Intercellular Communication and Heterosynaptic Metaplasticity in the Rodent Hippocampus
Long-term potentiation (LTP) and long-term depression (LTD) are two forms of synaptic plasticity thought to underlie learning and memory. LTP and LTD are characterised by a persistent increase or decrease, respectively, in synaptic transmission. Synaptic plasticity is vital for neural health and cognition, however unconstrained plasticity compromises learning and, in the extreme, may be injurious. Mechanisms must therefore be in place to prevent excessive LTP and LTD. Such regulation comes in part through processes termed metaplasticity, whereby neuronal activity influences later plasticity induction. The influential Bienenstock, Cooper and Munro (BCM) model of synaptic plasticity proposes a cell-wide plasticity threshold that shifts as a function of integrated postsynaptic action potential (AP) firing, such that greater AP firing raises the threshold for LTP induction and lowers the threshold for LTD induction. This homeostatic function therefore maintains synaptic plasticity within optimal bounds. Importantly, the BCM model (and more recent derivatives) posits that alterations to the plasticity threshold occur at all synapses on a given cell, whether active in generating APs or not. Recent evidence suggests that such heterosynaptic regulation of plasticity thresholds exists in the hippocampus, a brain region thought to be involved in certain forms of memory formation. In keeping with BCM predictions, strong ‘priming’ activity delivered to synapses in hippocampal CA1 inhibits subsequent LTP and facilitates subsequent LTD at a neighbouring pathway quiescent during priming. This metaplastic state is even able to spread heterodendritically, from the basal dendrites of stratum oriens (SO) to the apical dendrites of stratum radiatum (SR). The principal aim of this thesis was to expand upon the previous findings of heterosynaptic and heterodendritic metaplasticity in CA1, and in particular to determine the mode of long-distance communication which allows synaptic innervation to alter plasticity thresholds at sites hundreds of microns away. To this end, extracellular and intracellular electrophysiological experiments were conducted in acute hippocampal slices taken from rats or mice. Various patterns of priming stimulation were delivered to synapses in SO or SR, prior to testing LTP/LTD induction in a second pathway in either stratum. Inhibition of LTP in SR could be seen following several paradigms of repeated 100 Hz priming stimulation delivered to SR or SO. Further supporting the robustness of the effect, the metaplasticity was seen at two different ages and in two different rodent species. However, priming with theta-burst stimulation did not alter subsequent plasticity. Further, repeated 100 Hz priming delivered to SR did not alter subsequent LTP in SO, which contrasts with the cell-wide metaplasticity predicted by the BCM model. To test the importance of postsynaptic depolarization in generating the metaplasticity, individual pyramidal cells were hyperpolarized to -90 mV during SO priming via somatic injection of negative current. This procedure, delivered via sharp electrodes, abolished AP firing during priming and maintained the somatic membrane potential below resting values, thus blocking the spread of depolarization from SO to SR. However, priming under these conditions still triggered a reduction in later LTP in SR. Thus, the BCM prediction that cell-firing drives alterations to the plasticity threshold was not applicable to the current model. Certain derivatives of the BCM model implement altered membrane properties as an expression mechanism for altering LTP/LTD thresholds. However, no alterations were seen in the magnitude of the medium or slow afterhyperpolarisation, the h current or input resistance following priming. Instead, priming was found to require the activation of M1-acetylcholine receptors and release of Ca2+ from intracellular stores. To test the possibility that intercellular communication is required for the long-distance spread of the metaplasticity, pharmacological methods were utilised to target two common intercellular modes of communication; purinergic signalling and gap junctions. Heterosynaptic metaplasticity was found to require hydrolysis of extracellular ATP to adenosine, and activation of A2, but not A1 adenosine receptors. Further, heterosynaptic metaplasticity was blocked when priming occurred in the presence of the non-selective gap junction blockers carbenoxolone and meclofenamic acid, and by a connexin43-specific mimetic peptide. This latter result provides strong support that astrocytes are involved in the long-distance communication between dendritic layers. A model of neuron-astrocyte-neuron signalling is proposed as an explanation of these results. Possible function(s) of the heterosynaptic metaplasticity are considered. The departures from the predictions of the BCM model suggest that this form of metaplasticity is not a graded, homeostatic response to postsynaptic activity in single cells. Rather, heterosynaptic metaplasticity as described in this thesis appears well placed to regulate activity at the network level, and may serve as a neuroprotective mechanism.
Advisor: Abraham, Wickliffe Carson
Degree Name: Doctor of Philosophy
Degree Discipline: Psychology
Publisher: University of Otago
Keywords: Metaplasticity; Long-term potentiation; Synaptic plasticity; Intercellular communication; Hippocampus
Research Type: Thesis