Alzheimer’s disease (AD) causes progressive loss of memory, for which there is no cure. A growing understanding of memory impairment in AD suggests that alterations at the genetic and cellular levels contribute to circuit dysfunction. By restoring these vulnerable circuits and networks during early stages of the disease, it might be possible to reverse memory loss and slow disease progression. Recent studies have shown that GCs receive abundant subcortical inputs from the lateral supramammillary nucleus (SuM) of the hypothalamus, and the SuM inputs play a key role in modulating the strength of the EC inputs onto GCs. Importantly, we found that SuM is spared of Aβ deposition even in late-stage 5xFAD mice and exhibits intact anatomical inputs to GCs in early 5xFAD mice, thus placing SuM in an ideal position to modulate impaired EC-DG connections in AD. In addition, we found that stimulating SuM neurons or SuM-DG pathway is sufficient to increase GC activity and improve spatial memory performance during the NPR test through SuM glutamate release, suggesting that increased SuM glutamate transmission is beneficial to spatial memory performance. Furthermore, using an in vivo multi-fiber photometry recording system, we found highly correlated Ca2+ activities between SuM neurons and DG GCs during spatial memory retrieval in the NPR test, which is disrupted in early 5XFAD mice, suggesting that synchronized SuM-DG activity is critical for spatial memory retrieval. Based on these findings, we propose the following three aims to evaluate whether stimulation of SuM-DG activity can restore memory and slow disease progression. In Aim 1, we will determine whether stimulating SuM-DG activity in early AD mice restores EC-GC connections and DG activity; In Aim 2, we will determine whether stimulating SuM-DG activity restores interregional synchrony during spatial memory and improves memory performance in early AD mice; and In Aim 3, we will evaluate whether chronic stimulation of SuM neurons from the prodromal stage slows AD progression. Our proposed experiments will advance our understanding of the key neural circuits that modulate vulnerable networks during early AD pathology. Results from these studies will guide new therapeutic directions by selectively targeting these modulatory neurocircuits for treating memory loss associated with AD.