Past events

The vestibular system detects self-motion and in turn generates reflexes that are crucial for our daily activities, such as stabilizing the visual axis (gaze) and maintaining head and body posture. In addition, the vestibular system provides us with our subjective sense of movement and orientation in space. The loss vestibular function due to aging, injury, or disease produces dizziness, postural imbalance, and an increased risk of falls – all symptoms that profoundly impair quality of life. In this talk, I will describe how the brain processes vestibular information in natural conditions. Notably, our work has established how early stages of processing encode vestibular stimuli and integrate them with extra-vestibular cues – for example proprioceptive and premotor information to ensure accurate perception and behaviour. Our experiments have revealed that while vestibular afferents respond identically to externally-generated and actively-generated self-motion, this is not the case at first central stage of sensory processing. Neurons mediating the vestibulo-spinal reflexes, as well as ascending thalamocortical pathways, are robustly activated during externally-generated motion, however their sensory response are cancelled during actively-generated movements. Our work has further revealed that this cancellation of actively-generated vestibular input occurs only in conditions where the actual sensory signal matches the brain’s internal estimate of the expected sensory consequences of active movement. Moreover, when unexpected vestibular inputs becomes persistent during voluntary motion, a cerebellar-based cancellation mechanism is rapidly updated to re-enable the vital distinction between self-generated and externally-applied stimulation to ensure the maintenance of posture and stable perception. In contrast, vestibular pathways mediating the vestibulo-ocular reflex, employ a different strategy. In this pathway, head velocity is robustly encoded whenever the goal is to stabilize gaze, but when the goal is to voluntarily redirect gaze an efferent copy of the gaze command suppresses the efficacy of this reflex pathway. Taken together, these findings have important implications for understanding the neural basis of perception and action during self-motion.

Neighboring neurons in motor cortex exhibit diverse selectivity during sensation, movement preparation, and movement execution. Neuronal selectivity could emerge from diverse mechanisms, including selective connectivity and nonlinear interactions of synaptic inputs in dendrites. We studied dendritic integration in the anterior motor cortex of mice performing a tactile discrimination task with a delayed response (Guo and Li et al., 2014). We constructed a two-photon microscope that allows rapid (~15 Hz) imaging of up to 300 µm of contiguous dendrite while resolving calcium transients in individual dendritic spines. Two galvanometers and a remote focusing mirror (Botcherby et al., 2008) steer 16 kHz lines (24 µm extent) produced by a resonant mirror arbitrarily in three dimensions. Pyramidal neurons were labeled sparsely with GCaMP6f in transgenic mice. We imaged spine and dendritic calcium transients, as well as somatic calcium transients associated with action potentials. We developed methods to computationally remove the influence of backpropagating action potentials (bAPs), which allowed us to quantify the selectivity of spines and dendritic segments during sensation, movement preparation, and movement execution. Nearby spines and dendritic segments share similar selectivity (length constant of signal correlation, ~30 µm). This clustering was more often seen in distal than in proximal dendrites. Using a measure of local autocorrelation, we also found that this reflects distinct “hotspot” locations on the dendrite where nearby dendrite and spines are co-active in time. Hotspot selectivity was correlated with the behavioral selectivity of somatic spikes, suggesting that these locations may have privileged influence over the output of the cell.

I will discuss our efforts towards understanding synaptic and circuit dysfunction in Fragile X syndrome, the most common heritable cause of intellectual disability and the leading genetic cause of autism. I will describe our studies identifying major presynaptic defects in excitability and neurotransmitter release in Fragile X and the role of ion channel dysregulation in these deficits. Finally, I will present evidence for a direct link between presynaptic dysregulation and specific Fragile X phenotypes in a patient.

The neural network of the retina processes the stream of visual signals falling onto the eye. When a visual image is presented to the retina, retinal ganglion cells, which form the output of this network, encode changes in local visual contrast inside their receptive fields. In natural vision, however, images do not arrive in isolation, but are structured in rapid sequences, separated by frequent saccades, which activate some types of ganglion cells and suppress others. Yet, little is known about how the rapid succession of images induced by saccades affects the encoding of spatial visual information. We found that a specific type of retinal ganglion cells, recorded in mouse retina, displays unexpected responses to saccade-like image transitions; the cells elicit a distinct spike burst when the same visual pattern reappears after the transition, providing a special code for such transitions or image parts that lead to a recurrence of stimulus patterns. This sensitivity to image recurrence is mediated by a circuit of serial inhibition, allowing a rapid reappearance of the image to suppress transition-induced inhibition of the ganglion cell. Our results show that saccade-like image transitions trigger interactions in the complex inhibitory network of the retina that lead to a dynamical gating of the information flow through the retina and provide a mode of operation that differs from the processing of simple, standard laboratory stimuli.