Sema3C Signaling is an Alternative Activator of the Canonical WNT Pathway in Glioblastoma

Abstract

Wnt signaling maintains normal and cancer stem cells. The Wnt pathway is frequently dysregulated in many cancers, underscoring it as a therapeutic target. Although Wnt inhibitors appear promising in many preclinical studies, they have failed uniformly in clinical trials. Molecular mechanisms of resistance are poorly defined. Further dissection of the precise mechanisms of Wnt pathway activation in specific tumor types is needed to develop new Wnt pathway inhibitors with less toxicity. Here, we identify an alternative activator of the Wnt pathway that may mediate resistance to upstream Wnt inhibition in glioblastoma. Glioma stem-like cells (GSCs) were enriched in defined media. GSCs were transduced with lentiviruses to knockdown or overexpress Sema3C or Wnt pathway components. Cell viability, proliferation, apoptosis, and self-renewal were assessed. Expression of Sema3C and Wnt pathway components were assessed in GSCs, mouse models of GBM, and human glioblastoma by qPCR, Western blot, and/or immunostaining. Beta-catenin subcellular localization was assessed by cell fractionation and immunofluorescence. GSC-derived orthotopic models of GBM were used to assess the impact of genetic or pharmacologic inhibition of Sema3C or Wnt pathway components alone or in combination on tumor growth and animal survival. The axonal guidance protein Sema3C promotes the tumorigenicity of GSCs through binding its NRP/PlxnD1 receptor complex leading to Rac1 activation. Sema3C signaling directs beta-catenin nuclear accumulation in a Rac1-dependent process, leading to transactivation of Wnt target genes. Sema3C-driven Wnt signaling occurred despite suppression of Wnt ligand secretion, suggesting that Sema3C may drive canonical Wnt signaling independent of Wnt ligand binding. In human glioblastoma, Sema3C expression and Wnt pathway activation were highly concordant. In a mouse model of glioblastoma, combined depletion of Sema3C and beta-catenin partner TCF1 extended animal survival more than single target inhibition alone. Sema3C signaling may represent an alternative mechanism of WNT pathway activation even when WNT ligand-receptor interaction is inhibited. Since Sema3C is overexpressed in >85% glioblastoma and is used to maintain GSCs but not normal neural progenitor cells, this pathway may represent a major mechanism of Wnt pathway activation and resistance to upstream Wnt pathway inhibitors in GSCs. Our data provide a therapeutic strategy to achieve clinically significant Wnt pathway inhibition in GSCs potentially without the toxicity of currently available WNT inhibitors.