chromosome 5 [9]. Figure 12 presents a standard idiogram, based on G-banding and chromosome painting, on which the present and previous [10,11,[22][23][24][25] gene assignments are shown. Consideration of the contigs in Ensembl to which these genes belong greatly expands the size of the syntenic regions (Additional data file 1). Most of the conserved syntenies that we have observed between platypus chromosomes and chicken Z are in five groups.
The first group includes platypus homologues of chicken Z and human 9 (Gallus gallus (GGA)-Z/Homo sapiens (HAS)-9) genes. Three genes mapped to platypus X 5 , one mapped to X 2 and two mapped to X 3 ( Figure 12). Consideration of the contigs to which these genes belong greatly expands the size of the homologous regions (Additional data file 1).
The second group includes platypus homologues of chicken Z and human 5 (GGA-Z/HSA-5) genes. Ten genes are distributed over platypus chromosomes 1, 2, 3 and the sex chromosomes. Only one GGA-Z/HSA-5 gene (PDE6A) mapped to platypus X 1 by both PCR and BAC-clone mapping. This gene mapped on the pairing region of X 1 p and Y 1 q, distal to the   centromere on each (Figure 10b,c), and is contained in platypus contig 269 with nine other genes, four of which are homologous to GGA-13/HSA-5q (Additional data file 1). Two platypus genes (LMNB1 and DMXL1) homologous to chicken Z and human 5 (GGA-Z/HSA-5) mapped to platypus X 5 by both PCR and BAC-clone mapping. The homologue of the chicken gene LMNB1 PCR-mapped also to X 1 but not to Y 1 , suggesting that the PCR product of X 1 may represent a section of the X 5 gene that has a copy (paralogue) on X 1 . Likewise, DMXL1 was PCR-mapped also to Oan 10 as well as X 1 (but not Y 1 ), again indicating that a section of the DMXL1-X 5 gene has a copy (paralogue) on both Oan 10 and X 1 . The seven other GGA-Z/HSA 5 genes map to Oan 1, 2 and 3 by PCR or BAC-FISH.
The third group includes platypus homologues of chicken 13 and human 5 (GGA-13/HSA-5), and chicken 3 and human 2 (GGA-3/HSA-2) genes. Five GGA-13/HSA-5 genes and one GGA-3/HSA-2 gene mapped to the pairing regions of X 1 and Y 1 (Figures 11 and 12). PANK3 also PCR-mapped to Oan 8, which might be a paralogue. Thus, the three GGA-Z/HSA-5 homologues on X 1 are accompanied by GGA-13/HSA-5 genes     Thus, mapping platypus homologues of chicken Z genes revealed homology not only with platypus X 5 but also with platypus X 1 , Y 1 , X 2 and X 3 . Other genes from human chromosomes 5 and 9, which are not homologous to the chicken Z, also mapped to platypus sex chromosomes. A set of GGA-Z genes homologous to HSA-5, 8 or 18 (see above), but so far not HSA-9, mapped to platypus autosomes. So far, no chicken Z genes have been mapped to platypus X 4 , but this may not be surprising as X 4 is homologous to an echidna autosome.

Discussion
We confirm previous cytological studies of echidna mitotic and meiotic chromosomes that showed that the male shortbeaked echidna has 63 chromosomes -27 pairs of autosomes and 9 sex chromosomes [20,21,26] -and establish that the sex chromosome constitution is 5 Xs and 4 Ys. These chromosomes form a chain of nine at meiosis, expected (by analogy to platypus) to be in an alternating X-Y order. Adjacent members are expected to be held together by pairing within 16 pseudoautosomal regions (one per chromosome arm except for X 1 q and X 5 p, which terminate the chain). Each sex chromosome and 13 of the expected 16 pairing regions of the 5 Xs and 4 Ys (Figure 2, middle) were identified in this study. The alternating order in the chain is directly confirmed by painting in meiosis I metaphases (Figure 5e-g).

Homology between platypus and echidna chromosomes
Ten chromosomes are completely conserved between echidna and platypus. The non-conserved chromosomes differ between the two species by centric rearrangements only, which are charted in Figures 2 and 6. Without comparative data from an outgroup it is not possible to distinguish fusions in one lineage from fissions in the other.
Both platypus and echidna have prominent NORs on the short arm of chromosome 6, which was thought to be homologous, because of their similar size and morphology [21]. However, chromosome painting shows that platypus chromosome 6 is homologous to echidna chromosome16. In addition, echidna has NORs on X 4 and chromosome 3. Nonhomology of NOR-bearing chromosomes is not surprising in view of their rearrangement in many closely related species.

Differences between platypus and echidna sex chromosome chains
The constitution of the sex chromosome system in Tac differs from that in Oan in their number, order and in the identity of one XY pair. Firstly, chromosome painting shows that there are five Xs and five Ys in platypus but five Xs and only four Ys in the echidna. The missing platypus Y 5 is a very small chromosome, and it had been previously supposed to have been lost from the echidna lineage [21,27]. However, the presence of a strongly hybridizing region on the larger echidna Y 3 using the platypus Y 5 paint suggests that the content of this platypus Y is incorporated into echidnaY 3 . The surprising finding is that these two Y-chromosomes are at different locations in the chain. Secondly, Oan X 5 and Tac X 4 are shown to be homologous chromosomes by chromosome painting, and by sharing the DMRT1 gene cluster (Figure 10a). However, they occupy different positions in the chain. Thirdly, the most telling result we obtained was the finding that the platypus and echidna sex chromosome chains contain elements (Oan chromosomes Y 3 , X 4 and Y 4 , and Tac X 5 and Y 4 ) that are not homologous; platypus X 4 paints an autosome (chromosome 27) in echidna and echidna X 5 paints an autosome (chromosome 12) in platypus.
Thus, the chain in these monotreme species differs in both order and constitution, indicating that the chain continued to evolve after the divergence of platypus and echidna approximately 25 MYA [28]. One can speculate that the pri-Oan chromosome paints hybridized to male Tac metaphases      mary sex determining locus is more likely to be found on the non-pairing regions of the X or Y chromosomes that are shared between platypus and echidna rather than on those that are autosomal in either.
Although the comparative painting reported here is a genome-wide comparison between the two monotreme species, it cannot, of course, reveal the gene content of the chromosomes, which requires comparative gene mapping.

Mapping chicken-human homologous genes
Comparative gene mapping was used to establish that homologous genes are together in the same contiguous region in both species indicating chromosome homology. This does not exclude the possibility of non-orthology for some of the contiguous genes, but the likelihood of multiple, independent exceptional events is reduced as more homologues are discovered in the region. As the chance of independent evolution of syntenic regions is reduced, the likelihood of shared descent from the same chromosomal region in a common ancestor is increased.
The comparative mapping results show that monotreme sex chromosomes contain genes homologous to the chicken Z chromosome and chicken autosomes, with the implication that the sex determining system might be related to an ancestral sauropsid system. The early finding of the DMRT1 gene on platypus X 5 prompted our search for other chicken Z genes on this chromosome by mapping homologues of GGA-Z/HSA-9 genes followed by homologues of GGA-Z/HSA-5 genes. This led to the observation of chicken Z and autosomal homologues on other platypus sex chromosomes.
With regard to platypus X 1 , the preliminary results of the draft genome sequence of the female platypus (Ensembl release 44) and a recent comparative mapping study of therian X-linked genes (F Veyrunes, personal communication) show that platypus X 1 does not share homology with the therian X as previously reported [14][15][16][17][18]. X-linked genes on human Xq assigned by FISH localization of BACs so far all map to platypus chromosome 6 [10]. These on human Xp map to platypus chromosome 15 and 18 [23] (F Veyrunes, personal communication). The results presented here reveal instead that platypus X 1 shares homology with chicken chromosomes 3, 13, and Z and the corresponding human chromosomes 2, 5, 8, and 9 (Additional data file 1). Thus, our mapping data show that X 1 shares homology with those chicken chromosomes that are homologous to human autosomes.
The implications of some of the platypus gene assignments require further discussion. For example, the motivation to map RAPGEF6 was that this gene is close to the chicken Z HINT1 homologue in human (HSA-5q), although on a different chromosome in chicken (GGA-13). HINT1 might be involved in the chicken sex determining pathway, so may be a candidate for the primary sex determining locus in platypus; mapping a HINT1 homologue was unsuccessful. However, the localization of RAPGEF6 to Oan X 1 , Y 1 makes it unlikely that this region contains the primary sex determinant. Note that platypus X 1 p and Y 1 q contain homologues of genes on chicken 3, 13 and Z (mapped in this report). The chicken 13 and the chicken Z regions were presumably syntenic before the divergence of Prototheria and Theria as these regions are syntenic in both monotremes and human.
The platypus homologues that map to X 2 (MLLT3), X 3 (TSCOT, PALM2), and both X 3 and Y 2 (CDH12, DNAH5, TRIO, P15RS) assign their respective contigs to these sex chromosomes ( Figure 12 and Additional data file 1), indicating that parts of chicken 2 and chicken Z were fused before the divergence of Prototheria and Theria, and that the fission of chicken chromosome 2 into the regions on human chromosome 5, 9, and 18 occurred after the divergence of Prototheria and Theria.
Previous work showed that platypus chromosome X 5 contains the DMRT1-2-3 complex, whose homologue maps to the Z chromosome in chicken, and to human chromosome 9 [6,8].
We show here that several other genes with homologues on the chicken Z also map to the platypus X 5 . Identification of the contigs that include these genes maps a large fraction of the chicken Z to platypus X 5 , indicating that a large region of human 9 homologous to chicken Z is conserved on platypus chromosome X 5 . Platypus major histocompability complex (MHC) class III genes have been mapped recently to the pairing segments of X 5 and Y 4 and MHC class I and II genes have been mapped to the pairing regions of X 3 and Y 3 [29]. We note that platypus X 4 (homologous to the echidna autosome Tac 27) separates these two clusters, while in echidna the homologues of platypus X 5 and Y 4 (that is, Tac X 4 and Y 4 ) that carry the MHC genes are adjacent to X 3 and Y 3 , suggesting that the insertion of X 4 in platypus was a later event in the evolution of the monotreme meiotic chain. So far, no homologues of Tac chromosome paints hybridized to male Oan metaphases  chicken Z map to platypus X 4 , possibly because it is homologous to an echidna autosome.
The gene mapping results described here assign several genes to the pairing regions of platypus Y chromosomes. The sequence reported in Ensembl does not provide this information as it is based on the female genome, so Y-linked genes are not included.
Not all chicken Z homologous genes are located on the monotreme sex chromosomes. Seven additional GGA-Z/ HSA-5 genes map to platypus autosomes 1, 2, or 3 ( Figures 11  and 12). The GGA-Z/HSA-8 genes LPL and CHRNB3 map to platypus chromosome 5 and the GGA-Z/HSA-18 gene ATP5A1 maps to platypus chromosome 3. Platypus chromosome 3 contains homologues of genes localized on chicken Z and human 5 and 18. This means that these two human regions were syntenic before monotreme-therian divergence and became separated only in the mammalian lineage after this divergence.
The monotreme regions homologous to chicken Z are considerably rearranged and distributed over autosomes and at least four X chromosomes and the corresponding Ys. The (single) Z homologues on X 1 , Y 1 and X 2 are accompanied by other non-Z homologues in their respective contigs. X 3 seems to have a large region homologous to chicken Z as it contains contig 278 with a size around 0.8 Mb. This contig has chicken Z genes that are homologous only to human chromosome 9 and 5. In Ensembl release 45, contig 278 is part of ultracontig 84 with a size of around 8 Mb containing more chicken Z (human 5 and 9) homologous genes. Only chicken Z genes have so far been mapped to platypus X 5 and most of these are homologous to human 9 and a few to human 5. As both X 3 and X 5 contain human 5 and 9 genes (that are homologous to chicken Z), the separation into these human 5 and human 9 regions must have occurred later in the therian lineage. Finally, the platypus autosomes 1, 2 and 3 also contain chicken Z genes, suggesting that these chromosomes may represent the partners in original translocations involving a putative ancestral Z chromosome.
The results shown in Figure 12 all confirmed the homologies found by chromosome painting in this report and in our previous paper [7]. In particular, a number of FISH assignments confirmed the pairing regions between X and Y chromosomes. These early mapping results should help in the identification of the primary sex-determining locus, in the investigation of mechanisms of dosage compensation and in understanding the evolution of vertebrate sex chromosomes.
An exchange mechanism for the evolution of the multiple sex chromosomes of the platypus was postulated previously [7,21]. The chain development was suggested to start with an ancestral pair of differentiated sex chromosomes, one of which was repeatedly involved in exchanges with autosomes. An alternative view [27,30] suggests that the chain arose as the result of hybridization between two ancestral monotreme populations, each with a different set of Robertsonian translocations resulting in a male heterozygous for unpaired sex chromosomes. Common to all models is that the rearranged autosomes in the chain evolved into Y chromosomes. Our finding that different rearrangements occurred in the two monotreme lineages after the platypus-echidna divergence (25 MYA [28]) is easier to reconcile with a model of successive translocation, rather than the unlikely alternative of additional hybridizations between populations differing in other Robertsonian rearrangements.

Conclusion
Our cross-species painting studies of the monotreme sex chromosome complements shows that the platypus and echidna translocation chains share homology over four of the Location of mapped human 5 and human 9 genes in human, chicken and platypus Figure 11 Location of mapped human 5 and human 9 genes in human, chicken and platypus. Gene names in italic are mapped in platypus by PCR, gene names in bold are mapped by both PCR and BAC-clone FISH. EPB41L4B is in contig 29 (Additional data file 1, P15RS). five X chromosomes, but one in each species is entirely nonhomologous. This means that the chains continued to evolve after the divergence of platypus and echidna.
Our comparative mapping studies show chicken Z homologous genes in the sex chromosome system with the main clusters on platypus X 3 and X 5 and echidna X 3 and X 4 . Other Z homologous genes map to autosomes, indicating many rearrangements between the monotreme and avian lineages.
In combination with the mapping data available in current Ensembl release 44, our results also reveal homology of platypus X 1 to chicken 3, 13, Z, 11, and 12, which are homologous to human autosomes. This suggests that the monotreme's XY chromosome system is unrelated to the therian XY system. This is further explored by F Veyrunes (personal communication) in comparative studies with therian X-linked genes, and it may mean that the therian XY system evolved after the prototherian and therian divergence, but before the divergence of marsupials, and is, therefore, younger than previously anticipated [31].
It is important to note that three monotreme X chromosomes have large differential regions. The differential region on platypus X 5 (echidna X 4 ) is homologous to chicken Z, that of X 3 is homologous to chicken 2 and Z, and the large differential region on X 1 seems mostly homologous to chicken 3 and 12. These differential regions are completely different from those of the therian X and Y chromosomes, indicating again that the monotreme and therian sex chromosome systems have different origins. We believe that the comparative mapping results reported here will be useful in the continuing search for the monotreme sex determining switch, and in future studies on sex chromosome evolution and dosage compensation mechanisms.
It will be instructive to extend the genome comparison between birds and monotremes to other amniotes, such as snakes and lizards. These comparisons will enable the construction of the ancestral karyotype of sauropsids and mammals and reveal the chromosome evolutionary events that occurred at the origin of the sauropsid and mammalian lineages.
Idiogram showing location of genes in platypus Figure 12 Idiogram showing location of genes in platypus. Gene names in pink are human X-linked genes, gene names in green are homologues of genes imprinted in mouse, gene names in blue are homologues of genes in the mammalian sex determining pathway, gene names in black are Sox gene orthologues, and genes in grey are other previously mapped genes. Gene names in red under a chromosome are mapped in this report by PCR only. Gene names in red next to a chromosome are mapped in this report by PCR and BAC-clone FISH (DMRT1 mapped previously [6,8]). The numbers on the left refer to the gene location in human. The location in chicken is indicated as well, for example, FST located on platypus 1p is on human 5q, chicken Z.

Chromosome paint generation
Primary fibroblast cultures from the short-beaked echidna (T. aculeatus, 2n = 63 male, 64 female) were established routinely in standard medium at 32°C (AEEC permit no. R.CG.07.03 and AEC permit no S-049-2006, NSW P&W permit S10443). Flow sorting, chromosome paint production and FISH were performed according to the protocol described previously [7,32]. Platypus (2n = 52) chromosome paints and metaphases were generated as previously described for the characterization of the platypus sex chromosome complement [7].

Preparation of meiotic cells
Meiotic cells were obtained from animals captured at the upper Barnard river, New South Wales, Australia during breeding season (AEEC and AEC permits to FG as above). The captured animals were euthanased with an intraperitoneal injection of pentobarbitone sodium (Nembutal, Boehringer Ingelheim, NSW, Australia) at a dose of 0.1 mg/g body weight. Meiotic cells were obtained by disaggregating the testis. The material was either directly fixed in methanol/acetic acid (3:1) or incubated in 0.075 KCl M at 37°C as hypotonic treatment to improve spreading of metaphase cells and then fixed.

Fluorescence microscopy
Images were captured using the Leica QFISH software (Leica Microsystems, Milton Keynes, UK) and a cooled CCD camera (Photometrics Sensys, Photometrics, Tucson, AZ, USA) mounted on a Leica DMRXA microscope equipped with a 63×, 1.3 NA objective. Cy3, FITC 9 (fluorescein isothiocyanate) and DAPI (4',6-diamidino-2-phenylindole) signals were captured separately as 16 bit black and white images, and merged to a color image. The DAPI image was enhanced with a spatial filter to obtain enhanced chromosome bands. All image processing was performed with Leica CW4000 software.

Characterization of the short-beaked echidna karyotype
The Tac paints produced were hybridized to male (three individuals) and female (one individual) Tac metaphase preparations. Multicolor chromosome painting was used to ensure that different peaks represent different chromosomes, to define the order of the unpaired chromosomes and to determine the homologous parts that link these chromosomes in the meiotic chain.

Mapping chicken-human chromosome 5, 9, 8, 18 homologous genes on platypus and echidna chromosomes
Twenty-nine genes were mapped to platypus chromosomes (Table 1), and eight of these to echidna chromosomes. Two methods, PCR and BAC-clone mapping (indicated by 'a' and 'b' in Table 1) were used to localize the platypus homologues.

PCR
A human or chicken exon of the specific gene was blasted to find alignments with the NCBI trace archives of platypus using discontiguous megablast. The alignment was used to design platypus specific primers using PrimerQuest [35]. The primers were used first to amplify pools of chromosome specific DNA, and second to amplify chromosome specific DNAs of the positive pool. The exon was considered to be mapped to a single chromosome if only one pool was positive and only one 'chromosome' in that pool was positive. The size of the PCR product was checked to verify that it was as expected from the primer design section, and the product was sequenced for confirmation.
The sequence of the PCR products was blasted to find alignments with the Ensembl Platypus Ornithorhynchus anatinus database release 5. The platypus contigs in the database contain several predicted genes, which were identified by blasting to find alignments with the NCBI human genome database. Homologues of these genes were subsequently localized in chicken by BLAST alignment in Ensembl Chicken. Rens et al. R243.19 Genome Biology 2007, 8:R243

BAC-clone mapping
The above PCR product was used to screen a platypus BACclone library (Oa-Bb, Clemson University, South Carolina, USA). Positive clones were labeled by nick translation and positioned on platypus chromosomes by FISH. The presence of the target gene in the BAC clones above was confirmed by sequencing using the same gene specific primers.

Chromosome homology by comparative gene mapping
Comparative gene mapping was used for the assessment of chromosome homology. It is important to consider whether the homologous genes are likely to be true orthologues. The definition of orthologues is two genes from two different species that derive from a single gene in the last common ancestor of the species [36]. Absolute proof of orthology is difficult on this criterion and was not pursued in this report. Instead, support for chromosome homology was provided when homologous genes were together within the same contiguous region in both species. Instances of non-orthology in the syntenies may exist but the likelihood of multiple, independent exceptional events is reduced when several gene homologues are found in one region.

Authors' contributions
WR designed and performed most of the experiments and analyzed the data. PCMOB sorted the platypus and echidna chromosomes, FG and ETA undertook the meiotic analysis, OC, DG, VAT, HS, MCW, and FV performed other experi-