Distinct dopaminergic spike-timing-dependent plasticity rules are suited to different functional roles

Abstract

AbstractVarious mathematical models have been formulated to describe the changes in synaptic strengths resulting from spike-timing-dependent plasticity (STDP). A subset of these models include a third factor, dopamine, which interacts with the timing of pre- and postsynaptic spiking to contribute to plasticity at specific synapses, notably those from cortex to striatum at the input layer of the basal ganglia. Theoretical work to analyze these plasticity models has largely focused on abstract issues, such as the conditions under which they may promote synchronization and the weight distributions induced by inputs with simple correlation structures, rather than on scenarios associated with specific tasks, and has generally not considered dopamine-dependent forms of STDP. In this paper, we analyze, mathematically and with simulations, three forms of dopamine-modulated STDP in three scenarios that are relevant to corticostriatal synapses. Two of the models considered comprise previously proposed STDP rules with modifications to incorporate dopamine, while the third is a corticostriatal dopamine-dependent STDP rule adapted from a similar one already in the literature. We test the ability of each of the three models to maintain its weights in the face of noise and to complete simple reward prediction and action selection tasks, studying the learned weight distributions and corresponding task performance in each setting. Interestingly, we find that each of the three plasticity rules is well suited to a subset of the scenarios studied but falls short in others. These results show that different tasks may require different forms of synaptic plasticity, yielding the prediction that the precise form of the STDP mechanism may vary across regions of the striatum, and other brain areas impacted by dopamine, that are involved in distinct computational functions.Author summaryLearning from feedback is a crucial ability that allows humans and other animals to respond and adapt to their environments. One important locus for such learning is the basal ganglia, where dopamine-modulated corticostriatal plasticity shapes the dynamics of the cortico-basal ganglia-thalamic network in response to feedback signals to promote adaptive behavior. In this paper we ask, what learning rule is best suited to modeling this dopamine-modulated plasticity? To that end we investigate three learning rules that incorporate spike-timing-dependent plasticity as well as dopaminergic modulation. We study their performance in several settings meant to model the kinds of tasks and scenarios that striatal neurons are likely to be involved in. Each plasticity rule we examined performs well in some settings but fails in others. Different plasticity mechanisms may therefore be better suited to different functional roles and potentially to different regions of the brain.