Methylation status of genes flanking the fragile site in males with the fragile-X syndrome: a test of the imprinting hypothesis.

Laird has suggested that the mutation responsible for the fragile X (FraX) syndrome interferes with the process of X chromosome reactivation in oocytes, thus blocking the transcription of loci at or neighboring the fragile site (Xq27.3) and producing the clinical FraX phenotype; he has also suggested that the transcriptional block might result from inappropriate DNA methylation. We have explored the latter possibility by examining the methylation status of several genes flanking the fragile site in eight FraX males from seven unrelated families. These genes (HPRT, G6PD, P3, and GdX), contain 5' clusters of CpG dinucleotides which are differentially methylated in transcriptionally active and inactive loci. Using the methylation-sensitive restriction enzyme, HpaII, we observed no differences between FraX and normal males in the methylation either of CpG islands in any of these genes or in nonclustered CpGs within the body of the HPRT gene. The same was true for the CpG cluster in intron 22 of the clotting factor VIII gene. In each gene, the island is methylated on the inactive X chromosome and not on the active one, but in no case were these islands methylated in FraX males. The four anonymous loci (DXS98, DXS304, DXS52, and DXS15) that are closely linked to the FraX locus (the closest is within 5 centimorgans) are not differentially methylated on active and inactive X, nor are they unusually methylated in FraX males. Therefore, our observations provide no evidence that DNA methylation in the vicinity of the FraX locus has a role in producing the clinical phenotype of the FraX syndrome.