Development of a prognostic model based on m6A reader HNRNPA2B1 upregulation and immune infiltration in multiple malignant tumors

Introduction

Tumors are diseases caused by driver gene mutations and immune escape, in which precancerous cells lose control of their normal regulatory mechanisms and promote uncontrolled proliferation (1). Disruption of the homeostatic balance of the body leads to uncontrolled division of tumor cells, infiltration into surrounding tissues, and subsequent invasion and spread to distant organs (2). Given the heterogeneity and complexity of carcinogenesis and tumor progression, its specific mechanism of pathogenesis remains incompletely understood. N6-methyladenosine (m6A) is a frequently observed type of RNA modification in eukaryotes that has important effects on almost all vital cellular processes as well as tumorigenesis and progression (3,4). m6A modification is a dynamic and reversible process that is mediated by three homologous factors, namely, m6A-binding proteins (readers), demethylases (erasers), and methyltransferases (writers) (5-7). The deposition of m6A modification is catalyzed by the methyltransferase complex (MTC), while its removal is dynamically regulated by demethylases. By modulating the expression of target genes, m6A serves as a critical epigenetic regulator of cellular behaviors and systemic physiological functions. Mechanistically, m6A governs nearly all facets of RNA metabolism, spanning mRNA translation efficiency, stability control, alternative splicing, nucleocytoplasmic trafficking, and tertiary structure remodeling, with spatiotemporal specificity across biological contexts (8). To perform precise biological functions, m6A-modified messenger RNAs (mRNAs) are recognized by specific RNA-binding proteins, known as, m6A readers. The primary function of the reader protein is to specifically recognize the base and bind to the methylated region of m6A, weakening the homologous binding to RNA-binding protein, leading to alterations in protein-RNA associations by changing the RNA secondary structure and finally activating downstream regulatory signaling pathways (9-11).

HNRNPA2B1, which belongs to the HNRNP protein family, is a nuclear reader of the m6A epigenetic marker, which has an irreplaceable role in m6A modifications (12,13). HNRNPA2B1 binds to the RGAC motif of transcripts (13). It is highly expressed in different tumor types and regulates tumor progression through multiple metabolic processes, such as cytoplasmic RNA transport, mRNA alternative splicing (AS), transcription, and translation (13,14). HNRNPA2B1 has been previously linked to multiple myeloma (MM) genesis, development, and dismal prognostic outcomes (15). However, the immunoregulatory function of HNRNPA2B1 in promoting tumorigenesis and development remains incompletely understood (16,17). Consequently, the present work focused on elucidating the effects of HNRNPA2B1 on the genesis and progression of various tumor types from different perspectives, including transcriptional regulation, genetic variation, DNA phosphorylation and HNRNPA2B1 methylation, pathological staging, diagnostic efficacy, disease prognosis, immune infiltration, related gene enrichment, cellular pathways, and drug targets. HNRNPA2B1 may serve as a novel target for cancer therapy. Moreover, a model for diagnosis and disease prognosis prediction based on the expression level of HNRNPA2B1 was developed. We present this article in accordance with the TRIPOD reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2024-2616/rc).

Methods

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Gene expression, protein analysis, Tumor Node Metastasis (TNM) stages, and survival prognosis of patients with different tumors

The data collection time span of The Cancer Genome Atlas (TCGA; https://portal.gdc.cancer.gov) cohort started in 2006 and continued until the latest update. The Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/) is a high-throughput gene expression database created by the National Center for Biotechnology Information in 2000 and has since been maintained. The data used in this study were sourced from the public databases of TCGA and the GEO. A large number of tumor samples and corresponding normal tissue samples from medical institutions and research centers around the world have been collected from TCGA (in this study, 9,664 cancer tissues and 711 normal tissues were included). HNRNPA2B1 expression in multiple tumors was analyzed via the TCGA and GEO databases. The “Gene DE” module of TIMER2.0 (18) (http://timer.cistrome.org/) was used to compare the differential expression of the candidate genes between cancer tissues and the corresponding healthy tissues. We identified diverse types of tumor-associated genes whose expression was up- or downregulated in tumors. The expression levels of the candidate genes in different cancers are shown in boxplots.

We provide an overview of the TCGA/GTEx data available in GEPIA2. Tissue samples with complete gene and clinical information were included in this study, and comparisons between tumor and normal tissues were performed. The expression of HNRNPA2B1 was analyzed in tumor and para-tumor tissues from multiple tumor types. This work utilized the GEPIA2 (19) (http://gepia2.cancer-pku.cn/#analysis) and GTEx datasets to improve the analysis of tumor tissues for which very few or no adjacent healthy tissues were available [P=0.01, log₂ fold change (FC) =1, “Match GTEx and TCGA normal data”].

In addition, a violin plot showing HNRNPA2B1 expression levels across different pathological stages (stages I–IV) was generated via the “Pathological Stage Map” module via the TCGA and GEPIA2 datasets. Additionally, the Human Protein Atlas (HPA) (20,21) (https://www.proteinatlas.org/) was used to analyze HNRNPA2B1 protein expression in cells and tissues. The HNRNPA2B1 protein and mRNA expression levels within various tumors were assessed via the histology and pathology modules. This work also selected 6 different tumor datasets from UALCAN (22) (http://ualcan.path.uab.edu/analysis-prot.html), including ovarian cancer (OC), breast cancer (BC), renal clear cell carcinoma (RCCC), colorectal cancer (CRC), LUAD and UCEC datasets, to analyze protein expression levels via the Clinical Proteomics Tumor Analysis Consortium (CPTAC). The total protein expression levels and phosphorylation status (modified at S212, S225, S236, S259, S324, S341, S341, Y347, and Y347) of HNRNPA2B1 were further investigated in primary cancer and healthy tissue samples.

The “Survival Map” module of GEPIA2 (19) (http://gepia2.cancer-pku.cn/#survival) was used to perform survival analysis for HNRNPA2B1-expressing tumors. Survival analysis included overall survival (OS) and disease-free survival (DFS) for the TCGA tumor datasets. Cutoff-high (50%) and cutoff-low (50%) values were defined as thresholds for determining whether the tumor was in the high- or low-expression group. Moreover, log-rank tests were adopted for hypothesis tests, whereas the GEPIA2 “Survival analysis” module was utilized to obtain survival analysis plots.

Diagnostic efficiency analysis platform

Receiver operating characteristic (ROC) curves were drawn via ROC plotter. Online ROC analysis (23) (http://www.rocplot.org) was performed to link HNRNPA2B1 expression with response to therapy in BC, OC, CRC, and glioblastoma multiforme (GBM).

Genetic alteration analysis of HNRNPA2B1 in different tumor types

To explore genetic variation in HNRNPA2B1, this work utilized the cBioPortal (24) (https://www.cbioportal.org/) tool for analyzing the HNRNPA2B1 mutation signature in diverse tumors collected from the TCGA database. The results for the frequency of alterations, mutation types, and copy number alterations (CNAs) were observed under the “Cancer Types Summary” module. The “Mutation” module was used to acquire the three-dimensional (3D) structures of the mutated HNRNPA2B1 proteins. Samples from the TCGA database were used for survival analysis of patients with or without HNRNPA2B1 mutations. A Kaplan-Meier plot was used for visualization of different types of survival data, such as OS, DFS, and progression-free survival (PFS) data. Kaplan-Meier diagrams were generated by using the “Comparison/Survival” module. The expression of HNRNPA2B1 with driver mutations in tumors in the TCGA cohort and its correlation with drug sensitivity were analyzed using CAMOIP (http://camoip.net/) and OncoVar (https://oncovar.org/). Drug sensitivity related to HNRNPA2B1 mutation was analyzed via half maximal inhibitory concentration (IC₅₀) values in tumor cells using CAMOIP and is displayed in boxplots. AI-Driver, which was formulated recently, was adopted to predict driver mutations. A novel strategy was adopted for integrating known driver genes with driver estimation.

Analysis of the degree of immune infiltration

This work utilized the TIMER2 “immune genes” module (18) to explore the associations of HNRNPA2B1 expression levels with degrees of immune infiltration, as well as the potential relationships between HNRNPA2B1 gene expression levels and cancer-associated fibroblast (CAF) infiltration levels in diverse TCGA samples with different algorithms. Heatmaps and scatterplots were adopted for data visualization.

HNRNPA2B1-associated gene enrichment

The STRING database (https://string-db.org/) was used to determine the interacting protein partners of HNRNPA2B1. The main parameters used were as follows: protein name (“HNRNPA2B1”), organism (“Homo sapiens”), network edge meaning (“evidence”), maximal interactor number to display (“first-order interactions” ≤50 interactors), minimally required interaction score [“Low confidence (0.150)”], and sources of active interaction (“experiments”).

The “Similar Gene Detection” module (19) of GEPIA2 (http://gepia2.cancer-pku.cn/#survival) was used to screen the top 100 HNRNPA2B1-correlated target genes by comparing cancer samples and healthy samples. Pearson’s correlation coefficient was used to determine the relationship between the expression of HNRNPA2B1 and other genes via the GEPIA2 “Correlation Analysis” module (25). This work adopted Log₂ transcripts per million (TPM+1) to create a dot plot to show the correlation coefficients and the corresponding P values. Furthermore, heatmap information for the selected genes was provided by the “Gene_Corr” module of TIMER2 (18), which contains P values and partial correlations (cor) as assessed by Spearman’s rank correlation test adjusted by purity. By using the Jvenn interactive Venn diagram viewer, a crossover analysis was performed to compare the interacting partners of HNRNPA2B1.

Two datasets were integrated to conduct the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis. The gene list was uploaded into Database for Annotation, Visualization and Integrated Discovery (DAVID) to obtain information on functional annotation under the settings for the selected identifier (“OFFICIAL_GENE_SYMBOL”) as well as species (“Homo sapiens”). The “tidyr” and “ggplot2” R packages were used to visualize enriched pathways. Additionally, the “clusterProfiler” R package was used for the analysis of Gene Ontology (GO) functional annotations. The “cnetplot” module was adopted to visualize information regarding biological processes (BPs), molecular functions (MFs), and cellular components (CCs) via the cnetplot function (circular =F, node_label =T, colorEdge =T). Analysis was performed via R language software (R-3.6.3, 64-bit; https://www.r-project.org/), with P<0.05 indicating statistical significance. By integrating the GSE52834 dataset (26) (experimental context: siRNA-mediated knockdown of HNRNPA2B1 in GM19238 lymphoblastoid cells) and the GSE70061 dataset (27) [which combines high-throughput sequencing of RNA isolated by crosslinking immunoprecipitation (CLIP-seq) with nuclear RNA-seq to map HNRNPA2B1 binding sites and evaluate transcriptomic changes under METTL3 or HNRNPA2B1 depletion], we performed differential expression analysis. This comprehensive approach was employed within the RM2Target platform (http://rm2target.canceromics.org/#/home) (26), aimed at discovering significant motifs related to RNA modifications in human genes.

Phosphorylation level analysis of HNRNPA2B1

The CPTAC dataset from the UALCAN (22) data portal (http://ualcan.path.uab.edu/) was selected to analyze the phosphorylation status of HNRNPA2B1. Z values represent the sample standard deviation based on the median of a specific cancer. The log₂-specific values from CPTAC were first normalized within each sample profile and then normalized across samples. This study compared the phosphorylation status of HNRNPA2B1 between primary cancer samples and healthy samples from different cancer types.

Methylation level analysis of HNRNPA2B1

This work also utilized the Human Disease Methylation Database Version 2.0 (http://bioinfo.hrbmu.edu.cn/diseasemeth/) to compare the methylation status of HNRNPA2B1 between cancer and para-tumor samples. Additionally, this work investigated the association of HNRNPA2B1 expression levels with DNA methylation status by using MEXPRESS (28) (http://mexpress.be). Finally, multivariate survival analysis was performed via MethSurv (29) (https://biit.cs.ut.ee/methsurv/) (https://biit.cs.ut.ee/methsurv/) to assess the scattering of different CpG islands. We strengthened our analysis using publicly available methylation data from M6AREG database (30).

Analysis of traditional Chinese medicine (TCM) compounds targeting HNRNPA2B1

To determine the effects of traditional Chinese medicinal compounds on HNRNPA2B1 expression, data from 1,037 high-throughput experiments involving the use of different therapeutics were collected and reanalyzed from HERB (31) (http://herb.ac.cn/), which is a high-throughput sequencing (HTS) and reference database for TCM.

Statistical analyses

GraphPad Prism 8 (San Diego, CA, USA) was used for data generation, whereas the statistical software SPSS 23.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. P<0.05 represented statistical significance according to the Wilcoxon test. “*” indicates statistical significance. The degree of statistical significance is presented as follows: *, P<0.05; **, P<0.01. Thirty-three tumors were analyzed in the TCGA and GEO databases via R software (version 3.6.3).

Results

Gene expression profiles

HNRNPA2B1 mRNA levels in various cancer types collected from the TCGA database were assessed via the TIMER2.0 module. HNRNPA2B1 was significantly overexpressed in BRCA, BLCA, COAD, CHOL, LIHC, ESCA, LUAD, HNSC, LUSC, STAD, PRAD, UCEC, CESC, and READ tissues with variable statistical power (Figure 1A). In contrast, HNRNPA2B1 expression was downregulated in KICH, KIRC, and THCA (P<0.001). HNRNPA2B1 expression in metastatic skin cutaneous melanoma (SKCM) was greater than that in SKCM (P<0.05). Even though, for some tumor samples, very few or no matched healthy tissues were available, this work demonstrated high expression of HNRNPA2B1 in CHOL, DLBC, GBM, LGG, PAAD, and THYM and low expression in OV (P<0.05) (Figure 1B).

Figure 1 Expression levels of the HNRNPA2B1 gene in different tumors and pathological stages. The expression status of the HNRNPA2B1 gene in different cancers or specific cancers was analyzed by TIMER2 (A). The corresponding normal tissues in the GTEx database were used as controls, corresponding to BRCA, CESC, CHOL, DLBC, GBM, LAML, LGG, MESO, OV, PAAD, SARC, SKCM, TGCT, THYM, UCS and UVM in the TCGA project. Compared with those in normal tissues, HNRNPA2B1 expression levels were higher in CHOL, DLBC, GBM, LGG, OV, PAAD, and THYM tissues (B). On the basis of the CPTAC dataset, we also analyzed the expression levels of HNRNPA2B1 total protein in breast cancer, ovarian cancer, colon cancer, clear cell RCC, UCEC, lung adenocarcinoma and primary tissues (C). The TCGA data included ACC, SKCM, TGCT, COAD and LIHC data. The main pathological stages (stages I, II, III and IV) were analyzed for the expression level of the HNRNPA2B1 gene (D). *, P<0.05; **, P<0.01; ***, P<0.001. CPTAC, Clinical Proteomics Tumor Analysis Consortium; N, normal; T, tumor; TCGA, The Cancer Genome Atlas; TPM, transcripts per million.

Additionally, we analyzed total HNRNPA2B1 protein expression in different cancer types. According to CPTAC dataset analysis, total HNRNPA2B1 protein expression in tumor tissues collected from BC, OC, CRC, LUAD, and UCEC patients was greater than that in adjacent healthy tissues, whereas HNRNPA2B1 protein levels were lower in RCC patients than in adjacent healthy tissues (P<0.05) (Figure 1C).

Finally, the GEPIA2 “stage plot” module was used to observe the associations of HNRNPA2B1 levels with pathological tumor stages (stages I, II, III, and IV). In the TCGA dataset, HNRNPA2B1 gene expression levels in ACC, SKCM, TGCT, COAD, and LIHC were correlated with pathological stage (P<0.05) (Figure 1D).

Survival analysis and diagnostic efficiency

We aimed to determine the diagnostic and prognostic relevance of HNRNPA2B1 expression. Hence, the potential association of HNRNPA2B1 expression with patient survival was analyzed via survival analysis and independent diagnostic power (32-34). Tumor cases from the TCGA and GEO datasets were separated into low- and high-expression groups on the basis of HNRNPA2B1 expression to study the potential association between HNRNPA2B1 levels and the survival of patients with different tumor types. OS analysis suggested that HNRNPA2B1 upregulation was significantly correlated with a dismal prognosis in ACC (P<0.001), KICH (P=0.008), LGG (P=2.8e−05), LIHC (P=0.02), LUAD (P=0.003), MESO (P=0.03), and SARC (P=0.02) patients. Moreover, HNRNPA2B1 downregulation was significantly correlated with poor prognosis in THYM patients (P=0.004) (Figure 2A).

Figure 2 Correlation between HNRNPA2B1 gene expression in the TCGA cohort and cancer survival prognosis. We used the GEPIA2 tool to reveal the relationships between tumor HNRNPA2B1 gene expression and overall survival (A) and disease-free survival (B) by analyzing HNRNPA2B1 gene expression in different tumors in the TCGA. TCGA, The Cancer Genome Atlas.

DFS analysis from the TCGA database revealed that HNRNPA2B1 overexpression predicted poor patient survival in ACC (P=8.4e−06), CESC (P=0.009), KICH (P=0.03), LGG (P=0.001), LIHC (P=0.04), LUAD (P=0.01), and PRAD (P=0.006) (Figure 2B). Furthermore, HNRNPA2B1 downregulation predicted poor DFS in THCA patients (P=0.003).

ROC curve (http://www.rocplot.org) analysis was performed for OC [area under the curve (AUC) =0.603; P=1.7e−05], BC (AUC =0.625; P=5.4e−07), GBM (AUC =0.59; P=3.1e−02), and CRC (AUC =0.552; P=0.07) (Figure S1).

Genetic alteration data

Tumorigenesis is often driven by mutated or abnormally expressed oncogenes, leading to the immortal proliferation of cancer cells. Mutated oncogenes act in a variety of ways to drive tumorigenesis (35,36). Oncogenes also coordinate with a wide variety of protein complexes to switch specific genes “turn on” or “turn off” by binding to specific regions of DNA (37,38). Therefore, we further analyzed genetic alterations in HNRNPA2B1 in diverse cancer types from the TCGA cohort. According to Figure 3A, the most frequent alterations (>5%) in HNRNPA2B1 occurred in patients with bladder urothelial carcinoma, predominantly harboring mutations. Chromosomal amplification was more common in esophageal adenocarcinoma patients, and the alteration frequency was about 3.85% (Figure 3A). Notably, all patients with genetic alterations in cholangiocarcinoma, uterine carcinosarcoma, KIRP, and KIRC harbored mutations in HNRNPA2B1 (Figure 3A). All cases from sarcoma, PRAD, CESC, TGCT, LGG, PCPG, and PAAD harbored chromosomal amplifications in HNRNPA2B1 (Figure 3A).

Figure 3 Exploration of the genetic variation of HNRNPA2B1 in different tumor types included in TCGA. We analyzed the HNRNPA2B1 mutational signature of TCGA tumors via the cBioPortal tool. Mutation frequency (A) and mutation type and site (B) are shown. We show the mutation site with the highest alteration frequency (X293_splice) in the 3D structure of HNRNPA2B1 (C). We also analyzed potential correlations between mutation status and progression-free, disease specific, overall KIRC (D) using the cBioPortal tool. 3D, three-dimensional; CNA, copy number alteration; TCGA, The Cancer Genome Atlas.

Another question is whether genetic alterations in HNRNPA2B1 and structural alterations in m6A are important for the precise function of the HNRNPA2B1 protein. HNRNPA2B1 may bind to m6A at structured and unstructured/linear sites because it contains diverse RRM domains. The types of genetic alterations, domain loci, and case numbers analyzed are further presented in Figure 3B. From the TCGA database, this study revealed a total of 102 mutations in HNRNPA2B1, including 69 missense mutations, 12 truncation mutations, 3 in-frame mutations, and 13 splice-site mutations. Genomic structural variations and gene fusions were detected in 5 patients. It contains the RRM_1 domain and the PF14259 domain. In UCEC and COAD, X293 splice site mutations were detected in up to 5 cases.

The most frequently altered mutation site was X293_splice, which was confirmed from the 3D structure of HNRNPA2B1 (Figure 3C). Do mutations affect the prognosis of patients? We analyzed the potential associations of HNRNPA2B1 mutation status with PFS, DFS, and OS in KIRC patients (Figure 3D). HNRNPA2B1 mutation status was associated with poor prognosis in KIRC patients, including PFS (P=4.08e−7), disease-specific survival (P=4.787e−3), and OS (P=0.02).

Protein phosphorylation analysis

Posttranslational modifications (PTMs) participate in different cellular processes, among which protein phosphorylation, which regulates many cellular activities, such as proliferation, apoptosis, differentiation, and cell signaling pathways, has been extensively investigated (39-41). However, dysregulation of the phosphorylation/dephosphorylation cascade is observed in various tumor types (40,41). The addition or removal of phosphate groups from specific amino acid residues on proteins regulates many biological responses. Therefore, multiple phosphorylation sites are targeted for the development of cancer drugs (40,42). The difference in HNRNPA2B1 phosphorylation levels between cancer and healthy tissues was assessed in diverse tumor datasets from CPTAC, and specific phosphorylation sites were determined via UALCAN (NP_112533.1_S212, S236, S259, S341, etc.; NP_002128.1_S90, S137, T164, etc.). The schematic diagram in Figure 4A shows the putative phosphorylation sites in the HNRNPA2B1 protein. The expression of genes in primary cancer samples compared with that in healthy samples at selected sites was analyzed, including BC (Figure 4B), OC (Figure 4C), CRC (Figure 4D), RCCC (Figure 4E), UCEC (Figure 4F), and LUAD (Figure 4G) samples. Based on the CPTAC dataset, we analyzed phosphorylation sites of the HNRNPA2B1 protein (including NP_112533.1_S212, S236, S259, S341; and NP_002128.1_S90, S137, T164, among others) in both normal and primary tumor tissues using UALCAN. We examined expression levels across various cancer types, including BC (Figure S2A), OC (Figure S2B), colon cancer (Figure S2C), UCEC (Figure S2D), and LUAD (Figure S2E).

Figure 4 Comparison of HNRNPA2B1 phosphorylation levels in normal tissues and primary tumor tissues. Based on the HNRNPA2B1 protein phosphorylation analysis of different tumors in the CPTAC dataset, we analyzed HNRNPA2B1-phosphorylated proteins (NP_112533.1_S212, S236, S259, S341, etc.; NP_002128.1_S90, S137, T164, etc.) in normal tumor tissues and primary tumor cells by UALCAN. Expression levels among developing tissues. Schematic diagram of HNRNPA2B1 protein 4 (A) showing positively phosphorylated protein sites. Expression levels between normal and primary tumor tissues at selected sites, including breast cancer (B), ovarian cancer (C), colon cancer (D), clear cell RCC (E), UCEC (F) and LUAD (G) samples. RCC, renal clear cell.

Methylation analysis of HNRNPA2B1 in various tumors

Cancer genesis and development can be regulated via epigenetic and genetic events (7,43). Methyltransferase-catalyzed methylation is a predominant epigenetic mechanism that regulates cell growth, differentiation, apoptosis, transformation, and the cell cycle in eukaryotes (44). Recently, advances have been made in epigenetics, which helps to further understand the mechanism of tumorigenesis while serving as early diagnostic and prognostic biomarkers for different tumor types. Despite the potential of methylation biomarkers to advance precision medicine, there are still additional issues that should be addressed prior to approval for clinical applications (45). Therefore, biochemical evidence supporting the interaction between HNRNPA2B1 and methylated DNA was obtained to explore the possible mechanism underlying HNRNPA2B1 upregulation in the induction of cancer genesis and progression and to predict its potential clinical application in the future. This study suggested that a divergent HNRNPA2B1 methylation status was responsible for the altered gene expression observed in cancer and healthy samples. DiseaseMeth2.0 and MEXPRESS online datasets were used to explore the associations between HNRNPA2B1 levels and DNA methylation status. The chr7:26235418–26237418 region of the CpG island shelf revealed markedly decreased HNRNPA2B1 levels within PCPG, TGCT, and bile duct cancers (P<0.001, Figure 5A). HNRNPA2B1 expression was significantly downregulated in UCS and TGCT according to the chr7:26227555–26229555 region of Down2KB. The chr7:26236020–26236101 region of the exon revealed markedly decreased HNRNPA2B1 expression in PCPG, PRAD, TGCT, and bile duct cancers. Second, the methylation status of HNRNPA2B1 DNA sequences was analyzed in PRAD, PCPG, TGCT, and UCS via the MEXPRESS visualization dataset. Five methylation sites (cg23401436, cg02315870, cg11010242, cg17545334, and cg24274982) associated with decreased HNRNPA2B1 expression were identified (Figure 5B). The methylation sites of HNRNPA2B1 are presented in the heatmap. The NEURL1B-related methylation site, cg23401436, is located within the 3’UTR as well as in the open sea region.

Figure 5 Methylation analysis of HNRNPA2B1 in normal and primary tumor tissues. Differential methylation levels of HNRNPA2B1 in five types of primary tumor tissues and corresponding normal tissues. Methylation was assessed via DiseaseMeth 2.0 (A). The methylation site of the HNRNPA2B1 DNA sequence associated with gene expression in PRAD, PCPG, TGCT and UCS was visualized using MEXPRESS. The expression of HNRNPA2B1 is illustrated by the blue line at the center of the plot. Pearson’s correlation coefficients and P values for methylation sites and query gene expression (B). *, P<0.05; **, P<0.01; ***, P<0.001. BMI, body mass index; CpG, cytosine-phosphate-guanine; N, normal; OS, overall survival; PSA, prostate-specific antigen; T, tumor.

The M6AREG database clearly describes the molecular mechanisms of each type of regulation. The mechanisms of m6A methylation and demethylation on RNA and their impacts on RNA processing (30) (Figure S3). Key enzymes involved in adding m6A modifications include METTL3, METTL16, supported by auxiliary factors RBM15, RBM15B, and VIRMA, with WTAP facilitating the localization of the MTC to target RNA sequences (30). The resultant m6A marks are recognized by HNRNPA2B1, which plays a critical role in various RNA processing events, notably enhancing miRNA processing. Conversely, FTO and ALKBH5 function as demethylases that remove m6A modifications, thereby modulating the functional state of the RNA. Specifically, HNRNPA2B1 binds to m6A-modified RNA, promoting the processing of primary miRNAs (pri-miRNAs) (10,30). This dynamic interplay between methylation and demethylation influences RNA metabolism, including the regulation of specific RNAs such as LINC01833, Let-7B, and TP53, affecting their roles in cellular processes. In summary, the addition and removal of m6A modifications by these enzymes orchestrate the lifecycle and functionality of RNAs through intricate regulatory networks involving RNA stability, localization, and processing (30). Different methylated regions associated with HNRNPA2B1 were presented by a heat map using MethSurv (https://biit.cs.ut.ee/methsurv/) (Figure S4).

Immune infiltration analysis data

Cancer genesis and progression depend on the interaction of tumor cells with stromal components. The tumor microenvironment (TME) is the “home” for tumor cells and is composed mainly of the extracellular matrix (ECM) and cancer-associated fibroblasts (CAFs) (46). Cancer cells induce fibroblast activation by secreting transforming growth factor beta (TGF-β), and activated CAFs have auxiliary effects on tumor progression, including remodeling of the ECM (collagen crosslinking), angiogenesis, and induction of epithelial-mesenchymal transition, which occurs via the secretion of growth factors and cytokines (47,48). CAFs are also involved in tumor-host immune interactions (47,49). Tumor-infiltrating immune cells (TIICs) constitute a major component of the TME and are closely associated with tumor progression and metastasis. CAFs present in the TME are involved in modulating TIIC functions (50,51). Diverse algorithms could be adopted for exploring the possible correlations between HNRNPA2B1 gene expression and CAF infiltration levels in different tumor types (Figure 6A,6B). A positive correlation between HNRNPA2B1 expression levels and CAF infiltration values was found in the TCGA datasets for the CESC, MESO, LIHC, and HNSC-HPV datasets, whereas a negative correlation was found with those in the HNSC-HPV⁺ and STAD datasets. In addition, according to our results, the overall map of CD8⁺ T cells in BRCA-Basal, PRAD, THCA, THYM, and UVM was positively correlated with HNRNPA2B1 expression (Figure S5).

Figure 6 Exploration of the potential relationship between the infiltration level of cancer-associated fibroblasts and HNRNPA2B1 gene expression in different cancer types in the TCGA. Different algorithms were used to explore potential correlations between the expression levels of the HNRNPA2B1 gene and the infiltration levels of cancer-associated fibroblasts in all types of cancers in TCGA (A,B). TCGA, The Cancer Genome Atlas; TPM, transcripts per million.

HNRNPA2B1-related partner enrichment

To understand the possible molecular mechanism by which HNRNPA2B1 influences tumor progression, this study focused on identifying the selective binding partners of HNRNPA2B1 along with other HNRNPA2B1-related proteins via enrichment analysis.

We present the expression levels of the HNRNPA2B1 gene across various tumors and their pathological stages (Figure S6). Using the HPA (https://www.proteinatlas.org/?spm=5176.28103460.0.0.178d1db8uaYgn1) dataset, we analyzed the expression and distribution of total HNRNPA2B1 protein in pan-cancer tissues (Figure S6A), as well as in specific cancer types including CRC (Figure S6B), prostate cancer (Figure S6C), BC (Figure S6D), lung cancer (Figure S6E), and corresponding normal tissues. Additionally, we examined the mRNA expression patterns of HNRNPA2B1 across different cancers (Figure S6F). This study first used STRING to identify possible HNRNPA2B1-binding proteins (Figure 7A). All tumor datasets from TCGA were combined with GEPIA2, and 100 genes that were most significantly associated with HNRNPA2B1 expression were identified. The expression correlation of HNRNPA2B1 with selected target genes, such as DHX9, HNRNPD, HNRNPR, SRSF1, TARDBP, and other genes, was analyzed (Figure 7B). Corresponding data for detailed tumor types are displayed as a heatmap (Figure 7C). Cross-analysis of HNRNPA2B1 binding and related genes (Figure 7D). KEGG pathway enrichment according to HNRNPA2B1-interacting and HNRNPA2B1-binding genes (Figure 7E). KEGG and GO enrichment analyses suggested that the “Notch pathway”, “MAPK pathway”, “IL17 pathway” and “HIF-1 pathway” may be involved in the tumor pathogenesis driven by HNRNPA2B1.

Figure 7 Exploration of the proteins associated with HNRNPA2B1 and genes related to HNRNPA2B1 expression, enrichment analysis, and study of the molecular mechanism of HNRNPA2B1 in tumorigenesis. HNRNPA2B1-related gene enrichment analysis. We first obtained available experimentally determined HNRNPA2B1-binding proteins via the STRING tool (A). The GEPIA2 method was used to obtain the top 100 HNRNPA2B1-related genes in the TCGA project, and the expression correlation between HNRNPA2B1 and selected target genes, such as DHX9, HNRNPD, HNRNPR, SRSF1, and TARDBP, was analyzed (B). The corresponding data for detailed cancer types are displayed in the form of a heatmap (C). Crossover analysis of HNRNPA2B1 binding and related genes (D). KEGG pathway analysis was performed on genes with which HNRNPA2B1 binds and interacts (E). KEGG, Kyoto Encyclopedia of Genes and Genomes; TCGA, The Cancer Genome Atlas; TPM, transcripts per million.

HNRNPA2B1 mutations in cancer and correlation analysis with drug sensitivity

Small molecule inhibitor drugs can block tumorigenesis, providing a potential approach to treat tumors harboring genetic mutations. To explore the HNRNPA2B1 gene mutation profile in tumors, the most common gene mutations were detected in 8,655 (84.46%) of the 10,247 tumor samples (Figure 8A). The correlation of HNRNPA2B1 expression with driver mutations in tumors was analyzed via the CAMOIP module. Our results revealed that HNRNPA2B1 expression in tumors was associated with TP53, PIK3CA, CSMD3, LRP1B, KRAS, PTEN, and BRAF driver mutations in the TCGA cohort.

Figure 8 HNRNPA2B1 gene mutation analysis and drug sensitivity correlation analysis of HNRNPA2B1 mutations and their correlation with drug sensitivity. A total of 10,247 tumor samples had altered mutation frequencies in 8,655 (84.46%) genes. Correlations of mutation frequencies in 700 driver genes were integrated via the OncoVar system (A). Effect of CAMOIP HNRNPA2B1 gene mutation on drug sensitivity correlation analysis via CAMOIP (B-D). **, P<0.01. IC₅₀, half-maximal inhibitory concentration; MT, mutant; TMB, tumor mutational burden; WT, wild type.

Immunotherapy has greatly changed the way in which tumor therapies are carried out. The introduction of immune checkpoint inhibitors (ICIs) has significantly improved cancer management and survival outcomes. Comprehensive assessment of the immune status of patients with tumors, the formulation of individualized therapy, and the precise combination of therapeutic strategies are prerequisites for achieving favorable clinical outcomes under ICI treatment. We screened for the drug sensitivity of HNRNPA2B1 mutants and found that HNRNPA2B1 mutations significantly lowered the IC50 values for bexarotene, PD173074, fedratinib, NSC-87877, and KIN001–236 (P<0.05). However, HNRNPA2B1 mutation significantly increased the IC₅₀ values for IPA-3, EHT-1864, A2628, GW441756, and QL-VIII58 (P<0.05, Figure 8B-8D).

The HNRNPA2B1 gene targets HNRNPA2B1 in herbs/ingredients

With the increasing incidence of malignant tumors, the side effects and drug resistance associated with chemoradiotherapy and targeted therapy continue to increase. However, the use of TCM and its combination for treating cancer has received extensive attention. TCM has developed over thousands of years of clinical practice, has an important effect on disease treatment, and is increasingly used worldwide. However, unraveling the underlying molecular mechanism of TCM is still a daunting task because of the molecular diversity of TCM components and the complexity of TCM interactions with the human body. Therefore, finding effective antitumor drugs, such as small molecule inhibitors, from natural plants has become one of the primary goals of researchers.

Using the HERB database, we identified 9 TCMs that target HNRNPA2B1, such as dog kidney, wine, sheep whip, sheep external kidney, and pig kidney. In addition, we also identified 10 botanical herbal ingredients that target HNRNPA2B1, including 17-beta-estradiol, 17-betaoestradiol, 3,4-benzopyrene, allitridin, androgen, dandelion game ethyl alcohol, evoden, glycerin, meletin, and quercetin. These natural TCM ingredients have a wide range of antitumor effects on cultured tumor cells. Many HNRNPA2B1-related disorders have been identified in the HERB database, including 131 diseases such as Alzheimer’s disease and malignant neoplasms of the breast (Figure 9).

Figure 9 Traditional Chinese medicines/components targeting HNRNPA2B1. The BenCaoZuJian database was analyzed, and nine traditional herbal medicines/components that target HNRNPA2B1 were identified (A). Ten plant traditional Chinese medicine components that target HNRNPA2B1 (B). The diseases related to HNRNPA2B1 in the BenCaoZuJian database included types of tumors, Alzheimer’s disease and other diseases (C).

The analysis identified 2,446 differentially expressed genes (DEGs), including 468 upregulated and 616 downregulated genes, with a false discovery rate (FDR) <0.05 and an absolute log₂FC (|log2FC|) >1 (Figure S7A).

To identify the consensus motifs for target genes of HNRNPA2B1, we focused on the top 100 most reliable targets as determined by confidence scores. Our analysis included only those targets with binding peaks detected by CLIP-seq or modification peaks identified through MeRIP-seq. These peak clusters were designated as the target sequences. Additionally, a set of background clusters was generated using the shuffleBed tool from BEDTools, which facilitates the random shuffling of regions equivalent in size to the clusters across gene regions (Figure S7B).

• Enriched biological functions and pathways (Figure S8): proteolytic regulation—terms such as positive regulation of proteolysis (P=0.01) and regulation of peptidase activity (P=0.02) were prominently enriched, driven by genes like caspase-1 (CASP1, a key apoptosis executor), cathepsin D (CTSD, lysosomal protease), and ADAM9 (metallopeptidase involved in ECM remodeling), suggesting active roles in protein degradation and cellular remodeling.
• Cell cycle and cytoskeletal dynamics: pathways including spindle organization and actin filament bundle assembly were linked to aurora kinase A (AURKA, essential for mitotic spindle formation) and cell cycle regulator (CDC20), highlighting their contributions to mitotic fidelity and cytoskeletal integrity.
• Stress response: enrichment in apoptosis (e.g., BOK, TNFRSF10B) and MAPK signaling (e.g., MAPK9) underscored adaptive responses to cellular stress.
• Gene-function network insights: core gene nodes—the network (constructed using Cytoscape) identified CASP1 and BOK as central hubs in apoptosis, AURKA and CDC20 in cell cycle regulation, and CTSD/ADAM9 in proteolytic processes. These genes formed distinct functional modules, such as the “apoptosis-proteolysis cluster” and the “cell cycle-cytoskeleton module”, reflecting their coordinated roles in disease mechanisms.
• Expression trends: upregulated genes (e.g., AURKA, MYC) were enriched in proliferation-related pathways, while downregulated genes (e.g., certain apoptotic regulators) suggested loss of inhibitory control.
• Biological implications: dual regulation of cell fate—co-occurrence of pro-apoptotic (e.g., CASP1) and pro-proliferative (e.g., AURKA) genes implies a dynamic balance between cell survival and death. Microenvironment interaction—proteases (CTSD, ADAM9) and adhesion molecules (e.g., CLDN3) may mediate tumor-stroma crosstalk, influencing disease progression. Pathway crosstalk—overlap between MAPK signaling and apoptosis/cell cycle pathways (e.g., MAPK9) suggests synergistic regulatory mechanisms.

Discussion

Malignant transformation is an inherently complicated, highly dynamic, and multistep process that is driven primarily by mutations and/or aberrant expression of oncogenes. The malignant potential becomes prominent over time, thereby driving the unlimited proliferation of tumor cells (35,36,52). Heterogeneous interactions between tumor and nontumor cells are observed in the TME. The abnormal expression of driver genes affects the expression of other normal genes. The heterogeneity within tumors develops through the targeted regulation of driver genes. Gene methylation or protein phosphorylation eventually leads to the replication of tumor cells, tumor immune escape, and resistance to apoptosis (35,36,53-55). Abnormal driver genes cooperate with various protein complexes to switch target genes “on” or “off” by binding to specific regions of DNA, thereby promoting tumor angiogenesis, immune escape, local invasion, and distant metastases (37,38). Given the biological characteristics of tumor driver genes, such as genome instability, mutation, methylation and phosphorylation, signal transduction, and tumor immune escape, identifying new targets and understanding their mechanism of action may help improve the diagnosis and treatment of malignant tumors. Emerging evidence highlights the critical roles of m6A-binding proteins in cancer pathogenesis across multiple malignancies. In MM, elevated expression of the m6A reader HNRNPA2B1 demonstrates significant association with poorer clinical outcomes. Mechanistic studies reveal its oncogenic function through dual RNA stabilization mechanisms—maintaining ILF3 mRNA stability and subsequently prolonging AKT3 transcript longevity, thereby driving tumor cell proliferation and suppressing apoptotic pathways (56,57). HNRNPA2B1 binds to sites containing RGm6AC on nuclear RNA both in vivo and in vitro. HNRNPA2B1 participates in RNA transcription, splicing, stability, and translation by recognizing and binding to specific RNA substrates and DNA motifs, and regulates the expression of multiple genes. HNRNPA2B1 regulates the alternative splicing of exons in a group of transcripts in a manner similar to the m6A “writer” METTL3. The global impact of HNRNPA2B1 depletion on alternative splicing is highly correlated with the global impact of METTL3 depletion on alternative splicing. The depletion of HNRNPA2B1 also impairs the nuclear processing of a subset of microRNAs (miRNAs), whose maturation relies on METTL3 activity. Additionally, HNRNPA2B1 interacts with DGCR8 protein, which is a component of the pri-miRNA microprocessor complex and promotes the processing of pri-miRNAs. Our findings suggest that HNRNPA2B1 serves as a nuclear reader and effector of m6A marks (13).

m6A modification is one of the most frequently observed genetic events in eukaryotic RNA and has many important effects on tumors (58,59). m6A methylation is a reversible process that relies on the participation of m6A-binding proteins (readers), demethylases (erasers), and methyltransferases (words) (59-61). Methyltransferases catalyze the modification of m6A via adenylation of mRNAs (59,62,63). Examples of demethylases include FTO and ALKHB5. FTO demethylates single-stranded DNA and RNA (61,64). ALKBH5 demethylates mRNA and possesses an alanine-rich domain as well as a specific coiled-coil structure at the N-terminus (65,66). HNRNPA2B1 is an RNA-binding protein that acts as a m6A reader and binds to nascent RNA to subtly regulate aspects of RNA metabolism (15,67), such as mRNA splicing and miRNA maturation (27). HNRNPA2B1 is involved in multiple cellular signaling pathways, including the ERK/MAPK (68) and STAT3 pathways (69). Recognition of viral nucleic acid by HNRNPA2B1 via a pattern recognition receptor (PRR) triggers the host’s innate antiviral immunity; as a result, type I interferons and proinflammatory factors are produced (70). In MM, HNRNPA2B1 has been identified as a critical factor promoting disease advancement via TLR4 signaling. This study highlights how HNRNPA2B1 recognizes and enriches at the m6A sites of TLR4 mRNA, thereby affecting both its methylation status and transcriptional activity. These findings underscore the importance of RNA methylation in modulating immune responses and drug resistance in MM, suggesting HNRNPA2B1 as a promising therapeutic target (71). Similarly, in hepatocellular carcinoma (HCC), HNRNPA2B1’s overexpression is associated with poor prognosis and enhanced tumor characteristics such as proliferation, migration, and invasion. The use of CRISPR-Cas9 knockout experiments demonstrated that HNRNPA2B1 regulates key metabolic pathways like gluconeogenesis by downregulating PCK1 expression via m6A methylation. This indicates not only HNRNPA2B1’s impact on cellular processes but also its significant role in metabolic reprogramming within cancer cells, further supporting its potential as a therapeutic target (72). Moreover, the involvement of HNRNPA2B1 extends beyond direct cellular effects to include its role in regulating miRNAs. It was shown that m6A modification on specific miRNAs weakens their coupling with AGO2, impairs their function on target mRNAs, and determines their delivery into extracellular vesicles (EVs). This process is facilitated by the recognition of m6A-modified RNAs by HNRNPA2B1, emphasizing its dual role in intracellular regulation and intercellular communication (73). However, while these studies advance our knowledge of HNRNPA2B1’s functions, several aspects remain underexplored. Specifically, the detailed mechanisms underlying HNRNPA2B1’s interaction with target RNAs and how this interaction is influenced by methylation status require further investigation. Additionally, considering recent literature focusing on HNRNPA2B1’s activity as an m6A reader, more comprehensive discussions are needed to fully understand its functional implications across different cancers. Future research should aim to address these gaps, providing deeper insights into the complex regulatory networks governed by HNRNPA2B1 and potentially uncovering new avenues for targeted therapy. Furthermore, exploring the crosstalk between HNRNPA2B1 and other molecular pathways could reveal additional layers of complexity in cancer biology, paving the way for innovative treatment strategies.

The biological significance of HNRNPA2B1 phosphorylation lies in its role as a dynamic regulatory mechanism that fine-tunes diverse cellular processes. By modulating RNA binding affinity, phosphorylation governs mRNA splicing and miRNA biogenesis, directly impacting transcript diversity, protein isoform expression, and gene silencing. Under stress conditions, it promotes HNRNPA2B1’s aggregation into stress granules, halting translation to reprogram cellular recovery efforts. Phosphorylation also enhances its function in antiviral defense by boosting viral DNA sensing and interferon production through STING or MAVS pathways. Kinase-driven phosphorylation integrates extracellular signals with RNA processing, linking it to cell cycle control, proliferation, and apoptosis. Dysregulation of this modification contributes to disease pathogenesis, including neurodegeneration [via toxic protein aggregation in amyotrophic lateral sclerosis/frontotemporal dementia (ALS/FTD)] and cancer (through disrupted miRNA-mediated oncogene or tumor suppressor regulation). Additionally, phosphorylation acts as a molecular switch, directing its nuclear-cytoplasmic shuttling to balance roles in nuclear RNA metabolism and cytoplasmic immune signaling. Collectively, HNRNPA2B1 phosphorylation serves as a versatile orchestrator of RNA metabolism, stress adaptation, immune responses, and cellular homeostasis, with profound implications for both physiological processes and disease mechanisms.

Genetic alterations in HNRNPA2B1 may be closely related to oncogenic functions such as proliferation, apoptosis inhibition, tumor development, invasion, metastasis, and immune evasion of malignant tumors (74). In Alzheimer’s disease, downregulation of HNRNPA2B1 prevents oTau-c from binding to m6A, causing reduced protein synthesis and inhibiting tau fibril formation, nuclear envelope disruption, and oTau-induced neurodegeneration, confirming that oTau, HNRNPA2B1, and m6A form a complex that contributes to the comprehensive stress response of oTau (67). HNRNPA2B1 binds to m6A and induces RNA unfolding, increasing the accessibility of labeled RNA to HNRNPA2B1 (15,27,67). Genes encoding interacting partners of m6A methylation proteins have been shown to be involved in promoting tumorigenesis and metastasis (75-78). However, there are only a few reports on the role of HNRNPA2B1 in tumorigenesis. HNRNPA2B1 acts as a nuclear reader in tumors, but the molecular mechanism by which HNRNPA2B1 preferentially binds to downstream m6A markers and facilitates other nuclear processing events remains unclear.

Therefore, the present work analyzed HNRNPA2B1 expression profiles in 33 pairs of cancer and adjacent healthy tissues from the TIMER2 database combined with the GEPIA2 database. HNRNPA2B1 is significantly overexpressed in BRCA, BLCA, CHOL, CESC, DLBC, COAD, GBM, ESCA, LGG, HNSC, LUAD, LIHC, LUSC, OV, PRAD, PAAD, READ, THYM, STAD, and UCEC, which is comparable to the corresponding normal tissues. In contrast, significantly lower expression of HNRNPA2B1 was observed in KICH, KIRC, and THCA. Next, according to the HPA dataset analysis, weak to moderate HNRNPA2B1 staining was observed in most of the samples obtained from various tumor types, whereas strong positive HNRNPA2B1 staining was observed in colorectal, prostate, lung, and BC samples. HNRNPA2B1 mRNA and protein levels differ between tumor tissues and corresponding normal tissues in various cancer types. To explore its clinical significance, HNRNPA2B1 levels were positively correlated with TNM stage in ACC, SKCM, TGCT, COAD, and LIHC via the GEPIA2 database. Survival analysis revealed that high HNRNPA2B1 expression was associated with poor prognosis in ACC, LGG, LIHC, LUAD, and KICH, which indicated that HNRNPA2B1 overexpression was related to poor prognostic outcomes. Additionally, HNRNPA2B1 mutations were linked to a dismal prognosis in KIRC. In contrast, lower expression of HNRNPA2B1 was linked to poor DFS in THCA and KICH patients.

Cancer is one of the diseases with the fastest increases in morbidity and mortality rates. The mainstays of treatment for cancer are surgery, radiotherapy, and chemotherapy, but their side effects cannot be ignored. Therefore, combined treatment methods should be the focus of research and will have important clinical significance. The long-term goal is to find new therapeutic targets with low toxicity and effective killing of targeted tumor cells that can be combined with molecular radiotherapy (MRT), chemotherapy, immunotherapy, α/β combined MRT and multitarget MRT. This study explored the effects of HNRNPA2B1 on DNA methylation, protein phosphorylation, gene mutation, and drug sensitivity.

The occurrence of tumors involves a variety of mechanisms that are precisely regulated by different signaling pathways. Previous studies have shown that HNRNPA2B1 enhances CRC cell growth by regulating their apoptosis and cell cycle via the activation of the ERK/MAPK pathway (68,79-81). This study sheds new light on the development of HNRNPA2B1 as a potential candidate for anti-CRC therapy. Furthermore, HNRNPA2B1 can enhance tumorigenesis and metastasis of oral squamous cell carcinoma (OSCC) by modulating EMT via the LINE-1/TGF-β1/Smad2/Slug pathway (82). Another study revealed that HNRNPA2B1 negatively regulates BC metastasis, which contradicts other findings. HNRNPA2B1 may suppress BC metastasis, except for luminal A and triple-negative BC (TNBC) subtypes (83). HNRNPA2B1 may perform dual functions in different cancer subtypes.

To support tumor development and survival, the TME provides the necessary molecular components for tumor proliferation, metastasis, and invasion. The TME is very complex, and its interaction with tumor metabolism and the tumor immune system is very intimate. Tumor cells develop a specific TME that facilitates the immune escape of cancer cells and provides support for tumor evolution (84). Tumor antigens are occasionally recognized via the T cells of patients with tumors. Therefore, adoptive T-cell therapy may serve as a novel immunotherapeutic approach (85,86). CAFs have critical effects on promoting cancer occurrence, as they communicate with other stromal cells as well as tumor cells by secreting various cytokines, inhibiting the function of immune cells, regulating the remodeling of the ECM, and creating a barrier to the infiltration of drugs and immune cells (87,88). The present study also explored the associations of HNRNPA2B1 levels with the degree of fibroblast infiltration together with the number of CD8⁺ T cells in tumors. Therefore, therapeutic strategies targeting CAFs are expected to inhibit tumor progression.

Chinese herbal medicine is the predominant source of natural active molecules with different characteristics, including anti-inflammatory and anticancer effects with few side effects, but the underlying mechanism is poorly understood and under continuous research. The use of natural phytochemicals that target HNRNPA2B1 provides a theoretical basis for improving tumor treatment. In this study, we screened many natural plant herbal medicines that improve the therapeutic effect on tumors by regulating the expression of HNRNPA2B1.

This study, while providing valuable insights into the role of HNRNPA2B1 in multiple malignant tumors, is not without limitations. First, the data used in this research were drawn from public databases such as the TCGA and GEO. Although these databases contain a large amount of data, potential biases might exist. Second, the bioinformatics analysis methods employed have certain constraints. The algorithms used to analyze the degree of immune infiltration, gene enrichment, and other aspects are based on existing models and assumptions. These algorithms may not fully capture the complex biological interactions and regulatory mechanisms in tumors. Finally, the complexity of tumorigenesis and tumor progression is extremely high. This study focused mainly on HNRNPA2B1, but in reality, multiple genes and pathways interact with each other in the development of tumors. Ignoring the complex network of these interactions may lead to an incomplete understanding of the tumorigenic process.

Conclusions

In conclusion, the analysis of HNRNPA2B1 across cancers revealed that its expression was significantly correlated with pathological stage, protein phosphorylation, clinical prognosis, methylation, tumor mutation, immune infiltration, and various important signaling pathways involved in tumorigenesis and tumor development. Clinical research on tumor samples has determined the effects of HNRNPA2B1 on malignant transformation, including oncogenesis, development, metastasis, infiltration of immune cells, evolution, and prognostic outcomes of various cancer types, laying the theoretical foundation for future applications. This research identified potential targets for further drug development for the treatment of tumors.

Acknowledgments

We would like to thank the native English-speaking language specialists at Sagesci for their professional language editing services.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2024-2616/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2024-2616/prf

Funding: This study was supported partly by grants from the Science and Technology Development Fund of Shanghai Pudong New Area (No. PKJ2021-Y04); the Talents Training Program of Seventh People’s Hospital of Shanghai University of Traditional Chinese Medicine (No. QMX2022-04); the Budgetary Fund of Shanghai University of Traditional Chinese Medicine (No. 2021LK058); and the Key Discipline Construction Project of Pudong Health Burea of Shanghai (No. PWZxk2022-04).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2024-2616/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Siegel RL, Kratzer TB, Giaquinto AN, et al. Cancer statistics, 2025. CA Cancer J Clin 2025;75:10-45. [Crossref] [PubMed]
2. Hakala S, Hämäläinen A, Sandelin S, et al. Detection of Cancer Stem Cells from Patient Samples. Cells 2025;14:148. [Crossref] [PubMed]
3. Zhou Y, Jian N, Jiang C, et al. m(6)A modification in non-coding RNAs: Mechanisms and potential therapeutic implications in fibrosis. Biomed Pharmacother 2024;179:117331. [Crossref] [PubMed]
4. Kang Z, Li R, Liu C, et al. m(6)A-modified cenRNA stabilizes CENPA to ensure centromere integrity in cancer cells. Cell 2024;187:6035-6054.e27. [Crossref] [PubMed]
5. Liu C, Liang H, Wan AH, et al. Decoding the m(6)A epitranscriptomic landscape for biotechnological applications using a direct RNA sequencing approach. Nat Commun 2025;16:798. [Crossref] [PubMed]
6. Wang J, Zhang J, Liu H, et al. N6-methyladenosine reader hnRNPA2B1 recognizes and stabilizes NEAT1 to confer chemoresistance in gastric cancer. Cancer Commun (Lond) 2024;44:469-90. [Crossref] [PubMed]
7. Wang Y, Wang Y, Patel H, et al. Epigenetic modification of m(6)A regulator proteins in cancer. Mol Cancer 2023;22:102. [Crossref] [PubMed]
8. Lin H, Wang Y, Wang P, et al. Mutual regulation between N6-methyladenosine (m6A) modification and circular RNAs in cancer: impacts on therapeutic resistance. Mol Cancer 2022;21:148. [Crossref] [PubMed]
9. Bao Y, Zhai J, Chen H, et al. Targeting m(6)A reader YTHDF1 augments antitumour immunity and boosts anti-PD-1 efficacy in colorectal cancer. Gut 2023;72:1497-509. [Crossref] [PubMed]
10. Xiao S, Ma S, Sun B, et al. The tumor-intrinsic role of the m(6)A reader YTHDF2 in regulating immune evasion. Sci Immunol 2024;9:eadl2171. [Crossref] [PubMed]
11. Chen L, He Y, Zhu J, et al. The roles and mechanism of m(6)A RNA methylation regulators in cancer immunity. Biomed Pharmacother 2023;163:114839. [Crossref] [PubMed]
12. Zhang J, Liu B, Xu C, et al. Cholesterol homeostasis confers glioma malignancy triggered by hnRNPA2B1-dependent regulation of SREBP2 and LDLR. Neuro Oncol 2024;26:684-700. [Crossref] [PubMed]
13. An Y, Duan H. The role of m6A RNA methylation in cancer metabolism. Mol Cancer 2022;21:14. [Crossref] [PubMed]
14. Jiang X, Liu B, Nie Z, et al. The role of m6A modification in the biological functions and diseases. Signal Transduct Target Ther 2021;6:74. [Crossref] [PubMed]
15. Jiang F, Tang X, Tang C, et al. HNRNPA2B1 promotes multiple myeloma progression by increasing AKT3 expression via m6A-dependent stabilization of ILF3 mRNA. J Hematol Oncol 2021;14:54. [Crossref] [PubMed]
16. Hu T, Zeng C, Song Z, et al. HNRNPA2B1 and HNRNPR stabilize ASCL1 in an m6A-dependent manner to promote neuroblastoma progression. Biochim Biophys Acta Mol Basis Dis 2024;1870:167050. [Crossref] [PubMed]
17. Jiang J, Zhu J, Qiu P, et al. HNRNPA2B1-mediated m6A modification of FOXM1 promotes drug resistance and inhibits ferroptosis in endometrial cancer via regulation of LCN2. Funct Integr Genomics 2023;24:3. [Crossref] [PubMed]
18. Li T, Fu J, Zeng Z, et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res 2020;48:W509-14. [Crossref] [PubMed]
19. Li C, Tang Z, Zhang W, et al. GEPIA2021: integrating multiple deconvolution-based analysis into GEPIA. Nucleic Acids Res 2021;49:W242-6. [Crossref] [PubMed]
20. Karlsson M, Zhang C, Méar L, et al. A single-cell type transcriptomics map of human tissues. Sci Adv 2021;7:eabh2169. [Crossref] [PubMed]
21. Sjöstedt E, Zhong W, Fagerberg L, et al. An atlas of the protein-coding genes in the human, pig, and mouse brain. Science 2020;367:eaay5947. [Crossref] [PubMed]
22. Chandrashekar DS, Bashel B, Balasubramanya SAH, et al. UALCAN: A Portal for Facilitating Tumor Subgroup Gene Expression and Survival Analyses. Neoplasia 2017;19:649-58. [Crossref] [PubMed]
23. Fekete JT, Győrffy B. ROCplot.org: Validating predictive biomarkers of chemotherapy/hormonal therapy/anti-HER2 therapy using transcriptomic data of 3,104 breast cancer patients. Int J Cancer 2019;145:3140-51. [Crossref] [PubMed]
24. Gao J, Aksoy BA, Dogrusoz U, et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal 2013;6:pl1. [Crossref] [PubMed]
25. Pi J, Wang W, Ji M, et al. YTHDF1 Promotes Gastric Carcinogenesis by Controlling Translation of FZD7. Cancer Res 2021;81:2651-65. [Crossref] [PubMed]
26. Bao X, Zhang Y, Li H, et al. RM2Target: a comprehensive database for targets of writers, erasers and readers of RNA modifications. Nucleic Acids Res 2023;51:D269-79. [Crossref] [PubMed]
27. Alarcón CR, Goodarzi H, Lee H, et al. HNRNPA2B1 Is a Mediator of m(6)A-Dependent Nuclear RNA Processing Events. Cell 2015;162:1299-308. [Crossref] [PubMed]
28. Koch A, Jeschke J, Van Criekinge W, et al. MEXPRESS update 2019. Nucleic Acids Res 2019;47:W561-5. [Crossref] [PubMed]
29. Modhukur V, Iljasenko T, Metsalu T, et al. MethSurv: a web tool to perform multivariable survival analysis using DNA methylation data. Epigenomics 2018;10:277-88. [Crossref] [PubMed]
30. Liu S, Chen L, Zhang Y, et al. M6AREG: m6A-centered regulation of disease development and drug response. Nucleic Acids Res 2023;51:D1333-44. [Crossref] [PubMed]
31. Fang S, Dong L, Liu L, et al. HERB: a high-throughput experiment- and reference-guided database of traditional Chinese medicine. Nucleic Acids Res 2021;49:D1197-206. [Crossref] [PubMed]
32. Schober P, Vetter TR. Survival Analysis and Interpretation of Time-to-Event Data: The Tortoise and the Hare. Anesth Analg 2018;127:792-8. [Crossref] [PubMed]
33. Barakat A, Mittal A, Ricketts D, et al. Understanding survival analysis: actuarial life tables and the Kaplan-Meier plot. Br J Hosp Med (Lond) 2019;80:642-6. [Crossref] [PubMed]
34. Obuchowski NA, Bullen JA. Receiver operating characteristic (ROC) curves: review of methods with applications in diagnostic medicine. Phys Med Biol 2018;63:07TR01. [Crossref] [PubMed]
35. Martínez-Jiménez F, Muiños F, Sentís I, et al. A compendium of mutational cancer driver genes. Nat Rev Cancer 2020;20:555-72. [Crossref] [PubMed]
36. Porta-Pardo E, Valencia A, Godzik A. Understanding oncogenicity of cancer driver genes and mutations in the cancer genomics era. FEBS Lett 2020;594:4233-46. [Crossref] [PubMed]
37. Sapozhnikov DM, Szyf M. Unraveling the functional role of DNA demethylation at specific promoters by targeted steric blockage of DNA methyltransferase with CRISPR/dCas9. Nat Commun 2021;12:5711. [Crossref] [PubMed]
38. Ledford H. Reversal of biological clock restores vision in old mice. Nature 2020;588:209. [Crossref] [PubMed]
39. Pham L, Arroum T, Wan J, et al. Regulation of mitochondrial oxidative phosphorylation through tight control of cytochrome c oxidase in health and disease - Implications for ischemia/reperfusion injury, inflammatory diseases, diabetes, and cancer. Redox Biol 2024;78:103426. [Crossref] [PubMed]
40. Pan S, Chen R. Pathological implication of protein post-translational modifications in cancer. Mol Aspects Med 2022;86:101097. [Crossref] [PubMed]
41. Bhardwaj A, Panepinto MC, Ueberheide B, et al. A mechanism for hypoxia-induced inflammatory cell death in cancer. Nature 2025;637:470-7. [Crossref] [PubMed]
42. Liu R, Mathieu C, Berthelet J, et al. Human Protein Tyrosine Phosphatase 1B (PTP1B): From Structure to Clinical Inhibitor Perspectives. Int J Mol Sci 2022;23:7027. [Crossref] [PubMed]
43. Yue SW, Liu HL, Su HF, et al. m6A-regulated tumor glycolysis: new advances in epigenetics and metabolism. Mol Cancer 2023;22:137. [Crossref] [PubMed]
44. Yamaguchi K, Chen X, Rodgers B, et al. Non-canonical functions of UHRF1 maintain DNA methylation homeostasis in cancer cells. Nat Commun 2024;15:2960. [Crossref] [PubMed]
45. Papanicolau-Sengos A, Aldape K. DNA Methylation Profiling: An Emerging Paradigm for Cancer Diagnosis. Annu Rev Pathol 2022;17:295-321. [Crossref] [PubMed]
46. Biffi G, Tuveson DA. Diversity and Biology of Cancer-Associated Fibroblasts. Physiol Rev 2021;101:147-76. [Crossref] [PubMed]
47. Chandra Jena B, Sarkar S, Rout L, et al. The transformation of cancer-associated fibroblasts: Current perspectives on the role of TGF-β in CAF mediated tumor progression and therapeutic resistance. Cancer Lett 2021;520:222-32. [Crossref] [PubMed]
48. Kieffer Y, Hocine HR, Gentric G, et al. Single-Cell Analysis Reveals Fibroblast Clusters Linked to Immunotherapy Resistance in Cancer. Cancer Discov 2020;10:1330-51. [Crossref] [PubMed]
49. Ren Q, Zhang P, Lin H, et al. A novel signature predicts prognosis and immunotherapy in lung adenocarcinoma based on cancer-associated fibroblasts. Front Immunol 2023;14:1201573. [Crossref] [PubMed]
50. Geng S, Xiang T, Zhang Y, et al. Safe engineering of cancer-associated fibroblasts enhances checkpoint blockade immunotherapy. J Control Release 2023;356:272-87. [Crossref] [PubMed]
51. Chen C, Wang Z, Ding Y, et al. Tumor microenvironment-mediated immune evasion in hepatocellular carcinoma. Front Immunol 2023;14:1133308. [Crossref] [PubMed]
52. Elliott K, Larsson E. Non-coding driver mutations in human cancer. Nat Rev Cancer 2021;21:500-9. [Crossref] [PubMed]
53. Pirozzi CJ, Yan H. The implications of IDH mutations for cancer development and therapy. Nat Rev Clin Oncol 2021;18:645-61. [Crossref] [PubMed]
54. Dai X, Ren T, Zhang Y, et al. Methylation multiplicity and its clinical values in cancer. Expert Rev Mol Med 2021;23:e2. [Crossref] [PubMed]
55. Taddei ML, Pardella E, Pranzini E, et al. Role of tyrosine phosphorylation in modulating cancer cell metabolism. Biochim Biophys Acta Rev Cancer 2020;1874:188442. [Crossref] [PubMed]
56. Liu R, Zhong Y, Chen R, et al. m(6)A reader hnRNPA2B1 drives multiple myeloma osteolytic bone disease. Theranostics 2022;12:7760-74. [Crossref] [PubMed]
57. Wang J, Zuo Y, Lv C, et al. N6-methyladenosine regulators are potential prognostic biomarkers for multiple myeloma. IUBMB Life 2023;75:137-48. [Crossref] [PubMed]
58. Zhou Z, Lv J, Yu H, et al. Mechanism of RNA modification N6-methyladenosine in human cancer. Mol Cancer 2020;19:104. [Crossref] [PubMed]
59. Chen XY, Zhang J, Zhu JS. The role of m(6)A RNA methylation in human cancer. Mol Cancer 2019;18:103. [Crossref] [PubMed]
60. Sun T, Wu R, Ming L. The role of m6A RNA methylation in cancer. Biomed Pharmacother 2019;112:108613. [Crossref] [PubMed]
61. Jia G, Fu Y, Zhao X, et al. N6-methyladenosine in nuclear RNA is a major substrate of the obesity-associated FTO. Nat Chem Biol 2011;7:885-7. [Crossref] [PubMed]
62. Lin S, Choe J, Du P, et al. The m(6)A Methyltransferase METTL3 Promotes Translation in Human Cancer Cells. Mol Cell 2016;62:335-45. [Crossref] [PubMed]
63. Yang X, Zhang S, He C, et al. METTL14 suppresses proliferation and metastasis of colorectal cancer by down-regulating oncogenic long non-coding RNA XIST. Mol Cancer 2020;19:46. [Crossref] [PubMed]
64. Huang Y, Su R, Sheng Y, et al. Small-Molecule Targeting of Oncogenic FTO Demethylase in Acute Myeloid Leukemia. Cancer Cell 2019;35:677-691.e10. [Crossref] [PubMed]
65. Jin D, Guo J, Wu Y, et al. m(6)A demethylase ALKBH5 inhibits tumor growth and metastasis by reducing YTHDFs-mediated YAP expression and inhibiting miR-107/LATS2-mediated YAP activity in NSCLC. Mol Cancer 2020;19:40. [Crossref] [PubMed]
66. Guo X, Li K, Jiang W, et al. RNA demethylase ALKBH5 prevents pancreatic cancer progression by posttranscriptional activation of PER1 in an m6A-YTHDF2-dependent manner. Mol Cancer 2020;19:91. [Crossref] [PubMed]
67. Jiang L, Lin W, Zhang C, et al. Interaction of tau with HNRNPA2B1 and N(6)-methyladenosine RNA mediates the progression of tauopathy. Mol Cell 2021;81:4209-4227.e12. [Crossref] [PubMed]
68. Tang J, Chen Z, Wang Q, et al. hnRNPA2B1 Promotes Colon Cancer Progression via the MAPK Pathway. Front Genet 2021;12:666451. [Crossref] [PubMed]
69. Gao LB, Zhu XL, Shi JX, et al. HnRNPA2B1 promotes the proliferation of breast cancer MCF-7 cells via the STAT3 pathway. J Cell Biochem 2021;122:472-84. [Crossref] [PubMed]
70. Wang L, Wen M, Cao X. Nuclear hnRNPA2B1 initiates and amplifies the innate immune response to DNA viruses. Science 2019;365:eaav0758. [Crossref] [PubMed]
71. Jia C, Guo Y, Chen Y, et al. HNRNPA2B1-mediated m6A modification of TLR4 mRNA promotes progression of multiple myeloma. J Transl Med 2022;20:537. [Crossref] [PubMed]
72. Hao W, Chen Z, Tang J, et al. hnRNPA2B1 promotes the occurrence and progression of hepatocellular carcinoma by downregulating PCK1 mRNA via a m6A RNA methylation manner. J Transl Med 2023;21:861. [Crossref] [PubMed]
73. Garbo S, D'Andrea D, Colantoni A, et al. m6A modification inhibits miRNAs' intracellular function, favoring their extracellular export for intercellular communication. Cell Rep 2024;43:114369. [Crossref] [PubMed]
74. Dai S, Zhang J, Huang S, et al. HNRNPA2B1 regulates the epithelial-mesenchymal transition in pancreatic cancer cells through the ERK/snail signalling pathway. Cancer Cell Int 2017;17:12. [Crossref] [PubMed]
75. Chen XY, Liang R, Yi YC, et al. The m(6)A Reader YTHDF1 Facilitates the Tumorigenesis and Metastasis of Gastric Cancer via USP14 Translation in an m(6)A-Dependent Manner. Front Cell Dev Biol 2021;9:647702. [Crossref] [PubMed]
76. Dixit D, Prager BC, Gimple RC, et al. The RNA m6A Reader YTHDF2 Maintains Oncogene Expression and Is a Targetable Dependency in Glioblastoma Stem Cells. Cancer Discov 2021;11:480-99. [Crossref] [PubMed]
77. Yu D, Pan M, Li Y, et al. RNA N6-methyladenosine reader IGF2BP2 promotes lymphatic metastasis and epithelial-mesenchymal transition of head and neck squamous carcinoma cells via stabilizing slug mRNA in an m6A-dependent manner. J Exp Clin Cancer Res 2022;41:6. [Crossref] [PubMed]
78. Zhuang M, Li X, Zhu J, et al. The m6A reader YTHDF1 regulates axon guidance through translational control of Robo3.1 expression. Nucleic Acids Res 2019;47:4765-77. [Crossref] [PubMed]
79. Lu Y, Zou R, Gu Q, et al. CRNDE mediated hnRNPA2B1 stability facilitates nuclear export and translation of KRAS in colorectal cancer. Cell Death Dis 2023;14:611. [Crossref] [PubMed]
80. Sun YK, Wang JF, Sun XW, et al. hnRNPA2B1 drives colorectal cancer progression via the circCDYL/EIF4A3/PHF8 axis. Kaohsiung J Med Sci 2025;41:e12943. [Crossref] [PubMed]
81. Li Y, Li K, Lou X, et al. HNRNPA2B1-Mediated MicroRNA-92a Upregulation and Section Acts as a Promising Noninvasive Diagnostic Biomarker in Colorectal Cancer. Cancers (Basel) 2023;15:1367. [Crossref] [PubMed]
82. Zhu F, Yang T, Yao M, et al. HNRNPA2B1, as a m(6)A Reader, Promotes Tumorigenesis and Metastasis of Oral Squamous Cell Carcinoma. Front Oncol 2021;11:716921. [Crossref] [PubMed]
83. Liu Y, Li H, Liu F, et al. Heterogeneous nuclear ribonucleoprotein A2/B1 is a negative regulator of human breast cancer metastasis by maintaining the balance of multiple genes and pathways. EBioMedicine 2020;51:102583. [Crossref] [PubMed]
84. Binnewies M, Roberts EW, Kersten K, et al. Understanding the tumor immune microenvironment (TIME) for effective therapy. Nat Med 2018;24:541-50. [Crossref] [PubMed]
85. Duan Q, Zhang H, Zheng J, et al. Turning Cold into Hot: Firing up the Tumor Microenvironment. Trends Cancer 2020;6:605-18. [Crossref] [PubMed]
86. Wang W, Green M, Choi JE, et al. CD8(+) T cells regulate tumour ferroptosis during cancer immunotherapy. Nature 2019;569:270-4. [Crossref] [PubMed]
87. Elia I, Haigis MC. Metabolites and the tumour microenvironment: from cellular mechanisms to systemic metabolism. Nat Metab 2021;3:21-32. [Crossref] [PubMed]
88. Chen X, Song E. Turning foes to friends: targeting cancer-associated fibroblasts. Nat Rev Drug Discov 2019;18:99-115. [Crossref] [PubMed]