Low pyruvate carboxylase makes respiration-deficient tumours dependent on asparagine

Cells whose mitochondrial respiration has collapsed by mutation do not all starve in the same way. A study published in Nature Metabolism reports that whether a respiration-deficient cell can keep making proteins depends on one enzyme: mitochondrial pyruvate carboxylase, or PC. When PC levels are low, high levels of mutated mitochondrial DNA push cells into a state of asparagine auxotrophy, meaning they can no longer grow without an external supply of this amino acid. In patient-derived thyroid cancer cells carrying a high-heteroplasmy mitochondrial mutation, that dependency could be exploited with the drug L-asparaginase, which slowed tumour growth in mice.

The researchers, led by Ruobing Cui and Craig B. Thompson at Memorial Sloan Kettering Cancer Center, built their case using a panel of isogenic human retinal pigment epithelial (RPE) cell clones carrying a defined mitochondrial DNA deletion at heteroplasmy levels ranging from 3% to 99%. The team reconfirmed each clone’s heteroplasmy by droplet digital PCR and measured respiration with Seahorse oxygen-consumption analysis. Both basal and maximal oxygen consumption declined in a near-linear relationship as the mutant fraction rose.

Redox stress, however, did not track respiration smoothly. The cellular NADH/NAD+ ratio stayed unchanged in clones with up to 55% heteroplasmy and rose significantly only in clones with 85% or more. Similarly, protein components of electron transport chain complexes I through IV remained stable at low heteroplasmy and dropped only in the high-heteroplasmy clones, while the complex V protein ATP5A held constant. High-heteroplasmy cells also showed reduced citrate and malate, and glucose carbon’s contribution to those metabolites fell. The elevated NADH/NAD+ ratio correlated with reduced proliferation, and supplying pyruvate or the electron acceptor alpha-ketobutyrate restored redox balance and boosted growth: cells with 85% heteroplasmy more than doubled their growth with pyruvate, and those at 90% or above grew almost threefold.

Redox correction alone was not the whole story. When the team expressed a water-forming NADH oxidase from Lactobacillus brevis in 99% heteroplasmy cells, the NADH/NAD+ ratio fell but proliferation improved only modestly. That pointed to a second requirement. Because mitochondria also feed the synthesis of non-essential amino acids, the researchers supplemented cultures with the five non-essential amino acids missing from standard DMEM. The mix did nothing for clones at or below 85% heteroplasmy but significantly boosted proliferation at 90% and even more at 99%. Testing each amino acid individually revealed a single culprit: only asparagine matched the full supplement’s effect, and removing it from the mix eliminated the benefit.

Isotope tracing explained why. In respiratory-competent cells, glutamine carbon flows through the TCA cycle to aspartate and then to asparagine; in the 3% heteroplasmy clone, glutamine labelled about 25% of the asparagine pool. In 99% heteroplasmy cells, no glutamine-derived (m + 4) asparagine was detected over the same time frame, and glutamine’s contribution to aspartate, citrate and malate was also compromised. Notably, adding a physiologic 100 micromolar dose of aspartate did not help proliferation, even though supraphysiologic aspartate or the aspartate transporter SLC1A3 could rescue growth in other conditions. Asparagine, unlike its precursor, is used exclusively to charge tRNA for protein synthesis, and a fluorescent poly-asparagine reporter confirmed that asparagine availability directly limited translation in the high-heteroplasmy cells.

That translation block engaged a defined stress pathway. Proteomic profiling of asparagine-starved 99% heteroplasmy cells showed upregulation of ATF4 target genes, and GCN2 kinase, which responds to uncharged tRNAs, was phosphorylated only in high-heteroplasmy clones. Deleting the GCN2-encoding gene EIF2AK4 abolished ATF4 induction but made growth worse, while asparagine supplementation rescued proliferation in both control and GCN2-deleted cells. The authors conclude that above roughly 70% heteroplasmy, integrated stress response-associated translational defects emerge, and these can be reversed either by exogenous asparagine or by PC overexpression.

The link to pyruvate carboxylase came from comparing cell types. Mitochondrial PC carboxylates glucose-derived pyruvate into oxaloacetate inside mitochondria, providing an anaplerotic route that survives respiratory failure. The 143B cytB osteosarcoma line, which carries a comparable degree of respiratory compromise, expressed far more PC than the RPE cells and generated more than tenfold higher levels of glucose-derived labelled asparagine. Introducing a PC construct into 99% heteroplasmy RPE cells raised PC-derived labelling of aspartate from 40% to 72% and of asparagine from 25% to 59%, increased asparagine pools over 48 to 96 hours, suppressed ATF4 and rendered proliferation independent of exogenous asparagine. Conversely, deleting PC from 143B cytB cells cut glucose-derived aspartate and asparagine labelling, activated ATF4 and slowed growth in a way that asparagine supplementation or PC re-expression reversed.

The dependency followed heteroplasmy thresholds in a second model. Using mitochondrial base editors, the team created RPE clones carrying a complex I ND4 point mutation at heteroplasmy levels up to 81%. Clones at or below 52% heteroplasmy grew without added asparagine, while 72% and 81% clones showed respiratory deficiency, reduced aspartate and asparagine, high ATF4 and asparagine-dependent proliferation; PC expression relieved all of these defects. Genetic knockout of the nuclear-encoded ETC components UQCRC2 or COX4I1 in RPE cells produced the same asparagine dependence, indicating that impaired TCA-cycle-coupled electron transport, rather than a specific mitochondrial lesion, drives the auxotrophy in PC-low cells. Tissue context mattered too: epithelial-derived lines, including thyroid and colorectal cancer cells, expressed less PC than 143B cells and benefited from asparagine, whereas the high-PC breast cancer line MDA-MB-231 lost proliferation under respiratory inhibition only when PC itself was deleted.

The translational test centred on NCI-237, a patient-derived thyroid cancer line carrying a truncating MT-ND5 mutation at 90% heteroplasmy. These cells had low basal respiration, a NADH/NAD+ ratio four times that of respiratory-competent thyroid cells, low PC expression and high ATF4 that asparagine suppressed. PC expression in NCI-237 cells raised glucose- and pyruvate-derived aspartate labelling 3.4-fold and asparagine labelling 4.3-fold, and together with glutamine-derived synthesis accounted for up to 89% of the free aspartate pool and 80% of the free asparagine pool, while significantly promoting growth. In human plasma-like medium, L-asparaginase significantly reduced NCI-237 proliferation from day two onward, depleting asparagine without measurably lowering glutamine. The dependence was even sharper at physiologic 0.1 millimolar pyruvate, where asparagine raised proliferation more than fourfold. In immunodeficient NSG mice bearing NCI-237 xenografts, daily treatment with 80 international units of L-asparaginase for four weeks significantly reduced tumour volumes and endpoint weights compared with vehicle controls.

The authors suggest that tumours with high-heteroplasmy mtDNA mutations, which are enriched in thyroid, renal and colorectal cancers, may share this vulnerability when they arise in lineages with low PC expression. L-asparaginase is already a curative component of acute lymphoblastic leukaemia therapy, but other cancer types had not previously been shown to be asparagine auxotrophic. The findings are preclinical, resting on cell lines and xenografts, and the researchers note that PC expression varies widely across tissues and is not induced by the stress of respiratory compromise. Whether asparaginase can be optimized against mtDNA-mutant, PC-low tumours in patients remains a question for further study.

Subject of Research: Metabolic dependency on asparagine in respiration-deficient cells with low pyruvate carboxylase expression

Article Title: Respiration-deficient cells require pyruvate carboxylase to suppress asparagine auxotrophy

Article References: Cui, R., Ryu, K. W., Fu, Y., Bakouny, Z., Li, D., Kavlashvili, T., Sfeir, A., & Thompson, C. B. (2026). Respiration-deficient cells require pyruvate carboxylase to suppress asparagine auxotrophy. Nature Metabolism. https://doi.org/10.1038/s42255-026-01651-4

Image Credits: AI Generated

DOI: 10.1038/s42255-026-01651-4

Keywords: mitochondrial DNA, heteroplasmy, pyruvate carboxylase, asparagine auxotrophy, integrated stress response, ATF4, GCN2, L-asparaginase, thyroid cancer, oxidative phosphorylation, aspartate metabolism, xenograft