Statistics from Altmetric.com
During male meiosis the X and Y chromosomes pair along much of their length, with a single obligatory recombination event usually occurring in the pseudo-autosomal region (PAR) at the tip of Xp and Yp,1 thus maintaining identity of the sex chromosome PARs.2
Occasionally illegitimate crossover occurs outside the PAR, resulting in the transfer of Y specific sequences onto the X chromosome. Such translocations between distal Xp and Yp occur relatively frequently, resulting in the generation of 46,XX individuals, the majority of whom display an overtly male phenotype due to transfer of the SRY gene onto the short arm of the paternal X.3–5 However, a small number of Yp translocations are associated with hermaphroditism, defined as the presence of both testicular and ovarian tissue in the same individual.6–8 While the size of the translocated Yp fragment in XX males is variable,9 a recombination hotspot defined by the X/Y homologous genes PRKX and PRKY accounts for approximately one third of cases.10,11 A common inversion polymorphism in proximal Yp flanks this recombination hotspot.12–14
In XX individuals one of the two X chromosomes is inactivated in early embryonic development as a mechanism of dosage compensation for sex linked genes.15 This results in the conversion of the inactive X to a heterochromatic state and the transcriptional silencing of most of the genes upon it.16 Studies of X;autosome translocations have demonstrated that the X inactivation signal is also capable of spreading into cis linked chromatin in a variable manner,17–19 and it has been proposed that spreading of X inactivation into the translocated Yp segment carrying the SRY gene could account for the incomplete masculinisation which is occasionally observed in individuals with X/Y translocations.3,7,8 While evidence to support this view comes from studies of the Sxr mouse,20,21 only a single murine study has examined the spreading of X inactivation into Y chromatin,22 and no such studies have been performed in humans.
It is also becoming clear that chromosomal rearrangements can in themselves result in a disruption of normal gene expression, a phenomenon known as “position effect”.23 Studies of an increasing number of human diseases, including aniridia,24 campomelic dysplasia,25 and X linked deafness,26 have shown that rearrangements, some located up to 900 kb from the affected gene, can result in transcriptional repression, perhaps by removing essential regulatory elements or by alterations of local chromatin structure.
Here we describe detailed studies of 15 individuals with segments of Yp translocated onto the distal short arm of the X chromosome. In each case we have mapped X and Y breakpoints, determined X inactivation ratios, analysed the expression and methylation status of translocated Y chromosome genes, and performed phenotype/genotype correlations. Our analysis finds little evidence for spreading of X inactivation into Yp chromatin, and instead suggests that position effects resulting from chromosomal rearrangement account for the occurrence of incomplete masculinisation in some individuals with X/Y translocation.
We have studied causes of incomplete masculinisation in 15 individuals with segments of Yp translocated onto distal Xp. In each case we mapped X and Y breakpoints, determined X inactivation ratios, and performed expression studies of translocated Y genes by allele specific RT-PCR. We confirm the presence of a Yp inversion polymorphism predisposing to X/Y translocation.
Expression studies found very little evidence for spreading of X inactivation into Yp chromatin. However, in several cases disruption of gene expression occurred independently of X inactivation, suggesting “position effects” resulting from chromosomal rearrangement.
In particular, five of the six translocation carriers studied with an intersex phenotype showed either translocation breakpoints very close to SRY, or disrupted expression of genes near SRY in a manner unrelated to X inactivation. Furthermore, Southern analysis suggested the presence of a cryptic rearrangement 3–8 kb proximal to SRY in one such case.
Overall, our observations suggest that incomplete masculinisation in cases of X/Y translocation is a result of disruption of normal SRY expression by position effect rather than X inactivation.
The 15 patients included in this study were ascertained from a number of sources, and several have been described previously. Cases 1–4 and 7 are PG, AK, RZ, JW, and KM, respectively, in Kusz et al8; further analyses detailing the familial nature of the translocations in cases 1 and 2 are also described in Sharp et al27; case 5 is described in Coles et al28; and case 15 is patient 3 in Lindsay et al.29 With the exception of case 14 (a hermaphrodite), all the remaining patients presented with an overtly male phenotype or Klinefelter syndrome. Thus, patient referrals ranged from classic “XX male” to true hermaphroditism with ambiguous genitalia (fig 1).
Fibroblast and EBV transformed lymphoblastoid cell lines were established (cases 4 and 14) or obtained from the European Collection of Animal Cell Cultures, Porton Down, Wiltshire, UK (case 5, DD0144; case 8, DD3000; case 11, DD0053; case 12, DD0249; case 15, AL0049) and grown according to standard protocols. In all other cases samples of peripheral blood were obtained.
The translocated region of Y was determined in each case by the presence or absence of STS markers by PCR (fig 1). Parallel amplification of male and female control DNA confirmed Y chromosome specificity for each marker. X chromosome breakpoints were mapped by PCR amplification of polymorphic microsatellite markers (fig 1) in each proband and their parents. Breakpoints were assigned between the most proximal informative deleted marker and the most distal informative non-deleted marker. Where no parental DNA samples were available (cases 3, 6, 10, and 12–15) breakpoints were assigned distal to the most proximal heterozygous marker.
X inactivation ratios were determined by duplicate analysis using the AR gene PCR assay as described previously.30 This assay has shown to be both highly reproducible30 and quantitative over a wide dynamic range (A Sharp, unpublished data). Where this analysis was not informative, X inactivation ratios were assessed cytogenetically by combined detection of late replicating DNA and FISH using the probe pHuR14, specific to SRY.18 For each case, at least 50 cells were scored for late replication, corresponding to inactivation, of the normal X or the der(X;Y).
Allele specific semi-quantitative RT-PCR of PAR and X/Y homologous genes was performed as described previously.18 This method utilises transcribed sequence divergences or polymorphisms between X/Y homologous genes to differentiate the X and Y derived transcripts, allowing their relative levels of transcription to be quantified. For each X/Y homologous gene, pairwise BLAST alignment (http://www.ncbi.nlm.nih.gov/blast/) of mRNA reference sequences (http://www.ncbi.nlm.nih.gov/Web/Genbank) identified sequence divergences between the X and Y homologues. Webcutter 2.0 (http://www.firstmarket.com/firstmarket/cutter/cut2.html) was used to identify divergences which altered a restriction site, and spanning primers designed using Primer3 (http://www-genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi). All primers were located in regions of perfect X/Y identity to ensure similar amplification efficiency from the X and Y homologues. For genes located within the Yp PAR, putative transcribed single-nucleotide polymorphisms were identified from LocusLink (http://www.ncbi.nlm.nih.gov/LocusLink/) and screened by PCR, restriction digestion, and agarose gel electrophoresis for heterozygosity in each case. Details of each expression assay are shown in table 1.
For allele specific RT-PCR, RNA was extracted from peripheral blood, lymphoblast, or fibroblast cell lines using TRIZOL (Gibco BRL, Paisley, UK), DNAse treated (Promega, Southampton, UK), and cDNA synthesised using M-MLV RTase (Gibco BRL) with gene specific reverse-strand primers. Two rounds of PCR were performed using unlabelled and labelled primers, respectively, to avoid heteroduplex formation,31 which were digested and the products resolved using an ABI 377 sequencer for quantification of alleles by peak height. A gene was scored as inactive when there was a reduction in intensity of the Y specific allele, with the allele ratio gained using cDNA of the proband significantly different from that obtained using control cDNA (fig 2). Such results indicate that the copy on the der(X;Y) is inactive (fig 2). Similarly, a gene was scored as active when results obtained using cDNA of the proband were similar to those gained using control cDNA with no reduction in intensity of the Y specific allele.
Methylation analysis of CpG islands located at the 5′ end of SRY, ZFY, and PRKY was performed by restriction enzyme digestion with HpaII, CfoI, or MspI and subsequent PCR, as described previously.19
Probes for Southern blot analysis were generated by Expand Long Range PCR (Roche, Lewes, UK) using the following primer sequences: SRY1 for cgttccccggtactactgaa, SRY1 rev catcttgaaaggacatttgactc (4–6 kb 5′ of SRY); SRY2 for gcaggtcaccaaagacacaa, SRY2 rev tgaatcacattttcactggga (2–4 kb 5′ of SRY); ZFY for attgttgcttagtttgcatcccaga, ZFY rev taccaaatatgcacaggtttgaggg (spanning exon 2 of ZFY); RPS4Y for gagaggaactgggaaaaggtgtca, RPS4Y rev ggatcttgccttctacaagctcca (30–32 kb 3′ of RPS4Y).
To determine the orientation of a polymorphic inversion sequence within Yp we used the STS markers sY1242 and sY1243, which are specific to repeat sequences flanking the inversion (J Lange, H Skaletsky, and D Page, personal communication). Because the Yp breakpoint in cases 7–14 lay within the inversion region, its orientation could be assigned based on the presence of either one of the flanking repeats. Primer sequences are as follows: sY1242 for cgtcggtattttacgacacg, sY1242 rev gcatttgtttttcatgtgcg; sY1243 for atctgcacacttgggtaggc, sY1243 rev gaggaaatgcagaatttggg.
Although each of the 15 Yp translocations included SRY, as confirmed by positive PCR amplification using primers spanning the coding region of the gene, sexual phenotype varied. Patient referral reason and more detailed clinical data, where available, are shown in fig 1. While the majority of individuals studied presented with a clinical diagnosis of Klinefelter syndrome, four individuals (cases 1, 2, 3, and 14) were hermaphrodites, presenting with both male and female gonadal tissue and ambiguous genitalia, and two others (cases 5 and 7) had penile hypospadias. One individual (case 15) presented with MLS syndrome, resulting from functional nullisomy for the HCCS gene in Xp22.31.32
There was considerable variation in the size of the translocated Yp segment in the cases analysed (fig 1). While in one patient (case 15) almost the entire short arm of the Y chromosome was translocated onto distal Xp, in other individuals (cases 1 and 2) the Yp breakpoint occurred less than 20 kb proximal to the pseudoautosomal boundary. In two of the hermaphrodites the Y breakpoint occurred 6–7 kb proximal to SRY, and 63–70 kb proximal to SRY in a third case. However, there was clearly a hotspot for illegitimate X/Y recombination close to PRKX and PRKY, consistent with previous observations.10,11 In cases 7–14 the Yp breakpoints clustered within a region of <500 kb around PRKY, and the corresponding Xp breakpoints all occurred close to PRKX in Xp22.33. Similarly, where the Yp breakpoints were very distal (cases 1–3), Xp breakpoints occurred within the PAR, resulting in a der(X;Y) chromosome carrying two copies of the proximal part of the PAR.4
X inactivation ratios
There was considerable variation in the proportion of cells in which the der(X;Y) was inactive between different cases (fig 1). In the majority of cases, X inactivation ratios were determined using samples of peripheral blood. However, for several patients the only samples available were immortalised cell lines. These can show very different X inactivation ratios from those in vivo, as observed in cases 4, 5, and 8, where both fresh tissue and cultured cell lines were available and gave discordant results.
Expression analysis of translocated Y genes
Results of expression analysis of four PAR genes (PGPL, DXYS155E, SLC25A6, and MIC2) and three X/Y homologous genes (RPS4Y, ZFY, and PRKY) are shown in fig 1. SRY was not analysed because of its tissue limited expression. In addition, no RNA was available for cases 2, 6, 9, 10, and 13. Except in three patients (cases 7, 14, and 15) all Yp genes assayed were normally expressed.
In case 15 PRKY expression was reduced to <10% of normal levels. As X inactivation ratio analysis showed the der(X;Y) to be inactive in 94% of cells assayed, this suggested silencing of PRKY on the der(X;Y). Furthermore, analysis of a CpG island located at the 5′ end of PRKY using methylation sensitive restriction enzymes19 demonstrated the presence of high levels of methylation in case 15 which were not detected either in cases 11–14 or control individuals (fig 3), indicating that silencing of PRKY resulted from a spreading of X inactivation into the translocated Yp segment. However, PRKY expression was normal in cases 11, 12, and 14.
ZFY expression was disrupted in two individuals (shown in fig 2), although in neither case did this appear to be related to X inactivation. No ZFY transcripts were detectable in case 7, despite the der(X;Y) being active in a proportion of cells, and ZFY was expressed at approximately 50% of normal levels in case 14, with an X inactivation ratio of 96:4. Additionally, analysis of a CpG island at the 5′ end of ZFY did not detect any methylation differences between these two patients and 46,XY males or other X/Y translocation carriers in whom ZFY expression was normal (data not shown).
Southern blot analysis of Yp11.31
Because cases 7 and 14 presented with both reduced ZFY expression, which was apparently not due to spreading of X inactivation, and also intersex phenotype indicative of abnormal SRY expression, we hypothesised the presence of additional cryptic rearrangements within the translocated Yp11.31 region which disrupt both ZFY and SRY expression in these individuals. Consistent with this hypothesis, Southern analysis using a probe specific to the 5′ region of SRY hybridised to a novel NdeI fragment in case 7 which was not observed in other X;Y translocation carriers or normal controls (fig 4), suggesting the possible presence of a cryptic rearrangement 3–8 kb proximal to SRY. Hybridisation patterns in all other cases were normal.
Analysis of a Yp inversion polymorphism
There is an inversion polymorphism of ∼4 Mb in proximal Yp (fig 1) flanked by IR3 inverted repeats.13,14 Using STS markers specific to these flanking repeats (J Lange, H Skaletsky, and D Page, personal communication) we determined the orientation of the inversion in those patients in whom the Yp breakpoint lay within the inverted region (cases 7–14). All eight individuals studied carried the inversion. As previous haplotype analysis indicates the inversion is present on approximately one third of European Y chromosomes (27/81 normal males),12 its prevalence in Xp/Yp translocations is significantly increased (p = 0.0003, Fisher’s exact test), confirming the suggestion of Jobling et al12 that the inversion predisposes to Xp/Yp translocation.
Because X/Y translocations move SRY onto the X chromosome, spreading of X inactivation provides an attractive hypothesis to explain the variable sexual phenotype associated with Xp/Yp translocations.3,7,8 While studies of the Sxr mouse support this hypothesis,20,21 evidence from the study of human X/Y translocations is limited.
Our analysis allowed this hypothesis to be examined in more detail. If correct, we expected two criteria to be fulfilled: (i) patients with incomplete masculinisation would exhibit preferential silencing of the der(X;Y) when compared to classical XX males, and (ii) expression studies of translocated Yp genes would reveal evidence for spreading of X inactivation into Y chromatin. However, our data gave little support for either of these predictions.
We observed no apparent relationship between sexual phenotype and X inactivation ratios in the samples available for analysis. In particular, we did not find preferential silencing of the der(X;Y) in patients with incomplete masculinisation, as might be predicted from studies of the Sxr mouse.20,21 It should be noted, however, that in most cases the only tissue available for X inactivation analysis was peripheral blood, and it is possible that this may not be representative of other tissues.30
Secondly, our expression studies found very little evidence for spreading of X inactivation into Yp. Unfortunately direct study of SRY was not possible due to its tissue limited expression, and efforts to utilise CpG island methylation as a measure of the spread of X inactivation were inconclusive because of the apparent presence of SRY methylation in normal males (data not shown). While it is known that X inactivation may spread discontinuously and “skip” some genes and silence others located more distally,19 evidence for the silencing of other genes surrounding SRY by a spreading of X inactivation was almost completely lacking. Expression analysis of translocated Y chromosome genes showed that the vast majority were normally expressed from the der(X;Y), apparently suggesting that the ability of X inactivation to spread into Y chromatin is relatively limited when compared to X;autosome translocations.17–19 However, it should be noted that all of the genes we tested have X linked homologues which escape the spread of X inactivation, potentially biasing our results. Because of their high degree of sequence similarity, it might be predicted that these X/Y homologous genes, and in particular the PAR genes which are identical on the X and Y, would be resistant to the X inactivation signal. Thus our analysis may underestimate the true ability of X inactivation to spread through Y chromatin.
The inconsistency of our observations with the initial hypothesis led us to consider alternative explanations that could account for the occurrence of incomplete masculinisation in some cases of X/Y translocation. The clear association between translocation breakpoints in close proximity to SRY and the hermaphroditism observed in cases 1–3 suggested that disruption of sex determination might instead result from position effects interfering with normal SRY expression. In support of this hypothesis, a review of the literature revealed that almost all reported SRY positive 46,XX true hermaphrodites that have been studied molecularly have been found to have breakpoints very close to SRY.6,8,33–35 Furthermore, previous studies in both man and mouse have found that deletions of sequences flanking SRY, some of which overlap with the location of the translocation breakpoints in cases 1 and 2, are associated with sex reversal,36–38 providing strong evidence that position effects can disrupt SRY function.
Although the three other studied patients with incomplete masculinisation (cases 5, 7, and 14) all had translocation breakpoints distant from SRY, thus making it highly unlikely that the translocation breakpoint itself directly disrupts SRY, cases 7 and 14 both showed disrupted expression of ZFY (located ∼170 kb proximal to SRY) in a manner unrelated to X inactivation. In case 7 ZFY expression was completely abolished despite the fact that the der(X;Y) remained active in a proportion of cells. Such constitutional inactivation excludes X inactivation as the underlying cause, and instead suggests inactivation resulting directly from genomic rearrangement. Similarly case 14 ZFY exhibited reduced expression despite the der(X;Y) being almost exclusively inactive.
Because of this combination of reduced ZFY expression attributable to position effects and intersex phenotype indicative of abnormal SRY expression, we hypothesised the presence of additional cryptic rearrangements within the translocated Yp11.31 region in these two individuals which disrupt both ZFY and SRY expression. Many apparently balanced translocations actually involve more complex rearrangements close to the site of breakage,39–41 and a previous report described a Y;22 translocation associated with an interstitial deletion of SRY and sex reversal.42 Consistent with this hypothesis, Southern blot analysis of distal Yp identified an abnormal hybridisation pattern in case 7, suggesting the presence of a cryptic rearrangement 3–8 kb proximal to SRY in this individual. As our PCR studies found no evidence for any deletion within this region, and RPS4Y showed apparently normal expression, this may represent an inversion within Yp11.31 with breakpoints close to ZFY and SRY. Although identical Southern analysis did not identify any abnormalities in cases 5 or 14, our analysis covered <25% of the ∼170 kb region between SRY and ZFY, and thus the presence of cryptic rearrangements of Yp11.31 in these individuals cannot be excluded. Unfortunately, insufficient DNA from cases 7 and 14 and the very high density of repetitive DNA elements in this region precluded further such experiments. However, overall our data favour the suggestion that incomplete masculinisation in cases of X/Y translocation results from disruption of normal SRY expression by position effect, rather than X inactivation.
We also observed that a polymorphic Yp inversion13,14 occurs significantly more frequently on Xp:Yp translocation chromosomes than in controls, confirming the observations of Jobling et al that the inversion predisposes to Xp/Yp translocation.12 Furthermore, both the X and Y breakpoints on each der(X;Y) carrying the inversion were close to PRKX and PRKY, suggesting the inversion mediates illegitimate X/Y recombination by alignment of these homologous sequences during male meiosis. Consistent with this hypothesis, because PRKY is located within the inverted region it is brought into the same centromere to telomere orientation as PRKX in males carrying the Yp inversion (http://genome.cse.ucsc.edu/), thus aligning the X and Y homologues and providing an obvious mechanism by which the inversion mediates X/Y meiotic exchange. These results indicate that the Yp inversion polymorphism is an underlying factor in generating the PRKX/PRKY recombination hotspot observed in many XX males.10,11
These data represent the most detailed study to date of X/Y translocations, giving insight into the causes and effects of these rearrangements. Although we originally set out to study the spreading of X inactivation into Y chromatin, our results suggest that position effects disrupting gene expression also occur in a significant number of cases. When it is considered that we have only examined a relatively small number of genes on one side of each translocation breakpoint, we suggest that disrupted expression of genes in this way may be a relatively frequent occurrence, even in apparently balanced chromosomal rearrangements. In particular, we propose that incomplete masculinisation in cases of X/Y translocation is a result of disruption of normal SRY expression by position effect, rather than X inactivation.
We would like to thank the patients and their families involved in this study for their helpful co-operation, Julian Lange, Helen Skaletsky, and David Page for sharing unpublished data, and Christine Disteche and Kathleen Leppig for their generous provision of patient samples and clinical data.
This work was supported by grants from The Wellcome Trust (Ref. 058387) and Medical Research Council (Ref. G9801327).
Competing interests: none declared
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.