We performed a detailed and comprehensive study of the involvement of tumor suppressor genes in human prostate cancer. We utilized primers flanking either the restriction fragment length polymorphism (RFLP) or variable number of tandem repeat [VNTR; microsatellite or simple repeat site (SRS)] polymorphic sites to polymerase chain reaction (PCR) amplify the genomic DNA and detect loss of heterozygosity of the target genes. Quantitative reverse transcription (RT)-PCR was performed to measure the mRNA expression levels and PCR/single strand conformational polymorphism (SSCP) and DNA sequencing carried out to detect mutation of the tumor suppressor genes. We found that multiple tumor suppressor genes (e.g., p53, DCC, APC, MCC, BRCA1, and WAF1/CIP1) were inactivated at different frequencies via various mechanisms [e.g., loss of heterozygosity (LOH), loss of expression (LOE), mutation, and inactivation by cellular binding protein]. Several important and novel findings are as following: LOH and LOE of the DCC gene, LOH, LOE, and possible mutation of the APC/MCC genes, LOH of the BRCA1 locus, and mutation of the WAF1/CIP1 gene. For p53 tumor suppressor gene alone, multiple inactivation mechanisms (i.e., LOH, LOE, mutation, and amplification of the cellular inactivating protein MDM2) were identified. A possible involvement of genomic instability or mutator phenotype in human prostate cancer was investigated by microsatellite typing using PCR. A high frequency of microsatellite instability was detected and the microsatellite instability found to correlate with advanced stage and poor differentiation of prostate cancer, suggesting that genes functioning in DNA mismatch repair or general stabilization of the genome may be involved in prostate cancer. The results obtained in this study suggested that multiple tumor suppressor genes (both known and unknown genes) may share the role in prostate cancer; a pattern which has been found in a number of human malignancies such as cancers of the esophagus, colon and breast. In fact, we performed deletion studies aimed at localizing potential tumor suppressor loci on various chromosomal regions. A number of chromosomal regions (i.e., 6p12-24 and 17q21) were found to potentially harbor unidentified tumor suppressor genes. Detailed deletion mapping has localized the potential tumor suppressor loci to a < 2 Mb region centromeric to the BRCA1 gene on chromosome 17q. In addition, we identified a number of novel mechanisms of tumor suppressor gene inactivation, in prostate cancer such as loss of mRNA expression of the DCC, APC, MCC and p53 gene, and mutator phenotype. And for the very first time, we identified somatic mutations of the WAF1/CIP1 gene in primary human malignancy-human prostate cancer. This finding provides the first evidence in primary tumor that the WAF1/CIP1 gene may be a tumor suppressor gene and may be involved in prostate cancer. We identified 12-lipoxygenase (12-LOX) as a potential prognostic marker for human prostate cancer. mRNA expression levels of the 12-LOX gene was measured by quantitative reverse transcription-polymerase chain reaction (RT-PCR) and semi-quantitative in situ hybridization (ISH) in 122 pairs of matched normal and tumor tissues from prostate cancer patients. We found that 12-LOX expression levels were elevated in approximately half of the patients analyzed and the 12-LOX elevation correlates with advanced stage, poor differentiation, and surgical margin positivity. Our data suggest that 12-LOX may serve as a correlative marker for a more aggressive phenotype of prostate cancer and therefore for poor prognosis. We are currently refining our assays for possible clinical applicability. Since not all patients with loss of expression of the DCC gene showed LOH of the DCC locus, there must be other mechanism(s) responsible for loss of expression of the DCC gene. When we analyzed the relationship between DCC loss of expression and 12-LOX elevation in prostate cancer pati