Activity-Dependent Regulation of Inhibition from Different Inhibitory Subtypes
Neuronal activity, in the form of action potential firing, is critical in the maturation and maintenance of neocortical circuitry. A negative feedback mechanism by which neuronal circuits adapt to changing levels of average activity on a time scale of hours to days is known as homeostatic plasticity. At the simplest level, homeostatic adaptations occur to maintain firing rate of neurons at a particular set-point. To better understand homeostatic plasticity at the network level, one must understand the activity-dependent adaptations that occur in the different neocortical cells types. To this end, I examined the regulation of inhibitory neurons and their synapses. I used chronic pharmacological block of activity in a neocortical slice cultures to examine the role activity plays in regulating feedback inhibition defined by two biochemical inhibitory neuron subtypes - parvalbumin-positive (Parv+) and somatostatin-positive (Som+). The cellular and synaptic components of local feedback inhibition were examined. I found that chronic activity blockade caused the following: 1) an increase in the intrinsic excitability of Som+ neurons through the downregulation of 2 substhreshold currents. While not thoroughly examined in Parv+ neurons, a similar, but weaker, increase in excitability may occur in these neurons as well. These< changes are consistent with a homeostatic maintenance of firing rate in these neurons. 2) a differential regulation of monosynaptic inhibition based on subtype that was frequency dependent. At low frequency action potential firing, Parv+ mediated inhibitory drive was downregulated while Som+ was unchanged. Both subtypes were likely downregulated at high frequency firing. 3) an upregulation of excitatory drive onto both Parv+ and Som+ neurons. This was most dramatic at low frequency firing where both subtypes displayed an almost 3-fold increase. This is also consistent with homeostatic maintenance of firing rate in inhibitory neurons. 4) based on the above, a clear change in recurrent inhibition occurred at low frequency firing. First, net recurrent inhibition was increased for both subtypes, but the relative influence of the two changed, such that Som+ recurrent inhibition contributed more relative to that of Parv+ circuitry. At high frequency firing, a slight, but less resolvable, increase in net recurrent inhibition may have occurred in both subtypes without any change in relative contribution. 5) all of the synaptic changes were likely due to increases in presynaptic release probability and/or decreases in synapse number.