Molecular genetic studies on genes involved in hereditary nonpolyposis colorectal cancer (HNPCC)

Abstract

Colorectal cancer (CRC) is a major public health problem in Western countries, and it is the third leading cause of death from cancer in both males and females. Inherited predisposition can be divided into two subgroups based on the absence or the presence of multiple benign colorectal polyps. Hereditary nonpolyposis colorectal cancer (HNPCC) is one of the most common inherited cancer predisposing syndrome. It has an autosomal dominant transmission pattern, and it approximately accounts for up to 5% all CRC cases. Individuals with HNPCC are also at an increased risk of developing cancers of the small bowel, biliary tract, uroepithelium, kidney, and central nervous system (CNS). So far, several mismatch repair (MMR) genes are known to be involved in HNPCC, and germline mutations have been identified in five of these genes, hMSH2, hMLH1, hPMS1, hPMS2, and hMSH6. To determine and outline a mutation spectrum of MMR susceptibility genes as well as genes involved in CRC tumorigenesis pathway in the Swedish population, we have conducted a serial mutation analysis in big cohort of HNPCC or HNPCC-like families employing a number of molecular biological techniques. We have shown that 53% of the families that fulfilled stringent Amsterdam criteria harbor germline mutations in either hMLH1 or hMSH2, while a lower mutation detection frequency (18%) was found in non-Amsterdam families (Paper 1, IV).

Based on our studies of specific tumor phenotype and genotype correlation in a big series of samples, as well as several clinical parameters, we have defined sensitive and specific criteria to be used in preselection for mutation screening (Paper IV). Various mutation detection techniques, both RNA- and DNA-based, were used (Paper IV, V). Our studies demonstrated the importance of confirming RNA variants in genomic DNA (Paper V). Two MutL homologos of hPMS1 and hPMS2 have been suggested to explain a few HNPCC patients. However, we did not identify any germline mutations in 84 HNPCC or HNPCC-like families in which the presence of other known MMR genes had already been excluded (Paper VI).

Since no mutation in either of these two genes has been found to segregate with disease in 199 families studied so far, we conclude that it is possible that a mutation in hPMS1 and hPMS2 alone does not confer an increased risk for CRC. In a study in kindreds segregating endometrial and colorectal or endometrial cancer alone, a number of genes including all known MMR genes, PTEN, as well as beta-catenin gene were investigated. Our finding of MSI negative tumors and lack of clear-cut germline mutation in any of the genes suggest no significant role for these genes, and imply that other genes are more likely to predispose to both endometrial and colorectal cancer in these families (Paper VII).

Single nucleotide substitutional mutations were commonly seen in MMR genes, especially in hMLH1 We have identified two such missense mutations in hMLH1 and one in hMSH2. To determine whether these variants contribute pathogenic effects to CRC, we have investigated genotype-phenotype correlation in the families, studied the frequency of variants in general population, specific tumor phenotype as well as loss of heterozygosity (LOH).

Our results indicated that the variant in hMSH2 was considered a common polymorphism. In contrast, two hMLH1 mutations were associated with an increased risk of CRC. Lack of microsatellite instability (MSI)in these tumors suggested that they might be involved in an unknown tumorigenesis pathway, other than those generally hypothesized (Papers II, III).