Identification and characterization of an Xp22.33;Yp11.2 translocation causing a triplication of several genes of the pseudoautosomal region 1 in an XX male patient with severe systemic lupus erythematosus

Abstract

The X;Y translocation break point sequence in an XX male patient with prepubertal systemic lupus erythematosus (SLE) was characterized with the intention of identifying a predisposing gene(s) for SLE. Spectral karyotyping of the patient's metaphase chromosomes showed normal autosomes and 2 X chromosomes, one of which displayed a small portion of the Y chromosome. Using a Y chromosome polymerase chain reaction (PCR) walking strategy and inverse PCR, we found that the abnormal recombination occurred between retroviral long terminal repeats located at Xp22.33 (position 0.95 Mb; inside the pseudoautosomal regions) and Yp11.2 (4.20 Mb) downstream of the sex-determining region Y (SRY) gene. The complete DNA sequence of the break point was determined, revealing a partial duplication of the pseudoautosomal region 1 (PAR1) in the derivative X chromosome and causing a partial trisomy of the 12 known genes located between the interleukin-3 receptor α (IL3RA; position 1.1 Mb on the X and Y chromosomes) and CD99 (position 2.2 Mb) genes inclusively. All other X chromosome genes were present as 2 copies. Real-time quantitative PCR confirmed the presence of 3 copies of each of the 12 genes in the patient's genomic DNA. We also found that RNA for 1 of the candidate genes was indeed overexpressed in the patient's blood as compared with normal subjects. Taken together, the uniqueness of the translocation, the rarity of severe prepubertal SLE in males, and the presence of SLE in some patients with Klinefelter's syndrome (who have a triplication of the 2 PAR regions) point to a possible relationship between the partial triplication of the PAR1 region and the development of SLE.

Systemic lupus erythematosus (SLE; OMIM no. #152700) is a chronic autoimmune disease characterized by multisystem organ involvement (skin, joints, kidneys, and serosal membranes) and the production of autoantibodies directed against various cellular components. SLE occurs in 40–50 persons per 100,000 population, affects primarily adult women, and is rare in children, with only 10–17% of cases diagnosed before the age of 16 years. SLE in children appears to show more severe organ involvement than does SLE in adults (1). Lymphopenia, another typical feature of SLE, correlates with elevated disease activity and higher grades of systemic involvement (2).

Familial aggregation (3), monozygotic twin concordance (4), genetically engineered mutant mice susceptible to SLE (5), as well as the results of many genome scans support a genetic component of the disease. As with most complex traits, a number of susceptibility loci for SLE have been found in different populations, namely, in regions 5p15.3, 19p13.2, 18q21.1 (6), 10q22.3, 2q34–35, 11p15.6 (7), 1q21–22 (3, 8), 16q13 (9), 17p13, and 4p16 (10).

One of the most striking and intriguing particularities of SLE is its female predominance (11). The etiology of this predominance has generally been explained by the hormonal differences between males and females. Estrogens probably play a key role in triggering SLE manifestations (12). However, an X-linked hypothesis has also been suggested as an explanation of the female predominance in SLE. Furthermore, the reported presence of SLE in some male patients with XXY Klinefelter's syndrome and in female patients with 47XXX supports a possible contribution of the X chromosome to the pathogenesis of SLE. In fact, the prevalence of SLE in patients with Klinefelter's syndrome has never been measured, but many case reports have documented the presence of these 2 diseases in the same patients (13–21). This intriguing relationship suggests that overexpression of certain genes located on the X chromosome may predispose to the development of SLE.

Unfortunately, none of the genome scans reported thus far have detected an association between the X chromosome and SLE. However, most of the previous linkage studies have not included markers on the X chromosome, and the X-linked hypothesis for SLE has not been fully investigated. Also, potential susceptibility loci on the sex chromosomes are more likely to be missed because of their particular mode of transmission. If the X-linked hypothesis is true, then other methods should be used to examine the possible implication of the X chromosome in SLE.

We present herein the molecular characterization of a translocation between the X and Y chromosomes occurring in an XX male patient with SLE. The goal of this study was to identify an X-linked gene(s) that predisposes to the development of SLE or autoimmunity, since the analysis of individual patients with translocations affords a unique opportunity to identify causative genes.

CASE REPORT

Clinical history.

The patient, a 19-year-old man, was diagnosed as having SLE in 1992, when he was 6 years old. The fact that the onset of his disease was before pubertal development is evidence against the sex hormone hypothesis of SLE induction. His disease has been characterized by severe rash, severe mucositis, arthritis, and focal proliferative glomerulonephritis. His autoantibody profile has included antinuclear antibody, anti–double-stranded DNA antibody, and anti-Sm antibody positivity. During the course of his disease, he has also developed elevated liver enzyme levels and severe, protracted thrombotic thrombocytopenic purpura. The course of his disease has been further complicated by osteoporosis, with vertebral fractures and impaired growth.

In terms of pubertal development, the patient presented with primary hypogonadism, with partial Leydig's cell failure, suboptimal testosterone secretion, and elevated levels of both luteinizing hormone and follicle-stimulating hormone, as well as low testicular volume and incomplete seminiferous tubule development. Results of growth hormone testing have been normal. His hormone abnormalities have been treated with intramuscular injections of testosterone.

Treatment of the patient's SLE has included corticosteroids, hydroxychloroquine, methotrexate, intravenous pulse cyclophosphamide, azathioprine, and plasmapheresis. In addition, he has required treatment for hypertension and has received pamidronate treatment for osteoporosis.

Pertinent family history includes the presence of idiopathic thrombocytopenic purpura (ITP) in the patient's mother, which was diagnosed when she was 15 years of age. She was treated with corticosteroids and, subsequently, underwent splenectomy. A maternal aunt also had ITP, which was treated with corticosteroids followed by splenectomy. Both his mother and maternal aunt had antinuclear antibodies.

Biologic samples were obtained from the patient after informed consent was signed. The study was approved by the Institutional Review Board.

Findings of genetic analyses.

Karyotype analysis of the patient indicated 46 chromosomes, with 2 apparently normal X chromosomes and no Y chromosome. SkyPaint spectral karyotyping analysis (Applied Spectral Imaging, Vista, CA) confirmed the presence of 2 X chromosomes, but also revealed a t(X;Y) translocation, since 1 of the 2 X chromosomes harbored a small portion of the Y chromosome. No other chromosome abnormality was detected (Figure 1).

Figure 1. Human spectral karyotyping, designed to enable the simultaneous visualization of all human chromosomes, was performed on metaphase chromosomes from the XX male patient with systemic lupus erythematosus to eliminate the presence of other chromosomal translocations. Analysis was performed with the SkyVision Spectral Imaging system. The 22 autosomal chromosomes are intact, and 1 of the X chromosomes harbors a small portion of the Y chromosome.

Normally, in female cells, 1 of the 2 X chromosomes is inactivated in early embryonic life, a process known as X chromosome inactivation. Females are therefore mosaics for 2 cells lines, one with the maternal X as the active chromosome and the other with the paternal X as the active chromosome. The distribution of the 2 cell lines is variable, but is generally close to normal, with a 50:50 ratio. A skewed pattern is defined when >75% of cells express the same X chromosome. The X-inactivation status of the patient was determined on DNA extracted from peripheral white blood cells by analysis of the polymorphic CAG repeat in the human androgen receptor gene (HUMARA) using a method adapted from that described by Busque et al (22). This analysis showed a nearly random inactivation pattern (65% versus 35%), indicating that the translocation was not deleterious and did not cause a negative selection of cells harboring this active translocated X chromosome. Thus, in this patient, genes are expressed by both X chromosomes.

Fluorescence in situ hybridization (FISH) analysis performed using an X centromeric probe and a sex-determining region Y (SRY) chromosome gene probe demonstrated a t(X;Y) translocation, with the SRY gene transferred to the X chromosome, confirming that the translocation occurred downstream of this gene on the Y chromosome (2.30 Mb, according to the July 2003 assembly of the Genome Browser of the University of California, Santa Cruz [UCSC]; available at http://genome.ucsc.edu/). FISH analysis also detected 2 copies of the X chromosome gene steroid sulfatase microsomal, arylsulfatase C, isozyme S (STS), indicating that at this position (6.60 Mb), the patient still possessed material from both X chromosomes and that the translocation break point had to be upstream of this position on the X chromosome (data not shown).

To further refine the positioning of the translocation break point, 54 highly polymorphic microsatellite markers spanning the entire X chromosome were genotyped. Heterozygosity at microsatellite markers indicates the presence of 2 X chromosomes, whereas homozygosity suggests the presence of only 1 chromosome (loss of heterozygosity). An analysis of the patient's genomic DNA revealed heterozygosity for most of these markers, indicating the presence of almost 2 entire X chromosomes. Starting from the short arm (p) telomere, the first heterozygote marker detected was GATA164D10 (position 2.96 Mb). All upstream markers being homozygote, the X chromosome break point had to be located between the telomere and this position, probably in pseudoautosomal region 1 (PAR1). Moreover, GATA164D10 is located upstream of the protein kinase, X-linked (PRKX) gene, which eliminates the possibility that the translocation occurred by homologous recombination between the PRKX and PRKY genes, the usual X;Y translocation break point in XX males that accounts for one-third of all XX males (23). Three markers specific to the Y chromosome were also tested (amelogenin, Y-chromosomal [AMELY] at 6.44 Mb, DYS19 at 9.13 Mb, and DYS389 at 13.62 Mb) and were not detected in the DNA from the patient. Considering the presence of the SRY gene (2.30 Mb; detected by FISH) and the loss of the AMELY gene (6.44 Mb) on the translocated chromosome, we concluded that the Y chromosome translocation break point area was located between these 2 genes.

To determine the exact break point position, we walked the Y chromosome, using more than 50 specific polymerase chain reaction (PCR) markers. Because of the high degree of homology between chromosome Yp11.2 and Xq21.31 (99%), we used BLAST (National Center for Biotechnology Information) and/or BLAT (Genome Browser at UCSC) to compare the X chromosome sequence with numerous BAC sequences located between SRY (2.30 Mb) and AMELY (6.44 Mb). Primers specific to the Y chromosome region were designed. These markers were tested by PCR on DNA obtained from the patient (blood and cultured cells), from normal men (positive controls), and from normal women (negative controls). All markers located between positions 2.29 Mb and 3.98 Mb of the Y chromosome could be amplified in DNA from the patient and from the positive controls, but not in DNA from the negative controls. Markers located between 4.22 Mb and 8.60 Mb were detected in male controls, but not in the SLE patient.

Intensive analysis of the Y chromosome region between 3.98 Mb and 4.22 Mb allowed further characterization of the exact break point position, which was found to be located in a 2,130-bp portion of the BAC DNA sequence AC012077, between positions 4.205 Mb and 4.207 Mb of the Y chromosome. Interestingly, there are no genes in that region, and thus, no genes should be modified or truncated by the translocation. This area contains mainly repeat elements, especially, long terminal repeats (LTRs). The transforming growth factor β–induced transcription factor 2–like, Y-linked (TGIF2LY) gene (OMIM no. *400025) is found on the proximal side of the break point, at 3.15 Mb, and the protocadherin 11, Y-linked (PCDH11Y) gene (OMIM no. *400022) is distal to the break point, at 4.60 Mb. This latter gene was absent in our patient. Whereas the SRY, ribosomal protein S4, Y-linked (RPS4Y), zinc-finger protein, Y-linked (ZFY), and TGIF2LY genes were found in the patient, all other Y chromosome genes were absent.

To determine the precise break point position on the X chromosome, we used an inverse PCR method (24), which allows determination of an unknown sequence of the genome positioned beside a known sequence. In our patient, the known sequence was the 2 kb of the Y chromosome break point region determined by chromosome walking. Inverse PCR has previously been used to sequence break points for translocations or other rearrangements (25). Inverse PCR was performed on the genomic DNA from the patient, as well as on genomic DNA from normal men and women. A 1.6-kb DNA fragment specific to the patient (i.e., not observed in the normal male or female controls) was obtained and sequenced. This DNA fragment corresponded to the exact Y chromosome break point sequence region previously established (4.205 Mb) and contained ∼1 kb of new sequence, which was analyzed and found to be the expected X chromosome portion. We designed a pair of DNA primers, one of which was specific for the Y chromosome (5′-AATACTGTCTTCTCCCCAATG-3′) and the other specific for the X chromosome (5′-CGTCCTTGTCACCGCCAACTG-3′). These were expected to yield an amplification product only in individuals harboring the Xp22.33;Yp11.2 translocation, and indeed, the 1.6-kb amplicon could only be detected in the patient.

Using the same PCR primers and additional nested primers (5′-CTGTCTGCTGCCTGCCCCTGG-3′ and 5′-TCAGACACCAAGCTGTAGAAG-3′), we determined the complete sequence of the translocation (Figure 2). The first 473 nucleotides showed 100% homology with the BAC DNA sequence AC012077, which covers the Yp11.2 chromosome. Variations from this sequence were observed starting at position 474, which is an A→G transition, suggesting that the DNA sequence shifted to the X chromosome at this position. Thereafter, variations were increasingly frequent, confirming the shift in chromosomes. At position 967, a complete change in the homology sequence was observed. In fact, all homology between the X chromosome translocation sequence and the Y chromosome was lost. This sequence is thus specific to the X chromosome.

Figure 2. Alignment of the break point sequence found in the XX male patient with systemic lupus erythematosus with the human chromosome Yp11.2 sequence (GenBank accession no. AC012077). The XX male patient has a 100% Y-specific sequence for the first 473 nucleotides. Starting at position 474, several variations can be observed, indicating that the sequence may have switched to the X chromosome. We expect that the translocation occurred around position 473. Analysis with the RepeatMasker program indicated that 93% of the sequence consists of repeated elements and that the region between positions 1 and 966 is made up of long terminal repeat/human endogenous retroviral sequence K repeat sequences. The rest of the X chromosome break point sequence junction, from position 1,001 to position 1,668, is shown at the bottom. The X chromosome sequence was obtained by using an inverse polymerase chain reaction (PCR) technique. With this technique, genomic DNA from the patient was cut with a restriction enzyme and then circularized by ligation. PCR was performed around the circle using primers from the known sequence that amplified both the known and unknown parts of the template DNA. Sequencing of the PCR product then allowed determination of the break point sequence.

Analysis of the break point sequence using the RepeatMasker program (version 2; available at http://www.repeatmasker.org) indicated that the abnormal X and Y chromosome interchange occurred between repetitive elements, specifically, retroviral LTRs of the human endogenous retroviral sequence K (HERVK). This LTR/HERVK is part of a large family of repetitive elements that comprises ∼1% of the human genome. LTR/HERVK sequences have retained the potential to retrotranspose and, thus, to change genomic structure, function, and plasticity. Thus, the LTR sequences present on Xp22.33 and Yp11.2 caused an unusual homologous recombination event in the germ cells of the patient's father, resulting in the X;Y translocation seen in the SLE patient.

Surprisingly, both BLAST (GenBank) and BLAT (UCSC Genome Browser) failed to recognize the X chromosome break point in the human genome sequence, suggesting that it was located in an unknown region (i.e., a gap region that has not yet been sequenced by the Human Genome Project). From the telomere to the GATA164D10 marker, there are 3 gaps in the X chromosome sequence, all of which are located within the PAR1 region. Presumably, the translocation break point is positioned in one of those gaps. As illustrated in Figure 3, the translocation caused the transfer of the complete Y chromosome PAR1 region to the top of the X chromosome, within its own PAR1 region, resulting in the presence of a duplicated portion of the PAR1 region on the same chromosome. Considering the patient's other normal X chromosome, he therefore harbors a trisomy for that specific genomic region.

Figure 3. Ideogram of A, the X chromosome, B, the Y chromosome, and C and D, the derivative X chromosome present in the XX male patient with systemic lupus erythematosus. The derivative X chromosome harbors a partially pseudoautosomal region 1 (PAR1; black areas). Two copies of the genes in the PAR1 region between interleukin-3 receptor α (IL3RA; 1.100 Mb) and CD99 (2.204 Mb) are found on the derivative chromosome. Arrows indicate the position of the break point on the Xp22.33 and Yp11.2 chromosomes. PRKX = protein kinase, X-linked; STS = steroid sulfatase microsomal, arylsulfatase C, isozyme S; SRY = sex-determining region Y; AMELY = amelogenin, Y-chromosomal; PRKY = protein kinase, Y-linked; SLC25A6 = solute carrier family 25 (mitochondrial carrier) member A6; ASMT = acetylserotonin methyltransferase, X-chromosomal.

The PAR1 region, which is located on the telomeric part of the sex chromosomes, is characterized by complete homology between the X and Y chromosomes (26). It is the only portion of the sex chromosomes that can recombine during male meiosis. The PAR1 region is ∼2.2 Mb in length and contains ∼20 genes.

To determine the gap in which the translocation was situated and to confirm that some genes were present as 3 copies in our patient, an assay using SYBR Green real-time quantitative PCR was developed (adapted from the method described in ref. 27) to map the entire PAR1 region. Thirteen real-time quantitative PCRs covering the PAR1 region and X chromosome–specific regions were performed on the patient and on control DNA samples.

As shown in Figure 4, genes located outside of the PAR1 region (glycogenin 2 [GYG2] and transmembrane 4 superfamily member 2 [TM4SF2]) were present as 2 copies in our patient and in normal women, whereas in normal men, they were present as 1 copy. These results validated our method of discriminating the copy number of genes in the genome (1 in men versus 2 in women).

Figure 4. Graph showing the copy number of different X chromosome genes in the XX male patient with systemic lupus erythematosus and the average copy numbers obtained in 4 normal men and 4 normal women. The assay was developed using SYBR Green real-time quantitative polymerase chain reaction (PCR) run on a TaqMan gene expression assay and an ABI Prism 7000 Sequence Detection system. Thirteen PCRs covering the pseudoautosomal region 1 (PAR1) and X chromosome–specific regions were performed using identical amplification conditions and DNA dilutions. Briefly, the starting copy number was determined for the various target genes and normalized against the starting copy number of a reference gene (pseudoautosomal GTP binding protein–like [PGPL]). This normalized gene dose ratio (N) allows for the determination of the gene number in the genome. For example, knowing that the reference gene is present as 2 copies in the genome, a ratio of N = 1 indicates that the target gene was present as 2 copies, N = 1.5 implies presence as 3 copies, and hemizygosity is expected to yield a ratio of N = 0.5. SHOX = short stature homeobox; IL3RA = interleukin-3 receptor α; GYG2 = glycogenin 2; TM4SF2 = transmembrane 4 superfamily member 2.

Starting from the telomere, normal individuals (men and women) as well as our patient possess 2 copies of each marker up to the interleukin-3 receptor α (IL3RA) gene. From this position, our XX male SLE patient clearly showed 3 copies of each marker (including IL3RA) up to the end of the PAR1 region. Interestingly, in the public human genome sequence database, there is a gap of 100 kb (between 0.989 Mb and 1.089 Mb) just before the IL3RA gene. The translocation must have occurred in that gap. As a consequence, the SLE patient has 3 copies of each of the 12 known genes that are found between IL3RA and the end of the PAR1 region. This specific translocation is a rare event, is probably unique, and may have different functional consequences as compared with other XX males, since most of them have a balanced translocation outside PAR1, usually at PRKX/PRKY loci, causing no triplicated genes (23).

Interestingly, these 12 triplicate genes are also present as 3 copies in patients with Klinefelter's syndrome (XXY). Since it has been demonstrated that some patients with Klinefelter's syndrome have also the SLE phenotype, it can be argued that if there is a gene on the X chromosome that predisposes to the development of SLE, we may have pinpointed its location.

Based on the functions of the 12 genes, the IL3RA and CD99 genes are interesting candidates and should be considered for future genetic association studies in patients with autoimmune conditions. However, the 10 other genes or even genes not yet discovered in that region may also be implicated. The IL3RA gene encodes the α subunit of the receptor for interleukin-3, a pleiotropic growth factor affecting the proliferation and differentiation of hematopoietic stem cells. CD99 (OMIM no. *313470; also known as MIC2) encodes a 32-kd transmembrane protein expressed on all human tissues tested so far, with particularly high expression in cortical thymocytes and T cells. The CD99 protein is a cell surface molecule that is involved in adhesion processes and apoptosis of T cells (28, 29). The patient's family history of autoimmune disease suggests a genetic predisposition that may have been potentiated by the translocation. This multiple-hit hypothesis is consistent with the complex disease status of SLE.

To determine if the translocation might have functional consequences, we measured expression levels of 1 of the 12 triplicated genes (CD99) in the patient's lymphocytes and compared them with expression levels in normal subjects (male and female). Relative gene expression levels were quantified by real-time reverse transcription–PCR using the TaqMan gene expression assay (ABI database no. hs00365982_m1) and an ABI Prism 7000 Sequence Detection system (Applied Biosystems, Foster City, CA). A standard curve was prepared for CD99 and the endogenous control (ribosomal 18S and GAPDH). We found that the relative expression of the CD99 gene (GenBank accession no. NM_002414) was 3.3-fold higher in the patient as compared with 6 normal individuals (men and women) and, thus, may have consequences for T cell biology. In addition, the level of CD99 expression was also measured in peripheral white blood cells from our patient and from control subjects. In these cells, the relative expression of CD99 was 4.43-fold higher in the patient as compared with normal subjects.

The nucleotide sequence data reported herein have been deposited in the GenBank database (accession no. AY768705; homo sapiens t[X;Y] [p22.33;p11.2] translocation break point region genomic sequence).

DISCUSSION

The most important aspect of the characterization of the Xp22.33;Yp11.2 translocation in our unique XX male patient with severe SLE is in the identification of genes that are important in the predisposition to the development of SLE in the general population. Since this translocation is unique, it is unlikely that the same mechanism would be operational in a significant proportion of subjects. However, if one of the causes of SLE in this patient is the overexpression of the triplicated genes, one could hypothesize that certain nucleotide variations located in the regulatory regions of one or several of the triplicated genes could influence the level of expression of those genes and modulate susceptibility to the more common form of SLE. Sequencing and genotyping analyses of several single-nucleotide polymorphisms in a large cohort of SLE patients would be required to test this hypothesis. It would also be of great interest to study the expression of CD99 (and of the other triplicated genes) in a large cohort of lupus patients.

We can further speculate about other potential causes of SLE in this young male patient. First, the fact that the prevalence of SLE is higher in women suggests that some X chromosome genes could be implicated in the disease etiology. Consequently, a male who possesses 2 X chromosomes is genetically very similar to females and would therefore have a higher predisposition to the development of SLE as compared with normal XY men. However, the precocity and the severity of SLE in this patient set him apart from the classic postpubertal female presentation. There is also the possibility that the predisposition to SLE is related to the loss of some protective Y chromosome genes, since >90% of the patient's Y chromosome is missing. It is also possible that his XY chromosome abnormalities are irrelevant to the etiology of his SLE and are only coincidental, such that the patient is in fact affected by 2 extremely rare conditions. Estrogens or sex hormones should also be considered as a potential cause of the SLE. However, the fact that the patient developed the disease at the age of 6 years is evidence against the sex hormone hypothesis of SLE induction, although his hypogonadism may have contributed to its maintenance.

In view of the data presented herein, we conclude that the XY translocation we describe is a highly probable contributing factor to SLE in this patient with a family history of autoimmune disease. Analysis of a large cohort of male and female SLE patients will be necessary to determine if the identified candidate genes may predispose, and be relevant, to the development of SLE in the general population.

Acknowledgements

The authors are grateful to the patient for generously donating samples for study and to the patient's family members for their cooperation in this study. We also thank Dr. Anik Boudreau for helpful comments and corrections on the manuscript.