Search

Mutations in the KRAS gene is a key event in the carcinogenesis of many human cancers and may serve as a diagnostic marker and a target for therapeutic intervention. In this study we have applied three different techniques for mutation detection of KRAS exon 1 mutations: Allele specific polymerase chain reaction (AS-PCR), temporal temperature gradient electrophoresis (TTGE) and constant denaturant capillary electrophoresis (CDCE). Samples from 191 sporadic colon carcinomas were analyzed. AS-PCR were performed with oligonucleotides specific for know mutations in codon 12 and 13 of the KRAS gene. In TTGE analyses, linear ramping of the temperature were performed during electrophoresis in a constant denaturant gel. CDCE analyses were performed using fluorescin labeled PCR-products. Separation was achieved under constant denaturing conditions using high temperature in a gel-filled capillary followed by laser detection. A mutated KRAS gene was found in 42/191 (22.0%) of the samples using AS-PCR, in 62/191 (32.5%) using TTGE and in 66/191 (34.6%) of the samples using CDCE. In the TTGE and CDCE analyses the sequence of the mutant were determined by comparing the electrophoretic pattern to that of known mutations or by mixing the sample with known mutations prior to reanalysis. In a titration experiment mixing mutant and wild-type alleles prior to PCR, the sensitivity for mutation detection was shown to be 10(-2) for TTGE and under optimized conditions 10(-3) for CDCE.

Germline alterations in human DNA mismatch repair genes are associated with hereditary nonpolyposis colorectal cancer (HNPCC). Mutation analysis of the genes reveals carriers with a high risk of colorectal cancer, who will benefit from surveillance. We wanted to find the best predictive parameter of a germline mutation in those genes among patients with familial colorectal cancer. Affected members from a total of 83 unrelated colorectal cancer families previously analyzed for mutations in MSH2 and MLH1 were used to evaluate different parameters' ability to predict a germline mutation. We studied various clinical criteria such as family structure, age of onset, and prevalence of endometrial cancer, as well as microsatellite instability in the tumors from the families. In total, 124 tumors from 59 of the families were tested for microsatellite instability (MSI) using PCR-based mono- and dinucleotide markers to establish whether the families could be scored as MSI-positive or -negative. The finding of MSI-positive tumors in a family was the best predictor of a germline mutation, and was found in 73% of the MSI-positive, but in less than 3% of the MSI-negative families (P < 0.0001). In contrast, MSI in unselected colorectal cancer is not as useful, since most of these MSI-positive tumors are sporadic. The finding of microsatellite instability in colorectal tumors seems efficient enough even to select those with germline mutations among families fulfilling HNPCC Amsterdam criteria, once used in identification of the DNA mismatch repair genes. Genes Chromosomes Cancer 27:17-25, 2000.