A Hidden Brain Enzyme Emerges as the Secret Weapon Against Memory Decline in Aging

Deep inside the hippocampus, the brain’s memory-processing hub, a quiet chemical drama unfolds as we grow older. Synapses—the microscopic junctions that allow neurons to talk to one another—begin to disappear. Inflammation smolders. Fat molecules pile up where they should not. For decades, neuroscientists have catalogued this deterioration in detail, yet a central mystery has remained: why do some aging brains hold on to their memories despite these insults, while others slide toward mild cognitive impairment and Alzheimer’s disease? A new study published in Nature Neuroscience offers a striking answer, identifying a single secreted enzyme as a molecular guardian of the aging brain.

The research, led by Cinzia Vicidomini and Amar Sahay at Massachusetts General Hospital and Harvard Medical School, together with Gaurav Chopra at Purdue University and a broad multidisciplinary team, began with an in vivo genetic screen designed to catch compensatory factors in the act. Using an inducible mouse model in which the transcription factor KLF9 can be switched on to trigger loss of perforant path synapses in dentate granule cells, the researchers sequenced RNA from the dentate gyrus to see which genes flare up in response to synapse loss. One gene rose above the noise: Pla2g2f, which encodes a type II secreted phospholipase A2 enzyme known as PLA2G2F. In young adult mice, its expression in the brain is essentially undetectable—the highest natural expression in the body is in skin keratinocytes—but during aging it reappears in a sparse population of dentate granule cells, precisely as the synapses feeding into them begin to erode.

The biochemistry of PLA2G2F makes it an unusual suspect for neuroscience. Like other secreted phospholipases, it is a calcium-dependent enzyme that snips fatty acyl chains off the sn-2 position of membrane glycerophospholipids, liberating lysophospholipids and polyunsaturated fatty acids such as arachidonic acid and docosahexaenoic acid. These products are not mere metabolic scraps; they are bioactive signaling molecules that remodel synaptic membranes, alter membrane fluidity and broadcast signals between cells. The researchers hypothesized that aging neurons might deliberately deploy this lipid-hydrolyzing enzyme as a broadcast signal to their neighbors, mobilizing fat stores and calming inflammatory glia in the process.

Crucially, the human data pointed in the same direction. In single-nucleus RNA-sequencing datasets from human hippocampus, Pla2g2f expression is low and sparse in adolescence and adulthood but increases with age in mature granule cells and hippocampal excitatory neurons. Most tellingly, when the team examined dentate granule cells from aged donors—some cognitively healthy, some resilient despite harboring Alzheimer’s histopathology, and some with severe Alzheimer’s disease—the cognitively resilient group showed significantly higher Pla2g2f expression than the severely impaired group. In other words, the enzyme is naturally switched on during aging, and its higher expression tracks with preserved cognition even in brains riddled with disease pathology. This is precisely the signature that cognitive reserve theory predicts: a compensatory mechanism engaged to stave off decline.

To test causality, the team built conditional knockout mice and deleted Pla2g2f selectively in dentate granule cells of middle-aged animals. The consequences were dramatic and wide-ranging. By 20 months of age, the engineered mice had lost dendritic spines on their granule cells, particularly the large-diameter spines that serve as postsynaptic targets of the perforant path, the main input highway to the dentate gyrus. Adult-born granule cells in these animals also matured poorly, showing reduced spine density. Meanwhile, the glial environment deteriorated: the density of reactive GFAP-positive astrocytes climbed, microglia carried elevated levels of the lysosomal marker CD68 and adopted an amoeboid, less ramified shape, and expression of Serpina3n—a serine protease inhibitor linked to amyloid burden and poor cognition in Alzheimer’s—rose in the dentate gyrus. Behaviorally, the effects were equally clear. In the novel object location test, which depends on dentate gyrus function and on binding objects to their spatial contexts, middle-aged mice lacking Pla2g2f could no longer distinguish a moved object from an unmoved one, while their control littermates performed normally. Anxiety-like behavior and dentate-insensitive memory tasks were unaffected, pinpointing the deficit to hippocampal circuitry.

The converse experiment proved even more compelling. When the researchers boosted Pla2g2f expression in the dentate granule cells of aged mice using targeted lentiviral delivery, the aging pathology reversed. Spine density increased and large-diameter spines returned. Reactive astrocytosis receded, microglia shed their inflammatory CD68 load and regained their intricate branching architecture, and hippocampal-dependent memory was preserved in 17-month-old animals at levels comparable to much younger mice. Aged mice with elevated PLA2G2F also recruited their dentate gyrus more vigorously during exploration, showing greater c-FOS activation upon exposure to a novel enriched environment. Most remarkably, when the team boosted Pla2g2f in 12-month-old middle-aged mice and tested them seven months later, the treated animals showed strong trends toward enhanced performance in both novel object location and pre-exposure-dependent contextual fear conditioning, suggesting that enhancing this single pathway in midlife can prophylactically alter the entire trajectory of cognitive decline.

The deepest mechanistic insight concerns fat. Lipid droplets—organelles stuffed with neutral fats such as triacylglycerols and cholesterol esters—accumulate in aging neurons and microglia and are increasingly implicated in Alzheimer’s risk, particularly in cells carrying the APOE4 allele. Lipid-laden microglia lose their capacity for phagocytosis and debris clearance and can even promote neuronal death. Using unbiased automated imaging of lipid droplet burden in the dentate gyrus, the researchers found that deleting Pla2g2f increased lipid droplet area in both the granule cell layer and the molecular layer, while boosting Pla2g2f reduced it. To prove that the enzyme’s catalytic activity—and not some structural scaffolding role—was responsible, they engineered a catalytically dead mutant by substituting two conserved residues, glycine 30 to alanine and histidine 48 to lysine. The dead enzyme failed to suppress reactive gliosis, failed to calm microglia and failed to reduce lipid droplets, establishing that substrate hydrolysis itself drives the neuroprotection.

The team then took the question into human cells. They generated microglia-like cells and excitatory neuron forebrain organoids from human induced pluripotent stem cells, overloaded the microglia with lipid droplets using oleic acid, and applied conditioned medium from organoids expressing either a control construct or Pla2g2f. Only the medium from PLA2G2F-expressing organoids reduced the microglial lipid burden. Quantitative lipidomics using multi-reaction monitoring mass spectrometry revealed what was happening: the treated microglia massively increased their export of neutral lipids, with cholesterol esters showing the largest relative increase at 21.9 percent, alongside elevated triacylglycerols and expanded phospholipid synthesis pathways, particularly phosphatidylcholine. The interpretation is elegant: neuronal PLA2G2F signaling triggers microglia to break down their stored fat and ship it out, likely via lipoprotein particles, while rebuilding membrane phospholipids that support motility, phagocytosis and survival. In vivo, the researchers also documented that PLA2G2F acts at a distance—deleting or boosting the enzyme in the dentate gyrus altered astrocytes and microglia in downstream hippocampal regions CA3 and CA1 and even in the cortex, pointing to both short-range paracrine and long-range signaling.

The lipidomics of the aging dentate gyrus itself reinforced the picture. Tandem mass spectrometry profiling of 736 lipid species across 18 classes showed that loss of neuronal PLA2G2F elevated ceramides, diacylglycerols, lysophosphatidylethanolamine, phosphatidic acid and phosphatidylserine—lipid signatures associated with pathological aging, hippocampal atrophy and, intriguingly, ferroptosis, the iron-dependent cell death that accumulates in aging and neurodegeneration. A computational pipeline the authors call LIGER predicted disrupted lipid conversion reactions and altered expression of lipid-metabolism genes, including other secreted phospholipases and apolipoproteins whose human orthologs, such as APOE, are among the strongest genetic risk factors for Alzheimer’s disease.

The study has caveats, which the authors acknowledge candidly: the lipidomics cannot yet resolve which cell types harbor which lipid changes in vivo, fatty acid signaling mediators escape detection, and larger human cohorts will be needed to cement the link between Pla2g2f expression and cognitive status. Yet the convergence of evidence—aging-sensitive expression in mice and humans, association with cognitive resilience despite Alzheimer’s pathology, bidirectional causality in mice, catalytic dependence, and conservation of the mechanism in human neuron-microglia co-cultures—satisfies the strictest criteria for an endogenous compensatory resilience mechanism. Sahay and colleagues are named co-inventors on a patent covering vectors and methods for treating neurodegeneration based on this work, and the therapeutic implication is clear: if the aging brain already manufactures its own defense molecule in response to synapse loss, then amplifying that defense in midlife might shift the balance between resilience and vulnerability before decline takes hold. In an era when Alzheimer’s drug development has focused overwhelmingly on clearing amyloid plaques, the idea that a lipid-hydrolyzing broadcast signal from a handful of hippocampal neurons could protect memory opens a fundamentally new frontier—one where the goal is not to fight aging’s damage directly, but to strengthen the brain’s own quiet counteroffensive.

Subject of Research: The role of the secreted phospholipase PLA2G2F as a compensatory neuroprotective mechanism maintaining lipid homeostasis and cognitive resilience in the aging hippocampus

Article Title: An aging-sensitive secreted phospholipase confers neuroprotection and cognitive resilience

Article References: An aging-sensitive secreted phospholipase confers neuroprotection and cognitive resilience. (n.d.). https://doi.org/10.1038/s41593-026-02465-z

Image Credits: AI Generated

DOI: 10.1038/s41593-026-02465-z

Keywords: PLA2G2F, hippocampus, dentate gyrus, cognitive aging, cognitive reserve, Alzheimer's disease, lipid droplets, microglia, astrocytes, synapse loss, phospholipase A2, lipidomics