Two Molecular Machines, One Cancer Engine: How Deubiquitinases and m6A RNA Tags Talk to Each Other

Two of the most powerful control systems inside a cell have long been studied in separate worlds. One system decorates messenger RNA molecules with a chemical tag called N6-methyladenosine, or m6A, deciding how long RNAs survive, how efficiently they are translated into proteins, and which of them ever leave the nucleus. The other system, built around the small protein ubiquitin, stamps target proteins for destruction or shields them from it, tuning the abundance of nearly every regulatory molecule a cell owns. A review published in the journal Molecular Cancer by Xuan Zhang, Ning Lou, Yudan Wang and colleagues at Peking Union Medical College Hospital now argues that these two layers of regulation are not parallel tracks at all, but a tightly interwoven network whose crosstalk helps drive cancer progression, metastasis, metabolic reprogramming and drug resistance.

The m6A mark is the most abundant internal modification found in the messenger RNAs of higher organisms. It is installed by a consortium of enzymes that researchers colloquially call writers, a group that includes the catalytic methyltransferase-like proteins METTL3, METTL14 and METTL16, alongside METTL5, METTL4 and accessory components such as WTAP, VIRMA, RBM15, RBM15B, ZC3H13 and PRMT1. Once deposited, the mark does not simply sit idle. Reader proteins, including members of the YTH domain family such as YTHDF1, YTHDF2, YTHDF3, YTHDC1 and YTHDC2, as well as IGF2BP1, IGF2BP2, IGF2BP3, HNRNPC, HNRNPG, HNRNPA2B1, PRRC2A and FMRP, interpret the tag and route the tagged RNA toward translation, decay, splicing or export. A third class of enzymes, the erasers FTO and the AlkB homologues ALKBH5 and ALKBH3, can strip the methyl group away again, making the entire system dynamic and reversible. That reversibility is precisely what makes m6A such a potent lever in cancer, where the balance between writing, erasing and reading can tip a cell toward uncontrolled growth.

On the protein side of the equation sits the ubiquitin system. Ubiquitin is a small polypeptide that, once attached to a lysine residue on a target protein, acts as a molecular verdict: chains of a particular architecture command the proteasome to destroy the tagged protein, while other linkages reshape signalling, localization or interactions instead. The counterweight to this system is a family of roughly one hundred enzymes known as deubiquitinases, or DUBs, many of them belonging to the ubiquitin-specific peptidase, or USP, family. DUBs cleave ubiquitin tags back off their substrates, rescuing proteins from degradation and prolonging their functional lifespan. Because tumour cells depend on the stabilized activity of oncogenic proteins, DUBs have attracted intense interest as drug targets, and several inhibitors are already being evaluated in clinical development.

The central insight highlighted in the new review, published as an open-access article in Molecular Cancer, is that these two regulatory universes constantly feed back on one another. The first direction of crosstalk runs from proteins to RNA modifications. Specific DUBs can remove ubiquitin marks from the core components of the m6A machinery, including the writers that install the tag, the erasers that remove it and the readers that interpret it. By deubiquitinating and thereby stabilizing these enzymes, a DUB can effectively raise the activity of the entire m6A pathway without the cell needing to make a single new molecule of the machinery. The consequence is a shift in the methylation landscape of thousands of RNA transcripts at once, with downstream changes in gene expression that can favour proliferation, invasion or survival under stress.

The reverse direction of the conversation is equally consequential. m6A modification can itself regulate the expression of certain deubiquitinases, either by influencing the stability of the messenger RNAs that encode them or by controlling the translation of those RNAs into protein, as well as by managing associated regulatory factors upstream. In other words, the RNA modification layer can dial the abundance of DUBs up or down, and the DUB layer can in turn remodel the stability of the very enzymes that write and erase the RNA marks. The result is a reciprocal regulatory loop, a handshake in which each system continuously calibrates the other. Such coupled feedback loops are a classic motif in biology, and in cancer they can become locked into pathological configurations that entrench malignant behaviour.

The review, whose authors also include Haolong Li, Chuyan Nie, Yongzhe Li and senior author Xiaohong Han, systematically maps how this DUB–m6A axis impinges on the hallmark processes of tumour biology. Cell proliferation is an obvious beneficiary, since stabilized m6A enzymes and stabilized DUBs together can amplify the expression of growth-promoting transcripts while suppressing signals of cell-cycle arrest. Metastasis is another, with the epithelial–mesenchymal transition, the plasticity program by which epithelial tumour cells acquire migratory and invasive properties, standing out as a process particularly sensitive to epitranscriptomic tuning. Metabolic reprogramming, the Warburg-like shift in nutrient use that sustains rapidly dividing cells, and therapy resistance, the grinding ability of tumours to shrug off chemotherapy, targeted agents and radiation, both emerge from the analysis as downstream consequences of this crosstalk.

What lends the analysis particular technical weight is its level of resolution. Rather than treating the m6A machinery as a monolith, the authors distinguish among its individual components, and the abbreviations catalogued in the paper span the full breadth of the field: the writer complex built around METTL3 and METTL14, the demethylases FTO, ALKBH5 and ALKBH3, and the reader families YTHDF, YTHDC and IGF2BP, along with nuclear RNA-binding proteins such as HNRNPC and HNRNPA2B1 that respond to m6A indirectly by changing how RNA folds. This granularity matters because different cancers appear to co-opt different nodes. A deubiquitinase that stabilizes METTL3 may push one tumour type toward aggressive methylation of pro-oncogenic transcripts, while one that stabilizes FTO or ALKBH5 may have the opposite effect in another context, erasing marks that would otherwise recruit decay machinery to tumour-suppressive RNAs. The sign of the effect depends on which substrates dominate in a given tissue, which is why the review frames the axis as a network to be mapped rather than a switch to be flipped.

The therapeutic implications are the most forward-looking part of the analysis. Inhibitors of m6A writers and erasers are already in preclinical and early clinical development, and DUB inhibitors are advancing along a parallel track. The crosstalk described in the review suggests that monotherapy against either layer may be undermined by compensatory feedback from the other: block a DUB and the m6A machinery may re-equilibrate, or suppress a methyltransferase and DUB-driven stabilization of erasers may fill the void. Conversely, the coupling creates opportunities. Combination strategies that target both sides of the axis simultaneously, or that exploit a tumour’s dependency on a specific DUB–m6A wiring pattern that normal cells do not share, could deliver selective pressure that malignant cells find much harder to escape. The review explicitly positions the DUB–m6A axis as a critical regulatory hub whose deciphering unveils promising targets for overcoming drug resistance and developing innovative anti-cancer strategies.

There are, of course, substantial obstacles between this systems-level map and the clinic. The ubiquitin code is baroque, with different chain linkages carrying different meanings, and many DUBs are promiscuous, acting on dozens of substrates whose identities remain incompletely catalogued. The m6A landscape, meanwhile, is dynamic across the cell cycle, across tissues and across stages of tumour evolution, and measuring it at single-transcript resolution in patient samples remains technically demanding. Disentangling cause from consequence in a bidirectional loop requires perturbation experiments rather than correlation, and the review’s synthesis makes clear how much systematic mapping work remains. Yet the conceptual payoff is considerable. If cancer is, at bottom, a disease of misregulated information, then the discovery that its RNA-level and protein-level information systems are in constant negotiation offers researchers a genuinely new coordinate on the map, one where the two dancers, as the paper’s title suggests, truly do take two to tango.

Subject of Research: The bidirectional crosstalk between deubiquitinases and N6-methyladenosine RNA modification in cancer biology

Article Title: It takes two to tango: the intersection of deubiquitinases and m6A modification in cancer

Article References: It takes two to tango: the intersection of deubiquitinases and m6A modification in cancer. (n.d.). https://doi.org/10.1186/s12943-026-02808-1

Image Credits: AI Generated

DOI: 10.1186/s12943-026-02808-1

Keywords: deubiquitinases, m6A modification, epitranscriptomics, cancer, post-translational modification, RNA modification, METTL3, ubiquitin, therapy resistance, metastasis, tumorigenesis, Molecular Cancer