AbstractBackgroundArea CA1 of hippocampus is highly vulnerable to Alzheimer’s disease (AD), with denser pathology towards the subiculum versus the CA2 border. Spatial and non‐spatial memory processing in CA1 is also segregated along this transverse axis. Towards CA2, deep and superficial CA1 pyramidal neurons (PNs) uniquely process spatial information from medial entorhinal cortex (MEC) and CA3. Towards subiculum, deep and superficial CA1 PNs differentially integrate non‐spatial information from lateral entorhinal cortex (LEC) and CA3. It is unclear how AD affects these distinct processing architectures, and if it is reflective of the known gradient of AD pathology within CA1.MethodMEC or LEC of 3xTg‐AD (12‐24 month old) and age‐matched wild type (WT) mice were injected with an AAV expressing channelrhodopsin (ChrR2). Acute hippocampal slices were prepared two weeks later. Whole cell patch clamp recordings were performed on deep and superficial CA1 PNs, with MEC or LEC axons optogenetically activated by blue light and CA3 axons activated by extracellular stimulation in stratum radiatum. In separate mice, immunohistochemistry (IHC) was performed for AD pathology.ResultIHC confirmed increased AD pathology in 3xTg‐AD mice towards subiculum versus CA2. Near CA2, 3xTg‐AD mice (compared to WT) showed reduced MEC input responses in both deep (0.7mV vs. 1.8mV, n=7 each, p<0.05) and superficial PNs (0.8mV vs. 2.0mV, n=8 each, p<0.05). In contrast, CA3 input responses were unchanged in deep PNs but was surprisingly increased in superficial PNs (10.4mV vs. 6.6mV, n=8 each, p<0.05). Near subiculum, 3xTg‐AD mice (compared to WT) showed a reduction in LEC input response only in superficial PNs (0.9mV, n=8, vs. 1.7mV, n=7, p<0.05). Response to CA3 input was also selectively reduced in superficial PNs (3.0mV, n=8, vs. 8.2mV, n=7, p<0.01).ConclusionSpatial and non‐spatial memory circuits in CA1 are differentially affected in 3xTg‐AD mice, in a synapse‐ and cell type‐specific manner. Furthermore, they are not merely a direct reflection of AD pathology gradient in CA1, and may be also reflective of indirect or compensatory mechanisms related to the disease. These results also have implications on dysfunctional in vivo activity of CA1 PNs related to AD‐induced memory deficits.