Mutation of the androgen-receptor gene in metastatic androgen-independent prostate cancer.

BACKGROUND: Metastatic prostate cancer is a leading cause of cancer-related death in men. The rate of response to androgen ablation is high, but most patients relapse as a result of the outgrowth of androgen-independent tumor cells. The androgen receptor, which binds testosterone and stimulates the transcription of androgen-responsive genes, regulates the growth of prostate cells. We analyzed the androgen-receptor genes from samples of metastatic androgen-independent prostate cancers to determine whether mutations in the gene have a role in androgen independence. METHODS: Complementary DNA was synthesized from metastatic prostate cancers in 10 patients with androgen-independent prostate cancer, and the expression of the androgen-receptor gene was estimated by amplification with the polymerase chain reaction. Exons B through H of the gene were cloned, and mutations were identified by DNA sequencing. The functional effects of the mutations were assessed in cells transfected with mutant genes. RESULTS: All androgen-independent tumors expressed high levels of androgen-receptor gene transcripts, relative to the levels expressed by an androgen-independent prostate-cancer cell line (LNCaP). Point mutations in the androgen-receptor gene were identified in metastatic cells from 5 of the 10 patients examined. One mutation was in the same codon as the mutation found previously in the androgen-independent prostate-cancer cell line. The mutations were not detected in the primary tumors from of the two patients. Functional studies of two of the mutant androgen receptors demonstrated that they could be activated by progesterone and estrogen. CONCLUSIONS: Most metastatic androgen-independent prostate cancers express high levels of androgen-receptor gene transcripts. Mutations in androgen-receptor genes are not uncommon and may provide a selective growth advantage after androgen ablation.

Detection of sequence-specific antitumor alkylating agent DNA damage from cells treated in culture and from a patient.

Detection of sequence-specific DNA damage induced by antitumor alkylating agents might provide a mechanism for detecting and discriminating damage specific to one or more of these drugs. Using repetitive primer-extension and human alphoid DNA as a substrate, lesions specific for an activated form of cyclophosphamide, 4-hydroperoxycyclophosphamide, were detected at 32 of 33 guanines within a 200-base pair region in DNA from cells treated in culture. There was a marked variation in lesion site intensity among affected guanines. For instance, guanines flanked by cytosine were weak sites of 4-hydroperoxycyclophosphamide-induced damage. Damage at bases other than guanine induced by cisplatin, UV irradiation, and adozelesin were compared to drug-DNA lesions induced by 4-hydroperoxycyclophosphamide. Using this method it was possible to detect, and at some sites distinguish, between cyclophosphamide- and cisplatin-induced DNA damage within WBC DNA from a patient treated with both agents. There was a different damage pattern for DNA derived from cells treated in culture compared to DNA derived from the patient sample.

Differences in in vivo and in vitro sequence-specific sites of cisplatin-DNA adduct formation and detection of a dose-response relationship.

The cytotoxic and mutagenic properties of the anticancer drug cis-diammine-dichloroplatinum(II) (cisplatin) are mediated by bifunctional adducts between purines. Experiments performed in this study employed a new repetitive thermal-cycling technique to detect cisplatin adduct formation following exposure of cells in culture (in vivo) or following treatment of purified DNA (in vitro exposure). The initial goal of this study was to determine if cisplatin-DNA adduct formation could be measured accurately using phosphor-imaging over a broad concentration range. If this proved possible, it would then be feasible to determine if adduct formation differed within chromatin compared with purified DNA. There were no significant differences in the cisplatin-DNA adduct pattern induced in closed circular or linear double-stranded plasmids in vitro, suggesting that this type of tertiary structural change does not affect the formation of adduct sites. Sequence-specific DNA adduct formation within a human repetitive DNA target sequence, alphoid DNA, following cisplatin treatment of prostate cancer cells in culture (in vivo) and treatment of purified DNA in vitro revealed consistent increases in adduct formation over a broad concentration range, validating the experimental technique. Comparing preferences for cisplatin adduct site formation under these different conditions of exposure demonstrated statistically significant differences. Similar differences were detected for cisplatin repair-deficient Xeroderma pigmentosum cells treated in cell culture, indicating that in vivo/in vitro preferences for adduct site formation are not the result of DNA repair in vivo.