Quantitative Genexpressionsanalyse mittels Light Cycler an HNPCC-Tumoren

Background: Identification of HNPCC-patients can be difficult, mostly because pre-screening technics are based exclusively on tumor tissue. A mutation analysis of unaffected family members can consequently only be performed if a germline mutation has already been identified. This is a severe practical limitation in management of individuals at risk for HNPCC, because in about 60% of highly suspicious families a germline mutation is not found. We therefore established a real-time PCR method based on the LightCycler technology for fast and quantitative analysis of MMR-gene mRNA expression levels in formalin-fixed and paraffin embedded tumor and corresponding normal tissues. The hypothesis is, that mutation carriers have a low proteinexpression of the genproduct (haplotype- insufficiency). Our aim was to determine whether quantitative measurement of the mismatch repair proteins MSH2 and MLH1 on mRNA level by real-time-PCR allows better discrimination of subtle gene expression differences than with semiquantitative immunohistochemical staining of protein expression. Methods: In this study we used cDNA of the MMR-Genes MLH1 and MSH2 from paraffin embedded tissue. Extracted RNA was reverse transcribed to cDNA using the AMV-RT-Kit (Roche) with MLH1rev and MSH2rev as specific primers and G6PDH as the reference transcript. To analyze the expression status of the genes of interest, the relative amount of the synthesized cDNA-molecules was determined through quantitative real-time PCR (LightCycler System, Roche) using the RelQuant Software (Crossing point). Results: In three study groups, we analyzed the normal and the tumor tissue of 37 patients with colorectal carcinoma. The tumors of group1 were microsatellite stable (n = 15) and have a regular expression of the immunochemistry of the MMR-Gene group 2 (n-12) were microsatellite instable (MSI) with a loss of MLH 1 proteinexpression in immunochemistry. The third group is MSI with a loss of MSH2 proteinexpression. In group 1, non-MMR-deficient tumors MLH1 and MSH2 mRNAs were almost equally expressed as defined by MLH1 to MSH2 transcript ratio. As expected, in microsatellite instable (MSI-H) tumors deficient of MLH1 protein MLH1/MSH2 ratio was very low including one HNPCC case. In contrast, in MSH2 deficient carcinomas corresponding values were high all being HNPCC related. Most interestingly, in the corresponding normal tissues the MLH1/MSH2 transcript ratios were significantly elevated in MSH2 related HNPCC patients as compared to the ratio values of normal mucosa in non-MMR-deficient and MLH1 deficient cases. This finding indicates haplotype-insufficiency in normal colonic mucosa of HNPCC patients bearing MSH2 germline mutations. Conclusion: Real-time PCR allows quantitative expression analyses of MMR-gene mRNA in formalin fixed and paraffin-embedded tissue. Our data suggest that this approach provides a potential new tool for predictive testing of asymptomatic persons at risk for MSH2 mutation.