UCLA

Suppressor of Cytokine Signaling (SOCS) proteins are inhibitors of cytokine signaling that function via the Janus kinase (JAK)/ signal transducers and activators of transcription (STAT) pathway. Our lab has previously demonstrated that SOCS3-deficient mouse embryonic fibroblasts (MEFs) display enhanced sensitivity to DNA damaging agents, but the molecular mechanism of this effect is not well understood. In this study, we have confirmed SOCS3 as a radioprotector in fibroblasts and human solid tumors and identified a novel mechanism through which SOCS3 exerts its radioprotective effect. SOCS3 regulates the DNA damage response by promoting p21 induction and controlling STAT3-mediated proliferation. We have found that the transcriptional repressor BCL6 is targeted by SOCS3/STAT3 and this novel axis is responsible for specific control of the G1-S transition, as SOCS3-deficient cells display normal S-phase checkpoint. Further, by overexpressing SOCS3 in human pancreatic cancer cell lines that signal through insulin-like growth factor (IGF) receptor and the RON receptor tyrosine kinase, we have determined that the radioprotective effect of SOCS3 supercedes its ability to block specific kinase pathways that increase radiation resistance in tumor cells.

In addition to these novel effects of SOCS3 on the DNA damage response, we have also identified radiation-independent functions for SOCS3 in aggressive, metastatic solid tumors. Initial cDNA microarray studies in SOCS3 knockout MEFs revealed that the pro-inflammatory C-X-C family chemokines CXCL1 and CXCL2 are underrepresented in SOCS3-deficient fibroblasts, due to transcriptional repression by uncontrolled STAT3 signaling. On the basis of these observations, we have identified three molecular subtypes of human pancreatic ductal adenocarcinoma (PDA) based on SOCS3 expression and CXCL1/2 dependence. We have stratified human PDA cell lines into three groups in which SOCS3 is either highly expressed, partially repressed, or severely repressed/totally silenced in response to IL-6 or oncostatin M (OSM). Repression of the SOCS3 gene in PDA is not CpG island methylation-dependent, suggesting a possible novel mechanism through which SOCS3 expression is silenced in PDA. Our laboratory has shown that ectopic expression of SOCS3 can promote CXCL1 gene activation and tumor aggressiveness in the molecular subtype with partial SOCS3 repression but not the others. Thus, the three molecular subtypes of PDA that we have identified display distinct biological behaviors, responses to SOCS3 overexpression and CXCL1-dependence. Using an orthotopic xenograft model of human PDA in immunodeficient mice, we demonstrated the in vivo relevance of the STAT3/SOCS3/CXCL1 axis in tumor aggressiveness and metastatic behavior. Our results suggest a novel mechanism for controlling pro-inflammatory signaling in human solid tumors and that SOCS3 can regulate both metastasis and radiation sensitivity through distinct pathways. Our ability to stratify human PDA on the basis of SOCS3/STAT3/CXCL1/2 dependence has implications for the development of individualized therapy for cancer patients. By targeting STAT3, SOCS3, as well as CXCL1, we can regulate metastatic behavior in solid tumors, and increase radiation sensitivity of tumors in the clinic.