The scientific logic rests on a phenomenon called immunometabolic reprogramming. Microglia, the resident innate immune cells of the central nervous system, normally run on mitochondrial oxidative phosphorylation, an efficient mode of energy generation that sustains their quiet surveillance of the brain. Confronted with amyloid-β aggregates, oxidative stress, and persistent inflammatory signals, however, they shift toward aerobic glycolysis, the rapid but wasteful glucose-burning program familiar from cancer cells and acutely activated immune cells. Useful for a short burst of defense, this metabolic posture becomes destructive when it hardens into a permanent state, locking microglia into a pro-inflammatory phenotype characterized by excessive reactive oxygen species, crippled respiration, and amplified cytokine output. At the center of the malfunction sits a three-part signaling circuit. AMPK acts as the cell’s fuel gauge, preserving mitochondrial homeostasis and restraining inflammation. mTOR integrates nutrient and growth signals to drive anabolic, glycolytic metabolism, while HIF-1α, which mTOR activates, switches on the transcription of glycolytic enzymes and glucose transporters. In the Alzheimer’s brain, AMPK activity falters even as mTOR and HIF-1α remain stubbornly switched on, and the resulting inflammation-metabolism loop feeds on itself.

Rather than blocking a single inflammatory molecule, the team drew on the traditional Chinese medicine principle of “BuShen HuoXue,” tonifying the kidney and activating blood circulation, long applied to disorders of cognitive decline. From that framework they selected two compounds: icaritin, a flavonoid from Epimedium reported to support energy sensing and mitochondrial health, and tanshinone IIA, a diterpene from Salvia miltiorrhiza with documented anti-neuroinflammatory activity. Network pharmacology suggested the pair converges on AMPK and mTOR signaling, but both molecules are handicapped as drugs: they dissolve poorly in water, cross the blood-brain barrier inefficiently, and never reach meaningful concentrations inside diseased microglia. The researchers’ answer, developed by Ge Zhang, Ying Yang, Xue-tao Li, Yang Yu, and colleagues, was to encapsulate them within nanoparticles built from Panax notoginseng polysaccharide, a biocompatible, biodegradable macromolecule with intrinsic antioxidant and immunomodulatory activity that the team treated not as an inert wrapper but as a functional component of the therapy itself.

The resulting platform, named KPBIT@NPs, is a small feat of materials chemistry. The researchers first grafted 4-carboxyphenylboronic acid pinacol ester, or CPBA, onto the polysaccharide, adding hydrophobic domains that let the polymer chains self-assemble into spherical particles in water. They then attached KLVFFAED, a peptide taken from the amyloid-β sequence itself, exploiting its ability to engage RAGE, the receptor for advanced glycation end products, which is upregulated in Alzheimer’s disease and participates in amyloid transport across the blood-brain barrier. Proton nuclear magnetic resonance and Fourier transform infrared spectroscopy confirmed each modification step, and systematic variation of the CPBA feed ratio revealed a non-linear relationship between grafting density and assembly behavior, with a grafting degree of 8.95 percent giving the lowest critical aggregation concentration while preserving water solubility. The final formulation formed uniform, near-spherical particles about 79 nanometers in diameter with a near-neutral surface charge, encapsulating 92 percent of the icaritin and 87 percent of the tanshinone IIA and remaining stable in serum for two weeks.

The most elegant feature is the particles’ responsiveness to the disease environment. The phenylboronic ester bonds anchoring CPBA to the polysaccharide are cleaved by reactive oxygen species, which accumulate in the inflamed Alzheimer’s brain. When the nanoparticles were exposed to hydrogen peroxide in the laboratory, transmission electron microscopy showed them fragmenting, and drug release accelerated in a concentration-dependent manner from mild oxidative stimulation to high oxidative challenge; in neutral buffer, the cargo stayed locked inside. To test delivery, the team built a blood-brain barrier model with bEnd.3 endothelial cells cultured in Transwell chambers above BV2 microglia, then degraded the barrier with amyloid-β and lipopolysaccharide to mimic disease conditions while preserving measurable barrier integrity, verified by transendothelial electrical resistance, sodium fluorescein permeability, and tight-junction staining. KLV-modified nanoparticles crossed the injured barrier far more efficiently than unmodified versions and were taken up more avidly by the microglia below, an advantage that collapsed when cells were pretreated with the RAGE inhibitor FPS-ZM1. In living animals, near-infrared imaging showed the modified particles circulating for up to 72 hours and accumulating in APP/PS1 mouse brain well above non-targeted controls.

Inside microglia, the two drugs proved better together than apart. Using the SynergyFinder platform, the team screened combinations of free icaritin and tanshinone IIA in BV2 microglia injured by amyloid-β and lipopolysaccharide and identified a 1:1 molar ratio as optimal, with synergy scores above the threshold for significant cooperation. Nanoparticles loaded at that ratio outperformed single-drug and non-targeted formulations across every assay. They cut intracellular reactive oxygen species, lowered malondialdehyde, a marker of lipid peroxidation, and raised glutathione and catalase. They suppressed interleukin-6, interleukin-1β, and tumor necrosis factor-α while boosting the reparative markers arginase-1 and interleukin-10. Flow cytometry and immunofluorescence tracked a phenotypic migration from the pro-inflammatory M1 state, marked by CD86, toward the restorative M2 state, marked by CD206, and JC-1 staining showed damaged mitochondrial membrane potential substantially restored. Conditioned medium from treated microglia, transferred onto HT22 neurons, reduced neuronal apoptosis and oxidative stress, demonstrating that reprogrammed microglia become actively protective toward their neighbors.

The mechanism was then interrogated at the level of signaling proteins. Western blotting revealed that stressed microglia carried a depressed p-AMPK/AMPK ratio alongside elevated p-mTOR/mTOR and HIF-1α, the biochemical signature of a cell stuck in glycolytic overdrive, and KPBIT@NPs reversed all three markers. Downstream metabolic enzymes told the same story: expression of PFKFB3, a rate-limiting glycolytic regulator, fell, while inhibitory phosphorylation of PDHE1α declined, freeing pyruvate to enter the tricarboxylic acid cycle. Real-time flux analysis measured a drop in the extracellular acidification rate, a proxy for glycolysis, and a rise in the oxygen consumption rate, the direct readout of mitochondrial respiration, with basal respiration, ATP production, maximal respiration, and spare respiratory capacity all climbing. The glycolytic enzyme LDHA dimmed while ATP5A, a core subunit of mitochondrial ATP synthase, brightened. Crucially, the AMPK inhibitor Compound C blunted these effects and the mTOR activator MHY1485 reversed them, establishing that the AMPK-mTOR/HIF-1α axis is not merely correlated with the therapeutic effect but required for it.

Computational analyses reinforced the drug pairing. The team assembled a network spanning 48 active compounds from Epimedium and Salvia miltiorrhiza, 736 predicted drug targets, and 16,820 Alzheimer’s-related genes, converging on 688 shared targets. Pathway enrichment placed AMPK and mTOR signaling among the most significantly represented routes, with HIF-1α emerging as a hub node, and molecular docking with AutoDock Vina returned favorable binding energies for both compounds against both proteins: tanshinone IIA bound AMPK at −9.70 kilocalories per mole and mTOR at −8.83, while icaritin registered −8.17 and −7.17 respectively, with hydrogen bonding and hydrophobic contacts stabilizing the interactions. The computations suggest that this classical herb pair achieves, through complementary chemistry, precisely the dual regulation that Alzheimer’s pathology disrupts.

In APP/PS1 mice, the cellular effects translated into behavior and brain structure. Treated animals located the hidden platform faster in the Morris water maze, spent more time in the target quadrant, crossed the platform location more often, and built more coherent nests, a standard measure of daily function. Amyloid plaque burden in cortex and hippocampus shrank, Nissl staining revealed healthier neuronal layers, NeuN staining indicated greater neuronal survival, and Golgi staining showed dendritic trees with restored length and branching. Transmission electron microscopy of hippocampal tissue found mitochondria with intact cristae and membranes where model mice carried swollen, disrupted organelles. Whole-brain biochemical assays completed the picture: activities of the tricarboxylic acid cycle enzymes isocitrate dehydrogenase, α-ketoglutarate dehydrogenase, and succinate dehydrogenase, along with respiratory chain complexes I through V and total ATP content, all rebounded, while lactate, pyruvate, hexokinase, and lactate dehydrogenase, the fingerprints of glycolytic accumulation, receded. Markers of astrocyte and microglial activation, GFAP and IBA-1, fell, and co-staining confirmed the CD86-to-CD206 phenotypic flip within IBA-1-positive microglia in the brain itself.

Safety data were reassuring at this stage: hemolysis rates stayed below accepted thresholds, organ histology appeared normal, blood counts were unremarkable, and serum cytokines showed no peripheral inflammatory activation. The authors are candid about limitations. Their brain metabolic measurements relied on bulk homogenates that cannot separate microglial metabolism from that of neurons or astrocytes, direct intracerebral colocalization of the nanoparticles was not obtained, and long-term biodistribution, biodegradation, and clearance remain unexamined. They propose a three-stage follow-up combining single-cell multi-omics with spatial metabolic imaging, extracellular flux and isotope-tracing assays on purified primary microglia, and microglia-specific blockade of the AMPK-mTOR/HIF-1α axis to establish cell-autonomous causality. Even with those caveats, the study delivers a proof of concept with wider implications: a natural polysaccharide can be engineered into a carrier that is not passive cargo space but an active participant in therapy, delivering two synergistic plant compounds to the exact immune cells whose metabolic derailment helps drive neurodegeneration. If the framework survives more advanced preclinical testing, it could extend beyond Alzheimer’s disease to other disorders, from Parkinson’s disease to stroke, in which inflammation and metabolism fail together.

Atorvastatin, widely known for its cholesterol-lowering properties, has garnered increasing attention for its pleiotropic effects in the central nervous system. Prior studies have hinted at its ability to modulate neuroinflammation and enhance cerebral blood flow, but the underlying molecular details had remained elusive. In this latest investigation, the authors employed a naturally aging rat model to closely mimic human aging, which is pivotal given the complexity and multifactorial nature of cognitive deterioration in elderly populations. Their strategic use of this model allowed for the observation of atorvastatin’s effects over an extended period, revealing sustained cognitive benefits that correlate with biochemical modifications in neuronal structures.

Central to the study’s findings is the dynamic interplay between lysine crotonylation and ubiquitination at position 272 on the NFL protein. NFL, a fundamental component of the neuronal cytoskeleton, is integral to maintaining axonal integrity and facilitating proper nerve signal conduction. Modifications at the lysine 272 residue appear to act as molecular switches that regulate NFL’s stability and turnover. The researchers discovered that atorvastatin increases SIRT2-mediated decrotonylation at this site, which subsequently promotes ubiquitination. This orchestrated transition facilitates the clearance of damaged NFL proteins, thereby preserving cytoskeletal architecture and enhancing neuronal resilience.

SIRT2, a member of the sirtuin family of NAD+-dependent deacylases, emerges in this study as a pivotal enzymatic regulator orchestrating this modification cascade. Previous literature has established SIRT2’s involvement in neurodegeneration and metabolic regulation, but this particular study delves deeper into its nuanced role in modulating post-translational modifications relevant to aging neurons. By enhancing SIRT2’s de-crotonylase activity, atorvastatin appears to fine-tune the balance between protein modification states, enabling more effective proteasomal degradation of damaged or dysfunctional NFL molecules.

What makes these findings particularly compelling is the link between molecular modulation and actual cognitive improvements observed in the aging rats. Behavioral assays conducted over the duration of the study documented significant enhancements in memory, learning, and spatial navigation among atorvastatin-treated subjects compared to controls. This provides strong evidence that targeting the SIRT2-NFL modification axis does not merely represent an abstract biochemical phenomenon but translates into tangible neurocognitive benefits with potential clinical significance.

Further biochemical analyses revealed that untreated aging rats exhibited elevated levels of lysine 272 crotonylation on NFL alongside diminished ubiquitination, correlating with increased accumulation of misfolded NFL aggregates. These aggregates are hypothesized to disrupt axonal transport and synaptic function, underpinning various cognitive deficits. Atorvastatin treatment reversed this pattern, amplifying ubiquitination and promoting clearance of these neurotoxic protein forms, emphasizing the drug’s role in maintaining protein homeostasis through post-translational modification dynamics.

The study also addressed the broader implications of SIRT2’s role in cellular aging. Beyond its well-established functions in metabolic sensing and gene expression regulation, SIRT2’s involvement in modulating the proteostasis network represents an exciting frontier. The ability of atorvastatin to upregulate this pathway hints at potential cross-talk between lipid metabolism modulators and epigenetic-like enzyme activities, opening avenues for novel polypharmacological strategies to mitigate aging-related neurodegeneration.

This intersection between lipid-lowering therapies and epigenetic regulation of neuronal proteins represents a paradigm shift in understanding how systemic pharmacological interventions can impact brain aging. It positions atorvastatin as a candidate drug for repurposing in neurodegenerative therapeutics, especially considering its known safety profile and extensive clinical use. However, important questions remain regarding dosage optimization, the precise timing of intervention, and long-term consequences on neuronal function that subsequent studies will need to address.

Intriguingly, the research team also speculated on the possibility that modulating post-translational modifications on NFL might influence the interaction dynamics with other neurofilament subunits and associated cytoskeletal components. Such changes could have ripple effects on axonal transport efficiency and synaptic connectivity, hallmarks that degenerate in multiple neurodegenerative diseases including Alzheimer’s and Parkinson’s disorders. Thus, refining our understanding of these molecular switches might yield broader implications for neurobiology and aging research.

Moreover, the utilization of cutting-edge mass spectrometry techniques allowed for precise quantification and localization of lysine crotonylation and ubiquitination marks, providing an unprecedented molecular resolution. The rigorous temporal characterization of these modifications throughout the treatment timeline adds a dynamic dimension, underscoring that the post-translational landscape is fluid and tightly regulated during pharmacological intervention.

The significance of this study extends beyond the immediate context of atorvastatin and aging rats. It adds to a growing body of evidence affirming the importance of reversible acylations, such as crotonylation, in regulating protein function in health and disease. Unlike more traditional post-translational modifications, crotonylation is just beginning to be explored, and its dynamic crosstalk with ubiquitination suggests an intricate regulatory network poised to be a fertile ground for novel therapeutic approaches.

As the global population ages, the quest to preserve cognitive vitality takes on increasing urgency. With this study, the prospect of using a widely available drug to harness endogenous enzymatic machinery for proteome maintenance could represent a major stride forward. It also underscores the critical role of fundamental research in revealing unexpected drug actions and biological pathways that may translate into impactful clinical interventions.

While the results are promising, the authors are cautious in their interpretation and emphasize the necessity for subsequent validation in primate models and eventually human clinical trials. They advocate for integrative studies combining molecular biology, neuroimaging, and cognitive assessment to fully unravel the mechanistic underpinnings and therapeutic potential of targeting the SIRT2-crotonylation-ubiquitination axis.

In conclusion, this innovative study bridges pharmacology, epigenetics, and neurobiology to illuminate a previously unrecognized mechanism by which atorvastatin may confer neurocognitive benefits during aging. The discovery that SIRT2-mediated modulation of NFL lysine 272 crotonylation to ubiquitination enhances cognitive function opens new vistas in the development of therapeutic strategies aimed at ameliorating age-associated cognitive decline. As such, it invites a reassessment of the broader potential of statins beyond cardiovascular health and stimulates enthusiasm for further investigation into the complex regulatory networks governing neuronal longevity and plasticity.

Subject of Research: The study investigates the molecular mechanisms underlying the cognitive improvement induced by long-term atorvastatin treatment, focusing on the modulation of SIRT2-mediated dynamic transitions between lysine 272 crotonylation and ubiquitination on neurofilament light chain (NFL) in naturally aging rats.

Article Title: Long-term atorvastatin improves cognitive function by modulating SIRT2-mediated dynamic transition of NFL lysine 272 crotonylation to ubiquitination in naturally aging rats.

Article References:
Xu, TC., Cai, JR. & Chen, HS. Long-term atorvastatin improves cognitive function by modulating SIRT2-mediated dynamic transition of NFL lysine 272 crotonylation to ubiquitination in naturally aging rats. Cell Death Discov. 11, 463 (2025). https://doi.org/10.1038/s41420-025-02764-7

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41420-025-02764-7