Systematic mRNA analysis for the effect of MLH1 and MSH2 missense and silent mutations on aberrant splicing

Abstract

A substantial proportion of MLH1 and MSH2 gene mutations in hereditary nonpolyposis colon cancer syndrome (HNPCC) families are characterized by nucleotide substitutions, either within the coding sequence (missense or silent mutations) or in introns. The question of whether these mutations affect the normal function of encoding mismatch DNA repair proteins and thus lead to the predisposition to cancer is determinant in genetic testing. Recent studies have suggested that some nucleotide substitutions can induce aberrant splicing by disrupting cis-transcription elements such as exonic enhancers (ESEs). ESE disruption has been proposed to be the mechanism that underlies the presumed pathological missense mutations identified in HNPCC families. To investigate the prevalence of aberrant splicing resulting from nucleotide substitutions, and its relevance to predicted ESEs, we conducted a systematic RNA screening of a series of 60 patients who carried unrelated exonic or intronic mutations in MLH1 or MSH2 genes. Aberrant splicing was found in 15 cases, five of which were associated with exonic mutations. We evaluated the link between those splicing mutations and predicted putative ESEs by using the computational tools ESEfinder and RESCUE-ESE. Our study shows that the algorithm-based ESE prediction cannot be definitely correlated to experimental observations from RNA screening. By using minigene constructs and in vitro transcription assay, we demonstrated that nucleotide substitutions are the direct cause of the splicing defect. This is the first systematic screening for the effect of missense and silent mutations on splicing in HNPCC patients. The pathogenic splicing mutations identified in this study will contribute to the assessment of "unclassified variants" in genetic counseling. Our results also suggest that one must use caution when determining the pathogenic effect of a missense or silent mutation using ESE prediction algorithms. Analysis at the RNA level is therefore necessary.