When plants run out of food, they do something remarkable: they begin to eat themselves, in a controlled and precisely regulated way. This process, known as autophagy, allows a starving cell to recycle its own components, breaking down damaged proteins, worn-out organelles, and surplus molecules inside the vacuole and converting them into raw materials that keep the organism alive. Autophagy is an evolutionarily conserved survival strategy found in yeast, animals, and plants alike, and in crops it underpins tolerance to drought, flooding, darkness, and nutrient-poor soils. Yet for all that is known about how autophagy executes its recycling duties, one fundamental question has remained stubbornly open: how does the cell decide, at the molecular level, that it is time to switch the machinery on? A new study in Arabidopsis thaliana now provides a compelling answer, identifying a pore-forming membrane protein called MACP2 as a critical bridge between extracellular stress signals and the autophagy engine.
The research, led by Ying Zhou, Xue Zhang, Tiancong Qi, and Zi-Han Wang working with senior authors Lu-Jun Yu and Qin-Fang Chen at Sun Yat-Sen University and their collaborators, began with a deliberately simple strategy. The team used the central autophagy protein ATG8e as bait in a yeast two-hybrid screen, fishing an Arabidopsis cDNA library for any protein capable of physically binding to it. ATG8 sits at the heart of autophagosome formation, the process by which a double-membrane vesicle engulfs cellular cargo before delivering it to the vacuole for destruction. Proteins that dock onto ATG8 typically do so through a short peptide signature known as the ATG8-interacting motif, or AIM, defined by the consensus sequence of an aromatic residue followed by two variable positions and a bulky hydrophobic one. Among the fifteen candidate interactors that emerged from the screen, one stood out: MEMBRANE ATTACK COMPLEX/PERFORIN-LIKE 2, or MACP2, a member of a protein family better known for its role in immunity and programmed cell death.
The MACPF family is famous in animal biology. Members of this family, including the complement membrane attack complex and perforin, oligomerize into ring-shaped pores that punch holes in target membranes, a mechanism central to immune defense and development. Plants encode a small set of MACPF proteins, and in Arabidopsis four of them carry the conserved pore-forming domain: CAD1, NSL1, MACP1, and MACP2. Previous work had shown that MACP2 regulates programmed cell death and influences resistance to bacterial pathogens and susceptibility to necrotrophic fungi, and its transcript levels rise in response to a range of biotic and abiotic stresses. But its involvement in autophagy had never been suspected. The new study confirmed the physical interaction between MACP2 and ATG8e using two independent methods: co-immunoprecipitation in Arabidopsis mesophyll protoplasts, where tagged ATG8e pulled down tagged MACP2 from cell extracts, and bimolecular fluorescence complementation, in which the two proteins reconstituted a fluorescent signal at the plasma membrane only when co-expressed. Crucially, when the researchers mutated two conserved residues within the AIM, changing phenylalanine 134 and leucine 137 to alanines, the interaction vanished entirely, proving that MACP2 docks onto ATG8 through the canonical autophagy receptor motif.
With the interaction established, the team turned to genetics to ask what MACP2 actually does in a living plant. They compared wild-type Arabidopsis with macp2 knockout mutants and with lines overexpressing the protein, subjecting all genotypes to carbon starvation by transferring them to continuous darkness and to nitrogen starvation by growing them on nitrogen-free medium. The results were striking and, at first glance, counterintuitive. Plants lacking MACP2 were more tolerant of starvation, retaining greener leaves and higher chlorophyll content than wild type, while plants overproducing MACP2 senesced prematurely and wilted faster under the same conditions. Under normal growth the differences were subtle, appearing only as the plants aged, with overexpression lines yellowing by five to six weeks while the knockouts stayed green longer than their wild-type counterparts. This pattern, in which too much of a protein mimics the hypersensitivity of autophagy-defective mutants, suggested that MACP2 acts as a positive regulator of autophagy induction: without it, the starvation signal is blunted, and with too much of it, the signal becomes destructive.
The signal in question, the study reveals, is hydrogen peroxide. Reactive oxygen species have long been recognized as double-edged messengers in plant cells: at low concentrations they act as second messengers that coordinate growth and stress responses, but in excess they oxidize proteins, lipids, and DNA beyond repair. Outside the cell, in the apoplast, hydrogen peroxide is generated primarily by plasma membrane NADPH oxidases of the RBOH family, particularly RBOHD and RBOHF. Earlier work had hinted that these apoplastic reactive oxygen species are required for autophagosome formation during hypoxia, but the molecular channel connecting extracellular peroxide to the intracellular autophagy machinery had never been identified. Using an Amplex Red fluorescence assay, the researchers measured total hydrogen peroxide levels across their genotypes and found that MACP2-overexpressing plants accumulated significantly more peroxide during senescence and starvation, while macp2 knockouts accumulated significantly less. To visualize peroxide influx directly, they deployed HyPer, a genetically encoded fluorescent biosensor specific for hydrogen peroxide, and showed that the peroxide-dependent fluorescence ratio in root cells rose sharply after carbon starvation in wild-type seedlings but remained substantially lower in macp2 mutants both before and after starvation.
The picture that emerges is one of MACP2 acting as a gated doorway for stress chemistry. The authors propose that in response to senescence or nutrient deprivation, MACP2 proteins assemble into homo- or hetero-polymeric pores at the plasma membrane, structurally analogous to the membrane attack complexes of animal immunity, and that these pores allow extracellular hydrogen peroxide, produced by RBOH oxidases, to flood into the cytoplasm. There, the peroxide functions as a signal that triggers the initiation of autophagosome biogenesis. Supporting this model, supplying the antioxidant glutathione to starving seedlings completely rescued the starvation hypersensitivity of MACP2-overexpressing lines and partially rescued the atg5-1 autophagy mutant, demonstrating that the damaging phenotypes are chemically driven by peroxide rather than by some unrelated consequence of MACP2 abundance. The salicylic acid pathway was also implicated: overexpression lines accumulated elevated levels of salicylic acid and its glycoside SAG, showed upregulation of senescence-associated and defense genes including SAG13, SAG29, PR1, and WRKY53, and lost their accelerated senescence when crossed into eds1 or pad4 salicylic acid signaling mutants. Salicylic acid and reactive oxygen species, the authors note, appear to form a mutually reinforcing amplification loop that drives both autophagy induction and senescence.
Perhaps the most elegant evidence for MACP2’s role came from epistasis experiments with the autophagy mutant atg5-1. The atg5-1 mutant, defective in a core autophagy conjugation factor, displays the classic autophagy-defective phenotype: early leaf yellowing, runaway peroxide accumulation, and extreme sensitivity to carbon and nitrogen starvation. Remarkably, when the researchers crossed macp2 into the atg5-1 background, the double mutants were substantially rescued, showing delayed senescence, higher chlorophyll content, greater starvation tolerance, and dramatically lower hydrogen peroxide levels than atg5-1 alone. In other words, removing the peroxide doorway suppressed the damage caused by removing the recycling machinery itself, placing MACP2 genetically upstream of ATG5 as a regulator of the peroxide signal that both induces autophagy and, when unchecked, kills the cell. Consistent with this, direct measurements of autophagosome formation using an eGFP-ATG8e reporter showed that macp2 mutants formed fewer starvation-induced autophagic puncta and released less free eGFP, a readout of autophagic flux, while overexpression lines formed more. Treating seedlings with exogenous hydrogen peroxide or the salicylic acid agonist BTH likewise induced fewer puncta in macp2 and more in overexpressors, confirming that MACP2 channels the peroxide signal into autophagosome biogenesis.
The story acquires a final twist in the regulation of MACP2 itself. Under normal conditions, a MACP2-YFP fusion localizes to the plasma membrane, but after twelve hours of carbon or nitrogen starvation the fluorescent signal redistributes into cytoplasmic vesicles that colocalize with the autophagosome marker mCherry-ATG8f. Prolonged darkness causes MACP2 protein abundance to collapse after roughly 24 to 48 hours, and this degradation is blocked by concanamycin A, an inhibitor of vacuolar degradation, but not by the proteasome inhibitor MG132, indicating that MACP2 is dismantled by autophagy rather than by the 26S proteasome. Protein stability assays in protoplasts from atg5-1 and rpn10-1 mutants, the latter lacking a dual ATG8 and ubiquitin receptor, confirmed that MACP2 degradation requires the autophagy machinery and likely involves RPN10 acting as a selective autophagy receptor. The researchers also found that MACP2 lacking a functional AIM degrades more slowly and that its stability no longer depends on ATG8e, tying the feedback degradation directly to the ATG8 interaction. This creates a self-limiting circuit: early in starvation, MACP2 pores admit peroxide, peroxide induces autophagy, and autophagy preserves cellular homeostasis; but as peroxide continues to accumulate, MACP2 itself is loaded onto autophagosomes and destroyed, throttling the peroxide influx and protecting the cell from self-destruction during extended deprivation.
The implications of this work extend well beyond a single weed species. It identifies, for the first time, a molecular conduit through which extracellular reactive oxygen species gain access to the autophagy induction machinery in plants, and it adds MACP2 to a growing list of membrane proteins, including the aquaporins PIP1;4 and PIP2;1, that govern hydrogen peroxide trafficking across plant membranes. It also resonates with recent findings in animal cells, where the pore-forming protein gasdermin D mediates the release of mitochondrial reactive oxygen species, suggesting that pore-mediated peroxide transport may be a broadly conserved principle of cellular stress signaling. For agriculture, the prospect of tuning MACP2 or its relatives to adjust the sensitivity of crops to nutrient stress, darkness, or pathogen attack is an obvious long-term goal, though the authors caution that the precise architecture of the proposed MACP2 pore and the exact mechanism by which RPN10 targets MACP2 for degradation remain to be worked out. What is already clear is that the plant’s decision to consume itself in hard times is not a spontaneous collapse but a signal-driven choice, and that the gatekeeper of that signal is a protein borrowed from the ancient arsenal of membrane attack.
Subject of Research: Role of the Arabidopsis pore-forming protein MACP2 in linking extracellular reactive oxygen species to autophagy induction during nutrient starvation
Article Title: Arabidopsis MACP2 contributes to autophagy induction by modulating starvation-induced reactive oxygen species homeostasis
Article References: Zhou, Y., Zhang, X., Qi, T., Wang, Z.-H., Wang, Y., Wang, L.-N., Zeng, Y.-L., He, H., Jiang, L., Xie, D., Xiao, S., Yu, L.-J., & Chen, Q.-F. (2025). Arabidopsis MACP2 contributes to autophagy induction by modulating starvation-induced reactive oxygen species homeostasis. Advanced Biotechnology, 3(3), Article 25. https://doi.org/10.1007/s44307-025-00078-4
Image Credits: AI Generated
DOI: 10.1007/s44307-025-00078-4
Keywords: autophagy, MACP2, Arabidopsis, reactive oxygen species, hydrogen peroxide, ATG8, nutrient starvation, leaf senescence, membrane attack complex, salicylic acid, plant stress, RPN10
Cite Scienmag News
Drew Townsend. (October 4, 2026). Plant Pore Protein MACP2 Revealed as Missing Link Between Starvation Signals and Autophagy. Scienmag. https://scienmag.com/plant-pore-protein-macp2-revealed-as-missing-link-between-starvation-signals-and-autophagy/
Drew Townsend. "Plant Pore Protein MACP2 Revealed as Missing Link Between Starvation Signals and Autophagy." Scienmag, 4 October 2026, https://scienmag.com/plant-pore-protein-macp2-revealed-as-missing-link-between-starvation-signals-and-autophagy/. Accessed 4 October 2026.
Drew Townsend. "Plant Pore Protein MACP2 Revealed as Missing Link Between Starvation Signals and Autophagy." Scienmag. October 4, 2026. https://scienmag.com/plant-pore-protein-macp2-revealed-as-missing-link-between-starvation-signals-and-autophagy/

