Fig 2: EPHB1‐expressing tumor cells enhance contact‐dependent activation of platelets. (A) RNA‐sequencing data of liver metastatic tumors from the GSE71729 database were divided into two groups on the basis of platelet activation gene expression. GSEA plot based on the gene expression profiles of platelet activation genes. NES, normalized enrichment score. The FDR was set at 0.25. (B) Heatmap showing the expression of core axon guidance molecules in high‐ and low‐platelet activation score samples. (C) EPHB1 mRNA expression in the high and low platelet activation groups of liver metastatic tumors in GSE71729. Data were presented as the means ± SD). (D) Representative images of H&E staining and immunofluorescence staining for EPHB1 (red), CD42b (green) and DAPI (blue) in serial sections of liver metastatic tumors from PDAC patients. Patient 1 had high EPHB1 expression, while Patient 2 had low EPHB1 expression. (E) Identification of tdTomato‐positive platelets from Rosa26 mT/mG mice via flow cytometry analysis of CD41 expression. (F‐G) OV‐Ephb1 and OV‐Vector in both KPC1199 cells and Panc02 cells. Representative flow cytometry analysis of OV‐Vector and OV‐Ephb1 tumor cells aggregated with tdTomato‐positive platelets in a tumor cell‐platelet adhesion assay in suspensions of KPC1199 cells (F) and Panc02 cells (G). The right graphs show the percentage of tumor cells that aggregated with tdTomato‐positive platelets. Data were presented as the means ± SD with n = 3 per group. (H) Treatment scheme depicting the static tumor cell‐platelet adhesion assay. (I) Representative confocal images of tdTomato‐positive platelets adhering to GFP+ tumor cells. The right graph shows the quantification of the fluorescence signal corresponding to the number of platelets adhering to tumor cells. Data were presented as the means ± SD, with n = 5 per group. (J) PF4, 5‐HT and HexA concentrations in the supernatants of the indicated groups. (K) CCK8 analysis showing the viability of OV‐Vector and OV‐Ephb1 tumor cells cocultured with platelets at different ratios (platelets: tumor cells) for 72 h. Data were presented as the means ± SD), with n = 5 per group in (J) and (K). Statistical analyses were performed using the Wilcoxon test in (C), the two‐sided Student's t test in (D, F, G, I, J, K), and one‐way ANOVA followed by Tukey's multiple comparisons test in (G). *P<0.05, **P<0.01, ***P<0.001. Abbreviations: FDR, false discovery rate; OV‐Ephb1, lentivirus‐mediated Ephb1 overexpression; OV‐Vector, control; DAPI, 4ʹ,6‐diamidino‐2‐phenylindole; GSEA, Gene set enrichment analysis; H&E, Hematoxylin and eosin; NES, Normalized enrichment score; One‐way ANOVA, One‐way analysis of variance; SD, Standard deviations.

Fig 3: EPHB1 promotes PDAC outgrowth in the liver metastatic niche via platelet activation. (A) Mice were intrasplenically implanted with 2 × 106 OV‐Vector and OV‐Ephb1 luciferase‐expressing KPC1199 cells. In vivo imaging analysis of tumor progression on Days 7 and 14. The right graph shows the quantification of the total flux luminescence of the mice measured by IVIS at different time points. Data were presented as the means ± SD, with n = 5 mice per group. (B) Survival curve analysis of liver metastatic mice intrasplenically injected with OV‐Ephb1 or OV‐Vector KPC1199 cells (OV‐Vector group: n = 10; OV‐Ephb1: n = 8). (C) GO analysis of 1980 genes significantly upregulated in the OV‐Ephb1 group compared with that in the OV‐Vector group. (D) Representative flow cytometry analysis of CD62P expression in CD41+mCherry+ platelets after intrasplenic or intrapancreatic injection of Ephb1‐overexpressing or control cells in mice. The graphs show the percentage of CD62P‐expressing cells in CD41+mCherry+ platelets. Data were presented as the means ± SD, with n = 3 per group. (E‐F) Anti‐CD41 antibody or control IgG was applied to C57BL/6J mice harboring Ephb1‐overexpressing or control vector KPC1199 or Panc02 cells via intrasplenic injection. Representative bioluminescence image of the metastatic liver obtained via in vivo imaging analysis after KPC1199 cell injection (E, left). The quantification of the total flux luminescence of the liver is shown on the right (E, right). Data were presented as the means ± SD, with n = 5 mice per group. Representative images showing the gross morphology of the metastatic liver in mice 4 weeks after Panc02 cell injection (F). The yellow arrow indicates metastatic tumors. Statistical analyses were performed using two‐sided Student's t test in (A, D, E). The Kaplan‐Meier analysis (log‐rank test) was used to analyze the survival time of the mice in (B). *P<0.05, **P<0.01, ***P<0.001. Abbreviations: GO, Gene Ontology; IVIS, In vivo imaging system; SD, Standard deviations.

Fig 4: Platelet‐derived 5‐HT promotes metastatic PDAC outgrowth in the liver metastatic niche. (A) Relative levels of molecules released from thrombin‐ and EphB1‐Fc‐treated platelets or control platelets treated with vehicle or Fc. Data are presented as fold changes relative to Fc control platelets. (B) 5‐HT concentration in liver metastatic tumors from mice 7 and 14 days after intrasplenic injection of OV‐Vector or OV‐Ephb1 tumor cells. Data were presented as the means ± SD, with n = 3 mice per group. (C) CCK8 analysis showing the viability of OV‐Vector and OV‐Ephb1 tumor cells after coculture with platelets isolated from WT or Tph1 −/‐ mice for 72 h. Data were presented as the means ± SD, with n = 5 per group. (D) WT and Tph1 −/− mice were intrasplenically implanted with OV‐Vector or OV‐Ephb1 luciferase‐expressing KPC1199 cells and treated with an anti‐CD41 antibody or IgG. A representative bioluminescence image of the tumor obtained via in vivo imaging analysis is shown. The quantification of the total flux luminescence is shown in the lower panel. Data were presented as the means ± SD, with n = 5 per group. (E) Kaplan‐Meier analysis (log‐rank test) was used for survival curve analysis of WT, Tph1 −/− and Tph1 −/− + anti‐CD41 antibody‐treated mice injected with OV‐Vector or ov‐Ephb1 cells. Statistical analyses were performed using the two‐sided Student's t test in (B, D) and one‐way ANOVA followed by Tukey's multiple comparisons test in (C). *P<0.05, **P<0.01, ***P<0.001. Abbreviations: Tph1 −/−, Tph1 knockout mice; EphB1‐Fc, Recombinant EphB1 Fc chimera protein; CCK‐8 assay, Cell counting kit‐8 assay; One‐way ANOVA, One‐way analysis of variance; SD, Standard deviations.

Fig 5: Platelets promote liver metastatic PDAC cell growth via direct physical interaction with tumor cells. (A) Treatment scheme depicting the application of an anti‐CD41 antibody or control IgG in C57BL/6J mice intrasplenically implanted with 2×106 luciferase‐expressing KPC1199 cells. (B) In vivo imaging analysis of tumor progression on Days 7 and 14 (left). The right graph shows the quantification of the total flux luminescence of the mice measured by IVIS at different time points. Data were presented as the means ± SD, with n = 5 mice per group. (C) Representative images showing the gross morphology of the metastatic liver in mice 14 days after tumor cell injection. (D) Representative images of H&E‐stained livers of mice 14 days after tumor cell injection. The right graphs show the number of liver metastatic nodules, the percentage of liver metastatic nodules with diameters above or below 500 µm, and the percentage of liver area occupied by metastases in H&E‐stained sections of the liver. Data were presented as the means ± SD, with n = 5 mice per group. (E) Representative images showing immunochemical staining of Ki67 in liver metastatic tumors from anti‐CD41 antibody‐ or control IgG‐treated mice (left). The right graph shows the percentage of Ki67+ cells in the tumor lesions. Data were presented as the means ± SD, with n = 5 mice per group. (F) Schematic of contact coculture and noncontact coculture between platelets and tumor cells. For contact coculture, platelets and tumor cells were cultured in the same well in a 24‐well plate; for noncontact coculture, a separated coculture system in which a permeable membrane with 0.4 µm pores separated tumor cells in the lower chamber from platelets in the upper chamber was used. (G) Platelets were cultured with KPC1199 cells for 2 h at a ratio of 1,000:1 under both direct contact conditions (contact coculture) and noncontact conditions (noncontact coculture). Cell culture medium and untreated platelets served as negative controls, and thrombin‐treated platelets served as positive controls. PF4, 5‐HT and HexA concentrations in the media were determined. (H) CCK8 analysis showing the viability of KPC1199 tumor cells in contact or noncontact cocultures with platelets for 3 days in 2% FBS. Data were presented as the means ± SD, with n = 5 per group in (G) and (H). Statistical analyses were performed via two‐sided Student's t test in (B, D, E, H) and via one‐way ANOVA followed by Tukey's multiple comparisons test in (G). *P<0.05, **P<0.01, ***P<0.001. Abbreviations: IVIS, In vivo imaging system; PF4, Platelet factor‐4; 5‐HT, serotonin; HexA, Hexosaminidase A; DAPI, 4ʹ,6‐diamidino‐2‐phenylindole; FBS, Fetal bovine serum; One‐way ANOVA, One‐way analysis of variance; SD, Standard deviations.