Intranasal Insulin Reverses Memory Loss and Brain Inflammation in Aging Mouse Model

Age-related cognitive decline is closely tied to the loss of specific inhibitory neurons in the hippocampus. The dentate gyrus hilus contains somatostatin-positive (Sst+) GABAergic interneurons that decline naturally with aging, and their loss has been causally linked to memory impairment. This study from the University of Illinois Urbana-Champaign built on prior work by using a viral-genetic approach to selectively ablate these interneurons in young adult mice (3–5 months old), creating what the authors call 'pseudo-aged' mice — animals that recapitulate key molecular and behavioral features of hippocampal aging without the confounds of systemic aging.

The model was generated by injecting AAV5-EF1α-mCherry-flex-dtA bilaterally into the dentate hilus of Sst-IRES-Cre mice. This Cre-dependent diphtheria toxin construct selectively kills Sst+ neurons expressing Cre recombinase. After a 3-week recovery period to allow viral expression and interneuron loss, mice received intranasal insulin (2 IU/day in 20 µL saline) or vehicle for 9 consecutive days. Four groups were studied: controls with saline, pseudo-aged with saline (D/A + Veh), pseudo-aged with insulin (D/A + INS), and healthy mice given insulin alone (INS per se).

Behavioral outcomes were assessed across three validated paradigms. In the Y-maze spontaneous alternation test, pseudo-aged vehicle mice showed significantly impaired spatial working memory compared to controls, while D/A + INS mice performed comparably to controls, indicating full rescue of this deficit. In the novel object recognition test, pseudo-aged vehicle mice showed a recognition index near chance (50%), reflecting failure to distinguish novel from familiar objects, while INS-treated pseudo-aged mice showed significantly higher recognition indices. In trace fear conditioning — a hippocampus-dependent associative memory task — pseudo-aged vehicle mice displayed markedly reduced freezing during the trace interval and tone test, and INS treatment significantly restored freezing behavior. Importantly, insulin had no significant effect on any memory measure in healthy control mice, suggesting the cognitive benefits are state-dependent.

At the molecular level, the study examined three key markers in hippocampal tissue. Iba-1, a marker of microglial activation, was significantly elevated in pseudo-aged vehicle mice and was reduced back toward control levels by INS treatment. Phosphorylated TBK1 (pTBK1 at Ser172), a downstream effector of the cGAS-STING innate immune pathway, was similarly elevated in pseudo-aged mice and normalized by INS — providing the first evidence linking Sst+ interneuron loss to cGAS-STING pathway activation. BDNF protein levels, which were reduced in pseudo-aged vehicle mice, were restored to control levels following INS treatment. These three molecular changes collectively point to a neuroinflammatory and neurotrophic mechanism underlying both the cognitive deficits and their rescue.

The findings are significant for several reasons. They establish that intranasal insulin can reverse established cognitive deficits in a mechanistically defined aging model within just 9 days — a remarkably short treatment window. The suppression of the cGAS-STING pathway is particularly noteworthy, as this pathway has emerged as a major driver of chronic neuroinflammation in aging and neurodegeneration. The restoration of BDNF adds a neurotrophic dimension to insulin's mechanism of action in the brain. While the study is preclinical and uses a partial rather than complete interneuron ablation model, the results align with human clinical data showing intranasal insulin improves memory in Alzheimer's patients, strengthening the translational rationale for further investigation.