Theories from reinforcement learning have been highly influential for interpreting neural activity in the biological circuits critical for animal and human learning. Central among these is the identification of phasic activity in dopamine neurons as a reward prediction error signal that drives learning in basal ganglia and prefrontal circuits. However, recent findings suggest that dopaminergic prediction error signals have access to complex, structured reward predictions and are sensitive to more properties of outcomes than learning theories with simple scalar value predictions might suggest. Here, I will present recent work in which we probed the identity-specific structure of reward prediction errors in an odor-guided choice task and found evidence for multiple predictive “threads” that segregate reward predictions, and reward prediction errors, according to the specific sensory features of anticipated outcomes. Our results point to an expanded class of neural reinforcement learning algorithms in which biological agents learn rich associative structure from their environment and leverage it to build reward predictions that include information about the specific, and perhaps idiosyncratic, features of available outcomes, using these to guide behavior in even quite simple reward learning tasks.

I will discuss work with Jack Lindsey modeling reinforcement learning for action selection in the basal ganglia. I will argue that the presence of multiple brain regions, in addition to the basal ganglia, that contribute to motor control motivates the need for an off-policy basal ganglia learning algorithm. I will then describe a biological implementation of such an algorithm that predicts tuning of dopamine neurons to a quantity we call "action surprise," in addition to reward prediction error. In the same model, an implementation of learning from a motor efference copy also predicts a novel solution to the problem of multiplexing feedforward and efference-related striatal activity. The solution exploits the difference between D1 and D2-expressing medium spiny neurons and leads to predictions about striatal dynamics.

In the first half of my talk, I will discuss our recent work on the midbrain dopamine system. The hypothesis that midbrain dopamine neurons broadcast an error signal for the prediction of reward is among the great successes of computational neuroscience. However, our recent results contradict a core aspect of this theory: that the neurons uniformly convey a scalar, global signal. I will review this work, as well as our new efforts to update models of the neural basis of reinforcement learning with our data. In the second half of my talk, I will discuss our recent findings of state-dependent decision-making mechanisms in the striatum.

Human movement disorders and pharmacological studies have long suggested molecular dopamine modulates the pace of the internal clock. But how does the endogenous dopaminergic system influence the timing of our movements? We examined the relationship between dopaminergic signaling and the timing of reward-related, self-timed movements in mice. Animals were trained to initiate licking after a self-timed interval following a start cue; reward was delivered if the animal’s first lick fell within a rewarded window (3.3-7 s). The first-lick timing distributions exhibited the scalar property, and we leveraged the considerable variability in these distributions to determine how the activity of the dopaminergic system related to the animals’ timing. Surprisingly, dopaminergic signals ramped-up over seconds between the start-timing cue and the self-timed movement, with variable dynamics that predicted the movement/reward time, even on single trials. Steeply rising signals preceded early initiation, whereas slowly rising signals preceded later initiation. Higher baseline signals also predicted earlier self-timed movement. Optogenetic activation of dopamine neurons during self-timing did not trigger immediate movements, but rather caused systematic early-shifting of the timing distribution, whereas inhibition caused late-shifting, as if dopaminergic manipulation modulated the moment-to-moment probability of unleashing the planned movement. Consistent with this view, the dynamics of the endogenous dopaminergic signals quantitatively predicted the moment-by-moment probability of movement initiation. We conclude that ramping dopaminergic signals, potentially encoding dynamic reward expectation, probabilistically modulate the moment-by-moment decision of when to move. (Based on work from Hamilos et al., eLife, 2021).

The primary motor cortex (M1) is a major output center for movement execution and motor learning, and its dysfunction contributes to the pathophysiology of Parkinson’s disease (PD). While human studies have indicated that a loss of midbrain dopamine neurons alters M1 activation, the mechanisms underlying this phenomenon remain unclear. Using a mouse model of PD, we uncovered several shifts within M1 circuitry following dopamine depletion, including impaired excitation by thalamocortical afferents and altered excitability. Our findings add to the growing body of literature highlighting M1 as a major contributor in PD, and provide targeted neural substrates for possible therapeutic interventions.

In behaving rodents, CA1 pyramidal neurons receive spatially-tuned depolarizing synaptic input while traversing a specific location within an environment called its place. Midbrain dopamine neurons participate in reinforcement learning, and bursts of action potentials riding a depolarizing wave of synaptic input signal rewards and reward expectation. Interestingly, slice electrophysiology in vitro shows that both types of cells exhibit a pronounced reduction in firing rate (adaptation) and even cessation of firing during sustained depolarization. We included a five state Markov model of NaV1.6 (for CA1) and NaV1.2 (for dopamine neurons) respectively, in computational models of these two types of neurons. Our simulations suggest that long-term inactivation of this channel is responsible for the adaptation in CA1 pyramidal neurons, in response to triangular depolarizing current ramps. We also show that the differential contribution of slow inactivation in two subpopulations of midbrain dopamine neurons can account for their different dynamic ranges, as assessed by their responses to similar depolarizing ramps. These results suggest long-term inactivation of the sodium channel is a general mechanism for adaptation.

Animals are motivated to acquire knowledge. A particularly striking example is information seeking behavior: animals often seek out sensory cues that will inform them about the properties of uncertain future rewards, even when there is no way for them to use this information to influence the reward outcome, and even when this information comes at a considerable cost. Evidence from monkey electrophysiology and human fMRI studies suggests that orbitofrontal cortex and midbrain dopamine neurons represent the subjective value of knowledge during information seeking behavior. However, it remains unclear how the brain assigns value to information and how it integrates this with other incentives to drive behavior. We have therefore developed a task to test if information preferences are present in mice and study how informational value is imparted on stimuli. Mice are trained to enter a center port and receive an initial odor that instructs them to either go to an informative side port, go to an uninformative side port, or choose freely between them. The chosen side port then yields a second odor cue followed by a delayed probabilistic water reward. The informative port’s odor cue indicates whether the upcoming reward will be big or small. The uninformative port’s odor cue is uncorrelated with the trial outcome. Crucially, the two ports only differ in their odor cues, not in their water value since both offer identical probabilities of big and small rewards. We find that mice prefer the informative port. This preference is evident as a higher percentage choice of the informative port when given a free choice (67% +/- 1.7%, n = 14, p < 0.03), as well as by faster reaction times when instructed to go to the informative port (544ms +/- 21ms vs 795ms +/- 21ms, n = 14, p < 0.001). The preference for information is robust to within-animal reversals of informative and uninformative port locations, and, moreover, mice are willing to pay for information by choosing the informative port even if its reward amount is reduced to be substantially lower than the uninformative port. These behavioral observations suggest that odor stimuli are imparted with informational value as mice learn the information seeking task. We are currently imaging neural activity in orbitofrontal cortex with microendoscopes to identify changes in neural activity that may reflect value associated with the acquisition of knowledge.