Fig 2: Degradation of Endogenous Unmodified RAS Using a RAS-Specific Monobody(A) SDS-PAGE and western blot analysis of A549 cells transduced with indicated plasmids with indicated antibodies.(B) Post-puromycin selection, transduced cells, or non-selected A549 WT cells, were treated with 40 µM MG132, 10 µM bortezomib, or DMSO (all at 0.2% DMSO) for 14 h before harvest.(C) RNA from A549 cells transduced with and selected for the indicated plasmids was reverse transcribed and screened for mRNA levels by qRT-PCR for the indicated genes, normalized to GAPDH. Error bars (SD) are shown for n = 3 (except K-RAS4A VHL-aGFP16 and VHL, n = 2). Unpaired ordinary one-way ANOVA with Dunnett's multiple comparisons test has been performed.(D) Volcano plots of proteins identified in tandem mass tag total proteome analysis in A549 cells transduced with and selected for VHL-aHRAS compared with empty vector control (EV) (left) or VHL alone (right). Horizontal line shows significance level of p = 0.05. Vertical lines show 2-fold change. Positions of KRAS, HRAS, NRAS, and VHL are indicated.

Fig 3: SNARE interactions with KRAS are selectively SNORD50A/B-dependent. a Proximity ligation analysis (PLA) between KRAS and either SNAP23, SNAP29, or VAMP3, as well as single antibody controls in H23. Significance of WT cells determined by comparison to single antibody controls while significance of KO cells determined by comparison to WT. b Schematic of KRAS mutants used in proximity protein labeling experiments. c Strength of interaction between Raf1, SNAP23, SNAP29, or VAMP3 with KRAS mutants measured by proximity protein labeling followed by pulldown western blotting (n = 2) in H23. d Quantitation of c. e Interaction strength of Raf1 and SNAREs with KRASWT and KRASG13D in H23. f Quantification of e (n = 2). g Localization of HA-tagged KRAS and endogenous SNAP23, SNAP29, and VAMP3 visualized by super-resolution microscopy in A549. Scale bars are 5 µm. h Co-immunoprecipitation of endogenous RAS and SNARE proteins from A549 lysate. i Far western blot with spotted purified recombinant protein indicated by row name. Bound recombinant KRAS was detected with an anti-RAS antibody. Coomassie stain for total protein loading (middle) (n = 3, representative images shown). j Quantification of i. Error bars are s.e.m

Fig 4: Downstream KRAS signaling relies on SNARE proteins. a Representative PLA images of Ras-GTP interaction with EGFR, P110a, and Raf1 in A549. F-actin labeled in green. b Quantification of PLA signal in a. c Proximity of KRAS with membrane bound receptors and signaling effector proteins measured by PLA in H23 cells. d ERK1/2 and AKT activation in H23 cells; quantitated replicate western blots of phosphorylated pERK and pAKT over total levels of these kinases (tERK, tAkt) are shown. e GSEA plot of KRAS-dependent transcripts against RNA-sequencing datasets of SNARE wild-type and KO cells in H23. f Gene sets lost upon SNARE TKO in H23, A549, and CHL1 from RNA sequencing. Error bars are s.e.m

Fig 5: Anti-tumor effects of free Fasudil or Lip-Fasudil in hepatocellular carcinoma (HCC) mouse model induced by AKT and N-Ras oncogenes.Liver gross image (A), liver weight (B), body weight (C) and survival curve (D) of mice that treated with saline, free Fasudil (50 mg/kg) or Lip-Fasudil (50 mg/kg). (E) Protein expression of p-AKT (S473), t-AKT, p-ERK and t-ERK in WT, AKT/Ras saline, AKT/Ras free Fasudil and AKT/Ras Lip-Fasudil liver tissues, GAPDH was used as loading control. (F) Immunohistochemistry analysis of H&E, HA, p-AKT, N-Ras and Ki67 in different groups. Data are presented as average ± standard error (n = 3), and the statistical significance level is *p < 0.05, **p < 0.01 and ***p < 0.001.