ENRICHED SSCP - A HIGHLY SENSITIVE METHOD FOR THE DETECTION OF UNKNOWN MUTATIONS, APPLICATION TO THE MOLECULAR DIAGNOSIS OF LUNG-CANCER IN SPUTUM SAMPLES

Detection of gene mutations by sensitive polymerase chain reaction (PC R)-based methods can allow to identify occult neoplastic cells in a gr eat excess of nonmalignant cells. These molecular approaches have an e normous potential in terms of early diagnosis, detection of occult mic rometastases of solid tumors, and minimal residual disease in patients with hematopoietic malignancies. Currently, the applications of such methods are limited, mainly because the high sensitivity required for the identification of rare mutated alleles can be achieved only in cas es in which mutations occur in few specific codons of a gene or when t he mutation is already known, No methods are available by which few al leles with unknown mutations in tumor genes can be recognized in a gre at excess of wild-type alleles. We have developed an extremely sensiti ve method, termed enriched single-strand conformational polymorphism ( E-SSCP), which permits detection of a rare alleles with unknown mutati ons. The method is based on the observation that after a conventional SSCP analysis the vast majority of mutated bands migrate close to the wild-type bands. The area of the gel having the highest chance to hold mutated alleles is physically isolated and is used as a substrate for a second round of SSCP. Serially diluted DNA samples containing gene mutations demonstrated detection of 1 mutant/10(6) normal alleles. The E-SSCP assay was first applied to six sputum samples of patients affe cted by lung cancers with known p53 mutations showing in sputa the sam e mutations observed in tumors. The technique was then applied to eigh t cytologically negative sputum samples obtained from patients who lat er developed a clinically manifested lung carcinoma. In three cases, h arboring a p53 mutation in tumor tissue, the E-SSCP analysis allowed t he detection of the mutations in sputa months before clinical diagnosi s. In conclusion, we have presented a general, highly sensitive techni que for the detection of unknown mutations that may have several poten tial applications and may hold considerable promise for the early dete ction and study of cancer.