All of our work focuses on pyramidal cells in the CA3 subregion of the hippocampus because of their unique functional and network properties. CA3 is the site where sensory and mnemonic information from the entorhinal cortex, dentate gyrus, and CA3 itself (via extensive recurrent connections) is compared and integrated. Importantly, synaptic plasticity in these cells is required for rapid memory formation during single-trial learning.
To characterize the changes that occur in CA3 pyramidal cells during behavior, we made electrophysiological recordings from individual cells with simultaneous recordings of the LFP during naturally occurring brain states. From this data, we found a profound change in the membrane potential when theta oscillations were present in the LFP, a pattern that occurs during active brain states. Specifically, while the variance in the membrane potential decreased consistently in all cells, some cells experienced a consistent hyperpolarization while others a consistent depolarization. Accordingly, the depolarized cells fired action potentials and can thus be considered active. Since we know that memory formation occurs more readily during the theta state, and we also know that some CA3 cells are active in this process while many others are silent, we concluded that this shift in membrane potential reflects certain cells’ readiness to encode new information and other cells’ suppression in the circuit. This type of mechanism ensures efficient encoding of information with a high signal-to-noise ratio. In contrast, when there was large irregular activity (LIA) in the LFP, which occurs during resting and is associated with memory consolidation, cells tended to depolarize. This is similar to what happens in other cell types in the hippocampus, so we concluded that LIA is a state where the principal cell types in the hippocampus become activated, ready to replay and consolidate the recently encoded information. Thus, these data showed intracellular correlates of different brain states, which we know are associated with different memory processes. This observation is depicted in the figure: certain cells change their activity level (red, blue neurons) depending on behavior (walking, reflecting). We believe that our observed intracellular changes reflect mechanisms that allow individual cells to complete the different memory processes associated with different brain states.
To observe synaptic plasticity in CA3 pyramidal cells in vivo, we again recorded from individual cells. However, instead of observing the effects of the sum total of inputs through changing brain state, we instead measured the effects due to specific inputs. Area CA3 is different from other hippocampal subregions in that pyramidal cells receive excitatory input from other pyramidal cells in the same region. To take advantage of this unique property, we controlled the activity of CA3 pyramidal cells while recording their effect on other cells in the same region. We found that even though there was a detectible excitatory component to the input, a large and rapid inhibition dominated the signal in the postsynaptic cell, highlighting the importance of inhibitory regulation in this circuit. To induce synaptic plasticity, we controlled both the input and the firing of the recorded cell. It is known from in vitro studies that if the input occurs within milliseconds before or after firing, the amplitude of the excitation will increase. However, we found that in in vivo conditions, inhibition showed the greatest increase in amplitude. This contrasts with the results obtained in vitro, and highlights the importance of performing experiments under physiological conditions. It also brings interesting questions about how synaptic plasticity can underlie memory in intact brain circuits.
These results have been published in the form of a masters and PhD thesis, and are also being prepared for publication in a peer-reviewed scientific journal.
