Pan-cancer analysis identifies JMJD6 as an oncogene and prognostic biomarker

Abstract

JMJD6, a bifunctional epigenetic modulator within the Jumonji C family, has garnered increasing attention for its potential roles in tumorigenesis, yet its pan-cancer prognostic and mechanistic significance remain incompletely characterized. In this study, we performed a comprehensive pan-cancer analysis of JMJD6 using multi-omics data from public databases including TCGA, CPTAC, HPA, and GEPIA2. Our results revealed that JMJD6 is significantly overexpressed across cancers and positively correlated with advanced tumor stage and unfavorable patient survival outcomes. Mechanistically, JMJD6 overexpression was associated with promoter hypomethylation and showed close interactions with RNA modification regulators. Furthermore, JMJD6 expression correlated significantly with immune cell infiltration, elevated genomic instability (TMB, MSI, HRD), and differential sensitivity to targeted therapies such as EGFR inhibitors. Functional enrichment analysis underscored its involvement in spliceosome and chromatin remodeling pathways. Collectively, our findings establish JMJD6 as an influential oncogene and robust prognostic biomarker across cancer types, with implications for future therapeutic strategies targeting epigenetic and immune pathways.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-026-04847-y.

Introduction

Cancer remains one of the leading causes of mortality worldwide, characterized by multidimensional dysregulated mechanisms during its pathogenesis, encompassing genetic alterations, epigenetic abnormalities, and tumor microenvironment dysregulation [1]. Specifically, epigenetic mechanisms, such as DNA methylation, histone modifications, and alternative RNA splicing, govern gene expression dynamically without altering the DNA sequence, playing central roles in tumorigenesis and progression [2, 3]. The Jumonji C (JmjC) domain-containing family of histone demethylases, modulates chromatin accessibility and transcriptional activity by erasing methyl groups from histone tails (e.g., H3K4me3 and H3K27me3). Their dysregulation is closely associated with malignant transformation, establishing them as pivotal targets in cancer epigenetics [4, 5].

JMJD6 (Jumonji domain containing 6), a key member of the JmjC family, was initially identified as a phosphatidylserine receptor (PSR) localized on macrophage cell surface and involved in apoptotic cell clearance [6, 7]. Subsequent studies revealed its nuclear localization and dual enzymatic functions [8–10]. JMJD6 specially demethylates histone residues H3R2me2 and H4R3me2 to modulate chromatin structure [11], and also hydroxylates non-histone substrates including U2AF65 and p53 [12–14]. Additionally, JMJD6 has also been implicated in tyrosine kinase and protease activities [15, 16], underscoring its involvement in diverse biological processes such as transcriptional regulation, RNA splicing, and chromatin remodeling.

Accumulating evidence indicates that JMJD6 is frequently overexpressed in numerous cancers, including breast, lung, melanoma, liver, colon and oral cancers, suggesting a conserved oncogenic role across tumor types [17]. Mechanistic studies further reveal that JMJD6 promotes tumorigenesis via regulation of cell proliferation and survival pathways [17, 18]. Clinically, elevated JMJD6 expression correlates with aggressive phenotypes, metastasis, and poor patient survival [19–21]. Although these findings underscore context-specific roles of JMJD6, a systematic pan-cancer analysis is still lacking. A comprehensive investigation into its expression, functional implications, and clinical relevance across malignancies is essential to evaluate its potential as a therapeutic target.

In this study, we leverage large-scale public databases to conduct an integrative pan-cancer analysis of JMJD6. Our work aims to elucidate its clinicopathological significance, prognostic value, and molecular mechanisms, establishing JMJD6 as a pan-cancer biomarker and a candidate target for therapy, providing a foundation for subsequent experimental and clinical translation.

Results

Expression pattern of JMJD6 in pan-cancer

Analysis of the HPA database revealed that JMJD6 protein is predominantly localized in the nucleoplasm and exhibits specific high expression in bone marrow (Fig S1a, S1b, 1 A). Subsequent analysis revealed the RNA expression of JMJD6 was low immune cell and cancer specificity (Fig. 1B, S1c).

Fig. 1.

Expression of JMJD6 in normal and malignant tissues. A JMJD6 mRNA expression across 54 normal tissue and 7 blood cell types based on three consensus datasets (HPA, GTEx and FANTOM5); B JMJD6 mRNA expression in immune cell subtypes from HPA database; C JMJD6 mRNA expression in normal and tumor tissues across cancer types from TCGA via UALCAN; D JMJD6 protein expression in normal and tumor tissues from CTPAC via UALCAN; ns: not significant, * p < 0.05, ** p < 0.01, *** p < 0.001

Using TCGA data, we found that JMJD6 mRNA was significantly upregulated in 11 cancer types, including BRCA, CESC, CHOL, COAD, HNSC, KIRP, KIRC, LIHC, LUAD, LUSC, and READ, while only downregulated in 3 cancers (BLCA, KICH, PCPG) (Fig. 1C). Protein-level analysis via CPTAC confirmed elevated JMJD6 expression in BRCA, COAD, KIRC, LC, HNSC, GBM, and LIHC, consistent with mRNA trends, in contrast, PAAD showed reduced JMJD6 protein despite no significant changes in mRNA (Fig. 1D), suggesting post-transcriptional regulation and cancer-type-specific regulatory mechanisms.

Association between JMJD6 expression and clinicopathological features

Using the UALCAN platform, we observed that JMJD6 expression increased with advancing tumor stage in BRCA, CHOL, COAD, HNSC, KIRC, KIRP, LIHC, LUAD, LUSC and READ (all P < 0.05), with a more prominent upregulation trend in COAD, HNSC, LUAD, and KIRC (Fig. 2). Conversely, JMJD6 expression decreased with tumor stage in KICH and THCA (P < 0.05), although no significant reduction was observed in stage IV KICH (Fig. 2). These results indicate that JMJD6 expression is closely linked to disease progression in multiple cancers.

Fig. 2.

JMJD6 expression across tumor stages. JMJD6 expression levels in normal and different tumor stages (I-IV) tissues in various cancer types; ns: not significant, * p < 0.05, ** p < 0.01, *** p < 0.001

Prognostic value of JMJD6 in pan-cancer

Kaplan-Meier analysis showed high JMJD6 expression was associated with worse overall survival (OS) in patients with ACC, GBM, KIPAN, KIRC, KIRP, LGG, LIHC, MESO and UVM (all P < 0.05) (Fig. 3A). Only in PAAD was low JMJD6 expression correlated with better OS (P < 0.05). For disease-specific survival (DSS), high JMJD6 levels predicted shorter survival in ACC, GBM, KICH, KIPAN, KIRC, KIRP, LGG, LIHC, LUSC, MESO and UVW (all P < 0.05) (Fig. 3B). Similarly, elevated JMJD6 was associated with reduced progression-free interval (PFI) in nine cancers and decreased disease-free interval (DFI) in six cancers (P < 0.05; Fig S2a, S2b). These findings support JMJD6 as a risk factor and potential prognostic biomarker in most cancer types.

Fig. 3.

Survival analysis of JMJD6 in pan-cancer. A Overall survival (OS) and B Disease specific survival (DSS) of patients with high vs. low JMJD6 expression

Epigenetic regulation of JMJD6

DNA methylation plays a pivotal role in gene expression regulation [2, 22]. Through the GSCA database, we identified a significant negative correlation between JMJD6 mRNA expression and its DNA methylation level across nearly all cancer types (P < 0.05) (Fig. 4A). High JMJD6 methylation was correlated with better OS in ACC, BLCA, CESC, ESCA, and KIRC (P < 0.05) (Fig. 4B), suggesting that promoter hypomethylation may drive JMJD6 overexpression and contribute to poor prognosis.

Fig. 4.

Epigenetic regulation of JMJD6 in pan-cancer. A Correlation between JMJD6 mRNA expression and DNA methylation levels from GSCA; B OS analysis based on JMJD6 DNA methylation status from GSCA; C Correlation between JMJD6 expression and RNA modifications regulators (m1A, m5C, m6A) via SangerBox

Additionally, JMJD6 expression also correlated with various RNA modification regulators (Fig. 4C). For instance, it showed a positive correlation with the m5C reader ALYREF in all cancer types, while in THYM, DLBC and PRAD, it correlated negatively with several regulators. These results suggest JMJD6 may participate in post-transcriptional regulatory networks influencing tumor progression.

JMJD6 expression in relation to immune infiltration and genomic heterogeneity

Analysis via SangerBox indicated that JMJD6 expression showed diverse correlations with tumor-infiltrating immune cells across different cancer types (Fig. 5). Notably, it was significantly negatively correlated with the infiltration levels of several anti-tumor immune cells, including resting memory CD4 + T cells, gamma delta T cells, and activated NK cells. Conversely, a positive correlation was observed with regulatory T cells (Tregs), which are known to suppress anti-tumor immunity. Furthermore, JMJD6 expression positively correlated with several immune regulatory molecules, including the checkpoint genes PDCD1, CD276 and LAG3, as well as immunosuppressive cytokines/growth factors such as TGFB1, VEGFA and VEGFB (Fig S3). implying a role in immune evasion and a potential predictive value for immunotherapy response. The strength and direction of these associations were quantified using Pearson’s correlation coefficients (r), and statistical significance was assessed using FDR-adjusted p-values.

Fig. 5.

Correlation between JMJD6 expression and immune infiltration. Correlation between JMJD6 expression and 22 tumor-related immune cell infiltration calculated with the CIBERSORT algorithm via SangerBox. * p < 0.05. Correlation analyses were performed using Pearson’s correlation coefficients (r), and statistical significance was assessed using FDR-adjusted p-values. * FDR < 0.05

Tumor Mutational Burden (TMB), Microsatellite Instability (MSI) and Homologous Recombination Deficiency (HRD) are important signature of genomic heterogeneity which is closely related to tumor immunotherapy [23, 24]. Our results showed JMJD6 expression was positively correlated with TMB, MSI and HRD in most cancers (Fig. 6A-C). Given that HRD status influences sensitivity to PARP inhibitors [25]. we further analyzed drug response data from GDSC. Correlation strengths were evaluated using Pearson’s correlation coefficients (r), and statistical significance was determined based on FDR-adjusted p-values to account for multiple testing.

Fig. 6.

JMJD6 expression in relation to genomic heterogeneity and drug sensitivity. Correlation of JMJD6 expression with (A) TMB, (B) MSI and (C) HRD and (D) sensitivity to antitumor drugs from CTRP via GSCA. Correlations were evaluated using Pearson’s correlation coefficients (r), and statistical significance was determined based on FDR-adjusted p-values. * FDR < 0.05

High JMJD6 expression was associated with altered sensitivity to several anticancer compounds, including a positive correlation with resistance to multiple EGFR inhibitors (afatinib, gefitinib, erlotinib, and PD153035). These associations were derived from pharmacogenomic data obtained from the GDSC database and analyzed via the GSCA platform. Given the correlative and cell line–based nature of these data, these findings should be interpreted as exploratory and hypothesis-generating rather than as evidence of a direct causal role of JMJD6 in mediating drug resistance. Nevertheless, they raise the possibility that JMJD6 expression status may be linked to therapeutic response heterogeneity and warrant further experimental validation in controlled in vitro and in vivo models.

Functional enrichment of JMJD6-associated partners

Finally, we screened out the JMJD6-associated partners for a series of pathway enrichment analyses to understand the molecular mechanism of JMJD6 in carcinogenesis and progression. GEPIA2 database was applied to acquire the top 100 JMJD6 co-expressed genes in pan-cancer, NUP85, CCDC137, ALYREF, TSEN54, and ANAPC11 displayed strong correlations across cancers (Fig. 7A and B). Among these, CCDC137—an RNA-binding protein, ALYREF—an RNA export factor also m5C reader, and ANAPC11—a core subunit of the anaphase-promoting complex (APC/C), all linked to tumor progression and poor prognosis [26–30]. In addition, protein-protein interaction (PPI) network analysis via STRING revealed JMJD6 partners are mainly involved in RNA splicing (U2AF2, PUF60), chromatin remodeling (DEK, CHD4), and cell cycle regulation (CDK9, CCNT1) (Fig. 7C). KEGG enrichment highlighted the spliceosome pathway, while GO analysis indicated enrichment in chromatin remodeling, RNA splicing, ribosomal complexes, and histone demethylase activity. These results align with known JMJD6 functions and suggest spliceosome dysregulation and chromatin remodeling as key oncogenic mechanisms.

Fig. 7.

Enrichment analysis of JMJD6-associated partners. A, B Top 5 JMJD6-coexpressed genes and their pan-cancer correlations from GEPIA2. C PPI network of top 50 JMJD6-interacting proteins from STRING. D KEGG and E GO enrichment analyses of the top 100 JMJD6-coexpressed genes and top 100 JMJD6-interacting proteins

Functional enrichment analysis of JMJD6 co-expressed genes

To obtain deeper mechanistic insights into JMJD6-associated biological functions beyond individual pathway annotations, we performed Gene Ontology (GO) enrichment analysis based on genes most strongly co-expressed with JMJD6 across pan-cancer cohorts. The top 200 and top 500 JMJD6-correlated genes were selected for functional enrichment analyses. In the Biological Process (BP) category, enriched terms were predominantly related to RNA processing and splicing, chromatin organization, transcriptional regulation, DNA damage response, and cell cycle progression, including “RNA splicing,” “mRNA processing,” “chromatin modification,” “DNA-templated transcription,” and “mitotic cell cycle.” In the Molecular Function (MF) category, significantly enriched terms included “RNA binding,” “histone binding,” “chromatin binding,” “DNA-binding transcription factor activity,” and “protein serine/threonine kinase activity,” highlighting the involvement of JMJD6-associated gene networks in nucleic acid– and chromatin-related regulatory functions. In the Cellular Component (CC) category, enriched terms were mainly localized to nuclear and chromatin-associated compartments, such as “nucleus,” “chromatin,” “spliceosomal complex,” “nucleoplasm,” and “ribonucleoprotein complex.” Collectively, these GO enrichment results indicate that JMJD6 co-expressed genes are functionally concentrated in transcriptional regulation, RNA splicing, chromatin remodeling, and cell cycle–related processes, providing a more systematic functional context for JMJD6-associated transcriptional programs at the pan-cancer level.

Discussion

Epigenetic dysregulation is a hallmark of cancer and is commonly characterized by aberrant expression and activity of epigenetic modifiers [3]. As a member of the JmjC domain–containing histone demethylase family, JMJD6 has been reported to be upregulated in multiple malignancies, including breast cancer, lung cancer, melanoma, hepatocellular carcinoma, and colorectal cancer, and to be associated with pro-tumorigenic phenotypes [17, 18]. These prior studies have provided important evidence supporting a role for JMJD6 in cancer development and progression.

Building on these observations, the present pan-cancer study provides an integrative characterization of JMJD6 expression patterns, clinical associations, and multi-omics correlates across diverse tumor types using publicly available datasets. Rather than establishing a causal oncogenic role, our analysis systematically evaluates the associations between JMJD6 expression and key molecular, immunological, and clinical features, thereby extending previous findings within a broader pan-cancer framework. Our results corroborate earlier reports showing that JMJD6 is frequently overexpressed at both the mRNA and protein levels in a majority of cancer types. In addition, elevated JMJD6 expression was significantly associated with advanced tumor stage and unfavorable clinicopathological characteristics across multiple cancers. Furthermore, higher JMJD6 expression correlated with poorer outcomes across several survival endpoints, including overall survival (OS), disease-specific survival (DSS), progression-free interval (PFI), and disease-free interval (DFI). Together, these consistent pan-cancer associations suggest that JMJD6 expression is closely linked to tumor progression and adverse clinical outcomes. While these findings support the potential utility of JMJD6 as a prognostic biomarker, they remain correlative in nature and do not establish JMJD6 as a direct oncogenic driver. Instead, they position JMJD6 as a candidate molecular marker associated with aggressive disease biology whose functional and mechanistic roles warrant further experimental validation.

Although JMJD6 exhibited a predominantly oncogenic expression pattern across most cancer types, we observed notable exceptions in which JMJD6 was downregulated, such as in BLCA, KICH, and PCPG, and in which low JMJD6 expression was associated with better overall survival in PAAD. These findings suggest that the biological functions of JMJD6 may be context dependent and influenced by tissue-specific regulatory landscapes. One possible explanation is that JMJD6 engages in distinct protein–protein interaction networks across different tissue types, leading to divergent downstream effects. In certain cellular contexts, JMJD6 may preferentially interact with tumor-suppressive partners or participate in regulatory complexes that restrain oncogenic signaling, thereby attenuating its pro-tumorigenic potential. In addition, JMJD6 undergoes alternative splicing, generating multiple transcript isoforms with potentially different enzymatic activities, subcellular localizations, or substrate specificities. Tissue-specific expression of these isoforms may contribute to the observed heterogeneity in JMJD6 expression patterns and prognostic associations across cancer types. Furthermore, post-translational modifications of JMJD6, such as acetylation, phosphorylation, or hydroxylation, have been reported to modulate its stability, catalytic activity, and substrate interactions. Variations in the activity of upstream modifying enzymes across tissues may therefore alter JMJD6 function in a context-dependent manner. Together, these considerations indicate that JMJD6 may exert dual or even opposing biological roles depending on cellular context, molecular co-factors, and regulatory states. Elucidating these tissue-specific mechanisms will require systematic investigation of JMJD6 isoform usage, post-translational modification profiles, and interactomes across different tumor types.

Several recent pan-cancer studies published between 2021 and 2024 have reported that JMJD6 is frequently overexpressed across multiple tumor types and is associated with unfavorable prognosis and immune-related features. These prior works have laid an important foundation for understanding the oncogenic relevance of JMJD6 at a pan-cancer level. Compared with these earlier studies, the present work does not aim to merely replicate known expression–prognosis associations, but rather to provide a more integrative and functionally oriented extension of previous findings. First, we incorporated proteomic data from the CPTAC dataset to validate JMJD6 overexpression at the protein level, thereby strengthening the translational relevance of our conclusions beyond transcriptomic analyses alone. Second, we systematically linked JMJD6 expression to multiple layers of genomic instability, including tumor mutational burden (TMB), microsatellite instability (MSI), and homologous recombination deficiency (HRD), which have not been comprehensively explored in prior JMJD6 pan-cancer studies. Third, we extended the clinical relevance of JMJD6 by integrating large-scale drug sensitivity data from the GDSC database and demonstrated that high JMJD6 expression is associated with resistance to EGFR inhibitors and altered sensitivity to several other anticancer compounds. This pharmacogenomic dimension provides a direct therapeutic context that is largely absent from earlier reports. Finally, by jointly analyzing DNA methylation, RNA modification regulators, immune infiltration, immune checkpoint genes, and spliceosome-related co-expression networks, our study offers a more unified multi-omics framework that links epigenetic regulation, transcriptomic processing, immune microenvironment remodeling, and therapeutic response to JMJD6 activity. Together, these additional datasets and analytical dimensions position the present study as a substantive extension rather than a duplication of prior pan-cancer analyses, and they highlight new mechanistic and translational implications of JMJD6 that were not previously delineated.

Notably, we identified DNA hypomethylation as a likely mechanism driving JMJD6 overexpression in tumors. The significant negative correlation between JMJD6 mRNA expression and its promoter methylation levels, coupled with the favorable prognosis associated with hypermethylation, suggests that epigenetic derepression constitutes a key regulatory layer in JMJD6 activation. Beyond DNA methylation, JMJD6 expression also correlated with numerous RNA modification regulators—most notably the m5C reader ALYREF. This aligns with emerging evidence that JMJD6 can modulate epitranscriptomic pathways; for instance, it was recently shown to regulate the METTL14/m6A/SLC3A2 axis to suppress ferroptosis in lung cancer [31]. These observations position JMJD6 within a broader post-transcriptional regulatory network, offering new mechanistic avenues for understanding its oncogenic functions.

A particularly compelling dimension of JMJD6’s functionality lies in its impact on the tumor immune microenvironment. Prior studies indicate that JMJD6 can promote M2 macrophage polarization and suppress anti-tumor immunity through ANXA1 downregulation or activation of the STAT3/IL-10 axis [32, 33]. Our analysis extends these findings by demonstrating negative correlations between JMJD6 expression and infiltration levels of multiple immune cell subsets, including CD4⁺ T cells and NK cells. Moreover, JMJD6 expression was positively associated with immunosuppressive molecules such as VEGFA, TGFB1 and PDCD1, suggesting its involvement in establishing an immunosuppressive niche. These results imply that JMJD6 may facilitate immune evasion through dual mechanisms: by shaping immune cell composition and by regulating checkpoint molecule expression. Consequently, JMJD6 represents a promising predictive biomarker for immunotherapy response, and its targeting may synergize with immune checkpoint blockade strategies. Although JMJD6 expression was found to be positively correlated with multiple immune checkpoint genes, including PDCD1, LAG3, and CD276, it should be emphasized that these associations are correlative in nature and do not establish a direct regulatory relationship. At present, it remains unclear whether JMJD6 directly modulates the transcription or epigenetic status of these immune checkpoint genes, or whether their co-expression simply reflects broader changes in the tumor immune microenvironment. One plausible explanation is that high JMJD6 expression marks tumors with a more immunosuppressive or immune-infiltrated phenotype, in which immune checkpoint genes are upregulated as a secondary consequence of chronic immune activation or T-cell exhaustion. In this scenario, JMJD6 may indirectly associate with checkpoint gene expression by shaping the cellular composition or functional state of the tumor microenvironment, for example through its effects on macrophage polarization, cytokine signaling, or angiogenic pathways. Alternatively, JMJD6 may influence immune checkpoint expression more directly via its established roles in chromatin remodeling and RNA splicing. JMJD6 has been shown to regulate transcriptional elongation, histone arginine demethylation, and alternative splicing of multiple oncogenic and immune-related transcripts, raising the possibility that it could modulate checkpoint gene expression or stability at the epigenetic or post-transcriptional level. However, experimental evidence supporting such a direct mechanism is currently lacking. Therefore, the observed correlations between JMJD6 and immune checkpoint genes in this pan-cancer analysis should be interpreted with caution. Future studies integrating chromatin immunoprecipitation sequencing, perturbation experiments (e.g., JMJD6 knockdown or overexpression), and single-cell or spatial transcriptomic profiling will be required to disentangle whether JMJD6 acts as a direct upstream regulator of immune checkpoint genes or primarily serves as a marker of immune-rich and immunosuppressive tumor microenvironments.

Although our findings suggest that JMJD6 expression is associated with tumor progression, immune regulation, and therapeutic response heterogeneity, the clinical translation pathway of JMJD6 remains incompletely defined and requires careful consideration. At present, JMJD6 should be regarded as a candidate biomarker rather than a clinically validated diagnostic or therapeutic target. From a practical standpoint, several detection modalities could be employed to assess JMJD6 expression in clinical specimens. Immunohistochemistry (IHC) represents the most feasible and widely used approach for evaluating JMJD6 protein abundance and subcellular localization in formalin-fixed paraffin-embedded tumor tissues. Alternatively, reverse transcription quantitative PCR (RT-qPCR) or RNA sequencing could be used to quantify JMJD6 mRNA levels in fresh or frozen tissue samples, although these approaches are less amenable to routine clinical workflows. Beyond tissue-based assays, JMJD6 may also hold potential for non-invasive detection using liquid biopsy approaches. Given that tumor-derived RNA species can be detected in circulating cell-free RNA (cfRNA) and extracellular vesicles, future studies could explore whether JMJD6 transcripts or JMJD6-associated transcriptional signatures are detectable in plasma or serum samples. In addition, JMJD6 expression could potentially be evaluated in circulating tumor cells (CTCs) to enable dynamic monitoring of tumor burden or therapeutic response. With respect to therapeutic targeting, although JMJD6 possesses enzymatic activity as a JmjC domain-containing dioxygenase, the feasibility of pharmacologically inhibiting JMJD6 in patients remains uncertain. The development of selective JMJD6 inhibitors, together with careful assessment of on-target toxicity in normal tissues, will be essential before JMJD6 can be considered a viable drug target. Collectively, these considerations indicate that the primary near-term clinical utility of JMJD6 may lie in its potential role as a stratification or prognostic biomarker, rather than as an immediately actionable therapeutic target. Substantial experimental validation and prospective clinical studies will be required to establish the optimal detection modality, clinical context of use, and translational relevance of JMJD6.

Genomic instability markers such as tumor mutational burden (TMB), microsatellite instability (MSI), and homologous recombination deficiency (HRD) have emerged as important correlates of response to immunotherapy and selected targeted therapies [23–25]. In this study, we observed positive correlations between JMJD6 expression and these genomic instability markers in multiple cancer types. These associations suggest that elevated JMJD6 expression tends to co-occur with increased genomic instability, although they do not establish a direct role for JMJD6 in modulating genomic integrity.

The particularly strong association between JMJD6 expression and HRD is noteworthy, as it raises the hypothesis that JMJD6-high tumors may share molecular features with HRD-positive cancers, which are known to exhibit increased vulnerability to poly(ADP-ribose) polymerase (PARP) inhibition. However, this inference is based solely on correlative genomic data and does not demonstrate that JMJD6 expression itself confers sensitivity to PARP inhibitors. Prospective experimental and clinical validation will be required to determine whether JMJD6 expression has predictive value for PARP inhibitor responsiveness. Conversely, JMJD6 expression was negatively correlated with the estimated sensitivity to several epidermal growth factor receptor (EGFR) inhibitors, suggesting a potential association between high JMJD6 expression and reduced responsiveness to these targeted agents. These findings should be interpreted cautiously, as they are derived from in silico drug sensitivity analyses rather than from clinical treatment outcomes. Collectively, these observations indicate that JMJD6 expression is associated with distinct therapeutic response–related molecular features, supporting its potential utility as a hypothesis-generating stratification marker rather than as a validated predictor of treatment response. At the molecular level, integrative bioinformatics analyses revealed that JMJD6 co-expressed genes and predicted interacting partners are predominantly enriched in RNA splicing and chromatin remodeling–related pathways. This enrichment pattern is highly consistent with previously established molecular functions of JMJD6. Specifically, JMJD6 has been shown to physically interact with the splicing factor U2AF65, catalyze its hydroxylation, and cooperatively regulate alternative splicing of transcripts such as VEGFR1, PAK1, and GLS [12, 21, 34–36]. Through these reported interactions, JMJD6 has been implicated in the regulation of angiogenesis, metabolic programs, and migration-related signaling pathways. In the context of the present study, the enrichment of spliceosome- and chromatin-associated pathways among JMJD6-correlated gene networks provides functional coherence for the observed pan-cancer associations. These findings support the notion that JMJD6 expression is linked to transcriptional and post-transcriptional regulatory programs relevant to tumor biology, while remaining consistent with the correlative and hypothesis-generating nature of this pan-cancer analysis.

To further substantiate the biological relevance of JMJD6-associated transcriptional programs, we performed Gene Ontology (GO) enrichment analysis based on the top 200 and top 500 genes most strongly co-expressed with JMJD6 across pan-cancer cohorts. Importantly, these enrichment results were not limited to isolated pathway annotations but instead revealed coherent functional themes across all three GO categories (BP, MF, and CC). The dominant enrichment of RNA splicing, mRNA processing, chromatin organization, and transcriptional regulation terms provides systematic support for the proposed roles of JMJD6 in spliceosome function and chromatin remodeling. In addition, the enrichment of DNA damage response and cell cycle–related processes suggests that JMJD6-associated gene networks may also be linked to genomic stability and proliferative control. Together, these findings extend our earlier pathway-level observations and demonstrate that JMJD6 co-expression networks are functionally organized around nuclear regulatory processes central to tumor biology. These results strengthen the mechanistic plausibility of JMJD6 involvement in transcriptional and post-transcriptional regulation while remaining consistent with the correlative and hypothesis-generating nature of this pan-cancer analysis.

Despite these advances, several limitations merit consideration. First, as our study relied exclusively on bioinformatic analyses from public databases, future experimental validation using in vitro and in vivo models is essential to establish causality and elucidate detailed mechanisms. Second, while JMJD6 appears to function primarily as an oncogene, its putative tumor-suppressive properties and regulatory mechanisms in specific tumors await further exploration. Third, the drug sensitivity data derived from cell lines necessitate confirmation in clinical cohorts. Finally, the immune infiltration profiles inferred computationally should be validated using multiplex immunohistochemistry or spatial transcriptomics in actual tumor samples. Another important limitation of this study is that the immune infiltration analyses were entirely based on computational deconvolution algorithms, including TIMER and CIBERSORT, which infer immune cell proportions from bulk gene expression data. Although these tools are widely used and have been partially benchmarked against experimental measurements, they remain indirect estimates and are inherently sensitive to factors such as tumor purity, transcriptional heterogeneity, gene signature overlap among immune subsets, and platform-specific noise. As a consequence, the inferred associations between JMJD6 expression and immune cell infiltration levels should be interpreted cautiously and cannot be regarded as direct evidence of altered immune composition within the tumor microenvironment. In particular, the relative abundance estimates for functionally related immune populations such as CD4⁺ T cells, regulatory T cells (Tregs), and natural killer (NK) cells are subject to uncertainty when derived solely from bulk transcriptomic deconvolution. To directly validate these bioinformatics predictions, future studies should incorporate experimental approaches using tumor tissue specimens, such as multiplex immunofluorescence (mIHC) or flow cytometry, to quantitatively assess immune infiltration levels in tumors stratified by high versus low JMJD6 expression. Such validation would enable precise measurement of key immune subsets, including CD4⁺ T cells, Tregs, and NK cells, and would provide definitive evidence regarding whether JMJD6 expression is truly associated with immune cell recruitment or retention in the tumor microenvironment. Therefore, while the present findings support an association between JMJD6 expression and immune-related transcriptional features, they remain correlative and hypothesis-generating in nature. Experimental immune profiling will be essential to establish whether JMJD6 has a causal or functional role in shaping tumor immune landscapes.

A core limitation of the present study is the complete absence of direct experimental validation. All conclusions are derived from integrative bioinformatics analyses of publicly available multi-omics datasets and are therefore observational and correlative in nature. As such, the present findings do not establish causal relationships between JMJD6 expression and tumor progression, immune modulation, genomic instability, or therapeutic response. Consequently, the biological relevance and mechanistic interpretation of the observed associations remain provisional and require functional validation in appropriate experimental systems. In particular, future studies should incorporate loss- and gain-of-function experiments in representative cancer cell lines (e.g., breast, lung, or colorectal cancer), followed by phenotypic assays assessing proliferation, migration, invasion, apoptosis, and clonogenic capacity. In addition, quantitative PCR (qPCR) and immunohistochemistry (IHC) analyses in paired tumor and adjacent normal tissues will be essential to validate JMJD6 expression patterns at the mRNA and protein levels, respectively, and to confirm its subcellular localization in clinical specimens. Complementary in vivo models, such as xenograft or syngeneic mouse systems, will further be required to determine whether JMJD6 plays a causal role in tumor growth, metastatic dissemination, or therapeutic resistance. Therefore, we position the present work as an integrative, hypothesis-generating pan-cancer resource intended to systematically map JMJD6-associated molecular and clinical features, rather than as a definitive functional characterization of its oncogenic role.

A major limitation of the present study is the complete absence of direct experimental validation. All findings reported herein are derived from integrative in silico analyses of publicly available multi-omics datasets and therefore remain observational and correlative in nature. As such, the present results do not establish a causal oncogenic role for JMJD6 and should not be interpreted as definitive evidence that JMJD6 functions as an oncogenic driver. Consequently, the biological relevance and translational implications of the observed associations require experimental confirmation in clinically relevant tumor models. At a minimum, future studies should include quantitative PCR (qPCR) validation of JMJD6 expression in paired tumor and adjacent normal tissues, immunohistochemistry (IHC) analyses to assess protein abundance and subcellular localization in patient specimens, and functional perturbation experiments using JMJD6 knockdown or overexpression in representative cancer cell lines (e.g., breast, lung, or colorectal cancer). Such functional assays could evaluate the direct effects of JMJD6 on tumor cell proliferation, migration, invasion, and clonogenic potential. In addition, complementary in vivo xenograft or syngeneic mouse models would be required to determine whether JMJD6 plays a causal role in tumor growth and therapeutic resistance. Therefore, we position the present work as an integrative, hypothesis-generating pan-cancer resource that provides a systematic foundation for future mechanistic and translational investigations of JMJD6, rather than as a definitive functional characterization of its oncogenic role.

An important limitation of the present study is the absence of single-cell–level resolution. All expression analyses of JMJD6 were performed using bulk transcriptomic and proteomic datasets, which preclude precise delineation of the specific cellular compartments responsible for JMJD6 overexpression within the tumor microenvironment. As a result, it remains unclear whether elevated JMJD6 expression predominantly originates from malignant epithelial cells, stromal fibroblasts, endothelial cells, or infiltrating immune subsets. Given the growing recognition that tumor heterogeneity and cell type–specific gene regulation critically shape cancer progression and therapeutic response, single-cell RNA sequencing (scRNA-seq) represents an essential next step for refining the biological interpretation of our findings. Integrating scRNA-seq datasets from public resources such as GEO, CELLxGENE, and the Human Cell Atlas would enable direct assessment of JMJD6 expression across malignant, stromal, and immune cell populations and would clarify its functional context within the tumor microenvironment. In particular, single-cell–level profiling could help determine whether JMJD6 overexpression is primarily driven by tumor cells themselves or reflects expansion or activation of specific stromal or immune compartments. Such analyses would also facilitate more accurate mapping of JMJD6-associated transcriptional programs, immune modulatory effects, and cell type–specific regulatory networks. Therefore, while the present bulk-level pan-cancer analysis provides a global overview of JMJD6 expression patterns and clinical associations, future studies incorporating scRNA-seq and spatial transcriptomic approaches will be essential to validate the cellular sources of JMJD6 expression and to elucidate its context-dependent biological roles in cancer.

Beyond immune cell infiltration, a more comprehensive understanding of tumor immune regulation requires evaluating the associations between oncogenic factors and broader classes of immune modulators. In this study, we extended our immune analysis to include correlations between JMJD6 expression and key immune regulatory genes, including immune checkpoint molecules (e.g., PDCD1, CD274, CTLA4, LAG3, and TIGIT), cytokines and chemokines (e.g., CXCL9, CXCL10, IL6, and TGFB1), and components of the antigen presentation machinery (e.g., HLA class I genes and B2M). These analyses revealed that JMJD6 expression is broadly associated with immunoregulatory programs across multiple cancer types, supporting the notion that JMJD6 may be linked to immune microenvironment remodeling. Importantly, these findings are consistent with recent pan-cancer work [37], which demonstrated that epigenetic regulators can coordinately shape immune checkpoint expression, antigen presentation capacity, and T-cell–inflamed phenotypes across tumors. Together, these observations position JMJD6 within a broader immuno-epigenetic regulatory framework and further support its potential relevance to immune evasion and immunotherapy responsiveness, while emphasizing that these associations remain correlative and require mechanistic validation.

The absence of direct experimental validation represents an important limitation of the present study. Although our integrative pan-cancer analyses reveal consistent associations between JMJD6 expression and multiple molecular, immunological, and clinical features, these findings remain descriptive and correlative in nature. As such, they do not establish causality or define the mechanistic directionality of the observed relationships. Several biological interpretations proposed in this study—including the links between JMJD6 and immune modulation, genomic instability, and therapeutic response—are therefore best viewed as hypothesis-generating rather than mechanistically proven. While parts of these interpretations are supported by prior literature on JMJD6-mediated chromatin remodeling, transcriptional regulation, and RNA splicing, direct experimental evidence connecting JMJD6 to these specific pan-cancer phenotypes is currently lacking. Consequently, minimal laboratory confirmation will be required to substantiate key inferences drawn from the computational analyses. In particular, in vitro perturbation experiments using JMJD6 knockdown or overexpression models could be employed to assess its direct effects on tumor cell proliferation, immune-related gene expression, genomic instability markers, and sensitivity to anticancer agents. Complementary in vivo models may further clarify whether JMJD6 plays a causal role in tumor growth, immune evasion, and therapeutic resistance. In light of these considerations, we have moderated the interpretation of correlation-based findings throughout the manuscript and avoided implying direct causation. Future functional studies integrating molecular perturbation, epigenomic profiling, and pharmacological assays will be essential to validate the biological relevance of JMJD6 and to translate the present descriptive associations into mechanistically grounded insights.

It should be noted that the immune infiltration analyses in this study were based on computational deconvolution using the CIBERSORT algorithm implemented via the SangerBox platform. Although such approaches enable large-scale inference of immune cell composition from bulk RNA-sequencing data, they are subject to several inherent limitations. In particular, deconvolution accuracy depends on the quality of reference gene expression signatures and may be affected by tumor purity, stromal content, and transcriptional heterogeneity across cancer types. Moreover, closely related immune cell subsets (e.g., different T-cell activation states or macrophage polarization phenotypes) are often difficult to distinguish with high resolution using bulk transcriptomic data alone, which may limit cell type specificity and lead to partial misclassification of immune populations. As a result, the inferred abundance of certain immune cell subsets should be interpreted as approximate estimates rather than precise quantitative measurements. Therefore, while the observed associations between JMJD6 expression and immune cell infiltration provide valuable insights into its potential immunological relevance, these findings require independent validation. Future studies integrating orthogonal datasets and experimental approaches such as single-cell RNA sequencing, spatial transcriptomics, multiplex immunohistochemistry, or flow cytometry will be essential to confirm the cellular sources of JMJD6 expression and to validate its relationships with specific immune cell populations in the tumor microenvironment.

Conclusion

In summary, this pan-cancer analysis establishes JMJD6 as a multi-functional oncoprotein linked to aggressive clinicopathological features, poor prognosis, immune microenvironment remodeling, and therapy resistance. Its expression is regulated epigenetically via DNA hypomethylation and likely promotes tumorigenesis through spliceosome disruption and RNA processing dysregulation. These findings nominate JMJD6 as a promising prognostic biomarker and a candidate therapeutic target. Combinatorial strategies, such as JMJD6 inhibition alongside EGFR-targeting agents or immune checkpoint blockers, may offer novel avenues for precision therapy. Future work should prioritize functional validation of these mechanisms and prospective clinical assessment of JMJD6’s utility as a predictive biomarker.

Methods

Gene expression analysis

The subcellular localization and tissue-specific expression profiles of JMJD6 were acquired from The Human Protein Atlas (HPA) (https://www.proteinatlas.org/). Differential expression of JMJD6 mRNA between tumor and normal tissues was assessed using TCGA data through the UALCAN platform (http://ualcan.path.uab.edu/index.html). Protein-level expression differences were analyzed using the CPTAC module within UALCAN.

Clinicopathological and survival analysis

The association between JMJD6 expression and tumor clinicopathological stages was evaluated via UALCAN. Prognostic implications of JMJD6 expression, including overall survival (OS), disease-specific survival (DSS), disease-free interval (DFI), and progression-free interval (PFI), were examined using Kaplan–Meier survival curves using SangerBox tools (http://sangerbox.com/). Hazard ratios (HRs) and 95% confidence intervals were derived from univariate Cox regression models. To account for multiple testing across cancer types and survival endpoints, p-values derived from survival analyses were adjusted using the false discovery rate (FDR) method. FDR-adjusted p-values ≤ 0.05 were considered statistically significant.

Methylation and RNA modification analysis

Correlations between JMJD6 mRNA expression and DNA methylation levels across 33 cancer types were investigated using the Gene Set Cancer Analysis (GSCA) platform (https://guolab.wchscu.cn/GSCA/#/). Pearson’s correlation analysis was applied, with false discovery rate (FDR) ≤ 0.05 considered statistically significant. Associations between JMJD6 promoter methylation and patient OS were further evaluated via GSCA. Additionally, correlations between JMJD6 expression and regulators of RNA modifications (m1A, m5C, m6A) were analyzed via Pearson’s correlation and visualized using SangerBox tools. For all correlation analyses, including those involving immune infiltration, immune checkpoint gene expression, and genomic heterogeneity markers (TMB, MSI, and HRD), both Pearson’s correlation coefficients (r) and corresponding p-values were calculated. Where multiple comparisons were performed, p-values were adjusted using the false discovery rate (FDR) method, and FDR ≤ 0.05 was considered statistically significant.

Immune infiltration, genomic heterogeneity, and drug sensitivity analysis

The SangerBox platform was employed to assess relationships between JMJD6 expression and tumor-infiltrating immune cells, as well as immune checkpoint gene expression, based on data from the Tumor Immune Estimation Resource (TIMER) database (http://timer.cistrome.org/). Pearson’s correlation analysis was used to evaluate associations between JMJD6 expression and genomic heterogeneity markers, including tumor mutational burden (TMB), microsatellite instability (MSI), and homologous recombination deficiency (HRD). Drug sensitivity data were derived from the Genomics of Drug Sensitivity in Cancer (GDSC) database and accessed through the “Drug” module of the GSCA platform (https://guolab.wchscu.cn/GSCA/). The half-maximal inhibitory concentration (IC50) values of anticancer compounds across cancer cell lines were used to evaluate drug response. Pearson’s correlation analysis was performed to assess the associations between JMJD6 expression and drug sensitivity (IC50 values). A false discovery rate (FDR) ≤ 0.05 was considered statistically significant.

Gene enrichment analysis

The top 100 genes co-expressed with JMJD6 across cancers were identified using GEPIA2 (http://gepia2.cancer-pku.cn/). Pearson’s correlation analysis was performed for the top five correlated genes. JMJD6-interacting proteins were retrieved from the STRING database (https://cn.string-db.org/), and a protein–protein interaction (PPI) network was constructed for the top 50 candidates. Functional enrichment analyses, including KEGG pathway and Gene Ontology (GO) analyses, were conducted using WebGestalt (https://www.webgestalt.org/). Terms with p < 0.05 and FDR ≤ 0.05 were considered statistically significant.

To further explore the functional relevance of JMJD6-associated transcriptional programs, a pan-cancer co-expression analysis was performed. Pearson’s correlation coefficients were calculated between JMJD6 expression and all protein-coding genes across TCGA pan-cancer samples. Genes were ranked according to the absolute value of the correlation coefficient, and the top 200 and top 500 genes most strongly correlated with JMJD6 were selected for downstream functional enrichment analyses. Gene Ontology (GO) enrichment analysis, including Biological Process (BP), Molecular Function (MF), and Cellular Component (CC) categories, was conducted using the clusterProfiler package in R. Enriched GO terms with a false discovery rate (FDR)–adjusted p-value ≤ 0.05 were considered statistically significant.

Author contributions

LG: conceived and designed the study; LL, HJ: collected and analyzed data; LG, LT: wrote the manuscript. All authors reviewed, edited, and approved the final version.

Funding

This work was supported by the Natural Science Foundation of Jiangxi Province (20242BAB21043) and the Early-Career Young Scientists and Technologists Project of Jiangxi Province (20244BCE52223).

Data availability

The datasets used in this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.