All-Optical Electrophysiology of Excitation and Inhibition in Neural Circuits

Abstract

The stability of neural dynamics arises through a tight coupling of excitatory (E) and inhibitory (I) signals, and imbalance in these signals is implicated in many nervous system disorders such as schizophrenia and autism. Genetically encoded voltage indicators (GEVIs) can report both spikes and subthreshold dynamics, but voltage only reveals the combined effects of E and I currents, not their separate contributions individually. All-optical electrophysiology, simultaneous optical manipulation and recording of electrical activity of genetically defined neurons, would greatly facilitate studies of E/I balance in neuronal information processing. My PhD work has been focused on developing optical methods for probing E/I balance in cultured neurons for disease modeling and in awake, behaving mice to study attentional control of cortical layer 1 neurons. First, I developed optogenetic tools and methods for all-optical interrogation of synaptic electrophysiology (synOptopatch), and applied the technique to address an important and controversial neurobiological question: why does blockade of excitatory N-methyl-D-aspartate receptor (NMDAR)-mediated synaptic transmission by ketamine lead to overall enhanced neural activity? This counterintuitive phenomenon is important because NMDAR hypofunction and consequent network hyperactivity are hypothesized to occur in schizophrenia and are the basis of ketamine-induced model of schizophrenia. I developed genetic constructs to express a channelrhodopsin actuator and an archaerhodopsin-derived voltage indicator in disjoint subsets of neurons. Optically induced activity in the channelrhodopsin-expressing neurons generated excitatory and inhibitory postsynaptic potentials that could be optically resolved in reporter-expressing neurons. I demonstrated synOptopatch recordings in cultured rodent neurons and in acute rodent brain slice. In synOptopatch measurements of primary rodent cultures, acute ketamine administration suppressed disynaptic inhibitory feedbacks, mimicking the effect of this drug on network function in both rodents and humans. I discovered that this action of ketamine is through blocking E-to-I synapses. These results establish an in vitro all-optical model of disynaptic disinhibition, a synaptic defect hypothesized in schizophrenia-associated psychosis. Second, I developed optogenetic tools and methods for all-optical dissection of excitation and inhibition in vivo, and revealed the input-output properties of barrel cortical L1 circuit in awake mice during sensory processing. Our brain receives constant inputs from all of our sensory organs; yet we only attend to inputs that are important to us, either because we have learned that the input is important (top-down signals) or because the input is novel or salient (bottom-up signals). How does the brain control which inputs get processed and which get ignored? Cortical L1 interneurons have been hypothesized to be a hub for attentional control by integrating bottom-up and top-down inputs and controlling the underlying cortex through inhibition or disinhibition. However, it is unclear what their activity dynamics are in awake behaving mice during sensory processing and how they integrate the different inputs to produce the output. To study the input-output properties of L1 circuit, I developed all-optical electrophysiology in awake mice -- simultaneous optical manipulation and recording of membrane voltage -- to probe both spiking, and subthreshold excitation (E) and inhibition (I) individually, and neuromodulation in barrel cortex L1 neurons. Our studies reveal how the L1 microcircuit process sensory input by integrating thalamocortical excitation, lateral inhibition and top-down neuromodulatory inputs. We develop a simple computational model of the L1 microcircuit which captures the main features of our data. Together, these results suggest a model for computation in L1 interneurons consistent with their hypothesized role in attentional gating of the underlying cortex. My results demonstrate that all-optical electrophysiology can reveal basic principles of neural circuit function in vivo. This work provides a roadmap for how one can use all-optical electrophysiology to dissect circuit function in awake mice, a task which has been formidably difficult using conventional tools (patch clamp or calcium imaging). In the context of attentional control, the work opens the possibility of follow-on experiments to study in greater detail the role of different types of sensory and modulatory inputs to L1 as well as the downstream consequences of L1 activation.