Gasdermin D antagonizes immunosuppression in prostate cancer by inducing LAMC2 degradation to block M2 macrophage polarization

Qindan Du; Zhen Han; Yumeng Zhang; Ying Li; Ao Sun; Bin Wang; Yongquan Chen; Xiaoying Wang

doi:10.1136/jitc-2025-014533

. 2026 May 27;14(5):e014533. doi: 10.1136/jitc-2025-014533

Gasdermin D antagonizes immunosuppression in prostate cancer by inducing LAMC2 degradation to block M2 macrophage polarization

Qindan Du ¹, Zhen Han ¹, Yumeng Zhang ¹, Ying Li ¹, Ao Sun ¹, Bin Wang ¹, Yongquan Chen ^1,², Xiaoying Wang ^1,^2,^✉

PMCID: PMC13218149 PMID: 42203263

Abstract

Background

The N-terminal domain of gasdermin D (GSDMD) is a known suppressor of prostate cancer (PCa), primarily recognized for its tumoricidal pyroptotic function. However, research on the full-length GSDMD protein and its relationship with PCa remains unexplored. Therefore, the aim of this study was to investigate the potential non-pyroptotic mechanism of action of GSDMD in preventing PCa immune evasion.

Methods

The study used GSDMD deletion models to assess its impact on tumor progression and immune cell profiles. Molecular techniques, including protein interaction analysis, were employed to identify the binding partner of GSDMD and the subsequent signaling pathways. Correlative analysis was performed between GSDMD expression and immune markers in human PCa samples.

Results

GSDMD deletion increased tumor progression and reduced survival, cytotoxic T lymphocyte (CTL) activation/infiltration and tumor cell susceptibility to T-cell death, promoting an immunosuppressive tumor microenvironment (TME). GSDMD-bound laminin subunit gamma-2 (LAMC2), thus inducing its ubiquitin-mediated degradation. Consequently, GSDMD deficiency stabilized LAMC2, driving M2 macrophage polarization and CXC motif chemokine ligand 5 (CXCL5) secretion. CXCL5 interacted with CXCR2 to activate the AKT/NF-κB axis, increasing vascular endothelial growth factor (VEGF)/vascular endothelial growth factor receptor 2 (VEGFR2) and suppressing CTL activity. GSDMD expression in human PCa positively correlated with CD8⁺ T-cell infiltration and negatively with M2-phenotype tumor-associated macrophages.

Conclusion

This study reveals a novel non-pyroptotic pathway through which GSDMD inhibits PCa immune evasion by targeting LAMC2 for degradation, thereby preventing the establishment of an immunosuppressive TME. The findings suggest that activating GSDMD represents a promising therapeutic strategy to reprogram the TME from immunosuppressive (“cold”) to immunoreactive (“hot”), potentially enhancing the efficacy of immunotherapy in advanced PCa.

Keywords: Genitourinary Cancer, Macrophage, Prostate Cancer, T cell, Tumor microenvironment - TME

WHAT IS ALREADY KNOWN ON THIS TOPIC

Gasdermin D (GSDMD) is the executor of cell pyroptosis, being involved in host defense against pathogen infection and danger signals.
The loading of GSDMD N-terminal domain into extracellular vesicles targeting prostate-specific membrane antigen-positive prostate cancer (PCa) by genetic engineering inhibits tumor progression and activates antitumor immunity.

WHAT THIS STUDY ADDS

GSDMD interacted with laminin subunit gamma-2 (LAMC2) protein, leading to LAMC2 ubiquitination.
GSDMD knockout induced macrophage M2 polarization, secreted CXC motif chemokine ligand 5, promoted angiogenesis, and reduced the activation and infiltration of cytotoxic T lymphocytes into the tumor.
The critical GSDMD residue (D362A) involved in tumor suppression was identified.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

Using GSDMD activators may have a synergistic effect with existing immune checkpoint inhibitors.
Targeting GSDMD could be a novel strategy in treating PCa.

Introduction

Prostate cancer (PCa), a type of cancer closely linked with age, remains a major global health challenge, ranking among the most frequently diagnosed malignancies and a leading cause of cancer-related death in men.¹ Because of the changes in hormone levels and alterations in immune system activity caused by aging, early-onset PCa and late-onset PCa exhibit significant differences in clinical manifestations and pathological features² and despite advances in early PCa detection and treatment, advanced hormone-resistant PCa remains largely incurable.³ This emphasizes the need to identify additional molecular factors that drive disease progression to identify effective therapeutic targets.

PCa pathogenesis depends in part on the tumor microenvironment (TME),⁴ which is immunosuppressive in advanced PCa.5,7 Dynamic interactions among cancer cells, immune cells, stromal components, and the extracellular matrix (ECM) critically influence tumor growth, invasion, metastasis, and therapeutic response.^{8 9} A key feature is the dysfunction of critical immune cell populations. Tumor-associated macrophages (TAMs) that polarize to the M2 phenotype play important roles in the promotion of tumor growth, angiogenesis, and metastasis and the suppression of adaptive immunity.^{10 11} Additionally, T-cell dysfunction manifests as exhaustion, altered cytotoxicity, and reduced cytokine production, severely compromising effective antitumor immunity.¹² The dynamic interactions between cancer cells, M2-TAMs, and dysfunctional T cells create an environment that facilitates immune escape and disease progression.13,15

Beyond stromal and immune constituents, molecules that regulate cell death and inflammation can profoundly shape the TME. Gasdermin D (GSDMD) is the executioner protein that mediates pyroptosis downstream of inflammasome activation.16,18 In addition to its canonical lytic function, GSDMD has a non-pyroptotic function.^{19 20} Indeed, it can function as a signaling molecule or transcriptional regulator, modulating cellular pathways involved in inflammation, proliferation, and survival. Moreover, it regulates cytokine production (eg, facilitating interleukin (IL)-1β and interleukin (IL)-18 maturation and release) and influences signaling cascades such as the nuclear factor kappa-B (NF-κB) and MAPK pathways^{21 22}, which are fundamental for both oncogenesis and immune cell function. Dysregulation of GSDMD expression and activity is frequently observed in various cancers, suggesting that GSDMD is involved in tumorigenesis through mechanisms beyond pyroptotic cell death. However, the specific contribution of GSDMD, particularly its non-pyroptotic activity, to PCa pathogenesis and the establishment/maintenance of its immunosuppressive TME represents a critical knowledge gap. Thus, it remains unknown whether GSDMD expression or activation in PCa cells or in the TME directly promotes tumorigenesis or therapeutic resistance through inflammasome-independent, non-pyroptotic mechanisms; how GSDMD influences the balance of immune cell phenotypes in the PCa TME is also unclear.

This study aimed to investigate the functional significance of GSDMD in PCa development and progression, with a specific focus on its non-pyroptotic role and its impact on key immune cell populations in the TME. On the basis of emerging evidence of its non-canonical functions in signaling and immune regulation, it modulates macrophage polarization by inducing ubiquitination-mediated degradation of laminin subunit gamma-2 (LAMC2). The absence of GSDMD drives macrophage polarization toward the M2 phenotype, which promotes CXC motif chemokine ligand 5 (CXCL5) secretion and activates angiogenesis-related signaling pathways while significantly decreasing the activity of tumor-infiltrating CD8⁺ T cells. In murine models of PCa, this mechanism significantly regulates the progression of PCa. This discovery of a GSDMD-LAMC2 axis reveals a previously unknown immunoregulatory mechanism. Understanding these non-canonical functions of GSDMD might reveal new aspects of PCa biology and aid in the identification of innovative strategies for intervention.

Results

GSDMD is reduced in PCa and indicates poor prognosis

GSDMD is an executor of inflammatory pyroptosis and is increasingly involved in autoimmune and inflammatory diseases. Recombinant extracellular vesicles were constructed using genetic engineering to express a high-affinity prostate-specific membrane antigen (PSMA)-targeting single-chain variable fragment on their surface while being loaded with the N-terminal domain of GSDMD. This engineered platform enables specific delivery to PSMA-positive PCa, triggering inflammatory pyroptosis in tumors.¹⁸ This process results in the inhibition of tumor growth and the activation of an antitumor immune response. However, the relationship between the full-length GSDMD protein and PCa remains unexplored. Our preliminary bioinformatics analysis using the Gene Expression Profiling Interactive Analysis (GEPIA) revealed a consistent and significant downregulation of GSDMD messenger RNA (mRNA) expression in human PCa tissues compared with matched normal prostate tissues (figure 1A). Furthermore, the Cancer Cell Line Encyclopedia dependency screening indicated that the optimal survival or growth of multiple established PCa cell lines depended on GSDMD expression (figure 1B), suggesting a non-canonical role of GSDMD beyond the execution of pyroptosis in prostate carcinogenesis.

Next, we evaluated GSDMD protein expression by immunofluorescence (IF) on a human PCa tissue microarray to validate these findings and explore their clinical relevance. Statistical analysis revealed a significant inverse correlation between GSDMD expression and clinical stage, with lower GSDMD expression in more advanced disease (figure 1C). Patients with high GSDMD protein expression had significantly longer overall survival than those with low or no expression (figure 1D).

Therefore, while GSDMD could be used therapeutically to induce death through pyroptosis, its endogenous downregulation of PCa progression and its essential function in cancer cell survival revealed a fundamental yet undefined non-pyroptotic function in prostate carcinogenesis.

GSDMD expression inhibits PCa progression in both genetically engineered mouse and syngeneic tumor models

Having established a correlation between GSDMD loss and poor prognosis, we sought to determine its causal role in PCa progression using genetically engineered models. To investigate this, we generated PB-Cre4+PtenL^L/L Gsdmd^L/L double-knockout (pDKO) mice by crossing Gsdmd-loxP mice with PB-Cre4+PtenL^L/L single-knockout (pSKO) PCa model mice (figure 2A and online supplemental figure S1A). Prostate-specific deletion of Gsdmd was successfully achieved in pDKO mice (figure 2B). Statistical analysis of the survival data revealed that the survival time of pDKO mice was significantly shorter than that of pSKO mice (figure 2C). The prostates of 10-month-old pDKO and pSKO mice were dissected, and each prostate lobe was found to be significantly larger in pDKO mice than in pSKO mice (figure 2D). H&E staining and immunohistochemistry (IHC) indicated that the degree of malignancy and proliferation were significantly greater in pDKO tumors than in pSKO tumors (figure 2E and F). We then established Pten (PP) and Pten/Gsdmd (GPP) cell lines from pSKO and pDKO tumors to study the function of Gsdmd in tumor immune regulation. Compared with PP subcutaneous tumors, GPP subcutaneous tumors grew significantly faster after subcutaneous injection of the tumor cells into C57BL/6 male mice (figure 2G-I). The use of GSDMD inhibitors (such as LDC7559, DMF and disulfiram)23,25 had a procarcinogenic effect similar to that of prostate-specific knockout of Gsdmd (online supplemental figure S1B).

Histological examination of the end-stage prostate tumors revealed that the prostate tumors of the mice in the pDKO group were rich in blood vessels. Compared with that in the pSKO group, the expression of CD31 in each lobe of the prostate in the pDKO group was significantly greater, as detected by IHC and western blot analysis (figure 2J,K). The supernatants of PP and GPP cells were collected to stimulate human umbilical vein endothelial cells (HUVECs) for angiogenesis experiments. The results indicated that the supernatant of the GPP cells significantly increased the angiogenesis ability and vascular permeability of the HUVECs (figure 2L,M). These findings demonstrated that prostate-specific Gsdmd deletion promoted intratumoral angiogenesis. Validation of the PCa-suppressive activity of Gsdmd in our newly established genetically engineered and syngeneic mouse models prompted investigation into the previously undefined molecular mechanisms governing this function.

However, on deeper analysis of the tumor microenvironment as a whole, we observed an unexpected phenomenon: despite the specific knockout of GSDMD in prostatic epithelial cells, the overall pyroptotic signaling within the prostate tumors was paradoxically enhanced (online supplemental figure S2A). We therefore fractionated the prostate tumors into CD45⁺ and CD45⁻ cell populations. GSDMD-NT refers to the N-terminal domain released on cleavage of the GSDMD protein, which forms pores in the cell membrane and is the key executor of pyroptosis. The results showed that while the level of pyroptosis in CD45⁻ (largely tumor) cells remained virtually unchanged, the CD45⁺ compartment in the pDKO group exhibited a significant increase in pyroptosis compared with the pSKO control (online supplemental figure S2B). This observation was further corroborated by in vitro experiments demonstrating that modulating GSDMD expression levels in PCa cell lines did not impact their susceptibility to pyroptosis (online supplemental figure S2C and D). Collectively, these findings suggest that the loss of PCa cell-intrinsic GSDMD can extrinsically influence immune cells in a non-autonomous manner. This highlights that GSDMD assumes distinct, and even opposing, compartment-specific roles within the TME, revealing a functional network far more complex than previously appreciated.

De novo-designed binding partners reveal GSDMD–LAMC2 crosstalk and establish a role for GSDMD in LAMC2 ubiquitination

The tumor-suppressive phenotype prompted us to investigate the underlying molecular mechanisms. We hypothesized that GSDMD exerts its function through specific protein interactions. To further investigate how GSDMD influences the progression of PCa, RFdiffusion was used to further explore how these proteins interact with each other by designing high-affinity binding partners for GSDMD based on the three-dimensional structures of the full mouse proteome.^{26 27} First, the designed binding partners were subsequently compared with the three-dimensional structure library of the full mouse proteome to identify proteins that might interact with GSDMD with high binding energy. Next, we established a de novo protein design using RFdiffusion to identify new GSDMD interactors that potentially mediate its tumor-suppressive function, and we designed high-affinity binding partners targeting the active site region (A2-A96) of GSDMD (online supplemental figure S3A and B). Structural comparisons were then conducted between these designed binding partners and the co-immunoprecipitation (Co-IP) protein products identified by liquid chromatography-tandem mass spectrometry (LC-MS/MS) (online supplemental figure S3C and D, and online supplemental Tables S7 and S8). LAMC2 showed high structural similarity to the short peptide that binds to the generative formula. The results suggested a potential interaction between LAMC2 and GSDMD. Computational docking predicted strong interaction interfaces between both mouse and human GSDMD and LAMC2 (figure 3A,B). Western blot analysis revealed that in PCa cells, GSDMD and LAMC2 are predominantly localized in the cytoplasm (online supplemental figure S3E). This cytoplasmic localization of LAMC2 was further confirmed by immunofluorescence confocal microscopy (online supplemental figure S3F). Furthermore, Co-IP confirmed the physical interaction between GSDMD and LAMC2 (figure 3C). LAMC2 protein expression was significantly higher in pDKO tumors than in pSKO tumors (figure 3D). GSDMD overexpression (OE) mediated the downregulation of LAMC2 protein expression (figure 3D). Molecular docking was performed on the GSDMD and LAMC2 proteins to identify their interaction sites. On the basis of the results of the accessible surface area (ASA)/ buried surface area (BSA) analysis, solvation energy calculations, and the percentage of surface area buried on binding, amino acid residues 362 and 487 were prioritized (figure 3E). The D362A/C487A mutant (both tagged with 3xFlag) was transiently overexpressed in RM-1 cells by transfection with a plasmid encoding the mutated gene (online supplemental Tables S2 and S3). Western blot analysis confirmed successful mutant OE (figure 3F). Notably, OE of the D362A mutant produced effects comparable to those of the control, without altering LAMC2 protein levels (figure 3F). Co-IP assays further confirmed the absence of an interaction between the D362A mutant and HA-LAMC2 (figure 3G), whereas an interaction between the C487A mutant and HA-LAMC2 was detected (figure 3H). These results suggest that D362A is a key site in GSDMD for binding to LAMC2, which may mediate tumor suppression.

During the validation of the above results, we observed an interesting phenomenon: expression analysis of LAMC2 revealed that GSDMD did not alter the mRNA expression of LAMC2, confirming that GSDMD regulated LAMC2 expression at the post-transcriptional level (p=0.6298 and 0.8632, figure 3I). The critical D362 residue resides in a predicted functional domain of GSDMD (gasdermin, PUB domain (IPR041263)—InterPro entry—InterPro; online supplemental figure S3G), also known as “Peptidoglycan-binding U box”. The PUB domain is a key regulatory module in the C-terminal region. The C-terminal region of GSDMD primarily maintains a self-inhibitory state and interacts with other regulatory proteins, yet there has been limited development in this area. These findings prompted us to investigate whether GSDMD regulated LAMC2 stability through ubiquitination. A cycloheximide chase assay demonstrated faster LAMC2 degradation after GSDMD OE (figure 3J). In pharmacological inhibition experiments, the proteasome inhibitor MG132, but not the lysosome inhibitor chloroquine, restored LAMC2 protein expression in GSDMD-overexpressing cells (figure 3K), suggesting the involvement of the ubiquitin–proteasome system. Finally, Co-IP after MG132 treatment revealed LAMC2 polyubiquitination in the presence of GSDMD, which increased after GSDMD OE (figure 3L). Having established that GSDMD promotes the polyubiquitination of LAMC2, we sought to identify the specific E3 ubiquitin ligase(s) involved, as GSDMD itself lacks catalytic activity. We performed a two-step computational screening strategy. First, we conducted molecular docking of GSDMD against a comprehensive library of mouse ubiquitin ligases, ranking them by predicted binding affinity. The top 20 candidates from this screen were then docked against LAMC2. Notably, RNF121 and RNF148 ranked in the top two in both sequential screens (online supplemental Table S10 and S11), exhibiting high predicted affinity for both GSDMD and LAMC2 (online supplemental figure S3H). This convergent result suggested that RNF121 and/or RNF148 were strong candidates for forming a ternary complex, potentially serving as the bridge through which GSDMD, possibly via its PUB domain, recruited the ubiquitination machinery to target LAMC2 for degradation. Collectively, these results suggested that GSDMD promoted the ubiquitination and proteasomal degradation of LAMC2, affecting the stability of LAMC2 and thereby the activity of LAMC2-related pathways.

GSDMD regulates macrophage M2 polarization through LAMC2

Having identified LAMC2 as a key degradation target of GSDMD, we next asked how the GSDMD-LAMC2 axis promotes the progression of PCa. Although initial attempts to define a cell-intrinsic role for LAMC2 in PCa cells yielded limited effects (online supplemental figure S4A- E), existing evidence suggests that LAMC2 may recruit TAMs and promote their M2 polarization.28,31 We therefore hypothesized that GSDMD deficiency increased LAMC2 protein expression, thereby driving M2 macrophage polarization in PCa. This hypothesis was confirmed through experiments using culture supernatants from control and GSDMD-OE RM-1 cells to stimulate RAW264.7 macrophages. GSDMD-OE supernatants significantly decreased the expression of M2 macrophage markers (Arg-1, CD206 and IL-10; figure 4A) and the percentage of CD206 M2-TAMs (figure 4B) and concurrently decreased the secretion of IL-10 (figure 4C).

ELISA confirmed that shGSDMD RM-1 cells significantly increased the secretion of LAMC2 into the conditioned medium (online supplemental figure S5A). IF assays showed that macrophages cultured with supernatant from GSDMD-knockdown cells were surrounded by increased LAMC2 deposition (online supplemental figure S5B), supporting its paracrine action. We next directly tested the functional impact of soluble LAMC2 (sLAMC2) on macrophages. The results demonstrated that sLAMC2 could directly induce chemotactic migration of macrophages and promote their M2 polarization, as evidenced by upregulated expression of M2 markers. These effects were all blocked by LAMC2 neutralizing antibodies (online supplemental figure S5C-E).

Flow cytometry analysis revealed significantly greater M2-TAM infiltration in pDKO tumors than in pSKO control tumors, consistent with the above in vitro results (figure 4D). IF staining of GSDMD, CD68, and CD163 in tissue microarrays of human primary PCa revealed that the expression level of GSDMD was negatively correlated with the extent of M2-TAM infiltration (figure 4F). Migration assays further demonstrated increased chemotaxis in RAW264.7 cells compared with that in control cells (figure 4E). Bone marrow-derived macrophages from wild-type mice treated with the GSDMD inhibitor LDC7559 (10 µM) presented upregulated expression of M2 markers (IL-10, CD206 and Arg1; figure 4G and online supplemental figure S6A) and an increased proportion of CD206⁺ cells (figure 4H and online supplemental figure S6B) and secretion of IL-10 (online supplemental figure S6C), all of which occurred without alterations in the expression of pyroptosis markers (IL-18 and IL-1β; Online supplemental figure S6D). The interplay of GSDMD and LAMC2 in macrophages was established by stable OE studies confirming that GSDMD-OE reduced LAMC2 protein expression (online supplemental figure S6E–G), whereas LAMC2-OE did not affect GSDMD expression or cleavage (online supplemental figure S6H). In addition, increased LAMC2 expression promoted M2 macrophage polarization, whereas GSDMD overexpression suppressed M2 macrophage polarization. Combined manipulation (OE of GSDMD combined with knockdown of LAMC2, or knockdown of GSDMD combined with OE of LAMC2) failed to produce additive effects on the polarization phenotypes (online supplemental figure S6I-M), indicating that LAMC2 functions downstream of GSDMD, with potential compensatory mechanisms modulating their crosstalk.

The aforementioned results demonstrate that GSDMD-low tumor cells secrete higher levels of LAMC2. To understand how this increased LAMC2-mediated signaling to TAMs, we next investigated how LAMC2 mediates this signal to TAMs. Further mechanistic studies revealed that sLAMC2 functions through integrin α6β4 (ITGA6 and ITGB4) on the macrophage surface: immunofluorescence staining showed co-localization of LAMC2 with integrin α6β4 and the downstream signaling molecule FAK on macrophages (online supplemental figure S7A). Functionally, the use of integrin α4 inhibitors effectively inhibited sLAMC2-induced M2 polarization of macrophages (online supplemental figure S7B and C), further confirming the critical role of this receptor in LAMC2 signaling. Additionally, sLAMC2 stimulation significantly activated the FAK-AKT signaling pathway, as indicated by elevated phosphorylation levels (online supplemental figure S7D). Collectively, these findings demonstrated that GSDMD deficiency increased LAMC2 protein levels and secretion, and that the secreted LAMC2 acted via the α6β4 integrin-FAK-AKT axis on macrophages to promote their M2 polarization.

Myeloid-specific deletion of GSDMD exacerbates tumor growth by remodeling the immunosuppressive tumor microenvironment

Since LAMC2 can activate pro-M2 signaling pathways within macrophages, we further explore how macrophages respond to and execute this instruction. We hypothesize that their endogenous GSDMD may play a pivotal role. The dependence of GSDMD-mediated tumor suppression on macrophages was investigated by intraperitoneal administration of clodronate liposomes into mice, which resulted in approximately 50% depletion of peripheral blood macrophages (figure 5A). Macrophage ablation abrogated the tumor-promoting effect of GSDMD knockdown, as shown by the significant difference in tumor size/weight between the sh-GSDMD and control groups of clodronate liposome-treated mice (figure 5B).

To elucidate the functional significance of GSDMD in macrophage antitumor immunity, we generated a myeloid-specific Gsdmd knockout mice (Gsdmd^L/L;Lyz2-iCre hereafter referred to as Gsdmd^CKO) (figure 5C). Subsequent subcutaneous implantation of RM-1 PCa cells revealed a striking phenotype: tumors in Gsdmd^CKO mice exhibited significantly accelerated growth kinetics and reached a larger final volume and weight compared with those in their Gsdmd^L/L littermate controls (Gsdmd^L/L). Moreover, Gsdmd^CKO tumors exhibited distinct malignant features under H&E staining, with pathological necrosis at the tumor center and tumor angiogenesis (figure 5D). Simultaneously, tumors from Gsdmd^CKO mice exhibited a significantly higher level of LAMC2 protein expression compared with those from Gsdmd^L/L mice (figure 5E). This finding indicates that macrophage-intrinsic GSDMD is a critical negative regulator of tumor progression.

Given the pivotal role of macrophages in shaping the tumor immune landscape, we next profiled the phenotype of TAMs by flow cytometry. Similarly, the loss of GSDMD in macrophages drove a profound shift towards an M2-polarized, immunosuppressive state. Quantitative analysis demonstrated a significant increase in the proportion of CD206⁺ M2 macrophages within the total CD11b⁺ F4/80⁺ TAM population in Gsdmd^CKO tumors (figure 5F). Macrophages derived from tumor tissue displayed significantly enhanced IL-10 secretion, as assessed by ELISA; additionally, their expression of canonical M2 macrophage markers (IL-10, Arg-1, CD206) was markedly elevated at the transcriptional level (figure 5G,H). This suggests that GSDMD deficiency impairs the pro-inflammatory function of macrophages, favoring their differentiation into a protumoral phenotype.

Concomitantly, we assessed the adaptive immune cell compartment and discovered a significant reduction in the overall infiltration of T cells within the Gsdmd^CKO tumors (figure 5I,J). This decrease in T-cell abundance, coupled with the enhanced M2 polarization, provides a compelling cellular mechanism for the observed tumor growth advantage.

GSDMD activates the AKT/NF-κB/VEGF pathway in a CXCR2-dependent manner through CXCL5

Transcriptomic analysis via RNA sequencing (RNA-seq) enabled the comparison of gene expression patterns in pSKO and pDKO tumors to investigate the more extensive signal transduction mechanisms of GSDMD in PCa and the impact of M2 macrophage polarization. Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis revealed that IL-17-related and tumor necrosis factor (TNF)-α-related pathways were enriched in pDKO tumors (figure 6A). We used volcano plots and cluster analysis to screen for significantly differentially expressed genes, and the results revealed that the expression of the chemokine CXCL5 was significantly upregulated in pDKO tumors (figure 6B,C). Moreover, multiple studies have indicated that IL-17 and TNF-α synergistically increase CXCL5 expression.^{32 33} We also confirmed these results by western blotting in PCa cell lines (figure 6D). Differential expression analysis of the RNA-seq data revealed the upregulation of multiple IL and chemokine family members in pDKO tumors (figure 6E). Consistent with the RNA-seq data, the protein expression of cytokines in pDKO tumors was also upregulated, and the difference in CXCL5 expression was particularly significant (figure 6F).

M2 macrophages themselves secrete many signaling molecules that promote tumor growth, angiogenesis, and immunosuppression. We hypothesized that CXCL5, as a key upregulated chemokine, could be one of these “accomplice” molecules secreted by M2 macrophages. This hypothesis was based on the finding that the transcription and secretion of CXCL5 were specifically activated in polarized M2 macrophages stimulated by conditioned medium from GSDMD-deficient tumor cells. Mechanistically, this process depended on the activation of intracellular signaling initiated by the LAMC2-integrin α6β4 axis, particularly the FAK-AKT signaling pathway, indicating that CXCL5 was a direct downstream output of the M2 polarization program. Real-time quantitative PCR (RT-qPCR) and IHC confirmed that specific GSDMD knockout in PCa led to an increase in CXCL5 expression (figure 6G,H). Furthermore, preliminary experiments demonstrated that CXCL5 production is closely associated with LAMC2/GSDMD expression levels and M2 macrophage polarization, thereby confirming that it originates from this specific immune cell subset (online supplemental figure S8A-G). To elucidate the downstream signaling pathway of CXCL5, we used western blotting to detect the expression of pathway-related proteins in in situ carcinoma and subcutaneous tumors and revealed that CXCL5 increased the phosphorylation of AKT in a CXCR2-dependent manner; then, NF-κB translocated into the nucleus, increasing the expression of vascular endothelial growth factor (VEGF) and vascular endothelial growth factor factor 2 (VEGFR2) (figure 6I-K). While both OE of GSDMD and knockdown of CXCL5 exhibited inhibitory effects on subcutaneous tumor formation in mouse PCa models, the combined approach demonstrated significantly enhanced efficacy (figure 6L and online supplemental figure S8H). VEGF promotes abnormal tumor angiogenesis and vascular permeability. It also drives immunosuppression by inhibiting T-cell activity via VEGFR2 and expanding myeloid-derived suppressor cells (MDSCs).34,38 MDSC expansion is further stimulated by the upregulation of CXCL5 (recruited via CXCR2), TNF-α, and VEGF itself, with NF-κB signaling being crucial for their development and activation.39,42 Consistent with the increase in the level of CXCL5 and its role in MDSC recruitment, flow cytometry confirmed that the proportion of MDSCs was significantly greater in pDKO tumors than in pSKO tumors (online supplemental figure S8I) and that these cells counteract the immune response of antitumor T cells.⁴³ Prostate-specific knockout of GSDMD led to upregulation of CXCL5, which in turn activated the AKT/NF-κB/VEGF pathway in a CXCR2-dependent manner. This cascade promoted angiogenesis and remodeled the tumor immune microenvironment, thereby driving PCa initiation and progression. Therefore, CXCL5 was not merely a bystander in the microenvironment; it likely acted as a critical executor by which M2 macrophages actively reshaped an immunosuppressive and pro-angiogenic niche, thereby serving as an amplifier within the protumorigenic effects of the GSDMD-LAMC2 pathway.

GSDMD enhances immunotherapy by promoting CTL infiltration into PCa

Having established that GSDMD deficiency creates an immunosuppressive microenvironment rich in M2 macrophages and MDSCs, we next investigated how this affects antitumor immunity, particularly that involving CD8⁺ T cells. An analysis using the Tumor IMmune Estimation Resource (TIMER) V.2.0 database revealed that GSDMD expression was positively correlated with CD8⁺ T-cell infiltration in human prostate tumors (figure 7A). The greater infiltration of CD8⁺ T-cell subsets in pSKO tumors than in pDKO tumors was validated by IF performed on the prostate tissues of pDKO and pSKO mice (figure 7B). When the tumors from the pSKO and pDKO mice were compared using flow cytometry, the number of CD8⁺ T cells decreased in the Gsdmd-deficient tumors (figure 7C). This difference in the number of T cells could be due to the altered vasculature because pSKO and pDKO tumors had different microvessel densities (figure 2J–M). CD8⁺ T-cell activity was evaluated by isolation of tumor-infiltrating lymphocytes from size-matched pSKO and pDKO tumors using magnetic bead separation, followed by RNA purification and RT-qPCR analysis. Two distinct marker panels were examined (1): cytokine/effector molecules and (2) T-cell exhaustion markers (figure 7D). Compared with CD8⁺ T cells from pDKO mice, CD8⁺ T cells from pSKO mice expressed significantly more cytokines and effectors (eg, interferon (IFN)-γ, granzymes and perforin). With respect to exhaustion markers, compared with CD8⁺ T cells from pSKO mice, CD8⁺ T cells from pDKO mice expressed more markers.

The flow cytometry results consistently showed that the trends in protein expression align with the mRNA expression patterns detected by RT-qPCR (figure 7E). This concordance at both the transcriptional and translational levels strengthens our conclusions regarding the phenotypic and functional states of the T cells under investigation.

We postulated that GSDMD loss in PCa cells conferred resistance to cytotoxic T lymphocyte (CTL)-mediated killing because of the reduced CTL infiltration and increased tumor growth in Gsdmd-deficient tumors. GSDMD and chicken ovalbumin (OVA) were stably expressed in the murine PCa cell line RM-1, and these sublines were co-cultured with B3Z cells (a CD8⁺ T-cell hybridoma cell line). The cell line overexpressing GSDMD was more sensitive to the killing effect of B3Z cells (figure 7F). Carboxyfluorescein diacetate succinimidyl ester‌ (CFSE) proliferation assays revealed that GSDMD OE promoted the proliferation of B3Z cells (figure 7G). Therefore, the loss of GSDMD in PCa cells, which is linked to accelerated tumor growth and reduced T-cell infiltration, suggests a potential role for GSDMD-mediated processes in promoting an antitumor immune microenvironment.

Given that GSDMD enhances T-cell cytotoxicity and infiltration, we explored its potential to synergize with immunotherapeutic strategies. Since the knockdown of GSDMD in PCa cells reduces CTL infiltration in vivo and GSDMD OE enhances the killing of B3Z cells in vitro, upregulating the expression of GSDMD may enhance the efficacy of immunotherapy. Polymorphonuclear MDSCs constitute an obstacle to immunotherapy.³⁹ Blocking CXCR2 can reduce the infiltration of polymorphonuclear MDSCs and slow tumor progression.⁴⁴ SX-682 is an orally effective allosteric inhibitor of CXCR1 and CXCR2. CXCR2 is the receptor of CXCL5. SX-682 can block the recruitment of tumor MDSCs and promote T-cell activation and antitumor immunity.⁴⁴ We provided SX-682-supplemented drinking water to mice bearing subcutaneous tumors. SX-682 monotherapy showed moderate efficacy, while the combination of GSDMD OE and SX-682 treatment had the best efficacy (online supplemental figure S9A-C), and the tumors in this treatment group had the greatest degree of T-cell infiltration (online supplemental figure S9D).

Subcutaneous tumor formation experiments were performed simultaneously in C57BJ/6 mice and severely immunodeficient NCG mice. The results indicated that the inhibitory effect of GSDMD OE on the formation of subcutaneous PCa tumors was significantly reduced in the absence of immune cells (online supplemental figure S9E), confirming the synergistic effect of GSDMD and immune cells in the treatment of PCa. In an adoptive T-cell therapy setting, B3Z cells were injected into immune-deficient NCG mice harboring RM1-OVA tumors formed by control tumor cells or GSDMD-overexpressing cells. The results revealed that the infiltration of CD8⁺ T cells in the GSDMD OE group was two times as high as that in the control group (online supplemental figure S9F).

Next, C57BL/6 mice subcutaneously implanted RM-1 tumors reaching 100–150 mm³ in size were treated with isotype control IgG or ICB antibodies to evaluate the effect of GSDMD on immune checkpoint blockade (ICB). Treatment with either ICB alone or GSDMD OE induced some degree of tumor suppression. However, the combination of ICB and GSDMD OE resulted in the formation of the smallest tumors and the longest survival (online supplemental figure S9G and H). Overall, these findings suggest that targeting GSDMD might enhance the efficacy of both ICB and adoptive T-cell immunotherapy in PCa.

Human GSDMD overexpression is associated with increased CTL infiltration

To explore the correlation between GSDMD expression and CTL infiltration in human PCa, immunological correlation analysis was conducted using the TIMER V.2.0 database and Tumor-Immune System Interaction Database (TISIDB). The results revealed that GSDMD expression in patients with PCa was positively correlated with the degree of macrophage and CD8⁺ T-cell infiltration but inversely correlated with the degree of M2-TAM infiltration (figure 8A-C). Furthermore, GSDMD expression was significantly associated with distinct PCa immune subtypes, including those associated with wound healing, IFN-γ dominance, inflammation, and lymphocyte depletion (figure 8D). To further elucidate the relationship between GSDMD and CTL infiltration in human PCa, we performed Gene Set Enrichment Analysis of CTL-associated gene signatures across multiple transcriptomic datasets from primary and metastatic PCa. The results demonstrated an enrichment of CTL signatures in patient samples with elevated GSDMD expression (figure 8E). Subsequent IF staining of GSDMD and CD8 in human primary PCa tissue microarrays confirmed that samples with higher GSDMD expression exhibited a significantly greater abundance of CD8⁺ T cells (figure 8F). Collectively, these clinical findings support a functional role for GSDMD in enhancing intratumoral CTL infiltration in PCa.

Discussion

This study was motivated by the urgent need to understand the immunosuppressive nature of advanced PCa and to identify actionable targets for reprogramming the TME. Our results reveal the role of GSDMD as a significant tumor suppressor and immunomodulator in PCa. Low GSDMD expression is correlated with poor prognosis in patients with PCa, with its expression exerting a potent inhibitory effect on PCa progression in both genetically engineered mouse models (GEMs) and syngeneic tumor models. We identify GSDMD as a regulator of the TME, promoting CTL infiltration and enhancing CTL-mediated killing of PCa cells. This newly discovered role suggests that GSDMD loss or inhibition is a critical factor contributing to the immunosuppressive TME characteristic of advanced PCa, suggesting that dysregulated tumor-intrinsic pathways influence the immune system and determine resistance to immunotherapy.^{45 46}

Our mechanistic investigation revealed a new signaling axis (figure 8G, created with BioGDP.com) in which the loss of GSDMD in tumor cells leads to the accumulation and increased secretion of sLAMC2. This sLAMC2 acts as a pivotal paracrine signal, binding to the integrin α6β4 receptor on macrophages and activating the downstream AKT pathway. This cascade drives macrophages toward an M2-polarized state. A key functional consequence of this polarization was the significant upregulation of CXCL5 secretion from M2 macrophages, thereby establishing a feed-forward loop that reinforced an immunosuppressive and pro-angiogenic milieu. However, we acknowledged that the precise molecular link between AKT activation and the transcriptional induction of CXCL5 in macrophages—such as the definitive identification of the responsible transcription factors and their direct binding to the CXCL5 promoter—remained to be fully mapped. Future studies employing techniques like chromatin immunoprecipitation sequencing (ChIP-seq) in macrophages stimulated by sLAMC2 are needed to complete this signaling circuitry.

The apparent complexity of GSDMD’s role across models was resolved by appreciating its cell-type-specific function. In our pSKO versus pDKO models, the near-complete knockout of GSDMD in prostate epithelial cells (the origin of the tumors) was the driver of accelerated growth. A critical, yet not fully resolved, layer of complexity involves the cellular source of active GSDMD. While our data confirm the loss of full-length GSDMD in the tumor epithelium of pDKO mice, the active GSDMD-NT fragment detected within established tumors originates from infiltrating immune cells that retain a wild-type Gsdmd allele. A significant limitation of our current study was the lack of a comprehensive, single-cell resolution map of GSDMD-NT expression across specific immune subsets (eg, macrophages, neutrophils, dendritic cells) within the pSKO and pDKO TME. Such an analysis would precisely identify which myeloid or lymphoid populations contribute the functional, cleaved GSDMD that shapes the microenvironment. This gap limited our ability to fully deconvolute whether the protumor effects associated with total CD45⁺ GSDMD-NT and the antitumor function of macrophage-intrinsic GSDMD originate from distinct or overlapping cellular sources.

Furthermore, we uncovered a parallel tumor cell-intrinsic pathway where GSDMD loss activates the CXCL5/CXCR2 axis, leading to AKT/NF-κB-mediated upregulation of VEGF. This pathway provides a direct link between GSDMD’s tumor-suppressive function and the regulation of angiogenesis, another key process contributing to an immunosuppressive TME.

A pivotal discovery was that GSDMD promotes the ubiquitination and degradation of LAMC2. However, as GSDMD itself lacked intrinsic E3 ubiquitin ligase activity, the mechanism by which it facilitates this post-translational modification presents an intriguing question. Our preliminary in silico molecular docking analysis offered a testable hypothesis: RNF121 and RNF148 emerged as the top candidate E3 ligases displaying high binding affinity for both GSDMD and LAMC2. This computational prediction suggests that GSDMD may serve as a scaffold, potentially via its PUB domain, to recruit specific E3 ligases such as RNF121/RNF148 to LAMC2, thereby targeting it for proteasomal degradation. This hypothesis, while awaiting experimental validation (eg, Co-IP and functional rescue assays), provides a clear and specific downstream target to elucidate the complete ubiquitination machinery in this novel regulatory axis. The broader specificity of this regulation—whether GSDMD similarly governed the stability of other ECM proteins via analogous partnerships—remains an open and compelling question for future research.

The potential existence of the GSDMD-LAMC2 axis in other solid tumors was supported by the conserved function of its components. LAMC2 OE promoted invasion and metastasis in various cancers, including pancreatic, gastric, and lung cancers. Additionally, the regulation of macrophage polarization and CD8⁺ T-cell infiltration was a universal determinant of tumor progression. Thus, this pathway could plausibly function in other epithelial malignancies where basement membrane remodeling and immune suppression were prominent, such as gastrointestinal cancers or squamous cell carcinomas. However, tissue-specific environments (including alternative upstream activators of GSDMD or differences in tumor stroma) may modulate its activity. Future studies across multiple cancer types are warranted to validate its broader relevance.

The translational relevance of our findings is strongly supported by clinical data. Analyses of human PCa cohorts reveal that high tumor GSDMD expression is associated with increased intratumoral CTL infiltration. Moreover, in patients receiving ICB, high GSDMD expression correlates with significantly better clinical outcomes. This clinical validation corroborates our preclinical results and highlights GSDMD as a promising predictive biomarker for ICB response in PCa.

Despite these compelling findings, our study has additional limitations. While we used multiple models, certain validation cohorts were limited in size, warranting confirmation in larger, independent populations. Our mechanistic focused on the LAMC2/macrophage and CXCL5/CXCR2 pathways likely does not exhaust the means by which GSDMD shapes the TME; its pleiotropic effects suggest broader regulatory networks awaiting discovery.

In conclusion, the integrated preclinical and clinical evidence presented here provides a compelling rationale for targeting the GSDMD pathway in PCa. GSDMD’s dual capacity to directly restrain tumor progression and actively reshape the TME towards an immunostimulatory state—characterized by enhanced CTL activity, restrained M2 macrophage polarization, and improved response to immunotherapy—makes it an exceptionally attractive therapeutic node. The discovery of the GSDMD–LAMC2–CXCL5 axis opens direct avenues for intervention. Strategies aimed at stabilizing GSDMD function or mimicking its consequences (eg, promoting LAMC2 degradation via targeting the putative GSDMD-RNF121/148 complex or disrupting the sLAMC2-integrin interaction) could potentially reverse the immunosuppressive TME in GSDMD-low tumors. Future efforts must first address the noted cellular and mechanistic specifics—such as defining the immune source of GSDMD-N, completing the CXCL5 transcriptional cascade, and functionally validating the involvement of RNF121/RNF148—to enable precise therapeutic targeting. Subsequently, developing specific modulators of this pathway and evaluating their efficacy, both alone and in rational combination with immunotherapies, will be crucial to overcome the immunosuppressive barrier in advanced PCa.

Materials and methods

Mice

Animals were housed under pathogen-free conditions and treated in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Surgeries were performed after anesthesia with isoflurane. NOD/ShiLtJGpt-Prkdcem26Cd52Il2rgem26Cd22/Gpt (NCG) mice, a severely immunodeficient model, were supplied by GemPharmatech (Nanjing, China).

PB-Cre4 Pten^L/L mice⁴⁷ were bred with other mice carrying the Gsdmd allele, which were already congenic to C57BL/6, and then males with the homozygous status of Pten and Gsdmd, and hemizygous status of PB-Cre4+ were obtained. Lyz2-iCre mice (GemPharmatech) were bred with other mice carrying the Gsdmd allele, which were already congenic to C57BL/6, and then males with the homozygous status of Gsdmd. Mice with spontaneous prostate tumors were euthanized at designated time points for tumor collection. Signs of lethargy, reduced mobility, and morbidity due to the internal status of the tumors rather than the maximum tumor size, were used as factors in deciding the day of sacrifice.

Animal experiments

Pyroptosis experiment was performed by the injection of 1×10⁵ RM-1 sublines into the dorsal subcutaneous layer of C57BL/6 male mice. Mice were treated with a daily intraperitoneal injection of disulfiram (Ambeed, A912727), necrosulfonamide (Ambeed, A750842) or LDC7559 (Ambeed, A1263913) at a dose of 10 mg/kg each compound.

A total of 1×10⁵ tumor cells were subcutaneously injected into C57BJ/6 male mice to perform SX-682 treatment experiments; thus, SX-682-medicated drinking water (S414287, Aladdin) (equivalent to 200 mg/kg mouse body weight per day) was provided until the survival end point.

In the T-cell recruitment assay, tumor-bearing NCG mice are infused with 1×10⁷ B3Z cells, and the percentage of CD8⁺ T cells in tumor tissues was measured by flow cytometry after 18 hours.

A total of 1×10⁵ tumor cells were subcutaneously injected into the flanks of C57BL/6 male mice to establish the tumor model for ICB therapy. Tumor growth was monitored until the average volume reached 100–150 mm³, as measured with calipers using the formula: volume=length×width²×0.5. Once tumors attained this uniform starting size, mice were randomly allocated into treatment groups. Mice received intraperitoneal injections of anti-programmed cell death protein 1 (114116, BioLegend) and anti-cytotoxic T-lymphocyte associated protein 4 (CTLA-4) (106207, BioLegend) each at 10 mg/kg, two times per week, or an equivalent dose of isotype IgG control. Treatment was continued until the end of the survival observation period. ICB therapy was initiated simultaneously across all groups to ensure that any differences in therapeutic response were not confounded by variations in initial tumor burden.

As regards the macrophage depletion model, macrophages in the blood circulation of mice were depleted by an intravenous injection of chlorpropamide liposomes (200 µL every 3 days, 40337ES08, Yeasen Biotechnology) according to the manufacturer’s instructions. The control group was treated with liposomes without chlorpropamide to assess and rule out their effects. The efficiency of macrophage clearance was evaluated using flow cytometry.

Cell culture

PCa cell lines RM-1 (RRID: CVCL_B459) and PC-3 (RRID: CVCL_0035), mouse T-cell hybridoma cell line B3Z (RRID: CVCL_6227), HUVECs (RRID: CVCL_9Q53) and mouse mononuclear macrophages cells (RAW264.7, RRID: CVCL_0493), were purchased from BOSTER in January 2022, BDBIO in March 2023, BLUEFBIO in April 2024, IMMOCELL in October 2024, and BDBIO in December 2024, respectively. All cells were identity were authenticated by short tandem repeat profiling. These cells are widely used in related fields. Cells were cultured in a medium containing 10% fetal bovine serum (FBS, 16140071, Invitrogen) and 1% penicillin. PCa cell lines were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium (C22400500BT, Gibco), B3Z cells and RAW264.7 cells in Dulbecco’s Modified Eagle Medium (DMEM, 11995065, Gibco), and HUVECs in endothelial cell medium (ECM, Cat#1001, ScienCell). Cells were incubated in a humidified environment at 37°C and 5% CO₂. Cells were divided into subcultures when the cell confluence exceeded 90%.

Isolation of bone marrow-derived macrophages and in vitro analysis of macrophage polarization

6–8 weeks old C57BL/6 mice were euthanized, and the tibiae and femurs were aseptically harvested. After removing surrounding muscle tissue, the epiphyses of both bones were excised using sterile scissors. The bones were then placed in phosphate-buffered saline (PBS) to maintain sterility. Bone marrow was flushed from the medullary cavity using a syringe filled with PBS.

The collected bone marrow cells were centrifuged at 1,500 rpm for 5 min, followed by red blood cell lysis using ACK lysis buffer. After filtration and another centrifugation step, the cells were resuspended in DMEM supplemented with 10% FBS and 100 ng/mL macrophage colony-stimulating factor (PRP1136, Abbkine). The cell suspension was then plated in culture dishes for further differentiation. Then, cell polarization into M1 or M2 was performed by stimulation for 24 hours with IFN-γ (20 ng/mL, PRP1138, Abbkine) plus LPS (100 ng/mL, HY-D1056, MCE) or IL-4 (20 ng/mL, 214–4, PeproTech) plus IL-13 (20 ng/mL, PRP1051, Abbkine), respectively. RNA was extracted from cells and RT-qPCR was performed to measure the relative gene expression.

The sample size for the animal experiments was determined based on the variability observed in preliminary studies, which represented a part of the approval process for the Institutional Animal Care and Use Committee (IACUC) protocol. Additionally, criteria for halting data collection were established as a component of this approval process. Mice that died because of causes unrelated to tumor burden were excluded from the analysis. Once the tumor was developed and detected, animals were randomly divided into treatment and control groups. The initial tumor burden in both groups was comparable prior to treatment initiation. Blinding during these experiments was not consistently feasible, as the principal investigator had knowledge of mouse identification. Since this study focused specifically on PCa, only male mice were used throughout the research.

H&E, immunohistochemistry and immunofluorescence

H&E staining was performed on fresh tissues, which were fixed in a 4% paraformaldehyde before being embedded in paraffin. Then, 5 μm thick sections were cut and placed on SuperFrost Plus glass slides, deparaffinized using xylene and rehydrated with decreasing concentrations of ethanol. The slides were then stained with H&E, dehydrated using an anti-gradient ethanol-xylene mixture, mounted in neutral resin, and finally scanned using a Pannoramic MIDI slide scanner (MF53, Mingmei).

Animal tissues used for IHC were fixed overnight in 4% paraformaldehyde, embedded in paraffin and cut into 5 μm thick sections. Sections underwent deparaffinization in xylene, gradient ethanol rehydration and triple washing with distilled water. Sections were treated with 3% H₂O₂-methanol for 30 min to block endogenous catalase activity. Sections were treated with 0.01 M sodium citrate buffer (pH 6.0) (P019IH, Auragene) in a microwave oven at 95°C for 10 min for antigen retrieval, followed by a 30 min cooling period. The following primary antibodies were added, and the sections were incubated overnight at 4°C: Ki-67 (1:500 dilution, ab16667, Abcam), CXCL5 (1:100 dilution, DF9919, Affinity), CD31 (1:100 dilution, 77699T, Cell Signaling Technology), GSDMD (1:50 dilution, 39754T, Cell Signaling Technology), CD8 (1:100 dilution, 100701, BioLegend), LAMC2 (1:100 dilution, HA722815, HUABIO), FAK (1:100 dilution, 48643, SAB), ITGB4 (1:200 dilution, A-01015-2, Boster), ITGA6 (1:100 dilution, DF8323, Affinity). Signals were developed using a DAB staining kit (DA1015, Solarbio) for 5 min, followed by counterstaining with hematoxylin for 5 s. Next, each section was rinsed with 1×PBS, sealed, and examined under a microscope (MF53, Mingmei). ImageJ software (V.1.5.1, NIH, Bethesda) was used to quantify the number of positive cells.

Immunofluorescence staining was performed by tyramide signal amplification (TSA) technology (RK05903, ABclonal) with ‌horseradish peroxidase (HRP) -labeled secondary antibodies. HRP catalyzes the TSA-derived fluorescent dyes added to the system, generating activated fluorescent substrates. The activated substrates covalently bound to the tyrosine on the antigen, covalently attaching the signal to the antigen. Subsequently, the non-covalently bound antibodies were removed by heat repair. Then, a different primary antibody was used for the second round of incubation, along with another fluorescent substrate. Such repetition enabled multiple labeling. Specific steps were performed following the manufacturer’s instructions of the reagent.

Establishment of mouse prostate tumor cell lines

Tumors were dissected from prostates of Pten^L/LPB-Cre4+ and Gsdmd^L/L Pten^L/LPB-Cre4+ mice, minced, digested using 0.5% type I collagenase (17100017, Invitrogen), filtered through a 70 µm net, and the captured fragments were treated with red blood cell lysis buffer. Cells with typical epithelial morphology were collected. The independent cell lines were established and maintained in RPMI 1640 to complete the subsequent relevant experiments.

RNA-sequencing analysis

Tissues from three independent mice (pSKO and pDKO) were subjected to RNA-seq. All sequencing procedures and analyses were performed by Shanghai OE Biotech (Shanghai, China) after passing quality inspection. Detailed protocols are provided in the online supplemental materials.

RNA isolation and real-time quantitative PCR

Total cellular and tissue RNAs were extracted using FastPure Cell/Tissue Total RNA isolation Kit V2 (RC112-01, Vazyme). Subsequently, RNA quantification was performed using NanoDrop One (13–400-519, Thermo Fisher Scientific, Waltham, USA). 1 µg total RNA was reverse transcribed into complementary DNA in a 20 µL reaction volume using the 2×Taq Pro Universal SYBR qPCR Master Mix (Q712-02, Vazyme). RT-qPCR was performed using 2×ChanQ SYBR QPCR Green (Q311-02, Vazyme) on a LightCycler 480 instrument (LightCycler 480, Roche Diagnostics) as instructed by the manufacturer. The relative expression of each gene was determined using the 2^−∆∆Ct method, with β-actin as the endogenous control. Primers used in RT-qPCR are listed in table 1.

Table 1. List of primers used for real-time quantitative PCR.

Gene name	Forward	Reverse
Arg-1	CTCCAAGCCAAAGTCCTTAGAG	GGAGCTGTCATTAGGGACATCA
IL-10	GGCAGTGGAGCAGGTGAAGAATG	TGTCACGTAGGCTTCTATGCAGTTG
CD206	TTGCACTTTGAGGGAAGCGA	CCTTGCCTGATGCCAGGTTA
CXCL13	GGCCACGGTATTCTGGAAGC	GGGCGTAACTTGAATCCGATCTA
TIGIT	GAATGGAACCTGAGGAGTCTCT	AGCAATGAAGCTCTCTAGGCT
CTLA4	TTTTGTAGCCCTGCTCACTCT	CTGAAGGTTGGGTCACCTGTA
TNFRSF9	CGTGCAGAACTCCTGTGATAAC	GTCCACCTATGCTGGAGAAGG
HAVCR2	TCAGGTCTTACCCTCAACTGTG	GGGCAGATAGGCATTTTTACCA
LAG3	CTGGGACTGCTTTGGGAAG	GGTTGATGTTGCCAGATAACCC
CST7	GGAGCTGTACTTGCCGAGC	CATGGGTGTCAGAAGTTAGGC
GZMK	TGGCTGGCGTTTATATGTCTTC	GCTGCGGTACTGGATGGAC
GZMA	TGCTGCCCACTGTAACGTG	GGTAGGTGAAGGATAGCCACAT
NKG7	TCAAGTCCAGACATTCTTCTCCT	CACAAGGTTTCATACTCAGCCC
IFNG	ATGAACGCTACACACTGCATC	CCATCCTTTTGCCAGTTCCTC
PRF1	AAAAACTCCCTAATGAGAGACGC	ACACGCCAGTCGTTATTGATATT
PD1	ACCCTGGTCATTCACTTGGG	CATTTGCTCCCTCTGACACTG
TCF7	AGCTTTCTCCACTCTACGAACA	AATCCAGAGAGATCGGGGGTC
TOX	GCTCCCGTTCCATCCACAAA	TCCCAATCTCTTGCATCACAGA
CXCL5	CTCAGTCATAGCCGCAACCGAGC	CGCTTCTTTCCACTGCGAGTGC
ITGB4	GCAGACGAAGTTCCGACAG	GGCCACCTTCAGTTCATGGA

Open in a new tab

Western blot and co-immunoprecipitation

Cells and fresh tissues were separately lysed on ice using radio immunoprecipitation assay (RIPA) buffer supplemented with ProtLytic protease and phosphatase inhibitor cocktail (NCM Biotech, P002). Protein concentration determination and immunoblotting were performed as previously described.⁴⁸ The following primary antibodies were used: GSDMD (1:1000, 39754, Cell Signaling Technology), CXCL5 (1:1000, DF9919, Affinity), CXCR2 (1:1000, 20634–1 AP, Proteintech), NF-κB (1:1000, 6596, Cell Signaling Technology), NF-κB p65 (Ser536) (1:1000, 3033, Cell Signaling Technology), AKT (1:1000, ET1609-51, HUABIO), phospho-AKT (S473) (1:5000, ET1607-73, HUABIO), VEGF (1:1000, ET1604-28, HUABIO), VEGFR2 (1:1000, RT1650, HUABIO), PD-L1 (1:2000, 66248–1 lg, Proteintech), CD31 (1:1000, 77699S, Cell Signaling Technology), LAMC2 (1:1000, 19698–1 AP, Proteintech), GAPDH (1:5000, 60004–1 Ig, Proteintech), Lamin B1 (1:10000, 66095–1 Ig, Proteintech), FAK (1:1000, SAB, 48643), phospho-FAK (Tyr397) (1:1000, MCE, HY-P80460), HA-Tag (1:1000, 3724, Cell Signaling Technology) and DYKDDDDK Tag (1:1000, 14793, Cell Signaling Technology). Co-IP used protein samples prepared and quantified as described for western blotting. Antigen-antibody complexes were captured using Protein A/G Magnetic Beads (HY-K0202, MCE) per the manufacturer’s instructions. Supernatants were then resolved by sodium dodecyl sulfate-polyAcrylamide gel electrophoresis (SDS-PAGE). The primary antibodies included GSDMD (1:100, 39754, Cell Signaling Technology), LAMC2 (1:100, 19698–1 AP, Proteintech), HA-Tag (1:50, 3724, Cell Signaling Technology), Myc-Tag (1:50, 2278, Cell Signaling Technology), anti-rabbit IgG (1:100, 7074P2, Cell Signaling Technology) and DYKDDDDK Tag (1:50, 14793, Cell Signaling Technology). The blots were incubated with secondary antibody (1:10000, 511103/511203, Zen-Bio).

Lentiviral transduction

The lentiviral vectors encoding GSDMD-specific shRNA, HBLV-GSDMD-3xflag, HBLV-LAMC2-3xHA, HBLV-OVAL-3xflag, and CXCL5-specific shRNA (containing either puromycin resistance or blasticidin resistance genes) were provided by Zebrafish Biotech and Hanheng Biotechnology (online supplemental figure S10A-E). At 60% confluency, cells were transduced with lentiviral particles mixed with polybrene (5 µg/mL) in 50% volume of complete medium. Following 4-hour incubation, medium volume was restored, with complete replacement after 24 hours. Green fluorescent protein (GFP) expression was assessed at 48 hours post-transduction. On reaching 70–80% confluency, puromycin selection was initiated to establish stable cell lines.

Flow cytometry

Tumors were minced into small pieces in RPMI containing 2% FBS, 1% penicillin-streptomycin, DNase I (1 µg/mL; Sigma-Aldrich) and collagenase (0.5 µg/mL; Sigma-Aldrich) and digested for 60 min at 37°C, followed by filtration with a 70 µm cell strainer. As regards surface marker staining, cells were incubated with the following antibodies in PBS containing 2% FBS plus 2 mM EDTA for 45 min on ice: anti-CD8 (E-AB-F1104J, Elabscience, 5 µL/test; 566096, BD, Biosciences, 1 µL/test), anti-CD4 (E-AB-F1353D, Elabscience, 5 µL/test), anti-CD3 (553061/570560, BD Biosciences, 1 µL/test), anti-CD45 (103116, BioLegend, 1 µL/test), anti-CD11b (E-AB-F1081C, Elabscience, 5 µL/test), anti-F4/80 (E-AB-F0995D, Elabscience, 5 µL/test), anti-CD206 (E-AB-F1135E, Elabscience, 5 µL/test), anti-Ly6C (E-AB-F1121E, Elabscience, 5 µL/test), anti-Ly6G (127639, BioLegend, 1 µL/test), anti-CTLA4 (12–1522-81, Thermo, 1 µL/test), anti-TIGIT (142107, BioLegend, 1 µL/test), anti-LAG3 (740072, BD Biosciences, 1 µL/test). Then, the cells were washed two times with cold PBS and analyzed on flow cytometry. For samples requiring staining of intracellular or nuclear factors, the procedure was consistently performed in the following sequence: cell stimulation (550583, BD Biosciences, 37°C 5 hours), staining with a viability dye (Fixable Viability Dye 777 (S0D0026, STARTER)), application of an Fc receptor blocking reagent, surface marker staining, fixation and permeabilization (562725, BD Biosciences), intracellular staining (anti-IFNG (417–7311-80, Thermo, 1 µL/test), anti-GZMA, Thermo, 1 µL/test), anti-TCF7 (566692, BD Biosciences, 1 µL/test), anti-TOX (12–6502-80, Thermo, 1 µL/test)), and finally resuspension for flow cytometry analysis. All flow cytometry data were collected from BD FACSAriaIII flow cytometry and were analyzed using FlowJo software.

T-cell proliferation assay and T-cell killing assay

As regards the antigen-dependent T-cell killing assay, OVA-overexpressing cancer cells were seeded at a density of 5,000 cells/well in 96-well plates. Once the cells were attached to the bottom of the well, B3Z T cells were added to cancer cells at the specified ratios. After 72 hours of co-culture, a gentle, low-concentration trypsin treatment (0.05% trypsin for 1–2 min) was applied to dissociate any adherent T cells prior to harvesting, ensuring accurate measurement of remaining adherent cancer cell viability using the resazurin assay (HY-111391, MCE).

As regards the T-cell proliferation assay, co-culture of B3Z cells with non-specific cancer cells was used to assess the immunosuppressive activity. B3Z cells were collected after 48 hours of co-culture, labeled with CFSE fluorescent dye (S8269, Selleck), and analyzed by flow cytometry.

RFdiffusion to predict interspecies protein–protein interactions

RFdiffusion creates a diverse set of candidate binders by simulating protein backbone structures, focusing on regions likely to interact with protein GSDMD.²⁶ Binders with less than two significant contacts at the specified hotspots were excluded. Next, the generated binders were selected using the Foldseek database of protein structures of the target species. Structural similarity was assessed using TM-align, and only matches with at least 90% structural coverage and a bit score above 50 were retained. This ensured the identification of high-confidence natural proteins while minimizing false positives. The candidates identified by Foldseek were further evaluated using HDOCK to calculate their binding energies with the GSDMD protein.⁴⁹ Proteins with favorable binding energies were selected for the next step. AlphaFold was then used to model the complexes formed by the GSDMD protein and the selected candidates.⁵⁰ The structural confidence of these complexes was analyzed to ensure interaction stability. Finally, the most promising proteins were selected based on their structural similarity, binding energy and AlphaFold confidence score.

Cell migration assay

RAW264.7 cells (2×10⁵ cells in 400 µL serum-free DMEM) were plated in the upper chamber of transwell inserts (LABSELECT, 14242). The inserts were then transferred to 12-well plates containing conditioned medium from either RM-1 NC (negative control) or OE-GSDMD cells as chemoattractant, followed by 48-hour incubation at 37°C. After fixation, migrated cells were stained with 0.1% crystal violet solution. Non-migrated cells on the upper membrane surface were removed by gentle swabbing, while migrated cells adherent to the lower membrane surface were quantified by microscopic imaging and manual counting.

ELISA assay of tumor sites and cell culture supernatant

Tumor lysates and cell culture supernatant were collected and cytokine levels in the tumor after different treatments were analyzed using a mouse ELISA kit according to the manufacturer’s instructions. The specific cytokines were the following: IL-10 (EK0417, BOSTER), CXCL5 (EK0919, BOSTER), TNF-α (EK0527, BOSTER), IL-1β (EK0934, BOSTER), and IL-18 (EK0433, BOSTER).

Tube formation assay

Briefly, 60 µL Matrigel were added to a 96-well plate, which was incubated at 37°C for 45 min. Subsequently, 2×10⁴ HUVECs in 100 µL of specific supernatant were added to the Matrigel-coated wells and incubated for 12 hours. Next, HUVECs were cultured in conditioned supernatant and allowed to undergo tube formation. The resulting vascular network structures were visualized and imaged using a microscopy. Quantitative analysis of tubular networks was performed by measuring the number of meshes and branching points using ImageJ software.

In vitro permeability assay

HUVECs were cultured on 0.4 µm pore transwell inserts (353095, Falcon) and treated with either PP CM or GPP CM for 72 hours. Following treatment, FITC-dextran (1 mg/mL, 61220ES, Yeasen) was introduced to the apical chamber and allowed to diffuse for 45 min. The fluorescence intensity of translocated FITC-dextran in the basolateral chamber was quantified at 590 nm using a Varioskan LUX microplate reader (Thermo Scientific).

Cell transfection

Cells were seeded into a 12-well plate at a density of 2×10⁵ cells in complete culture medium 24 hours before transfection. Cell transfection was carried out employing Lipofectamine 3000 Transfection Reagent (L3000001, Thermo) according to the manufacturer’s instructions. Then, the cells were incubated at 37°C under 5% CO₂ for 48 hours to allow gene expression. RAW264.7 cells were pre-transfected with 1x NATE (Lyec-nate, Invivogen) working fluid under shaking to ensure the uniform distribution of cells and incubated under normal cell culture conditions for at least 30 min to increase the transfection efficiency. The relevant plasmids involved in the transfection are shown in the online supplemental material Table S1-6.

Statistical analysis

Statistical analysis was performed using GraphPad Prism software (V.8.0, GraphPad Software, San Diego, California, USA) and Microsoft Excel (V.2020, Microsoft Corporation, Redmond, Washington, USA). Each experiment was repeated at least three independent times to ensure reproducibility. Student’s t-test was used to compare two groups. A value of p<0.05 was considered statistically significant.

Supplementary material

online supplemental file 1

jitc-14-5-s001.docx^{(1.8MB, docx)}

DOI: 10.1136/jitc-2025-014533

Footnotes

Funding: This research was supported in part by grants from the National Natural ScienceFoundation of China: 81902857 (XYW).

Provenance and peer review: Not commissioned; externally peer reviewed.

Patient consent for publication: Not applicable.

Ethics approval: All human prostate cancer tissue cores were approved by the Bioethics Review Committee of Hunan Aifang Biotechnology Co., Ltd. The approval ID is HN20250401. Participants gave informed consent to participate in the study before taking part.

Data availability statement

Data are available upon reasonable request.

References

1.He YD, Xu WD, Xiao YT, et al. Targeting signaling pathways in prostate cancer: mechanisms and clinical trials. Sig Transduct Target Ther . 2022;7 doi: 10.1038/s41392-022-01042-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Cheng Y, Liu B, Xin J, et al. Author Correction: Single-cell and spatial RNA sequencing identify divergent microenvironments and progression signatures in early- versus late-onset prostate cancer. Nat Aging . 2025;5:1370–1. doi: 10.1038/s43587-025-00892-4. [DOI] [PubMed] [Google Scholar]
3.Hawley JE, Obradovic AZ, Dallos MC, et al. Anti-PD-1 immunotherapy with androgen deprivation therapy induces robust immune infiltration in metastatic castration-sensitive prostate cancer. Cancer Cell. 2023;41 doi: 10.1016/j.ccell.2023.10.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Chen S, Zhu G, Yang Y, et al. Single-cell analysis reveals transcriptomic remodellings in distinct cell types that contribute to human prostate cancer progression. Nat Cell Biol. 2021;23:87–98. doi: 10.1038/s41556-020-00613-6. [DOI] [PubMed] [Google Scholar]
5.Hegde PS, Chen DS. Top 10 Challenges in Cancer Immunotherapy. Immunity. 2020;52:17–35. doi: 10.1016/j.immuni.2019.12.011. [DOI] [PubMed] [Google Scholar]
6.Beer TM, Kwon ED, Drake CG, et al. Randomized, Double-Blind, Phase III Trial of Ipilimumab Versus Placebo in Asymptomatic or Minimally Symptomatic Patients With Metastatic Chemotherapy-Naive Castration-Resistant Prostate Cancer. J Clin Oncol. 2017;35:40–7. doi: 10.1200/JCO.2016.69.1584. [DOI] [PubMed] [Google Scholar]
7.Hage Chehade C, Ozay ZI, Ostrowski M, et al. T-cell Engagers in Prostate Cancer. Eur Urol. 2025;87:553–8. doi: 10.1016/j.eururo.2025.02.001. [DOI] [PubMed] [Google Scholar]
8.Sutherland TE, Dyer DP, Allen JE. The extracellular matrix and the immune system: A mutually dependent relationship. Science. 2023;379:eabp8964. doi: 10.1126/science.abp8964. [DOI] [PubMed] [Google Scholar]
9.Hu Q, Zhu YF, Mei J, et al. Extracellular matrix dynamics in tumor immunoregulation: from tumor microenvironment to immunotherapy. J Hematol Oncol . 2025;18 doi: 10.1186/s13045-025-01717-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Chen M, Zhao S, Xie X, et al. Role of Tunneling Nanotubes in Arachidonic Acid Transfer and Macrophage Function Reprogramming in Intrahepatic Cholangiocarcinoma. Adv Sci (Weinh) 2025;12 doi: 10.1002/advs.202500148. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Xiang W, Shi R, Kang X, et al. Monoacylglycerol lipase regulates cannabinoid receptor 2-dependent macrophage activation and cancer progression. Nat Commun. 2018;9:2574. doi: 10.1038/s41467-018-04999-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Pritykin Y, van der Veeken J, Pine AR, et al. A unified atlas of CD8 T cell dysfunctional states in cancer and infection. Mol Cell. 2021;81:2477–93. doi: 10.1016/j.molcel.2021.03.045. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Li Y, Shen Z, Chai Z, et al. Targeting MS4A4A on tumour-associated macrophages restores CD8+ T-cell-mediated antitumour immunity. Gut. 2023;72:2307–20. doi: 10.1136/gutjnl-2022-329147. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Yamaguchi Y, Gibson J, Ou K, et al. PD-L1 blockade restores CAR T cell activity through IFN-γ-regulation of CD163+ M2 macrophages. J Immunother Cancer. 2022;10:e004400. doi: 10.1136/jitc-2021-004400. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Wang S, Wang J, Chen Z, et al. Targeting M2-like tumor-associated macrophages is a potential therapeutic approach to overcome antitumor drug resistance. NPJ Precis Oncol . 2024;8:31. doi: 10.1038/s41698-024-00522-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Lv TT, Xiong XF, Yan W, et al. Targeting of GSDMD sensitizes HCC to anti-PD-1 by activating cGAS pathway and downregulating PD-L1 expression. J Immunother Cancer. 2022;10:e004763. doi: 10.1136/jitc-2022-004763. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Ning H, Huang S, Lei Y, et al. Enhancer decommissioning by MLL4 ablation elicits dsRNA-interferon signaling and GSDMD-mediated pyroptosis to potentiate anti-tumor immunity. Nat Commun. 2022;13 doi: 10.1038/s41467-022-34253-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Gao K, Xi W, Ni J, et al. Genetically modified extracellular vesicles loaded with activated gasdermin D potentially inhibit prostate-specific membrane antigen-positive prostate carcinoma growth and enhance immunotherapy. Biomaterials. 2025;315 doi: 10.1016/j.biomaterials.2024.122894. [DOI] [PubMed] [Google Scholar]
19.Fontana P, Du G, Zhang Y, et al. Small-molecule GSDMD agonism in tumors stimulates antitumor immunity without toxicity. Cell. 2024;187:6165–81. doi: 10.1016/j.cell.2024.08.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Li M, Yang D, Yan H, et al. Gasdermin D maintains bone mass by rewiring the endo-lysosomal pathway of osteoclastic bone resorption. Dev Cell. 2022;57:2365–80. doi: 10.1016/j.devcel.2022.09.013. [DOI] [PubMed] [Google Scholar]
21.Fan X, Han J, Zhong L, et al. Macrophage-Derived GSDMD Plays an Essential Role in Atherosclerosis and Cross Talk Between Macrophages via the Mitochondria-STING-IRF3/NF-κB Axis. ATVB . 2024;44:1365–78. doi: 10.1161/ATVBAHA.123.320612. [DOI] [PubMed] [Google Scholar]
22.Bradfield CJ, Liang JJ, Ernst O, et al. Biphasic JNK signaling reveals distinct MAP3K complexes licensing inflammasome formation and pyroptosis. J Immunol. 2023;210:167. doi: 10.4049/jimmunol.210.Supp.167.01. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Ho L-C, Chen Y-H, Wu T-Y, et al. Phosphate burden induces vascular calcification through a NLRP3-caspase-1-mediated pyroptotic pathway. Life Sci. 2023;332:122123. doi: 10.1016/j.lfs.2023.122123. [DOI] [PubMed] [Google Scholar]
24.Rathkey JK, Zhao J, Liu Z, et al. Chemical disruption of the pyroptotic pore-forming protein gasdermin D inhibits inflammatory cell death and sepsis. Sci Immunol. 2018;3:eaat2738. doi: 10.1126/sciimmunol.aat2738. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Hu JJ, Liu X, Xia S, et al. FDA-approved disulfiram inhibits pyroptosis by blocking gasdermin D pore formation. Nat Immunol. 2020;21:736–45. doi: 10.1038/s41590-020-0669-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Watson JL, Juergens D, Bennett NR, et al. De novo design of protein structure and function with RFdiffusion. Nature New Biol. 2023;620:1089–100. doi: 10.1038/s41586-023-06415-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Zhao S, Luo J, Xu P, et al. Designed peptide binders and nanobodies as PROTAC starting points for targeted degradation of PCNA and BCL6. Int J Biol Macromol. 2025;308:142667. doi: 10.1016/j.ijbiomac.2025.142667. [DOI] [PubMed] [Google Scholar]
28.Liu M, Cai R, Wang T, et al. LAMC2 promotes the proliferation of cancer cells and induce infiltration of macrophages in non-small cell lung cancer. Ann Transl Med. 2021;9:1392. doi: 10.21037/atm-21-4507. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Nguyen NM, Pulkkinen L, Schlueter JA, et al. Lung development in laminin γ2 deficiency: abnormal tracheal hemidesmosomes with normal branching morphogenesis and epithelial differentiation. Respir Res . 2006;7 doi: 10.1186/1465-9921-7-28. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Fu R, Zhang C, Song M-M, et al. A single-cell map of patients with non-small cell lung cancer harboring rare-driver mutations after anti-PD-1 treatment. Cancer Lett. 2025;616:217595. doi: 10.1016/j.canlet.2025.217595. [DOI] [PubMed] [Google Scholar]
31.Lu C, Gao Z, Wu D, et al. EP.12E.04 Understanding the Dynamics of TKI Induced Changes in the Tumor Immune Microenvironment for Improved Therapeutic Effect. J Thorac Oncol. 2024;19:S654. doi: 10.1016/j.jtho.2024.09.1235. [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Griffin GK, Newton G, Tarrio ML, et al. IL-17 and TNF-α sustain neutrophil recruitment during inflammation through synergistic effects on endothelial activation. J Immunol. 2012;188:6287–99. doi: 10.4049/jimmunol.1200385. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Liu Y, Mei J, Gonzales LW, et al. IL-17a and tnf-± exert synergistic effects on expression of cxcl5 by alveolar type ii cells in vivo and in vitro. Am J Resp Crit Care. 2010;181 doi: 10.1164/ajrccm-conference.2010.181.1_MeetingAbstracts.A6787. [DOI] [PubMed] [Google Scholar]
34.Zhang L, Lin Y, Hu L, et al. Transient intracellular expression of PD-L1 and VEGFR2 bispecific nanobody in cancer cells inspires long-term T cell activation and infiltration to combat tumor and inhibit cancer metastasis. Mol Cancer. 2025;24 doi: 10.1186/s12943-025-02253-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Nair S, Boczkowski D, Moeller B, et al. Synergy between tumor immunotherapy and antiangiogenic therapy. Blood. 2003;102:964–71. doi: 10.1182/blood-2002-12-3738. [DOI] [PubMed] [Google Scholar]
36.Tada Y, Togashi Y, Kotani D, et al. Targeting VEGFR2 with Ramucirumab strongly impacts effector/ activated regulatory T cells and CD8 T cells in the tumor microenvironment. J Immunother Cancer. 2018 doi: 10.1186/s40425-018-0403-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Bhattacharya R, Ye X-C, Wang R, et al. Intracrine VEGF Signaling Mediates the Activity of Prosurvival Pathways in Human Colorectal Cancer Cells. Cancer Res. 2016;76:3014–24. doi: 10.1158/0008-5472.CAN-15-1605. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Karakhanova S, Link J, Heinrich M, et al. Characterization of myeloid leukocytes and soluble mediators in pancreatic cancer: importance of myeloid-derived suppressor cells. Oncoimmunology. 2015;4:e998519. doi: 10.1080/2162402X.2014.998519. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Wang G, Lu X, Dey P, et al. Targeting YAP-Dependent MDSC Infiltration Impairs Tumor Progression. Cancer Discov. 2016;6:80–95. doi: 10.1158/2159-8290.CD-15-0224. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Ma S, Cheng Q, Cai Y, et al. IL-17A Produced by γδ T Cells Promotes Tumor Growth in Hepatocellular Carcinoma. Cancer Res. 2014;74:1969–82. doi: 10.1158/0008-5472.CAN-13-2534. [DOI] [PubMed] [Google Scholar]
41.Wang M, Han Y, Yao X, et al. Hyperexpression of tumor necrosis factor receptor 2 inhibits differentiation of myeloid‐derived suppressor cells by instigating apolarity during ageing. MedComm . 2024;5 doi: 10.1002/mco2.605. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Liu X, Tang R, Xu J, et al. CRIP1 fosters MDSC trafficking and resets tumour microenvironment via facilitating NF-κB/p65 nuclear translocation in pancreatic ductal adenocarcinoma. Gut. 2023;72:2329–43. doi: 10.1136/gutjnl-2022-329349. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Gabrilovich DI, Ostrand-Rosenberg S, Bronte V. Coordinated regulation of myeloid cells by tumours. Nat Rev Immunol. 2012;12:253–68. doi: 10.1038/nri3175. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Lu X, Horner JW, Paul E, et al. Effective combinatorial immunotherapy for castration-resistant prostate cancer. Nature New Biol. 2017;543:728–32. doi: 10.1038/nature21676. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.de Visser KE, Joyce JA. The evolving tumor microenvironment: From cancer initiation to metastatic outgrowth. Cancer Cell. 2023;41:374–403. doi: 10.1016/j.ccell.2023.02.016. [DOI] [PubMed] [Google Scholar]
46.Xin Z, Lin M, Hao Z, et al. The immune landscape of human thymic epithelial tumors. Nat Commun. 2022;13 doi: 10.1038/s41467-022-33170-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Lesche R, Groszer M, Gao J, et al. Cre/loxP-mediated inactivation of the murine Pten tumor suppressor gene. Genesis. 2002;32:148–9. doi: 10.1002/gene.10036. [DOI] [PubMed] [Google Scholar]
48.Wang X, Du Q, Chen J, et al. Fatty acid dehydratase HACD3 poses protein kinase activity and promotes the malignant progression of colorectal cancer. Int J Biol Macromol. 2024;283:137414. doi: 10.1016/j.ijbiomac.2024.137414. [DOI] [PubMed] [Google Scholar]
49.Yan YM, Tao HY, He JH, et al. The HDOCK server for integrated protein–protein docking. Nat Protoc. 2020;15:1829–52. doi: 10.1038/s41596-020-0312-x. [DOI] [PubMed] [Google Scholar]
50.Abramson J, Adler J, Dunger J, et al. Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature New Biol. 2024;630:493–500. doi: 10.1038/s41586-024-07487-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

online supplemental file 1

jitc-14-5-s001.docx^{(1.8MB, docx)}

DOI: 10.1136/jitc-2025-014533

Data Availability Statement

Data are available upon reasonable request.

[R1] 1.He YD, Xu WD, Xiao YT, et al. Targeting signaling pathways in prostate cancer: mechanisms and clinical trials. Sig Transduct Target Ther . 2022;7 doi: 10.1038/s41392-022-01042-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R2] 2.Cheng Y, Liu B, Xin J, et al. Author Correction: Single-cell and spatial RNA sequencing identify divergent microenvironments and progression signatures in early- versus late-onset prostate cancer. Nat Aging . 2025;5:1370–1. doi: 10.1038/s43587-025-00892-4. [DOI] [PubMed] [Google Scholar]

[R3] 3.Hawley JE, Obradovic AZ, Dallos MC, et al. Anti-PD-1 immunotherapy with androgen deprivation therapy induces robust immune infiltration in metastatic castration-sensitive prostate cancer. Cancer Cell. 2023;41 doi: 10.1016/j.ccell.2023.10.006. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R4] 4.Chen S, Zhu G, Yang Y, et al. Single-cell analysis reveals transcriptomic remodellings in distinct cell types that contribute to human prostate cancer progression. Nat Cell Biol. 2021;23:87–98. doi: 10.1038/s41556-020-00613-6. [DOI] [PubMed] [Google Scholar]

[R5] 5.Hegde PS, Chen DS. Top 10 Challenges in Cancer Immunotherapy. Immunity. 2020;52:17–35. doi: 10.1016/j.immuni.2019.12.011. [DOI] [PubMed] [Google Scholar]

[R6] 6.Beer TM, Kwon ED, Drake CG, et al. Randomized, Double-Blind, Phase III Trial of Ipilimumab Versus Placebo in Asymptomatic or Minimally Symptomatic Patients With Metastatic Chemotherapy-Naive Castration-Resistant Prostate Cancer. J Clin Oncol. 2017;35:40–7. doi: 10.1200/JCO.2016.69.1584. [DOI] [PubMed] [Google Scholar]

[R7] 7.Hage Chehade C, Ozay ZI, Ostrowski M, et al. T-cell Engagers in Prostate Cancer. Eur Urol. 2025;87:553–8. doi: 10.1016/j.eururo.2025.02.001. [DOI] [PubMed] [Google Scholar]

[R8] 8.Sutherland TE, Dyer DP, Allen JE. The extracellular matrix and the immune system: A mutually dependent relationship. Science. 2023;379:eabp8964. doi: 10.1126/science.abp8964. [DOI] [PubMed] [Google Scholar]

[R9] 9.Hu Q, Zhu YF, Mei J, et al. Extracellular matrix dynamics in tumor immunoregulation: from tumor microenvironment to immunotherapy. J Hematol Oncol . 2025;18 doi: 10.1186/s13045-025-01717-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R10] 10.Chen M, Zhao S, Xie X, et al. Role of Tunneling Nanotubes in Arachidonic Acid Transfer and Macrophage Function Reprogramming in Intrahepatic Cholangiocarcinoma. Adv Sci (Weinh) 2025;12 doi: 10.1002/advs.202500148. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R11] 11.Xiang W, Shi R, Kang X, et al. Monoacylglycerol lipase regulates cannabinoid receptor 2-dependent macrophage activation and cancer progression. Nat Commun. 2018;9:2574. doi: 10.1038/s41467-018-04999-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R12] 12.Pritykin Y, van der Veeken J, Pine AR, et al. A unified atlas of CD8 T cell dysfunctional states in cancer and infection. Mol Cell. 2021;81:2477–93. doi: 10.1016/j.molcel.2021.03.045. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R13] 13.Li Y, Shen Z, Chai Z, et al. Targeting MS4A4A on tumour-associated macrophages restores CD8+ T-cell-mediated antitumour immunity. Gut. 2023;72:2307–20. doi: 10.1136/gutjnl-2022-329147. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R14] 14.Yamaguchi Y, Gibson J, Ou K, et al. PD-L1 blockade restores CAR T cell activity through IFN-γ-regulation of CD163+ M2 macrophages. J Immunother Cancer. 2022;10:e004400. doi: 10.1136/jitc-2021-004400. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R15] 15.Wang S, Wang J, Chen Z, et al. Targeting M2-like tumor-associated macrophages is a potential therapeutic approach to overcome antitumor drug resistance. NPJ Precis Oncol . 2024;8:31. doi: 10.1038/s41698-024-00522-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R16] 16.Lv TT, Xiong XF, Yan W, et al. Targeting of GSDMD sensitizes HCC to anti-PD-1 by activating cGAS pathway and downregulating PD-L1 expression. J Immunother Cancer. 2022;10:e004763. doi: 10.1136/jitc-2022-004763. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R17] 17.Ning H, Huang S, Lei Y, et al. Enhancer decommissioning by MLL4 ablation elicits dsRNA-interferon signaling and GSDMD-mediated pyroptosis to potentiate anti-tumor immunity. Nat Commun. 2022;13 doi: 10.1038/s41467-022-34253-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R18] 18.Gao K, Xi W, Ni J, et al. Genetically modified extracellular vesicles loaded with activated gasdermin D potentially inhibit prostate-specific membrane antigen-positive prostate carcinoma growth and enhance immunotherapy. Biomaterials. 2025;315 doi: 10.1016/j.biomaterials.2024.122894. [DOI] [PubMed] [Google Scholar]

[R19] 19.Fontana P, Du G, Zhang Y, et al. Small-molecule GSDMD agonism in tumors stimulates antitumor immunity without toxicity. Cell. 2024;187:6165–81. doi: 10.1016/j.cell.2024.08.007. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R20] 20.Li M, Yang D, Yan H, et al. Gasdermin D maintains bone mass by rewiring the endo-lysosomal pathway of osteoclastic bone resorption. Dev Cell. 2022;57:2365–80. doi: 10.1016/j.devcel.2022.09.013. [DOI] [PubMed] [Google Scholar]

[R21] 21.Fan X, Han J, Zhong L, et al. Macrophage-Derived GSDMD Plays an Essential Role in Atherosclerosis and Cross Talk Between Macrophages via the Mitochondria-STING-IRF3/NF-κB Axis. ATVB . 2024;44:1365–78. doi: 10.1161/ATVBAHA.123.320612. [DOI] [PubMed] [Google Scholar]

[R22] 22.Bradfield CJ, Liang JJ, Ernst O, et al. Biphasic JNK signaling reveals distinct MAP3K complexes licensing inflammasome formation and pyroptosis. J Immunol. 2023;210:167. doi: 10.4049/jimmunol.210.Supp.167.01. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R23] 23.Ho L-C, Chen Y-H, Wu T-Y, et al. Phosphate burden induces vascular calcification through a NLRP3-caspase-1-mediated pyroptotic pathway. Life Sci. 2023;332:122123. doi: 10.1016/j.lfs.2023.122123. [DOI] [PubMed] [Google Scholar]

[R24] 24.Rathkey JK, Zhao J, Liu Z, et al. Chemical disruption of the pyroptotic pore-forming protein gasdermin D inhibits inflammatory cell death and sepsis. Sci Immunol. 2018;3:eaat2738. doi: 10.1126/sciimmunol.aat2738. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R25] 25.Hu JJ, Liu X, Xia S, et al. FDA-approved disulfiram inhibits pyroptosis by blocking gasdermin D pore formation. Nat Immunol. 2020;21:736–45. doi: 10.1038/s41590-020-0669-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R26] 26.Watson JL, Juergens D, Bennett NR, et al. De novo design of protein structure and function with RFdiffusion. Nature New Biol. 2023;620:1089–100. doi: 10.1038/s41586-023-06415-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R27] 27.Zhao S, Luo J, Xu P, et al. Designed peptide binders and nanobodies as PROTAC starting points for targeted degradation of PCNA and BCL6. Int J Biol Macromol. 2025;308:142667. doi: 10.1016/j.ijbiomac.2025.142667. [DOI] [PubMed] [Google Scholar]

[R28] 28.Liu M, Cai R, Wang T, et al. LAMC2 promotes the proliferation of cancer cells and induce infiltration of macrophages in non-small cell lung cancer. Ann Transl Med. 2021;9:1392. doi: 10.21037/atm-21-4507. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R29] 29.Nguyen NM, Pulkkinen L, Schlueter JA, et al. Lung development in laminin γ2 deficiency: abnormal tracheal hemidesmosomes with normal branching morphogenesis and epithelial differentiation. Respir Res . 2006;7 doi: 10.1186/1465-9921-7-28. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] 30.Fu R, Zhang C, Song M-M, et al. A single-cell map of patients with non-small cell lung cancer harboring rare-driver mutations after anti-PD-1 treatment. Cancer Lett. 2025;616:217595. doi: 10.1016/j.canlet.2025.217595. [DOI] [PubMed] [Google Scholar]

[R31] 31.Lu C, Gao Z, Wu D, et al. EP.12E.04 Understanding the Dynamics of TKI Induced Changes in the Tumor Immune Microenvironment for Improved Therapeutic Effect. J Thorac Oncol. 2024;19:S654. doi: 10.1016/j.jtho.2024.09.1235. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R32] 32.Griffin GK, Newton G, Tarrio ML, et al. IL-17 and TNF-α sustain neutrophil recruitment during inflammation through synergistic effects on endothelial activation. J Immunol. 2012;188:6287–99. doi: 10.4049/jimmunol.1200385. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R33] 33.Liu Y, Mei J, Gonzales LW, et al. IL-17a and tnf-± exert synergistic effects on expression of cxcl5 by alveolar type ii cells in vivo and in vitro. Am J Resp Crit Care. 2010;181 doi: 10.1164/ajrccm-conference.2010.181.1_MeetingAbstracts.A6787. [DOI] [PubMed] [Google Scholar]

[R34] 34.Zhang L, Lin Y, Hu L, et al. Transient intracellular expression of PD-L1 and VEGFR2 bispecific nanobody in cancer cells inspires long-term T cell activation and infiltration to combat tumor and inhibit cancer metastasis. Mol Cancer. 2025;24 doi: 10.1186/s12943-025-02253-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R35] 35.Nair S, Boczkowski D, Moeller B, et al. Synergy between tumor immunotherapy and antiangiogenic therapy. Blood. 2003;102:964–71. doi: 10.1182/blood-2002-12-3738. [DOI] [PubMed] [Google Scholar]

[R36] 36.Tada Y, Togashi Y, Kotani D, et al. Targeting VEGFR2 with Ramucirumab strongly impacts effector/ activated regulatory T cells and CD8 T cells in the tumor microenvironment. J Immunother Cancer. 2018 doi: 10.1186/s40425-018-0403-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R37] 37.Bhattacharya R, Ye X-C, Wang R, et al. Intracrine VEGF Signaling Mediates the Activity of Prosurvival Pathways in Human Colorectal Cancer Cells. Cancer Res. 2016;76:3014–24. doi: 10.1158/0008-5472.CAN-15-1605. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R38] 38.Karakhanova S, Link J, Heinrich M, et al. Characterization of myeloid leukocytes and soluble mediators in pancreatic cancer: importance of myeloid-derived suppressor cells. Oncoimmunology. 2015;4:e998519. doi: 10.1080/2162402X.2014.998519. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R39] 39.Wang G, Lu X, Dey P, et al. Targeting YAP-Dependent MDSC Infiltration Impairs Tumor Progression. Cancer Discov. 2016;6:80–95. doi: 10.1158/2159-8290.CD-15-0224. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R40] 40.Ma S, Cheng Q, Cai Y, et al. IL-17A Produced by γδ T Cells Promotes Tumor Growth in Hepatocellular Carcinoma. Cancer Res. 2014;74:1969–82. doi: 10.1158/0008-5472.CAN-13-2534. [DOI] [PubMed] [Google Scholar]

[R41] 41.Wang M, Han Y, Yao X, et al. Hyperexpression of tumor necrosis factor receptor 2 inhibits differentiation of myeloid‐derived suppressor cells by instigating apolarity during ageing. MedComm . 2024;5 doi: 10.1002/mco2.605. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R42] 42.Liu X, Tang R, Xu J, et al. CRIP1 fosters MDSC trafficking and resets tumour microenvironment via facilitating NF-κB/p65 nuclear translocation in pancreatic ductal adenocarcinoma. Gut. 2023;72:2329–43. doi: 10.1136/gutjnl-2022-329349. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R43] 43.Gabrilovich DI, Ostrand-Rosenberg S, Bronte V. Coordinated regulation of myeloid cells by tumours. Nat Rev Immunol. 2012;12:253–68. doi: 10.1038/nri3175. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R44] 44.Lu X, Horner JW, Paul E, et al. Effective combinatorial immunotherapy for castration-resistant prostate cancer. Nature New Biol. 2017;543:728–32. doi: 10.1038/nature21676. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R45] 45.de Visser KE, Joyce JA. The evolving tumor microenvironment: From cancer initiation to metastatic outgrowth. Cancer Cell. 2023;41:374–403. doi: 10.1016/j.ccell.2023.02.016. [DOI] [PubMed] [Google Scholar]

[R46] 46.Xin Z, Lin M, Hao Z, et al. The immune landscape of human thymic epithelial tumors. Nat Commun. 2022;13 doi: 10.1038/s41467-022-33170-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R47] 47.Lesche R, Groszer M, Gao J, et al. Cre/loxP-mediated inactivation of the murine Pten tumor suppressor gene. Genesis. 2002;32:148–9. doi: 10.1002/gene.10036. [DOI] [PubMed] [Google Scholar]

[R48] 48.Wang X, Du Q, Chen J, et al. Fatty acid dehydratase HACD3 poses protein kinase activity and promotes the malignant progression of colorectal cancer. Int J Biol Macromol. 2024;283:137414. doi: 10.1016/j.ijbiomac.2024.137414. [DOI] [PubMed] [Google Scholar]

[R49] 49.Yan YM, Tao HY, He JH, et al. The HDOCK server for integrated protein–protein docking. Nat Protoc. 2020;15:1829–52. doi: 10.1038/s41596-020-0312-x. [DOI] [PubMed] [Google Scholar]

[R50] 50.Abramson J, Adler J, Dunger J, et al. Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature New Biol. 2024;630:493–500. doi: 10.1038/s41586-024-07487-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Gasdermin D antagonizes immunosuppression in prostate cancer by inducing LAMC2 degradation to block M2 macrophage polarization

Qindan Du

Zhen Han

Yumeng Zhang

Ying Li

Ao Sun

Bin Wang

Yongquan Chen

Xiaoying Wang

Abstract

Background

Methods

Results

Conclusion

WHAT IS ALREADY KNOWN ON THIS TOPIC

WHAT THIS STUDY ADDS

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

Introduction

Results

GSDMD is reduced in PCa and indicates poor prognosis

GSDMD expression inhibits PCa progression in both genetically engineered mouse and syngeneic tumor models

De novo-designed binding partners reveal GSDMD–LAMC2 crosstalk and establish a role for GSDMD in LAMC2 ubiquitination

GSDMD regulates macrophage M2 polarization through LAMC2

Myeloid-specific deletion of GSDMD exacerbates tumor growth by remodeling the immunosuppressive tumor microenvironment

GSDMD activates the AKT/NF-κB/VEGF pathway in a CXCR2-dependent manner through CXCL5

GSDMD enhances immunotherapy by promoting CTL infiltration into PCa

Human GSDMD overexpression is associated with increased CTL infiltration

Discussion

Materials and methods

Mice

Animal experiments

Cell culture

Isolation of bone marrow-derived macrophages and in vitro analysis of macrophage polarization

H&E, immunohistochemistry and immunofluorescence

Establishment of mouse prostate tumor cell lines

RNA-sequencing analysis

RNA isolation and real-time quantitative PCR

Table 1. List of primers used for real-time quantitative PCR.

Western blot and co-immunoprecipitation

Lentiviral transduction

Flow cytometry

T-cell proliferation assay and T-cell killing assay

RFdiffusion to predict interspecies protein–protein interactions

Cell migration assay

ELISA assay of tumor sites and cell culture supernatant

Tube formation assay

In vitro permeability assay

Cell transfection

Statistical analysis

Supplementary material

Footnotes

Data availability statement

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases