Starvation enzyme SIRT2 strips a mitochondrial gatekeeper to trigger cell death

When cells are starved of nutrients for long enough, they do not simply fade away quietly. Behind the scenes, a precisely choreographed molecular program pushes them toward death, and researchers have now identified a critical switch in that program. A team at Army Medical University in Chongqing, China, reports in Cell Death Discovery that the enzyme SIRT2, best known as a nutrient-sensing deacetylase, can relocate to mitochondria during prolonged starvation and chemically modify a key component of the mitochondrial permeability transition pore, setting off a cascade that ultimately dismantles the cell. The finding adds an unexpected layer of control to one of the most debated structures in cell biology and suggests that a single acetyl group on a single amino acid can help decide whether a cell survives metabolic crisis or succumbs to it.

The protein at the center of the study is adenine nucleotide translocase 3, or ANT3, a transporter embedded in the inner mitochondrial membrane that normally shuttles ATP out of the mitochondrial matrix in exchange for ADP from the cytosol. Beyond its day job as an exchange factor, ANT3 has long been implicated as a core constituent of the mitochondrial permeability transition pore, or mPTP, a high-conductance channel whose opening dissipates the proton gradient that powers ATP synthesis, swells the mitochondria, and can launch cell death. Disruption of ANT3 protein homeostasis, or conformational changes in the protein itself, are sufficient to trigger this catastrophic opening. What has remained murky is how the cell regulates ANT3’s abundance and shape in the first place, and that is precisely the question the Chongqing group set out to answer.

The researchers’ starting point was prolonged starvation, a physiological stress that cells in metabolically demanding tissues experience during nutrient deprivation. They found that extended starvation drives SIRT2, a NAD+-dependent deacetylase that typically resides in the cytoplasm and nucleus, to translocate into the mitochondria. This relocation matters because it places SIRT2 in the same compartment as ANT3, allowing the two proteins to bind directly. Once docked onto ANT3, SIRT2 performs its signature reaction: it removes an acetyl group from lysine 96, a specific residue on the ANT3 protein. Acetylation at lysine residues is a widespread post-translational modification that can alter a protein’s charge, structure, interactions, and stability, and the new work shows that the acetyl state of this one lysine carries enormous consequences for ANT3 behavior.

The consequences unfold along two parallel tracks. First, deacetylation of ANT3 at K96 promotes conformational changes in the protein, shifting it into a state that favors the formation of a complex with VDAC1, the voltage-dependent anion channel that sits in the outer mitochondrial membrane. The ANT3-VDAC1 pairing spans the contact sites between the inner and outer membranes and is considered a plausible architecture for the permeability transition pore, so enhancing this complex effectively assembles more of the death apparatus at the mitochondrial boundary. Second, and perhaps more surprisingly, deacetylated ANT3 binds more strongly to USP15, a deubiquitinating enzyme that removes ubiquitin tags from its substrates. Strengthened binding to USP15 enhances ANT3 deubiquitination, which stabilizes the ANT3 protein and raises its steady-state levels inside the mitochondria.

The logic of this dual mechanism is elegant and somewhat ruthless. By simultaneously reshaping ANT3 and protecting it from degradation, SIRT2 ensures that more of the transporter is present and more of it is locked into the pore-forming conformation with VDAC1. The result is an accelerated assembly of the permeability transition pore complex and, ultimately, an accelerated onset of cell death. In other words, starvation does not merely weaken the cell passively; it actively recruits a deacetylase to the mitochondria, which then stabilizes and reconfigures the very channel that will destroy the organelle. The study’s authors describe this as a rarely reported molecular mechanism in which SIRT2-mediated acetylation-dependent regulation governs programmed cell death.

Technically, the work illustrates how a single deacetylation event can be amplified into a systems-level outcome through feed-forward loops. The K96 deacetylation does not act alone; it recruits USP15, whose deubiquitinating activity further increases ANT3 abundance, which in turn increases the pool of ANT3 available to pair with VDAC1. This kind of positive feedback is characteristic of switch-like biological decisions, where a graded stress signal, in this case the duration of starvation, must be converted into a binary outcome: live or die. SIRT2’s dependence on NAD+ also ties the pathway directly to cellular metabolism, since NAD+ availability fluctuates with nutrient status and redox state. A starving cell, with its altered NAD+/NADH balance, thus provides both the signal and the enzymatic cofactor for the death-promoting modification.

The findings carry weight for a field that has struggled for decades with the molecular identity of the permeability transition pore. Genetic ablation of individual candidate components, including the ANT isoforms, has produced confusing and sometimes contradictory results, and the pore’s exact composition remains contested. By focusing on post-translational regulation rather than mere presence or absence of the protein, the new study offers a way to reconcile some of that ambiguity: ANT3 may be innocuous in one acetylation state and lethal in another. Conformational state, protein stability, and partner binding, all governed here by K96 acetylation, could explain why simply deleting a pore component does not always abolish permeability transition, since compensatory mechanisms and residual complexes may fill the gap.

There are also broader implications for human disease. Excessive mPTP opening is a hallmark of ischemia-reperfusion injury, where restored blood flow after a heart attack or stroke paradoxically kills cells that survived the initial insult, and it contributes to neurodegeneration, muscle wasting, and liver injury. Conversely, blocking pore opening can allow damaged cells to evade death, a trait exploited by some cancers. A pathway in which SIRT2, USP15, ANT3, and VDAC1 cooperate provides multiple potential intervention points. Small molecules that modulate SIRT2 activity are already under investigation in oncology and neurology, and deubiquitinase inhibitors are an active area of drug development. The Chongqing study does not test therapeutic compounds, but it maps the wiring that such compounds would need to respect.

The research, led by corresponding authors Liangbo Sun, Jiqin Lian, Mingzhen Yang, and Haojun Xiong, with Dong Liu, Meng He, and Jiahong Liu contributing equally as first authors, was supported by the National Natural Science Foundation of China and the Chongqing Natural Science Foundation. Animal experiments were conducted under the guidelines of the Army Medical University Laboratory Animal Center with ethics committee approval, and the authors declare no competing interests. The article was published open access on 29 September 2026, with raw western blotting data and supplementary materials made available alongside the paper, a transparency practice that allows independent researchers to scrutinize the key biochemical evidence.

What makes the study resonate beyond its immediate findings is the picture it paints of cellular death as an actively regulated, enzymatically driven process rather than a passive collapse. A starving cell does not merely run out of fuel; it deploys SIRT2 to the mitochondria, strips an acetyl group from ANT3 at lysine 96, stabilizes the protein through USP15-mediated deubiquitination, and assembles the ANT3-VDAC1 complex that breaches the mitochondrial barrier. Each step is reversible in principle, which means each step is a potential checkpoint at which the cell, or a clinician, could intervene. As researchers continue to dissect the permeability transition pore and the acetylation networks that tune it, the K96 switch on ANT3 stands out as a vivid example of how the smallest chemical changes can command the largest cellular decisions.

Subject of Research: SIRT2-mediated deacetylation of the mitochondrial protein ANT3 and its role in regulating programmed cell death during starvation

Article Title: SIRT2-mediated deacetylation of ANT3 at Lys96 promotes cell death

Article References: Liu, D., He, M., Liu, J., Shan, M., Ma, X., lu, L., Bi, X., Xiong, H., Yang, M., Lian, J., & Sun, L. (2026). SIRT2-mediated deacetylation of ANT3 at Lys96 promotes cell death. Cell Death Discovery. https://doi.org/10.1038/s41420-026-03330-5

Image Credits: AI Generated

DOI: 10.1038/s41420-026-03330-5

Keywords: SIRT2, ANT3, VDAC1, USP15, mitochondrial permeability transition pore, deacetylation, lysine 96, starvation, programmed cell death, deubiquitination, mitochondria, cell death discovery