Skip to main content
Download PDF

Signatures of H3K4me3 modification predict cancer immunotherapy response and identify a new immune checkpoint-SLAMF9

Respiratory Research volume 26, Article number: 17 (2025) Cite this article

Abstract

H3 lysine 4 trimethylation (H3K4me3) modification and related regulators extensively regulate various crucial transcriptional courses in health and disease. However, the regulatory relationship between H3K4me3 modification and anti-tumor immunity has not been fully elucidated. We identified 72 independent prognostic genes of lung adenocarcinoma (LUAD) whose transcriptional expression were closely correlated with known 27 H3K4me3 regulators. We constructed three H3K4me3 modification patterns utilizing the expression profiles of the 72 genes, and patients classified in each pattern exhibited unique tumor immune infiltration characteristics. Using the principal component analysis (PCA) of H3K4me3-related patterns, we constructed a H3K4me3 risk score (H3K4me3-RS) system. The deep learning analysis using 12,159 cancer samples from 26 cancer types and 725 cancer samples from 5 immunotherapy cohorts revealed that H3K4me3-RS was significantly correlated with cancer immune tolerance and sensitivity. Importantly, this risk-score system showed satisfactory predictive performance for the ICB therapy responses of patients suffering from several cancer types, and we identified that SLAMF9 was one of the immunosuppressive phenotype and immunotherapy resistance-determined genes of H3K4me3-RS. The mice melanoma model showed Slamf9 knockdown remarkably restrained cancer progression and enhanced the efficacy of anti-CTLA-4 and anti-PD-L1 therapies by elevating CD8 + T cell infiltration. This study provided a new H3K4me3-associated biomarker system to predict tumor immunotherapy response and suggested the preclinical rationale for investigating the roles of SLAMF9 in cancer immunity regulation and treatment.

Introduction

The growing field of immunotherapies reforms cancer treatment strategies, achieving remarkable clinical benefits for patients [1,2,3]. Especially immune checkpoint blockades (ICBs), which target inhibitory checkpoint molecules to enhance T cell functions, have been approved as the first-line treatment for various solid and hematological cancer types [4], despite many patients displaying primary hyporesponsiveness or only temporary disease control [5, 6]. Programmed cell death-ligand 1 (PD-L1) expression level within tumor tissues, tumor mutation burden (TMB), and tumor-infiltrating lymphocytes (TILs) are the mainstream predictive marker for ICB therapy response [7,8,9]. However, due to the complexity and heterogeneity of individual cancer immunity, the predictive competence of these biomarkers is difficult to keep reliable across different tumors. Therefore, the exploitation of more stable and feasible biomarkers that can accurately predict the therapeutic benefits is of significance for improving treatment outcomes and economic benefits.

Epigenetic dysregulation is one of the hallmarks of cancers [10, 11], the fundamental regulatory mechanisms mainly comprise chromosome modification and epitranscriptome regulation. DNA methylation and various histone modifications are responsible for gene transcriptional modulation by regulating chromosome accessibility and nucleosome stability, thereby engaging in downstream cellular behaviors [12,13,14]. Unlike DNA methylation, histone can be modified into various patterns, including methylation, acetylation, and phosphorylation, and play distinct roles in regulating transcriptional activity [15, 16]. Notably, the regulated targets of histone modifications include numerous cancer immunity-related genes, thus understanding the regulatory network of epigenetic modification is crucial for clarifying the epistatic control mechanism of anti-tumor immunity [17].

Among the multifarious histone modifications, H3K4me3 is a pivotal histone modification mode, linked to active transcription and euchromatins [18,19,20]. Previous studies uncovered that H3K4me3-targeted genes extensively regulate the fate of immune cells, cancer immunity activation, and immunotherapy efficiency [21,22,23,24]. For example, broad H3K4me3 domains have been found to regulate T cell differentiation in healthy individuals and activate oncogenes in neoplastic T cells [25]. Cxxc finger protein 1 (CFP1)-mediated H3K4me3 contributed to the homeostasis and function of group 3 innate lymphoid cells (ILC3s) in the field of innate immunity [26]. In addition, the functional effects of H3K4me3 are also found in other tumor-infiltrating immune cells [24]. Especially, H3K4me3 deposition also promoted PD-L1 expression in pancreatic cancer cells or PD-1 expression in T cells, suggesting the potential synergy effects of harnessing H3K4me3 on ICB therapies [27, 28]. These findings are the critical rationales for exploring targets and regulators of H3K4me3 to improve immunotherapy efficiency.

However, considering the crosstalk of epigenetic regulation, including mutually antagonistic and synergistic effects, it is necessary to explore cancer epigenetics from a multigene perspective, instead of a single target or modifier. Based on this thought, some studies have identified efficient epigenetic biomarkers and therapeutic targets in cancers. For example, cancer cell-specific DNA methylation patterns have been identified as the biomarker of cancer diagnosis and anti-tumor immunity characteristics [29,30,31], aiming to distinguish T cell effector status [32] and predict the clinical benefits of immunotherapy [33]. The histone acetylation-related gene signatures were identified as prognostic markers in various cancer types [34,35,36]. We also reported in a previous study that the total H3K4me3 level was correlated with poor survival of LUAD patients [37]. Nonetheless, whether the H3K4me3-associated transcriptional profile can be exploited as biomarkers for patients’ prognosis and immunotherapy responses is unknown.

In this study, we explored the relationship between H3K4me3 modification and anti-tumor immunophenotype across pan-cancers. We found that different H3K4me3 modification patterns based on the transcriptome signatures of H3K4me3-related histone lysine methyltransferases and lysine demethylases were associated with cancer immune phenotypes and tumor immune microenvironment (TIME) composition. To evaluate individual H3K4me3 modification for each patient, we designed a scoring system, termed H3K4me3-RS, which is competent in predicting patients’ responses to ICBs and survival across several cancer types. Crucially, we further screened the underlying genes that mediate H3K4me3-related immunosuppressive phenotypes and ICB therapy resistance, and we identified SLAMF9-mediated immunosuppressive effects. Public data analysis and in vivo experiments validated that tumor Slamf9 regulated CD8 + T cell infiltration and PD-1 expression within TME, and Slamf9 deficiency significantly sensitized mice with melanoma to anti-PD-L1 as well as anti-CTLA-4 therapies. Taken together, we shed light on the potency of H3K4me3 modification patterns as a predictor of prognosis and immunotherapy efficacy in cancers.

Results

Landscape of genetic variation of H3K4me3 regulators in lung adenocarcinoma

Non-small cell lung cancer (NSCLC) is identified as a kind of sensitive cancer type to ICB therapy. Pembrolizumab, Atezolizumab, and Nivolumab have been approved as the first-line therapy for advanced and metastatic NSCLC patients with PD-L1 ≥ 1% [38, 39]. We first assessed the gene expression and mutation signatures of all H3K4me3-related regulators in LUAD. These regulators include writers, erasers, and readers validated to mediate H3K4me3 modification, including MLL and KDM families, and other reported proteins associated with H3K4me3 methylation activity [13, 40]. In addition, H3K27me3 and H3K36me3 are the dominant histone markers cooperating with H3K4me3 to modulate transcriptional activities, thus we also showed the mutation profiles of their major histone modifiers including EZH2 (H3K27 methylation) and SETD2 (H3K36 methylation) in LUAD [41, 42]. In line with the infrequent mutation of histone regulators in most solid tumors, the waterfall plot showed that the mutation frequency of most genes is low in the TCGA LUAD cohort (Supplemental Fig. 1A-B). Notably, some H3K4me3 regulators have high copy number variation (CNV) frequencies, such as KMT2B, KDM5B, and PHF13 (Supplemental Fig. 1C), and their expression levels exhibit significant differences between normal tissues and tumor tissues of LUAD (Supplemental Fig. 1D). Interestingly, their tumor-promoting roles have been reported in cancer, suggesting that the high CNV frequency is linked to oncogenesis [43,44,45,46].

Fig. 1
figure 1

The differential enrichment pathways and immune cell infiltration of three H3K4me3 modification clusters. A Principal component analysis to distinguish three H3K4me3 modification clusters in the TCGA LUAD cohort. B Survival analysis of patients from the three H3K4me3-clusters. C Venn plot for differentially expressed genes between the three clusters. “B-A” means the DEGs between cluster A and cluster B, and “C-A”, “C-B” means in the same manner. D The GO and E KEGG analysis for the 467 intersection genes of the three DEG groups (B-A, C-A, C-B). F, G Comparison of the immune cell infiltration in the three H3K4me3 clusters through ssGSEA and MCP-counter respectively. The two immune cell categories and algorithms are referred to in the two pieces of literature [67, 69]

Full size image

H3K4me3 modification patterns and biological characteristics

To explore the relationship between the transcriptional profiles of H3K4me3-related regulators and biological characteristics of cancers, we performed a correlation analysis for the 27 H3K4me3-related regulators and all other gene expressions based on the bulk transcriptome sequencing data of the TCGA LUAD cohort. According to the screened criteria Person |r|> 0.5 and p value < 0.0001, 1470 genes whose transcriptional levels are significantly correlated with H3K4me3-related regulators were screened (Supplemental Table 1). We assumed that genes with prognostic effects should play more pivotal roles in oncogenic phenotype determination, so we further performed survival analysis for each gene using univariate and multivariate regression models. We identified 72 genes significantly correlated with LUAD patients’ prognosis (Supplemental Table 2). To distinguish different H3K4me3 modification patterns determined by the 72 prognosis-related gene transcriptional profiles scattered in the LUAD cohort, we performed PCA analysis based on the 72 gene transcriptional data and classified LUAD patients into three clusters (Fig. 1A). The three clusters termed cluster A, B, and C had significant prognostic differences, especially patients in cluster C had the worst survival, while the survival probabilities of patients in cluster A and B were similar (Fig. 1B). And over half of the 27 H3K4me3-related genes were differentially expressed in the three clusters (Supplemental Fig. 2A). To identify the biological differences between the three clusters, we came to 467 overlapping genes in pairs of differential gene sets within three clusters (Fig. 1C). The GO and KEGG analyses of the 467 genes showed that the underlying biological pathways responsible for differences in prognosis (Fig. 1D-E). Notably, these genes were greatly enriched at cancer immunity-related pathways, including T cell activation, leukocyte adhesion, leukocyte proliferation, T and B cell receptor signaling pathway, and cytokine-cytokine receptor interaction. These findings implied that H3K4me3 modification patterns were associated with cancer immunity regulation in LUAD. In addition, analyzed with three algorithms, the infiltration estimation of most immune cells in TME of patients from cluster C was less than those in the other two clusters (Fig. 1F-G, Supplemental Fig. 2B), especially anti-cancer immune cells, including CD8 + T cells, activated CD4 + T cells, M1-like macrophages, and activated dendritic cells. And the transcriptional score of cancer immunity-related signaling pathways, including CD8 T effector, pan-fibroblast TGF-β response signature (Pan-F-TBRS), antigen processing machinery (APM), Immune checkpoint (IC), IFN score, T cell-inflamed gene expression profiles (GEP), COX-associated cancer-promoting (COX-CP), COX-2-associated cancer-inhibitory (COX-CI), and COX-2-associated inflammatory-signature (COX-IS), are significantly differentiated in the three clusters, and patients from cluster C had the lowest score (Supplemental Fig. 2C). COX-CP and COX-CI represent tumor inflammation-related pathways. IFN score, Pan-F-TBRS, T cell-inflamed GEP pathways, and COX-IS (calculated as COX-CP divided by COX-CI) are associated with anti-PD-L1/anti-CTLA-4 response of patients [47,48,49]. Therefore, we speculated that the prognostic difference among the three clusters at least partially resulted from the differential activation of immunity/immunotherapy-related pathways. Moreover, we found that gene transcriptions of oncogenic signaling pathways were also differentially activated in the three clusters, including Epithelial-Mesenchymal Transition (EMT), DNA damage repair (DDR), Angiogenesis, cell cycle, DNA replication, Nucleotide excision repair, Mismatch repair, CCR interaction, p53 signaling pathway, Base excision repair, Oxidative phosphorylation, which implied that the different H3K4me3-modified profiles linked to different malignance-associated signaling pathways (Supplemental Fig. 2D) (for details and literature resources of these gene sets see Supplemental Table 3).

Fig. 2
figure 2

Relationship between H3K4me3-RS and anti-tumor immunophenotype of LUAD patients. A The expression and mutation signature of immune and oncogenic signaling pathway genes for LUAD patients (TCGA) grouped by high or low H3K4me3-RS. The rows of the heatmap represent gene expressions (z-scores). B Survival analysis for LUAD patients (TCGA) in high and low H3K4me3-RS subgroups. C Survival analysis for LUAD patients (TCGA) classified into four subgroups based on the TMB levels and H3K4me3-RS scores. D Immune cell population in the TME of two patient subgroups evaluated by MCP-counter. E The correlation analysis between H3K4me3-RS and anti-tumor immunity-related pathways. The red circle represents positive correlation and the blue circle represents negative correlation. The number represents the Pearson coefficient between two variables. The enrichment score of pathways was calculated with ssGSEA analysis. The asterisk (*) represents the range of p values. One * means p < 0.05, two * means p < 0.01, and three * means p < 0.001

Full size image

Relationship between H3K4me3-RS and core biological pathways

To further clarify the individual H3K4me3 modification patterns, we utilized the 72 genes to design a scoring system named H3K4me3-risk score (H3K4me3-RS) through PCA and calculated the H3K4me3 modification score for each patient from TCGA LUAD cohort. The computational formula was given in the Method. Given the optimal cut-off score of H3K4me3-RS in survival analysis, LUAD patients were cataloged into high and low H3K4me3-RS groups respectively, and we observed that the expression of some well-described histone modifiers was significantly different between the two groups. Patients with high H3K4me3 risk scores had a higher expression level of oncogenic regulators than patients with low H3K4me3 risk scores, such as EZH2 and KDM1A. By contrast, patients in the low H3K4me3-RS group had higher expression levels of tumor suppressor-related genes like KMT2A and KMT2C (Supplemental Fig. 3A). In addition, high H3K4me3-RS scores were positively correlated with advanced TNM stages, lymphatic metastasis, and dead cases (supplemental Fig. 3B).

Fig. 3
figure 3

The immune characteristics of H3K4me3-RS and its predictive ability in pan-cancers. A, B Heatmap showing the Pearson correlation coefficient of the correlation analysis between H3K4me3-RS and A immune cell infiltration (MCP-counter) or B enrichment score of anti-tumor immunity-related pathways across multiple cancer datasets from TCGA. Testicular germ cell tumors(TGCT; n = 155), lung adenocarcinoma (LUAD; n = 512), head and neck squamous cell carcinoma(HNSC; n = 517), uterine corpus endometrial carcinoma (UCEC; n = 530), primary skin cutaneous melanoma (PSKCM; n = 115), sarcoma (SARC; n = 259), kidney renal clear cell carcinoma (KIRC;, n = 516), cervical and endocervical cancers (CESCs; n = 305), lung squamous cell carcinoma (LUSC; n = 487), stomach adenocarcinoma (STAD; n = 412), esophageal carcinoma (ESCA; n = 183), metastatic skin cutaneous melanoma (MSKCM; n = 357), pancreatic adenocarcinoma (PAAD; n = 178), glioblastoma multiforme (GBM; n = 166), bladder urothelial carcinoma (BLCA; n = 408),ovarian cancer (OV; n = 305), acute myeloid leukemia (LAML; n = 173), thymoma (THYM; n = 120), rectum adenocarcinoma (READ; n = 156), prostate adenocarcinoma (PRAD; n = 495), pheochromocytoma and paraganglioma (PCPG; n = 183), kidney renal papillary cell carcinoma (KIRP; n = 286), brain lower grade glioma (LGG; n = 528), colon adenocarcinoma (COAD; n = 445), breast invasive carcinoma (BRCA; n = 976), liver hepatocellular carcinoma (LIHC; n = 370), and thyroid carcinoma(THCA; n = 507). C Kaplan–Meier survival plots showing the survival status of SKCM, KIRC, LIHC, SARC, and LAML patients classified as high or low H3K4me3-RS. D Multivariate Cox regression analysis of risk factors in SKCM, LUAD, KIRC, LIHC, SARC, and LAML patients

Full size image

To identify the underlying biological pathways contributing to the poor clinical outcomes of patients with high H3K4me3-RS scores, we compared the expression level and mutation status of the pivotal oncogenes and tumor suppressors of LUAD in the two H3K4me3-RS groups. Patients with driver mutations in RET, KRAS, NTRK2, and TP53 genes had higher H3K4me3 risk scores (Supplemental Fig. 4B, E) and relatively higher TMB levels (Fig. 2A, Supplemental Fig. 4A). TMB is a canonical biomarker for ICB therapy efficacy and is positively correlated with patient survival in the LUAD TCGA cohort [50, 51] (Supplemental Fig. 4D). LUAD patients with high H3K4me3-RS had a worse survival (Fig. 2B). Interestingly, Although H3K4me3-RS is directly proportional to TMB level in LUAD patients (Supplemental Fig. 4C), high H3K4me3-RS significantly withdrew the survival advantage of patients with high TMB, and patients with low TMB and high H3K4me3-RS showed the worst prognosis (Fig. 2C). This finding demonstrated that H3K4me3-RS combined with TMB had a superior prognostic capability than TMB only.

Fig. 4
figure 4

H3K4me3-RS is a novel predictor for ICB therapy response. A The Kaplan–Meier plot for the overall survival (OS) of the 24 melanoma patients treated with anti-PD-1 [52] with high or low H3K4me3-RS. B The Kaplan–Meier plot for the OS of the 43 melanoma patients [53] treated with anti-PD-1 with high or low H3K4me3-RS. C The Kaplan–Meier plot for the OS of the 41melanoma patients [54] treated with anti-PD-1 or anti-PD-1/anti-CTLA-4 combined therapy with high or low H3K4me3-RS. D–F Waterfall plots and bar plots of ICB therapy response for patients from three cohorts in A–C respectively. In the left waterfall plots, each column represents a patient, the column length represents the H3K4me3-RS values, and the color of the columns indicates the ICB response of the patients. In the right bar chart, the column represents the proportion of non-response (NR), response (R), or complete response/partial response (CR/PR), progressive disease (PD), and stable disease (SD) patients after ICB therapy in low H3K4me3-RS and high H3K4me3-RS subgroup. G–I Time-dependent ROC curves for the performance of H3K4me3-RS in predicting ICB responses of patients from the cohorts in A–C respectively. J The Kaplan–Meier plot for the OS of the 336 mUC patients [48] treated with anti-PD-L1 in high or low H3K4me3-RS. K The left bar chart for three TIME type proportions of two patient subgroups in (J). The right box plot for H3K4me3-RS values of patients classified in the three TIME types. L The correlogram for H3K4me3-RS and other immune-related molecular signatures based on the patient cohort in (J). The Blue ellipse represents negative correlation, and the red ellipse represents positive correlation. The value in the box is the Pearson correlation coefficient of two corresponding variables

Full size image

Based on the findings that three H3K4me3 clusters had distinct activation status of cancer immunity/immunotherapy-related pathways, we further demonstrated the genes from CD8 + T effector, APM, IC, pan-F-TBRS, IFN, T cell-inflamed GEP, and COX-IS pathways were highly expressed in the low H3K4me3-RS group (Fig. 2A, Supplemental Fig. 5B). The infiltration of anti-tumor effector immune cells in the low H3K4me3-RS subgroup was also significantly higher than those in the high subgroup (Fig. 2D, Supplemental Fig. 5A). Moreover, the correlation analysis validated that the H3K4me3-RS was negatively correlated with the activated status of CD8 effector, APM, IC, IFN, T cell-inflamed GEP, COX-CI, and COX-IS pathways (Fig. 2E, Supplemental Fig. 5B), suggesting high H3K4me3-RS is associated with immunosuppressive phenotypes in LUAD. In addition, we found that patients in the high H3K4me3-RS group also had high expression of genes grouped in oncogenic pathways, including DDR, EMT, and cell cycle (Fig. 2A, Supplemental Fig. 6A). High H3K4me3-RS was positively correlated with most tumor-promoting pathways in correlation analysis, except for the EMT processes (Supplemental Fig. 6B). The differential gene analysis of RNA-seq data from TCGA LUAD patients demonstrated that there were 379 differential expression genes between the high and low H3K4me3-RS subgroup, and upregulated genes in patients with low H3K4me3-RS were enriched to immune activity pathway, including INTERFERON_ALPHA_RESPONSE, IL6_JAK_STAT3_SIGNALING, IL2_STAT5_SIGNALING, INFLAMMATORY_RESPONSE, and INTERFERON_GAMMA_RESPONSE (Supplemental Fig. 7A-C).

Fig. 5
figure 5

Slamf9 mediates high H3K4me3-RS-related ICB therapy resistance in melanoma. A Sequencing of the expression log-fold change (log[FC]) value of genes in anti-CTLA4-resistant B16 murine tumors. The gene set is derived from the overlapping genes between the intersecting H3K4me3-RS-associated genes in the four cancer types (SARC, KIRC, SKCM, LUAD, p < 0.01, |r|> 0.1, Supplemental Table 4) and the DEGs between anti-CTLA4-resistant and parental murine B16 tumors [58]. B, C The Kaplan–Meier plot for the survival status of advanced clear cell renal cell carcinoma B and metastatic urothelial cancer C patients stratified by SLAMF9 expression level [48, 55]. The bar plot shows the proportion of patients with various ICB responses in the high and low SLAMF9 subgroups. D Comparison of Slamf9 expression level between anti-CTLA4-resistant and parental B16 murine tumors. Each dot represents a sample; the middle line represents the median expression value; the top and bottom of the boxes are the 75th and 25th percentiles (interquartile range), and the whiskers encompass 1.5 times the interquartile range. (E) The differentially expressed IFN-related genes in SLAMF9 knockout vs wild-type plasmacytoid dendritic cells. F, G The tumor volume curve and the final tumor weight after subcutaneous injection of the Slamf9-NC (n = 12) or Slamf9-knockdown (n = 12) B16F10 cells on C57BL/6 mice. H The representative image of immunofluorescent staining for tumor tissues from the Slamf9-NC and Slamf9-knockdown subcutaneous tumor model (F–G). The white fluorescence is CD8, the green fluorescence is CD3, and the red fluorescence is PD-1. The right bar plot for the number of cells stained by different fluorescences. (I) The schematic diagram of dosing regimen in Slamf9-knockdown or not B16F10 tumor-bearing mice. J, K Kaplan–Meier survival plots for Slamf9-NC (n = 24) or Slamf9-knockdown (n = 24) B16F10 tumor mice model. Randomly selecting mice for either anti-CTLA-4 inhibitor (n = 12) or anti-PD-L1 antibody (n = 12) treatment in the two groups respectively

Full size image

The immune characteristics of H3K4me3-RS and its predictive ability in pan-cancers

To validate the universality and reliability of the correlation between anti-tumor immunophenotype and H3K4me3-RS showed in the LUAD cohort, we performed the correlation analysis for immune signatures and H3K4me3-RS values across various cancer types from the TCGA database. Immune cell infiltrations (Fig. 3A) and immunotherapy response-related signatures (Fig. 3B) were positively correlated with H3K4me3-RS in some cancer types including pancreatic adenocarcinoma (PAAD) and bladder urothelial carcinoma (BLCA). In contrast, they showed inverse correlations in kidney renal clear cell carcinoma (KIRC), LUAD, sarcoma (SARC), and skin cutaneous melanoma (SKCM). To investigate whether H3K4me3-RS is an independent prognostic factor for cancer types in which H3K4me3-RS is negatively associated with anti-tumor immunity, we performed the survival analysis with the KIRC, SARC, SKCM, LAML, LIHC, and LUAD cohorts from TCGA database. Kaplan–Meier curves showed that patients with high H3K4me3-RS had a poor prognosis (Fig. 3C). Furthermore, the multivariate Cox regression model validated that H3K4me3-RS was an independent prognostic factor across the six cancer types after taking age, stage, sex, smoking, TMB, grade, and therapies into consideration (Fig. 3D).

H3K4me3-RS predicts ICB response

Based on the above findings that H3K4me3-RS is associated with the immunosuppressive phenotype and poor survival in some cancers, we further explored whether H3K4me3-RS could predict patients’ response to ICB therapy. We selected five published patient cohorts receiving ICB therapies for downstream analyses. H3K4me3-RS displayed a positive correlation with poor survival across the five cohorts, including three melanoma cohorts (Fig. 4A-C, Supplemental Fig. 8A) [52,53,54], one metastatic urothelial cancer (mUC) cohort (Fig. 4J)[48], and one advanced renal cell carcinoma cohort (Supplemental Fig. 8C-D) [55]. Among the five cohorts, patients determined as non-response (NR) or progression of disease (PD)/stable disease (SD) following ICB treatment accounted for high proportions in the high H3K4me3-RS group, and fewer patients achieved complete remission (CR) or partial remission (PR) (Fig. 4D-F, Supplemental Fig. 8E), and melanoma patients as in Fig. 4C with hyporesponsiveness to ICB treatment had a higher H3K4me3-RS (Supplemental Fig. 8B). These findings suggested that high H3K4me3-RS was a poor biomarker for the benefits of ICB therapies. The time-dependent ROC curves revealed that H3K4me3-RS displayed a satisfactory predictive ability for ICB responses of melanoma patients (Fig. 4G-I). TIME can be classified into inflamed, excluded, and desert phenotypes according to immune cell infiltration status, and these classifications are also biomarkers for ICB therapy [56]. The proportion of mUC patients characterized by desert phenotype TME was higher in the high H3K4me3-RS group than in the low H3K4me3-RS group. And mUC patients with desert phenotype had higher H3K4me3-RS scores than patients with excluded or inflamed phenotypes (Fig. 4K). To concretely present the infiltrating cells, we conducted the correlation analysis and found that H3K4me3-RS scores were negatively correlated with anti-tumor immunity effector cell infiltration in the mUC cohort (Fig. 4L). In addition, the intensity of PD-L1 expression within tumor tissues is also recognized as a predictor of ICB response. TME can be cataloged into IC0, IC1, and IC2 according to the IHC scoring of PD-L1 [57]. Likewise, mUC patients from the IMvigor210 study stratified into IC0 TME had the highest H3K4me-RS than patients stratified into IC1 and IC2, and IC0 patients were majority in the high H3K4me3-RS group (Supplemental Fig. 8F-G). As well, the gene signatures of PD1, PD-L1, IC, IFNG, T cell inflamed, and CD8 T effector also displayed opposite correlation with H3K4me3-RS (Supplemental Fig. 8H). These findings implied that the few infiltration of immune cells and constrained expression of immunotherapy-related molecules were potential mechanisms for H3K4me3-RS-related ICB therapy resistance.

SLAMF9 predicted by H3K4me3-RS serves as a regulator of ICB resistance

We presumed that some potential genes dominate the high H3K4me3-RS-related poor prognosis and ICB resistance of cancer patients. To get a shared gene set that correlated with H3K4me3-RS in the four cancer cohorts (SARC, KIRC, SKCM, LUAD) where H3K4me3-RS acted as the immunosuppressive biomarker, we selected the overlapping genes that were screened from the correlation analysis between H3K4me3-RS scores and the whole transcriptome gene expressions of the four TCGA cohorts, and the genes were filtered by criteria p value < 0.01 and Pearson |r|> 0.1 (Supplemental Table 4). We focused on those genes implicated in ICB resistance, so we intersected the screened gene set with the differentially expressed genes between anti-CTLA-4-resistant and parental B16 mouse tumors from the study of Victor et.al. [58]. We ranked the newly acquired intersection genes by their expression fold change and identified SLAMF9 as the most up-regulated gene (Fig. 5A). SLAMF9 was highly expressed within tumor tissues compared with normal tissues across most cancer types (Supplemental Fig. 9A), and IHC stainings displayed that SLAMF9 was expressed in all tumor tissues (Supplemental Fig. 10A-B). Similar to H3K4me3-RS, high SLAMF9 expression is associated with poor prognosis and anti-PD-L1 therapy resistance in advanced clear cell renal cell carcinoma cohorts (Fig. 5B) [55] and metastatic urothelial cancer cohorts (Fig. 5C) [48], which further confirmed the affinity between H3K4me3-RS and SLAMF9. Additionally, Slamf9 expression was markedly upregulated in anti-CTLA-4-resistant murine B16 melanoma compared with parental tumors (Fig. 5D). These findings implied that high SLAMF9 was associated with immunosuppressive phenotype and immunotherapy resistance, and the nature was shown among different tumors. Nonetheless, its prognostic ability was not consistent. For example, high SLAMF9 expression level was poor prognostic factor for patients from the KIRC, LUAD, SARC, and SKCM cohorts (Supplemental Fig. 9B), whereas positively correlated with patients’ survival in ovarian serous cystadenocarcinoma (OV), HNSC, KIRP, PRAD, or didn’t show prognostic ability in other cancer types (Supplemental Fig. 9C-D).

SLAMF9 is a member of the signaling lymphocytic activation molecule (SLAMF) family which includes nine receptors. Expressing on distinct hematopoietic cells, SLAMF1-9 molecules are implicated in immune regulation [59]. The functional characteristic of SLAMF9 is poorly described. A previous study disclosed that genes contributing to regulating IFN levels would be downregulated following SLAMF9 knockout in plasmacytoid dendritic cells (pDCs) [60](Fig. 5E). Studies have reported that these interferon-stimulated genes participated in ICB therapy resistance [61], so SLAMF9 and cancer immunity were likely linked. Intriguingly, with different immune cell infiltrating algorithms, we found that SLAMF9 expression was generally negatively correlated with CD8 + T cell infiltration score across various cancers (Supplemental Fig. 11A-B), especially the negative correlation was significant in metastatic SKCM compared to the primary tumor.

Furthermore, we performed in vivo experiments to confirm the roles of SLAMF9 in TME reshaping and ICB responses. B16F10 cell line with Slamf9-knockdown was constructed and injected subcutaneously into mice. The knockdown efficiency was verified by quantitative PCR (Supplemental Fig. 12). Slamf9 elimination significantly attenuated tumor growth in the mouse melanoma model (Fig. 5F-G; Supplemental Fig. 13A-C). Importantly, immunofluorescence staining showed that the intensity of CD3, PD-1, and CD8 expression remarkably increased in the isolated Slamf9-knockdown cancer tissues (Fig. 5H). Randomly selecting mice (n = 12 for each group) bearing Slamf9-knockdown or empty vector tumors for receiving anti-CTLA-4 inhibitor or anti-PD-L1 antibodies therapy (Fig. 5I), and we found that Slamf9-knockdown significantly improved the survival benefits of tumor-bearing mice after ICB treatment (Fig. 5J-K). Taken together, SLAMF9 inhibited CD8 + T cell infiltration in TME and markedly impaired ICB efficacy.

Discussion

In this study, we constructed a score method H3K4me3-RS to assess the correlation between the individual H3K4me3 modification signature of patients and cancer phenotypes. We found that the H3K4me3-RS was negatively correlated with cancer immune pathways and was an independent prognostic marker for patients. Critically, analyses with large cohorts suggested that patients with high H3K4me3-RS were resistant to ICB therapy, and SLAMF9 was a crucial molecule that mediated immunosuppressive phenotype and ICB therapy resistance. SLAMF9 has been found to express in different immune cells, such as plasmacytoid dendritic cells and macrophages, and participates in regulating immune cell functions and reactions to exogenous stimulation [60, 62,63,64]. Yet, its roles in cancer cells and TME regulation are unclear. We demonstrated that Slamf9 knockdown significantly sensitized mouse melanoma allograft model to anti-PD-L1 and anti-CTLA-4 therapy, suggesting that SLAMF9 inhibition is a promising strategy for overcoming ICB resistance.

Epigenetic dysregulation mechanisms provide opportunities to develop novel targets for cancer treatment, such as the archetypal DNA methylation inhibitors and histone acetylation inhibitors, which have achieved great clinical benefits in hematologic malignancies and several solid tumors. Moreover, the combinatorial therapy of epigenetic drugs coupled with other treatments including immunotherapies expands the available options for individual therapies and offers the opportunity for the improvement of immunotherapy efficacy [65]. However, due to the complexity and pleiotropy of epigenetic-mediated phenotype regulation, especially for histone modification networks, targeting single histone modification enzymes or regulators to correct the aberrant transcriptional profiles of multiple driver genes is underoptimized. Furthermore, although the functional elucidation of specific epigenetic modifiers increases our understanding of epigenetic regulation in cells, it is unlikely to offer integrated information on how the targeted modification patterns are modulated by synergetic or antagonistic modifiers. The complementary or adverse regulatory effects between various modifiers may be the critical obstacle to the effectiveness of epigenetic drugs. Therefore, the correlational exploration between all genotypes of specific epigenetic modification-associated modifiers and cellular transcriptome helps establish the integrated epigenetic modification-phenotype regulatory axis. With this strategy, we can reliably manipulate the epigenome of cells to regulate cellular functions and reveal the epigenetic regulation mechanisms of some genes that are not caught by conventional epigenetic detection techniques. Our study confirms the robusticity and university of the H3K4m3-central regulatory network for predicting prognosis and ICB responses and finds the potential role of SLAMF9 in cancer immunity regulation.

However, the specific mechanisms by which SLAMF9 performs the aforementioned regulatory function are unknown. Previous studies indicated that SLAMF9 lacks signaling motifs [66], so how the cancer cell-derived SLAMF9 is involved in cytoplastic tumor-promoting pathways is confusing. In addition, the effects of SLAMF9 on ICB therapies warrant more evidence in other cancer types.

Collectively, we design an evaluation system to characterize the individual H3K4me3 modification. H3K4me3-RS is a reliable predictor for prognosis and ICB therapies. SLAMF9 is a novel molecule associated with immunosuppressed phenotype and acts as a promising target for improving ICB therapy.

Methods

Cancer dataset sources and preprocessing

RNA sequencing data of all cancers and somatic mutation data of LUAD were downloaded from The Cancer Genome Atlas (TCGA) (https://portal.gdc.cancer.gov/) database, along with clinical annotation. There were 26 cancer types and corresponding datasets. Patients without survival data were excluded. The immunohistochemistry (IHC) images and staining scores of SLAMF9 in different cancer types were downloaded from The Human Protein Atlas (https://www.proteinatlas.org/). Patient cohorts with ICB efficacy and survival data were obtained from the five studies [48, 52,53,54,55], and patients without complete data were excluded. RNA sequencing data of anti-CTLA-4 resistant and parental B16 murine melanoma were obtained from this study [58].

TME cell evaluation

We used different algorithms to quantify the proportions of each cell type infiltrating TME in different analyses. The algorithms included CIBERSORT, MCP-counter, and single sample gene set enrichment analysis (ssGSEA). Patients’ RNA sequencing data for TME cell estimation were acquired from TCGA datasets or individual publications. The gene sets of TME-infiltrating cells for ssGSEA were obtained from the study of Zhang et al. [67]. The enrichment scores were calculated and visualized by R package “CIBERSORT”, “gsva”, “MCPcounter”, “pheatmap”. TIMER2.0 (http://timer.cistrome.org/) [68] was also used to analyze the immune cell infiltration.

Survival analysis

The survival information of patients from various cancers was downloaded from the TCGA database or individual publications. Survival analysis compared the prognostic status of patients stratified in different subgroups, and data analysis and the drawing of Kaplan–Meier curves were conducted by R package “survival” and “survminer”.

Functional enrichment analysis

The Gene Ontology (GO) and Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathway analyses of differentially expressed genes (DEGs) were performed using R packages “clusterProfiler”, “org.Hs.eg.db”, “enrichplot”, and “ggplot2”. GSEA hallmark term analysis and GO term analysis were performed with GSEA software (V.4.1.0) for the DEGs of high and low H3K4me3-RS subgroups. For each sample or collection of patients from specific cancer types, the enrichment scores of gene sets were calculated by ssGSEA. Gene sets included anti-tumor immunity-related gene sets, immunotherapy-related gene sets, and oncogenic pathway-related gene sets, and the gene lists were concluded in Supplemental Table 3.

Principal component analysis for H3K4me3 clusters and calculation of H3K4me3 risk score

Based on the RNA sequencing data of the TCGA LUAD cohort, the expression correlation analysis between a total of 27 H3K4 methylation-related regulators and other genes was conducted to screen genes whose expressions were correlated with the H3K4me3 modification signature. Then the prognostic ability of these genes was evaluated by univariate and multivariate Cox regression model and 72 prognosis-related genes were screened. PCA was applied to distinguish different H3K4me3 modification patterns in the LUAD cohort according to the expression level of the 72 genes. This method for identifying H3K4me3 modification features can down-weight the contribution of genes that didn’t track in a similar way as other gene sets, thereby avoiding making the distinction ambiguous. H3K4me3-RS for each patient was calculated by the formula [67]: H3K4me3-RS = Σ(PC1i + PC2i), the principal component 1 and 2 were used to generate the H3K4me3 risk score, and i means the expression of 72 H3K4me3-related genes.

Fluorescent multiplex immunohistochemistry

Mice B16F10 tumors were removed and mounted in OCT embedding medium (Thermo, USA) and stored at −80℃. After consecutive sections, fluorescent multiplex immunohistochemistry was carried out using AlphaTSA Multiplex IHC Kit (AXT37100031, Beijing, China) according to kit protocol. Primary antibodies included anti-CD8 (AB_2890649, 1:500, RRID: AB_2890649), anti-PD-1 (CST84651, 1:400, RRID: AB_2800041), and anti-CD3 (Abcam5690, 1:100, RRID: AB_305055), and each sections incubation in at 37℃ for 1 h. Negative controls didn’t receive primary antibody treatment. DAPI (alphaxbio, Beijing, China) was used for nuclei detection. Sections were analyzed with Leica Aperio Versa 200. Cell number was calculated with Huygens software. The cell number data is stored in Supplemental Table 5.

Cell transfection

Cell lines were purchased from ATCC and confirmed to be mycoplasma-free. B16F10 cell lines were cultured in basal medium DMEM (Corning, USA) supplemented with 10% fetal bovine serum FBS (Gibco, USA) and 1% penicillin–streptomycin (Gibco, USA) at 37℃, 5% CO2. 293 T cell lines were cultured in the same manner. For Slamf9 shRNA knockdown, three kinds of Slamf9 shRNAs were inserted into the lentiviral vectors (Genechem, Shanghai, China) respectively to construct target plasmids. The 293 T cells were transfected with target plasmids, lentivirus Helper plasmids (Genechem, Shanghai, China), Lipo3000 (Invitrogen), and P3000 (Invitrogen) in Optim-MEM medium (Invitrogen) at 37℃, 5%CO2, and replaced new medium after 8 h. The culture medium of co-transfected 293 T cells was collected and centrifuged to obtain the supernatant containing virus particles, and the virus was stored at −80℃. B16F10 cells were seeded onto a six-well plate at 50–60% confluence in the complete medium supplemented with polybrene, and virus liquids were added. 8 h later, the culture was replaced with complete medium. After 36 h, puromycin (2.5 μg/ml) was used for screening transfected B16F10 cells in complete medium for one week.

Construction of murine melanoma models

Slamf9-knockdown and Slamf9-negative control (NC) B16F10 cells were harvested by trypsinization and rinsed with PBS three times, finally diluted to 1 × 106/ml cells in PBS. 100μL cell suspension was subcutaneously injected into the flank of 8–10 week-old female C57BL/6 mice (each group, n = 12). Tumor sizes were measured every four days. In the experiment for ICB efficacy measurement, a total of 48 mice were included in the study and randomly allocated to Slamf9-knockdown or Slamf9-NC groups, 100μL 5 × 106/ml tumor cell suspensions were subcutaneously injected into each mouse. After tumor inoculation as described above, 12 mice from each group were treated with i.p. injection of 200 mg anti-CTLA-4 inhibitor (MDK 24720) or 200 mg anti-PD-L1 antibodies (Atezolizumab A2004, Selleck) respectively on day 3, and survival recording for 31 days until most mice were euthanized after the tumor maximum diameter greater than 15 mm.

Quantitative PCR

3 × 106 Slamf9-knockdown or Slamf9-NC B16F10 cells were collected and washed with PBS three times, then total RNA was extracted with the RNA-Quick Purification Kit (ES science, Shanghai, China) following the manufacturer's protocol. After measuring RNA concentration, cDNA was obtained using 10 μg RNA by reverse transcription PCR using RT-PCR Kit (Trans, Beijing, China). Quantitative real-time PCR was performed to detect the transcriptional level of Slamf9 in two kinds of B16F10 cells using primers (GENEray, Shanghai, China) and SYBER Green PCR Supermix Kit (Trans, Beijing, China). Data were analyzed in the ∆2CT method.

Statistical analysis

Pearson correlation coefficient was used for correlation analyzes. The optimal cut-off value for the H3K4me3-RS dichotomies of each dataset was determined by the cut points with the most statistically significant difference in survival analysis. The hazard ratios were calculated by the univariate or multivariable Cox regression model. The R package “forestplot” was used to visualize the result of regressions. The sensitivity and specificity of H3K4me3-RS predicting immunotherapy response were assessed by the receiver operating characteristic (ROC) curve and the area under the curve was calculated by the R package “pROC”. The waterfall plots of the somatic gene mutation for 27 H3K4methylation-related genes in the TCGA LUAD cohort were presented by R package “maftools”. The copy number variation plot of the 27 genes in the LUAD cohort was drawn with the R package “RCircos”. Tumor mutation burden of TCGA LUAD patients was calculated by R package “maftools”. Difference comparisons of two groups were analyzed with two-tailed unpaired Student’s t-test and three or more groups were analyzed with one-way ANOVA. Data were represented as mean ± SEM. Statistical P values < 0.05 was significant. Data were analyzed with the R (version 4.2.0) and GraphPad Prism 8.4.3.

Availability of data and materials

No datasets were generated or analysed during the current study.

References

  1. Cortiula F, Reymen B, Peters S, Van Mol P, Wauters E, Vansteenkiste J, De Ruysscher D, Hendriks LEL. Immunotherapy in unresectable stage III non-small-cell lung cancer: state of the art and novel therapeutic approaches. Ann Oncol. 2022. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.annonc.2022.06.013.

    Article  PubMed  Google Scholar 

  2. Huang AC, Zappasodi R. A decade of checkpoint blockade immunotherapy in melanoma: understanding the molecular basis for immune sensitivity and resistance. Nat Immunol. 2022;23:660–70.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Fenis A, Demaria O, Gauthier L, Vivier E, Narni-Mancinelli E. New immune cell engagers for cancer immunotherapy. Nat Rev Immunol. 2024;24:471–86.

    Article  CAS  PubMed  Google Scholar 

  4. Morad G, Helmink BA, Sharma P, Wargo JA. Hallmarks of response, resistance, and toxicity to immune checkpoint blockade. Cell. 2021;184:5309–37.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Zhang Q, Green MD, Lang X, Lazarus J, Parsels JD, Wei S, Parsels LA, Shi J, Ramnath N, Wahl DR, et al. Inhibition of ATM increases interferon signaling and sensitizes pancreatic cancer to immune checkpoint blockade therapy. Cancer Res. 2019;79:3940–51.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Cui JW, Li Y, Yang Y, Yang HK, Dong JM, Xiao ZH, He X, Guo JH, Wang RQ, Dai B, Zhou ZL. Tumor immunotherapy resistance: revealing the mechanism of PD-1 / PD-L1-mediated tumor immune escape. Biomed Pharmacother. 2024;171: 116203.

    Article  CAS  PubMed  Google Scholar 

  7. Doroshow DB, Bhalla S, Beasley MB, Sholl LM, Kerr KM, Gnjatic S, Wistuba II, Rimm DL, Tsao MS, Hirsch FR. PD-L1 as a biomarker of response to immune-checkpoint inhibitors. Nat Rev Clin Oncol. 2021;18:345–62.

    Article  CAS  PubMed  Google Scholar 

  8. McGrail DJ, Pilié PG, Rashid NU, Voorwerk L, Slagter M, Kok M, Jonasch E, Khasraw M, Heimberger AB, Lim B, et al. High tumor mutation burden fails to predict immune checkpoint blockade response across all cancer types. Ann Oncol. 2021;32:661–72.

    Article  CAS  PubMed  Google Scholar 

  9. Presti D, Dall’Olio FG, Besse B, Ribeiro JM, Di Meglio A, Soldato D. Tumor infiltrating lymphocytes (TILs) as a predictive biomarker of response to checkpoint blockers in solid tumors: a systematic review. Crit Rev Oncol Hematol. 2022;177: 103773.

    Article  PubMed  Google Scholar 

  10. Hanahan D. Hallmarks of cancer: new dimensions. Cancer Discov. 2022;12:31–46.

    Article  CAS  PubMed  Google Scholar 

  11. Pang L, Zhou F, Liu Y, Ali H, Khan F, Heimberger AB, Chen P. Epigenetic regulation of tumor immunity. J Clin Invest 2024:134.

  12. Li S, Peng Y, Panchenko AR. DNA methylation: precise modulation of chromatin structure and dynamics. Curr Opin Struct Biol. 2022;75: 102430.

    Article  CAS  PubMed  Google Scholar 

  13. Zhao S, Allis CD, Wang GG. The language of chromatin modification in human cancers. Nat Rev Cancer. 2021;21:413–30.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Davalos V, Esteller M. Cancer epigenetics in clinical practice. CA Cancer J Clin. 2023;73:376–424.

    Article  PubMed  Google Scholar 

  15. Bajbouj K, Al-Ali A, Ramakrishnan RK, Saber-Ayad M, Hamid Q. Histone modification in NSCLC: molecular mechanisms and therapeutic targets. Int J Mol Sci. 2021;22:11701.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Zhang Y, Zhang Q, Zhang Y, Han J. The role of histone modification in dna replication-coupled nucleosome assembly and cancer. Int J Mol Sci 2023, 24.

  17. Hogg SJ, Beavis PA, Dawson MA, Johnstone RW. Targeting the epigenetic regulation of antitumour immunity. Nat Rev Drug Discov. 2020;19:776–800.

    Article  CAS  PubMed  Google Scholar 

  18. Zhang T, Cooper S, Brockdorff N. The interplay of histone modifications - writers that read. EMBO Rep. 2015;16:1467–81.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Fan T, Zhu M, Muhammad S, Xiao C, Li S, Tian H, Liu Y, Xue L, Zheng B, Li C, He J. H3K4me3-related lncRNAs signature and comprehensive analysis of H3K4me3 regulating tumor immunity in lung adenocarcinoma. Respir Res. 2023;24:122.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Fan T, Xiao C, Liu H, Liu Y, Wang L, Tian H, Li C, He J. CXXC finger protein 1 (CFP1) bridges the reshaping of genomic H3K4me3 signature to the advancement of lung adenocarcinoma. Signal Transduct Target Ther. 2023;8:369.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Piatek P, Namiecinska M, Lewkowicz N, Kulińska-Michalska M, Jabłonowski Z, Matysiak M, Dulska J, Michlewska S, Wieczorek M, Lewkowicz P. Changes within H3K4me3-marked histone reveal molecular background of neutrophil functional plasticity. Front Immunol. 2022;13: 906311.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Stamos DB, Clubb LM, Mitra A, Chopp LB, Nie J, Ding Y, Das A, Venkataganesh H, Lee J, El-Khoury D, et al. The histone demethylase Lsd1 regulates multiple repressive gene programs during T cell development. J Exp Med. 2021. https://doiorg.publicaciones.saludcastillayleon.es/10.1084/jem.20202012.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Zhang J, Zhou Q, Xie K, Cheng L, Peng S, Xie R, Liu L, Zhang Y, Dong W, Han J, et al. Targeting WD repeat domain 5 enhances chemosensitivity and inhibits proliferation and programmed death-ligand 1 expression in bladder cancer. J Exp Clin Cancer Res. 2021;40:203.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Xiao C, Fan T, Zheng Y, Tian H, Deng Z, Liu J, Li C, He J. H3K4 trimethylation regulates cancer immunity: a promising therapeutic target in combination with immunotherapy. J Immunother Cancer. 2023. https://doiorg.publicaciones.saludcastillayleon.es/10.1136/jitc-2022-005693.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Belhocine M, Simonin M, Abad Flores JD, Cieslak A, Manosalva I, Pradel L, Smith C, Mathieu EL, Charbonnier G, Martens JHA, et al. Dynamics of broad H3K4me3 domains uncover an epigenetic switch between cell identity and cancer-related genes. Genome Res. 2022;32:1328–42.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Shen X, Gao X, Luo Y, Xu Q, Fan Y, Hong S, Huang Z, Liu X, Wang Q, Chen Z, et al. Cxxc finger protein 1 maintains homeostasis and function of intestinal group 3 innate lymphoid cells with aging. Nat Aging. 2023;3:965–81.

    Article  CAS  PubMed  Google Scholar 

  27. Lu C, Paschall AV, Shi H, Savage N, Waller JL, Sabbatini ME, Oberlies NH, Pearce C, Liu K. The MLL1-H3K4me3 axis-mediated PD-L1 expression and pancreatic cancer immune evasion. J Natl Cancer Inst. 2017. https://doiorg.publicaciones.saludcastillayleon.es/10.1093/jnci/djw283.

    Article  PubMed  PubMed Central  Google Scholar 

  28. Redd PS, Lu C, Klement JD, Ibrahim ML, Zhou G, Kumai T, Celis E, Liu K. H3K4me3 mediates the NF-κB p50 homodimer binding to the pdcd1 promoter to activate PD-1 transcription in T cells. Oncoimmunology. 2018;7: e1483302.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Nishiyama A, Nakanishi M. Navigating the DNA methylation landscape of cancer. Trends Genet. 2021;37:1012–27.

    Article  CAS  PubMed  Google Scholar 

  30. Papanicolau-Sengos A, Aldape K. DNA methylation profiling: an emerging paradigm for cancer diagnosis. Annu Rev Pathol. 2022;17:295–321.

    Article  CAS  PubMed  Google Scholar 

  31. Wu Y, Seufert I, Al-Shaheri FN, Kurilov R, Bauer AS, Manoochehri M, Moskalev EA, Brors B, Tjaden C, Giese NA, et al. DNA-methylation signature accurately differentiates pancreatic cancer from chronic pancreatitis in tissue and plasma. Gut. 2023;72:2344–53.

    Article  CAS  PubMed  Google Scholar 

  32. Frias AB, Boi SK, Lan X, Youngblood B. Epigenetic regulation of T cell adaptive immunity. Immunol Rev. 2021;300:9–21.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Kim JY, Choi JK, Jung H. Genome-wide methylation patterns predict clinical benefit of immunotherapy in lung cancer. Clin Epigenetics. 2020;12:119.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Jing Z, Ziwang F, Yinhang W, Yani Z, Jian C, Jingwen W, Shuwen H. Novel acetylation-related gene signatures for predicting the prognosis of patients with colorectal cancer. Hum Cell. 2022;35:1159–73.

    Article  CAS  PubMed  Google Scholar 

  35. Dai Q, Ye Y. Development and validation of a novel histone acetylation-related gene signature for predicting the prognosis of ovarian cancer. Front Cell Dev Biol. 2022;10: 793425.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Hervás-Corpión I, Navarro-Calvo J, Martín-Climent P, Iriarte-Gahete M, Geribaldi-Doldán N, Castro C, Valor LM. Defining a correlative transcriptional signature associated with bulk histone H3 acetylation levels in adult glioblastomas. Cells. 2023. https://doiorg.publicaciones.saludcastillayleon.es/10.3390/cells12030374.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Fan T, Liu Y, Liu H, Wang L, Tian H, Zheng Y, Zheng B, Xue L, Li C, He J. Transmembrane protein-based risk model and H3K4me3 modification characteristics in lung adenocarcinoma. Front Oncol. 2022;12: 828814.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Ettinger DS, Wood DE, Aisner DL, Akerley W, Bauman JR, Bharat A, Bruno DS, Chang JY, Chirieac LR, D’Amico TA, et al. Non-small cell lung cancer, version 3.2022, NCCN clinical practice guidelines in oncology. J Natl Compr Canc Netw. 2022;20:497–530.

    Article  PubMed  Google Scholar 

  39. Lahiri A, Maji A, Potdar PD, Singh N, Parikh P, Bisht B, Mukherjee A, Paul MK. Lung cancer immunotherapy: progress, pitfalls, and promises. Mol Cancer. 2023;22:40.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Chao YL, Pecot CV. Targeting epigenetics in lung cancer. Cold Spring Harb Perspect Med. 2021. https://doiorg.publicaciones.saludcastillayleon.es/10.1101/cshperspect.a038000.

    Article  PubMed  PubMed Central  Google Scholar 

  41. Mei H, Wu H, Yang J, Zhou B, Wang A, Hu C, Qi S, Jiang Z, Zou F, Wang B, et al. Discovery of IHMT-337 as a potent irreversible EZH2 inhibitor targeting CDK4 transcription for malignancies. Signal Transduct Target Ther. 2023;8:18.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Weng Y, Xue J, Niu N. SETD2 in cancer: functions, molecular mechanisms, and therapeutic regimens. Cancer Biol Med. 2024;21:725–30.

    Article  PubMed  PubMed Central  Google Scholar 

  43. Leng X, Wang J, An N, Wang X, Sun Y, Chen Z. Histone 3 lysine-27 demethylase KDM6A coordinates with KMT2B to play an oncogenic role in NSCLC by regulating H3K4me3. Oncogene. 2020;39:6468–79.

    Article  CAS  PubMed  Google Scholar 

  44. Fu YD, Huang MJ, Guo JW, You YZ, Liu HM, Huang LH, Yu B. Targeting histone demethylase KDM5B for cancer treatment. Eur J Med Chem. 2020;208: 112760.

    Article  CAS  PubMed  Google Scholar 

  45. Duan R, Du W, Guo W. EZH2: a novel target for cancer treatment. J Hematol Oncol. 2020;13:104.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Sun Y, Li D, Liu H, Huang Y, Meng F, Tang J, Li Z, Xie W. PHF13 epigenetically activates TGFβ driven epithelial to mesenchymal transition. Cell Death Dis. 2022;13:487.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  47. Ayers M, Lunceford J, Nebozhyn M, Murphy E, Loboda A, Kaufman DR, Albright A, Cheng JD, Kang SP, Shankaran V, et al. IFN-gamma-related mRNA profile predicts clinical response to PD-1 blockade. J Clin Invest. 2017;127:2930–40.

    Article  PubMed  PubMed Central  Google Scholar 

  48. Mariathasan S, Turley SJ, Nickles D, Castiglioni A, Yuen K, Wang Y, Kadel EE III, Koeppen H, Astarita JL, Cubas R, et al. TGFbeta attenuates tumour response to PD-L1 blockade by contributing to exclusion of T cells. Nature. 2018;554:544–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  49. Bonavita E, Bromley CP, Jonsson G, Pelly VS, Sahoo S, Walwyn-Brown K, Mensurado S, Moeini A, Flanagan E, Bell CR, et al. Antagonistic inflammatory phenotypes dictate tumor fate and response to immune checkpoint blockade. Immunity. 2020;53(1215–1229): e1218.

    Google Scholar 

  50. Chan TA, Yarchoan M, Jaffee E, Swanton C, Quezada SA, Stenzinger A, Peters S. Development of tumor mutation burden as an immunotherapy biomarker: utility for the oncology clinic. Ann Oncol. 2019;30:44–56.

    Article  CAS  PubMed  Google Scholar 

  51. van de Haar J, Mankor JM, Hummelink K, Monkhorst K, Smit EF, Wessels LFA, Cuppen E, Aerts J, Voest EE. Combining genomic biomarkers to guide immunotherapy in non-small cell lung cancer. Clin Cancer Res. 2024;30:1307–18.

    Article  PubMed  PubMed Central  Google Scholar 

  52. Hugo W, Zaretsky JM, Sun L, Song C, Moreno BH, Hu-Lieskovan S, Berent-Maoz B, Pang J, Chmielowski B, Cherry G, et al. Genomic and transcriptomic features of response to anti-PD-1 therapy in metastatic melanoma. Cell. 2016;165:35–44.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Riaz N, Havel JJ, Makarov V, Desrichard A, Urba WJ, Sims JS, Hodi FS, Martín-Algarra S, Mandal R, Sharfman WH, et al. Tumor and microenvironment evolution during Immunotherapy with Nivolumab. Cell. 2017;171:934-949.e916.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Gide TN, Quek C, Menzies AM, Tasker AT, Shang P, Holst J, Madore J, Lim SY, Velickovic R, Wongchenko M, et al. Distinct immune cell populations define response to anti-PD-1 monotherapy and anti-PD-1/Anti-CTLA-4 combined therapy. Cancer Cell. 2019;35:238-255.e236.

    Article  CAS  PubMed  Google Scholar 

  55. Am R, Ml W, Mm K, Sk W. T B, T B, RH D: RASSF10 is frequently epigenetically inactivated in kidney cancer and its knockout promotes neoplasia in cancer prone mice. Oncogene. 2020;39:3114–27.

    Article  Google Scholar 

  56. Du F, Li H, Li Y, Liu Y, Li X, Dang N, Chu Q, Yan J, Fang Z, Wu H, et al. Identification of m(6)A regulator-associated methylation modification clusters and immune profiles in melanoma. Front Cell Dev Biol. 2021;9: 761134.

    Article  PubMed  PubMed Central  Google Scholar 

  57. Nishino M, Ramaiya NH, Hatabu H, Hodi FS. Monitoring immune-checkpoint blockade: response evaluation and biomarker development. Nat Rev Clin Oncol. 2017;14:655–68.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  58. Twyman-Saint Victor C, Rech AJ, Maity A, Rengan R, Pauken KE, Stelekati E, Benci JL, Xu B, Dada H, Odorizzi PM, et al. Radiation and dual checkpoint blockade activate non-redundant immune mechanisms in cancer. Nature. 2015;520:373–7.

    Article  CAS  PubMed  Google Scholar 

  59. Dragovich MA, Mor A. The SLAM family receptors: Potential therapeutic targets for inflammatory and autoimmune diseases. Autoimmun Rev. 2018;17:674–82.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  60. Sever L, Radomir L, Stirm K, Wiener A, Schottlender N, Lewinsky H, Barak AF, Friedlander G, Ben-Dor S, Becker-Herman S, Shachar I. SLAMF9 regulates pDC homeostasis and function in health and disease. Proc Natl Acad Sci U S A. 2019;116:16489–96.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  61. Benci JL, Xu B, Qiu Y, Wu TJ, Dada H, Twyman-Saint Victor C, Cucolo L, Lee DSM, Pauken KE, Huang AC, et al. Tumor Interferon Signaling Regulates a Multigenic Resistance Program to Immune Checkpoint Blockade. Cell. 2016;167:1540-1554.e1512.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  62. Dollt C, Michel J, Kloss L, Melchers S, Schledzewski K, Becker K, Sauer A, Krewer A, Koll F, Schmieder A. The novel immunoglobulin super family receptor SLAMF9 identified in TAM of murine and human melanoma influences pro-inflammatory cytokine secretion and migration. Cell Death Dis. 2018;9:939.

    Article  PubMed  PubMed Central  Google Scholar 

  63. Wilson TJ, Clare S, Mikulin J, Johnson CM, Harcourt K, Lyons PA, Dougan G, Smith KGC. Signalling lymphocyte activation molecule family member 9 is found on select subsets of antigen-presenting cells and promotes resistance to Salmonella infection. Immunology. 2020;159:393–403.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  64. Yang Y, Huang J, Li X, Lin R, Wang X, Xiao G, Zeng J, Wang Z. Periplaneta americana extract promotes infectious diabetic ulcers wound healing by downregulation of LINC01133/SLAMF9. Chin J Nat Med. 2024;22:608–18.

    PubMed  Google Scholar 

  65. Dai E, Zhu Z, Wahed S, Qu Z, Storkus WJ, Guo ZS. Epigenetic modulation of antitumor immunity for improved cancer immunotherapy. Mol Cancer. 2021;20:171.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  66. Zeng X, Liu G, Peng W, He J, Cai C, Xiong W, Chen S, Yang M, Dong Z. Combined deficiency of SLAMF8 and SLAMF9 prevents endotoxin-induced liver inflammation by downregulating TLR4 expression on macrophages. Cell Mol Immunol. 2020;17:153–62.

    Article  CAS  PubMed  Google Scholar 

  67. Zhang B, Wu Q, Li B, Wang D, Wang L, Zhou YL. m(6)A regulator-mediated methylation modification patterns and tumor microenvironment infiltration characterization in gastric cancer. Mol Cancer. 2020;19:53.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  68. Li T, Fu J, Zeng Z, Cohen D, Li J, Chen Q, Li B, Liu XS. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res. 2020;48:509–14.

    Article  Google Scholar 

  69. Becht E, Giraldo NA, Lacroix L, Buttard B, Elarouci N, Petitprez F, Selves J, Laurent-Puig P, Sautès-Fridman C, Fridman WH, de Reyniès A. Estimating the population abundance of tissue-infiltrating immune and stromal cell populations using gene expression. Genome Biol. 2016;17:218.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We are very grateful to the TCGA database for providing valuable data resources.

Funding

This work was supported by the National Natural Science Foundation of China (82473443), the Beijing Natural Science Foundation (L248050, 7242119), the National Key R&D Program of China (2021YFF1201303), and the CAMS Innovation Fund for Medical Sciences (CIFMS) (2021–1-I2M-012).

Author information

Author notes

  1. Tao Fan and Chu Xiao contributed equally to this work.

Authors and Affiliations

  1. Department of Thoracic Surgery, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, 100021, China

    Tao Fan, Chu Xiao, Ziqin Deng, He Tian, Yujia Zheng, Chunxiang Li & Jie He

  2. Department of Colorectal Surgery, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China

    Shuofeng Li

  3. Department of Pathology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, 100021, China

    Bo Zheng

Authors
  1. Tao Fan
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Chu Xiao
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Ziqin Deng
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Shuofeng Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. He Tian
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Yujia Zheng
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Bo Zheng
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Chunxiang Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Jie He
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

Conceptualization: CXL, JH; Methodology: TF, CX, ZQD; Visualization: TF, CX; Project administration: JH; Writing – original draft: TF, CX; Writing – review & editing: ZQD, SFL, HT, YJZ, BZ.

Corresponding authors

Correspondence to Chunxiang Li or Jie He.

Ethics declarations

Ethics approval and consent to participate

The animal experiment was performed following a protocol approved by the Institutional Animal Care and Use Committee of the Cancer Hospital, Chinese Academy of Medical Sciences.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Fan, T., Xiao, C., Deng, Z. et al. Signatures of H3K4me3 modification predict cancer immunotherapy response and identify a new immune checkpoint-SLAMF9. Respir Res 26, 17 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12931-024-03093-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12931-024-03093-6

Respiratory Research

ISSN: 1465-993X

Contact us