Identification of the prognostic value of elevated ANGPTL4 expression in gallbladder cancer‐associated fibroblasts

FIGURE 1. The expression of ANGPTL4 is upregulated in GCAFs in vitro. (A) Gene expression profiles of GCAFs and NFs were identified by Affymetrix GeneChip Human 1.0ST array. Volcano maps depict the genes that were significantly affected in GCAFs and NFs. Downregulated genes are represented in blue, and upregulated genes in red. (B) Gene ontology (GO) enrichment analysis classified genes in selected genesets: the biological processes involved. The top 20 with the most significant differential gene function and the number of genes they contain are listed here. (C) Based on the inclusion criteria, 466 upregulated co‐expression genes (FC > 1.5) were identified between GCAFs and NFs, and the ANGPTL4, as one of the 16 genes related to tumor angiogenesis was significantly upregulated in GCAFs (FC = 4.41)

FIGURE 2. ANGPTL4 RNA expression was detected by RT‐qPCR. Compared with the adjacent gallbladder NFs, the mRNA of ANGPTL4 was significantly upregulated in all GCAFs groups. *** p < 0.001 and **** p < 0.0001

FIGURE 3. ANGPTL4 expression in the co‐culture of GBC‐SD with GCAFs or NFs in vitro. (A) Immunocytochemical staining of ANGPTL4 expression in a co‐culture system of GBC‐SD + GCAFs/NFs in vitro. (B) Western blotting was used to detect the ANGPTL4 expression in a co‐culture system of GBC‐SD + GCAFs/NFs in vitro. ** p < 0.01

FIGURE 4. Upregulated ANGPTL4 expression in the xenografts of nude mice obtained from the injected cells co‐cultured of GBC‐SD with GCAFs may promote xenograft growth in vivo. (A) Immunohistochemical staining of ANGPTL4 expression in two groups of the xenografts of nude mice in vivo. (B) Western blotting was used to detect the protein expression of ANGPTL4 in two groups of the xenografts of nude mice in vivo. (C) Tumor xenografts of nude mice obtained from the injected cells co‐cultured of GBC‐SD with CAFs/NFs and the growth curves of the tumor xenografts of each group. The tumor xenograft volume in group GBC‐SD + GCAFs was larger compared with the GBC‐SD + NFs group. * p < 0.05 and ** p < 0.01

FIGURE 5. ANGPTL4 expression is upregulated in the stroma of GBCs. (A) ANGPTL4 expression was analyzed by immunohistochemistry staining, E, epithelium; S, stroma. Magnified insets show representative ANGPTL4 staining in stroma. The ANGPTL4 expression in the stroma of GBC (cytoplasm and/or nuclear brown staining) was significantly higher than that in GBPLs (light brown or negative staining) and GBBLs (light brown or negative staining). (B) Immunohistochemistry staining for ANGPTL4 expression in the stroma was scored using SI. ANGPTL4 expression in stroma of GBCs was significantly higher than that in GBPLs stroma (5.918 ± 0.412 vs. 2.800 ± 0.554, **p < 0.01) or GBBLs stroma (5.918 ± 0.412 vs. 2.100 ± 0.458, ***p < 0.001). ANGPTL4 expression in GBPLs stroma was higher than GBBLs stroma, however, the difference was not significant (2.800 ± 0.554 vs. 2.100 ± 0.458, #p > 0.05)

FIGURE 6. ANGPTL4 expression is co‐localized with both α‐SMA‐ and FSP‐1‐positive stroma. Co‐immunofluorescence staining was performed on GBC patient samples to analyze the expression of ANGPTL4 (green) and α‐SMA or FSP‐1 (red). Arrows show positive staining in stromal fibroblasts

FIGURE 7. Kaplan–Meier analysis of high and low ANGPTL4 expression in stroma of GBCs. GBC patients with high ANGPTL4 expression in stroma had a shorter survival time compared with patients with low ANGPTL4 expression (log‐rank test; p < 0.05)