Role of Recurrent Inhibition in Perceptual Decision-Making

Our lives are full of decisions. These may have immediate or life-long consequences. Despite the fact that we are constantly faced by the task of deciding what to do next, the process of decision-making remains one of the mysteries in neurobiology. The formation of a decision can be studied at different levels of abstraction, from the level of neural cells to the behavioral level. The latter has been reliably described by phenomenological models, which can quantify the test-subjects behavior. Theories have also been formulated to find a causal relationship between the formation of a decision and the underlying neural activity. However, the complexity of mammalian brains has impeded the identification of the causal neural substrate for decision-making. The fruit fly Drosophila melanogaster must make decisions too, but has a relatively simple brain. I will take advantage of the numerical simplicity of this model system to study the neural basis of decision-making, focusing on the role of so-called recurrent inhibition. This mechanism has been suggested to have a role in various forms of decision-making, but a causal link between the responsible cells and the formation of a decision has been lacking, simply because of the sheer complexity of the mammalian brain. The brain region, which is responsible for adaptive behavior and decision-making in the fly, receives recurrent inhibition by just one single inhibitory neuron. This makes the fly brain an ideal model system to study the role of recurrent inhibition at the single cell level. I propose that this neuron participates in a neural circuitry that encodes the evolution from a choice-neutral towards a committed state. I will test this hypothesis by measuring the activity of this neuron while the fly must make decisions in a virtual environment. This experimental system is not only suited for monitoring neural activity; it also allows for selective manipulation of the neurons activity, while observing the behavioral effects. I therefore hope to establish a direct causal link between neural activity and the formation of a decision at the single cell level.