We are pursuing the hypothesis that prostate tumors from current, past, and never smokers exhibit differences in their gene expression profiles that are consistent with distinct oncogenic molecular alterations in tumors of current smokers. We are also exploring the effects of nicotine in human prostate cancer cells and TRAMP mice, and are evaluating whether those resemble smoking-associated alterations in prostate tumors. This research is aimed at identifying the mechanisms by which cigarette smoking induces prostate cancer progression, and to define the specific role of nicotine in this process. This proposal combines novelty with a high-impact concept. If we find that nicotine induces disease metastasis, the results could have significant public health implications. Tobacco smoke contains numerous chemicals, including many that are DNA-damaging and carcinogenic. Nitrosamines that are produced from the alkaloid nicotine during post-harvest processing and the burning process of cigarettes are an important group of carcinogens in tobacco smoke. Recently, nicotine and nitrosamines were found to activate signaling pathways in non-neuronal mammalian cells by receptor-mediator mechanisms. Several of these pathways are cancer-related and promote cell survival, angiogenesis, and metastasis. For example, nicotine activates the Akt pathway, which is a key pathway in the development and progression of many cancers, including prostate cancer. In addition, nicotine can reach high nanomolar steady-state concentrations in the blood of current smokers that may activate a signaling pathway in organs other than lung if the appropriate receptors are expressed by the target cells. We collected 67 prostate tumors from 16 current, 28 past, and 23 never smokers for the study, which we obtained from the NCI CPCTR, our resource contract, and Johns Hopkins Medical Institutions. The clinical characteristics of these tumors are similar among current, past, and never smokers. In a pilot, we analyzed the gene expression profiles of tumors from 9 current, 21 past, and 17 never smokers. This analysis revealed a very distinct signature that differentiated tumors from current smokers from those of never and past smokers. Because the first dataset contained tumors from only 9 current smokers, additional prostate tumors were collected at the Department of Urology, Johns Hopkins Medical Institutions (in collaboration with Jun Luo and William Isaacs) and combined with the existing samples to increase the statistical power of our study to identify additional genes that are differentially expressed between current and never/past smokers. The comparison of tumors from current and never smokers yielded 98 transcripts encoding 73 differentially expressed genes. A second comparison, current versus past/never smokers, resulted in a shorter list of only 70 transcripts encoding 40 differentially expressed genes. Likely a residual effect of smoking on the tumor gene signature in past smokers, this study yielded fewer genes when the current and combined past/never smokers were compared. Many of the differentially expressed genes have known immune-regulatory functions. The list also contained interleukin 8, and several others of the differentially genes were found to have an association with hepatocyte growth factor (HGF). The latter is intriguing because both nicotine and HGF activate common pathways, e.g., the PI3 kinase-Akt axis. Some of our observations are preliminary and will need further validation. Nevertheless, the data show that a current smoking status generates a gene signature in prostate tumors that could reveal the mechanism by which smoking causes the metastatic spread of prostate cancer. These observations are being studied. In FY09, we performed invasion and migration assays with human prostate cancer cells (22Rv1, PC3) using physiologically relevant concentrations of nicotine, HGF, or both, in the presence and absence of a nicotinic acetylcholine receptor antagonist. These experiments revealed an enhancing effect of nicotine on invasion in these cells and an additional additive effect of nicotine and HGF in PC3 cells that are consistent with the promotion of metastasis. We also performed alpha-/beta-integrin binding assays in the presence and absence of nicotine. These assays revealed that nicotine modulates alpha-integrin, but not beta-integrin, binding and adhesion of prostate cancer cells to matrix molecules. These data support our hypothesis that nicotine has oncogenic properties in prostate cancer by altering adhesion properties of cancer cells. Finally, we tested whether plasma interleukin-8 levels are higher in prostate cancer patients who are current smokers when compared to patients who never smoked. Consistent with our expression profiling data in the tumors of current smokers which showed increased interleukin-8 expression, we found higher plasma interleukin-8 levels in plasma of smokers when compared with never smokers. The data indicate that increased interleukin-8 could be one of the mechanisms that lead to the increased prevalence of metastatic prostate cancer among current smokers. To further advance our observations from tumors and human cancer cells, we begun to study the effect of nicotine on prostate cancer progression in TRAMP mice. This animal study is being conducted in collaboration with Arthur Hurwitz (Laboratory of Molecular Immunoregulation, CCR). In our pilot study, we treated TRAMP mice (n = 25 per group) with 100 and 250 microgram/ml nicotine in the drinking water for 12 weeks. Analysis of blood nicotine levels in these mice showed that the treatment leads to nicotine steady state levels that are comparable with nicotine steady state levels in blood of current smokers. The end points of the pilot were tumor incidence and metastasis. Our analysis of these two parameters showed that nicotine-treated mice develop more prostate tumors than mock-treated animals although this difference was not statistically significant. In addition, tumor metastasis was only observed among nicotine-treated TRAMP mice after 12 weeks of treatment. To further corroborate the findings, we will repeat the study with more animals per treatment group. In summary, our preliminary findings indicate that nicotine may increase tumor burden and metastasis in an animal model of prostate cancer.
