Broken mitochondria in the brain's immune cells may be a hidden driver of depression
A new review links stress-damaged mitochondria in microglia to chronic brain inflammation and depression, pointing to new therapeutic targets beyond serotonin.
Summary
Depression affects over 300 million people, yet many do not respond well to current treatments. This review argues that part of the problem lies in microglia, the brain's resident immune cells, and in their mitochondria, the cell's energy producers. Under chronic stress, microglial mitochondria switch from efficient energy production to a faster, inflammatory fuel mode. They also leak reactive oxygen species and mitochondrial DNA, which sets off alarm pathways that keep microglia activated. Activated microglia release inflammatory molecules and prune healthy synapses, which damages circuits that regulate mood. Most evidence comes from stressed rodents, backed by brain imaging and postmortem work in people with depression. The authors suggest future treatments should restore microglial balance rather than simply suppress these cells, because microglia also protect the brain.
Detailed Summary
Major depressive disorder affects more than 300 million people worldwide, and antidepressants and psychotherapy often work slowly or incompletely. Monoamine-deficiency and stress-hormone (HPA axis) theories do not explain how varied the disease is. This narrative review proposes that microglia, the brain's resident immune cells (about 10–15% of glial cells), and the mitochondria inside them form a unifying pathogenic axis linking chronic stress to neuroinflammation and mood symptoms.
The authors first summarize evidence that microglia are activated in depression. PET imaging shows microglial activation in the prefrontal and anterior cingulate cortex of people with depression, and the degree of activation tracks with episode severity. Postmortem tissue shows higher microglial density with elevated Iba1, CD45, and MCP-1. Similar activation appears in rodent models of lipopolysaccharide exposure, chronic unpredictable mild stress, social defeat, and restraint stress. The picture is not uniform, however. One transcriptomic study found depression-associated microglia in gray matter with an immune-suppressed profile, with lower MKI67, SPP1, C1Q, CD14, and CD163, unlike the pro-inflammatory profile in Alzheimer's disease.
The review then describes how activated microglia cause harm. They release cytokines that destabilize NMDA receptors and cause excitotoxicity. They pass inflammatory microRNAs to astrocytes through exosomes, pushing them toward a toxic A1 state. Under stress they also engulf excitatory synapses via CX3CR1 and complement pathways, reducing spine density in the hippocampus and prefrontal cortex. TNF-α and IL-1β additionally impair long-term potentiation, the cellular basis of learning and resilience.
Mitochondria sit upstream of much of this. Activated microglia shift from oxidative phosphorylation to aerobic glycolysis through HIF-1α, which upregulates HK2, PFKFB3, and LDHA. In stressed mice, brain glucose, glycolytic enzymes, and hexokinase activity rise. Lactate accumulation activates the NLRP3 inflammasome. The glycolysis inhibitor 2-deoxy-D-glucose reduced ROS and cytokines in corticosterone-exposed microglia and eased depression-like behavior in stressed mice. Succinate buildup also drives reverse electron transport at complex I, generating mitochondrial ROS. N-acetylcysteine or microglial depletion with PLX5622 blocked social-defeat-induced behaviors.
Damaged mitochondria also release their 16,569 base-pair genome, which acts as a danger signal. Cytosolic mtDNA activates the cGAS-STING pathway, TBK1, IRF3, and type I interferon, and engages NLRP3. Through TLR9-MyD88 it also activates NF-κB and MAPK, spreading inflammation to neighboring microglia. The abstract also names fission-fusion imbalance and impaired autophagy and mitophagy, although the supplied text ended before those sections.
The authors caution that microglial activation is not purely harmful. Depleting microglia with PLX3397 worsened stress-induced depressive behavior, so the goal should be restoring microglial homeostasis rather than blanket suppression. Caveats include a heavy reliance on rodent models, heterogeneous human microglial signatures, unclear causality in people, and some mechanistic support borrowed from non-brain tissue such as cardiac reperfusion studies.
Key Findings
- Major depressive disorder affects more than 300 million people worldwide, and current treatments often show limited efficacy and delayed response.
- PET imaging shows microglial activation in the prefrontal cortex and anterior cingulate cortex of people with depression, positively correlated with episode severity.
- Postmortem depression brains show elevated Iba1, CD45, and MCP-1, but gray matter microglia may show an immune-suppressed 'DepDAM' profile, unlike Alzheimer's-type disease-associated microglia.
- In stressed mice, brain glucose, glycolytic enzymes, and hexokinase activity rise; the glycolysis inhibitor 2-DG reduced ROS, cytokines, and depression-like behavior.
- Succinate-driven reverse electron transport at complex I generates mitochondrial ROS; N-acetylcysteine or PLX5622 microglial depletion blocked social-defeat-induced depressive behaviors.
- Released mitochondrial DNA (16,569 bp, 37 genes) activates cGAS-STING, TBK1, IRF3, NLRP3, and TLR9-MyD88-NF-κB signaling, amplifying microglial inflammation.
- Depleting microglia with PLX3397 worsened chronic stress-induced depressive behavior, indicating microglia also have protective roles.
Methodology
This is a narrative review integrating molecular, cellular, animal, and human imaging and postmortem studies on microglia, mitochondrial function, and depression. No systematic search protocol, pooled sample size, or quantitative meta-analysis is described. Evidence is drawn largely from rodent stress models (LPS, CUMS, CSDS, CRS) and in vitro microglial cultures, plus PET and postmortem studies in people with depression. The supplied full text ended partway through the mitochondrial DNA section, so later sections were not available for this summary.
Study Limitations
The evidence is largely preclinical, with human findings on microglial phenotypes inconsistent and heterogeneous, and causality in people unproven. The authors note that mechanisms by which microglia mediate depression remain unclear and that some mitochondrial evidence derives from non-brain contexts. No conflicts of interest were available in the supplied text.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: