The diagnostic and prognostic value of MTA3 in pan-cancer analysisTo assess whether MTA3 serves as a potential diagnostic indicator across multiple cancer types, we systematically curated and analyzed pan-cancer datasets (including both paired and unpaired samples) to examine the expression level variations of the MTA3 gene (Fig. 1A). Our analysis revealed significantly elevated expression of MTA3 in 23 tumor types, including GMB, GBMLGG, LGG, UCEC, BRCA, LUAD, ESCA, STES, KIPAN, COAD, STAD, HNSC, KIRC, LUSC, LIHC, WT, BLCA, PAAD, TGCT, UCS, ALL, LAML, and CHOL, compared to normal tissues. Conversely, MTA3 exhibited significantly lower expression in 6 tumor types, namely SKCM, THCA, OV, PCPG, ACC, and KICH.For the prognostic assessment of MTA3, we conducted a comprehensive analysis of its expression levels in relation to various survival prognosis indicators among cancer patients, utilizing the Cox proportional hazards model, including overall survival (OS), disease-specific survival (DSS), disease-free interval (DSI), and progression-free survival (PSI). Statistical significance was determined using the Logrank test, with the criterion of no fewer than 3 significant survival indices among the 4 indicators to denote the prognostic significance of MTA3 (Fig. 1B). Our findings revealed that high expression levels of MTA3 were associated with poorer prognosis in 5 tumor types, namely ACC, KIPAN, LIHC, PAAD, and STES, whereas low expression levels were associated with poorer prognosis in 3 tumor types, including GBMLGG, LGG, and THYM.To further explore the potential of MTA3 in indicating genomic variations and epigenetic alterations, we conducted analyses using the cBioPortal online platform (TCGA, Pan-cancer Atlas). Our findings revealed that approximately 1.2% of the samples exhibited MTA3 mutations, with amplifications and missense mutations being the most prevalent mutation types (Fig. 1C). Notably, the mutation frequency of MTA3 was highest in UCEC, reached approximately 5% (Fig. 1D). Moreover, the somatic mutation frequency of MTA3 was 0.6%, with the S351F/C alteration being the most significant amino acid sequence variation observed (Fig. 1E). Considering the role of RNA modifications in cancer biology, we investigated the association between MTA3 expression and genes involved in RNA methylation modifications. Our analysis demonstrated a robust positive correlation between MTA3 expression and most genes implicated in RNA methylation modifications across various cancer types. However, certain genes did not exhibit significant correlations with cancer types, such as UCS, THYM, DLBC, and CHOL, suggesting potential heterogeneity in the regulatory mechanisms (Fig. 1F). Furthermore, utilizing the UALCAN database, we examined the methylation levels of MTA3 across pan-cancer and their corresponding tissues. Our analysis unveiled reduced MTA3 promoter methylation levels in BLCA, BRCA, LIHC, LUAD, and LUSC compared to adjacent normal tissues. Conversely, COAD, KIRP, PRAD, PCPG, and THCA demonstrated higher levels of MTA3 promoter methylation relative to their adjacent normal tissues (Fig. 1G). In summary, these analyses underscore the significance of MTA3 in indicating genomic and epigenetic variations across multiple cancer types, shedding light on its potential implications in cancer pathogenesis and progression.Figure 1Pan-cancer charactersitics of MTA3 expression profile in pan-cancer. (A) The expression profile of MTA3 in multiple carcinoma tissues compared with normal tissue was analyzed using the Wilcoxon test based on TCGA and GTEx datasets, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; (B) Using the Cox proportional hazards model and confirmed the prognostic significance using the Logrank test, MTA3 expression level associated with OS, DSS, DFI, and PFI in pan-cancer, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; (C, D) Mutation types and frequencies of MTA3 in pan-cancer revealing approximately 1.2% of the samples exhibited MTA3 mutations with amplifications and missense mutations being the most common mutation types based on cBioPortal platform (TCGA, Pan-cancer Atlas); (E) Landscape of MTA3 mutations revealed the somatic mutation frequency of 0.6%, with S351F/C being the most significant amino acid sequence alteration; (F) Relationship between MTA3 expression levels and genes associated with RNA modifications in pan-cancer tissues, suggesting a strong positive correlation with most genes involved in RNA methylation modifications across pan-cancer, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; (G) Promoter methylation levels of MTA3 in Pan-Cancer analysis, including BLCA, BRCA, COAD, KIRP, LIHC, LUAD, LUSC, PCPG, PRAD and THCA.Immunological characteristics of MTA3 in pan-cancerTo comprehensively investigate the potential role of MTA3 in the tumor microenvironment, we examined the correlation between MTA3 expression and immune scores, including stromal score, immune score, and the ESTIMATE score. Considering at least 2 scores with statistical significance as qualifying indicators of MTA3, our analysis revealed a negative correlation between MTA3 expression and the immune score across 25 cancers, including BRCA, CESC, CHOL, DLBC, ESCA, GBM, GBMLGG, HNSC, KIPAN, KIRC, KIRP, LGG, LIHC, LUAD, LUSC, MESO, OV, PAAD, SARC, SKCM, STAD, STES, TGCT, THCA, and UCEC. Conversely, MTA3 exhibited a positive correlation with the immune score in BLCA, COAD, COADREAD, and PRAD (Fig. 2A), suggesting its potential role in the regulation of the immune microenvironment. Given the critical link between tumor stemness, malignant tumor resistance, metastasis, and microenvironmental regulation, we assessed the correlation between MTA3 expression and various tumor stemness scores. Our analysis indicated significant correlations between MTA3 expression and DNAss in 23 different tumor types, with positive correlations observed in 14 tumor types, including TGCT, SKCM, HNSC, DLBC, UVM, STES, STAD, LUAD, READ, ESCA, CESC, PAAD, LUSC, and BLCA, and negative correlations in 9 cancers including THYM, KICH, LIHC, GBMLGG, KIPAN, MESO, THCA, LGG, and BRCA (Fig. 2B). Additionally, correlations with RNAss revealed positive associations in 15 tumor types, including TGCT, THYM, LGG, GBMLGG, LUAD, ESCA, LUSC, PRAD, STES, LIHC, OV, PCPG, HNSC, BLCA, and STAD, while negative correlations were significant in LAML, KIRP, and KIPAN (Fig. 2C). Furthermore, to assess the potential relevance of MTA3 in immunotherapy, we analyzed its correlation with microsatellite instability (MSI) scores across different cancer types. Our findings demonstrated a positive correlation between MTA3 expression and MSI scores in patients with GBM, GBMLGG, HNSC, LIHC, LUAD, LUSC, STAD, and STES, while negative correlations were observed in BRCA, KIPAN, and ACC (Fig. 2D).Through our investigation, we observed significant overexpression of MTA3 in LIHC and its close association with diagnostic indicators, prognostic markers, genomic variations, and immunological characteristics. Consequently, we proceeded to explore and screen potential marine-derived drugs targeting MTA3 in LIHC (Fig. 2E).Figure 2Immunological characteristics of MTA3 in pan-cancer analysis. (A) Illustrating the correlation between MTA3 expression and Stromal scores, immune scores, and ESTIMATE scores, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; (B, C) Correlation of MTA3 with Stemness Scores of DNAss and RNAss; (D) Correlation of MTA3 expression with MSI, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; (E) Venn diagram was depicted according to Figs. 1 and 2.MTA3 is an independent biomarker for hepatocellular carcinomaInitially, paired differential analysis of MTA3 expression in LIHC tumor tissues versus normal tissues suggested its diagnostic and prognostic relevance in HCC. Validation using the ICGC-LIHC-US cohort confirmed elevated MTA3 expression in both unmatched (Fig. 3A) and matched (Fig. 3B) tumor samples. Kaplan-Meier (KM) survival curves further indicated worse survival outcomes for LIHC patients with elevated MTA3 expression in the ICGC cohort (Fig. 3C). To further explore into the pivotal role of MTA3 in LIHC prognosis, we analyzed its association with various clinical characteristics. MTA3 exhibited increased expression in advanced pathological stages (Fig. 3D) and histological grades (Fig. 3E). Moreover, MTA3 expression correlated with age, sex, AFP levels, and vascular invasion (Fig. 3F-G), suggesting its association with malignant pathological progression in LIHC. Subsequent univariate and multivariate COX regression analyses, in conjunction with clinical data, identified MTA3’s status as an independent prognostic factor for LIHC (Fig. 3J, K). Incorporating pathological staging and MTA3 expression levels into a nomogram underscored the clinical utility of MTA3 (Fig. 3L), with calibration curves indicating strong predictive accuracy (Fig. 3M). ROC analysis further validated the prognostic performance of MTA3 in LIHC, yielding AUCs of 0.706 and 0.610 for 1-year and 3-year survival, respectively (Fig. 3N). Thus, our specific analysis of MTA3 in LIHC diagnosis and prognosis indicates that elevated MTA3 expression could serve as a reliable indicator for the occurrence and progression of hepatocellular carcinoma.Figure 3Expression of MTA3 in LIHC patients and its clinical relevance. (A) Expression of MTA3 in the ICGC cohort; (B) Paired comparison of MTA3 expression between paired normal and tumor tissues; (C) Relationship between MTA3 expression in the ICGC cohort and overall survival rate; (D-I) Analysis of MTA3 expression changes based on different pathological stages (D), histological grades (E), age (F), gender (G), alpha-fetoprotein (AFP) levels (H), and vascular invasion (I), utilizing the Wilcoxon test for statistical evaluation. *p < 0.05, **p < 0.01, ****p < 0.0001; (J, K) Univariate COX analysis (J) and multivariate COX analysis (K) revealing the prognostic significance of MTA3 in LIHC; (L) Nomogram based on MTA3 expression and pathological stage; (M) Calibration plot evaluate the accuracy of the nomogram; (N) ROC curve for the prognostic evaluation of 1-year and 3-year OS in LIHC patients based on MTA3 expression.Association analysis of MTA3 with the tumor microenvironment in hepatocellular carcinoma tissuesTo explore the underlying mechanisms through which MTA3 influences the occurrence and progression of LIHC, we identified differentially expressed genes (DEGs) associated with MTA3. GO enrichment analysis of these DEGs revealed their involvement primarily in biological processes such as cell division (mitosis, cytokinesis) and cell-cell adhesion. Moreover, these DEGs were implicated in the formation of ion channel complexes (Fig. 4A).Subsequent KEGG analysis identified associations of these DEGs with cell cycle regulation, cAMP signaling pathway, Hippo signaling pathway, ECM-receptor interaction, and Cell adhesion molecules (Fig. 4B). Notably, the Hippo signaling pathway is particularly noteworthy, given its established roles in maintaining tumor stem cell stemness and regulating the immune microenvironment. Additionally, the involvement of cell cycle and cAMP signaling pathway indicates that MTA3 may influence tumor cell proliferation, migration, and invasion through multiple mechanisms. GSEA further validated the potential functions of MTA3. The high MTA3 expression group showed significant enrichment in biological processes related to cell adhesion (Fig. 4C), and immune response processes involving immunoglobulin complexes, reflecting MTA3 potential role in regulating the tumor immune microenvironment. High MTA3 expression was additionally associated with the activity of ion channels, particularly sodium ion transmembrane transport, suggesting its involvement in tumor cell signaling and electrophysiological characteristics.Furthermore, correlation analysis between MTA3 expression and the expression of immune checkpoint genes in LIHC revealed significant positive correlations with several immune checkpoint genes, including CTLA-4, HAVCR2, PD-1, and TIGIT. These genes were significantly upregulated in the high MTA3 expression group, suggesting a potential role for MTA3 in regulating the tumor microenvironment (Fig. 4D, E). Given the enrichment analysis results suggesting MTA3 involvement in the tumor immune microenvironment of LIHC, we further analyzed the correlation between MTA3 expression and TME scores. The high MTA3 group exhibited significantly lower TME scores (Fig. 4F), including Immune Score, ESTIMATE Score, and Stroma Score. This suggests that MTA3 may contribute to the formation of an immune-suppressive microenvironment, adversely affecting treatment response and prognosis by promoting immune escape mechanisms.Furthermore, the correlation between MTA3 expression and the infiltration of various immune cells was analyzed. The results showed that the infiltration of immune-active cells such as B cells and NK cells was showed negative correlation with MTA3 expression levels. By implanting Hepa1-6 cells transfected with siNC or siMTA3 into mice and performing HE staining, tumors with siMTA3 showed increased immune infiltration (Figure S1), which verified the aforementioned analysis. This further demonstrates that MTA3 may promote the formation of an immune escape and immune-suppressive environment in LIHC by regulating the infiltration of specific immune cells (Fig. 4G-J).Figure 4Analysis of the association between MTA3 and the tumor microenvironment in LIHC. (A) Gene Ontology (GO) enrichment analysis of differentially expressed genes (DEGs) associated with MTA3; (B) Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of DEGs associated with MTA3; (C) Gene Set Enrichment Analysis (GSEA) of MTA3 high expression levels; (D, E) Correlation analysis between MTA3 and immune checkpoint genes in LIHC; (F) Differential analysis of tumor microenvironment (TME) scores between high-MTA3 and low-MTA3 groups, utilizing the Wilcoxon test for statistical evaluation.; (G-J) Correlation analysis of MTA3 with immune cells in LIHC.Drug screening from the natural product database of marine origin by regulating MTA3Through comprehensive analysis from pan-cancer to hepatocellular carcinoma, we have evaluated the potential application value of MTA3 as a disease biomarker. To demonstrate the impact of MTA3 on hepatocellular carcinoma cells, we utilized the HepG2 and Hepa1-6 cell lines as the study models. By employing MTA3 siRNA to reduce MTA3 expression in hepatocellular carcinoma cells (Figure S2A), inhibiting MTA3 significantly suppressed the cell viability of hepatocellular carcinoma cells (Fig. 5A). As MTA3 is recognized as metastasis-associated protein, we then examined the effect of MTA3 knockdown on cell metastasis. The results showed that knockdown of MTA3 suppressed the migration of HepG2 and Hepa1-6 cells (Fig. 5B), indicating the key role of MTA3 in the regulation of hepatocarcinoma cell. This finding is consistent with our previous bioinformatics analysis, which indicates a negative correlation between MTA3 levels and the immune system’s impact on tumor progression.Figure 5Loss-of-function analysis of MTA3 regulates HepG2 and Hepa1-6 cell viability and migration. (A) Hepatocellular cell lines were transfected with siMTA3 or its negative control (siNC). CCK8 assay was performed to detect the viability of the hepatocellular carcinoma cell line HepG2 and Hepa1-6, n = 5 in each batch; (B) Transwell assay was used to detect cell migration, n = 5 in each batch, ***p < 0.001 compared with siNC group. Scale bar equals 100 μm.To validate our assessment and search for potential targeted drugs, we obtained active ingredients from marine-derived natural products (HY-L143) from the MCE website. Since the biological function of MTA3 primarily involves transcriptional regulation, we focused on the protein domain of MTA3 binding to gene transcription sites (PDB: 2CRG). Subsequently, we employed molecular docking to calculate the binding scores of 38 molecules from the natural product library to MTA3. Following Autodock scoring criteria, we selected small molecules with binding scores lower than − 4.0 kcal/mol. A total of 9 small molecules were screened and highlighted in blue font in Table 1. Additionally, we conducted a drug-likeness evaluation among these 38 molecules (Table 2), based on drug-likeness principles from Lipinski, Pfizer, GSK, and the Golden Triangle. Consequently, we selected 13 molecules that met drug-likeness criteria, also highlighted in blue font in Table 2. To ensure a good balance between drug-likeness and MTA3 binding properties, we cross-referenced Tables 1 and 2, resulting in 7 qualified molecules listed in Table 3. Subsequently, we evaluated these 7 candidates, considering absorption, distribution, metabolism, excretion, and toxicity properties, prioritizing the exclusion of carcinogenicity. Finally, we selected Jaspamycin as a potential marine-derived drug targeting MTA3 for hepatocellular carcinoma treatment, based on its favorable characteristics and potential efficacy.Table 1 Molecular docking in marine-derived natural components targeting MTA3.Table 2 Drug likeness evaluation among the components of marine-derived natural products.Table 3 Marine-derived drug screen based on MTA3 combination and drug likeness.Jaspamycin inhibits the growth of hepatocellular carcinoma cells by regulating MTA3To explore the potential of Jaspamycin in inhibiting hepatocellular carcinoma, we initially administered Jaspamycin according to its half maximal effective concentration in Trypanosoma brucei (EC50 = 6.5 nM) with reference to MedChemExpres (https://www.medchemexpress.cn/Jaspamycin.html). Thus, we set a series of concentration at the nM level, from 0 to 40 nM. We observed that Jaspamycin significantly inhibited the activity of hepatocellular carcinoma cells around 5 nM for 48 h (EC50 for 24 h = 8.263 nM, EC50 for 48 h = 4.988 nM, Figure S3), which led to suppressed proliferation and reduced numbers of hepatocellular carcinoma cells. Considering MTA3 acts as a transcription factor for itself, we investigated the impact of Jaspamycin on MTA3 expression. We found that Jaspamycin significantly inhibited the expression of MTA3 (Figure S2B). Growth curve analysis further revealed a significant inhibition of cell viability in HepG2 and Hepa1-6 following 48 h of treatment, reaching its peak inhibition at 72 h (Fig. 6A). Moreover, Jaspamycin exerted effects on the migration of hepatocellular carcinoma cells (Fig. 6B and C) and cell cycle regulation, as indicated by cell cycle analysis demonstrating direct effects of Jaspamycin on the S and G2 phases (Fig. 6D). In the in vivo experiments, saline or Jaspamycin was administered via oral gavage to Hepa1-6 tumor-bearing mice for 4 weeks. Tumor tissues were harvested at the end of the fourth week (Fig. 7A), and tumor volume changes were recorded throughout the progression (Fig. 7B). The results demonstrate a significant reduction in tumor progression in the group treated with Jaspamycin compared to the control group. Furthermore, Jaspamycin treatment suppressed both tumor size and weight (Fig. 7C), indicating the efficacy of Jaspamycin in inhibiting hepatocellular carcinoma. This also validates the feasibility of our approach to identifying targets through bioinformatics and discovering drugs using CADD.To validate whether Jaspamycin acts through regulating MTA3 to inhibit the activity of hepatocellular carcinoma cells, HepG2 and Hepa1-6 cells were transfected with an MTA3 overexpression plasmid along with Jaspamycin treatment, and the effectiveness of Mta3 overexpression was validated (Figure S2C). Overexpression of MTA3 effectively counteracted the inhibitory effects of Jaspamycin on cell viability (Fig. 8A), migration (Fig. 8B and C), and cell cycle regulation (Fig. 8D), when compared to the vector control group. This suggests that the inhibitory effects of Jaspamycin on hepatocellular carcinoma cells involve regulating MTA3, and the observed effects are dependent on MTA3 expression levels.Taken together, our study has not only identified MTA3 as a promising therapeutic target for hepatocellular carcinoma but also has unveiled Jaspamycin as a potential therapeutic agent by regulating MTA3, thereby inhibiting hepatocellular carcinoma cell activity. These findings offer valuable insights into therapeutic targets and drugs for hepatocellular carcinoma and potentially other cancers characterized by high MTA3 expression. This comprehensive approach highlights the potential of MTA3-based therapies in cancer treatment.Figure 6Jaspamycin inhibits HepG2 and Hepa1-6 proliferation and migration. (A) CCK8 assay was applied to detect cell viability of HepG2 and Hepa1-6 over 1–3 days after Jaspamycin treatment, n = 5 in each time spot, ***p < 0.001, ****p < 0.0001; (B, C) Cell migration was assessed using scratch assay (×100 magnification) and transwell assay (×200 magnification), respectively. n = 5, ****p < 0.0001. (D) Cell cycle was detected by flow cytometry after 48 h Jaspamycin treatment, n = 5, *p < 0.05, ****p < 0.0001.Figure 7Jaspamycin inhibits hepatocellular carcinoma progression in tumor bearing BALB/c mice. (A) Representative images of tumor tissue in Saline group and Jaspamycin group, n = 5; (B) Tumor volume was measured and record once a week during tumor progression, n = 5; (C) From left to right are the results of tumor weight, tumor volume, and tumor diameter (length) in the 4th week after tumor tissue harvest, n = 5 in each batches, **p < 0.001, ****p < 0.0001.Figure 8MTA3 mediates the inhibitory effects of Jaspamycin on HepG2 and Hepa1-6 proliferation and migration. (A) CCK8 assay was applied to evaluate cell proliferation and viability of HepG2 and Hepa1-6 for 72 h with Jaspamycin treatment and MTA3 overexpression plasmid or its negative control vector, n = 5, ***p < 0.001,****p < 0.0001; (B, C) Cell migration was detected by scratch assay (×100 magnification) and transwell assay (×200 magnification), with scale bars of 50 μm and 40 μm, n = 5, ****p < 0.0001; (D) Cell cycle was detected by flow cytometry after 48 h Jaspamycin treatment with MTA3 overexpression plasmid or its negative control vector, n = 5, ***p < 0.001, ****p < 0.0001.Figure 9Graphic abstract illustrating key concepts of the research. Study design includes pan-cancer analysis, hepatocellular carcinoma analysis, marine-derived anti-cancer agent discovery, and verification of anti-cancer effectiveness.
