Conjunctive processing of locomotor signals by the ventral tegmental area neuronal population

Abstract

The ventral tegmental area (VTA) plays an essential role in reward and motivation. How the dopamine (DA) and non-DA neurons in the VTA engage in motivation-based locomotor behaviors is not well understood. We recorded activity of putative DA and non-DA neurons simultaneously in the VTA of awake mice engaged in motivated voluntary movements such as wheel running. Our results revealed that VTA non-DA neurons exhibited significant rhythmic activity that was correlated with the animal's running rhythms. Activity of putative DA neurons also correlated with the movement behavior, but to a lesser degree. More importantly, putative DA neurons exhibited significant burst activation at both onset and offset of voluntary movements. These findings suggest that VTA DA and non-DA neurons conjunctively process locomotor-related motivational signals that are associated with movement initiation, maintenance and termination.

Figure 1. Muti-tetrode recording and spike sorting.

(A) Electrode array track shown on an example coronal brain section (top-right) and locations of the electrode array tip from 10 mice on the atlas section diagrams . (B) Left panel, an example of spike sorting using principle component analysis (Plexon OfflineSorter). Here, multiple units were recorded simultaneously from one tetrode; red/purple/blue dots represent isolated units 1–3, respectively; yellow dots represent un-isolated spikes and noises; black dots represent overlapping spike waveforms, which would be manually assigned to units 1–3. Middle and right panels, representative spike waveforms for the units 1–3. Note that only unit 3 was classified as putative DA neuron, while units 1 and 2 were classified as non-DA neurons.

Figure 2. Classification of VTA putative DA and non-DA neurons.

(A) Left panels, peri-event rasters and histograms of two simultaneously recorded VTA neurons (unit 1: putative DA neuron; units 2: non-DA neuron) in response to the conditioned tone (5 kHz, 1 sec) that reliably predicted food delivery. Right panels, cumulative spike activity of the same two neurons before and after (-60-0 and 0–60 min, respectively) the injection of the dopamine receptor agonist apomorphine (1 mg/kg, i.p.). (B) Percentages of classified putative DA (100%; n  =  25) and non-DA neurons (13%; 6/47) that were significantly activated by the conditioned tone that reliably predicted food delivery. (C) Normalized firing rates of putative DA and non-DA neurons after the injection of apomorphine (firing rates averaged for 30 min). Note that 9 out 10 putative DA neurons tested showed significant suppression (≤ 20% baseline firing rates), while the majority of the non-DA neurons (20/22) showed limited or no change of firing rate by apomorphine (1 mg/kg, i.p.).

Figure 3. Rhythmic activity of VTA non-DA neurons during wheel running.

(A) Rasters of two simultaneously recorded VTA non-DA neurons during voluntary wheel running and quiet wakefulness. (B and C) Auto-correlation and power spectral density (PSD) analyses suggest that the same two neurons (as shown in A) showed strong rhythmic activity (at a cycle of 0.31 sec on average) during wheel running.

Figure 4. Activity of the VTA non-DA neuron correlates with the wheel running rhythm.

(A and B) Rasters (upper panels) and smoothed auto-correlation histograms (lower panels) of two simultaneously recorded VTA non-DA neurons during high-speed (A) and low-speed (B) voluntary wheel running. (C) VTA neuron activity correlates with the limb-movement rhythm during wheel running.

Figure 5. Phase-specific firing of VTA non-DA neurons during wheel running.

(A) Rasters of five simultaneously recorded VTA non-DA neurons during voluntary wheel running. (B) Cross-correlation histograms show that the same five neurons (as shown in A) fire preferentially at specific phases of each cycle and in sequence during voluntary wheel running (unit 4 was used as the reference for cross-correlation calculation).

Figure 6. Rhythmic activity of both putative DA and non-DA neurons during wheel running.

(A) Peri-event rasters and histograms of four simultaneously recorded VTA neurons in response to the conditioned tone that reliably predicted food delivery (units 1 & 2: putative DA neurons; units 3 & 4: non-DA neurons). (B) Smoothed cross-correlation histograms of the same four neurons (as shown in A) during voluntary wheel running. Unit 3 was used as the reference for cross-correlation calculations. (C) Smoothed cross-correlation histograms of shuffled units 1 and 2 (randomized spikes). The same unit 3 (as shown in B) was used as the reference for cross-correlation calculations. (D) Correlation coefficient analyses and comparisons between the recoded and shuffled spikes. A set of simulated sine oscillation curves y  =  sin (ax + b) were used as the reference for the correlation coefficient analysis (see Materials and Methods). n = 14 and 29 for DA and non-DA neurons, respectively; *P<0.001, Student's paired t-test. Error bars represent s.d.

Figure 7. Burst activation of the VTA putative DA neuron at wheel running initiation/termination.

(A) Rasters of two simultaneously recorded putative DA (unit 1) and non-DA (unit 2) neurons during voluntary wheel running. Red triangles indicate the start and stop of wheel running, respectively. It was noted that the two neurons were recorded from one tetrode in the VTA. (B) Peri-event rasters and histograms of the same two neurons (as shown in A) referenced to the starts (left panels) and stops (right panels) of wheel running. (C) VTA putative DA neurons exhibit significant higher probability of burst activity at starts and stops of wheel running in compare with the sustained wheel running. (n  =  14; **P < 0.01, ***P < 0.001, Student's paired t-test). Error bars represent s.e.m.