Distinct tonic and phasic anticipatory activity in lateral habenula and dopamine neurons - PubMed (original) (raw)

Distinct tonic and phasic anticipatory activity in lateral habenula and dopamine neurons

Ethan S Bromberg-Martin et al. Neuron. 2010.

Abstract

Dopamine has a crucial role in anticipation of motivational events. To investigate the underlying mechanisms of this process, we analyzed the activity of dopamine neurons and one of their major sources of input, neurons in the lateral habenula, while animals anticipated upcoming behavioral tasks. We found that lateral habenula and dopamine neurons anticipated tasks in two distinct manners. First, neurons encoded the timing distribution of upcoming tasks through gradual changes in their tonic activity. This tonic signal encoded rewarding tasks in preference to punishing tasks and was correlated with classic phasic coding of motivational value. Second, neurons transmitted a phasic signal marking the time when a task began. This phasic signal encoded rewarding and punishing tasks in similar manners, as though reflecting motivational salience. Our data suggest that the habenula-dopamine pathway motivates anticipation through a combination of tonic reward-related and phasic salience-related signals.

PubMed Disclaimer

Figures

Figure 1

Figure 1. Lateral habenula and dopamine neurons tonically encode variable intervals before rewarding tasks

(A) Events during the inter-trial interval (ITI) of the reward-biased saccade task. During the ITI the screen was blank. After a randomized 2.2–3.2 second delay, a cue appeared marking the start of the next trial. (B,C) Average firing rate of lateral habenula neurons (B) and dopamine neurons (C) aligned on the start of the ITI (left) and the onset of the trial start cue (right). The light gray line indicates baseline firing rate. The yellow shaded area indicates the deviation from baseline firing rate. See also Figure S1.

Figure 2

Figure 2. Prevalence of tonic activity in single neurons

(A) Spike activity of a lateral habenula neuron with strong tonic activity during the ITI. Each row is a trial and each dot is a spike. Trials are sorted by the time that the trial start cue appeared. Top: shortest 30 ITIs; bottom: longest 30 ITIs. Arrows mark the earliest and latest trial start times (gray arrow). (B) Average firing rate of the lateral habenula neuron during the ITI. Black histogram: ITI firing rate in 40 ms bins. Red dots: fitted start and end firing rates. (C) Left: Fitted start firing rate (x-axis) and end firing rate (y-axis) for each lateral habenula neuron. Right: Histogram of changes in firing rate, (end rate − start rate). Text indicates the mean change in firing rate across the population. Asterisk indicates statistical significance (p < 0.05, t-test). (D–F) Same as A–C, for dopamine neurons. See also Figure S2.

Figure 3

Figure 3. Lateral habenula and dopamine neurons encode the temporal distribution of rewarding tasks

(A–D) Average activity of lateral habenula neurons during tasks with a constant ITI (A) and variable ITIs (B–D). Text indicates the range of ITIs. Arrows mark the first possible time the trial could start. Black lines are mean baseline-subtracted firing rate in non-overlapping bins. The bin width for each plot was adjusted based on the range of trial start times during the ITI and the number of recorded neurons. The bin widths for A–D were 150 ms, 150 ms, 200 ms, and 250 ms. Error bars are ± 1 SE. Data in (D) are combined from ITI distributions of 3.1–6.1 s and 3.1–7.1 s, and the last error bar in (C) is cropped above. Red lines indicate a linear least-squares fit to the plotted data points and red text indicates the linear correlation. All correlations were significant (p < 0.005, permutation test). (E–H) Same format as (A–D) for dopamine neurons. The bin widths for E–H were 150 ms, 150 ms, 150 ms, and 250 ms. See also Figure S3.

Figure 4

Figure 4. Tonic activity preferentially encodes rewarding tasks

(A) Pavlovian conditioning procedure. In the appetitive block, a conditioned stimulus (CS) predicted juice rewards (US). In the aversive block, CSs predicted airpuffs. Different CSs indicated different outcome probabilities. On a small number of ITIs an uncued ‘free’ outcome was delivered, either a reward during the appetitive block or an airpuff during the aversive block. (B,C) Average neural ITI activity during the Pavlovian procedure, plotted separately for the appetitive block (red) and aversive block (blue). Same format as Figure 3. All correlations are significant (p < 0.02, permutation test). (D,E) Changing intensity of tonic ITI ramping activity over the course of the aversive block (blue) and appetitive block (red). Activity is plotted for the 1st-4th quarters of each block, defined as trial numbers 2–12, 13–22, 23–32, and 33–42. For each quarter of each block we fit each neuron’s ITI activity using a ramp function (as in Figure 2), and calculated the neuron’s tonic effect as the difference (end firing rate) − (start firing rate). Each data point is the mean of the single-neuron tonic effects. Error bars are ± 1 SE. Symbols indicate statistical significance and trends (+ indicates p < 0.10, * indicates p < 0.05, ** indicates p < 0.01). See also Figure S4.

Figure 5

Figure 5. Timecourse of tonic activity in multiple types of dopamine neurons

(A,B) Left: Tonic ITI activity in multiple types of dopamine neurons with different responses to aversive events. Neurons were sorted into types based on their excitatory and inhibitory responses to airpuff CSs and USs. Same format as Figure 4E, except that due to the relatively small number of neurons for each type, activity in each block was analyzed using two bins representing the first and second halves of each block (trials 2–22 and 23–42). Right: Phasic responses of each neuron type to reward and airpuff CSs and USs. (C) Classification of dopamine neuron types based on responses to aversive cues (x-axis, response to 100% Airpuff CS) and aversive outcomes (y-axis, response to Free Airpuff US). Responses were defined as the firing rate in a window after event onset minus the rate in a window before event onset (Methods). Dots represent neurons and colors represent types of neurons: cells inhibited by the CS and US (“Inhibited”, brown), excited by the CS and US (“Excited”, blue), excited by the CS and inhibited by the US (“Mixed”, black), and non-significantly responsive (“Non-sig”, gray). Open circles indicate two neurons that had a rare mixed pattern of inhibition by the CS and excitation by the US. See also Figure S5.

Figure 6

Figure 6. Phasic trial start activity encodes both rewarding and aversive tasks

(A,B) Population average activity in response to the trial start cue during the appetitive block (red) and aversive block (blue), separately for lateral habenula neurons (A) and dopamine neurons (B). Activity is baseline-subtracted. Shaded region indicates ± 1 SE. Neurons had similar responses in both appetitive and aversive blocks. (C,D) Comparison between response to the trial start cue during the appetitive block (x-axis) and aversive block (y-axis). The response is the rate difference between a post-cue window (gray bar below x-axis) and a pre-cue window (250 ms before the cue). Each dot is a single neuron. Colors indicate neurons with responses significantly different from zero during the appetitive block (red), aversive block (blue), or both (purple) (p < 0.05, signed-rank test). Text indicates the rank correlation and its p-value (permutation test). See also Figure S8.

Figure 7

Figure 7. Trial start activity encodes rewarding and aversive tasks in all neuron types

(A) Left: Population average activity in response to the trial start cue, shown separately for lateral habenula neurons that responded with inhibition (top, “Trial start inhibited”) or excitation (bottom, “Trial start excited”; signed-rank test, p < 0.05). Same format as Figure 6A. Right: Timecourse of trial start responses during the aversive block (blue) and appetitive block (red). Each data point is the mean response to the trial start cue during a selected group of trials within each block; blocks were divided into 7 bins each containing 6 trials. Asterisks indicate statistical significance (p < 0.05, signed-rank test). To prevent selection bias, this plot displays cross-validated data: the data displayed for each bin only includes neurons whose inhibitory (top) or excitatory (bottom) responses were statistically significant when that bin’s data was excluded from the analysis. (B) Timecourse of trial start responses for the four types of dopamine neurons, using the classification in Figure 5. Each data point is the mean response to the trial start cue during a selected group of trials within each block; blocks were divided into 7 bins each containing 6 trials. Asterisks indicate statistical significance (p < 0.05, signed-rank test). See also Figure S6.

Figure 8

Figure 8. Trial start activity is correlated with orienting reactions

(A) Mean distance between the eye and the center of the trial start cue, plotted separately for the appetitive block (red, left) and aversive block (blue, right) and for the half of saccades when the animal’s saccadic reaction time was fastest (solid lines, “Fast RT”) or slowest (dashed lines, “Slow RT”). (B) Same format as (A), plotting the mean firing rate of the lateral habenula neurons that were inhibited by the trial start cue. The firing rate was quantified using the trial start analysis window (gray bar below the x-axis in (C)); text indicates the mean difference in firing rate between fast and slow trials, its standard error, and the _p_-value (signed-rank test; asterisks indicate p < 0.05). (C) Same as (B), for dopamine neurons that were excited by the trial start cue. See also Figure S7.

Similar articles

Cited by

References

    1. Anderson AK, Christoff K, Stappen I, Panitz D, Ghahremani DG, Glover G, Gabrieli JDE, Sobel N. Dissociated neural representations of intensity and valence in human olfaction. Nat Neurosci. 2003;6:196–202. - PubMed
    1. Badia P, Harsh J, Abbott B. Choosing Between Predictable and Unpredictable Shock Conditions: Data and Theory. Psychological Bulletin. 1979;86:1107–1131.
    1. Balci F, Freestone D, Gallistel CR. Risk assessment in man and mouse. Proceedings of the National Academy of Sciences of the United States of America. 2009;106:2459–2463. - PMC - PubMed
    1. Berridge KC, Robinson TE. What is the role of dopamine in reward: hedonic impact, reward learning, or incentive salience? Brain research. 1998;28:309–369. - PubMed
    1. Bromberg-Martin E, Hikosaka O. Midbrain dopamine neurons signal preference for advance information about upcoming rewards. Neuron. 2009;63:119–126. - PMC - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources