Are CpG sites mutation hot spots in the dystrophin gene?

Abstract

Duchenne muscular dystrophy (DMD) and the milder allelic Becker muscular dystrophy (BMD) are X-linked neuromuscular disorders resulting from mutations in the 2.4-megabase gene encoding the protein dystrophin (1,2). Two thirds of the diseasecausing mutations are large deletions which occur in the 5' and the central portions of the gene; most of them can be detected by PCR using eighteen primer pairs targetting deletion-prone regions (3—5). The remaining mutations, which are presumably due to small insertions/deletions or nucleotide substitutions, can be examined by SSCP (single strand conformational polymorphism) technique (6,7). Using SSCP we analyzed eighteen dystrophin fragments (see Legend to Fig. 1) in 27 DMD or BMD patients without detectable deletions or insertions (8). In two DMD patients we observed unique SSCP profiles for fragments spanning dystrophin exons 8 and 19 (Fig. 1). The corresponding DNA fragments were cloned and sequenced. In both cases we identified a nonsense mutation caused by a C-toT transition. The first mutation, in exon 8 at cDNA position 932 (numbering according to ref. 9), changes a CAA (GIU242) codon and has already been described in an Italian patient (10). The second, at position 2510 (exon 19), occurs within a CGA codon (Arg768) and has been already seen in two instances (11,12). Interestingly, among thirteen STOP codons reported to date in unrelated DMD/BMD patients, four were found in exon 19, two in exon 8 and two in exon 70 (11 -16 ; this study). Is the clustering of nonsense mutations in certain exons expected? To answer this question we examined the distribution of 1738 nonsense substitutions that are possible within the dystrophin coding region. Assuming that they are all equally probable, the probability of finding a STOP codon in a particular exon corresponds to the fraction of such substitutions within that exon. As shown in Fig. 2A, these probabilities are similar for most of the exons (an average of 0.013 with a standard deviation of 0.006). It may be argued that the distribution of the reported nonsense mutations is biased by an unequal sampling of exons as summarized in Fig. 2B. However, 23 exons have been examined in more than 100 patients and STOP codons were found in only three of these exons. Although some mutations might have escaped detection there is no reason to assume that this would introduce a significant distribution bias.