HMCN1 as a conserved biomarker of epithelial–mesenchymal transition: a cross-cancer analysis

Background: The extracellular matrix (ECM) critically regulates tumor progression, but the systematic role of its large constituent hemicentin-1 (HMCN1) across cancers remains poorly defined. Although implicated in cell invasion and migration, a comprehensive pan-cancer understanding of HMCN1 is lacking.

Methods: We performed an integrative multi-omics analysis of HMCN1 across 33 cancer types using data from TCGA, GTEx, and CPTAC. Based on prominent functional importance in bone tumors, osteosarcoma was selected for functional validation. In vitro experiments, including knockdown and overexpression, were conducted to assess effects on migration and invasion.

Results: HMCN1 was frequently upregulated in tumors and associated with poor prognosis. Its mutations correlated with genomic instability. HMCN1-high tumors were enriched in inflammatory immune subtypes and exhibited a dysfunctional tumor microenvironment. Pathway analysis consistently linked HMCN1 expression to epithelial-mesenchymal transition (EMT). Functional validation in osteosarcoma confirmed that HMCN1 knockdown suppressed, while its overexpression enhanced, cell migration and invasion through regulation of key EMT markers.

Conclusions: This study identifies HMCN1 as a novel, conserved regulator of EMT and a candidate prognostic biomarker across cancers. The findings in osteosarcoma provide a foundational rationale for developing HMCN1 as a potential therapeutic target.

1 Introduction

Cancer, a major global public health challenge arising from complex genetic and epigenetic regulatory networks, is characterized by high degrees of heterogeneity and molecular complexity, which result in an incomplete understanding of its regulatory mechanisms. In recent years, advancements in high-throughput sequencing technologies and the establishment of large-scale data-sharing platforms have positioned pan-cancer research as a key paradigm for systematically uncovering the commonalities and heterogeneity among cancers, and these technologies have led to remarkable progress (1). By integrating genomic, transcriptomic, epigenomic, and clinical data across multiple cancer types, this approach has the potential to identify shared driver mutations, signaling pathways, and potential therapeutic targets, and thereby offers new perspectives for early cancer detection, prognostic stratification, and personalized treatment. Despite these advantages, there exist certain challenges with integrating multi-omics data and interpreting their complex biological significance (2–4). Identifying key molecules with roles across multiple cancers remains a crucial goal of pan-cancer research. Such molecules can reveal previously unrecognized pan-cancer mechanisms and hold promise as broad-spectrum biomarkers or therapeutic targets (5). Further, validating their universality and effectiveness as therapeutic targets will provide a critical theoretical foundation for developing universal or adaptable targeted therapies for a variety of malignancies, as well as offer significant translational value, particularly in drug discovery and combination therapy strategies. Therefore, the identification and validation of such functional hubs or molecular markers would not only represent a crucial breakthrough in understanding the “common rules” of cancer biology but also drive the construction of a cross-cancer precision medicine framework and innovate clinical practice models. The identification of pan-cancer prognostic biomarkers represents a major focus in cancer research. Such biomarkers refer to molecular features significantly associated with patient outcomes (e.g., overall survival and progression-free survival) across different cancer types. Investigating these biomarkers aims to transcend the limitations of single-cancer-type studies, reveal shared cancer biological processes, and thereby provide a scientific basis for developing universal prognostic tools, guiding personalized treatment strategies, and improving clinical outcomes for cancer patients (6, 7).

Tumorigenesis and tumor progression are driven by genetic mutations and signaling aberrations within cancer cells, as well as by complex interactions among diverse cellular and molecular components of the tumor microenvironment (TME). Notably, the extracellular matrix (ECM) constitutes a fundamental and frequently dysregulated compartment of the TME, making it a prime source for the pan-cancer biomarkers discussed above. The extracellular matrix (ECM), a core constituent of the TME, provides structural support and actively participates in regulating cell behavior and tissue homeostasis in normal tissues. However, in tumors, its composition, structure, and mechanical properties undergo significant alterations, including aberrant expression, increased cross-linking, and enhanced degradation activity (8, 9). Collectively, these changes promote tumor cell proliferation, migration, and invasion, while also influencing immune cell infiltration and function, thereby fostering a microenvironment conducive to tumor progression (10, 11). However, given the vast complexity and heterogeneity of the ECM across different cancers, a systematic, pan-cancer approach is essential to identify which specific ECM components act as conserved regulators of malignancy. Such an analysis can pinpoint central nodes within the tumor ECM that hold promise as broad-spectrum biomarkers or therapeutic targets. Hemicentin-1 (HMCN1), a prominent member of the fibulin family distinguished by its large molecular weight and widespread distribution in the basement membranes of various tissues, has emerged as a key molecule in ECM research.

HMCN1 was initially identified in Caenorhabditis elegans for its role in tethering adjacent basement membranes and was, subsequently, proven to play a crucial role in maintaining the integrity of the basement membrane-connective tissue boundary through studies on HMCN1-deficient mice (12). Further, functional studies have demonstrated that HMCN1 directly binds to Keratin 14, and its loss was found to reduce keratin intermediate filament formation and induce subepidermal blistering in 3D skin equivalent models (13). This underscores its vital function in linking the epidermal cytoskeleton to the ECM. Emerging evidence indicates an active role for HMCN1 in tumor progression. For example, Liu et al. reported co-upregulation of sushi repeat−containing protein, X−linked (SRPX) and HMCN1 in cancer-associated fibroblasts from ovarian cancer patients. Functionally, knocking down HMCN1 significantly suppressed Transwell invasion of cancer cells, a mechanism shown to be mediated through the activation of the RhoA signaling pathway in fibroblasts (14). Furthermore, in bovine leukemia virus infection models, HMCN1 upregulation was associated with tumor cell migration (15). Together, this evidence suggests that HMCN1 exerts multifaceted regulatory functions in the TME that extend beyond mere structural support.

However, a critical and unresolved question remains: does HMCN1 act as a conserved, central regulator of cancer progression across diverse tumors, or merely as a context-dependent passenger? Establishing HMCN1’s fundamental role is of paramount scientific importance because, as a large ECM glycoprotein, it is uniquely positioned at the interface between the cellular cytoskeleton and the basement membrane. This strategic location suggests it could be a master orchestrator of tissue architecture and cell signaling. Clinically, resolving this ambiguity is crucial. If HMCN1 is a conserved oncogenic driver, it would emerge as a novel and compelling therapeutic target, particularly for aggressive, metastatic cancers driven by EMT and immune evasion – two processes it appears to influence. Currently, few ECM components have been successfully targeted, making HMCN1 a promising candidate for pioneering new classes of stromal therapy. Critically, it is unknown whether HMCN1 represents a conserved oncogenic driver across cancers or a context-dependent factor, and its potential role as a master regulator of central processes like epithelial-mesenchymal transition (EMT) is poorly defined. To test the hypothesis that HMCN1 is a pan-cancer tumor-promoting gene that promotes malignancy primarily through activating EMT, the present study conducted a comprehensive analysis of HMCN1 characteristics across 33 cancer types by integrating data from The Cancer Genome Atlas (TCGA), Genotype-Tissue Expression (GTEx), and DepMap, as well as single-cell and spatial transcriptomic atlases. We delineated the genomic, transcriptomic, and proteomic landscapes of HMCN1 and used the resulting data to evaluate its prognostic value and immunomodulatory functions across cancer types. Given the mesenchymal origin of osteosarcoma and its well-characterized reliance on ECM remodeling and EMT for progression, coupled with our preliminary data indicating high HMCN1 expression and functional essentiality in bone cancer contexts, we selected this aggressive tumor type for subsequent experimental validation of HMCN1’s oncogenic role (16–18). Our research demonstrates its tumor-promoting properties based on published pan-cancer data as well as in vitro experiments in an osteosarcoma model. Importantly, HMCN1 emerged as a novel regulator of epithelial-mesenchymal transition (EMT) in the cancer milieu. Thus, these findings establish HMCN1 as a pan-cancer ECM-derived biomarker and a potential therapeutic target and provide a crucial foundation for future research into HMCN1 and related ECM mechanisms.

2 Methods

2.1 Bibliometric analysis

A comprehensive literature search was conducted using the Web of Science Core Collection (WOSCC) database to systematically review studies on HMCN1 in the field of oncology. The following Boolean search strategy was employed: [(HMCN1 OR Hemicentin-1 OR “Hemicentin 1” OR FIBL-6 OR FIBL6 OR FBLN6 OR ARMD1) AND (cancer* OR carcinoma* OR tumor* OR tumour* OR neoplas* OR malignan* OR oncolog*)]. The search was limited to articles published between January 1, 2000, and September 1, 2025. The article types included were research articles and reviews. Comments, letters, conference abstracts, and retracted publications were excluded, and only English-language documents considered. Visualizations were generated to highlight trends in keywords and their expansions within the HMCN1 research theme (19).

2.2 Pan-cancer data sources and data preprocessing

Transcriptomic data and corresponding clinical information, including overall survival (OS), progression-free interval (PFI), disease-free interval (DFI), and disease-specific survival (DSS), were retrieved from TCGA. To compare HMCN1 expression between tumor and normal tissues, we also obtained matched normal tissue data from the GTEx project and integrated the cancer and normal datasets. Specifically, we utilized the TPM (transcripts per million) expression data from the UCSC Xena database (datasets: tcga_RSEM_gene_tpm and gtex_RSEM_gene_tpm). To ensure accuracy and account for anatomical factors, only primary tumor tissues from TCGA were paired with GTEx data. The expression data were then standardized by converting them into a unitless Z-score for each cancer type using the formula (x-μ)/σ. Outliers, defined as Z-scores less than -3 or greater than 3, were removed. A cancer type was included in the subsequent analysis only if at least three normal samples remained after outlier removal. Samples with incomplete clinical data, poor sequencing quality, or undetectable HMCN1 expression were excluded to ensure data integrity and comparability. Using the integrated TCGA and GTEx dataset, we systematically compared differences in HMCN1 expression between normal and malignant tissues across various cancer types.

Protein-level expression of HMCN1 was assessed using the Human Protein Atlas platform, and three-dimensional structural information was retrieved from Protein Data Bank. Gene dependency or essentiality was assessed using CRISPR screening data from the DepMap portal. Gene essentiality was quantified using the Chronos score: a score <0 indicates that HMCN1 knockdown is detrimental to cell survival and is suggestive of its potential role in tumorigenesis (20, 21). Pan-cancer single-cell RNA sequencing data were obtained from the TISCH database (http://tisch.compbio.cn/home/). Using the pheatmap R package, we generated a pan-cancer single-cell expression heatmap. Euclidean distance was used as the similarity metric, and Ward’s minimum variance method was employed for hierarchical clustering to investigate conserved patterns of gene expression origin. Further dimensionality reduction and visualization of the high-dimensional data were performed using the Uniform Manifold Approximation and Projection method (22).

Spatial transcriptomics data were sourced from the Sparkle database [with sample identifiers as previously described (23)]. For the spatial transcriptomics analysis, we first identified each microregion characterized by a dominant cell type and detected its gene expression profile. After Z-score normalization, the data were visualized using the pheatmap package. Next, the spatial distribution of major cell types across microregions was computed and visualized using the SpatialPlot function of the Seurat package. Finally, Spearman correlation analysis was used to determine the correlation between cellular composition and gene expression data across all the detected spots with the linkET package, and the results were visualized to illustrate correlation patterns.

2.3 Analysis of the diagnostic and prognostic potential of HMCN1

The diagnostic utility of HMCN1 was evaluated by constructing receiver operating characteristic (ROC) curves using the pROC package, and its prognostic value was assessed through Kaplan-Meier survival analysis (survival and survminer packages) and univariate/multivariate Cox regression analyses (forestplot package). The restricted cubic spline (RCS) method was employed to analyze potential non-linear relationships between variables. This statistical approach flexibly captures complex relationships between a continuous independent variable (HMCN1 expression in this case) and a dependent variable (survival time in this case) by fitting cubic polynomials at multiple knots and connecting them smoothly. In this study, RCS was incorporated into the Cox proportional hazards model to test for non-linear associations between HMCN1 expression and the hazard ratio (HR). The significance of any non-linear effect was assessed via hypothesis testing to enhance model accuracy and interpretability (24).

Based on data extracted from the TIDE database, the association of HMCN1 expression levels with response to immunotherapy and prognosis was analyzed to evaluate its potential as a predictor of immunotherapy response. The association of HMCN1 with immunotherapy response was assessed using data from the TIDE web platform. This analysis integrated transcriptomic and clinical data from a pan-cancer cohort (comprising non-small cell lung cancer, melanoma, and bladder cancer) who received anti-PD-1 or anti-CTLA-4 therapy. Patient response was defined based on the objective response rate according to RECIST criteria, as pre-computed by the TIDE framework.

2.4 Genomic mutation analysis

Single nucleotide variation (SNV) data for 33 cancer types were collected from TCGA. The analyzed mutation types were Missense_Mutation, Silent, 5’ Flank, 3’ UTR, RNA, In_Frame_Del, Nonsense_Mutation, Splice_Site, Intron, 5’ UTR, In_Frame_Ins, Frame_Shift_Del, Nonstop_Mutation, 3’ Flank, Frame_Shift_Ins, and Translation_Start_Site. The SNV mutation frequency (expressed as a percentage) for each gene coding region was calculated using the following formula: (number of mutated samples/total number of samples for that cancer type) (25). Somatic copy number alterations for HMCN1 were analyzed using GISTIC 2.0 (26) on data from the TCGA pan-cancer cohort, with thresholds set as previously described (|G-score| > 0.1) (27). Data on structural variations (SVs) were obtained from the pre-computed TCGA which aggregates calls from established algorithms. Mutation spectrum oncoplots and lollipop plots were generated utilizing the cBioPortal database and the maftools R package, respectively, with protein domain information sourced from the PFAM database (https://www.ebi.ac.uk/interpro/). To further investigate the correlation between different gene sets and HMCN1 expression across the pan-cancer datasets, Spearman correlation analysis was employed. The resulting correlation coefficients and their corresponding statistical significance were visualized using radar plots (28, 29).

2.5 Analysis of the role of HMCN1 in the TME

To systematically investigate the role of HMCN1 in the TME and its clinical implications, we integrated multi-dimensional analytical approaches. Samples were stratified into high- and low-HMCN1 groups based on the median expression level. The association between these groups and the six established immune subtypes (C1–C6) was then assessed using a Chi-square test. Subsequently, to more precisely delineate the relationship between HMCN1 expression gradients and immune features, patients were further stratified into four quartiles (Q1–Q4, where Q1 represents the highest 25% and Q4, the lowest 25%) based on HMCN1 expression levels. The average values of various immune scoring metrics were calculated for each quartile (with missing values ignored), and the expression patterns of immune features were visualized using the pheatmap package to comprehensively display the continuous trends between HMCN1 expression levels and the immune status of the TME (25). By leveraging the TIMER2.0 database, Spearman correlations between HMCN1 and different immune-infiltrating cells were computed using multiple algorithms, and the results were intuitively presented in a heatmap.

2.6 Identification of potential HMCN1-targeting anti-tumor drugs

To explore potential therapeutic strategies targeting HMCN1-mediated tumor promotion, we performed a Connectivity Map (CMAP) analysis. By comparing patients with high versus low HMCN1 expression, we identified the 150 most significantly upregulated and 150 most significantly downregulated genes to construct an HMCN1-related gene signature. This signature was then matched against the perturbation profiles of 1,288 compounds in the CMAP database. Similarity scores were calculated using the XSum (eXtreme Sum) algorithm, and compounds with lower scores were considered potential candidates for reversing the HMCN1-mediated pro-tumor phenotype (30).

2.7 Enrichment analysis and pathway activity assessment

This study utilized the CancerSEA database, a resource that systematically integrates single-cell transcriptomic data reflecting 14 distinct functional states of cancer cells. The comprehensive score for each functional state was calculated using the z-score method from the Gene Set Variation Analysis (GSVA) R package. After standardization, this score was defined as the respective gene set activity index. Subsequently, the Pearson correlation between individual gene expression and these gene set activity scores was analyzed (31). Protein expression data were obtained from The Cancer Proteome Atlas. Patients were stratified into high and low HMCN1 protein expression groups based on the median expression level. Differences in pathway activity scores between these two groups were compared using the Wilcoxon test. Additionally, Gene Set Enrichment Analysis (GSEA) was performed using the clusterProfiler R package to further investigate the underlying biological mechanisms (32).

2.8 Cell culture and real-time quantitative PCR analysis of the mRNA expression of HMCN1

Human osteosarcoma (Saos-2, MG63, HOS, U-2OS, 143B) and osteoblastic (HOB) cell lines were obtained from the Cell Bank, Chinese Academy of Sciences (all certified) and cultured in Dulbecco modified Eagle medium supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin at 37°C in a 5% CO₂ atmosphere.

The mRNA expression level of HMCN1 in cells was determined by real-time quantitative PCR (RT-qPCR). Total RNA was extracted from cells using the TRIzol reagent (Vazyme Biotech, Nanjing, China, Cat# R401-01). cDNA was synthesized from 1 µg of total RNA using the HiScript III RT SuperMix for qPCR (+gDNA wiper) kit (Vazyme Biotech, Cat# R323-01) according to the manufacturer’s instructions. RT-qPCR amplification was performed on a QuantStudio 5 Real-Time PCR System (Applied Biosystems, USA) using the ChamQ Universal SYBR qPCR Master Mix (Vazyme Biotech, Cat# Q711-02). The reaction conditions were as follows: initial denaturation at 95°C for 30 s, followed by 40 cycles of 95°C for 10 s and 60°C for 30 s. GAPDH served as the internal reference gene. The relative gene expression levels were calculated and analyzed using the 2^–ΔΔCt method. All primers utilized in this study were synthesized by Wuhan Servicebio Technology Co. Ltd. The primer sequences were as follows: HMCN1 (NM_031935.3) forward: ACTGATGCTCGGTCCAAAGATTA, reverse: CTTGACCAGAACTTGTAGAGGCA (amplicon: 163 bp); GAPDH (NM_002046.7) forward: GGAAGCTTGTCATCAATGGAAATC, reverse: TGATGACCCTTTTGGCTCCC (amplicon: 168 bp).

2.9 Western blot assay of HMCN1 and EMT-related proteins

Total proteins were extracted from a panel of osteosarcoma cell lines (Saos-2, MG63, HOS, U-2OS, 143B) and a human osteoblast cell line (HOB) as a control. Cells were washed with cold PBS and lysed on ice using RIPA Lysis Buffer (Beyotime Biotechnology, Shanghai, China, Cat# P0013B) supplemented with 1 mM phenylmethylsulfonyl fluoride (PMSF, Beyotime, Cat# ST506) and a protease inhibitor cocktail (MedChemExpress, China, Cat# HY-K0010). The lysates were centrifuged at 12,000 × g for 15 min at 4°C, and the supernatants were collected. To manipulate HMCN1 expression, cells were transfected with three distinct HMCN1-specific siRNAs (si-HMCN1-1, si-HMCN1-2, si-HMCN1-3) designed and synthesized by GWuhan Servicebio Technology Co. Ltd (China) or an HMCN1 overexpression construct (OE-HMCN1), using non-targeting siRNA and an empty vector as negative controls. The siRNA sequences were as follows: si-HMCN1-1: 5’-GCAUCAAGAUUACCGUUUATT-3’; si-HMCN1-2: 5’-CCUGUACAAGAGGACCAUUTT-3’; si-HMCN1-3: 5’-GGAGAAGUUCUACAACAAATT-3’. The protein concentration of all lysates was quantified using a bicinchoninic acid (BCA) assay kit (Beyotime, Cat# P0012). Subsequently, equal amounts of protein (10-30 µg per lane) were separated by 8-12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electrophoretically transferred onto polyvinylidene difluoride membranes (Millipore, USA) under a constant current of 300 mA. The membranes were blocked with 5% non-fat milk for 2 h at room temperature and then incubated overnight at 4°C with specific primary antibodies diluted in TBST containing 5% bovine serum albumin. The primary antibodies and their dilutions were as follows: anti-HMCN1 (1:1000, Thermo Fisher Scientific, Cat# PA5-149932), anti-E-cadherin (1:2000, CST, #3195), anti-N-cadherin (1:1500, Proteintech, 22018-1-AP), anti-Vimentin (1:1000, CST, #5741), anti-Snail (1:1000, CST, #3879), and anti-GAPDH (1:5000, Proteintech, 60004-1-Ig), as previously described (33). After thorough washing, the membranes were incubated with appropriate horseradish peroxide-conjugated secondary antibodies (1:5000, Abbkine, China). Protein bands were visualized using an enhanced chemiluminescence (ECL) detection kit (New Cell & Molecular Biotech, Suzhou, China, Cat# P10300) and the band intensity was quantified with the ImageJ software (National Institutes of Health, USA). GAPDH was used as an internal loading control for normalization.

2.10 Wound healing and transwell invasion and migration assays

For the wound healing assay, 5 × 10⁵ cells were seeded per well in 6-well plates. Once the cells reached 100% confluence, uniform scratches were created across pre-marked lines using a sterile 200 µL pipette tip. The plates were washed with phosphate-buffered saline to remove detached cells and then replaced with serum-free medium to minimize cell proliferation. Images of the same fields were captured at 0 h and 24 h using an inverted microscope (Olympus, Japan). The wound closure area was quantified using ImageJ software. For the Transwell invasion assay, Transwell chambers (Corning, USA, Cat# 3422) with 8-µm pores were pre-coated with Matrigel (Corning, Cat# 356234) diluted 1:8 in serum-free medium, and solidified at 37°C for 4 hours. Serum-free medium containing 1 × 10⁵ cells was added to the upper chamber, while the lower chamber was filled with medium supplemented with 20% fetal bovine serum as a chemoattractant. Following 24 h of incubation, non-invading cells on the upper surface were carefully removed with a cotton swab. The cells that had invaded through the Matrigel were fixed with 4% paraformaldehyde for 30 minutes, stained with 0.1% crystal violet for 20 minutes, and subsequently quantified by counting cells in five random fields under the microscope. The migration assay was performed similarly but without Matrigel coating. All assays were performed in triplicate and repeated three times independently.

2.11 Statistical analysis

All statistical analyses were conducted using the R 4.4.1 software and GraphPad Prism 8.4.0. Group comparisons were performed as follows: Student’s t-test for parametric data, Wilcoxon test for non-parametric data, Chi-square or Fisher’s exact test for categorical variables, and one-way ANOVA or Kruskal-Wallis test for comparison of multiple groups. Correlations were assessed using Pearson’s (normal distribution) or Spearman’s (non-normal distribution) methods. Survival was analyzed using Cox proportional hazards regression and log-rank tests. In the meta-analysis, heterogeneity was quantified with Cochran’s Q and I². I² >50% was considered to indicate significant heterogeneity (p < 0.05), in which case a random-effects model was applied. In case of I² <50%, a fixed-effects model was used. Statistical significance was defined as p < 0.05.

3 Results

3.1 Multi-dimensional characterization of HMCN1 localization and its pan-cancer expression patterns at the gene, protein, cellular, and tissue levels

Bibliometric analysis indicated that HMCN1 oncological research remains nascent, even though it has recently gained traction. Moreover, the published literature largely focuses on prognosis, mutation, and therapy (Supplementary Figures 1A, B). Accurate identification of the subcellular localization of HMCN1 is crucial for a deeper understanding of its biological significance, as gene function is dependent on its subcellular localization (34). Localization analysis based on the UniProt database indicated that HMCN1 is primarily distributed in the Golgi apparatus and cytoplasm, which is consistent with its biosynthesis and secretion pathway as a large extracellular matrix glycoprotein. Its functional deposition into the extracellular matrix and basement membrane is well-established, as demonstrated in models of skin and tendon junctions (12). Concurrently, we obtained three-dimensional structural information of this protein from Protein Data Bank (Figures 1A, B), and this was used as a structural basis for further functional investigations, which is critical to note that this model is not experimentally verified and covers only a portion of the full-length protein. Structural information should be regarded as a putative visual guide rather than a definitive structural basis.

Figure 1

Figure 1. Expression and localization of HMCN1 across cancers. (A) Subcellular localization of HMCN1 as annotated in the UniProt database. (B) Three-dimensional structure of HMCN1 from the PDB database. (C) HMCN1 mRNA expression levels in tumor tissues from The Cancer Genome Atlas (TCGA) compared to normal tissues from the Genotype-Tissue Expression (GTEx) project. Statistical significance was determined using the unpaired two-tailed Student’s t-test. (D) Paired comparison of HMCN1 mRNA levels between tumor tissues and their matched adjacent normal tissues from the TCGA pan-cancer dataset. Statistical significance was determined using the paired two-tailed Student’s t-test. (E) Differential HMCN1 protein expression levels between tumor and normal tissues across various cancer types in the Clinical Proteomic Tumor Analysis Consortium (CPTAC) dataset. Statistical significance was determined using the unpaired two-tailed Student’s t-test. (F) The landscape of HMCN1 expression (nTPM: normalized transcripts per million) in cancer cell lines from the Human Protein Atlas (HPA) database. (G) Gene essentiality of HMCN1 across cell lines from various organs, as quantified by Chronos loss-of-function scores from the DepMap portal. Data are presented as mean ± SD. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001.

Next, the mRNA expression level of HMCN1 was compared between various tumor tissues and corresponding normal tissues using data from the TCGA and GTEx databases. The results showed that HMCN1 mRNA expression was significantly elevated in cancers such as glioblastoma multiforme (GBM), skin cutaneous melanoma (SKCM), and kidney renal clear cell carcinoma (KIRC) (Figure 1C). Furthermore, paired analysis using data from the TCGA pan-cancer dataset consistently demonstrated that HMCN1 mRNA expression was markedly upregulated in most tumor tissues compared to adjacent normal tissues (Figure 1D). To determine whether this genetic dysregulation also manifests at the protein level, we integrated and analyzed pan-cancer proteomics data from CPTAC (https://pdc.cancer.gov/pdc/browse). The results revealed that although HMCN1 protein expression showed an upward trend in most cancer types, this dysregulation was statistically significant only in a few cases (Figure 1E). To further clarify its expression characteristics at the protein level, we performed supplementary validation using the Human Protein Atlas dataset. The results suggested that HMCN1 expression was highest in bone cancer cell lines (Figure 1F). This explains the lack of significant results in the CPTAC dataset. Analysis using Chronos loss-of-function scores indicated that HMCN1 plays important functional roles in various malignancies, including osteosarcoma (Figure 1G). These findings indicate its potential involvement in carcinogenesis as a tumor-associated gene.

Multi-omics technologies, particularly the application of single-cell sequencing, have significantly enhanced the precision of detecting gene expression patterns (35). Therefore, we applied this technology to determine the expression of HMCN1 in specific cell types by analyzing pan-cancer single-cell RNA sequencing data from the TISCH database. The results showed that HMCN1 is primarily expressed in fibroblasts, endothelial cells, and a subset of tumor cells (Supplementary Figure 1C). To further validate this expression pattern and confirm it spatially, we integrated and analyzed pan-cancer spatial transcriptomics data. The results confirmed significant HMCN1 expression in the aforementioned cell types (Supplementary Figure 1D).

Building on the above results, we investigated the spatial distribution heterogeneity of HMCN1 within the TME. Using spatial transcriptomics data from GBM, SKCM, and KIRC, we categorized tissue regions into three types based on the proportion of malignant cells within each microregion (spot): regions with a malignant cell proportion of 1 were defined as malignant regions; those with a proportion of 0, as normal regions; and the rest, as mixed malignant regions (23). The SpatialPlot function of the Seurat package was used to visualize the major cell types and HMCN1 expression distribution across these microregions (36). The results demonstrated that HMCN1 expression levels were significantly higher in malignant regions than in normal and mixed regions. Further, Spearman correlation analysis of the correlation between cellular composition and HMCN1 expression levels across all microregions, performed using the linkET package, revealed a significant positive correlation between HMCN1 expression and malignant cell content in tumor cell-enriched areas (Supplementary Figure 2). In summary, HMCN1 exhibits an overexpression pattern in various cancer tissues, and its expression is spatially concentrated in malignant areas in a way that is highly consistent with the distribution of tumor cells. This suggests that HMCN1 holds potential as a clinical diagnostic biomarker, a possibility that was explored in the following experiments.

3.2 HMCN1 as a valuable prognostic and diagnostic biomarker across cancer types

Using data from the TCGA database, we first assessed the association between HMCN1 expression and tumor status using univariate logistic regression analysis. The results showed that in multiple cancer types, including GBM, SKCM, and KIRC, the odds ratios were all greater than 1. This indicates that elevated HMCN1 expression is significantly associated with the tumor state, reflecting its differential expression between malignant and normal tissues. (Table 1). Subsequently, OS, DSS, PFI, and DFI were analyzed using Cox proportional hazards regression models and the log-rank test. OS analysis indicated that high HMCN1 expression was significantly associated with shortened OS in patients with BLCA, KIRP, MESO, OV, and stomach adenocarcinoma (STAD). Furthermore, HMCN1 was also a significant risk factor for worsened DFI in these cancers. With regard to PFI, high HMCN1 expression was an independent risk factor for COAD and KIRP (Figure 2A). To further validate the prognostic value of HMCN1, we performed validation in multiple independent external datasets. The results consistently demonstrated that HMCN1 is a significant unfavorable prognostic factor in various cancers. However, in KIRC, despite upregulated HMCN1 expression across multiple datasets, it exhibited a protective prognostic effect (Figure 2B). This paradoxical association might reflect tissue-specific biological functions of HMCN1, potentially related to its role in maintaining structural integrity in renal tissues, though the precise mechanisms warrant further investigation.

Table 1

Table 1. Univariate logistic regression.

Figure 2

Figure 2. Diagnostic and prognostic value of HMCN1 in various cancers. (A) Heatmap of hazard ratios (HRs) from univariate Cox regression analysis, illustrating the association between HMCN1 mRNA expression and overall survival (OS), disease-specific survival (DSS), progression-free interval (PFI), and disease-free interval (DFI) across cancer types in the TCGA cohort. Red indicates HR > 1 (risk factor), blue indicates HR < 1 (protective factor). (B) Correlation of HMCN1 mRNA levels with patient overall survival in independent Gene Expression Omnibus (GEO) datasets. (C) Restricted cubic spline (RCS) analysis evaluating the continuous relationship between HMCN1 expression levels and the hazard ratio for overall survival (OS) in bladder urothelial carcinoma (BLCA), stomach adenocarcinoma (STAD), kidney renal papillary cell carcinoma (KIRP), and pancreatic adenocarcinoma (PAAD). The solid red line indicates the hazard ratio (HR), and the dashed black lines represent the 95% confidence interval (CI). P-overall and P-nonlinear values from the Cox proportional hazards model are shown. (D) Diagnostic performance of HMCN1 expression in distinguishing tumor from normal tissues, as shown by area under the curve (AUC) values of receiver operating characteristic (ROC) curves for various cancer types.

The potential non-linear relationship between HMCN1 expression levels and OS was thoroughly investigated with the flexible RCS method (24). The results showed that in cancers such as BLCA, STAD, KIRP, and PAAD, the effect of HMCN1 expression on survival risk followed a linear trend. The red solid line in the figure indicates that the HR increases linearly with rising HMCN1 expression, with a statistically significant overall association observed (P-overall < 0.05) (Figure 2C). The result of the non-linearity test was not significant (P-non-linear > 0.05); thus, there was no apparent non-linear effect of this gene in the aforementioned cancers. This finding suggests a stable and consistent exposure-response relationship between HMCN1 expression levels and patient prognosis in the case of these cancers. Its impact pattern demonstrates good interpretability and predictability and supports its use as a reliable linear prognostic indicator for subsequent research and clinical risk modeling (Figure 2C).

Previous studies have indicated that in STAD, the phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA) and HMCN1 co-mutation significantly reduced patient mortality risk. When this combination further co-occurred with low-density lipoprotein receptor-related protein 1B (LRP1B) or AHNAK nucleoprotein 2 (AHNAK2) mutations, distinct survival curves were observed. The survival results implied that HMCN1 mutation might exert a protective effect by influencing signaling pathways or microenvironment stability in the context of synergistic mutations. However, the role of HMCN1 alone remained unclear in the previous study (37). Therefore, this study used a comprehensive supplementary analysis to investigate further the role of HMCN1 in STAD. Analysis based on the GSE13861, GSE15459, and GSE26901 datasets revealed that high HMCN1 expression was significantly associated with higher mortality (Supplementary Figure 3A). A meta-analysis further confirmed that high HMCN1 expression was significantly correlated with poorer overall survival in patients with STAD (pooled HR = 1.23, 95% confidence interval [CI] = 1.16–1.30, p < 0.01; Supplementary Figure 3B). Multivariate Cox regression analysis showed that after adjusting for multiple clinical variables, high HMCN1 expression remained significantly associated with adverse prognosis (HR > 1, p < 0.05); thus, it could serve as an independent prognostic factor in STAD (Supplementary Figure 3C). Finally, ROC curve analysis confirmed that HMCN1 has high accuracy in distinguishing tumor from normal tissue and is applicable to most cancer types (Figure 2D). Overall, these findings demonstrate that HMCN1 holds diagnostic and prognostic value across different cancer types.

3.3 Genomic alteration analysis of HMCN1

We first mapped the mutation spectrum of HMCN1 across pan-cancer datasets by color coding the circles corresponding to each mutation type in the plot (Figure 3A). Further analysis based on the cBioPortal database revealed differential mutation frequencies of HMCN1 across various cancer types. Specifically, HMCN1 gene amplification was primarily observed in cholangiocarcinoma, while deep deletions and structural variants were more prevalent in sarcoma (Figure 3B). Analysis of SNV data from the TCGA database indicated that HMCN1 exhibits a moderate mutation frequency in several cancers, with the frequency being particularly high in SKCM (Figure 3C). A mutation landscape generated by the maftools package revealed that missense mutations constitute the predominant mutation type for HMCN1 (Figure 3D). Furthermore, Spearman correlation analysis revealed a significant positive association between HMCN1 expression levels and various genomic instability scores across multiple cancers. It is important to note that this observed association could be confounded by underlying factors such as tumor purity and proliferation rate. Nonetheless, this pan-cancer pattern suggests a potential link between HMCN1 expression and a genomic context characterized by higher chromosomal instability. These associations were intuitively presented via a radar chart, which highlighted the potential value of HMCN1 as a therapeutic target (Figure 3E).

Figure 3

Figure 3. Genomic alteration profiles of HMCN1. (A) Panoramic view of HMCN1 mutation sites across different cancers. (B) Stacked bar plot representing the frequency of HMCN1 mutations in pan-cancer analyses. (C) Heatmap displaying the co-occurrence of HMCN1 mutations with key oncogenic signaling pathways. (D) Mutation distribution profile (oncoplot) of HMCN1 across cancer types. (E) Radar plot illustrating the Spearman correlation coefficients between HMCN1 expression and multiple genomic instability scores (e.g., aneuploidy score, fraction altered, homologous recombination deficiency) across the TCGA pan-cancer dataset.

3.4 Role of HMCN1 in the tumor immune microenvironment and therapeutic implications

Analysis of the relationship between HMCN1 expression levels and the distribution of immune subtypes revealed that in tumors with high HMCN1 expression, the proportion of the C3 subtype (an inflammatory subtype, characterized by high expression of Th17- and Th1-related genes and typically associated with weaker tumor proliferation inhibition) was significantly increased, whereas the proportion of the C5 subtype (an immunologically quiet subtype) was lower (Figure 4A) (25). This distribution pattern, combined with the poor prognosis of HMCN1-high patients, led us to hypothesize that this immune-infiltrated microenvironment is functionally impaired. To test this, we assessed the relationship between HMCN1 and markers of T-cell exhaustion. We found that HMCN1 expression was significantly positively correlated with the levels of key immune checkpoint and exhaustion genes, including PDCD1 (PD-1) and CTLA4 (Supplementary Figure 4). This indicates that the HMCN1-high TME is not one of effective anti-tumor immunity, but rather a state of T-cell dysfunction and exhaustion, which facilitates immune evasion and explains the associated poor outcomes.

Figure 4

Figure 4. HMCN1 in immune regulation, therapy resistance, and drug targeting. (A) Stacked bar chart showing the distribution of six immune subtypes (C1-C6) across tumors stratified by high and low HMCN1 expression. Association was assessed by Chi-square test. (B) Positive associations between HMCN1 levels and infiltration of multiple immune cell types in various cancers. (C) Integrated analysis linking HMCN1 expression with genomic traits and immune infiltration patterns. (D, E) High HMCN1 expression is associated with adverse outcomes in the absence of immunotherapy and reduced treatment response, indicating a possible role in therapy resistance. (F) Identification of arachidonyltrifluoromethane as a potential therapeutic agent targeting HMCN1.

Previous studies have emphasized the crucial role of the TME in tumor progression and treatment response (38). Based on this, we employed multiple algorithms to further evaluate the correlation between HMCN1 and the abundance of various cell types in the tumor microenvironment. The results showed that HMCN1 had significant positive correlations with the infiltration levels of multiple immune cell types as well as stromal cells, such as endothelial cells, across the pan-cancer datasets (Figure 4B). By integrating this information with the aforementioned immune subtype characteristics, we hypothesize that this inflammatory microenvironment might represent a “dysfunctional” or “ineffective immune activation” state. That is, the substantial immune cell infiltration does not lead to effective killing of tumor cells but, instead, indirectly promotes tumor progression through sustained pro-inflammatory signaling. To systematically validate this hypothesis, we analyzed the association between HMCN1, genomic status, and immune infiltration patterns based on immunogenicity scores, DNA damage scores, and HMCN1 expression quartiles, and the results further supported the above observations (Figure 4C).

Our analysis revealed that patients who did not respond to immunotherapy exhibited high HMCN1 expression and were associated with a worse prognosis (Figures 4D, E). This observation suggests that high HMCN1 expression might be linked to immunotherapy resistance. Conversely, patients with low HMCN1 expression showed the highest rate of effective response to immunotherapy. This finding implies that targeted inhibition of HMCN1 could represent a novel sensitization strategy. Thus, HMCN1 could be regarded as a potential combination target for immunotherapy and offers a new direction for pan-cancer immunotherapy strategies. Following this, we conducted predictive screening for HMCN1-specific drugs with tumor sensitivity. Pan-cancer analysis indicated that arachidonyltrifluoromethane (also known as Arachidonyl trifluoromethyl ketone, AACOCF3; CAS# 149301-79-1) had the highest sensitivity across multiple cancers and might have potential as a targeted intervention drug (Figure 4F). This compound is a recognized inhibitor of cytosolic phospholipase A2 (cPLA2), with a known mechanism involving the blockade of arachidonic acid release. It is critical to clarify that this identification stems from a computational analysis of the Connectivity Map (CMAP) database, and the precise connectivity score and false discovery rate (FDR) for this specific prediction are unavailable. Most importantly, this prediction has not been experimentally validated and thus remains a hypothesis-generating finding for future research.

3.5 Multi-dimensional validation of HMCN1 association with pan-cancer progression and the EMT pathway

Analysis based on protein expression data from The Cancer Proteome Atlas and pathway activity scores revealed that in the vast majority of cancer types, the high HMCN1 expression groups showed significant enrichment in EMT-related pathways, while cell cycle pathway activity was relatively lower (Figure 5A). This suggests that high HMCN1 expression is associated with malignant tumor progression and is linked to the EMT process rather than directly by driving cell proliferation during cancer advancement. Further enrichment analysis indicated that multiple oncogenic signaling pathways were significantly enriched in the high HMCN1 expression group, whereas cell cycle-related pathways were more active in the low expression group, consistent with the above findings (Figure 5B). To delve deeper into its functional states, we analyzed the activity of 14 cancer functional states using the CancerSEA database. Detection via GSVA and the z-score-based algorithm confirmed that HMCN1 expression was positively correlated with pro-tumor processes such as EMT, angiogenesis, invasion, and metastasis, while simultaneously confirming its negative correlation with cell cycle activity (Figure 5C).

Figure 5

Figure 5. HMCN1 as a candidate conserved regulator of epithelial–mesenchymal transition (EMT). (A) Pathway activity analysis reveals elevated EMT scores in tumors with high HMCN1 expression. (B) Enrichment of multiple oncogenic signaling pathways in HMCN1-high expressing groups. (C) Correlation analysis from the CancerSEA database showing HMCN1 association with functional states such as angiogenesis, EMT, invasion, and stemness. (D) Pan-cancer analysis confirms a significant positive correlation (assessed by Spearman correlation) between HMCN1 expression and EMT activity across 33 cancer types in the TCGA dataset.

Previous studies have shown that the ECM can drive EMT through various mechanisms, thereby promoting tumor progression and metastasis, and indicate that ECM remodeling is a key driver of the EMT process in the TME. To investigate the role of HMCN1 in this process, we further analyzed the correlation between its expression levels and EMT scores across 33 cancers. The results demonstrated that, except in the case of uveal melanoma, HMCN1 showed a positive correlation with EMT scores in all the other cancer types examined. This suggests that HMCN1 expression is a highly conserved indicator of EMT activity across cancer types. (Figure 5D).

3.6 Role of HMCN1 as an tumor-promoting gene in osteosarcoma by driving malignant progression through EMT

No studies have yet investigated HMCN1 expression in osteosarcoma or its impact on patient prognosis. Therefore, we focused our analysis on its mechanism of action in this tumor. We integrated mRNA expression profiles from 10 data platforms comprising a total of 540 cases, including 414 osteosarcoma and 126 non-osteosarcoma cases, and the results showed that HMCN1 mRNA expression was significantly higher in the osteosarcoma group than in the non-osteo-sarcoma group (standardized mean difference = 0.80, 95% CI: 0.47–1.14; p = 0.024) (Figure 6A). Further, prognostic analysis indicated that high HMCN1 expression was significantly associated with poor prognosis in patients with osteosarcoma (Figure 6B).

Figure 6

Figure 6. HMCN1 in osteosarcoma: prognostic and functional insights. (A) Meta-analysis demonstrates significant upregulation of HMCN1 mRNA in osteosarcoma tissues compared to non-osteosarcoma controls across multiple gene expression datasets. (B) Elevated HMCN1 expression correlates with reduced patient survival. (C, D) Single-cell RNA sequencing reveals highest HMCN1 expression in malignant cells, with enrichment in HMCN1-positive populations. (E) HMCN1 expression is not influenced by cell cycle phase. (F) Positive correlation between HMCN1 levels and EMT functional scores in osteosarcoma, assessed by Spearman correlation analysis in two independent GEO datasets.

Next, we investigated the cellular heterogeneity of osteosarcoma and distribution of HMCN1 expression across cell types through single-cell transcriptome sequencing analysis. The results revealed that HMCN1 expression was the highest in malignant osteosarcoma cells (Figure 6C). We further identified all cell subpopulations and classified them into HMCN1-positive and HMCN1-negative groups based on HMCN1 expression levels. We then calculated the proportion of each cell subpopulation within these two groups to identify the primary cell types contributing to HMCN1 expression. Malignant cells constituted a significantly higher proportion of HMCN1-positive cell populations (48.1%) than HMCN1-negative cell populations (18.5%); this implies that malignant cells are probably the primary source of HMCN1 expression (Figure 6D). These results provide important cellular-level evidence for further investigating the role of HMCN1 in osteosarcoma development and progression.

We note that the RPPA platform employed by TCPA utilizes a predefined antibody panel, whose coverage for specific EMT-related phospho-proteins may be limited. Therefore, to more directly and definitively test the relationship between HMCN1 and EMT, we turned to functional validation. Our preliminary findings suggest that HMCN1 might drive tumor progression primarily by promoting the EMT process rather than directly regulating cell proliferation. To validate this mechanism in the context of osteosarcoma, we utilized the GSE146773 dataset. We sorted 1,152 U2OS FUCCI cells via fluorescence-activated cell sorting and performed single-cell RNA sequencing using the SMART-seq2 technology to simultaneously obtain gene expression profiles and cell cycle phase information for individual cells (39). The results showed that there was no significant variation in HMCN1 expression levels across different cell cycle phases (Figure 6E), indicating that its expression is not regulated by the cell cycle. Furthermore, in both the GSE39055 and GSE21257 datasets, HMCN1 expression showed significant positive correlations with EMT functional state scores (Figure 6F). This strongly supporting a potential role for HMCN1 in the EMT process during osteosarcoma progression (as observed from the pan-cancer data analysis).

3.7 In vitro evidence of HMCN1-mediated malignant behavior via regulation of EMT in osteosarcoma

To functionally validate the role of HMCN1 in driving malignancy via regulation of EMT in osteosarcoma, we conducted a series of in vitro experiments. We first confirmed that HMCN1 was significantly overexpressed at both the mRNA and protein levels in osteosarcoma cell lines using RT-qPCR and western blot analysis: MG63 and Saos-2 cells showed the most prominent upregulation of HMCN1 in comparison to cells (Figures 7A, B). Based on this finding, we constructed three distinct HMCN1-targeting siRNAs and one overexpression vector. si-HMCN1–2 demonstrated the most potent knockdown efficiency at the HMCN1 protein level and was, therefore, selected for all subsequent functional experiments (Figure 7C). Wound healing and Transwell assay results indicated that knocking down HMCN1 significantly inhibited the migration and invasion capabilities of osteosarcoma cells, whereas its overexpression produced the opposite, promotive effect (Figures 7D–F). To further explore the underlying mechanism, we examined the impact of HMCN1 on key EMT markers and found that HMCN1 overexpression significantly upregulated the expression of N-cadherin, vimentin, and Snail, while downregulating E-cadherin. Conversely, HMCN1 knockdown had the opposite effect. These functional gains and losses, coupled with the consistent shifts in EMT marker expression, demonstrate that HMCN1 promotes osteosarcoma cell invasion and migration by regulating the EMT process (Figure 8).

Figure 7

Figure 7. HMCN1 promotes migration and invasion of osteosarcoma cells in vitro. (A, B) HMCN1 expression is upregulated in osteosarcoma cell lines at the (A) mRNA and (B) protein levels. (A) Data are presented as mean ± SD (n=3). Statistical significance was determined using one-way ANOVA. (C) Western blot confirming efficient HMCN1 knockdown in Saos-2 and MG63. (D) Functional assays reveal that HMCN1 knockdown significantly inhibited the migration and invasion capabilities of MG63 and Saos-2 cells. Data are presented as mean ± SD (n=3). (E, F) Western blot confirming efficient HMCN1 overexpression in Saos-2 and MG63 and Functional assays reveal that HMCN1 overexpression significantly enhances the migration and invasion capabilities of MG63 and Saos-2 cells. Functional assays reveal that HMCN1 overexpression significantly enhances the migration and invasion capabilities of MG63 and Saos-2 cells. Data are presented as mean ± SD (n=3). Statistical significance was determined using the unpaired two-tailed Student’s t-test.

Figure 8

Figure 8. HMCN1 modulates the expression of key EMT markers. Protein levels of E-cadherin, N-cadherin, Vimentin, and Snail in osteosarcoma cells after HMCN1 knockdown or overexpression, as determined by western blot. The blots are representative of three independent experiments. * indicates a P-value less than 0.05, meaning the result is significant at the 0.05 level. **: indicates a P-value less than 0.01, meaning the result is significant at the 0.01 level. ***: indicates a P-value less than 0.001, meaning the result is highly significant at the 0.001 level.

4 Discussion

Our pan-cancer analysis establishes HMCN1 as a critical oncogenic ECM component that is frequently upregulated across cancers, with its high expression serving as a potential prognostic biomarker that is associated with poorer survival in a linear manner. Genomic analyses revealed that HMCN1 mutations are predominantly missense and positively correlate with genomic instability. Furthermore, its high expression was associated with immune cell infiltration and a tumor microenvironment compatible with a dysfunctional phenotype. The apparent discrepancy between HMCN1’s ECM function and its Golgi/cytoplasmic localization is reconciled by recognizing that the latter reflects its biosynthetic processing as a secreted protein, while its mature, functional form operates in the ECM. Thus, it may promote tumor development by modulating an immune context compatible with an immunosuppressive niche, a finding that aligns with and substantiates the broader role of the ECM. Mechanistically, the ECM component HMCN1 expression is strongly associated with the pro-metastatic EMT pathway across cancers. This finding is grounded in the established role of the ECM in driving EMT across cancers, such as pancreatic (40) and breast malignancies (41, 42). Our pan-cancer analysis demonstrated a significant association between HMCN1 expression, EMT pathway activity, and higher EMT scores. We specifically extended this finding to osteosarcoma, where the ECM’s role in promoting an EMT-like, invasive phenotype is recognized (16–18). We experimentally confirmed that HMCN1 promotes osteosarcoma malignancy by regulating EMT. Collectively, our work elevates HMCN1 from a mere ECM constituent to a key regulatory node and potential therapeutic target within the central ECM–EMT axis of cancer progression.

Another interesting finding that emerged from this study is that high HMCN1 expression may not directly play a role in regulating cell cycle activity. This phenomenon may be related to the overall dysfunction of the ECM in the TME, with potential mechanisms including dysregulated ECM component expression, excessive protease degradation, abnormal cell surface receptors, and altered ECM physical properties (8, 43). In this context, as one of many components of the ECM, the potential independent regulatory function of HMCN1 may be masked or compensated for by broader ECM dysregulation, thereby explaining the lack of significant association with cell cycle control. Transwell migration, invasion, and wound healing assays demonstrated that HMCN1 overexpression significantly enhanced malignant behaviors in osteosarcoma. This suggests its tumor-promoting effect is probably achieved indirectly by inducing EMT, rather than through directly accelerating cell cycle progression. Numerous studies have confirmed that the EMT process is closely associated with maintaining cancer stem cell properties, and stem-like cells typically possess enhanced self-renewal capacity and proliferative potential (44). Therefore, HMCN1 may activate the EMT program, enabling cells to acquire an invasive phenotype that subsequently manifests as a proliferative advantage overall-a mechanism that was experimentally validated in our study. Finally, drug sensitivity analysis indicated that arachidonyltrifluoromethane shows promise as a potential therapeutic agent targeting HMCN1. This compound has demonstrated anti-tumor potential in various cancer models, possibly through specifically modulating the immune microenvironment or inhibiting relevant signaling pathways to impede tumor progression. (45–47). However, it is crucial to note that our computational prediction identifies this compound based on its functional antagonism of the HMCN1 gene signature; the precise molecular mechanism remains to be elucidated. A key question for future research is whether arachidonyltrifluoromethane exerts its effects by directly binding to the HMCN1 protein itself or indirectly through upstream regulators or co-components within the ECM-EMT axis. These findings suggest its potential value for treating tumors with high HMCN1 expression, although this requires additional experimental validation.

5 Limitations

While our study has the enormous advantage of a pan-cancer analysis of data from multiple databases and provides valuable insights into the broad role of HMCN1, there are several limitations that need to be acknowledged. First, the analytical approaches used may be inherently limited by batch effects or systematic discrepancies across platforms. Furthermore, our proteomic analysis is constrained by the coverage of the TCPA antibody panel. Second, although our study confirms the significant role of HMCN1 in osteosarcoma progression, its precise molecular mechanisms require further elucidation. For instance, the dysregulation mechanisms of HMCN1 under the influence of upstream regulators and its interaction patterns with other core molecules within the EMT pathway remain to be clarified. Finally, our functional findings lack in vivo validation, and future translational research should build upon our findings by exploring two key avenues. First, the anti-tumor efficacy of the identified candidate, arachidonyltrifluoromethane, requires functional validation in vitro to directly assess its ability to inhibit HMCN1-mediated oncogenic phenotypes. Second, expanding on this, the therapeutic potential of combining such HMCN1-targeting strategies with existing conventional treatments for osteosarcoma should be thoroughly investigated.

6 Conclusion

Through a systematic pan-cancer analysis, this study establishes for the first time that elevated HMCN1 expression is a conserved biomarker positively associated with EMT across cancer types. Further, its role as an independent prognostic biomarker and oncogenic driver was validated in multiple cancers. In osteosarcoma in particular, HMCN1 was highly expressed and found to promote malignant phenotypes, and its function as an EMT-associated tumor-promoting gene was validated. Moreover, high HMCN1 expression was correlated with an immunosuppressive microenvironment and resistance to immunotherapy; thus, it may hold potential as a combination target for immunotherapy. Integrating data on the expression levels of this protein into clinical decision-making could provide a basis for precise treatment strategies and potentially improve patient outcomes.

Data availability statement

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

Author contributions

YXL: Writing – original draft. YYL: Writing – review & editing, Funding acquisition. LZ: Writing – review & editing.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the Science and Technology Plan of Jiangxi Provincial Health Commission (No. 202130005), and The Science and Technology Plan of Jiangxi Provincial Administration of Traditional Chinese Medicine (2023B0027).

Acknowledgments

The authors thank all the public data sources involved in the present study.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. The author(s) used generative AI to enhance language clarity, particularly for complex sentences. All content was subsequently reviewed and edited, and the author(s) take full responsibility for the final manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fonc.2025.1730887/full#supplementary-material

Supplementary Figure 1 | (A, B) Bibliometric analysis of HMCN1 in oncology, showing the rising publication trend and the current research focus on prognosis, mutation, and therapy. (C, D) Single-cell RNA sequencing and spatial transcriptomics reveal the expression pattern of HMCN1.

Supplementary Figure 2 | HMCN1 expression is spatially correlated with malignant cells in the tumor microenvironment. Spatial transcriptomics analysis of GBM, SKCM, and KIRC shows that HMCN1 expression is significantly higher in malignant regions than in normal or mixed regions. Spearman correlation analysis reveals a significant positive correlation between HMCN1 expression and malignant cell content across all microregions.

Supplementary Figure 3 | High expression of HMCN1 is an independent prognostic factor for poor survival in STAD. (A) Analysis of the GSE13861, GSE15459, and GSE26901 datasets reveals that high HMCN1 expression is significantly associated with higher mortality. (B) A meta-analysis confirms that high HMCN1 expression is significantly correlated with poorer overall survival in STAD patients (pooled HR = 1.23, 95% CI = 1.16–1.30, p < 0.01). (C) Multivariate Cox regression analysis demonstrates that high HMCN1 expression remains an independent prognostic factor for adverse outcomes after adjusting for clinical variables.

Supplementary Figure 4 | The heatmap shows the correlation analysis between HMCN1 and PDCD1 and CTLA4 at the pan-cancer level.

References

1. Liu Z and Zhang S. Toward a systematic understanding of cancers: a survey of the pan-cancer study. Front Genet. (2014) 5:194. doi: 10.3389/fgene.2014.00194

PubMed Abstract | Crossref Full Text | Google Scholar

2. Sathyanarayanan A, Tanha HM, Mehta D, and Nyholt DR. Integrative multi-omic analysis identifies genetically influenced DNA methylation biomarkers for breast and prostate cancers. Commun Biol. (2022) 5:594. doi: 10.1038/s42003-022-03540-4

PubMed Abstract | Crossref Full Text | Google Scholar

3. Kar SP, Considine DPC, Tyrer JP, Plummer JT, Chen S, Dezem FS, et al. Pleiotropy-guided transcriptome imputation from normal and tumor tissues identifies candidate susceptibility genes for breast and ovarian cancer. HGG Adv. (2021) 2. doi: 10.1016/j.xhgg.2021.100042

PubMed Abstract | Crossref Full Text | Google Scholar

4. Scarbrough PM, Weber RP, Iversen ES, Brhane Y, Amos CI, Kraft P, et al. A cross-cancer genetic association analysis of the DNA repair and DNA damage signaling pathways for lung, ovary, prostate, breast, and colorectal cancer. Cancer Epidemiol Biomarkers Prev. (2016) 25:193–200. doi: 10.1158/1055-9965.EPI-15-0649

PubMed Abstract | Crossref Full Text | Google Scholar

5. Yu K, Zhang Q, Liu Z, Du Y, Gao X, Zhao Q, et al. Deep learning based prediction of reversible HAT/HDAC-specific lysine acetylation. Brief Bioinform. (2020) 21:1798–805. doi: 10.1093/bib/bbz107

PubMed Abstract | Crossref Full Text | Google Scholar

6. Xu X, Jaberi-Douraki M, and Wallace NA. Predicting the prognostic value of POLI expression in different cancers via a machine learning approach. Int J Mol Sci. (2022) 23. doi: 10.3390/ijms23158571

PubMed Abstract | Crossref Full Text | Google Scholar

7. Iqbal S, Karim MR, Mohammad S, Mathiyalagan R, Morshed MN, Yang DC, et al. Multiomics analysis of the PHLDA gene family in different cancers and their clinical prognostic value. Curr Issues Mol Biol. (2024) 46:5488–510. doi: 10.3390/cimb46060328

PubMed Abstract | Crossref Full Text | Google Scholar

8. Najafi M, Farhood B, and Mortezaee K. Extracellular matrix (ECM) stiffness and degradation as cancer drivers. J Cell Biochem. (2019) 120:2782–90. doi: 10.1002/jcb.27681

PubMed Abstract | Crossref Full Text | Google Scholar

9. Eble JA and Niland S. The extracellular matrix in tumor progression and metastasis. Clin Exp Metastasis. (2019) 36:171–98. doi: 10.1007/s10585-019-09966-1

PubMed Abstract | Crossref Full Text | Google Scholar

10. Cho Y and Doh J. The extracellular matrix in solid tumor immunotherapy. Trends Immunol. (2024) 45:705–14. doi: 10.1016/j.it.2024.07.009

PubMed Abstract | Crossref Full Text | Google Scholar

11. Du W, Xia X, Hu F, and Yu J. Extracellular matrix remodeling in the tumor immunity. Front Immunol. (2023) 14:1340634. doi: 10.3389/fimmu.2023.1340634

PubMed Abstract | Crossref Full Text | Google Scholar

12. Welcker D, Stein C, Feitosa NM, Armistead J, Zhang JL, Lutke S, et al. Hemicentin-1 is an essential extracellular matrix component of the dermal-epidermal and myotendinous junctions. Sci Rep. (2021) 11:17926. doi: 10.1038/s41598-021-96824-4

PubMed Abstract | Crossref Full Text | Google Scholar

13. Bergson S, Sarig O, Giladi M, Mohamad J, Mogezel-Salem M, Smorodinsky-Atias K, et al. HMCN1 variants aggravate epidermolysis bullosa simplex phenotype. J Exp Med. (2025) 222. doi: 10.1084/jem.20240827

PubMed Abstract | Crossref Full Text | Google Scholar

14. Liu CL, Pan HW, Torng PL, Fan MH, and Mao TL. SRPX and HMCN1 regulate cancer−associated fibroblasts to promote the invasiveness of ovarian carcinoma. Oncol Rep. (2019) 42:2706–15. doi: 10.3892/or.2019.7379

PubMed Abstract | Crossref Full Text | Google Scholar

15. Joris T, Jouant T, Jacques JR, Gouverneur L, Saintmard X, Vilanova Mana L, et al. Reduction of antisense transcription affects bovine leukemia virus replication and oncogenesis. PloS Pathog. (2024) 20:e1012659. doi: 10.1371/journal.ppat.1012659

PubMed Abstract | Crossref Full Text | Google Scholar

16. Cortini M, Baldini N, and Avnet S. New advances in the study of bone tumors: A lesson from the 3D environment. Front Physiol. (2019) 10:814. doi: 10.3389/fphys.2019.00814

PubMed Abstract | Crossref Full Text | Google Scholar

17. Yang G, Yuan J, and Li K. EMT transcription factors: implication in osteosarcoma. Med Oncol. (2013) 30:697. doi: 10.1007/s12032-013-0697-2

PubMed Abstract | Crossref Full Text | Google Scholar

18. Tzanakakis GN, Giatagana EM, Berdiaki A, Spyridaki I, Hida K, Neagu M, et al. The role of IGF/IGF-IR-signaling and extracellular matrix effectors in bone sarcoma pathogenesis. Cancers (Basel). (2021) 13. doi: 10.3390/cancers13102478

PubMed Abstract | Crossref Full Text | Google Scholar

19. Jian E, Wang M, Zhang Z, Heng Y, Zhang C, Chen Q, et al. The metabolic endotoxemia and gut microbiota: research trajectories and hot trends across the centuries, (1999-2024). Front Microbiol. (2025) 16:1634803. doi: 10.3389/fmicb.2025.1634803

PubMed Abstract | Crossref Full Text | Google Scholar

20. Kim Y and Lee D. Unsupervised cell line embedding using pairwise drug response correlation. Comput Struct Biotechnol J. (2025) 27:2566–73. doi: 10.1016/j.csbj.2025.06.018

PubMed Abstract | Crossref Full Text | Google Scholar

21. Man YN, Xu H, Chen PJ, Sun Y, and He ML. Comprehensive analysis of the clinical significance and molecular mechanism of T-box transcription factor 3 in osteosarcoma. J Cancer. (2024) 15:4007–19. doi: 10.7150/jca.96168

PubMed Abstract | Crossref Full Text | Google Scholar

22. Han Y, Wang Y, Dong X, Sun D, Liu Z, Yue J, et al. TISCH2: expanded datasets and new tools for single-cell transcriptome analyses of the tumor microenvironment. Nucleic Acids Res. (2023) 51:D1425–31. doi: 10.1093/nar/gkac959

PubMed Abstract | Crossref Full Text | Google Scholar

23. Shi J, Wei X, Xun Z, Ding X, Liu Y, Liu L, et al. The web-based portal spatialTME integrates histological images with single-cell and spatial transcriptomics to explore the tumor microenvironment. Cancer Res. (2024) 84:1210–20. doi: 10.1158/0008-5472.CAN-23-2650

PubMed Abstract | Crossref Full Text | Google Scholar

24. Austin PC, Fang J, and Lee DS. Using fractional polynomials and restricted cubic splines to model non-proportional hazards or time-varying covariate effects in the Cox regression model. Stat Med. (2022) 41:612–24. doi: 10.1002/sim.9259

PubMed Abstract | Crossref Full Text | Google Scholar

25. Thorsson V, Gibbs DL, Brown SD, Wolf D, Bortone DS, Ou Yang TH, et al. The immune landscape of cancer. Immunity. (2018) 48:812–830.e14. doi: 10.1016/j.immuni.2018.03.023

PubMed Abstract | Crossref Full Text | Google Scholar

26. Mermel CH, Schumacher SE, Hill B, Meyerson ML, Beroukhim R, and Getz G. GISTIC2.0 facilitates sensitive and confident localization of the targets of focal somatic copy-number alteration in human cancers. Genome Biol. (2011) 12:R41. doi: 10.1186/gb-2011-12-4-r41

PubMed Abstract | Crossref Full Text | Google Scholar

27. Cancer Genome Atlas Research, Network, Weinstein JN, Collisson EA, Mills GB, Shaw KR, Ozenberger BA, et al. The Cancer Genome Atlas Pan-Cancer analysis project. Nat Genet. (2013) 45:1113–205. doi: 10.1038/ng.2764

PubMed Abstract | Crossref Full Text | Google Scholar

28. Mayakonda A, Lin DC, Assenov Y, Plass C, and Koeffler HP. Maftools: efficient and comprehensive analysis of somatic variants in cancer. Genome Res. (2018) 28:1747–56. doi: 10.1101/gr.239244.118

PubMed Abstract | Crossref Full Text | Google Scholar

29. Mistry J, Chuguransky S, Williams L, Qureshi M, Salazar GA, Sonnhammer ELL, et al. Pfam: The protein families database in 2021. Nucleic Acids Res. (2021) 49:D412–9. doi: 10.1093/nar/gkaa913

PubMed Abstract | Crossref Full Text | Google Scholar

30. Yang C, Zhang H, Chen M, Wang S, Qian R, Zhang L, et al. A survey of optimal strategy for signature-based drug repositioning and an application to liver cancer. Elife. (2022) 11. doi: 10.7554/eLife.71880

PubMed Abstract | Crossref Full Text | Google Scholar

31. Hanzelmann S, Castelo R, and Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinf. (2013) 14:7. doi: 10.1186/1471-2105-14-7

PubMed Abstract | Crossref Full Text | Google Scholar

32. Yu G, Wang LG, Han Y, and He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. (2012) 16:284–7. doi: 10.1089/omi.2011.0118

PubMed Abstract | Crossref Full Text | Google Scholar

33. Tian Y, Liu R, Hou X, Gao Z, Liu X, and Zhang W. SIRT2 promotes the viability, invasion and metastasis of osteosarcoma cells by inhibiting the degradation of Snail. Cell Death Dis. (2022) 13:935. doi: 10.1038/s41419-022-05388-2

PubMed Abstract | Crossref Full Text | Google Scholar

34. Bridoux L, Gofflot F, and Rezsohazy R. HOX protein activity regulation by cellular localization. J Dev Biol. (2021) 9. doi: 10.3390/jdb9040056

PubMed Abstract | Crossref Full Text | Google Scholar

35. Wang J, Ye F, Chai H, Jiang Y, Wang T, Ran X, et al. Advances and applications in single-cell and spatial genomics. Sci China Life Sci. (2025) 68:1226–82. doi: 10.1007/s11427-024-2770-x

PubMed Abstract | Crossref Full Text | Google Scholar

36. Xun Z, Ding X, Zhang Y, Zhang B, Lai S, Zou D, et al. Reconstruction of the tumor spatial microenvironment along the Malignant-boundary-nonmalignant axis. Nat Commun. (2023) 14:933. doi: 10.1038/s41467-023-36560-7

PubMed Abstract | Crossref Full Text | Google Scholar

37. Yu Y, Xie Z, Zhao M, and Lian X. Identification of PIK3CA multigene mutation patterns associated with superior prognosis in stomach cancer. BMC Cancer. (2021) 21:368. doi: 10.1186/s12885-021-08115-w

PubMed Abstract | Crossref Full Text | Google Scholar

38. Bejarano L, Jordao MJC, and Joyce JA. Therapeutic targeting of the tumor microenvironment. Cancer Discov. (2021) 11:933–59. doi: 10.1158/2159-8290.CD-20-1808

PubMed Abstract | Crossref Full Text | Google Scholar

39. Mahdessian D, Cesnik AJ, Gnann C, Danielsson F, Stenstrom L, Arif M, et al. Spatiotemporal dissection of the cell cycle with single-cell proteogenomics. Nature. (2021) 590:649–54. doi: 10.1038/s41586-021-03232-9

PubMed Abstract | Crossref Full Text | Google Scholar

40. Bulle A and Lim KH. Beyond just a tight fortress: contribution of stroma to epithelial-mesenchymal transition in pancreatic cancer. Signal Transduct Target Ther. (2020) 5:249. doi: 10.1038/s41392-020-00341-1

PubMed Abstract | Crossref Full Text | Google Scholar

41. Brandao-Costa R,M, Helal-Neto E, Vieira AM, Barcellos-de-Souza P, Morgado-Diaz J, and Barja-Fidalgo C. Extracellular Matrix Derived from High Metastatic Human Breast Cancer Triggers Epithelial-Mesenchymal Transition in Epithelial Breast Cancer Cells through alphavbeta3 Integrin. Int J Mol Sci. (2020) 21. doi: 10.3390/ijms21082995

PubMed Abstract | Crossref Full Text | Google Scholar

42. Han SJ, Kwon S, and Kim KS. Contribution of mechanical homeostasis to epithelial-mesenchymal transition. Cell Oncol (Dordr). (2022) 45:1119–36. doi: 10.1007/s13402-022-00720-6

PubMed Abstract | Crossref Full Text | Google Scholar

43. Chanpanitkitchote P, Nuanpirom J, Pongsapich W, Asavapanumas N, Mendler S, Wiesmann N, et al. EMILIN-1 suppresses cell proliferation through altered cell cycle regulation in head and neck squamous cell carcinoma. Am J Pathol. (2025) 195:995–1012. doi: 10.1016/j.ajpath.2025.01.010

PubMed Abstract | Crossref Full Text | Google Scholar

44. Lambert AW and Weinberg RA. Linking EMT programmes to normal and neoplastic epithelial stem cells. Nat Rev Cancer. (2021) 21:325–38. doi: 10.1038/s41568-021-00332-6

PubMed Abstract | Crossref Full Text | Google Scholar

45. Zhu Z, Liu H, Fu H, Luo Y, Chen B, Wu X, et al. CMTM7 shapes the chronic inflammatory and immunosuppressive tumor microenvironment in hepatocellular carcinoma as an M2 macrophage biomarker. Sci Rep. (2024) 14:29659. doi: 10.1038/s41598-024-75538-3

PubMed Abstract | Crossref Full Text | Google Scholar

46. Dong C, Lin Z, Hu Y, and Lu Q. KLRD1 (CD94): A prognostic biomarker and therapeutic candidate in head and neck squamous cell carcinoma. J Cancer. (2025) 16:982–95. doi: 10.7150/jca.104762

PubMed Abstract | Crossref Full Text | Google Scholar

47. Zhao J, Li L, Wang Y, Huo J, Wang J, Xue H, et al. Identification of gene signatures associated with lactation for predicting prognosis and treatment response in breast cancer patients through machine learning. Sci Rep. (2025) 15:13575. doi: 10.1038/s41598-025-98255-x

PubMed Abstract | Crossref Full Text | Google Scholar