Dyschondrosteosis (DCS) has been recently ascribed to mutations of the SHOX gene on the pseudoautosomal region of the X and Y chromosomes.1 2 Most cases are accounted for by large scale deletions3-7 and only two point mutations have been hitherto identified in exon 4 (R195 X and Y199X1 2). Here, we show that point mutations in various regions of the SHOX gene also play an important role in the pathogenesis of the disease.

A total of 22 affected subjects belonging to eight families were included in the study. Inclusion criteria for affected status were short stature (2 SD below normal) with short forelimbs and distal radioulnar deformity on forearm x rays.

The 22 patients and their relatives were genotyped using microsatellite DNA markers of the pseudoautosomal region (CA-SHOX, DXYS233, DXYS234, DXYS228). Linkage studies supported the mapping of the disease gene to Xp22.3 in all families and hemizygosity at the CA-SHOXlocus was observed in three families (families 1-3, data not shown).

In families 4-8, PCR amplification and sequence analysis of the five translated exons of the SHOXgene (table 1) led to the detection of five different mutations which consistently cosegregated with the disease (table 2). In family 4, a two base pair deletion in exon 2 led to a frameshift and a translation termination at codon 77. In families 5 and 6, base changes at nucleotides 334 and 445 respectively created a stop codon in exon 3 (Q112X and E149X). In family 7, a GT insertion at codon 161 led to a premature stop codon. All four mutations resulted in a premature translation termination. Finally, in family 8, a C→T transition in exon 4 changed an arginine into a cysteine in the protein (R173C). This mutation was found in the affected mother and her three affected children but neither in the healthy father nor in 90 control chromosomes.

We have reported on SHOXdeletions and a nonsense mutation in DCS (seven deletions, one nonsense mutation1). Studying eight additional DCS families, we describe here three deletions and five point mutations ofSHOX. Taken together, these data suggest that SHOX deletion is the most frequent common disease causing mechanism in DCS (10/16, 62.5%) and that point mutations also account for a significant fraction of our patients (6/16, 37.5%).

This study also supports the view that haploinsufficiency is the most frequent mechanism in DCS (10 deletions and five nonsense and frameshift mutations in 15/16 families). Among the five novel mutations reported here, only one was a missense mutation resulting in an amino acid change in the homeodomain ofSHOX. Interestingly, this family was not clinically different from the others. Indeed, all patients had short stature and x ray deformity of the forearms although intrafamilial variability was consistently observed with males being usually less severely affected.

In conclusion, this study shows that point mutations of theSHOX gene account for a significant number of DCS patients and shows that haploinsufficiency is the most frequent disease causing mechanism in DCS.