The hippocampus is critical for the acquisition and retrieval of episodic memory. Sensory information from cortical regions is conveyed to the hippocampus primarily via the lateral and medial perforant path (LPP, MPP) projections originating in the lateral and medial entorhinal cortex (LEC, MEC) respectively. The LPP projection has been shown to be involved in the transfer of information regarding cue identity, while the MPP conveys spatial information (Hargreaves et al., 2005; Reagh and Yassa, 2014). Within the hippocampus, the LPP sends its projections to both the dentate gyrus (DG) and field CA3 with evidence suggesting that these terminals can be extensions from the same axon (Yeckel and Berger, 1990; Tamamaki and Nojyo, 1993). However, comparisons between the two LPP branches, with regard to their release dynamics and synaptic plasticity mechanisms, are scarce and will thus be the focus of this dissertation. The manner in which the two LPP branches transform (i.e., amplify, filter) incoming signals to the hippocampus at behaviorally- relevant frequencies and patterns is relatively unknown. Multiple factors regulate frequency-dependent signal transformations (e.g. vesicle release dynamics) that can result in markedly different responses to similar inputs within the same network (Trieu et al., 2015). Chapter 1 characterizes the signal transformations at the LPP-DG branch located at the head-stage of the circuit. These studies demonstrate that the LPP-DG synapse operates as a low-pass filter, where responses to repetitive activation at ≥50 Hz (i.e., frequency) display within-train suppression. This operation appears to be governed by factors located in the presynaptic terminal associated with release probability and vesicle recycling. Indeed, the induction of a unique form of presynaptic long-term potentiation (LTP) at LPP-DG terminals increases the release probability and, as a result, enhances the low-pass filtering at this terminal.
Given that the two LPP branches originate from the same axon, it would appear plausible that the unique release dynamics and LTP mechanisms evident at the LPP-DG terminal are also present at those synapses innervating CA3 pyramidal cells. However, as described in Chapter 2, recordings from the distal apical dendrites of field CA3 reveal that responses to repetitive stimulation are radically different than those seen in the DG. Within this node, 50 Hz frequency stimulation results in a frequency following as opposed to a depression, with lower frequencies acting similarly. These differences in frequency facilitation at the two LPP terminals suggest distinctions in vesicle release and recycling at the two sites. Furthermore, while LTP is expressed at presynaptic sites at the LPP-DG synapse, a postsynaptic variant of LTP that does not require endocannabinoid signaling is evident at LPP-CA3 terminals. These observations support a target-cell specification hypothesis for the production of LTP variants and suggest the preferential routing to, and encoding by, CA3.
Finally, to emphasize the mechanistic differences in plasticity at these two terminals, a separate study, presented in Chapter 3 was conducted in which the effects that microglia had on LTP was assessed. Removal of microglia with the Colony Stimulating Factor-1 antagonist, PLX5622, had no effects on baseline synaptic responses at either LPP termination site (i.e., LPP-DG or LPP-CA3). However, when LTP was evaluated, mice that underwent PLX treatment had severe impairments at LPP-DG but not LPP-CA3 synapses. Although, the levels of the endocannabinoid, 2-AG, were elevated four-fold in the DG, the prevention of its breakdown fully restored LTP at these synapses. These results demonstrate that disruption of endocannabinoid signaling following microglia loss impairs LPP-DG LTP with no effect upon plasticity at LPP-CA3 terminals.