In vivo microendoscopy and computational modeling studies of mammalian brain circuits


Professor Mark Schnitzer


Depart of Biological Sciences, Department of Applied Physics, Stanford University

 Professor Schnitzer will describe two biophysical approaches to the study of mammalian learning and memory that his lab is pursuing:
                

First,Stanford researchers are developing fluorescence microendoscopy, an emerging optical modality providing cellular level imaging in deep brain tissues that have been inaccessible to in vivo microscopy. One- and two-photon fluorescence microendoscopy based on minimally invasive micro-lenses (350-1000 micron diameter) offer micron-scale resolution and have enabled visualization of neurons and blood cells in deep areas of the live mammalian brain. They have recently developed a chronic mouse preparation that has enabled in vivo microendoscopy imaging of fluorescent CA1 hippocampal pyramidal cells over several months after an initial surgery. They have also built a compact (3.9 gram) two-photon fluorescence microendoscope that is intended for brain imaging in freely moving mice, based on a flexible photonic bandgap fiber for delivery of fluorescence excitation, a DC micro-motor for fine focal control, and a dichroic micro-prism.

               

Second, they are performing computational studies of cerebellum-dependent motor learning, including classical conditioning of motor reflexes and adaptation of the vestibulo-ocular reflex (VOR), to address basic questions about how cerebellar brain circuits process timed stimuli and produce well-timed motor outputs. The leading theoretical framework invokes a long-term depression (LTD) of cerebellar parallel fiber to Purkinje cell synapses as a mechanism underlying learning. This does not account for several temporal aspects of motor behavior, and assumes GABAergic projections from Purkinje cells to deep cerebellar nuclei (DCN) neurons are purely inhibitory in effect. However, it is well established that hyperpolarization of DCN neurons commonly leads to subsequent depolarization due to rebound conductances. They have developed a theory of learning based on such rebound excitation, which makes testable predictions and provides a consistent account of previously unexplained aspects of classical conditioning and VOR adaptation during acquisition and expression phases of learning.