RAS is the most frequently mutated oncogene in human cancers. Despite decades of effort, anti-RAS therapies have remained elusive. Isoprenylcysteine carboxylmethyltransferase (ICMT) methylates RAS and other CaaX-containing proteins, but its potential as a target for cancer therapy has not been fully evaluated. We crossed a Pdx1-Cre;LSL-KrasG12D mouse, which is a model of pancreatic ductal adenocarcinoma (PDA), with a mouse harboring a floxed allele of Icmt. Surprisingly, we found that ICMT deficiency dramatically accelerated the development and progression of neoplasia. ICMT-deficient pancreatic ductal epithelial cells had a slight growth advantage and were resistant to premature senescence by a mechanism that involved suppression of cyclin-dependent kinase inhibitor 2A (p16INK4A) expression. ICMT deficiency precisely phenocopied Notch1 deficiency in the Pdx1-Cre;LSL-KrasG12D model by exacerbating pancreatic intraepithelial neoplasias, promoting facial papillomas, and derepressing Wnt signaling. Silencing ICMT in human osteosarcoma cells decreased Notch1 signaling in response to stimulation with cell-surface ligands. Additionally, targeted silencing of Ste14, the Drosophila homolog of Icmt, resulted in defects in wing development, consistent with Notch loss of function. Our data suggest that ICMT behaves like a tumor suppressor in PDA because it is required for Notch1 signaling.
Authors
Helen Court, Marc Amoyel, Michael Hackman, Kyoung Eun Lee, Ruliang Xu, George Miller, Dafna Bar-Sagi, Erika A. Bach, Martin O. Bergö, Mark R. Philips
(A) Notch signaling in mammalian cells quantified with a CSL firefly luciferase reporter. Shown are the ratios of CSL firefly/CMV Renilla in U2OS cells containing doxycycline-inducible FLAG-Notch1-GFP. Cells were cocultured with the indicated “stromal cells” in the presence or absence of doxycycline (Dox) and siRNA targeting ICMT or a nontargeting control. The stromal cells were NIH3T3 (3T3) with or without expression of the Notch1 ligand Jagged-2 (J2). The γ-secretase inhibitor compound E was included as a control and completely blocked Notch signaling. All ratios were normalized to the maximum in each experiment. Data shown are mean ± SEM, n = 3. Panel on right shows a representative knockdown by immunoblotting for ICMT and a control (ERK1/2). (B) Wings of D. melanogaster transgenic for UAS-shSte14, a GAL4-responsive hairpin that silences Icmt. GAL4 expression was driven by a wing-specific (Ap-GAL4) promoter. ICMT deficiency in the developing wing phenocopies the terminal vein bifurcation (arrows) and thickened cross-vein (arrowhead) observed in the wings of Delta (Dl) flies deficient for the Notch ligand. (C) GAL4 expression driven with an eye-specific (GMR-GAL4) promoter. The rough eye phenotype is also consistent with that seen in Notch loss-of-function alleles, including Dl. (D) Wing imaginal discs from third instar larvae expressing shSte14 in heat shock–induced clones marked with GFP and stained for CUT, a Notch-dependent gene product. Where the GFP-positive clone intersects the line of CUT staining there is a decrease in the number of CUT-positive cells (enlargement). Scale bar: 50 εm.