1518-P: HMGA1 Deficiency—A Common Causal Link between Tau Pathology and Insulin Resistance?

Consistent evidence supports a link between neurodegeneration, insulin resistance (IR) and type 2 diabetes (T2D). The molecular mechanisms for this association are unclear. It has been previously shown that the transacting factor HMGA1 is needed for proper expression of the insulin receptor. Also, IR and T2D were reported earlier in humans and mice with HMGA1 gene defects. Here, we focused on HMGA1 as a potential novel pathogenic factor of Tauopathies, including Alzheimer’s disease. Investigations were performed in vitro, in human and mouse neuronal cell models, and in vivo, in Hmga1-/- mice. Immunoblot analyses, combined with protein-DNA interaction studies, ChIP-qPCR, and reporter gene assays in living cells were carried out, in addition to Tau immunohistochemistry in Hmga1-/- mouse brain. Also, a case-control study was used to explore the relationship between HMGA1 and Tau pathology in Tauopathy patients, carrying or not the rs146052672 variant, which negatively impacts HMGA1 protein production and increases the risk of IR and T2D. We show that HMGA1 plays a role in controlling Tau protein expression. This function was mediated primarily through the specific repression of MAPT gene transcription. Thus, Tau mRNA and protein levels were inversely related to HMGA1 in both human neuronal cells and primary mouse neurons. This inverse relationship was confirmed in the brain of Hmga1-/- mice, in which Tau was overexpressed, compared to Hmga1+/+ mice. Furthermore, genetic analysis indicated that the rs146052672 variant was more common among patients with Tauopathy (17%) than among controls (5%). Adjusted logistic regression analysis revealed a strong association of this variant with Tauopathy (OR, 5.56 [95% CI, 2.08-14.87]; P = 0.001). HMGA1 is a repressor of the MAPT gene. As such, HMGA1 deficiency may be a causative factor of Tauopathy. We hypothesize that a shared causal link between Tau pathology and insulin resistance is plausible, that is, via downregulation of HMGA1 function. Disclosure M. Mirabelli: None. E.D. Chiefari: None. F. Brunetti: None. D. Foti: None. A. Brunetti: None.