Correction: Targeting PDPN enhances antitumor T-cell activity by disrupting β-catenin-mediated PD-L1 expression in melanoma

Abstract

Melanoma is a highly immunogenic malignancy originating from melanocytes and is distinguished by a significantly higher propensity for metastasis compared to other cutaneous cancers (1)(2). This strong immunogenicity is evidenced by dense lymphocytic infiltration within the tumor microenvironment (TME) and an increased incidence among immunocompromised individuals (3)(4), rendering melanoma particularly amenable to immunotherapeutic strategies. While immunotherapies-such as immune checkpoint blockade and adoptive cell transfer-exhibit superior therapeutic indices and fewer off-target toxicities relative to conventional surgery, radiotherapy, or chemotherapy, their clinical application in melanoma has advanced more slowly than in breast, lung, or liver cancers. This slower pace is largely attributable to the absence of robust, melanoma-specific predictive biomarkers. Consequently, the identification of reliable biomarkers to guide patient stratification, optimize treatment selection, and inform rational design of combination immunotherapeutic approaches remains an urgent and critical objective in melanoma management. Advancing the discovery and clinical utilization of such markers is essential for realizing the full promise of immune-based therapies against this aggressive disease.Podoplanin (PDPN) is a mucin-type transmembrane glycoprotein with pivotal roles in embryonic development and platelet aggregation (5). PDPN is aberrantly overexpressed across various malignancies-including hepatocellular carcinoma (6-7), glioma (8)(9), breast cancer (10), and lung squamous cell carcinoma (11)(12)-where it actively drives tumor initiation, progression, and metastatic spread. Mechanistically, PDPN associates with CLEC-2 receptors on platelets via its extracellular PLAG3 domain, thereby facilitating platelet activation and promoting tumor cell invasion (13).Within tumor cells, PDPN interacts with the ERM protein family, contributing to epithelial-mesenchymal transition (14). Beyond direct tumor cell functions, PDPN also plays a critical role in shaping the immunosuppressive tumor microenvironment (TME) (5,8,15). In melanoma, expression of PDPN by intratumoral cancer-associated fibroblasts (CAFs) is correlated with increased sentinel lymph node metastases, with PDPN-expressing CAFs enhancing local immunosuppression through cytokine secretion and immune modulation (16)(17). Tumor cell-intrinsic PDPN impairs the cytotoxic activities of CD8⁺ T cells, natural killer (NK) cells, and macrophages, potentially through IL-27-dependent processes (5,15). The PDPN-CLEC-2 axis further supports the recruitment of immunosuppressive macrophages, and pharmacologic blockade of this pathway reduces macrophage accumulation and alleviates local immune suppression (18). Additionally, previous studies, including our own, have demonstrated that the PDPN-targeting inhibitory peptide CY12-RP2 augments antitumor immunity by enhancing the infiltration and activation of cytotoxic CD8⁺ T cells, NK cells, and M1-polarized macrophages, with concurrent increases in pro-inflammatory cytokine secretion (e.g., IFN-γ, TNF-α) (19).To gain a comprehensive understanding of PDPN-driven immune evasion, we used the same tissue microarray (HMelC112CD01). This investigation is necessary to realize PDPN's potential as a therapeutic target for improving antitumor immunity, especially in melanoma.Tumor cells evade T cell-mediated immune surveillance primarily through upregulation of PD-L1 (20). Emerging evidence indicates that anti-PD-1 therapy can trigger adaptive upregulation of PD-L1, contributing to acquired resistance to immune checkpoint blockade (ICB) (21). Delineating the molecular mechanisms that control PD-L1 expression is therefore crucial for optimizing ICB efficacy and improving patient outcomes. The regulation of PD-L1 is multifactorial: extrinsic regulation is mediated by cytokines in the TME (such as TNF-α, IFN-γ, and IL-1β) (22), while intrinsic regulation involves oncogenic signaling pathways. Although several oncogenic drivers-including MYC, EGFR, and RAS-are established regulators of PD-L1 (23)(24)(25), the precise molecular mechanisms through which PDPN modulates PD-L1 remain incompletely defined.The β-catenin signaling pathway represents a critical mechanism facilitating tumor immune evasion, primarily through its regulation of PD-L1 expression. In glioblastoma, activation of β-catenin directly induces transcription of CD274 (encoding PD-L1) via β-catenin/LEF1 binding to the CD274 promoter, a process that is maintained by AKT signaling and is associated with reduced intratumoral CD8⁺ T cell accumulation (26). In addition to direct transcriptional control, β-catenin signaling broadly suppresses antitumor immune responses in the TME by impairing T cell priming by dendritic cells (DCs) and fostering immune tolerance (27)(28).Melanoma model systems have demonstrated that activation of the β-catenin pathway leads to the exclusion of CD103⁺ dendritic cells and T cells, conferring resistance to anti-PD-L1 therapy (29). Although phosphorylation of β-catenin at S552 facilitates its nuclear translocation, its specific role in regulating PD-L1 in melanoma and other tumors remains insufficiently characterized.In the present study, we establish that inhibition of PDPN using the CY12-RP2 inhibitory peptide attenuates melanoma progression in multiple murine models, concomitant with reduced functional exhaustion of tumor-infiltrating CD8⁺ T cells. A375, B16-F10, and HEK293T cells were maintained in DMEM (Gibco, South America) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (Beyotime, Shanghai), and cultured at 37°C in a humidified atmosphere containing 5% CO₂. A375 and B16-F10 cells were obtained from the Army Medical University (Chongqing, China), while the HEK293T line was sourced from ATCC. PDPN-targeted shRNA plasmids, cDNA fragments for PDPN and negative control particles were purchased from Shangwei Biotechnology (Shenzhen, China). Lentiviral packaging was conducted in HEK293T cells using UltraFection 3.0 transfection reagent (4A Biotech, Beijing), following the manufacturer's protocol.Harvested lentiviral particles were transduced into A375 and B16-F10 cells in the presence of 10 μg/mL polybrene (Sigma), and transduced cells were selected with 5 μg/mL puromycin (Sigma). For PDPN overexpression, pcDNA3.1-CTNNB1(human/mouse) or empty vector was stably integrated into A375 and B16-F10 cells under 5 μg/mL puromycin selection. All PDPN knockdown and CTNNB1 overexpression events were verified by western blotting.A375, B16-F10, A375 scramble, A375 PDPN shRNA, B16-F10 scramble, and B16-F10 PDPN shRNA cells were plated into 60-mm dishes. Cells were left untreated or subjected to CY12-RP2 (Qiang Yao, Shanghai) at indicated concentrations (20 µM) for 24 hours. Following treatment, cells were lysed in ice-cold RIPA buffer (MCE, Cat# HY-K1001) with protease inhibitor cocktail (MCE, Cat# HY-K0010). Proteins were separated by SDS-PAGE and transferred to PVDF membranes (Thermo Fisher Scientific). Blots were incubated with primary antibodies (see Supplementary Table 4)and HRP-conjugated, species-matched secondary antibodies. Bands were visualized using an ECL substrate (4A Biotech, Beijing, China) and imaged with a ChemiDoc instrument (Bio-Rad). Densitometric quantification was performed in ImageJ, and protein levels were normalized to β-actin as loading controls.Membrane-associated PD-L1 expression was quantified by flow cytometry using a fluorochrome-conjugated anti-PD-L1 antibody. Multiplex immunofluorescence staining was performed on a human melanoma tissue microarray (Cohort ID: HMelC112CD01) containing 112 cores-benign nevi (n=1), primary cutaneous melanomas (n=94), and distant metastases (n=17)-to assess the associations between PDPN expression, immune infiltration, and PD-L1 levels. FFPE sections underwent antigen retrieval (10 mM citrate buffer, pH 6.0, 95°C, 30 min), then simultaneous incubation with primary antibodies against PDPN (1:200, 11629-1-AP, Proteintech), CD8 (1:200, 66868-1-Ig, Proteintech), and PD-L1 (1:200, 28076-1-AP, Proteintech) for 2 hours at room temperature, followed by incubation with ABflo® 488-, 594-, or 647-conjugated goat anti-rabbit IgG (1:500 each, AS037, AS039, AS060) in PBS/1% BSA/1% goat serum (30 min, RT), all performed under coverslips with three 5-min PBS washes after each step. Nuclei were stained with DAPI (0.1 µg/mL, 5 min), and slides mounted with ProLong Diamond Antifade.Quantitative analysis of tissue microarray immunofluorescence was performed using the AQUA technique of QIF (NavigateBP). This method quantifies the target signal by calculating the quotient of the total target pixel intensity over the area of the molecularly designated compartment .For parallel assessment of CY12-RP2-mediated immune modulation in C57BL/6 melanoma models, lymph nodes, spleens, and tumors were harvested, sectioned into sequential 12 µm cryosections, fixed in ice-cold acetone, blocked with 5% goat serum antibodies for 2 hours at room temperature, followed by multiplex ABflo® secondary antibody staining using identical protocols as above. Imaging was performed using a Nikon confocal microscope, and cell quantification was carried out in QuPath v0.4.3 using tissue segmentation (DAPI threshold), cellular phenotyping (≥50 cells/mm² threshold), and spatial proximity analysis (≤15 µm between cells).A375 melanoma cells were maintained in DMEM supplemented with 10% FBS and 1% penicillin-streptomycin at 37°C with 5% CO₂. For luciferase assays, A375 cells seeded in 24-well plates were co-transfected at 70-80% confluency using Lipofectamine 3000. Each well received the pGL3-CD274 promoter construct (0.5 μg), pRL-TK (0.05 μg), and increasing amounts of β-catenin overexpression plasmid (0-0.5 μg; empty vector balanced total DNA) (Shangwei Biotechnology, Shenzhen).After 48 hours, cells were lysed, and luciferase activity was measured using the Dual-Luciferase Reporter Assay (11402ES60, YEASEN). Firefly luciferase readings were normalized to Renilla to determine relative promoter activity.The levels of major inflammatory cytokines, including TNF-α (BMS607-3TEN, Invitrogen), IL-1β (BMS6002, Invitrogen), IL-2 (BMS601, Invitrogen), and IFN-γ (BMS606-2, Invitrogen) were measured using an ELISA kit. Briefly, serum of C57BL/6 melanoma models was collected and added to an antibody pre-coated 96-well plate, and then followed up according to the manufacturer's instructions.RNA-seq libraries were generated from 1 μg total RNA per sample, using the Illumina TUNEL assay was conducted to evaluate apoptosis in paraffin-embedded tumor tissue sections with an cell death detection kit (MCE, HY-K1078). Following deparaffinization and rehydration, sections were subjected to antigen retrieval in sodium citrate buffer (pH 6.0) at 60°C for 4 hours, and then permeabilized with 0.3% Triton X-100 (in PBS, pH 7.4) for 15-30 min. Each tissue section was covered with 50 µL of TUNEL working solution, and incubated at 37°C for 1 hour. After PBS washes the next day, nuclei were stained with DAPI. Imaging was performed using a fluorescence microscope ( Nikon, Japan).All quantitative analyses were performed using GraphPad Prism 9 and results are presented as mean ± SD. Cell-based measurements, including migration distance and proliferation counts, were quantified using ImageJ-based morphometric analysis.Statistical comparisons between groups used unpaired, two-tailed Student's t-tests for normally distributed data, with significance thresholds: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; ns. (not significant) = p ≥ 0.05.PDPN functions as a pivotal orchestrator of immune checkpoint networks and contributes to the establishment of immunosuppressive landscapes (5,8). Cluster analysis using VOSviewer identified PDPN-associated modules encompassing platelet aggregation, tumor microenvironment, inflammation, lymphatic metastasis, and CD8⁺ T lymphocyte biology (Fig. 1A). Evaluation of melanoma data from The Cancer Genome Atlas (TCGA) revealed strong positive correlations between PDPN expression and several immune checkpoint receptors, including PD-L1, CTLA4, LAG3, TIGIT, and BTLA, with the association with PD-L1 (CD274) being the most prominent (r=0.504, p<0.001; Fig. 1B). STRING database analysis further substantiated the direct physical and functional interactions between PDPN and these immune checkpoints. Additionally, PDPN demonstrated significant network connectivity (FDR<0.05) with major immune lineage markers (CD86, CD8A, CD4, CD80; Fig. S1B). TIMER2.0 analysis of 471 melanoma samples (368 metastatic, 103 primary) by single-cell RNA-seq revealed that PDPN expression was inversely correlated with infiltration by antitumor immune populations (CD4⁺ T cells, CD8⁺ T cells, NK cells, and M1 macrophages), and positively correlated with immunosuppressive populations (M2 macrophages, regulatory T cells), most notably with M2 macrophages (Table 1). These associations were independently validated using a separate single-cell RNA-seq TCGA cohort (n=42), which demonstrated that PDPN modulates immune cell infiltration, suppresses cytotoxic immune cell functions, and inversely correlates with T cell accumulation within the tumor (Fig. 1C-E).TTo directly evaluate the immunomodulatory role of PDPN, the inhibitory peptide CY12-RP2 was tested in melanoma xenograft models using both immunodeficient BALB/c nude mice and immunocompetent C57BL/6 mice. As shown in Fig. 1G, administration of CY12-RP2 led to a significant reduction in tumor growth rates relative to controls in both animal models. CY12-RP2-treated cohorts showed substantial decreases in both tumor volume and terminal tumor weight (Fig. 1F and 1H), achieving tumor inhibition rates of 42.22% in BALB/c nude mice and 60.62% in C57BL/6 mice (Fig. 1I). These data indicate that CY12-RP2 exerts enhanced antitumor effects in immunocompetent hosts compared to immunodeficient counterparts. Integrated multi-database analysis revealed a significant positive association between PDPN and the immunosuppressive checkpoint receptor PD-L1 in melanoma transcriptomes. To confirm this, a human tissue microarray (HMelC112CD01, Shanghai Xinchao Biotechnology) comprising 112 melanoma specimens was assessed (19). Triple immunofluorescence staining showed limited co-expression of PDPN and PD-L1 in benign nevi, whereas marked upregulation and co-expression were observed in primary (e.g., #3, #22) and metastatic (e.g., #110) melanoma lesions (Fig. 2A).Semi-quantitative assessment of 111 melanoma specimens, of which 102 were valid, revealed a correlation between PDPN and PD-L1 expression (overall co-expression 46.1%, p < 0.005), with disease stage-specific disparities: PDPN prevalence was higher in primary lesions (61.6%) than in metastases (43.8%), whereas PD-L1 expression was elevated in metastatic lesions (87.5%) compared to primary tumors (61.6%). Notably, the co-expression rates of PDPN and PD-L1 were consistent across disease stages (46.5% in primary versus 43.8%