We report the parental origin, and where possible the chromosomal origin of 115 de novo unbalanced structural chromosome abnormalities detectable by light microscopy. These consisted of 39 terminal deletions, 35 interstitial deletions, 8 rings, 12 duplications and 21 unbalanced translocations. In all categories the majority of abnormalities were of paternal origin, although the proportions varied from a high of 84% in the interstitial deletions and rings to a low of 58% in the duplications. Among the interstitial deletions and duplications, there were approximately equal numbers of intra- and interchromosomal abnormalities, while the majority of unbalanced translocations were isodisomic for the duplicated chromosome. The examination of the parental ages in the four main classes of abnormality showed terminal deletions of maternal origin to be associated with a significantly reduced maternal age. Thus, there is a clear propensity for structural chromosome abnormalities to occur in male germ cells, although the chromosomal origin seems similar irrespective of the parental origin.

Non-allelic homologous recombination between chromosome-specific LCRs is the most common mechanism leading to recurrent microdeletions and duplications. To look for locus-specific differences, we have used microsatellites to determine the parental and chromosomal origins of a large series of patients with de novo deletions of chromosome 7q11.23 (Williams syndrome), 15q11-q13 (Angelman syndrome, Prader -Willi syndrome) and 22q11 (Di George syndrome) and duplications of 15q11-q13. Overall the majority of rearrangements were interchromosomal, so arising from unequal meiotic exchange, and there were approximately equal numbers of maternal and paternal deletions. Duplications and deletions of 15q11-q13 appear to be reciprocal products that arise by the same mechanisms. The proportion arising from interchromosomal exchanges varied among deletions with 22q11 the highest and 15q11 -q13 the lowest. However, parental and chromosomal origins were not always independent. For 15q11 -q13, maternal deletions tended to be interchromosomal while paternal deletions tended to be intrachromosomal; for 22q11 there was a possible excess of maternal cases among intrachromosomal deletions. Several factors are likely to be involved in the formation of recurrent rearrangements and the relative importance of these appear to be locus-specific.

otide repeat (FRAXA) located in the 5Ј untranslated region of the FMR1 gene gives rise to the fragile X syndrome, an inherited disorder associated with mental retardation (Fu et al. 1991;Verkerk et al. 1991). Distal to FMR1 (approximately 0.77 Mb) is FMR2, which contains a GCC triplet repeat (FRAXE) in the 5Ј untranslated region. Full mutations (over 200 repeats) in FRAXE are rare (Sutherland and Baker 1992) but appear to be associated with nonsyndromic mental impairment (Knight et al. 1994).Instability at FRAXA is thought to be due at least in part to loss of an AGG interspersion from the CGG repeat, resulting in slippage during replication (Eichler et al. 1994). However, haplotype studies using microsatellites suggest that other factors may also be important determinants of risk for FRAXA expansion (Murray et al. 1997). Eichler et al. (1996) proposed three independent mechanisms for repeat expansion: (1) loss of 3Ј stabilizing AGG interspersion followed by rapid expansion to a pre-or full mutation, (2) slow expansion from common to pre-and full mutation without loss of an AGG interspersion, and (3) generalized instability that affects other polymorphic microsatellite markers. Instability at FRAXE usually appears to increase with increasing repeat number. In contrast to FRAXA, the repeat does not contain any interspersions, suggesting a simple relationship between size of repeat and risk of expansion. However, work by Ennis et al. (2001) suggested that, although triplet repeat size was a significant predisposing factor for FRAXE expansion, there were other genetic determinants involved. For example, they described a significant association between unusual FRAXA and unusual FRAXE alleles. FRAXA repeat sizes of more than 50 were positively associated with FRAXE repeats of less than 11. Although these associations occurred on similar haplotype backgrounds, founder effects could not be confirmed without looking at comparative data from different ethnic backgrounds.With the near completion of the human genome sequencing project, there has been much interest in the use of single-nucleotide polymorphisms (SNPs) for studying genetic factors associated with complex disease traits. SNPs G. Brightwell (*) · R. Wycherley · G. Potts · A. Waghorn Wessex Regional Genetics Laboratory, Salisbury District Hospital, Salisbury, Wiltshire SP2 8BJ, UK Tel. ϩ44-1722-425047; Fax ϩ44-1722-338095 e-mail: galebrightwell@hotmail.com Abstract Single-nucleotide polymorphisms (SNPs) are the most common type of genetic variation within the human genome, occurring approximately once every kilobase. However, for association studies, SNPs are not as informative as microsatellite markers and a large number of SNPs and substantial population sizes are required for linkage and mapping studies. A SNP map was generated for the FRAX region of the X chromosome, approximately 0.8 Mb proximal and 1.8 Mb distal to the FRAXA repeat, at a density of at least 1 SNP every 100 kb. SNPs were identified in a population of 28 women with a FRAXA expansion (includi...