The transcription factor STAT3 is a point of convergence for multiple signaling pathways and is activated in both tumor and non-tumor stromal cells in the tumor microenvironment. STAT3 activation mediates the expression of genes involved in many aspect of tumorigenesis. Thus, STAT3 has been hotly pursued as a target for cancer. Although inhibitors of STAT3 upstream signaling pathways have shown anti-tumor activity in preclinical animal models, the relative contributions of STAT3 in tumor cells versus tumor-associated stromal cells to the observed antitumor effects remains unclear.
To address this question we have developed potent next generation (Generation 2.5) antisense oligonucleotides (ASOs) that are selective for either mouse or human STAT3 [This is the first generation 2.5 antisense program added to the ISIS pipeline] and we have assessed their specificity, potency, and antitumor activities in several preclinical cancer models. The safety of targeting STAT3 with ASOs was first confirmed in both rodents and non-human primates.
STAT3 ASOs were very well tolerated at doses that resulted in complete abrogation of STAT3 expression in the liver of treated animals, Next, we evaluated the relative contribution of STAT3 in tumor cells and tumor-associated stromal cells in a human ovarian cancer xenograft model (SKOV3), using both mouse and human specific STAT3 ASOs. Mouse and human STAT3 ASOs selectively reduced their respective target STAT3 mRNA levels in SKOV3 tumors. However, when we evaluated the level of IL-6, a key cytokine in ovarian cancer and known to be regulated by STAT3, it was significantly reduced only by the inhibition of the STAT3 in the tumor associated-stromal cells (mouse STAT3). Moreover, the antitumor activity of the mouse STAT3 was superior to that observed with inhibition of STAT3 in the tumor cells themselves, suggesting that the tumor promoting effects of STAT3 in this model is mediated through STAT3 activation in tumor-associated stromal cells. To evaluate the effects of STAT3 inhibition in a mouse model of cancer we employed the APCmin model in which mice develop intestinal neoplasia at a young age. Treatment of APCmin mice with mouse STAT3 ASOs led to a >90% inhibition of STAT3 mRNA levels in both the small intestine and colon. In addition, STAT3 ASO treatment resulted in a significant decrease in the number of polyps in the intestine. Finally, we evaluated the effects of human-specific STAT3 ASO in NSCLC s.c xenograft models using either PC9 or primary human tumor explants. Human STAT3 ASO treatment resulted in >50 % inhibition in STAT3 mRNA levels in the tumor cells and resulted in significant tumor growth inhibition (~50% and 90% >, respectively), suggesting that relative contributions of mouse stromal cells to tumor growth may vary in different tumor types.
Taken together, these results demonstrate that STAT3 ASO can selectively downregulate STAT3 in both tumor and stromal compartments, affect anti-tumor activity in vivo and that stromal derived STAT3 activity is important contributor to tumor growth and survival.