Bidirectional crosstalk between RNA modification networks and metabolic reprogramming in hepatocellular carcinoma

Metabolic reprogramming is a defining feature of hepatocellular carcinoma (HCC), enabling malignant cells to sustain proliferation, adapt to nutrient and oxygen limitation, evade antitumor immunity, and survive therapeutic stress. Emerging evidence indicates that RNA modification networks provide an important interface between metabolic states and post-transcriptional gene regulation. RNA modifications and their associated writers, erasers, and readers can reshape glucose, lipid, and amino acid metabolism by regulating RNA stability, translation, processing, and RNA–protein interactions. Conversely, nutrient availability, glycolytic activity, and metabolites such as lactate, acetyl-CoA, and S-adenosylmethionine can influence RNA-modifying enzymes, modification abundance, reader activity, and RNA fate, thereby establishing reciprocal metabolic–RNA regulatory circuits. Importantly, however, the functions of RNA-modification-associated proteins are not invariably modification-dependent; several regulators also exert catalytic-independent or non-canonical effects through protein interactions, chromatin regulation, or metabolic signaling. Distinguishing these mechanisms is therefore essential for defining genuine epitranscriptomic regulation. These reciprocal networks extend beyond tumor-cell-autonomous metabolism to influence redox homeostasis, ferroptosis and cuproptosis, cancer stemness, tumor immunity, and therapeutic resistance. Long non-coding RNAs and circular RNAs further intersect with RNA modification machinery as regulators, scaffolds, and, in some contexts, modification substrates, adding another layer of metabolic plasticity. Here, we critically discuss how major RNA modifications, with particular emphasis on m⁶A and the emerging but less extensively characterized roles of m⁵C, m⁷G, ac⁴C, and other modifications, interface with glucose, lipid, and amino acid metabolism in HCC. We further examine how metabolic remodeling feeds back on RNA-modification machinery and modification-dependent RNA regulation, while distinguishing these effects from modification-independent functions of the same regulators. Non-coding RNA crosstalk, tumor immunometabolism, regulated cell death, and therapeutic adaptation are also integrated into this framework. Finally, we evaluate therapeutic strategies targeting RNA-modification machinery or downstream metabolic dependencies in combination with targeted therapy, chemotherapy, ferroptosis or cuproptosis induction, and immune checkpoint blockade. We highlight persistent challenges, including uneven mechanistic evidence across RNA modification classes, limited site-specific causal validation, insufficient metabolic-flux measurements, catalytic-independent functions of modification-associated proteins, and unresolved cell-type and spatial heterogeneity. A more rigorous distinction between RNA modification-dependent and modification-independent mechanisms will be critical for identifying context-specific vulnerabilities and translating metabolic–RNA regulatory networks into biomarker-guided precision therapy for HCC.