In this study, we examined the effect of NE on the expression of factors important in melanoma tumor progression. We found that exposure of the human melanoma tumor cell line, C8161, to NE resulted in the induction of VEGF, IL-8, and IL-6 protein levels. This NE-dependent effect was not as strong in the 1174MEL and Me18105 cell lines as compared to the C8161 cell line. We show that the upregulation of the levels of these proteins in culture supernatants is not due to a stimulation of cell proliferation. In contrast to the observed difference in the NE-dependent modulation of protein levels in the three cell lines, we show that NE efficiently upregulated VEGF, IL-8, and IL-6 mRNA levels in all three cell lines. We further show that the expression of the three genes in C8161 cells is mainly due to increased de novo transcription. We show that C8161, 1174MEL and Me18105 cells expressed different levels of the β1- and β2-AR proteins and the mRNA that encode them (the ADRB1 and ADRB2, respectively). It is currently not known what accounts for the difference in the transcriptional and translational effects of NE on these three proteins. Evidence supporting the role of β-ARs in the NE-dependent effect is provided by the results showing that propranolol completely inhibited the NE-dependent upregulation of gene expression in all three cell lines, and that the agonists dobutamine and terbutaline stimulated VEGF, IL-8, and IL-6 gene expression in C8161 cells.
The activation of the hypothalamic-pituitary-adrenal (HPA) and the SAM axes that results in an increase of stress hormone levels have been shown to induce immune dysregulation significant enough to produce health outcomes, such as the slowing of wound healing (Glaser and Kiecolt-Glaser, 2005
; Kiecolt-Glaser et al., 1995
; Mercado et al., 2002
). Stress-associated immune dysregulation can upregulate the production of proinflammatory cytokines, such as IL-1, TNF-α, and IL-6 (Glaser and Kiecolt-Glaser, 2005
; Gomez-Merino et al., 2005
; Johnson et al., 2005
; Kiecolt-Glaser et al., 2005
; Segerstrom and Miller, 2004
). Inflammation has been linked to a significant percentage of cancers suggesting another alternative whereby stress could be a co-factor for modulating the growth and progression of tumor cells. Stress-induced changes in the cellular immune response results in decreased antigen-specific T-cell and natural killer (NK) cell responses and have implications for the immune response to immunogenic tumors (Glaser and Kiecolt-Glaser, 2005
; Padgett and Glaser, 2003
). Studies have also implicated stress-related effects on immune function as having an influence on tumor progression (Ben-Eliyahu, 2003
; Goldfarb and Ben-Eliyahu, 2007
; Lamkin et al., 2008
; Reiche et al., 2004
; Saul et al., 2005
). The data obtained in this study support the hypothesis of an alternative pathway linking stress with cancer progression via effects on other biological processes involving the SAM axis (Antoni et al., 2006
; Reiche et al., 2004
; Saul et al., 2005
; Thaker et al., 2007
Work by Entschladen and others have indicated a role for catecholamines in tumor progression (Entschladen et al., 2004
). For example, catecholamines have been shown to promote the migration of breast and colon cancer cells in vitro
(Drell et al., 2003
; Masur et al., 2001
) and the metastasis of human PC-3 prostate cancer cells in nude mice (Palm et al., 2006
), effects that could be inhibited by the addition of β-blockers. The data from this study suggest that NE may further enhance tumor progression by stimulating the process of angiogenesis. The processes promoted by the NE-dependent upregulation of VEGF, IL-8, and IL-6 expression are yet to be fully understood. However, these factors have been implicated in several processes associated with melanoma tumor progression including the angiogenic process (Mahabeleshwar and Byzova, 2007
). For example, it has been shown that VEGF released by melanoma cells is an important mediator of neo-vascularization and has been shown to be a marker for melanoma progression (Löffek et al., 2006
; Osella-Abate et al., 2002
). In addition, serum levels of IL-6 have been associated with advanced stages of the disease, the ability to discriminate progressive from non-progressive disease, and have been used as a serum marker for monitoring response to chemotherapy and survival. Furthermore, IL-8 and IL-6 are released in the culture supernatants of highly invasive melanoma cells (Bar-Eli, 1999
; Payne and Cornelius, 2002
). Both these factors are involved in angiogenesis by either directly stimulating the proliferation and differentiation of endothelial cells or by inducing the production of additional factors in other cells (Bar-Eli, 1999
Since the recent studies by Sood and others suggest that the NE-dependent effects on the growth of ovarian tumor cells in nude mice are mediated through the tumor cell cAMP-PKA signaling pathway (Thaker et al., 2006
), we focused on this signaling pathway in our study. Our results show that isoproterenol is a potent inducer of VEGF, IL-8 and IL-6 gene expression in C8161 cells supporting the role of β-ARs in this pathway. In addition, our results show that the adenylate cyclase agonist forskolin stimulated the expression of the VEGF, IL-8 and IL-6 genes while the cAMP analogs 8-CPT and 6-Bnz-cAMP did not efficiently stimulate the expression of the three genes. Furthermore, our observation that H-89 effectively inhibited the NE-dependent upregulation of IL-8 and IL-6 mRNA levels in C8161 cells while PKI did not can be explained by the fact that H-89 has inhibitory effects on the activities of protein kinase G (PKG) and PKC as well (Chijiwa et al., 1990
; Hidaka and Kobayashi, 1992
). These observations suggest that the cAMP-PKA signaling pathway may play a limited role in mediating these effects and further elucidation of the signaling pathway involved has yet to be addressed.
It is of interest that we are seeing differences in the expression of pro-angiogenic proteins in cells derived from different kinds of tumors in spite of the fact that the tumor cells express β-ARs and that the these proteins can be induced by NE in each study (Yang et al., 2008
; Yang et al., 2006
). Although NPC, multiple myeloma-derived, and melanoma cell lines all express immunoreactive bands of the same relative molecular weights in Western blots detecting β1- and β2-ARs, the abundance of the specific bands differed among these types of cancers. We still do not know how these differences play a part in their varying responses to NE. Our data add to the growing literature showing that the modulatory effects of psychological factors may have important clinical relevance and may lead to new approaches for treating cancer.
Consistent with our NPC study, we also observed the expression of β2-AR on tumor cells in a majority of melanoma biopsies supporting the possible clinical relevance of our in vitro results. It is important to note that in this study we examined different classes of melanoma, i.e., primary melanoma (nodular, superficial spreading and desmoplastic) and metastatic melanoma (lymph node and visceral metastasis). We found that these classes of melanoma express the β1-AR as well. That a majority of the melanoma tumors we examined expressed β-ARs supports our hypothesis that melanoma tumor cells at various stages of the disease have the potential to respond to NE.
Derived from neural crest stem cells, human malignant melanoma is one of the most deadly of human neoplasms, exhibiting aggressive metastatic behavior even when its primary tumor is very small (Safarians et al., 1996
). Whereas most human cancers metastasize when they reach a size of 1 cm diameter, melanomas can metastasize when they are only 1 mm in size demonstrating a 1000 fold increase in inherent metastatic potential (Barsky et al., 1997
). Our traditional methods of staging and predicting distal metastasis and melanoma recurrence are based on measurements of tumor thickness in the skin and presence of tumor in regional lymph nodes but these prognostic markers fall short in the majority of cases where the likelihood of tumor recurrence and metastasis remains indeterminate (Balch et al., 2001
). Although a number of genetic abnormalities have been identified in human melanomas including significant alterations involving the BRAFpathway (Haluska et al., 2007
), to date no molecular subclassification scheme has improved upon the prognostic information imparted by thickness alone (Palmieri et al., 2007
). Since it can be assumed that the overall tumor burden at presentation is low in the majority of cases, it is likely that host rather than tumoral factors alone regulate recurrence or cure. Stress may be one important host cofactor. Certainly the presence of adrenergic receptors and signaling in human melanoma cell lines and cases of human melanoma support this hypothesis.
Our data support the hypothesis that the stress hormone, NE, can stimulate the aggressive potential of melanoma tumor cells by inducing the release of proteins, including VEGF, IL-8, and IL-6. The role(s) that psychological factors may play in tumor progression is not fully understood but the connection between the HPA and SAM axes that have been described by our laboratory and others may add to our understanding of the role of psychological stressors on the progression of some tumors. The fact that these three factors are upregulated in C8161, a cell line known for its highly metastatic phenotype, and not to the same extent in 1174MEL and Me18105, supports the hypothesis that different tumor cell populations may respond differently to the release of stress hormones and that stress can be a co-factor for the progression of a highly metastatic cancer. It is not known whether 1174MEL and Me18105 produce metastatic tumors in a nude mouse. To our knowledge, this is the first report to suggest that stress-related activation of the SAM axis may have a role in human melanoma tumor progression.