Fig 1: LAMB3 derived from PDAC ductal cells promotes LRRC15 expression in fibroblasts by activating the ITGB1/FAK/MAPK/FOSL2 pathway. (A) Western blot analysis of LAMB3 protein level in various cell lines. (B) ELISA‐based quantification of secreted LAMB3 level from different cells (n = 3 in each group). (C) RT‐qPCR analysis of LRRC15 mRNA levels in primary CAFs treated with rhLAMB3 (n = 3 per group). (D) Western blot analysis of LRRC15 protein level in CAFs treated with varying concentrations of rhLAMB3. (E) Western blot analysis of LRRC15 level in CAFs treated with conditioned medium from LAMB3‐OE PANC‐1 and MIA PaCa‐2. (F) Flow cytometric analysis of cytotoxic activity in activated primary human CD8+ T cells directly co‐cultured with CAFs pretreated with rhLAMB3 (n = 3 per group). (G) human CD8+ T cells were directly co‐cultured with CAFs pretreated with conditioned medium from LAMB3‐OE MIA‐PaCa‐2, and the cytotoxicity was assessed by flow cytometry (n = 3 in each group). (H) Heatmap comparing TF regulon activity between Fib_LRRC15 and other fibroblast subpopulations. (I) Violin plot illustrating FOSL2 expression in fibroblasts across stages (UNIN: n = 611; IPMN: n = 1899; PDAC: n = 3086). (J) Scatter plot showing the Spearman correlation between FOSL2 expression in fibroblasts and the relative proportion of Fib_LRRC15 among total fibroblasts across samples in the scRNA‐seq dataset. Shading represents the 95% confidence interval. (K) Heatmap displaying Pearson correlations between FOSL2 expression and fibroblast phenotypes across stages. (L) Western blot analysis of FOSL2 and LRRC15 protein levels in CAFs following FOSL2 overexpression or siRNA‐mediated FOSL2 knockdown. (M) Schematic diagram depicting truncated luciferase reporter constructs for the LRRC15 promoter and engineered mutant FOSL2‐binding sites within the LRRC15 promoter. (N) Luciferase activity in HEK293T cells transfected with distinct LRRC15 promoter fragments under FOSL2 overexpression or control conditions (n = 3 in each group). (O) Luciferase activity in HEK293T cells transfected with constructs harboring wild‐type or mutant FOSL2‐binding sites within LRRC15 promoter fragments under FOSL2 overexpression vs. control conditions (n = 3 in each group). (P, Q) Bar plots showing FOSL2 enrichment at LRRC15 promoter in HEK293T cells (P) and CAFs (Q). ChIP assays were performed using anti‐IgG and anti‐FOSL2. Subsequent qPCR analysis was conducted using primers targeting distinct regulatory regions within the LRRC15 promoter, and GAPDH was used as negative control. Primer #1 was specifically designed for putative FOSL2 binding site 2, while primer #2 was designed for putative binding sites 3 and 4 (n = 3 in each group). (R) Western blot analysis of ITGB1/FAK/MAPK/FOSL2 signaling activation in CAFs treated with varying concentrations of rhLAMB3. (S) Western blot analysis of ITGB1/FAK/MAPK/FOSL2 pathway activation in CAFs treated with conditioned medium from PANC‐1 or MIA PaCa‐2 overexpressing LAMB3. (T) Western blot of LRRC15 expression in CAFs with siITGB1 or FOSL2‐OE following rhLAMB3 treatment. Statistical analysis was performed using two‐tailed Student's t test (B, C, F, G, N, O, P, and Q) and two‐tailed Wilcoxon rank‐sum test (I). For B, C, F, G, N, O, P, and Q, data are shown as mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001.
Fig 2: Distribution and clinical relevance of co‐occurrence of LAMB3+ ductal cells and LRRC15+ fibroblasts in PDAC. (A) Bar plots showing stroma_LRRC15 H‐scores in patients stratified by tumor_LAMB3 H‐score (left) (low: n = 73; high: n = 74) and tumor_LAMB3 H‐scores in patients stratified by stroma_LRRC15 H‐score (right; low: n = 73; high: n = 74) in a PDAC TMA cohort comprising 147 patients with resected PDAC. (B) Pearson correlation between tumor_LAMB3 and stroma_LRRC15 H‐scores across the TMA PDAC cohort. Shading represents 95% confidence intervals. (C) Representative IHC images from two patients with either concordantly high or low tumor_LAMB3 and stroma_LRRC15 H‐scores. Scale bars: low magnification, 200 µm; high magnification, 50 µm. (D) Representative mIHC images from the patient with concordantly high tumor_LAMB3 and stroma_LRRC15 H‐scores shown in C. Scale bars: low magnification, 100 µm; high magnification, 25 µm. (E) Kaplan–Meier survival curves stratified by tumor_LAMB3 H‐score (left) or stroma_LRRC15 H‐score (right) in TMA PDAC cohort. (F) Kaplan–Meier curve showing overall survival based on combined tumor_LAMB3 and stroma_LRRC15 H‐scores in TMA‐PDAC cohort. (G) Bar plots quantifying the fraction of LAMB3+ cells among PanCK+ epithelial cells and the density of LRRC15+PDPN+ fibroblasts per mm2 across response groups (iPR/iSD: n = 4; iPD: n = 3). (H) Spearman correlation between LAMB3+/PanCK+ epithelial cell fraction and LRRC15+PDPN+ fibroblasts per mm2 in each case. Shading represents 95% confidence intervals. (I) Representative mIHC images depicting the spatial co‐localization of LRRC15+ fibroblasts and LAMB3+ epithelial cells in the immunotherapy cohort. Scale bars, 25 µm. Statistical analysis was performed using two‐tailed Wilcoxon rank‐sum test (A), two‐tailed Student's t test (G), log‐rank test (E and F), and Tarone–Ware test (F). For A and G, data are shown as mean ± SD.
Fig 3: LAMB3 drives stromal‐immune remodeling and impairs the cytotoxic activity of CD8+ T cells in vivo. (A) Western blot analysis of LAMB3 protein expression in KPC cells with LAMB3‐OE. (B) Representative images of orthotopic tumors from mice implanted with vector control or LAMB3‐OE KPC cells and treated with vehicle or FAKi. (C) Tumor weight and tumor burden relative to body weight in the vector + vehicle (n = 6), vector + FAKi (n = 6), LAMB3‐OE + vehicle (n = 6), and LAMB3‐OE + FAKi (n = 6) groups. (D) Representative images of H&E staining and LRRC15 IHC of tumor sections from the vector + vehicle, vector + FAKi, LAMB3‐OE + vehicle, and LAMB3‐OE + FAKi groups. Scale bars: low magnification, 100 µm; high magnification, 25 µm. (E) Quantification of the LRRC15 positive area in tumor sections from the vector + vehicle (n = 6), vector + FAKi (n = 6), LAMB3‐OE + vehicle (n = 6), and LAMB3‐OE + FAKi (n = 6) groups. (F) Representative mIHC images showing PDPN and LRRC15 expression in tumor sections from the vector + vehicle, vector + FAKi, LAMB3‐OE + vehicle, and LAMB3‐OE + FAKi groups. Scale bar, 25 µm. (G) Flow cytometric analysis of GZMB MFI and the proportion of GZMB+ cells among tumor‐infiltrating CD8+ T cells in the vector + vehicle (n = 4), vector + FAKi (n = 3), LAMB3‐OE + vehicle (n = 4), and LAMB3‐OE + FAKi (n = 4) groups. (H) Western blot analysis confirming LAMB3 knockdown in KPC cells transduced with shRNAs targeting LAMB3. shLAMB3‐2 was used for subsequent in vivo experiments. (I) Representative images of orthotopic tumors from mice implanted with shCtrl or shLAMB3 KPC cells and treated with or without anti‐PD‐1 antibody. (J) Tumor weight and tumor burden relative to body weight in the shCtrl (n = 7), shCtrl + anti‐PD‐1 (n = 7), shLAMB3 (n = 7), and shLAMB3 + anti‐PD‐1 (n = 7) groups. (K) Quantification of the LRRC15 positive area in the shCtrl (n = 7), shCtrl + anti‐PD‐1 (n = 7), shLAMB3 (n = 7), and shLAMB3 + anti‐PD‐1 (n = 7) groups. (L) Flow cytometric analysis of GZMB MFI and the proportion of GZMB+ cells among tumor‐infiltrating CD8+ T cells in the shCtrl (n = 4), shCtrl + anti‐PD‐1 (n = 4), shLAMB3 (n = 4), and shLAMB3 + anti‐PD‐1 (n = 4) groups. Statistical analysis was performed using one‐way ANOVA with Tukey's post hoc test (C, E, G, J, K, and L). For C, E, G, J, K, and L, data are shown as mean ± SD. The orthotopic mouse experiment was performed once.
Fig 4: Glycolytic reprogramming in PDAC ductal cells upregulates LAMB3 expression in PDAC. (A) Heatmap of hierarchical clustering based on Jaccard similarity between transcriptional programs identified by non‐negative matrix factorization in PDAC tissue samples. (B) Violin plot depicting glycolysis scores of ductal cells across different stages (UNIN: n = 522; IPMN: n = 9100; PDAC: n = 7671). (C) Dot plot showing scaled average expression of glycolysis‐associated genes in ductal cells. Dot size represents the proportion of expressing cells in each group. Color represents expression intensity. (D) Pearson correlation between LAMB3 expression and glycolytic activity in individual ductal cells. (E) Heatmap displaying Spearman correlations between glycolytic features in ductal cells and fibroblast phenotypes. (F) Bar plots depicting fractions of SLC2A1+ cells and SLC2A1+LAMB3+ cells among PanCK+ epithelial cells across stages (UNIN: n = 20; IPMN: n = 25; PDAC: n = 25). (G) Spearman correlation between LAMB3+ and SLC2A1+ cell fractions out of PanCK+ cells (top), and between LRRC15+ fibroblast density per mm2 and SLC2A1+LAMB3+/PanCK+ cells per ROI (bottom). Shading represents 95% confidence intervals. (H) Representative mIHC images showing the presence of SLC2A1+LAMB3+PanCK+ cells during PDAC development. (I) Representative mIHC images showing the co‐enrichment of SLC2A1+LAMB3+ epithelial cells and LRRC15+ fibroblasts in a patient from the TMA PDAC cohort. (J) Bar plots showing the fractions of LAMB3+ cells out of PanCK+ cells in patients stratified by the fractions of SLC2A1+ cells out of PanCK+ cells (left) (low: n = 79; high: n = 79) and the fractions of SLC2A1+ cells out of PanCK+ cells in patients stratified by the fractions of LAMB3+ cells out of PanCK+ cells (middle) (low: n = 79; high: n = 79) in TMA PDAC cohort. Scatter plot (right) depicting Pearson correlation between SLC2A1+ epithelial cells and LAMB3+ epithelial cells across patients in the TMA PDAC cohort. Shading represents 95% confidence intervals. (K) Bar plot (left) showing the density of LRRC15+ fibroblasts per mm2 in patients stratified by the fractions of SLC2A1+LAMB3+ cells out of PanCK+ cells (low: n = 79; high: n = 79). Scatter plot (right) depicting Pearson correlation between LRRC15+ fibroblast density per mm2 and the fractions of SLC2A1+LAMB3+ cells out of PanCK+ cells in TMA PDAC cohort. Shading represents 95% confidence intervals. (L) Western blot analysis of LAMB3 expression in MIA PaCa‐2 and PANC‐1 cells under normoxic or hypoxic (3% O2) conditions with different concentrations of 2‐DG treatment. (M) Western blot analysis of LAMB3 expression in MIA PaCa‐2 and PANC‐1 cells with or without HIF‐1α knockdown. Statistical analysis was performed using a two‐tailed Wilcoxon rank‐sum test (B, F, J, and K). For F, J, and K, data are shown as mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001.
Fig 5: LAMB3+ PDAC ductal cells and LRRC15+ fibroblasts are in close proximity within PDAC tissue. (A) Bar plots comparing predicted information flow from ductal cells to fibroblasts across disease stages. (B) Dot plot revealing the intensity of LAMININ pathway‐mediated signaling from ductal cells to fibroblasts across disease stages. Dot color and size indicate communication probability and statistical significance, respectively. (C) Violin plots illustrating the expression levels of LAMB3 (UNIN: n = 522; IPMN: n = 9100; PDAC: n = 7671) in ductal cells and ITGB1 (UNIN: n = 611; IPMN: n = 1899; PDAC: n = 3086) in fibroblasts by stages. (D) Box plot showing LAMB3 expression across major cell types. (E) Representative IHC images showing LAMB3 protein across stages in the same patient. Scale bars: low magnification, 200 µm; high magnification, 50 µm. (F) Paired box plot illustrating LAMB3 H‐score in ductal cells across distinct disease stages (UNIN: n = 11; IPMN: n = 11; PDAC: n = 11). (G) Scatter plots showing the Spearman correlations between LAMB3 expression in ductal cells and LRRC15+ fibroblast features across samples in the scRNA‐seq dataset. Shading represents the 95% confidence interval. (H) Line diagram showing the spatial distance between LRRC15posfibroblast signature and LAMB3 expression in P05 sample. The x‐axis represents distance from spots with high LRRC15posfibroblast signature, and the y‐axis indicates LAMB3 expression probability. (I) Bar plots quantifying the fraction of LAMB3+ cells among PanCK+ epithelial cells and the density of LRRC15+PDPN+ fibroblasts per mm2 across stages (UNIN: n = 20; IPMN: n = 25; PDAC: n = 25). (J) Spearman correlation between LAMB3+/PanCK+ epithelial cell fraction and LRRC15+PDPN+ fibroblasts per mm2 in each ROI. Shading represents 95% confidence intervals. (K) Representative mIHC images showing spatial proximity of LRRC15+ fibroblasts to LAMB3+ epithelial cells within PDAC tissue. Scale bar, 25 µm. Statistical analysis was performed using two‐tailed Wilcoxon rank‐sum test (C and I) and paired two‐tailed Wilcoxon signed‐rank test (F). For C, F, and I, data are shown as mean ± SD.
Supplier Page from Fine Biotech Co., Ltd. for Recombinant Human LAMB3