Single neuron and network computations in escape decisions

Computing decisions is a problem that animals must constantly solve, which requires analysing information and selecting appropriate actions. Previous work has demonstrated a role for cortical circuits in learned decision-making paradigms, but we do not know how elementary decision processes, such as integrating evidence, map onto circuit and cellular mechanisms. Our goal is to uncover these mechanisms by investigating instinctive escape decisions in mice. Recent work has identified the superior colliculus and periaqueductal gray as key circuits for escape initiation, providing a unique entry point for investigating decision-making. We aim to explain at the cellular and circuit level the processes of threat stimulus integration and selection of escape behaviour. Our experimental strategy is to record and stimulate neural activity in the midbrain, while manipulating escape decisions by varying the threat level of sensory stimuli, past experience and the spatial environment. At the circuit level we will make recordings with high-density silicon probes and calcium imaging, while at the single neuron level we will use whole-cell recordings to investigate connectivity and the biophysics of synaptic integration during decision-making. The results from this project will produce mechanistic models of how neurons integrate sensory information and past experience to generate behavioural choices.