Proteomic profiling identifies a stromal TGF-β1/podoplanin axis as a driver of colorectal cancer progression

Abstract

Background: The tumor microenvironment (TME) plays a pivotal role in the development and progression of colorectal cancer (CRC), yet the complex crosstalk among its components remains incompletely understood. Cancer-associated fibroblasts (CAFs) and tumor-associated macrophages (TAMs) have emerged as key regulators of CRC progression, but their specific contributions, particularly given their heterogeneity, are not fully elucidated. This study identifies podoplanin (PDPN), a transmembrane glycoprotein enriched in CAFs, as highly expressed in the CRC TME, in particular surrounding the tumor, and associated with macrophage infiltration and cancer progression.

Methods: We performed mass spectrometry-based proteomic analysis on matched CRC and adjacent normal tissues from patients to identify altered signaling pathways and protein expression. The clinical relevance of PDPN expression was evaluated in CRC samples from two independent cohorts using immunohistochemistry and immunofluorescence analysis. Publicly available data from the Gene Expression Omnibus (GEO) database were analyzed to assess the association between PDPN expression and patient survival. Functional assays using direct and indirect co-culture systems investigated the influence of macrophage infiltration on stromal PDPN expression and its effect on colon adenocarcinoma cell growth.

Results: PDPN expression was significantly elevated in the stroma of the colorectal tumor tissues compared to normal tissues and correlated with M2-like macrophage infiltration. High PDPN expression was associated with reduced relapse-free survival in CRC patients. Stromal cells pre-conditioned with M2-like macrophages upregulated PDPN and more effectively supported the growth of three colon adenocarcinoma cell lines. PDPN depletion impaired the ability of stromal cells to promote tumor cell proliferation. Mechanistically, M2-like macrophage pre-conditioning induced a TGF-β1-dependent increase in YAP/TAZ nuclear localization, RhoA/ROCK/myosin-driven cytoskeletal contractility, and extracellular matrix (ECM) production in stromal cells. Inhibition of TGF-β1 signaling or ROCK activity reduced stromal support for cancer cell growth.

Conclusion: This study reveals a novel mechanism by which the TME facilitates CRC progression and highlights PDPN as a potential prognostic biomarker and therapeutic target in CRC.

Keywords: Cancer associated fibroblasts; Colorectal cancer; Proteomic; Transforming growth factor beta; Tumor associated macrophages; Tumor microenvironment; YAP/TAZ.

Fig. 1

Pathways associated with significantly upregulated proteins in CRC compared to matched healthy tissue. (A) Label-free LC–ESI–MS/MS experimental design and main results. Image created with BioRender. (B) Volcano plot displaying the log 2 fold change (x axis) against the − log 10 statistical q value (y axis) for all proteins differentially expressed between NDT and CRC tissues. Proteins with significantly decreased levels in CRC (q < 0.05) are shown in blue, while the proteins with significantly increased levels in CRC are noted in red. (C) Enrichment pathway analysis of significantly up-regulated proteins in CRC versus NDT from the REACTOME 2024 and MSigDB Hallmark 2020 databases. (D) Heatmap showing protein expression data of innate immune system, IL-4 and IL-13 signaling, inflammatory response, TGF-b signaling, EMT and ECM organization. Green and red colors refer to the z-score of the data by row

Fig. 2

PDPN is increased in the stroma surrounding CRC. (A) RT-qPCR analysis of PDPN mRNA expression in matched normal tissue distant to the tumor (NDT) and CRC samples. Mann Whitney test, p < 0.0001, n = 9. (B) Western blot analysis of PDPN protein expression in matched NDT and CRC samples. β-actin was used as loading control. (C) The intensity of PDPN/β-actin ratio measured by densitometry is shown. Upaired t-test, p = 0.0199, n = 6. (D) Representative immunohistochemistry staining of PDPN in matched NDT, normal tissue adjacent to the tumor (NAT) and CRC tissues. Scale bar is 50 μm. (E) PDPN immunohistochemistry staining score was compared between NDT, NAT and CRC tissues. One-way ANOVA, n = 20. (F) PDPN immunohistochemistry staining score in NAT and CRC tissues from the tissue microarray. Paired t-test, n = 40. (G) Receiver Operating Characteristic (ROC) curve representing PDPN score sensitivity and specificity for the classification of tumoral vs. normal tissue. (H) Representative immunofluorescence staining of PDPN and a-SMA in CRC and NDT tissues. Scale bar is 50 μm. (I) Quantification of PDPN and a-SMA co-localization rates in normal and CRC tissues. Paired t-test, n = 20. (J) PDPN immunohistochemistry staining score in CRC tissues of patients with different tumor size. Unpaired t-test, n = 6 T1-T2, n = 14 T3-T4. K) PDPN mRNA expression in CRC tissues of patients with different tumor size. Data have been extracted from the Colon Sidra-LUMC AC-ICAM dataset [39]. Unpaired t-test, n = 69 T1-T2, n = 279 T3-T4. L) Kaplan-Meier survival analysis showing the relapse-free survival of patients with colon cancer stratified by the expression of PDPN. (adapted from KmPlot using the data from 1336 tumor samples from 16 independent cohorts) [43]. * p < 0.05, ** p < 0.01, *** p < 0.0001, **** p < 0.00001

Fig. 3

CD68 + macrophages are increased in the tissue surrounding CRC and are associated to increased stromal PDPN. (A) Representative immunohistochemistry staining of CD68 in matched NDT, NAT and CRC tissues. Scale bar is 50 μm. (B) CD68 immunohistochemistry staining score was compared between NDT, NAT and CRC tissues. One-way ANOVA, n = 20 (C) CD68 immunohistochemistry staining score in NAT and CRC tissues from the tissue microarray. Paired t-test, n = 40. (D) Representative PDPN and CD68 immunohistochemistry staining on consecutive serial sections of CRC tissues. Scale bar is 50 μm, inlet scale bar is 25 μm. (E) Pearson correlation analysis of PDPN and CD68 staining area coverage in CRC tissues. Data have been obtained from tissues deriving from 8 patients. A minimum of 30 randomly selected ROI per patient have been analyzed. (F) Pearson correlation analysis of PDPN mRNA expression and macrophage M1 and M2 infiltration level in CRC samples. Data derive from the CRC TCGA datasets analyzed by the CIBERSORT algorithm. (G) ssGSVA from Colon Sidra-LUMC AC-ICAM dataset showing Pearson correlation analysis of PDPN and M2 macrophage signature. * p < 0.05, ** p < 0.01, *** p < 0.0001, **** p < 0.00001

Fig. 4

Alternatively activated macrophages increase PDPN expression in stromal cells. A) Set up of indirect macrophages (Mφ)/mesenchymal stromal cells (MSC) co-cultures. Image created with BioRender. B) RT-qPCR analysis of PDPN, ACTA2 (a-SMA) and CDH2 (N-Cadherin) mRNA expression in MSC cultured alone or co-cultured with classically activated M1 or alternatively activated M2 macrophages. One-way ANOVA, n=3. C) Representative immunofluorescence staining of PDPN in MSC cultured alone or co-cultured with classically activated M1 or alternatively activated M2 macrophages. Nuclei were counterstained with Hoechst. Scale bar is 50 mm. D) Quantification of PDPN staining intensity of MSC treated as in C. One-way ANOVA, n=3.  E)Western blot analysis of cancer associated fibroblast markers Vimentin, a-SMA and PDPN in MSC cultured alone or co-cultured with classically activated M1 or alternatively activated M2 macrophages. β-actin was used as loading control. F) The intensity of PDPN, vimentin or a-SMA/β-actin ratio measured by densitometry is shown. One-way ANOVA, n=3. G) RT-qPCR analysis of PDPN, ACTA2 and CDH2 mRNA expression in MSC cultured alone or in presence of 10 ng/mL TGF-β1, 40 ng/mL IFNγ or 40 ng/mL IL-6. One-way ANOVA, n=3. H) Representative immunofluorescence staining of PDPN in MSC cultured alone or in presence of 10 ng/mL TGF-β1, 40 ng/mL IFNγ or 40 ng/mL IL-6. Scale bar is 50 mm. I) Quantification of PDPN staining intensity of MSC treated as in I. One-way ANOVA, n=3. J) Western blot analysis of cancer associated markers Vimentin, a-Sma and PDPN in MSC cultured alone or in presence of 10 ng/mL TGF-β1, 40 ng/mL IFNγ or 40 ng/mL IL-6. K) The intensity of PDPN/β-actin ratio measured by densitometry is shown. One-way ANOVA, n=3. L) RT-qPCR analysis of PDPN, ACTA2 (a-Sma) and CDH2 (N-Cadherin) mRNA expression in MSC cultured alone or co-cultured with classically activated M1 or alternatively activated M2 macrophages in presence or not of 10 μM TGF-β1 receptor (ALK5) inhibitor. One-way ANOVA, n=3. * p < 0.05, ** p < 0.01, *** p < 0.0001.

Fig. 5

Macrophages and TGF-β1 induced PDPN stromal expression through YAP/TAZ. (A) Representative immunofluorescence staining of YAP/TAZ (green) in MSC cultured alone or co-cultured with classically activated M1 or alternatively activated M2 macrophages. Nuclei were counterstained with Hoechst. Scale bar is 50 μm. (B) Percentage of MSC displaying preferential nuclear YAP/TAZ localization. One-way ANOVA, n = 3. (C) RT-qPCR of YAP/TAZ targets CYR61 and CTGF in MSC cultured alone or co-cultured with classically activated M1 or alternatively activated M2 macrophages. One-way ANOVA, n = 3. (D) Representative immunofluorescence staining of YAP/TAZ (green) in MSC cultured alone or in presence of 10 ng/mL TGF-β1, 40 ng/mL IFNγ or 40 ng/mL IL-6. Nuclei were counterstained with Hoechst. Scale bar is 50 μm (E) Percentage of MSC displaying preferential nuclear YAP/TAZ localization. One-way ANOVA, n = 3. (F) RT-qPCR of YAP/TAZ targets CYR61 and CTGF in MSC cultured alone or in presence of 10 ng/mL TGF-β1, 40 ng/mL IFNγ or 40 ng/mL IL-6. One-way ANOVA, n = 3. (G) RT-qPCR of PDPN in MSC cultured alone or co-cultured with classically activated M1 or alternatively activated M2 macrophages in presence or not of 250 nM YAP/TAZ inhibitor verteporfin (VP). One-way ANOVA, n = 3. (H) Western blot analysis of PDPN protein expression in in MSC cultured alone or co-cultured with classically activated M1 or alternatively activated M2 macrophages in presence or not of 250 nM YAP/TAZ inhibitor verteporfin (VP). (I) RT-qPCR of PDPN in MSC cultured or not in presence of 10 ng/mL TGF-β1. Where indicated, MSC were pre-treated with 250 nM YAP/TAZ inhibitor verteporfin (VP). One-way ANOVA, n = 3. (J) Western blot analysis of PDPN protein expression in MSC cultured or not in presence of 10 ng/mL TGF-β1. Where indicated, MSC were pre-treated with 250 nM YAP/TAZ inhibitor verteporfin (VP). (K) Schematic representation of the PDPN gene promoter evaluated in the ChIP analysis. Light blue boxes represent TEAD-binding sites. +1 position indicates transcription start site. (L) The cross-linked chromatin purified by CRC and matched healthy tissues from three patients was used in ChIP experiments using the antibodies indicated in the figure. The TEAD binding sequences A and B (TEAD seq A and B) were analyzed by quantitative real time PCR. Normalization was performed to the amount of input chromatin. The ChIP samples were further tested by qPCR on a region that was negative for transcriptional factor recruitment. Bars represent mean ± SD from technical replicates. Unpaired t-test. * p < 0.05, ** p < 0.01, *** p < 0.0001

Fig. 6

Stromal PDPN sustains colon adenocarcinoma cells growth. (A) Set up of indirect macrophages (MΦ)/mesenchymal stromal cells (MSC) co-cultures followed by direct MSC/colon adenocarcinoma (AC) cell co-culture. Image created with BioRender. MSC feeders were pre-conditioned or not with classically activated M1 or alternatively activated M2 macrophages. Where indicated MSC were pretreated with 10 µM TGF-β1 receptor ALK5 inhibitor or siRNAs. Colon AC cells were seeded as single cell suspension on MSC feeders and colony size measured after 48 h. (B) Colon AC cell colony size on MSC feeders pre-conditioned or not with classically activated M1 or alternatively activated M2 macrophages and in presence or not of 10 µM TGF-β1 receptor ALK5 inhibitor. One-way ANOVA, n = 3. (C) Representative immunofluorescence staining of actin (green) and nuclei counterstaining (blue) of MSC/Caco-2 co-cultures as in B. Scale bar is 50 μm. (D) Colon AC cell colony size on MSC pre-conditioned or not with 10 ng/mL TGF-β1, 40 ng/mL IFNγ or 40 ng/mL IL-6. One-way ANOVA, n = 3. (E) Representative immunofluorescence of actin (green) and nuclei counterstaining (blue) in MSC/Caco-2 co-cultures as in D. Scale bar is 50 μm. (F) Colon AC cell colony area after 48 h of direct co-cultures on MSC pre-conditioned with scramble (Scr) or PDPN specific siRNA and with 10 ng/mL TGF-β1. Unpaired t-test, n = 3. (G) Colon AC cell colony area after 48 h of direct co-cultures on MSC pre-conditioned with scramble (Scr) or PDPN specific siRNA and cultured alone or in indirect co-cultures with classically activated M1 or alternatively activated M2 macrophages. One-way ANOVA, n = 3. (H) Representative immunofluorescence of actin (green) in MSC/Caco-2 co-cultures as in G. Scale bar is 50 μm. * p < 0.05, ** p < 0.01, *** p < 0.0001, **** p < 0.00001

Fig. 7

Stromal PDPN expression favors cancer associated fibroblast differentiation and extracellular matrix synthesis. (A) Normalized enriched score (NES) bubbleplot of GSEA results from tissue proteomics with MSigDB REACTOME genesets. (B) Immunofluorescence analysis of active RhoA-GTP in MSC treated with scamble (Scr) and PDPN specific siRNA and cultured alone or in indirect co-cultures with classically activated M1 or alternatively activated M2 macrophages. Scale bar is 20 μm. (C) Quantification of RhoA-GTP staining intensity of MSC treated as in B. One-way ANOVA, n = 3. (D) Western blot analysis of myosin light chain (Mlc2) phosphorylation in MSC cultured alone or in indirect co-cultures with classically activated M1 or alternatively activated M2 macrophages. (E) Immunofluorescence analysis of active RhoA-GTP in MSC treated with scamble (Scr) and PDPN specific siRNA and in presence of 10 ng/mL TGF-β1. (F) Quantification of RhoA-GTP staining intensity of MSC treated as in E. Unpaired t-test, n = 3. Scale bar is 20 μm. (G) Western blot analysis of Mlc2 phosphorylation in MSC cultured alone or in presence of 10 ng/mL TGF-β1, 40 ng/mL IFNγ or 40 ng/mL IL-6. (H) Western blot analysis of Mlc2 phosphorylation in MSC cultured alone or in presence of 10 ng/mL TGF-β1. Where indicated, MSC were treated with scramble (Scr) or PDPN specific siRNA. (I) RT-qPCR analysis of cancer associated fibroblast markers CDH2, ACTA2 and PDPN in MSC treated with scramble (Scr) or PDPN specific siRNA and 10 ng/mL TGF-β1. (J) Western blot analysis of a-Sma and PDPN in MSC treated with scramble (Scr) or PDPN specific siRNA and 10 ng/mL TGF-β1. K) RT-qPCR analysis of FN1, COL1A1, COL1A2, COL6A1, COL6A2, COL6A3 in MSC treated with scramble (Scr) or PDPN specific siRNA and 10 ng/mL TGF-β1. L) Western blot analysis of fibronectin (FN), type I collagen (COL1) and type VI collagen (COL6) in MSC treated with scramble (Scr) or PDPN specific siRNA and 10 ng/mL TGF-β1. M) Colon AC cell colony area after 48 h of direct co-cultures on MSC pre-conditioned with 10 ng/mL TGF-β1 and in presence or of 10 µM ROCK inhibitor Y27632. * p < 0.05, ** p < 0.01, *** p < 0.0001

Fig. 8

Role of stromal PDPN in CRC progression. During colorectal cancer (CRC) progression, tumor-associated macrophages (MΦ2) release soluble factors, including TGF-β1, which may induce upregulation of podoplanin (PDPN) in stromal cells via YAP/TAZ pathway activation. PDPN overexpression in these stromal cells promotes their differentiation into cancer-associated fibroblasts (CAF) by enhancing Rho/ROCK/myosin-dependent cytoskeletal remodeling and extracellular matrix (ECM) production. This, in turn, contributes to the establishment of a tumor-supportive microenvironment that facilitates CRC cell proliferation. Inhibition of PDPN expression in activated CAF, or treatment with either a TGF-β receptor inhibitor or a ROCK inhibitor, reduces their capacity to support the growth of colon adenocarcinoma cells