HPRT mutational spectra and microsatellite DNA instability in HNPCC and lung cancer patients

Abstract

The, general aims of this work have been to explore the use of microsatellite DNA length variation and mutational spectra of the hypoxanthine guanine phosphoribosyltransferase (HPRT) locus in T-cells as tools for a better understanding of human somatic mutagenesis in vivo. In particular (a) how inherited mismatch repair (MMR) deficiency may affect the stability of microsatellite DNA and the mutational spectrum at the HPRT locus, and (b) how the mutational spectrum at the HPRT locus is influenced by smoking, and by predisposition to lung cancer among never smokers. Microsatellite DNA length variation (MSDLV) was studied in DNA from T-cell clones and peripheral blood lymphocytes, using suitable markers for PCR analysis and polyacrylamide gel electrophoresis. T-cell cloning in medium containing 6-thioguanine was used to select for HPRT mutant clones, and the mutations were further classified and characterised by PCR-based methods and DNA sequencing.

The background frequency of MSDLV in peripheral lymphocytes was determined using three microsatellite markers (D2S123, D9S180, D10S197). 3 out of 1028 alleles studied in T-cell clones of normal healthy subjects, showed altered microsatellite size compared to other clones from the same individual. Thus the background MSDLV was estimated to 2.9 x 10-3. We then analysed the MSDLV and HPRT mutant frequency (MF) in a breast cancer patient belonging to a hereditary nonpolyposis colorectal carcinoma (HNPCC) family, with two different mutations in her hMLH1 genes. This compound heterozygote showed a frequency of microsatellite length variation of 18.9 % per allele, which was 50 times higher than the background frequency. The HPRT MF of 34.5 x 10-6 was elevated 2-3 times compared to controls. The HPRT mutational spectrum of this patient was significantly different from normal, with a shift from base pair substitutions towards frameshifts, especially + 1bp insertions, and deletions. Also two new basepair mutations not reported earlier were seen and two of the clones studied had two mutations each, which is very unusual. We concluded that the patient was likely to have a mild MMR deficiency in her somatic cells due to the mutations in both of her hMLH1 genes, and that this was the cause of her microsatellite instability (MSI), increased HPRT MF and abnormal HPRT mutational spectrum. The HPRT mutational spectrum was studied in 73 T-cell clones each from smoking and nonsmoking lung cancer patients. The proportions of different types of mutations, were not significantly different between smokers and nonsmokers, although the smokers had less deletions. The distribution of basepair substitutions was nonrandom, with clustering at previously identified hotspots at positions 143, 197 and 617 of the HPRT coding sequence. One additional hotspot at position 606 was observed, in smokers only. The frequency of GC>TA transversions (13%) was higher in smokers than in nonsmokers (6%). Conversely smokers had a lower frequency of GC>AT transitions (24 %) than nonsmokers (35 %).

We concluded that there was a minor effect of smoking on the HPRT mutational spectra, with a trend for increase of GC>TA transversions and decrease of GC>AT transitions, in the smokers compared to the nonsmokers. This is consistent with the in vitro mutagenicity of benzo(a)pyrene, one of the prominent carcinogens of tobacco smoke. In conclusion, these results show that analysis of HPRT mutations may contribute to the understanding of somatic mutagenesis in vivo, and that the mutational spectrum at the HPRT locus may reflect abnormalities of repair and extensive environmental exposure, such as tobacco smoking.