Animals must be able to take external stimuli from their environment and create an internal representation that can be used for survival. The neocortex conveys general representations of sensory events to the hippocampus, and the hippocampus is thought to classify and sharpen the distinctions between these events. We recorded populations of dentate gyrus granule cells (DG GCs) and lateral entorhinal cortex (LEC) neurons across days to understand how sensory representations are modified by experience. We found representations of odors in DG GCs that required synaptic input from the LEC. Odor classification accuracy in DG GCs correlated with future behavioral discrimination. In associative learning, DG GCs, more so than LEC neurons, changed their responses to odor stimuli, increasing the distance in neural representations between stimuli, responding more to the conditioned and less to the unconditioned odorant. Thus, with learning, DG GCs amplify the decodability of cortical representations of important stimuli, which may facilitate information storage to guide behavior.

Parkinson’s disease (PD) is a neurodegenerative disease, in which the progressive loss of dopamine neurons is associated with prominent motor deficits, including a loss of voluntary movement. Although the mechanisms underlying PD are poorly understood, these motor deficits can be effectively treated pharmacologically through dopamine replacement therapy with the dopamine precursor, levodopa. However, while levodopa is highly effective, the majority of patients develop motor complications, including levodopa-induced dyskinesia (LID) after 5-10 years of treatment. The excessive, involuntary movements observed in LID, which are triggered by levodopa administration, limit its therapeutic use. In Chapter 1, I will discuss prominent theories and experimental evidence suggesting aberrant activity of neurons in the striatum, the primary input nucleus of the basal ganglia, plays a crucial role in the abnormal movements seen in both PD and LID. This will be focused on the two main cell-types in the striatum, whose activity is thought to underlie the generation and coordination of movement in health and disease: the direct and indirect pathway medium spiny neurons (dMSNs and iMSNs, respectively). However, the precise cellular and circuit mechanisms of striatal dysfunction in PD and LID, or how aberrant striatal activity relates to behavioral dysfunction, are poorly understood. To gain a better understanding of these circuit mechanisms, in Chapter 2, I use a mouse model of PD and LID combined with in vivo electrophysiological recordings of dMSNs and iMSNs. Using this approach, I identified the changes in striatal physiology that correlate with the abnormal movements seen in parkinsonism and LID. In particular, this work converges on a distinct subpopulation of dMSNs with aberrant levodopa-evoked activity. In Chapter 3, I investigate how the differential regulation of the intrinsic and synaptic properties of these LID-associated dMSNs shape their response to levodopa, using synaptic tracing, combined with in vivo and ex vivo electrophysiology experiments. Together, these studies elucidate the cellular and circuit dysfunction that arises following chronic changes in dopamine and shed light on how the heterogeneous properties of direct pathway neurons in LID contribute to the differential therapeutic and dyskinetic effects of levodopa.

It is commonly hypothesized that hyperexcitability of local circuits within the rostral anterior cingulate cortex (rACC) underlies the affective features of chronic neuropathic pain, but the contribution of specific inputs to the rACC remains unexplored. Here we used optogenetics to study how inputs from the mediodorsal thalamus (MD) and basolateral amygdala (BLA) to the rACC influence pain-related aversion in sciatic nerve injury and chemotherapy-induced mouse models of neuropathic pain. Activating MD inputs elicited conditioned place aversion in mice with chronic neuropathic pain, whereas activating BLA inputs elicited conditioned preference. Strikingly, inhibiting BLA inputs elicited the same effects as activating MD inputs, i.e., exacerbated pain-related aversion. To assess whether these behavioral effects reflect a general pathological dysregulation of rACC neurons vs. changes at specific synapses, we recorded intrinsic properties of layer V rACC pyramidal neurons and their response to light-mediated activation of MD or BLA inputs. We focused on two subtypes of layer V rACC pyramidal projection neurons: intratelencephically-projecting (IT) and subcortically-projecting (SC). In slices from animals with chronic neuropathic pain, we found that both IT and SC neurons are hyperexcitable compared to control. Unexpectedly, however, excitatory responses of these neurons to MD input activation are significantly weakened in pain conditions compared to control. In fact, in mice with chronic neuropathic pain, the ratio between excitation and inhibition elicited by MD inputs shifted towards inhibition, specifically within SC neurons of the rACC. Responses of IT and SC neurons to BLA input activation are, in contrast, strengthened compared to control. We proposed a model positing that activating MD inputs and inhibiting BLA inputs both drive pain-related aversion through a net inhibition of subcortically-projecting (SC) rACC neurons. Indeed, direct inhibition of SC neurons elicited conditioned place aversion, specifically in mice with chronic neuropathic pain. Inhibition of IT neurons, in contrast, has no effect on pain-related aversion. Our findings reveal a novel role for specific rACC cell types and inputs on pain-related aversion. We conclude that chronic pain-related aversion may be better understood in terms of activity of specific input-output pathways of the rACC, rather than overall activity of the rACC.

The basal ganglia are an evolutionarily conserved set of nuclei that are crucial for context-dependent learning and selection of appropriate behavioral outputs. The striatum is considered the input nucleus of the basal ganglia because it not only receives the bulk of excitatory synaptic drive from sources outside the basal ganglia, but also contains two populations of medium spiny neurons that give rise to the direct and indirect basal ganglia pathways. The activity of direct and indirect pathway medium spiny neurons (dMSNs and iMSNs) is sufficient to promote or suppress behavioral output, respectively. Changes in synaptic strength at excitatory inputs to MSNs are thought to underlie context-dependent learning, whereas the acute modulation of MSN intrinsic excitability affects motivation or vigor of an ongoing behavior. Previous research has revealed that the neuromodulator dopamine appears to be in a unique position to exert both acute and chronic modulatory effects over MSNs and their excitatory inputs. The differential expression of Gs-coupled D1 and Gi-coupled D2 receptors by dMSNs and iMSNs, respectively, results in opposing changes in direct and indirect pathway function with changes in striatal dopamine levels. However, previous technical limitations have restricted the study of dopamine's acute actions in the striatum and the long-term effects of dopamine on MSN subtypes and excitatory inputs of distinct origins. Here we use recently developed optogenetic and chemogenetic techniques combined with behavior and ex vivo brain slice electrophysiology to study the mechanisms of acute and chronic dopamine signaling in MSNs and their excitatory inputs. In Chapter One, we review the evidence for the dual-pathway model of basal ganglia function and its modulation by dopamine in the context of controlling behavioral output. In Chapter Two, we perform an electrophysiological comparison of common optogenetic proteins to determine the optimal tool for our experiments. We then use optogenetic control of dopamine release to show specific modulation of dMSN intrinsic excitability and D1-mediated invigoration of behavior in Chapter Three. In Chapter Four, we provide evidence that the chronic loss of dopamine in Parkinson's disease results in a reorganization of the thalamostriatal system that has negative outcomes on motor control. Finally, we discuss the overarching implications of these findings in the context of basal ganglia function in health and disease in Chapter Five. These findings not only represent a long overdue reevaluation of the assumptions surrounding the role of dopamine in striatal function, but also provide new testable hypotheses addressing the cellular and synaptic bases of an evolutionarily conserved system for the dynamic regulation of ethologically relevant behaviors.

Fast spiking interneurons (FSIs) are thought to play a critical role in the normal function of the striatum, which is the input nucleus to the basal ganglia. In vitro studies have determined that FSIs are a major source of inhibition onto striatal principal cells, or medium spiny neurons (MSNs). However, in vivo studies have produced conflicting results regarding how FSI activity modulates MSN activity in the awake animal. Here, we use optogenetic techniques and acute head-fixed extracellular recordings to resolve existing controversy in our understanding of FSI-MSN interactions. We find that the net effect of FSI activity is to suppress MSNs, and in the process, discover an opsin-independent effect of light on physiology and behavior. We find strong evidence that this effect is mediated by temperature, and determine parameters for the safe use of light in neuroscience experiments. We then investigate the specific manner in which FSIs influence MSN activity by recording the sensory responses of striatal units in the auditory striatum. Here, we find that MSNs are more frequency selective for pure tone stimuli than neighboring FSIs, and exhibit differences in their responses to continuous stimuli when compared with FSIs. Optogenetic inhibition of FSIs results in a disinhibition of MSN responses to pure tones, increasing the magnitude of their responses and reducing their frequency selectivity. These results suggest that FSI activity in the auditory striatum serves to shape the activity of surrounding MSNs and increase their frequency selectivity.

The ability to locomote through the environment is fundamental to most animal species’ survival. Vertebrates ranging from lamprey to humans share common circuitry involved in sending signals from the brain to the spinal cord to control central pattern generators (CPGs). These CPGs in turn control the alternating muscle activations that result in walking, slithering, swimming or flying. The mesencephalic locomotor region (MLR) is putatively the final node in the central nervous system that integrates locomotor information from various brain regions and sends locomotor commands to the periphery. Originally defined functionally as a brainstem region in which electrical stimulation drives locomotion with short latencies, the MLR has been found in lamprey, stingray, dogs, birds, guinea pigs, mice, rats, cats, and dogs among other species. The MLR receives input from a variety of areas which all likely affect whether or not and how fast an animal locomotes. In this dissertation we will explore both the MLR and two specific afferent connections, one from the basal ganglia (BG) and one from the central amygdala (CeA) which are demonstrated to control locomotion through the MLR. In Chapter 1 we will provide an introduction to the MLR, the BG and the CeA. In Chapter 2 we will dissect the circuitry of the mouse MLR to provide evidence for which nuclei and MLR cell types are responsible for the results observed during electrical stimulation. In Chapter 3 we will look at how the BG controls locomotion through the MLR, taking time to dissect the two canonical pathways which are hypothesized to bidirectionally control locomotion as well as movement in general. Chapter 4 will provide evidence for a “stop” signal sent from the CeA to the MLR that determines whether or not an animal will run in response to aversive stimuli. Finally, in Chapter 5 we discuss the significance of the work and look at how it might shape the direction of future experiments.

The basal ganglia are a group of interconnected, subcortical nuclei that mediate a wide array of behavioral functions. The primary input nucleus of the basal ganglia, the striatum, is a region implicated in behaviors ranging from motor control, reward learning, decision making, habit formation, to social interaction. How these diverse behavioral functions map onto various striatal cell types and circuit connections remains a major question within the basal ganglia field. Classical models of striatal circuitry that divide output neurons into populations called the direct and indirect pathways have proven extremely useful in understanding how the striatum mediates function such as movement initiation and behavioral reinforcement. Furthermore, such models have played a foundational role in understanding the pathophysiology of disorders linked to the basal ganglia, such as Parkinson’s disease and Huntington’s disease. These models, however, struggle to explain how the striatum mediates other well-established functions, and can even lead to contradictory predictions of behavior in other types of disease. To explore how alternative models of striatal circuitry might be used to fill in these gaps, we took two parallel approaches. First, we investigated whether striosome and matrix neurons—an alternative model for dividing striatal neurons—differ in the electrophysiological properties and synaptic connectivity. Early anatomical work suggested that striosome and matrix may represent two parallel circuits, relaying limbic and motor information to downstream nuclei, respectively. However, recent attempts to genetically isolate these population cast doubt upon these earlier findings. Using a developmental approach to isolate striosome and matrix neurons, we found that these population receive differing cortical input consistent with early anatomical studies. Furthermore, we determined that striosome and matrix differ in how they process cortical input and where they project information within the midbrain. Next, we investigated how changes in striatal activity relate to a disorder called tardive dyskinesia. Resulting from chronic administration of antipsychotic medications, tardive dyskinesia is characterized by involuntary movements of the face, mouth, and tongue that can be irreversible once developed. Several lines of evidence implicate a causal role for the striatum in tardive dyskinesia, and classical models predict that excessive involuntary movements may be driven by hypoactivity of indirect pathway neurons. Unfortunately, little is known about how striatal activity changes over the course of the disease. To address this gap in knowledge, we used a combination of genetic manipulations, in vivo recordings, and slice electrophysiology to determine how striatal activity changes of the development of TD. Currently in progress, our findings indicate that alternative changes outside that predicted by classical models of basal ganglia function likely underlie TD. Mainly, we find that tardive dyskinesia may arise as the result of changes in dopaminergic signaling that affect both direct and indirect pathway neurons. In combination, the results of these projects demonstrate that additional models of striatal circuitry are necessary to describe its function, and additionally identify new cellular populations on which to build such models.

The basal ganglia are a network of subcortical brain nuclei important for action selection and motor learning. Excitatory synapses onto projection neurons in the striatum, the input nucleus of the basal ganglia, are a major site of long-term synaptic plasticity. This plasticity has the potential to powerfully regulate motor function by setting the gain on the incoming cortical signals that drive basal ganglia circuits. Endocannabinoid-dependent long-term depression (eCB-LTD) is the best characterized form of striatal plasticity, but the mechanisms governing its normal regulation and pathological dysregulation are not well understood. This thesis examines the molecular mechanisms underlying striatal eCB-LTD. It focuses on eCB-LTD at excitatory synapses onto a subset of striatal projection neurons, or medium spiny neurons (MSNs), the "indirect-pathway" MSNs. These MSNs express dopamine D2 and adenosine A2A receptors and project to the external segment of the globus pallidus. In this thesis, I characterize two distinct biochemical signaling pathways mediating eCB production in striatal indirect-pathway MSNs: a PLCβ- and DAG lipase-dependent pathway that is engaged by low-frequency stimulation, and a novel Src- and PLD-dependent pathway that is engaged by high-frequency stimulation. Endocannabinoid production through both signaling pathways is modulated by dopamine D2 and adenosine A2A receptors, acting through cAMP/PKA. I identify Regulator of G-protein Signaling 4 (RGS4) as a key link between D2/A2A signaling and eCB mobilization pathways. Behaviorally, I find that indirect-pathway eCB-LTD likely underlies at least some of the stimulant properties of A2A receptor antagonists and that the loss of indirect-pathway eCB-LTD may mediate the development of parkinsonian motor deficits following dopamine depletion. In experiments using RGS4 knockout mice, which in contrast to wildtype mice exhibit normal eCB-LTD following dopamine depletion, I show that the loss of RGS4 leads to fewer behavioral deficits after dopamine depletion, suggesting that inhibition of RGS4 may be an effective non-dopaminergic strategy for treating Parkinson's disease.