1. ¹Institute of Molecular Pathology, University of Heidelberg, Germany
2. ²Department of General Surgery, University of Göttingen, Germany
3. ³Department of Pathology, Klinikum Kassel, Germany
4. ⁴Institute of Pathology, University Hospital Bonn, Germany
5. ⁵Institute of Pathology, University of Heidelberg, Germany
6. ⁶Department of Surgery, University of Heidelberg, Germany
7. ⁷Department of Internal Medicine, Knappschaftskrankenhaus, University of Bochum, Germany
8. ⁸Human Genetics, University of Bochum, Germany
9. ⁹Institute of Human Genetics, University Hospital Bonn, Germany
10. ¹⁰Department of Surgical Research, Dresden University of Technology, Germany
11. ¹¹Department of Surgery, University of Düsseldorf, Germany
12. ¹²Department of Surgery, University of Technology, Munich, Germany

Correspondence: JF Gebert, Institute of Molecular Pathology, University of Heidelberg, Im Neuenheimer Feld 220/221, D-69120 Heidelberg, Germany. E-mail: johannes.gebert@med.uni-heidelberg.de

¹³These two authors contributed equally to this work

Received 13 September 2004; Revised 13 December 2004; Accepted 15 December 2004; Published online 7 February 2005.

Abstract

Microsatellite instability (MSI) occurs in most hereditary nonpolyposis colorectal cancers (HNPCC) and less frequently in sporadic tumors as the result of DNA mismatch repair (MMR) deficiency. Instability at coding microsatellites (cMS) in specific target genes causes frameshift mutations and functional inactivation of affected proteins, thereby providing a selective growth advantage to MMR deficient cells. At present, little is known about Selective Target Gene frameshift mutations in preneoplastic lesions. In this study, we examined 30 HNPCC-associated MSI-H colorectal adenomas of different grades of dysplasia for frameshift mutations in 26 cMS-bearing genes, which, according to our previous model, represent Selective Target genes of MSI. About 30% (8/26) of these genes showed a high mutation frequency (50%) in colorectal adenomas, similar to the frequencies reported for colorectal carcinomas. Mutations in one gene (PTHL3) occurred significantly less frequently in MSI adenomas compared to published mutation rates in MSI carcinomas (36.0 vs 85.7%, P=0.023). Biallelic inactivation was observed in nine genes, thus emphasizing the functional impact of cMS instability on MSI tumorigenesis. Some genes showed a high frequency of frameshift mutations already at early stages of MSI colorectal tumorigenesis that increased with grade of dysplasia and transition to carcinoma. These include known Target Genes like BAX and TGFBR2, as well as three novel candidates, MACS, NDUFC2, and TAF1B. Overall, we have identified genes of potential relevance for the initiation and progression of MSI tumorigenesis, thus representing promising candidates for novel diagnostic and therapeutic approaches directed towards MMR-deficient tumors.