Fig 2: Diosmin alleviates inflammation response. The relative expression of ICAM-1 and VCAM-1 in serum was presented and normalized to the average of the sham/NS group. Sham/NS (n = 3), sham/diosmin A (n = 3), sham/diosmin B (n = 3), IVS/NS (n = 6), IVS/diosmin A (n = 7), and IVS/diosmin B (n = 7). *p <0.05 vs. sham/NS; &p < 0.05 vs. IVS/NS; #p < 0.05 vs. IVS/diosmin A (ordinary one-way ANOVA's multiple-comparisons test). IVS, iliac vein compression; ICAM-1, intracellular adhesion molecule-1; VCAM, vascular cell adhesion molecule-1.

Fig 3: Model: FGFR blockade reverses T cell exclusion and ICT resistance by modulating CAFs. Blocking FGFR by FGFR inhibitor (FGFRi) suppresses the activation of MAPK/ERK signaling pathway in CAFs, thereby inhibiting the proliferation, migration and secretion of VCAM-1 of CAFs, leading to the breakage of physical and chemical barriers built by CAFs to prevent T cell infiltration. Notably, FGFRi improves ICT efficacy by increasing the infiltration of anti-tumor immune cells such as CD8+ T cells and M1 macrophages, inhibiting the infiltration of pro-tumor immune cells such as MDSC and M2 macrophages, and enhancing anti-tumor activity of CTLs in tumors.

Fig 4: MAPK/ERK activation maintains FGFR function in CAFs in vitro. A) GSEA (Reactome pathway analysis) of FGFR score in breast cancer from GEO. B) The effect of FGFRi Erdafitinib on p-ERK1/2 and total ERK1/2 expression of human CAFs and mouse CAFs was detected by western blot. C) The effect of MAPK pathway inhibitor U0126 on cell proliferation of human CAFs and mouse CAFs for 48 h was detected by CCK-8 assay (n=3 biological replicates, one-way ANOVA). D) The effect of U0126 on cell migration of human CAFs and mouse CAFs was detected by transwell migration assay (n=3 biological replicates, one-way ANOVA). E) The effect of U0126 on VCAM-1 expression of human CAFs and mouse CAFs was detected by western blot. F) The effect of U0126 on VCAM-1 level in cell supernatant of human CAFs and mouse CAFs was detected by ELISA (n=3 biological replicates, one-way ANOVA).

Fig 5: Blocking FGFR pathway inhibited cell proliferation, migration and VCAM-1 secretion of CAFs. A) The effect of FGFRi Erdafitinib on cell proliferation of human CAFs and mouse CAFs for 48 h was detected by CCK-8 assay (n=3 biological replicates, one-way ANOVA). B) The effect of FGFRi Erdafitinib on cell migration of human CAFs and mouse CAFs was detected by transwell migration assay (n=3 biological replicates, one-way ANOVA). C) Cytokine arrays for vehicle-treated versus FGFRi-treated mouse CAFs. Boxes indicate the cytokines with significant changes. D) The effect of FGFRi Erdafitinib on VCAM-1 level in cell supernatant of human CAFs and mouse CAFs was detected by ELISA (n=3 biological replicates, one-way ANOVA). E) The effect of different concentrations of FGFRi Erdafitinib on VCAM-1 expression in human CAFs and mouse CAFs was examined by western blot. F) The effect of different durations of FGFRi Erdafitinib on VCAM-1 expression in human CAFs and mouse CAFs was examined by western blot. G) Representative IF staining of a-SMA, VCAM-1 and CD3 in 4T1 tumors from vehicle- and FGFRi-treated mice. H) The effect of recombinant VCAM-1 (10 µM) or/and Erdafitinib (1 µM) on CD4+ and CD8+ T cell migration in presence of CAFs was detected by transwell migration assay (n=3 biological replicates, one-way ANOVA). I) 4T1 tumor growth and CD8+ T cell infiltration in tumors of BALB/c mice treated with vehicle or anti-VCAM1 antibody (n=5 mice/group, two-way ANOVA). J) 4T1 tumor growth and CD8+ T cell infiltration in tumors of BALB/c mice. 4T1 cells were co-transplanted with 3T3 shNC control cells or 3T3 shVCAM1 cells (n=5 mice/group, two-way ANOVA).