Fig 2: FLRT2 silencing affects genes related to oxidative stress and intercellular adhesion.(A) RT-PCR analysis of HUVECs treated with siRNA (n = 3 each). (B) Principal component analysis of RNA-Seq data from HUVECs treated with si-Cont or si-FLRT2. (C) Volcano plot of genes differentially expressed by HUVECs treated with si-Cont or si-FLRT2. Genes significantly upregulated or downregulated (q < 0.01, logFC > 5 or < –2) in si-FLRT2–treated HUVECs are marked in red or green, respectively. (D) Correlation matrix showing the Pearson correlation coefficients for each RNA-Seq experiment. (E and F) GSEA plots and heat maps of up- or downregulated genes in si-FLRT2–treated HUVECs. (G) Ingenuity Pathway Analysis (IPA) of differentially expressed genes between HUVECs treated with si-Cont or si-FLRT2. Terms underscored by red lines are related to the phenotype in Flrt2iΔEC mice. Data are presented as the mean ± SD. Comparisons between mean values of 2 groups were evaluated using a 2-tailed Student’s t test.

Fig 3: Endothelial Flrt2 sustains abnormalized tumor vessels in mice.(A) Protocol for 4OHT injections and tumor inoculations. (B) Representative IHC of B16 tumors on day 10 (n = 3) and representative PCR results (n = 3) showing expression of Flrt2 in isolated CD31+ tumor endothelial cells. (C) Tumors resected from mice 10 days after transplantation of B16 cells into the back skin (arrowheads). (D and E) Measurement of tumor weight and volume (n = 13 [control], 15 [Flrt2iΔEC]). (F–K) Immunohistochemical analysis of tumor sections (n = 7 [control], 9 [Flrt2iΔEC]). Intratumoral hemorrhage (open arrowheads) in tumors from control mice is greater than that in Flrt2iΔEC mice. (L–W) Immunohistochemical analysis of tumor sections (n = 7, 9, 7, and 5 [control], 3, 5, 4, and 7 [Flrt2iΔEC]). Tumors from Flrt2iΔEC mice show reduced endothelial proliferation (closed arrowheads) and increased pericyte coverage (open arrowheads). Scale bars: 1 cm (C [top]); 2 mm (C [bottom], F, G, P, and Q); 200 μm (R and S); 50 μm (B, H, I, and L–O). Data are presented as the mean ± SD. Comparisons between mean values of 2 groups were evaluated using a 2-tailed Student’s t test.

Fig 4: Endothelial expression of FLRT2 in advanced human colorectal cancers.(A) Schematic diagram depicting the location of the progressive and superficial areas in human colorectal cancer. (B–G) Immunohistochemical analysis of FLRT2 and CD34 expression in serial sections of paraffin-embedded samples cut from resected tumors from stage IV cases. Endothelial cells in the progressive area (closed arrowheads), but not those in the superficial area (open arrowheads), express FLRT2. (H–K) Immunohistochemical analysis of FLRT2 expression in the progressive area of tumors from individuals with various stages of colorectal cancer. Endothelial cells in tumors at advanced stages show FLRT2 expression (arrowheads). (L) Quantification of FLRT2 expression, as measured by the IHC score (n = 46 each). (M) Quantification of FLRT2 expression in the progressive area of tumors at various stages (n = 13 [stage I], 23 [stage II/III], 10 [stage IV]). The red line represents regression. (N) Kaplan-Meier curve showing recurrence-free survival of patients with stage II or III colorectal cancer, stratified according to FLRT2 expression (high- and low-scoring tumors). The log-rank test was used to compare differences between groups (n = 66). (O) Relative expression of FLRT2 and FLRT3 under various culture conditions (n = 4 each). (P–U) Immunocytochemistry analysis of HUVECs. Localization of FLRT2 proteins at the intracellular junctions (arrowheads) increased markedly after oxidative stress induced by KU55933. Representative images for 3 independent experiments are shown. Scale bars: 500 μm (B and C); 50 μm (D–K and P–U). Data are presented as the mean ± SD. Comparisons between mean values of 2 groups were evaluated using a 2-tailed Student’s t test. Comparisons among multiple groups (M and O) were evaluated using 2-way ANOVA followed by Bonferroni’s multiple-comparison test. Kaplan-Meier curves and the log-rank test were used to compare survival among the groups.

Fig 5: Deletion of Flrt2 prevents tumor metastases and spontaneous tumor formation.(A) Protocol for 4OHT injections and tumor inoculation. (B) Measurement of tumor volume in mice 28 days after transplantation of AXT cells into the back skin (arrowheads) (n = 12 [control], 7 [Flrt2iΔEC]). (C) Kaplan-Meier survival analysis of mice inoculated subcutaneously with AXT cells (n = 13 [control], 17 [Flrt2iΔEC]). (D–L) Immunohistochemical analysis of primary tumors, lungs, or livers from mice 28 days after transplantation of AXT cells into the back skin (n = 4 [control primary], 4 [control lung], 12 [control liver], 7 [Flrt2iΔEC primary], 12 [Flrt2iΔEC lung], 7 [Flrt2iΔEC liver]). Intravasation of tumor cells (open arrowheads) in primary tumors and their metastasis to distant organs (closed arrowheads) are less pronounced in Flrt2iΔEC mice. (M) Protocol for tamoxifen injection into APCMin/+ model mice. (N–T) Representative images showing the macroscopic appearance of the colon, along with the sectioned colon and small intestine specimens stained with H&E (n = 6 [control], 5 [Flrt2iΔEC]). APCMin/+Flrt2iΔEC mice have fewer and smaller tumors (arrowheads) in both the colon and small intestine than APCMin/+ mice. Scale bars: 1 cm (B); 2 mm (F–I, N, and O); 50 μm (D and E). Data are presented as the mean ± SD. Comparisons between mean values of 2 groups were evaluated using a 2-tailed Student’s t test. Kaplan-Meier curves and the log-rank test were used to compare survival among the groups.