Fig 2: TPC‐EV‐induced EPC vasculogenic properties are regulated by Gas6 in vitro. (a) LC‐MS/MS (nanoLC‐QE) proteomic analysis for TPC‐EVs. (b) The protein level of Gas6 in tumor perivascular cell lysates and TPC‐EVs were determined by Western blotting assay. No., number order. (c) Gas6 mRNA and protein levels in tumor perivascular cells after transfection with Gas6 siRNA were analyzed by RT‐PCR and Western blotting assays. Quantification of Gas6 mRNA expression in tumor perivascular cells after transfection with Gas6 siRNAs (n = 3). *** P < 0.001 versus the NC siRNA‐transfected group by one‐way ANOVA and Tukey's multiple comparison test. (d) EVs were harvested from tumor perivascular cells transfected with Gas6 siRNA and EV‐Gas6 protein levels were determined by Western blotting analysis. CD63 served as a loading control. Quantification of Gas6 expression in TPC‐EVs (n = 3). *** P < 0.001 versus the NC siRNA‐transfected groups by unpaired two‐tailed t‐test. (e) Flow cytometry analysis of Axl expression in EPCs. (f) The phosphorylated and total forms of Axl, Akt and Erk1/2 in EPCs after the indicated treatments were determined. Representative blots and quantification of the ratio of p‐Axl/Axl (n = 3). TPC‐EV‐(siNC), EVs derived from TPCs transfected with NC siRNA. TPC‐EV‐(siGas6), EVs derived from TPCs transfected with Gas6 siRNA. (g) Representative images and quantification of the number of migrated EPCs after the indicated treatments (n = 5). (h) Representative images and quantification of the number of EPC adhered to FN. FN, fibronectin. (i) Quantification of the number of EPC adhesion to HUVECs (n = 5). ECs, HUVECs. (j) Quantitative assessment of EPC viability after the indicated treatments (n = 3). (k) EPCs were treated with TPC‐EV‐(siNC), TPC‐EV‐(siGas6), and Gas6 (200 ng/ml) for 48 h, and then EPC proliferation was examined. Quantification of EdU incorporation in EPCs (n = 5). Data are presented as the mean ± SEM. Scar bar, 200 μm. ** P < 0.01, and *** P < 0.001 versus the Vehicle groups; # P < 0.05, ## P < 0.01, and ### P < 0.001 versus the TPC‐EV‐(siNC)‐treated groups

Fig 3: Regorafenib plus R428 therapy prolongs survival and decreases the tumor burden in mice with colorectal cancer liver metastases. (a) Therapeutic schedule for continuous treatment with vehicle, regorafenib, R428, and regorafenib + R428 in mice bearing HCT116 liver metastases. (b) Kaplan‐Meier estimates of overall survival for mice with HCT116 hepatic tumors after the indicated treatment (n = 7). OS, overall survival. MS, median survival. Reg., regorafenib. NS, no significance. * P < 0.05 and ** P < 0.01 versus indicated groups by Log‐rank (Mantel‐Cox) test. (c) Representative ultrasound images of HCT116 hepatic tumors after the indicated treatments. Mice bearing HCT116 hepatic tumors were treated as indicated and imaged at days 15 and 28 after tumor implantation. Scale bar, 5 mm. (d) Representative images of CD31+ (red) endothelial cell immunostaining and DAPI (blue) for nuclear staining. Scale bar, 50 μm. Quantification of microvessel density (CD31 density) in HCT116 hepatic tumors (n = 8). Data are presented as mean ± SEM. *** P < 0.001. (e) Diagram depicting the mechanism by which Gas6+ TPC‐EVs activates the Axl pathway and consequently promotes EPC recruitment for rapid tumor revascularization after withdrawal of AA‐TKI therapy

Fig 4: TPC‐EV‐mediated EPC recruitment and tumor revascularization are regulated by Gas6 in vivo. (a) Quantification of the proportional changes of recruited EPCs in HT‐29 tumors (n = 6). (b) Representative images of CD31+ (red) endothelial cell, α‐SMA+ (green) perivascular cell immunostaining, and DAPI (blue) for nuclear staining. (c) Quantification of microvessel density (CD31 density) and perivascular cell number (α‐SMA density) in HT‐29 tumors (n = 6). NS, no significance. (d) Representative images of cells with p‐Histone H3 positive staining (green) and DAPI (blue) for nuclear staining. (e) Quantification of cell proliferation (p‐Histone H3 index) in HT‐29 tumors (n = 6). Data are presented as the mean ± SEM. Scale bars, 50 μm. * P < 0.05, ** P < 0.01, and *** P < 0.001

Fig 5: Activity of INCB081776 in sarcoma PDX models. Athymic nude mice bearing (A) CTG-2041 (B) CTG-1302, or (C) CTG-1339 tumors were dosed orally twice a day (BID) with INCB081776 at 30 mg/kg. N = 10 mice per group in panels (A–C). Statistical analysis was performed using two-way analysis of variance with Sidak’s multiple comparisons test (D) CTG-2041 or CTG-1339 tumors from mice treated with INCB081776 at 30 mg/kg twice a day (BID) or vehicle were lysed and processed by Western blot for pAXL, AXL, pMERTK, MERTK, GAS6, pAKT, AKT, and ß-Actin. Three mice per group were analyzed. Western blot data were obtained from different gels. *p < 0.05; **p < 0.0001; compared with control. SEM, standard error of the mean; TGI, tumor growth inhibition.