We screened human prostate cancer tissues for the presence of somatic mutations in the hormone binding domain of the androgen receptor (AR) gene. Exons E-H were amplified from genomic DNA using the polymerase chain reaction and analyzed by denaturing gradient gel electrophoresis (DGGE), which separates DNA fragments that differ by only a single base. We detected a mutation in exon E of the hormone binding domain in 1 of 26 specimens of untreated organ-confined stage B prostate cancer. The mutation was not detectable in peripheral blood lymphocyte DNA. Lymphocyte DNA (wild-type AR) migrated in DGGE as a single band. The tumor DNA migrated in DGGE as four bands, consistent with the presence of cells with mutant AR plus cells with wild-type AR and indicating that the tumor contained a somatic mutation. To our knowledge, a somatic AR gene mutation has not been reported previously. Sequencing revealed a G --A substitution in codon 730, changing valine to methionine. Codon 730 is in a region highly conserved among all steroid receptors. The abundance of the mutated fragment (about 50% of the DNA in the specimen) indicates its presence in cells with a growth advantage. A somatic mutation could be detected by DGGE if it represented at least 10% of the sample. Failure to detect mutations in other specimens analyzed may be due to this limit of sensitivity, the presence of mutations in other parts of the AR, or a low frequency of mutations in early stage disease.Prostate cancer is the most frequently diagnosed cancer and the second leading cause of cancer deaths in men in the U.S.(1). Recognition that androgen is required for the development of prostate cancer (2) and its growth (3) has been the basis for continuing interest in the role of the androgen receptor (AR) in prostate cancer (1,(4)(5)(6). The AR is a member of the superfamily of genes that code for the steroid and thyroid hormone receptor family of ligand-dependent nuclear transcription factors, all of which have an N-terminal domain that affects transcription efficiency, a central DNA

Transforming growth factor-beta 1 (TGF beta 1) is a potent modulator of cell proliferation, differentiation, angiogenesis, and the immune system. TGF beta 1 messenger RNA (mRNA) levels were much higher in several rat prostate adenocarcinomas (Dunning R3327 MATLyLu, AT2, G, HI, and H sublines) than in normal prostate. Normal prostate and the well differentiated H and HI tumors produced two TGF beta 1 mRNA transcripts, 2.4 kilobases (major) and 1.6 kilobases (minor). The poorly differentiated MATLyLu and AT2 sublines produced these plus additional TGF beta 1 mRNA transcripts that were present in the primary tumors, metastases, and cultured cell lines. TGF beta 1 mRNA levels were unchanged 2 weeks after castration. Immunohistochemical staining of TGF beta 1 protein was more prominent and more extensive in prostate cancer than in normal prostate. Only extracellular TGF beta 1 staining was detected. In normal prostate and in well differentiated tumors (H and HI), extracellular TGF beta 1 staining was located in the interacinar stroma, suggesting that it may be produced there. In contrast, in the poorly differentiated tumors (MATLyLu, AT2, and G) that contain sheets of epithelial cells, extracellular TGF beta 1 staining was present throughout the tumor, suggesting that TGF beta 1 may be made and secreted by the tumor epithelial cells. MATLyLu, AT2, and G tumor cells were cultured in vitro, and the conditioned medium was analyzed for the presence of TGF beta using a bioassay. TGF beta 1 is produced and secreted as an inactive latent precursor and must be activated to release bioactive TGF beta 1. Cells secreted about 100-500 pg TGF beta/10(6) cells.24 h and were able to activate about 50% of the total TGF beta secreted. Because TGF beta 1 mRNA and protein expression are higher in cancerous than normal tissue and because prostate cancer cells themselves can activate latent TGF beta 1 to a bioactive form, TGF beta 1 produced endogenously by prostate cancer has the potential to affect tumor behavior in vivo. Therefore, TGF beta 1 may represent a new therapeutic target in prostate cancer.

No abstract