The RBM family is defined by a shared structural module: the RNA recognition motif, a compact domain of roughly ninety amino acids folded into a four-stranded beta sheet packed against two alpha helices. This domain binds short stretches of RNA, typically through aromatic residues that stack against the bases, and most RBM proteins carry multiple copies of it, often connected by flexible linkers that allow tandem domains to recognize longer or composite RNA elements. Many members also harbor auxiliary domains, such as zinc fingers, OCRE motifs, or arginine-rich regions, that extend their binding repertoire or recruit interaction partners. Structural studies, including nuclear magnetic resonance analyses of RBM10’s OCRE-containing domain and crystallographic work on RBM5’s RNA-binding surface, have revealed how these proteins achieve specificity despite the modest selectivity of any single RRM domain. The result is a versatile toolkit: RBM proteins can clasp pre-messenger RNAs, mature transcripts, long noncoding RNAs, and even ribosomal RNA, positioning them to regulate nearly every step of the RNA lifecycle.

Among the best-characterized family members are RBM5 and RBM10, which the review describes as tumor suppressors. Both proteins function as components of the U2 small nuclear ribonucleoprotein particle, engaged with intron branch sites on chromatin, where they steer the splicing machinery toward particular exon choices. Their tumor-suppressive influence is vividly illustrated by their control of NUMB alternative splicing: RBM5, RBM6, and RBM10 differentially regulate which NUMB isoform is produced, and the resulting isoform balance determines whether cancer cells proliferate or restrain themselves. In lung adenocarcinoma, RBM10 loss promotes proliferation and metastasis through a feedback loop involving microRNA-224-5p and p53, and the protein also restrains the RAP1/AKT/CREB signaling pathway. RBM10 mutations, which are frequent in lung adenocarcinoma, have been proposed as negative prognostic and predictive markers, and a specific C761Y mutation in cholangiocarcinoma generates oncogenic ASPM isoforms that activate beta-catenin signaling. Recent work has even exploited this vulnerability therapeutically: RBM10-deficient lung adenocarcinoma cells show heightened sensitivity to agents that induce DNA replication stress, opening a synthetic-lethal strategy against tumors that have lost this suppressor.

RBM5 itself displays a strikingly context-dependent personality. In colorectal cancer, it suppresses proliferation, metastasis, and glycolysis by stabilizing the messenger RNA of the phosphatase and tensin homolog, PTEN, a master brake on growth signaling. It also recruits the factor MGC32805 in a sandwich mode to induce a novel FAS neoantigen, switching the death receptor’s properties in a way that may sensitize tumors to immune attack. Yet in acute myeloid leukemia, RBM5 plays the villain, activating the oncogenic protein HOXA9 and supporting leukemic growth. This duality, the same protein acting as friend in one tissue and foe in another, recurs throughout the family and underscores a central theme of the review: RBM proteins cannot be labeled simply as oncogenes or suppressors. Their function depends on the RNA landscape of the cell, the signaling pathways in play, and the selective pressures of the tumor microenvironment.

A second axis of RBM-mediated cancer regulation runs through m6A, the most abundant internal chemical modification of messenger RNA. RBM15 and its paralog RBM15B are core components of the m6A writer complex, the machinery that deposits methyl groups onto adenosines across the transcriptome. By recognizing the histone mark H3K79me2, RBM15B guides selective m6A modification of messenger RNAs and enhances oncoprotein translation in MLL-rearranged leukemia, a finding reported in the EMBO Journal in 2026 that links chromatin state directly to RNA modification patterns. In triple-negative breast cancer, RBM15 drives cell growth by stimulating serine and glycine metabolism, while in cervical cancer it promotes epithelial-mesenchymal transition through m6A modification of the EZH2 transcript and supports stemness by stabilizing the long noncoding RNA HEIH. RBM15 also enhances paclitaxel resistance in triple-negative breast cancer by methylating TNFSF9 and polarizing tumor-associated macrophages toward an immunosuppressive M2 state, connecting RNA modification to both drug resistance and the immune microenvironment.

Other family members exert their influence through equally inventive mechanisms. RBM3, a cold-shock-inducible protein, suppresses stemness remodeling of prostate cancer in the bone microenvironment by modulating m6A on CTNNB1 messenger RNA, halts the cell cycle of cutaneous squamous cell carcinoma through the PI3K/AKT pathway, and has repeatedly been associated with favorable prognosis and cisplatin sensitivity in epithelial ovarian cancer and muscle-invasive bladder cancer. RBM38, meanwhile, is a key lieutenant of the p53 tumor-suppressor network: it stabilizes the p53-MDM2 loop in hepatocellular carcinoma, and its genetic ablation promotes lymphomagenesis in the context of mutant p53 by downregulating PTEN. RBM47 restrains renal cell carcinoma by interacting with the long noncoding RNA HOXB-AS1 and acts as an anti-oncogene in colorectal cancer, yet in pancreatic cancer it promotes proliferation and immune evasion by upregulating PDIA6, and in nasopharyngeal carcinoma it drives tumorigenesis through multiple pathways. RBM45 reprograms lipid metabolism in hepatocellular carcinoma via Rictor and the acyl-CoA synthetases ACSL1 and ACSL4, and stabilizes the glutamine transporter ASCT2 through phosphorylation to fuel tumor progression.

The therapeutic implications of this biology are beginning to crystallize. The most advanced example involves RBM39, a related RNA-binding protein that can be degraded by molecular-glue compounds such as indisulam and E7820, which recruit the DCAF15 ubiquitin ligase. A 2024 study in npj Precision Oncology showed that an RBM39-degrader induces synthetic lethality in cancer cells deficient in homologous recombination repair, suggesting a rational combination with PARP inhibitors or platinum chemotherapy. For the RBM family proper, a first-in-class small-molecule inhibitor of the eIF4E-RBM38 complex has been developed; by disrupting this interaction, the compound enhances translation of wild-type p53 and suppresses tumor growth, validating protein-protein interfaces involving RBM proteins as druggable targets. A synthetic peptide called Pep8 achieves a similar effect by camouflaging the same interface. These proof-of-concept successes matter because the RNA recognition motif has long been considered difficult to drug, its shallow RNA-binding grooves resisting conventional small-molecule approaches.

The review also highlights how RBM proteins intersect with clinical decision-making beyond drug targets. RBM3 expression levels correlate with chemosensitivity in ovarian and bladder cancers, potentially guiding platinum-based treatment choices. RBM10 mutation status may stratify patients with non-small-cell lung cancer, flagging tumors with distinct vulnerabilities and therapy responses. RBM11 serves as a prognostic biomarker in ovarian cancer, RBM19 promotes prostate cancer progression under docetaxel treatment through the SNHG21/PIM1 axis, and RBMX contributes to sorafenib resistance in hepatocellular carcinoma by stabilizing the long noncoding RNA BLACAT1. RBM38, conversely, reverses sorafenib resistance by promoting the tumor-suppressive lncRNA GAS5. As liquid biopsy platforms for circulating tumor cells mature, RNA-binding protein signatures detected in blood may eventually complement tissue-based diagnostics, though the review stops short of claiming clinical readiness for such applications.

What emerges from this synthesis is a portrait of a protein family whose complexity mirrors that of cancer itself. The same RBM protein can suppress one tumor while feeding another, can be drugged in one context and mimicked in another, and can act through splicing, m6A methylation, messenger RNA stability, translation, or long noncoding RNA scaffolding depending on circumstance. The Qingdao University team argues that disentangling these context-dependent roles, through systematic mapping of RNA targets, structural dissection of binding domains, and careful correlation with patient outcomes, is the essential next step. If that effort succeeds, the RBM family could yield a generation of biomarkers that predict treatment response and therapeutic strategies that exploit the very RNA-processing dependencies cancers have built into themselves. For now, the review stands as both a map of what is known and a catalog of what remains to be tested, from the branch sites of chromatin-bound introns to the methylated transcripts that decide a tumor’s fate.

Subject of Research: The roles of the RBM RNA-binding protein family in cancer occurrence, progression, and therapy

Article Title: RBM protein family: Focusing on the occurrence, progression and treatment of cancer

Article References: Zhang, H., Zheng, W., Li, L., Geng, C., Wang, Y., Zhang, J., Sun, Y., Zhang, X., & Li, B. (2026). RBM protein family: Focusing on the occurrence, progression and treatment of cancer. Medical Oncology, 43(11), Article 322. https://doi.org/10.1007/s12032-026-03439-8

Image Credits: AI Generated

DOI: 10.1007/s12032-026-03439-8

Keywords: RBM proteins, RNA-binding proteins, alternative splicing, m6A methylation, tumor suppressor, oncogene, RBM10, RBM15, RBM38, therapeutic target, biomarker, cancer