A balanced complex chromosomal rearrangement (BCCR) in a family with reproductive failure

Balanced complex chromosomal rearrangements are very rare events in the human population. Translocations involving three or more chromosomes frequently lead to a severe reproductive impairment secondary to meiotic disturbance in males and to chromosomal imbalance in gametes of females. We report a new familial case of complex chromosome anomaly involving chromosomes 13, 14 and 22. Cytogenetic investigations showed a complex chromosomal chromosome rearrangement involving: (i) a Robertsonian translocation between chromosomes 13 and 14; and (ii) a reciprocal translocation between the long arms of chromosome 14 and the long arm of chromosome 22. The aetiology of the translocation was characterized by conventional ﬂuorescence in-situ hybridization (FISH) studies and routine R- and G-banding (RTBG and GBTG) combined with a and b satellite centromeric FISH probes. Predicted conﬁguration of the hexavalent at pachytene stage of meiosis was used to consider the modes of segregation; only two conﬁgurations resulted in a normal or balanced gamete karyotype. Reproductive management and genetic counselling are discussed.


Introduction
Complex chromosome rearrangements (CCR) are structural abnormalities involving at least three chromosomes with three or more breakpoints. Balanced CCR (BCCR) are infrequent and usually occur de novo (Batavian and Eswara, 1998). Rarely, BCCR are of familial origin and, in these cases, transmission is more frequently from the mother (Farrell et al., 1994). In males, BCCR are thought to lead to severe reproductive impairment through meiotic disturbance or chromosomal imbalance in gametes. We report here a new case of complex chromosomal anomaly involving chromosomes 13, 14 and 22 with familial transmission.

Family report
The proband (II 2 -D) is a 40 year old woman who has had nine ®rst-trimester miscarriages (Figure 1). She has a 34 year old sister (II 4 -A) who has had three miscarriages and a 36 year old sister (II 3 -B) who has had two miscarriages and two healthy children. A 44 year old brother (II 1 -C) was infertile with oligoasthenozoospermia. Two other brothers (II 5 and II 6 ) of II 2 -D could not be studied. The proband's mother (I 1 -M) had three miscarriages. She underwent menopause at age 55 years. The proband's father (I 2 -A) died at 37 years of age.

Cytogenetic analysis
Cytogenetic investigations were carried out using standard methods on lymphocytes from phytohaemagglutinin (PHA)stimulated peripheral blood cultures and from immortalized Blymphocytes. Chromosome spreads were processed for RHG, QFQ, GTG, CBG bands and Nuclear Organizer Region (NOR) staining. High resolution banding (RTBG, GBTG) was obtained according to the standard technique (Dutrillaux and Vigas-Pequignot, 1981). RHG, GTG, RTBG and GBTG banded metaphases from lymphocytes were interpreted at Figure 1. Pedigree of proband's family. Human Reproduction Vol.18, No.10 pp. 2058±2066, 2003DOI: 10.1093 resolution levels of 450 and 650 bands. Conventional¯uorescence in-situ hybridization (FISH) was carried out using human probes on metaphases of the transformed lymphoblast cell line according to standard protocols and to the manufacturer's manuals (Table I).

Genotype analysis
Genomic DNA was puri®ed from peripheral blood lymphocytes according to standard sodium dodecyl sulphate±proteinase-K and phenol/chloroform extraction methods. DNA polymorphisms in the mother and the four children were analysed by PCR ampli®cation of tandem short sequence repeats. The selected markers on chromosomes 13, 14 and 22 were chosen from the Ge Âne Âthon and CHLC collections included in the screening set, version 6.0, distributed by Research Genetics Murray et al., 1994;Dib et al., 1996). One oligonucleotide primer for each marker was labelled with 5 mCi of [g 32 P]ATP using T4 polynucleotide kinase. PCR ampli®cations were performed using 60±90 ng of genomic DNA in a total volume of 15 ml mixture per reaction containing 0.4 pmol/l of labelled forward primer, 2.6 pmol/l of unlabelled reverse primer, 1.3 mmol/l of each dNTP, and 0.25 IU Taq polymerase. Radioactive PCR products were separated by electrophoresis on a 6% denaturing polyacrylamide/50 percentage urea gel (Blouin et al., 1995). Two different investigators independently determined genotypes after autoradiography.

DNA genotypes
In order to assess the parent-of-origin of the rearranged chromosome segments and to detect infra-microscopic duplication or deletion that may have arisen from these chromosome rearrangements, we genotyped individual I 1 -M and her children (II 2 -D, II 3 -B, II 4 -A, II 1 -C). PCR ampli®cations were performed for a series of microsatellites markers that map in the regions where the translocation breakpoints occur (Table II). Heterozygous bi-allelic genotypes were obtained for most markers in the four children. A single allele genotype was obtained for only four markers, three of them mapping to chromosome 22. Interestingly only the most centromeric marker D22S420 shows a single allelic pattern in three out of the four children (II 2 -D, II 4 -A, II 1 -C). However, since the father's DNA (I 2 -A) was not available, it is not possible to discern between a homozygous disomic genotype and a monosomic status for marker D22S420 that could have arisen  from a partial deletion of this region of chromosome 22. Other markers also showed single allelic patterns for D22S689 and D22S683 in individual II 4 -A, D22S445 in individual II 3 -B, and D14S749 in individual II 2 -D. Similarly, a partial duplication of this region of chromosome 22 cannot be excluded in individual II 2 -D due to lack of information in the case of single allelic pattern. The analysis of these 12 markers did not show any evidence of DNA duplication or deletion in the four children analysed. In addition, all informative markers in the three chromosomes do not provide any evidence of uniparental disomy.
Further analysis of genotypes suggests that the children displaying the same rearrangements preferentially share spe-ci®c haplotypes. For example, the three children having a translocation between chromosomes 13 and 14 show the same alleles on proximal markers of chromosome 13. Moreover, genotypes of chromosome 14 markers are compatible with a common paternal haplotype but not with the maternal haplotype. A compatibility with a common paternal haplotype is also seen with the three most proximal markers of chromosome 22 in the sibs sharing the same chromosomal pattern, whereas sib II 4 -A does not share.

Familial BCCR and segregation
In this family, we propose that the proband's father (I 2 -A) was carrying the BCCR, in either all cells or in mosaic, since the complete BCCR is present in at least two of his offspring: II 2 -D and II 4 -A ( Figure 5). However, this hypothesis could not be con®rmed. The normal blood mother's karyotype cannot exclude a maternal germ cell mosaiscism. By using FISH with TTAGGG repeats as the probe, we were able to exclude the possibility of a jumping translocation (JT) in the mother. JT is a rare chromosomal abnormality in which a speci®c chromosomal segment translocates onto the ends of various  chromosomes and may predispose to chromosomal imbalance via non-disjunction. JT mostly involve the acrocentric chromosomes in the Robertsonian translocations.
Different classi®cations of CCR have been proposed (Kausch et al., 1988;Lurie et al., 1994). CCR can be divided into two types: a three-way translocation, and a rearrangement with more than one breakpoint per chromosome. This BCCR results from a veritable three-way exchange in which three segments, generated by three chromosome breaks, translocate and unite. Three-way BCCR are rarely transmitted by the father (Farrell et al., 1994). At meiosis, the chromosomes involved in the rearrangement form a hexavalent. This con®guration would allow full synapsis of six homologous segments except for the proximal segments adjacent to the breakpoints. To our knowledge, formal analysis of the breakpoints of cases described in the literature has not been reported. Two gametes arising from alternate segregation would be balanced. The remaining gametes would be unbalanced to a greater or lesser degree (Table III). A tendency to favour symmetric alternate segregation in the ®rst generation, probably during male gametogenesis, appears to characterize this family.
We cannot use the`adjacent or alternates' models generally used for translocations between two chromosomes to analyse BCCR. For II 3 -B and II 1 -C, there was recombination at the time of parental meiosis between the chromosome 14 involved in the translocation and the normal chromosome 14. II 3 -B received part of the (13;14) with the translocation± recombination±telomeric extremity of normal chromosome 14 and the normal chromosome 22 of her father. II 1 -C received the centromeric extremity of chromosome 14 normal± recombination±translocation (14;22) and the 13 normal chromosome of his father.
The result of the microsatellite analysis is not easily interpretable. The molecular analysis of the genotypes shows that the proximal part of chromosome 22 is non-informative. We might surmise the existence of a common haplotype in subjects with a translocation (14;22). However, con®rmation of this is impossible, the mother (I 1 -M) being homozygote for marker D22S446, the father's (I 2 -A) deduced haplotype being non-informative for D22S420. The data for the three sisters (II 2 -D, II 3 -B, II 4 -A) with the Robertsonian translocation t(13;14) is less ambiguous.
Another hypothesis is possible. I 2 -A could be a carrier of two translocations: a Robertsonian translocation der(13;14) (p11;p11) and a reciprocal translocation t(14;22) (q32.3;q11.2). During meiosis, a crossing-over would have occurred between the two chromosomes 14 giving rise to a derivative der(13;14)t(14;22) and another der(22)t(14;22); II 3 -B would have received only the translocation t(13q;14q) whereas II 1 -C would have received the derivative 14 of translocation t(14;22) without the derivative 22. The other question is the probability of a second crossing-over identical to the index case (II 2 -D). However, such a relatively simple mechanism is insuf®cient to explain the identical rearrangement in the sib of the proband, which would seem to suggest that the same crossing-over occurred twice in succession.
Review of the literature suggests that severely unbalanced con®gurations often occur in female gametogenesis. Paternal origin was, however, very frequently shown in de-novo CCR informative reports. The probability of such uniparental origin occurring by chance alone is 1/256 (Batista et al., 1993). This low probability leads us to speculate that mechanisms resulting in BCCR occur preferentially during spermatogenesis.

BCCR and fertility
BCCR rarely occur in phenotypically normal persons (Fukushima et al., 1986). The impact of CCR on fertility is important. Anomalies of the acrocentric chromosomes increase the risk of sterility (Gabriel-Rodez et al., 1986). The fact that individual I 1 -M had six children is surprising. The possibility of a germinal parental mosaicism should be considered. In the female, gametogenesis can accommodate the complexity of CCR. The female may be fertile and have pregnancies that produce phenotypically normal children. In the father's case (I 2 -A), we suspect that spermatogenesis produces phenotypically normal children. In contrast, in the literature, male carriers are often subfertile (Johannisson et al., 1985;Saadallah and Hulten, 1985) or sterile due to spermatogenic arrest (Rodriguez et al., 1985).
Studies of the pachytene stage of meiosis have provided clues to the underlying mechanisms responsible for male sterility associated with some autosomal translocations. Three features are regularly observed in such male-sterilizing rearrangements: (i) synaptic failure around breakpoints, (ii) association of the translocation ®gure with the sex chromosomes, (iii) frequent occurrence of an acrocentric chromosome in the translocation.
Two main models have been proposed to explain gametogenic failure in the male. Burgoyne and Baker (1994) have argued that impairment of spermatogenesis might be attributed to generalized pairing disruption along the genome, an extension of the earlier hypothesis of Miklos (1974) in which XY-pairing failure was suggested as a primary cause of germ cell failure. Alternatively, the defect could result from XY± multivalent interaction, as originally proposed for the mouse by Forejt (1974) and later suggested by Chandey (1979) to explain human spermatogenic failure. Each mechanism in itself may be suf®cient to cause spermatogenic failure, but the two could interact, where partial asynapsis between normal and translocated chromosomes would favour attraction between the translocation ®gure and the differential segment of the Xautosome (Rosenmann et al., 1985).
Studies of three-way translocation (Johannisson et al., 1985;Saadallah and Hulten, 1985) gave few indications of XY association, all arms of the hexavalents being fully paired during the pachytene stage. Extensive asynapsis around the breakpoints was a feature, but there was very little evidence of spermatogenic depression or arrest, with the sperm count being within normal limits. Our case presents a hexavalent formed by three acrocentric chromosomes (one Robertsonian translocation and one reciprocal translocation). Meiotic studies on human infertile male carriers of Robertsonian translocation have shown that X-autosome association was attained by the central asynapsis and/or by the terminal chromomere of the acrocentric chromosome involved in the translocation. It was proposed that the acrocentric chromosome favours the contact between the quadrivalent and the sex vesicle, and increases the risk of sterility in male carriers of Robertsonian translocations and of reciprocal translocation involving almost one acrocentric chromosome.
In women, without sex vesicle, an involved cause of infertility does not exist and by itself could explain the different effect on fertility between male and female. Moreover, all the studies on infertile males with a balanced Robertsonian translocation show a slightly reduced number of chiasma. Variations in pattern of maternal recombination have been identi®ed as a risk factor for meiotic chromosome nondisjunction. Recent studies have con®rmed the large difference in recombination frequency between human oocytes and spermatocytes and demonstrate a clear between-sex variation in distribution of crossing-over (Tease et al., 2002). They observed an abnormal pattern of meiotic recombination in abnormal oocytes that showed chromosome-pairing errors. These facts could explain the high rate of conceptuses with presumed severely unbalanced karyotypes (spontaneous miscarriages) present in women of this family.

BCCR and genetic counselling
The nature of CCR implies that different unbalanced combinations might be expected to be viable. By attachment to centromeres, the meiotic spindle ensures attachment at the two poles and thus successful segregation of homologous chromosomes to opposite poles (Kallio et al., 1998). Therefore, the complex meiotic con®guration disturbs the chromosome orientation and causes abnormal spindle attachment leading to chromosome malsegregation. Moreover, normal meiosis requires crossing-over during homologous chromosome pairing at the pachytene stage: these chromatid exchanges, in the case of complex meiotic con®gurations, increase the risk of chromosome rearrangement, as for patient II 4 -A, and of segmental aneuploidy, as for III 9 -SA. A theoretical prediction of chromosomal segregation in gametes is possible, giving 30 different karyotypes. The empirical estimated risk for spontaneous abortion is 75±100% for some BCCR (Creasy, 1989). The chance of carrying a pregnancy to term with an abnormal and developmentally delayed child is possible and has been estimated at 50% (Wang et al., 1993). We think that this risk can be higher depending of the type of BCCR (Ruiz et al., 1996). Viability thresholds for chromosomal imbalances have been estimated at 5% of haploid autosomal length for pure trisomies and 3% for pure monosomies. In a monosomy± trisomy combination, the haploid autosomal length represented by the trisomy should not be >3.6% and should not be >0.6% for the monosomy . The resulting viability area has a step shape out of which every chromosomal imbalance is considered as lethal. The risk of serious congenital malformation with de-novo balanced reciprocal translocation between two chromosomes was estimated at~7% (3.5% per each break) on the basis of published data (Warburton, 1991). For apparently balanced CCR arising de novo, an empirical risk of up to 90% has been proposed for phenotypic abnormality and mental retardation (Gardner and Sutherland, 1989), although the exact prevalence is impossible to establish. The risk is undoubtedly much smaller. We can speculate on 3.5% per break whatever the number of breakpoints. These values vary slightly with the segregation mode, the sex of the carrier parent and the genomic content of unbalanced chromosomal segments. An international registry of minimal chromosomal imbalances should be considered in order to assist in the counselling of these patients. Preimplantation genetic diagnosis (PGD) has been used for couples with normal fertility but at high risk of having a child with chromosomal abnormalities. PGD increases the implantation rate in human IVF by avoiding the transfer of chromosomally abnormal embryos. Here, the complexity of these BCCR makes the preimplantation diagnosis impossible.

Conclusion
We report here a familial case of CCR possibly inherited from the father. In this family, CCR resulted in fertility. Thus, the risk for miscarriages appears to be higher than that of a simple balanced reciprocal translocation carrier. The risk for a liveborn child with an unbalanced rearrangement does not appear to differ signi®cantly. Our data con®rm that it is impossible to predict the risk of unbalanced progeny to carriers of BCCR. In conclusion, cytogenetic analysis is a useful tool to investigate miscarriages, to give adequate genetic counselling and to discuss the choice of an appropriate assisted reproduction technique.