Identification of p53 mutations in archival prostate tumors. Sensitivity of an optimized single-strand conformational polymorphism (SSCP) assay.

To correlate molecular changes with clinical information in prostate tissue, it is necessary to have accurate methods for screening for mutations in clinically available specimens. We have refined the polymerase chain reaction-single-strand conformation polymorphism (PCR-SSCP) analysis for detection of p53 mutations in routine pathology specimens. These improvements help overcome technical barriers that interfere with SSCP analysis of archival tissues when only small amounts of poorly preserved formalin-fixed DNA are available. Furthermore, prostate samples are heterogeneous in containing tumor, normal tissue, and hyperplastic tissue. To address the first issue, the method included an initial selection of PCR products using exonuclease I, followed by a second-step selection using nested PCR. This step ensures adequate amplification of the target sequence while minimizing artifactual products that could otherwise interfere with mutation analysis. For the second issue, in addition to morphologic selection of appropriate tissue areas, we improved the sensitivity of detection of mutations by using restriction enzyme digestion of products prior to SSCP analysis. Detection of mutations in heterogeneous tissue was evaluated by determining the minimal detectable mutant allele frequencies in exons 4, 5, 6, 7, 8-9, and 10 by using mixtures of known mutant and wild-type cell lines, which were found to be 17.6, 9.1, 12.5, 8.1 14.0, and 14.3%, respectively. To determine the utility of this method when used on heterogeneous clinical samples, we performed study of 19 archival prostate specimens (14 primary prostate cancers, three benign prostatic hyperplasia and two metastases) and detected abnormally migrating products in six of the prostate cancer specimens (four primaries and two metastases). In five of these samples, there was sufficient DNA to perform sequencing, which disclosed single-base change mutations in all five samples.