Intranasal Human NSC-Derived EVs Therapy Can Restrain Inflammatory Microglial Transcriptome, and NLRP3 and cGAS-STING Signalling, in Aged Hippocampus

Madhu et al., Journal of Extracellular Vesicles, 2026, gave late middle-aged mice, 18 months old, two intranasal doses of extracellular vesicles from human induced pluripotent stem cell derived neural stem cells, and looked at the aged hippocampus at 20.5 months of age. The treated animals showed less astrocyte hypertrophy, fewer microglial clusters, less oxidative stress, and lower levels of the proteins that drive the NLRP3 inflammasome and the cGAS to STING to interferon axis. Single-cell RNA sequencing of microglia found a broad shift toward oxidative phosphorylation and away from proinflammatory signalling, and the animals performed better on cognitive tests.

That left the question of which cargo inside the vesicles was doing the work. Two microRNAs were the candidates, miR-30e-3p for the inflammasome and miR-181a-5p for STING. The test was to make vesicles without them. Human neural stem cells at passage 11 were given an antagomir against one microRNA or the other at 100 nM, with a scrambled antagomir as the control, and the vesicles they secreted were collected and measured. The antagomirs came from AUM BioTech.

The vesicles from treated cells carried about 79.61% less miR-30e-3p and about 72.81% less miR-181a-5p than vesicles from untreated cells, and the scrambled controls produced no such fall. Vesicles missing miR-30e-3p no longer held down interleukin 1 beta and interleukin 18 in RAW-ASC cells given nigericin, and vesicles missing miR-181a-5p no longer suppressed the interferon reporter in RAW-Lucia-ISG cells given cyclic GMP-AMP. That is what turned two abundant microRNAs into the identified active components. One point of care for anyone citing this paper: the antagomirs were delivered with a lipid transfection reagent, so it is not a self-delivery citation, and no AUM material went into an animal or into the RAW cells.