Introduction

SRY-positive 46,XX testicular disorders of sex development (46,XX-TDSD) (previously known as XX maleness) is a relatively rare condition resulting from unbalanced Xp;Yp translocations during paternal meiosis.1, 2 This condition is associated with normal male external genitalia in 80–90% of patients and undermasculinized external genitalia in the remaining 10–20% of patients.1, 2 Since undermasculinized genitalia are prone to occur when translocated Yp materials are small (<100 kb of the Y-differential region),3 it has been suggested that spreading of X-inactivation into SRY or change in the SRY position relative to chromosomal environments (position effect) has compromised SRY expression, leading to undermasculinization.1, 2, 3 By contrast, when large Yp materials (several Mb of the Y-differential region) are translocated onto Xp, the presence of abundant Yp materials would have protected SRY from being subjected to X-inactivation or position effect, permitting normal male development.1, 2, 3 However, some exceptional patients exhibit normal male genitalia under translocations of small Yp materials,4, 5 and other exceptional patients show undermasculinized external genitalia under translocations of relatively large Yp materials.3, 6, 7 Thus, a straightforward explanation is difficult for the variable sex development in 46,XX-TDSD.

The Xp;Yp translocations can be caused by non-allelic homologous recombination (NAHR).8 In particular, previous studies have revealed the recurrent NAHR events between X-Y homologous sequences around PRKX and PRKY, especially at the ‘hot spot A’ on the 5′ sequence and the ‘hot spot B’ on the C-terminal coding sequence, in a subgroup of males with a common ∼3.5 Mb paracentric Yp inversion mediated by Y-specific long inverted repeats.9 Since PRKX and PRKY are aligned in the same direction during paternal meiosis in such Yp inversion-positive males, this would have permitted the occurrence of NAHR around PRKX and PRKY.10 However, the positions of Xp and the Yp breakpoints are variable among patients, and the genomic basis for Xp;Yp translocations remains to be clarified in most patients, especially in those born to Yp inversion-negative males.

Here, we report on molecular studies in four patients with SRY-positive 46,XX-TDSD, and discuss on variable sex development and genomic rearrangements in SRY-positive TDSD.

Materials and methods

Patients and ethical approval

We studied hitherto undescribed four patients with 46,XX-TDSD (cases 1–4) in whom fluorescence in situ hybridization delineated SRY on the tip of one of the two X chromosomes and direct sequencing showed normal SRY sequence. This study was approved by the Institutional Review Board Committee of Hamamatsu University School of Medicine, and performed after obtaining written informed consent from the parents of the child subjects (cases 1, 2 and 4) and from the adult subject (case 3).

Samples and primers

Molecular studies were performed using peripheral leukocytes of cases 1–4 and the parents of cases 1 and 2. The primers utilized in this study are shown in Supplementary Table 1.

Array comparative genomic hybridization analysis

Oligonucleotide-based array comparative genomic hybridization (CGH) was performed, using a custom-build array containing 522 888 probes for an ∼67-Mb distal Xp region and 87 464 probes for an ∼15-Mb distal Yp region, as well as ∼10 000 reference probes for other chromosomal regions. In case 2 with branchial arch syndrome phenotype, we also performed array CGH using the Agilent G4447A Sure Print G3 Human CGH 1 × 1M Oligo Microarray kit (Agilent Technologies, Santa Clara, CA, USA) containing 1 million catalog probes for the whole genome. Obtained copy number variants/polymorphisms were screened with the Agilent Genomic Workbench software (Agilent Technologies) using the Database of Genomic Variants (http://dgv.tcag.ca/dgv/app/home). The procedure was as described in the manufacturer’s instructions.

Determination of the Xp;Yp translocation fusion point

We performed long PCR amplification using multiple primers that were designed to hybridize to single copy sequences around the breakpoints indicated by array CGH. Single copy sequences were identified by Repeatmasker (http://www.repeatmasker.org). When long PCR products were obtained, they were sequentially sequenced, and obtained sequences were compared with the X and the Y reference sequences (GRCh37/hg19, http://genome.ucsc.edu/). The homology between the Xp and the Yp sequences harboring the breakpoints was examined by Blast search (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

X-inactivation analysis

X-inactivation pattern was analyzed by the methylation assay for AR (exon 1), as reported previously.11 The parental origin of the two X chromosomes was determined by genotyping for CAG repeat length polymorphism within the amplified segments before HpaII digestion.

Results

Clinical findings

Clinical findings are summarized in Table 1. Case 1 had undermasculinized external genitalia and was raised as a male, after thorough consultation with the parents. He received testosterone therapy for microphallus (a small penis associated with other external genital abnormalities such as hypospadias)12 during infancy and surgical operation for the correction of hypospadias and orchidopexy at 16 months of age. Macroscopic examination at the time of operation showed apparently normal testes and a uterine remnant; the uterine remnant was not removed, because there were no clinical problems such as infection. Testicular biopsy for microscopic examination was refused by the parents. Case 2 had male external genitalia with micropenis (a small penis without other external genital abnormalities)12 that required testosterone therapy. Cases 3 and 4 exhibited normal male external genitalia, while adult case 3 had apparently small testes. Case 2 also had branchial arch syndrome phenotype. The reason for karyotyping was variable in cases 1–4. Endocrine studies revealed apparently normal findings with sufficiently elevated basal serum testosterone value during mini-puberty in case 1, age-appropriated data in case 2 and relative hypergonadotropic hypogonadism in adult case 3.

Table 1 Clinical and molecular findings in cases 1–4 with 46,XX-TDSD
Full size table

Array CGH analysis

The results are shown in Figure 1 and summarized in Table 1. Case 1 had two copies of the Xp/Yp short arm pseudoautosomal region (PAR1), deletion of an ∼3.8-Mb X-differential region with the breakpoint proximal to NLGN4X, and translocation of an ∼0.75-Mb Y-differential region with the breakpoint proximal to ZFY. Case 2 had two copies of the PAR1, deletion of an ∼1.0-Mb X-differential region with the breakpoint on PRKX, and complex translocations of an ∼3.8-Mb and an ∼2.7-Mb Y-differential regions with the middle breakpoint on PRKY and the centromeric and the telomeric breakpoints around the long inverted repeats for the common ∼3.5 Mb paracentric Yp inversion. Cases 3 and 4 had an increased copy number of a proximal part of the PAR1 and a normal copy number of a distal part of the PAR1, and possessed the entire Xp-differential region and a very distal part of the Yp-differential region, with the different breakpoints. In case 2, no other copy number variation, which was not registered in the Database of Genomic Variants, was detected by the whole genome catalog array.

Figure 1
figure 1

Representative results of array CGH analysis. The black, the red, and the green dots denote signals indicative of the normal, the increased (log2 signal ratio>+0.5), and the decreased (log2 signal ratio<–0.8) copy numbers, respectively. The log2 signal ratios of +0.5 and –1.0 indicate the presence of three copies and a single copy of the corresponding regions, respectively. The segments highlighted with red, blue, yellow, and green rectangles represent the deleted X-differential region, the translocated Y-differential region, the PAR1 with the normal copy numbers, and the PAR1 with increased copy numbers, respectively.

Full size image

Determination of the Xp;Yp translocation fusion point

Long PCR products harboring the fusion points were obtained in cases 2 and 4, and the precise translocation fusion points were determined by direct sequencing (Figure 2); in case 2, long PCR amplification was carried out on the assumption of paternal Yp inversion. The fusion point in case 2 resided within an 87-bp homologous segment at intron 4 of PRKX and PRKY. The fusion point in case 4 was associated with 18 nucleotides of unknown origin that was inserted between the ‘A’ nucleotide at the position 2 453 897 bp from the Xptel and the ‘A’ nucleotide at the position 2 727 998 bp from the Yptel. There was no sequence homology between the Xp and the Yp breakpoint regions in case 4.

Figure 2
figure 2

Precise translocation fusion points. The translocation fusion point in case 2 is located within an X-Y homologous 87-bp segment (indicated by a yellow rectangle) between PRKX- and PRKY-specific nucleotides that are highlighted in red and blue, respectively; although only a single PRKX-specific nucleotide is shown in this figure, multiple PRKX-specific nucleotides are identified in the more centromeric portion. The translocation fusion point in case 4 resides between the X- and the Y-specific sequences that are highlighted in red and blue, respectively, and is accompanied by an 18 nucleotide sequence of unknown origin (indicated by a green rectangle).

Full size image

In cases 1 and 3, long PCR products were not obtained, probably because of the presence of repetitive sequences around the Yp breakpoint in case 1 and the Xp breakpoint in case 3; in addition, the Xp breakpoint in case 3 resided on a copy number variant region.

X-inactivation analysis

The two X chromosomes were randomly inactivated in cases 1–4, although the paternally derived X chromosome harboring SRY was somewhat preferentially inactivated in cases 1 and 2 (Table 1; Supplementary Figure 1).

Discussion

The present study indicates that the Xp;Yp translocations occurred between the X- and the Y-differential regions in case 1, between PRKX and inverted PRKY in case 2, and between the X-chromosomal PAR1 and the Y-differential regions in cases 3 and 4 (Figure 3). This suggests that the Xp;Yp translocations can take place between various Xp and Yp regions, and provides further support for a critical role of a common paracentric Yp inversion in the occurrence of the PRKX/PRKY-mediated Xp;Yp translocation.10 Notably, the combination of the relatively large Xp deletion and the relatively large Yp translocation observed in case 1 has not been described previously, although the translocations between homologous sequences encompassing PRKX and PRKY observed in case 2 and between the Xp-PAR1 and the very distal part of the Yp-differential region observed in cases 3 and 4 have been reported previously.3, 10, 13 Furthermore, the results of case 2 represent that the PRKX/PRKY-mediated Xp;Yp translocation can occur outside the ‘hot spot A’ and the ‘hot spot B’.

Figure 3
figure 3

Schematic representation of the aberrant translocations. The yellow, the orange, the blue, and the green segments denote the PAR1, the X-differential region, the Y-differential region, and the Y-differential inverted region, respectively.

Full size image

Clinical and molecular findings in this study argue against the notion that undermasculinization is prone to occur when translocated Yp materials are small. The translocated Yp material was relatively large in case 1 with undermasculinized external genitalia and quite small in case 4 with normal male genitalia. In addition, while the SRY-positive paternally derived X chromosome was somewhat preferentially inactivated in case 1, previous studies have indicated no correlation between the degree of masculinization and X-inactivation patterns.3, 14 Indeed, if X-inactivation is spread into SRY, then this would cause ovotesticular disorders of sex development and undermasculinized genitalia under random X-inactivation. One may argue that the location of Xp breakpoints have an important role in sex development, because SRY might escape or undergo X-inactivation when the Yp materials are attached to normally active or inactive X regions, respectively. However, the Xp breakpoints of cases 1–4 resided in the regions that primarily, if not completely, escape X-inactivation.15

Thus, the underlying primary factors for undermasculinization may be variable among patients with SRY-positive 46,XX-TDSD, including spreading of X-inactivation and position effect. In this regard, case 1 had macroscopically normal testes and sufficiently elevated basal serum testosterone during mini-puberty, although defective testis formation and resultant testosterone hyposecretion are postulated for undermasculinization in SRY-positive 46,XX-TDSD.16 It might be possible, however, that SRY expression was transiently dysregulated during fetal life because of a position effect caused by an altered three-dimensional chromatin structure that is more or less specific to the Xp;Yp translocation in case 1.17 Such possible transient dysregulation of SRY expression might also be relevant to micropenis in case 2. Furthermore, since postnatal endocrine status can grossly be normal in patients with mutations or deletions of testis formation genes such as DMRT1 and NR5A1 (SF1),18, 19 the apparently normal endocrine data would not necessarily exclude the macroscopically undetectable defective testis structure in case 1. Similarly, while case 4 exhibited apparently normal external genitalia, this would not necessarily exclude the possibility of impaired testis formation and resultant testosterone hyposecretion of a certain degree. Since the translocations of small Yp materials onto the Xp-PAR1 are associated with both apparently normal external genitalia and undermasculinized external genitalia,3, 13 diverse genital phenotype may be regarded as a continuous spectrum.

The precise Xp;Yp translocation fusion points were determined in cases 2 and 4. Although the PRKX/PRKY-mediated translocation in case 2 is ascribed to NAHR, the translocation in case 4 occurred between non-homologous regions and was associated with an 18-bp sequence of unknown origin. Such genomic structure at the fusion point is characteristic of non-homologous end joining that can lead to various genomic rearrangements including translocations.8 It is likely, therefore, that Xp;Yp translocations are caused by several mechanisms for genomic rearrangements, although Xp;Yp translocations resulting from replication-based mechanisms such as fork stalling and template switching and microhomology-mediated break-induced replication remain to be identified at present.8

In addition to 46,XX-TDSD, case 2 had branchial arch syndrome. Thus, case 2 would have a branchial arch syndrome-related genetic aberration that could not be identified by the whole genome array CGH. In this regard, an X-linked recessive form of branchial arch syndrome has been reported,20 although there is no other report describing branchial arch syndrome in patients with 46,XX-TDSD. Thus, it might be possible that the gene for branchial arch syndrome was lost from the paternally derived X chromosome, and that the homologous gene on the maternally derived X chromosome was mutated or selectively inactivated in tissues expressing that gene.

In summary, the results exemplify the complexity of underlying factors involved in sex development in 46,XX-TDSD, and imply that the Xp;Yp translocations are caused by several mechanisms including NAHR and non-homologous end joining. For sex development in 46,XX-TDSD, while the overall data currently available would still support the notion that undermasculinization in 46,XX-TDSD is prone to occur when translocated Yp materials are small, several exceptional patients have been identified including cases 1 and 4 in this study. Further studies will permit to clarify the underlying factors for variable sex development and genomic basis of Xp;Yp translocations in 46,XX-TDSD.