How Microglial Metabolism and Mitochondria Drive Brain Damage After Stroke

When a blood clot chokes off the supply of oxygen and glucose to a region of the brain, the neurons caught in the ischemic zone begin to die within minutes. But the story of stroke injury does not end with the neurons. Embedded throughout the brain is a population of resident immune cells called microglia, and in the hours and days that follow an ischemic insult, these cells undergo a dramatic transformation. A new review published in the Journal of Translational Medicine by Yujie Ding, Nan Li, Qian Li and colleagues at Anhui University of Chinese Medicine weaves together three previously separate strands of stroke biology—mitochondrial dysfunction, metabolic reprogramming, and activation of the NLRP3 inflammasome—into a single, integrated framework that could reshape how researchers think about neuroinflammation after cerebral ischemia.

Microglia are the sentinels of the central nervous system. In the healthy brain they patrol their territory with ramified processes, sampling the microenvironment for signs of danger. When ischemia strikes, they rapidly shift into an activated state, and for years researchers described this shift using a binary model: pro-inflammatory M1-like microglia that amplify tissue damage, and anti-inflammatory M2-like microglia that promote repair. The new review argues that this tidy dichotomy has outlived its usefulness. Single-cell sequencing studies have revealed a far richer spectrum of microglial subtypes, each with distinct transcriptional profiles, metabolic preferences, and functional consequences. Polarization, in this view, is not a switch but a continuum—a dynamic landscape of states that cells occupy, abandon, and re-enter as the post-stroke environment evolves.

At the heart of this transformation lies a profound metabolic reprogramming. Resting microglia rely primarily on oxidative phosphorylation, the efficient mitochondrial process that converts nutrients into ATP. After ischemia, however, oxygen and glucose scarcity forces a shift toward aerobic glycolysis, the same fermentative strategy that Warburg observed in cancer cells. This metabolic pivot is orchestrated in large part by hypoxia-inducible factor 1-alpha, or HIF-1α, which stabilizes under low-oxygen conditions and drives expression of glycolytic enzymes such as hexokinase 2. The result is a cell that burns glucose wastefully but rapidly, generating the biosynthetic intermediates and NADPH it needs to fuel an immune response—while simultaneously producing lactate and succinate as metabolic byproducts with signaling roles of their own.

The review pays particular attention to mitochondria, which are not merely passive casualties of ischemia but active participants in the inflammatory cascade. Under normal conditions, a network of mitochondrial quality control mechanisms—collectively termed MQC—keeps these organelles healthy. Mitochondrial dynamics, the constant fission and fusion of the network, allow damaged segments to be segregated. Mitophagy, guided by the PINK1-Parkin pathway, then removes severely compromised mitochondria before they can do harm. Ischemia disrupts this surveillance system. The mitochondrial permeability transition pore opens, membrane potential collapses, and the quality control machinery becomes overwhelmed. The consequence is the release of a battery of mitochondrial damage-associated molecular patterns: reactive oxygen species generated at the electron transport chain, fragments of oxidized mitochondrial DNA, and cardiolipin, a phospholipid normally confined to the inner mitochondrial membrane that becomes a potent danger signal when exposed to the cytosol.

These mitochondrial alarm signals converge on one of the most consequential immune complexes known: the NLRP3 inflammasome. NLRP3, a cytosolic pattern-recognition receptor, assembles with the adaptor protein ASC and procaspase-1 into a multi-protein platform that activates caspase-1. The review highlights how several metabolic and mitochondrial cues converge to trigger this assembly. Succinate, which accumulates during ischemia, can drive reverse electron transport through mitochondrial complex I, supercharging the production of mitochondrial reactive oxygen species. Oxidized mitochondrial DNA binds directly to NLRP3. Extracellular lactate and cytosolic cardiolipin provide additional priming and activation signals. Once assembled, the inflammasome executes its inflammatory program: caspase-1 cleaves pro-interleukin-1β and pro-interleukin-18 into their mature, secreted forms, and cleaves gasdermin D to release its N-terminal fragment, GSDMD-NT.

What happens next is where the review makes its most striking conceptual contribution. GSDMD-NT is not simply a bystander in inflammation—it migrates to the mitochondria and perforates both the inner and outer mitochondrial membranes, forming pores that further destabilize these organelles. This establishes a self-amplifying positive feedback loop: mitochondrial damage releases the signals that activate the NLRP3 inflammasome, and inflammasome activation inflicts further mitochondrial damage. In the ischemic brain, where microglia are already metabolically stressed, this loop can lock the cell into a hyperinflammatory state that persists long after the initial insult. The authors also note that activated mitochondrial DNA can engage the STING pathway, adding an interferon-mediated dimension to the inflammatory response and further entrenching the pro-inflammatory phenotype.

Crucially, the review insists that this axis cannot be understood without considering time. The authors stratify the MQC–metabolism–inflammation network into three temporal phases. In the hyperacute phase, within minutes to hours of vessel occlusion, energy failure and mitochondrial permeability transition dominate, and the earliest danger signals are released. In the subacute phase, spanning hours to days, microglial polarization reaches its peak, glycolytic reprogramming is fully established, and NLRP3-driven cytokine release shapes the evolving lesion. In the chronic phase, weeks to months later, a lingering population of chronically activated microglia sustains low-grade inflammation that interferes with tissue remodeling and repair. Each phase, the authors argue, presents distinct therapeutic windows and distinct molecular targets, and interventions that ignore this temporal structure are unlikely to succeed.

The review is also refreshingly candid about the translational bottlenecks that have stalled progress from bench to bedside. Numerous NLRP3 inhibitors and mitochondrial protective agents have shown promise in rodent models of stroke, yet clinical trials have repeatedly disappointed. One reason, the authors emphasize, lies in fundamental differences between mice and humans. The NLRP3 inflammasome and its regulatory feedback mechanisms differ in important ways between the two species, meaning that regulatory circuits mapped in mouse models must be interpreted with caution before being extrapolated to patients. The authors call for humanized models and bidirectional clinical validation to close this gap, alongside single-cell multi-omics approaches to map the full diversity of microglial subpopulations in the human ischemic brain, and metabolic flux analysis to quantify the actual thresholds at which metabolites such as succinate and lactate become pathogenic.

Two testable hypotheses emerge from the framework. The first proposes that the strength and duration of the mitochondrial damage–inflammasome feedback loop determines whether microglia resolve toward a reparative phenotype or remain trapped in a chronic inflammatory state—suggesting that early stabilization of mitochondrial quality control could steer the entire trajectory of post-stroke inflammation. The second proposes that specific metabolite thresholds, rather than bulk shifts in metabolic pathway activity, are the critical determinants of NLRP3 activation, implying that precise quantification of metabolite flux in individual microglial subtypes will be essential for rational drug design. Both hypotheses lend themselves to direct experimental interrogation with the tools now available, from metabolomics to live imaging of mitochondrial dynamics.

For a field that has long treated metabolism, mitochondria, and inflammation as separate chapters of the stroke story, this integrated framework offers something genuinely new: a systems-level map of how the pieces fit together in space and time. If the hypotheses hold up, the implications extend beyond ischemic stroke to any neurological condition in which microglial metabolism and inflammasome activation run amok, from traumatic brain injury to neurodegenerative disease. The therapeutic challenge remains formidable—timing interventions to the right phase, achieving selectivity for pathogenic microglial states, and bridging the species divide—but the review provides a clear conceptual scaffold on which the next generation of experiments, and perhaps eventually therapies, can be built.

Subject of Research: Microglial metabolic reprogramming, mitochondrial dysfunction, and NLRP3 inflammasome activation in cerebral ischemia

Article Title: Metabolic reprogramming and polarization of microglia in cerebral ischemia: the roles of mitochondria and inflammasomes

Article References: Ding, Y., Li, N., Li, Q., Zhang, H., He, L., Gao, F., & Li, P. (2026). Metabolic reprogramming and polarization of microglia in cerebral ischemia: the roles of mitochondria and inflammasomes. Journal of Translational Medicine. https://doi.org/10.1186/s12967-026-08947-9

Image Credits: AI Generated

DOI: 10.1186/s12967-026-08947-9

Keywords: ischemic stroke, microglia, mitochondrial quality control, metabolic reprogramming, NLRP3 inflammasome, neuroinflammation, mitochondrial dysfunction, aerobic glycolysis, gasdermin D, single-cell sequencing, HIF-1alpha, neuroimmunology