ELUCIDATING THE NETWORK MECHANISM OF MULTISENSORY INTEGRATION DEFICITS IN A MOUSE MODEL OF TAUOPATHY

Alzheimer’s disease (AD) is characterized by the accumulation of several pathophysiological markers, one of which is hyperphosphorylated tau, contributing to synaptic loss and cognitive decline. Prior to the onset of cognitive decline, the senses also deteriorate until all are compromised. This affects multisensory integration (MSI), the process of combining inputs from different sensory modalities to guide behavior, which is preserved in healthy aging but disrupted in AD. Recently, a number of therapeutic approaches utilizing multisensory stimulation have been reported, however, what remains to be understood is the mechanism by which tau pathology disrupts the temporal coordination and large-scale network interactions required for multisensory integration. Here, we tested rTg4510 mice in a spatial choice task involving visual, auditory, and audiovisual cues. Bayesian modeling and neural network modeling is used to determine how sensory information was weighted and integrated during decision making. Tauopathy mice showed reduced accuracy in sensory-guided choices and abnormal cue weighting, consistent with impaired multisensory computation. Notably, female tauopathy mice assigned lower reliability to visual cues, suggesting early sensory reweighting associated with tau-induced circuit dysfunction. Meanwhile, male tauopathy mice yielded more heterogeneous responses, suggesting that not every animal integrates cues effectively. We propose that tauopathy alters neural computations supporting MSI, and our results highlight modality- and sex-specific changes in sensory processing. Ongoing work incorporating real-time imaging during head-fixed tasks will further clarify the neural dynamics underlying MSI impairment in AD to identify early behavioral and computational biomarkers of circuit dysfunction and support the development of circuit-based interventions.