Role of heterosynaptic plasticity in the modification of sensory responses of mouse visual cortex neurons

Abstract

Synaptic plasticity is a critical factor in neural network function during development, perception, learning, and memory. However, current ideas on the role of synaptic plasticity in cortical network mechanisms are mainly correlative because cellular and molecular mechanisms are predominantly studied in reduced preparations. Currently, the majority of research on synaptic plasticity mechanisms is focused on studying homosynaptic (associative, Hebbian) plasticity. This type of plasticity involves modifying the same synapses that are directly involved in the induction process. However, heterosynaptic plasticity, which occurs in synapses that were inactive during induction, plays a crucial role in the function of neural networks, in addition to the more extensively researched homosynaptic plasticity [1, 2]. In this study, we examined how heterosynaptic plasticity, triggered by intracellular tetanization of pyramidal neurons in the visual cortex of mice, affects their response to visual stimulation in vivo. In the initial stage of the experiment, we recorded intracellularly the visual cortex neurons of anesthetized mice by means of the whole-cell patch clamp approach. We employed an intracellular tetanization procedure to assess the effect of heterosynaptic plasticity on visual responses. Specifically, we applied ten bursts of 5 action potentials each second at a frequency of 100 Hz during the tetanization process in the recorded neuron, repeating the procedure five times at 60-second intervals. Previous studies in brain slices have shown that this protocol leads to substantial plastic changes in the synaptic inputs of a given neuron, including potentiation, depression, and no change (see review [3]). As visual stimuli, we used vertical and horizontal bars moving in opposite directions on the computer screen. A small hyperpolarizing current was continuously applied to the cell to prevent the generation of action potentials. As a result, the changes in the cell membrane potential represented the responses to the visual stimuli. Intracellular tetanization led to a noteworthy amplification in the amplitude and area of the response to the optimal stimulus, which was based on the orientation and movement direction. The reaction to other stimuli, on the other hand, did not encounter any substantial change. A consequence of this was an elevation in the simplified index of directional selectivity of tetanized neurons, computed as the ratio of the optimal stimulus response amplitude to the response amplitude in the opposite direction (null direction). To minimize the influence of intracellular perfusion, which unavoidably arises during whole-cell patch clamp recordings, and to extend the duration of cell response recording after tetanization, we conducted additional experiments involving extracellular recording of cell activity and optogenetic stimulation through an optical fiber inserted into the recording microelectrode. Two weeks prior to the experiment, the pyramidal neurons in the 2/3 layer of the mouse visual cortex underwent viral transduction to express the fast channel rhodopsin oChiEF. During the experiment, visual responses of neurons were recorded for 15–40 minutes. Subsequently, we induced bursts of action potentials with a frequency of 75 to 100 Hz in the recorded neuron using otptogenetic tetanization and continued to record the visual responses for at least 40 minutes. In this series of experiments, we recorded action potentials induced in neurons by visual stimulation, which was similar to that used in experiments with intracellular recording. Finally, we calculated total post-stimulus histograms from the responses. Intracellular tetanization was found to cause a significant reduction in response amplitude to the optimal stimulus, while responses to stimuli with other orientations and directions of movement remained unaffected. As a result, the index of cell directional selectivity decreased in this series of experiments, indicating a direct opposition to the results obtained through intracellular recording experiments. To clarify this discrepancy, we conducted theoretical simulations using the Leaky Integrate and Fire (LIF) model neuron. We used a model in which orientation is determined by different peak positions of inhibitory and excitatory components of visual responses when moving in optimal and opposite orientations [4]. Simulations were carried out at two different resting potentials: –90 mV, which simulated experiments involving intracellular recordings and injections of hyperpolarizing currents, and –65 mV, which simulated experiments involving extracellular registrations of spiking cell responses in the UP-state mode. Our findings suggest that tetanization causes potentiation of both excitatory and inhibitory components of the responses, leading to the observed situation. During our experiments on visual cortex slices, we discovered that intracellular tetanization of pyramidal neurons in layer 2/3 of the visual cortex causes balanced heterosynaptic changes in excitatory inputs to dendritic regions distant from the soma. This means the net change in all inputs after tetanization equals zero, balancing potentiation and depression. Conversely, it causes unbalanced potentiation of excitatory perisomatic inputs. Furthermore, previous research has demonstrated that the occurrence of high-frequency action potentials in layer 5 pyramidal neurons within the neocortex results in the strengthening of their inhibitory perisomatic inputs originating from adjacent parvalbumin interneurons [5]. Based on the above-cited work, our experiment suggests that intracellular tetanization of the pyramidal neuron in the 2/3 layer of the visual cortex may result in the potentiation of perisomatic inhibitory inputs and the development of simultaneous perisomatic excitatory inputs through heterosynaptic plasticity mechanisms. If the changes in excitatory and inhibitory inputs are balanced, our model experiments show that this can lead to changes in the directional selectivity of the observed cells. Thus, high-frequency spike activity in the absence of specific sensory activation, such as during sleep, may decrease the directional selectivity of visual cortical neurons. This prepares the neurons to adjust their visual responses more finely to new scenes during wakefulness. The potentiation of perisomatic excitatory and inhibitory synaptic inputs, through heterosynaptic plasticity, may be the mechanism responsible for such tuning.