POSTN+ CAFs facilitate gastric cancer peritoneal metastasis by promoting ICAM-1-dependent tumor cell adhesion and CD8+ T-cell exhaustion

Figure 1. Integrative machine learning analysis uncovers exhaustion-associated mechanisms and develops a prognostic model for peritoneal metastasis in GC. (A) Cell type classification of single-cell RNA-seq data was conducted according to the expression patterns of canonical marker genes. (B) The activity scores of exhaustion-associated genes were calculated and visualized for each single cell. (C) Violin plots illustrating the expression patterns of exhaustion-associated genes among different cell types. (D) Volcano plot illustrating differential gene expression between peritoneal metastatic and primary GC samples, with the top five genes labeled. (E) A total of 101 predictive models were developed using a rigorous 10-fold cross-validation framework, with the top ten predictive models shown. (F) Univariate analysis of clinical characteristics along with predictive model regarding OS. (G) Assessment of the correlation between risk scores derived from the GSVA-based exhaustion model and the activity of representative pathways. (H) GADD45B, DUSP1, TSC22D3, POSTN, EMP3 and TAP1 were analyzed by scRNA-seq. (I) Correlation analysis was performed to assess the relationship between the six model…

Figure 2. POSTN+ CAFs are significantly associated with peritoneal metastasis and poor prognosis in patients with GC. (A) Representative distribution of POSTN+ CAFs across different invasion depths of primary lesions in GCPM (GSE167297, n = 5 per group). (B, C) Frequency (B) and ratio (C) of POSTN+ CAFs and POSTN- CAFs in GCPM. (D) Pseudotime trajectory analysis illustrating the distinct developmental dynamics and transcriptional states of POSTN+ versus POSTN- CAFs in superficial and deep tissues. (E, F) Analysis of the distribution and relative proportions of immune cell populations across different invasion depths of primary lesions in GC. (G) Functional enrichment analysis illustrating the biological processes and signaling pathways associated with POSTN+ CAFs. (H) Western blot analysis of FAP, αSMA, and POSTN expression in primary CAFs isolated from GC patients with postoperative PM. Representative results from three independent experiments are shown. n = 3 per group. (I) Prognostic impact of POSTN expression and CAFs infiltration in GC patients analyzed using the TIMER 2.0 database. (J) Multi-immunostaining images of CK8 (red), POSTN (green), αSMA (red), FAP (red), CD68 (red) and CD8…

Figure 3. POSTN+ CAFs induced CD8+ T-cell exhaustion in vitro and in vivo. (A) Multi-immunostaining of FAP, αSMA, and POSTN in primary CAFs and NFs (scale bars: 100 µm). (B) Western blot analysis of POSTN in CAFs, NFs and GC cell lines. Representative results from three independent experiments are shown. (C) qRT-PCR analysis of S100A4, SMA and FAP expression level in CAFs and NFs. (D) Construction of stable POSTN up-regulated 3T3-L1 cell lines. n = 3 per group. (E) Schematic illustration of CD8+ T-cell isolation from spleens of C57BL/6 mice and the direct co-culture system with 3T3-L1 cells. CD8+ T cells were activated with anti-CD3/CD28 antibodies and directly co-cultured with 3T3-L1 cells at a 1:1 ratio (1 × 105 cells each) for 72 h prior to subsequent analyses. (F) Proliferation of CD8+ T-cells assessed by CFSE dilution after co-culture with POSTN-overexpressing 3T3-L1POSTN cells. n = 3 per group. (G) Immunofluorescence staining of Ki67 expression in CD8+ T-cells following co-culture with POSTN-overexpressing 3T3-L1 cells (scale bars: 10 µm). n = 3 per group. (H) Apoptosis analysis of CD8+ T-cells after co-culture with POSTN-overexpressing 3T3-L1 cells. n = 3 per group. (I)…

Figure 4. POSTN+ CAFs facilitate GC tumor spheroid formation through ICAM-1 dependent regulation. (A) qRT-PCR analysis of the mRNA expression levels of proliferation and adhesion markers in 3T3-L1POSTN and 3T3-L1NC cells. (B) Western blot analysis showing the protein expression levels of MMP9 and ICAM-1 in 3T3-L1 cells and NFs following treatment with increasing concentrations of POSTN protein. n = 3 per group. (C) Western blot analysis of ICAM-1 in 3T3-L1POSTN cells after treatment with SB273005. n = 3 per group. (D, E) Western blot analysis of ICAM-1 in 3T3-L1NC (D) and NFs (E) cells after treatment with SB273005 and POSTN protein. n = 3 per group. (F) Western blot analysis of ICAM-1 in 3T3-L1POSTN cells after treatment with A-205804. Representative results from three independent experiments are shown. (G) Cell adhesion assay assessing 3T3-L1 cell attachment at 50, 60, and 70 minutes after plating. A total of 1 × 105 cells were seeded into 12-well plates for each assay. n = 3 per group. (H) Adhesion assay evaluating the attachment of FITC-labeled MFC and YTN16 tumor cells to adherent 3T3-L1POSTN cells. Tumor cells and fibroblasts were co-cultured at a 2:1 ratio (2 × 105 tumor cells and 1…

Figure 5. POSTN induces CD8+ T-cell exhaustion through regulation of the ICAM-1–ETV3 signaling pathway. (A) Western blot analysis showing the AKT-NF-kB pathway in 3T3-L1POSTN and 3T3-L1NC cells. n = 3 per group. (B) Western blot analysis showing the expression levels of AKT-NF-κB pathway in 3T3-L1 cells and NFs following treatment with increasing concentrations of POSTN protein. n = 3 per group. (C, D) Western blot analysis showing the expression of ICAM-1 in 3T3-L1POSTN cells following treatment with AKT (MK-2206) and NF-κB (JSH-23) inhibitors. Representative results from three independent experiments are shown. (E) Fibroblast–tumor cell adhesion assay assessing the effects of AKT (MK-2206) and NF-κB (JSH-23) inhibitors on cell adhesion. FITC-labeled MFC and YTN16 tumor cells were co-cultured with PKH26-labeled 3T3-L1POSTN cells at a 2:1 ratio (2 × 105 tumor cells and 1 × 105 fibroblasts per well) in 12-well plates for 2 h (scale bars: 100 µm). (F) qRT-PCR analysis showing the mRNA expression levels of exhaustion-associated genes in CD8+ T cells following co-culture with 3T3-L1POSTN cells treated with A286982 (LFA-1/ICAM-1 inhibitor). n = 3 per group. (G) Representative flow cytometry…

Figure 6. Combination therapy with anti-PD-1, anti-CTLA-4, and Cilengitide inhibits peritoneal metastasis of GC driven by POSTN+ CAFs. (A) A GCPM model was established in C57BL/6 mice by intraperitoneal injection of YTN16 cells and 3T3-L1POSTN cells. Starting on day 5 post-implantation, mice were administered anti-PD-1 (10 mg/kg) and anti-CTLA-4 (10 mg/kg), with or without Cilengitide (5 mg/kg), every other day (n = 5 per group). (B) Representative images showing intraperitoneal implantation of tumors in the GCPM mouse model. (C) Statistical analysis of the number of peritoneal metastatic nodules in each group. (D) Representative flow cytometry plots showing PD-1 and CTLA-4 expression in CD8+ T-cells isolated from the peritoneal lavage fluid of mice. n = 5 per group. (E) Quantification of exhaustion marker expression in CD8+ T-cells. n = 5 per group. T-cell exhaustion subsets were gated and defined according to previous studies: Ly108+ CD69- (Tex prog) and Ly108- CD69- (Tex int); Ly108+ CD62L+ (Tex prog), and Ly108- CD62L- (Tex int). (F) IHC analysis of POSTN, ICAM-1, and TOX expression in peritoneal tumor nodules. (G) Schematic diagram illustrating the molecular mechanism by which POSTN+…