Commentary: Burst Firing in a Motion-Sensitive Neural Pathway Correlates with Expansion Properties of Looming Objects That Evoke Avoidance Behaviors

What is the neural code? This essential question has been the driving force behind much research in sensory and motor neuroscience, spurring investigations of diverse animals, brain areas, and behaviors. Given that neurons generally transmit information with trains of voltage spikes, what information are these spikes representing, and how can they be interpreted? Do downstream neurons respond to spike rates, counts, times, or some combination of these parameters? How does changing the pattern of spikes change behavior? These are fundamental questions in computational and behavioral neuroscience, and the answers have been as diverse as the neurons themselves. The reflexive response of locusts to looming stimuli is a classic model in neuroethology, with a small number of neurons encoding the expanding stimulus and with spike trains resulting in a robust jump response or wing steering maneuvers. Because the neural circuit is relatively direct, from retina to muscle in only a handful of synapses, it is an excellent candidate for the study of neural coding. In this pathway, the visual expansion of a dark circle on the locust's retina, representing a looming threat, results in a train of spikes in the lobula giant motion detector (LGMD) that signal to a descending interneuron, the descending contralateral motion detector (DCMD) (Gabbiani et al., 1999). The DCMD then stimulates thoracic interneurons and motoneurons to initiate jump or wing-steering responses (Burrows and Fraser Rowell, 1973). In previous work, various parameters of the DCMD spike train, including firing rate, time of peak firing rate, and total spike count, were found to control different aspects of the jump response (Fotowat et al., 2011), indicating that a train of spikes from a single neuron can modulate responses in multiple ways. This multiplexing suggests that the DCMD spike train is a powerful means of motor control. Previous analysis of the DCMD response to a looming stimulus found that spike rate increases as the stimulus expands, peaking within 200 ms of the predicted collision and peaking sooner when the stimulus is moving faster. In a new study, McMillan and Gray analyzed DCMD spike trains looking for evidence of neural bursting, which they defined using specific statistics, most importantly an inter-spike interval of <8 ms. They found that the DCMD neuron does indeed burst, and most of these bursts occur when the looming stimulus becomes imminent (>200 ms