Androgenic regulation of growth factor and growth factor receptor expression in the CWR22 model of prostatic adenocarcinoma.

The effects of androgen manipulation on epidermal growth factor (EGF) receptor, p185erbB-2 and transforming growth factor-alpha (TGF-alpha) levels were examined in prostatic adenocarcinoma. Male nude mice were inoculated with the CWR22 androgen-dependent human prostatic tumor or an androgen-independent (CWR22R) derivative. Mice with CWR22 tumors were castrated and subsequently killed at 3, 7, 21, 28 or 42 days post-castration. Other CWR22-bearing mice received s.c. testosterone pellets at 21 days post-castration and were killed 7 days later. EGF receptor, p185erbB-2 and TGF-alpha levels were examined by immuno-histochemistry. Strong EGF receptor and p185erbB-2 immunostaining was detected in CWR22 tumors from intact controls. EGF receptor immunostaining decreased by 65% to 70% at 21 to 42 days post-castration. Testosterone treatment at 21 to 28 days post-castration resulted in a 2-fold increase in EGF receptor immunostaining. p185erbB-2 immunostaining within CWR22 tumors did not decrease following castration and, in fact, was slightly increased at 7 days post-castration. The effects of castration on EGF receptor and p185erbB-2 levels were confirmed by Western blot analysis. Fewer than 10% of CWR22 tumor cells demonstrated strong TGF-alpha immunostaining, and androgen manipulation did not effect TGF-alpha immunostaining. In contrast, 30% of androgen-independent CWR22R tumor cells were strongly immunostained for TGF-alpha. Our findings indicate that EGF receptor levels, but not p185erbB-2 levels, are strongly dependent on testosterone in CWR22 tumors. The co-localization of TGF-alpha and the EGF receptor in CWR22R tumors suggests that these factors may constitute an autocrine pathway that regulates androgen-independent growth.

Changes in cyclin dependent kinase inhibitors p21 and p27 during the castration induced regression of the CWR22 model of prostatic adenocarcinoma.

PURPOSE: The expression of the cyclin dependent kinase inhibitors p21 and p27 was examined in prostatic adenocarcinomas following castration. MATERIALS AND METHODS: Male nude mice inoculated with the androgen dependent human prostatic tumor CWR22 were castrated when the tumors reached a volume of 0.8 to 1.1 cm.3 and were sacrificed at 3, 7, 21, 28 and 42 days post-castration. An additional group of mice received a subcutaneous testosterone pellet at 21 days post-castration and was sacrificed at 28 days post-castration. The expression of the Ki-67 antigen, p21 and p27 was examined by immunohistochemistry. RESULTS: The mitotic rate as well as the number of Ki-67 antigen positive cells decreased to 3% of intact control values by 7 days post-castration and were less than 0.01% of intact control values at 21, 28 and 42 days post-castration. The percentage of p21 expressing cells decreased from 15+/-2% in intact controls to less than 1% by 42 days post-castration. In contrast, the percentage of cells that expressed p27 increased from 25+/-3% in intact controls to 51+/-8% at 3 days post-castration and to 80 to 95% at days 7, 21, 28 and 42 days post-castration. Testosterone treatment from 21 to 28 days post-castration resulted in an increase in Ki-67 antigen positive cells to 200% of intact controls and a concomitant reduction in p27 expressing cells to about 50% of intact controls. Castration-induced changes in p27 expression were not observed in the CWR22R tumor, a transplantable relapsed derivative of the CWR22 tumor. CONCLUSION: These findings suggest that p27 expression is regulated negatively by androgens and that increased expression of p27 in CWR22 xenografts may be involved in the suppression of proliferation following castration.

Polynucleotide immunization of nonhuman primates against carcinoembryonic antigen.

In preparation for a Phase I trial of DNA immunization against carcinoembryonic antigen (CEA) in patients with colorectal carcinoma, we have produced a single plasmid DNA encoding CEA and hepatitis B surface antigen (HBsAg) under transcriptional regulatory control of two separate cytomegalovirus promoters within separate eukaryotic expression cassettes, designated pCEA/HBsAg. Hepatitis B surface antigen was included to provide an internal positive control for the efficacy of this immunization strategy without regard to the issue of breaking tolerance to a self-antigen. In the present work, we sought to examine the immunogenicity of this plasmid in a nonhuman primate model with close phylogenetic relationship to humans. Groups of pig-tailed macaques were immunized with pCEA/ HBsAg by i.m. injection or particle bombardment of the skin according to a dose and schedule thought to be optimal for the respective technique of DNA immunization. Both administration techniques produced humoral and lympho-proliferative responses of comparable magnitude. However, delayed type hypersensitivity to CEA and CEA-specific interleukin-2 release were observed only in the i.m. group, suggesting a qualitative difference in the character of the immune response elicited by the two techniques of DNA immunization. The antibody responses to CEA and HBsAg were surprisingly persistent in that all immunized animals maintained moderate antibody titers against both antigens for more than 15 months after the last boost. No toxicity was observed during 2 years of follow-up, including no measurable levels of anti-DNA antibody. This antitumor immunization strategy is presently being examined in patients with metastatic colorectal carcinoma using pCEA/HBsAg administered by i.m. injection.