A defined network of fast-spiking interneurons in orbitofrontal cortex: responses to behavioral contingencies and ketamine administration

Abstract

Orbitofrontal cortex (OFC) is a region of prefrontal cortex implicated in the motivational control of behavior and in related abnormalities seen in psychosis and depression. It has been hypothesized that a critical mechanism in these disorders is the dysfunction of GABAergic interneurons that normally regulate prefrontal information processing. Here, we studied a subclass of interneurons isolated in rat OFC using extracellular waveform and spike train analysis. During performance of a goal-directed behavioral task, the firing of this class of putative fast-spiking (FS) interneurons showed robust temporal correlations indicative of a functionally coherent network. FS cell activity also co-varied with behavioral response latency, a key indicator of motivational state. Systemic administration of ketamine, a drug that can mimic psychosis, preferentially inhibited this cell class. Together, these results support the idea that OFC-FS interneurons form a critical link in the regulation of motivation by prefrontal circuits during normal and abnormal brain and behavioral states.

Keywords: depression; fast-spiking interneuron; motivation; parvalbumin; psychosis; schizophrenia.

Figure 1

Identification of putative interneurons. (A) Average firing rate of a neuron plotted as a function of the cell's average spike width yielding two distinct populations of cortical cells corresponding to putative narrow-spiking (NS) interneurons (blue) and wide-spiking pyramidal cells (P, gray). Example of average waveforms (upward deflections indicate inward current) for a single putative interneuron (blue) and pyramidal cell (gray) are shown as an inset with scale bar. Distributions of firing rates and spike widths are shown to the left and bottom of the scatter plot respectively. (B) Dendrogram depicting the clustering of cortical neurons into two groups corresponding to narrow-spiking (NS, blue) and wide-spiking (WS, gray) cells. Scale bar indicates average Euclidian distance between sub-groups.

Figure 2

Physiological classification of narrow-spiking interneurons. (A) Spike shape features extracted from an average spike waveform: Peak (P), Width (W), pre-peak valley (V1), post-peak valley (V2). (B) Dendrogram depicting the clustering of narrow-spiking cells based on physiological similarities and differences. Two main clusters are evident: a relatively homogeneous cell population of narrow-spike type 1 cells (NS1; red) and a more heterogeneous cluster of narrow-spike type 2 cells (NS2; black). Scale bar indicates Euclidean distance between sub-groups.

Figure 3

Physiological distinct properties for two populations of narrow-spiking neurons. (A) Average peak-normalized waveforms for NS1 (red) and NS2 (black) cell classes (upward deflections indicate depolarization). Cumulative distributions of activity dependent properties for NS1 (red) and NS2 (black) neurons. (B) Log firing rate, (C) CV2, (D) fractional spike broadening (SB).

Figure 4

Cell-type specific local network interactions. (A) Representative cross-correlation histogram (left) for a single NS1–WS cell pair. Firing probability of wide-spiking cell, measured as number of standard deviations from baseline, is plotted relative to the occurrence of spikes from the NS1 neuron. High pass filtering of the raw CCH (right) reveals a significant increase in the firing probability of the putative pyramidal cell just prior to a NS1 spike. Note the change in time axis. (B) Representative cross-correlation histogram (left) for a single NS2–WS cell pair. Firing probability of wide-spiking cell is plotted relative to the occurrence of spikes from the NS2 neuron (time = 0). High pass filtering of the CCH (right) reveals a lack of short latency spike train interactions.

Figure 5

Cell-type specific interneuronal network interactions. (A) Average CCH for pairs of simultaneously recorded NS1 neurons showing a significant increase in correlated firing [x-axis: time lag of correlation (ms); y-axis: correlation value relative to baseline]. (B,C) Flat CCHs for populations NS2–NS2 and NS1–NS2, respectively, indicate a lack of synchronous firing. (D) Percentage of cell pairs exhibiting positive (+), negative (−), mixed (+/−), or flat cross-correlation histograms. Criteria for significance was at least one bin within ±5 ms of the zero bin exceeding three standard deviations in either the positive or negative direction.

Figure 6

Activity profiles of NS1 cells during behavior. (A) Raster plot for a single NS1 neuron time-locked to an animal's exit from the odor port. Each line of dots corresponds to a single trial. (B) Post-stimulus time histogram (PSTH) for NS1 neuron in (A), time-locked to an animal's withdrawal from odor port. Red trace reflects average firing rate across all trials for this cell; horizontal line indicates baseline-firing rate of cell. (C) Average peak-normalized PSTHs for populations of NS1 (red) and NS2 (black) neurons time-locked to animals’ withdrawal from odor port. Shading represents ±1 SEM. X-axis: time lag relative to withdrawal from odor port; y-axis: average peak-normalized firing rate. (D) Trial-by-trial variations firing rate (y-axis) for cell in (A) plotted as a function of movement time. (E) Distributions of movement time-firing rate correlations for all NS1 neurons. Movement time was calculated from the time in which the animal left odor port until time in which it entered reward port. Firing rate was calculated over the last 300 ms of the movement time. Arrow indicates the correlation value for the cell shown in (A,B,D).

Figure 7

Ketamine selectively suppresses the spiking activity of NS1 neurons. (A) Mean ± SEM response of NS1 neurons following subcutaneous of either saline (black) or 30 mg/kg ketamine as function of time relative to injection (indicated by arrow). Firing rates calculated in 3-min bins. (B) Fractional change in firing rate relative to baseline for three cell populations (WS, NS1, and NS2) following either saline (black) or ketamine (blue) administration. Bar graphs reflect mean ± SEM fractional firing rate change calculated by dividing, for each cell, mean firing rate 2–17 min post-injection by the cell's mean firing rate during the last 15 min of the baseline session.

Figure 8

Proposed role of OFC interneurons in regulation of behavior. (A) The activity of OFC–FS type interneurons is negatively correlated with level of arousal or motivation, as indicated by correlations between NS1 firing and response latency. (B) An intermediate level of activity of OFC–FS interneurons is expected to be associated with optimal behavioral functionality, with too low levels of interneuron leading to symptoms associated with hyperarousal (e.g., mania, positive symptoms) and too high levels leading to symptoms of hypoarousal (e.g. depression, negative symptoms).