The centerpiece of the new work is the pyruvate dehydrogenase complex, or PDHc, a multi-enzyme assembly that sits at one of the most consequential crossroads in cellular metabolism. Glycolysis in the cytoplasm converts glucose into pyruvate, and the fate of that pyruvate determines which metabolic program the cell will run. When PDHc is active, it catalyzes the oxidative decarboxylation of pyruvate into acetyl-coenzyme A, which then feeds the tricarboxylic acid cycle, or TCA cycle, inside mitochondria, driving the production of NADH and FADH2 that power oxidative phosphorylation. When PDHc is inhibited, pyruvate is instead reduced to lactate, the TCA cycle slows, and the cell becomes dependent on glucose fermentation and on the diversion of glycolytic intermediates into biosynthetic pathways that support rapid proliferation. PDHc activity is therefore a metabolic master switch, and its regulation is a matter of life and death for a growing tumor.

The canonical off switch for PDHc is phosphorylation. Pyruvate dehydrogenase kinase 1, or PDK1, attaches phosphate groups to the E1 alpha subunit of the complex, PDHE1α, and this inhibitory mark suppresses the enzyme’s decarboxylase activity. Pharmaceutical efforts to reverse this blockade have centered on dichloroacetate, or DCA, a small molecule that inhibits PDK enzymes and thereby restores PDHc flux. What has remained less clear is how tumor cells recruit and amplify PDK1 activity in the first place, and whether noncoding RNAs, a class of regulatory molecules long suspected of shaping cancer metabolism, participate directly in this process. The new study provides a striking answer to both questions.

The researchers began with an observation that challenges the traditional division of labor between proteins and RNAs: PDHE1α, an enzyme long considered a purely metabolic protein, can bind RNA. Among the transcripts that associate with it is SNHG15, a long noncoding RNA that has previously been implicated in the progression of several cancers but whose metabolic functions were poorly defined. This RNA-protein interaction suggested that a noncoding transcript might physically insert itself into the machinery that governs mitochondrial fuel oxidation, a possibility the team pursued through a series of mechanistic experiments in hepatocellular carcinoma cells.

The mechanism that emerged is a scaffold model of metabolic control. According to the authors’ analysis, SNHG15 facilitates the interaction between phosphoglycerate kinase 1, PGK1, a glycolytic enzyme, and PDK1, and it simultaneously promotes the association between PDK1 and its substrate, PDHE1α. In effect, the RNA acts as a molecular matchmaker, bringing the kinase into proximity both with its upstream activator and with the enzyme it phosphorylates. The consequence is increased phosphorylation of PDK1 itself and of PDHE1α, which suppresses the catalytic activity of the pyruvate dehydrogenase complex. With PDHc throttled back, pyruvate oxidation into the TCA cycle declines, glycolysis accelerates, and lactate production rises, the classic biochemical signature of the Warburg phenotype.

This reprogramming is not merely a correlative curiosity; the study connects it directly to malignant behavior. In cell models, SNHG15-driven suppression of the TCA cycle promoted the glycolytic flux that fuels hepatocellular carcinoma progression. The team then extended the findings into xenograft models, in which tumor cells are implanted into experimental animals. Tumors with elevated SNHG15 grew more aggressively, and biochemical analysis of those tumors revealed increased lactate production alongside heightened phosphorylation of both PDK1 and PDHE1α, confirming that the RNA-protein phosphorylation axis operates in living tumors and not only in culture dishes.

Perhaps the most therapeutically compelling result came when the researchers administered dichloroacetate to the xenograft-bearing animals. DCA, which inhibits pyruvate dehydrogenase kinases and thereby releases the brake on PDHc, significantly reversed the oncogenic effects of SNHG15. Tumor growth was blunted, and the metabolic markers of the Warburg state receded. This pharmacological rescue is important because it demonstrates causality: if blocking PDK1 activity negates the tumor-promoting effects of the RNA, then the RNA’s oncogenic power flows substantially through that kinase, and the PGK1-PDK1-PDHE1α axis becomes a defined, druggable vulnerability rather than a loose correlation.

The study also broadens the conceptual repertoire of long noncoding RNA biology. LncRNAs have typically been cast as regulators of gene expression, modulating transcription, chromatin state, or messenger RNA stability. Here, SNHG15 acts instead at the level of enzyme activity, functioning post-translationally by organizing protein-protein interactions. The finding that PDHE1α itself is an RNA-binding protein adds to a growing recognition that core metabolic enzymes moonlight as participants in RNA-based regulatory networks, blurring the boundary between the genome’s informational layer and its metabolic machinery. In this view, noncoding RNAs can serve as intracellular scaffolds that rewire enzyme kinetics on demand, offering tumors a rapid and reversible means of adapting their metabolism to the demands of proliferation.

For hepatocellular carcinoma specifically, the clinical implications are worth weighing carefully. Liver cancer is frequently diagnosed at advanced stages and remains among the leading causes of cancer mortality worldwide, and metabolic vulnerabilities have long been proposed as therapeutic entry points. The SNHG15-PDK1-PDHE1α pathway offers several potential intervention surfaces: targeting the RNA itself with antisense or degrading strategies, disrupting its interaction with PDHE1α or PGK1, or inhibiting PDK1 pharmacologically, an approach already tested clinically with DCA in various contexts. The authors position the PGK1-PDK1-PDHE1α axis as a candidate target for clinical intervention, and the xenograft data with DCA provide an early proof of principle, though the gap between animal models and human therapy remains substantial and will require validation in patient-derived systems and ultimately clinical trials.

As with any study, important questions remain open. The precise structural basis of the SNHG15-PDHE1α interaction, the determinants of SNHG15 overexpression in liver tumors, and the relationship between this metabolic scaffold function and any other roles the RNA may play in HCC cells all await further definition. The work was supported by the National Natural Science Foundation of China, the Natural Science Foundation of Shandong Province and related institutional funds, and the authors declare no competing interests. What the study establishes, however, is a clear mechanistic narrative: a long noncoding RNA can hijack a phosphorylation cascade at the glycolysis-mitochondria junction, silence the TCA cycle, and push liver cancer cells into the fermentative metabolism on which their growth depends, and that same cascade can be chemically disarmed. In doing so, it adds a new layer to the metabolic playbook of hepatocellular carcinoma and offers researchers a concrete molecular handle on one of cancer’s oldest and most stubborn adaptations.

Subject of Research: Metabolic reprogramming of hepatocellular carcinoma by the lncRNA SNHG15 through PDK1 and PDHE1α phosphorylation

Article Title: SNHG15 inhibits the tricarboxylic acid cycle and promotes HCC progression by facilitating the phosphorylation of PDK1 and PDHE1α

Article References: Chen, H., Wang, Y., Chong, W., Wang, X., Wang, L., Zhang, J., Zhao, X., & Wang, C. (2026). SNHG15 inhibits the tricarboxylic acid cycle and promotes HCC progression by facilitating the phosphorylation of PDK1 and PDHE1α. Cell Death Discovery. https://doi.org/10.1038/s41420-026-03380-9

Image Credits: AI Generated

DOI: 10.1038/s41420-026-03380-9

Keywords: hepatocellular carcinoma, SNHG15, long noncoding RNA, pyruvate dehydrogenase complex, PDK1, PDHE1α, PGK1, TCA cycle, glycolysis, Warburg effect, dichloroacetate, cancer metabolism