UC Irvine

The hippocampus has long been known to play a central role in various behavioral and cognitive functions. Despite extensive behavioral, anatomical studies postulating roles of the hippocampus and its sub-regions, the underlying dynamics of the neural circuits responsible for region-specific roles in the hippocampus remains poorly understood. In addition, the communication between the hippocampal sub-regions also plays a major role in learning and memory but has not been explored as much due to the lack of access to the individual axons. This paper seeks to understand the role of slow waves (4-100 Hz) known as Local Field Potentials (LFP) in the hippocampal sub-regions, their dynamics and role in information transmission to bind cell assemblies through synchrony by timed-bursts and individual spikes, by space (EC, DG, CA3, CA1) in the hippocampal sub-regions and by frequency (4-11 Hz theta and 30-100 Hz gamma bands). We used a four-chambered polydimethysiloxane (PDMS) micro-tunnel device over a multi-electrode array (MEA) with the engineered living reconstruction of the hippocampal sub-regions. The 5 x 10 μm micro-tunnels between each chamber allow axonal growth and inhibit the migration of 15 μm diameter cell bodies or traversal by dendrites providing insight into the role of axons in inter-regional communication. A significant advantage of our system is the ability to separate activity in the axons in the tunnels from the computations of the neural cells in the wells of the 4-chambered device. Functional connectivity within hippocampal sub-regions and the axonal tunnels helped in ascertaining the role of LFP in information processing and transmission. A computationally simple and linear method, Cross-Correlation of the theta power and the cross correlation of the gamma power was used to estimate synchrony of LFP. Comparison between the theta-theta and gamma-gamma LPF’s from the different sub-regions indicates that the CA3 has the strongest correlation. CA3 sub-region is characterized by recurrent network wiring pattern and such strong correlation can be associated with this kind of network pattern. The spatial distribution of the theta and gamma correlation between electrodes was estimated by computing the correlation as a function of the inter-electrode distance within each sub-region. A common relationship of LFP power with the inverse square of distance during bursting events suggested LFP transmission through synapses of branching dendritic trees in the plane of the network. These first measures of LFP’s in axons suggests axonal transmission of slow waves, independent of spikes and synapses. Understanding these network properties within the hippocampus could provide important information for steps in memory formation, neuromorphic computing and artificial intelligence systems in addition to better detection of hippocampal diseases such as epilepsy and Alzheimer’s disease.