Genomic context of paralogous recombination hotspots mediating recurrent NF1 region microdeletion.

Recombination between paralogs that flank the NF1 gene at 17q11.2 typically results in a 1.5-Mb microdeletion that includes NF1 and at least 13 other genes. We show that the principal sequences responsible are two 51-kb blocks with 97.5% sequence identity (NF1REP-P1-51 and NF1REP-M-51). These blocks belong to a complex group of paralogs with three components on 17q11.2 and another on 19p13.13. Breakpoint sequencing of deleted chromosomes from multiple patients revealed two paralogous recombination hot spots within the 51-kb blocks. Lack of sequence similarity between these sites failed to suggest or corroborate any putative cis-acting recombinogenic motifs. However, the NF1 REPs showed relatively high alignment mismatch between recombining paralogs, and we note that the NF1REP hot spots were regions of good alignment bordered by relatively large alignment gaps. Statistical tests for gene conversion detected a single significant tract of perfect match between the NF1REPs that was 700 bp long and coincided with PRS2, the predominant recombination hot spot. Tracts of perfect match occurring by chance may contribute to breakpoint localization, but our result suggests that perfect tracts at recombination hot spots may be a result of gene conversion at sites at which preferential pairing occurs for other, as-yet-unknown reasons.

The ichq mutant mouse, a model for the human skin disorder harlequin ichthyosis: mapping, keratinocyte culture, and consideration of candidate genes involved in epidermal growth regulation.

Harlequin ichthyosis (HI) is a rare and usually fatal scaling skin disorder. The HI mutant mouse (ichq/ichq) has many similarities to the human disorder and provides an important model to identify candidate genes. In this study, we report refined mapping of the mouse ichq locus and consideration of the candidate genes: calpain 1 (Capn1), phospholipase C beta 3 (Plcb3), and Rela and Ikka/Chuk that encode components of the nuclear factor-kappa B (NF-kappaB) pathway. Each are strong candidates because of epidermal expression and/or changes in expression in human HI. All candidates are linked to the ichq locus on mouse Chromosome 19, although Ikka is located more distally. Genetic mapping in mouse has narrowed the ichq critical region to 4 cM. Keratinocytes from skin of +/+, +/ichq and ichq/ichq mice were cultured; all genotypes had similar expression of epidermal differentiation markers. RT-PCR amplification and sequence analysis of each candidate gene did not reveal any mutations in the ichq mouse. Mutational screening of CAPN1 cDNA from different human HI cases revealed a R433P change, but analysis of 50 normal samples demonstrated that this was an apparent polymorphism. Sequence of RELA in five unrelated human HI cases was normal. The results provide compelling evidence that none of these genes are the primary defect in the ichq mouse and that CAPN1 and RELA are not mutated in the human disorder.