To the editor: We thank Dr Berg and Dr Steigen for their letter in response to our article on DNA methylation in breast and colorectal cancer.1 We acknowledge the errors that were not caught during the proofreading of the paper. We agree that the term epigenetic refers to a heritable change in gene expression that is mediated by mechanisms other than alterations in the primary nucleotide sequence. It is also correct that the frequency of methylation of the BRCA1 gene is indeed 9.1% and we apologize for our mistake.2 However, it does not change the conclusion regarding the functional importance of the methylation of BRCA1, the well-known breast cancer-related gene.1 Indeed, it is the gene that is methylated and not the gene product.

In regard to the HNPCC, we referred to the findings of Gazzoli et al3 who found a direct association between the methylation of the region of MLH1 and the silencing of the gene in HNPCC. In a subset of sporadic colorectal cancers, increased microsatellite instability is caused by the inactivation of the mismatch repair gene MLH1 due to promoter methylation. We agree that in HNPCC, which also shows increased microsatellite instability, mismatch repair inactivation results primarily from germline mutations. However, MLH1 promoter methylation has also been found in a subset of HNPCC, and this is inversely associated with loss of heterozygosity.4 Chan et al5 reported a family with inheritance, in three successive generations, of germline allele-specific and mosaic hypermethylation of the MSH2 gene, without evidence of DNA mismatch repair gene mutation. In this family, three siblings carrying the germline methylation developed early-onset colorectal or endometrial cancers, all with microsatellite instability and MSH2 protein loss.5 This suggests that promoter hypermethylation or point mutation could be responsible, at least in some cases, for the somatic loss of the mismatch repair genes, MLH1 or MLH2 in HNPCC.

In response to the request for an explanation of the X-chromosome inactivation and DNA hypomethylation, we refer to the published articles by Paning and Jaenisch6 and by Beard et al7 on the hypomethylation and silencing of X-linked genes.