Multitimescale Computation by Astrocytes

Abstract

The basic computational unit of the brain has long been defined as the neuron. However, mounting evidence suggests that other cells, especially astrocytes, also perform computations. Here, we demonstrate that cerebellar astrocytes decompose norepinephrine input into slow and fast calcium ion activities through differential adrenergic receptor engagement. During reward learning in mice, these slow and fast activities selectively target and modulate distinct synaptic pathways. Causal manipulations reveal that fast 2-adrenergic signals govern event-triggered responses and reinforcement learning, whereas slow 1-adrenergic signals maintain behavioral states and coordinate transitions. Remarkably, an actor-critic neural network trained on a similar sequence task spontaneously recapitulates these multitemporal dynamics, suggesting that astrocytes implement critic-like computations that evaluate states and modulate neuronal learning. These findings establish astrocytes as multilevel processors that transform univariate neuromodulatory inputs into multivariate, pathway-specific circuit controls operating in parallel with neuronal processing.