Intranasal Mn3O4 nanozymes reprogram microglial inflammation in Alzheimer’s disease

Reactive oxygen species accumulate when amyloid deposition and inflammatory signaling reinforce one another, and under those conditions microglia drift away from debris clearance and repair toward a state that feeds neuronal stress. That shift has been recognized for years in Alzheimer’s disease, but it has remained difficult to decide whether changing microglial phenotype can do more than transiently quiet inflammation.    Microglia do not just react to amyloid deposits as passive bystanders. They read protein aggregates, stress signals, and cytokine cues, and then choose from functional programs that can differ sharply in consequence. A pro-inflammatory program amplifies cytokine release, nitric oxide production, and oxidative injury; a repair-oriented program supports resolution and tissue maintenance. The difficulty is that these programs are not clean switches in diseased tissue. They are shaped by continuous exposure to fibrillar β-amyloid, by oxidative pressure, and by receptor systems such as TLR4 that can convert extracellular pathology into intracellular inflammatory signaling. When that loop persists, NOX2-derived oxidants help keep NF-κB activity alive, and the brain pays the price in synaptic and neuronal vulnerability. In a recent research paper published in Theranostics, Dr. Jun Xie, Kai Cao, Dr. Luman Liu, Dr. Liding Zhang, Dr. Ying Yang, Dr. Hui Gong, and Professor Haiming Luo from the Wuhan National Laboratory for Optoelectronics at Huazhong University of Science and Technology, developed a water-soluble Mn3O4 nanozyme modified for intranasal delivery and brain access. They used it as a catalytic antioxidant to suppress TLR4/NOX2-associated ROS production and shift microglia toward an M2-like state.

The research team began by building a water-soluble Mn3O4 nanozyme through PEG-lipid modification of a small hausmannite Mn3O4 core. They measured a core size near 4.8 nm, observed a larger hydrodynamic diameter in water, and found that the material behaved mainly as a superoxide-dismutase-like catalyst while also scavenging superoxide and hydroxyl radicals. A broad but not unlimited antioxidant repertoire gave the investigators a tool capable of testing whether selective oxidant control was enough to push microglia away from inflammatory bias. They then challenged N9 microglia with LPS, identified 2 μg/mL as an effective condition for M1-skewing, and showed that the nanozyme suppressed the rise in the CD86/CD206 ratio without introducing appreciable toxicity at the concentrations carried into the phenotype experiments. The investigators also tracked the response over time, which turned out to be informative: LPS drove a pronounced rise in the M1/M2 marker ratio around 24 hours, while nanozyme pretreatment blunted that shift instead of abolishing it outright.

The authors then moved into mouse models in a way that tightened the mechanistic argument. In an acute hippocampal LPS model, they selected an injection dose that maximized M1 polarization and paired it with intranasal nanozyme administration. They confirmed that the material reached the brain within hours, distributed beyond the nasal cavity, and was largely cleared from the body within 36 hours, while hemolysis and routine liver and kidney chemistry did not reveal acute toxicity. In the inflamed hippocampus, the investigators observed lower CD86, higher CD206, reduced TLR4 abundance, and recovery of NeuN-positive neuronal signal after nanozyme treatment.

The researchers then turned to the 5×FAD mouse model and documented elevated TLR4 expression in the diseased hippocampus and prefrontal cortex, which set up the relevance of the same pathway in amyloid-bearing tissue. After one week of intranasal treatment, microglial markers changed little. After four weeks, M2-like microglia increased significantly, while fibrillar β-amyloid burden did not yet shift in a measurable way which may indicate that inflammatory reprogramming may precede visible plaque improvement, and it also warns against demanding immediate amyloid readouts from interventions aimed first at the inflammatory machinery. When the investigators extended dosing to eight weeks, they found better working and spatial memory in Y-maze and Morris water maze tasks, reduced plaque burden, and improved neuronal signal. At later time points, the phenotype response evolved instead of remaining fixed in its earlier form. The authors connected that attenuation to immune tolerance in chronically stimulated microglia, an interpretation consistent with their scheme and with the observed loss of sustained anti-inflammatory suppression at later time points. Parallel molecular measurements supported the pathway argument: after four weeks, the nanozyme reduced TLR4 and NOX2 expression, lowered ROS, and damped downstream NF-κB-associated inflammatory signaling.

To summarize, Professor Haiming Luo and colleagues demonstrated that an intranasal nanozyme improved phenotype in a mouse model of Alzheimer’s disease. Their work linked in an elegant way the pathway control to a staged biological response, with early inflammatory modulation preceding later plaque reduction and cognitive improvement. Indeed, an important contribution here is the ordering of events. The data place microglial reprogramming early, amyloid and cognitive improvement later, and persistent pathway suppression somewhere in the middle of that sequence. That ordering reframes how anti-inflammatory nanomedicines might be evaluated. A therapy aimed at receptor-linked oxidant production may not need to reduce plaques first; it may instead act by changing the inflammatory setting in which plaques continue to injure tissue. That is a different design logic from the classical drug development which target direct anti-amyloid removal, and it has practical consequences for how dosing schedules, biomarkers, and response windows should be interpreted.

The reduced persistence of M2 polarization after prolonged treatment adds a temporal dimension to the response and suggests that microglial behavior changes as treatment continues. Repeated exposure to pathology can push them toward a tolerant state in which inflammatory readouts diminish but functional rescue may still emerge through cumulative changes already set in motion.   It hints that brain-directed antioxidant therapy may work through phase-specific biology, with an early window dominated by inflammatory reset and a later window dominated by tissue-level consequence. Any attempt to translate such a system would need to respect that temporal structure. Continuous dosing, intermittent dosing, or combination therapy with direct amyloid-targeting agents could perform very differently if microglial memory changes the tissue response after the first few weeks.

Intranasal delivery can give brain access without obvious long-term injury in the reported safety assays, the nanozyme retained stability under simulated nasal conditions, manganese ion release remained limited, and no obvious accumulation signal emerged in the brain over the studied interval. Those properties strengthen the translational relevance of the platform by addressing several early practical concerns surrounding brain-directed nanomaterials. More than that, the authors’ work highlights a design principle: a durable catalytic antioxidant can be useful because it can both removes oxidants and also interferes with a receptor-linked inflammatory circuit that helps define microglial identity in disease.