A Novel Mechanism for Activation of GLI1 by Nuclear SMO That Escapes Anti-SMO Inhibitors.

Small-molecule inhibitors of the Hedgehog (HH) pathway receptor Smoothened (SMO) have been effective in treating some patients with basal cell carcinoma (BCC), where the HH pathway is often activated, but many patients respond poorly. In this study, we report the results of investigations on PTCH1 signaling in the HH pathway that suggest why most patients with BCC respond poorly to SMO inhibitors. In immortalized human keratinocytes, PTCH1 silencing led to the generation of a compact, holoclone-like morphology with increased expression of SMO and the downstream HH pathway transcription factor GLI1. Notably, although siRNA silencing of SMO in PTCH1-silenced cells was sufficient to suppress GLI1 activity, this effect was not phenocopied by pharmacologic inhibition of SMO, suggesting the presence of a second undefined pathway through which SMO can induce GLI1. Consistent with this possibility, we observed increased nuclear localization of SMO in PTCH1-silenced cells as mediated by a putative SMO nuclear/nucleolar localization signal [N(o)LS]. Mutational inactivation of the N(o)LS ablated this increase and suppressed GLI1 induction. Immunohistologic analysis of human and mouse BCC confirmed evidence of nuclear SMO, although the pattern was heterogeneous between tumors. In PTCH1-silenced cells, >80% of the genes found to be differentially expressed were unaffected by SMO inhibitors, including the putative BCC driver gene CXCL11. Our results demonstrate how PTCH1 loss results in aberrant regulation of SMO-independent mechanisms important for BCC biology and highlights a novel nuclear mechanism of SMO-GLI1 signaling that is unresponsive to SMO inhibitors.Significance: This study describes novel noncanonical Hedgehog signaling, where SMO enters the nucleus to activate GLI1, a mode that is unaffected by SMO inhibitors, thus prompting re-evaluation of current BCC treatment as well as new potential therapies targeting nuclear SMO. Cancer Res; 78(10); 2577-88. ©2018 AACR.The authors gratefully acknowledge the funding from the British Skin Foundation. The authors also thank Anna Saran (ENEA, Rome, Italy) for providing mouse BCC tissue for immunohistochemistry and Dr. Monika Cichon for the shRNA control sequences. The authors also thank the Dr. Hadwen Trust (DHT) for Humane Research for additional funding to develop human models for human disease and confirm that no funding from the DHT was used for any animal research in this article