Insulin resistance induces earlier initiation of cognitive dysfunction mediated by cholinergic deregulation in a mouse model of Alzheimer's disease

**Background:** Alzheimer's disease (AD) is a neurodegenerative disorder characterized by cognitive decline and amyloid β (Aβ) deposition. Type 2 diabetes mellitus (T2DM) doubles the risk of AD, and insulin resistance is considered a key contributor. However, the mechanisms by which insulin resistance exacerbates AD pathology, particularly in the preclinical stage, are not fully understood. This study investigates the role of insulin resistance in accelerating cognitive dysfunction using a novel double-knock-in mouse model (APP/IR-dKI) that combines AD pathology (APP-KI mice with Aβ accumulation) with insulin resistance (IR-KI mice with P1195L mutation in the insulin receptor) without persistent hyperglycemia.

**Methods:** Male APP-KI and APP/IR-dKI mice were used. Metabolic tests (insulin tolerance test, glucose tolerance test, serum insulin levels, and blood glucose monitoring) were performed at 3 months of age. Cognitive function was assessed at 4 months using the novel object recognition test (with and without the nAChRα7 antagonist MLA), novel place preference test, three-arm maze test, and eight-arm maze test. Cerebral blood flow (CBF) responses to electrical stimulation of the ulnar nerve were measured using laser speckle imaging under spinalization, with pharmacological blockade of muscarinic (atropine) and nicotinic α7 (MLA) receptors. Cortical gene expression of neuronal activity markers (Egr1, Nptx2, Creb) and cholinergic system components (Chrna7, Tmem35a, Ric3, Chrnm1, Chrnb2, Slc5a7, Slc33a1) was analyzed by quantitative RT-PCR. Aβ burden was assessed by immunohistochemistry and ELISA for Aβ40 and Aβ42 in soluble and insoluble fractions.

**Key Results:** APP/IR-dKI mice exhibited insulin resistance (ITT: p=0.0174 at 10 min, p=0.0047 at 30 min, p=0.0033 at 60 min; GTT: p=0.0350 at 30 min, p=0.0281 at 60 min), hyperinsulinemia (p=0.0007), and larger glucose fluctuations (maximum levels p=0.0263; gap p=0.0174) without persistent hyperglycemia (fasting glucose p=0.182). In the novel object recognition test, APP/IR-dKI mice showed significantly lower exploratory preference than APP-KI mice (p=0.0046). MLA administration reduced preference in APP-KI mice (p=0.0002) but not in APP/IR-dKI mice (p>0.9999). Novel place preference was also lower in APP/IR-dKI mice (p=0.0006). No differences were found in working memory (three-arm maze p=0.699; eight-arm maze p=0.646). CBF responses to ulnar nerve stimulation were similar between genotypes under spinalization (p>0.05 at all time points). Atropine reduced CBF response in both APP-KI (p=0.0476) and APP/IR-dKI mice (p=0.0108), but MLA reduced CBF only in APP-KI mice (p=0.0431) and not in APP/IR-dKI mice (p=0.311). Cortical gene expression showed reduced Egr1 (p=0.0041), Nptx2 (p=0.0170), Chrna7 (p=0.0120), and Slc5a7 (p=0.0329) in APP/IR-dKI mice, with no change in Creb (p=0.458), Tmem35a (p=0.0877), Ric3 (p=0.0670), Chrnm1 (p=0.984), Chrnb2 (p=0.0856), or Slc33a1 (p=0.714). Aβ burden did not differ between genotypes (hippocampal deposition p=0.722; Aβ40 soluble p=0.704, insoluble p=0.840; Aβ42 soluble p=0.439, insoluble p=0.773).

**Clinical Implications:** This study demonstrates that insulin resistance, independent of hyperglycemia, accelerates cognitive dysfunction in a preclinical AD model through cholinergic deregulation, specifically via downregulation of nAChRα7. The findings suggest that nAChRα7 dysfunction mediates early cognitive decline and altered neurovascular coupling, without accelerating Aβ accumulation. This highlights the importance of targeting insulin resistance and cholinergic pathways, particularly nAChRα7, as potential therapeutic strategies to prevent or delay AD onset in patients with T2DM. The APP/IR-dKI mouse model provides a valuable tool for studying AD exacerbation by insulin resistance and developing interventions for high-risk populations.