Stephen G. Rayport - US grants

Synapses are central to understanding cellular mechanisms underlying learning, memory, certain disease states and more broadly aspects of mental organization. As yet, identified mammalian central synapses have been largely inaccessible to mechanistic study. Inspired by recent success in establishing in vitro synaptically coupled neurons from the marine mollusc Aplysia mediating learning, this proposal seeks to apply the Aplysia dissociated cell culture technique to neonatal rat brain slices. In this way cultures will be made containing two or at most three cells identified by origin and type. Not only will such cultures permit simultaneous recordings from synaptically coupled cells, but they will have relevance to actual synapses in vivo. To complement the cell culture approach, novel ways of approaching slices will be pursued that offer the possibility of examining the same synapses with dual recordings closer to the undisturbed state. The thrust of the work focuses on learning and schizophrenia. Organizing questions are: (1) What are the cellular bases of long- term potentiation -- are the mechanisms pre or post-synaptic or both? (2) What are the characteristics of dopaminergic synaptic transmission? (3) How are dopaminergic synapses influenced by neuroleptics in the short and long term? (4) How might dopaminergic synapses and neuroleptics impact on long-term potentiation? Understanding long-term potentiation in cellular terms promises the ability to target specific pharmacologic agents so as to relate the phenomenon to memory and behavior. Likewise, more explicitly correlating the behavioral effects of neuroleptics drugs with their sites of synaptic action may bridge between neuronal mechanisms and psychosis.

Schizophrenia is a prevalent disease of profound morbidity. Its treatment was revolutionized by the advent of the neuroleptics. These drugs reduce psychotic symptoms by blocking the synaptic action of dopamine. Consequently, pathologic dopaminergic synaptic transmission, in particular, that of mesocorticolimbic dopamine neurons has become the focus of attempts to elucidate the neurophysiology of the schizophrenic psychosis. Mechanistic understanding of the functioning of these synapses, however, has remained inferential, contributing to the controversy over how neuroleptics act at these critical sites. What has been needed to gain a deeper understanding of the function of mesocorticolimbic dopamine neurons has been a method of examining their synapses individually. In preliminary studies, one subclass of mesocorticolimbic dopamine neurons, those projecting to the nucleus accumbens, has been successfully labelled in midbrain cell cultures, using retrograde transport of fluorescent latex microspheres. In vitro, these cells show their characteristic electrophysiology, known from studies in intact preparations and brain slice work. Examining the synapses formed by such identified cells now makes approachable four seminal questions: (1) What are the functional properties of mesocorticolimbic dopamine synapses? (2) How do neuroleptics act at these synapses, both acutely and chronically? (3) How does dopamine interact with its autoreceptors, found on the terminals of certain mesocorticolimbic dopamine neurons? (4) And, what role do the co-transmitters cholecystokinin and neurotensin play? Characterization of mesocorticolimbic dopamine synapses would provide a significant step toward discerning the role of the mesocorticolimbic dopamine system in the schizophrenic psychosis. Moreover, understanding neuroleptic action on dopamine synapses might lead to the development of improved pharmacologic treatments. In the longer term, insight might be gained into treatment of neuroleptic-refractory symptoms, those of the deficit state of chronic schizophrenia, which bring about the tremendous morbidity of the illness.

DESCRIPTION (provided by applicant): The goal of this competing continuation is to further the development of the Candidate as a translational neuroscientist. In this competing continuation, the Candidate plans to focus on the role of glutamatergic synaptic transmission in sensitization. It is well known that psychostimulants act by increasing synaptic dopamine levels, principally in the nucleus accumbens. When taken repeatedly, constant doses produce an increasing behavioral response -- known as psychostimulant sensitization. This animal model of drug dependence is mediated by neuroplastic changes both at the level of the dopamine neuron cell bodies in the ventral tegmental area and at their synapses in the nucleus accumbens. These changes require glutamatergic synaptic transmission. Recently, this laboratory has made the striking observation that dopamine neurons corelease glutamate. If so, glutamatergic synapses of dopamine neurons are likely to be important in sensitization. To test this, mice generated in this laboratory with fluorescent dopamine neurons will be used to examine the relationship between the dopaminergic and glutamatergic terminals of single dopamine neurons and to assess the plastic capabilities of the glutamatergic synapses as a basis for sensitization. To address the role of the glutamatergic cotransmission in the behaving animal, another line of mice have been made that lack glutaminase -- the enzyme principally responsible for the production of neurotransmitter glutamate. Preliminary results confirming the importance of glutaminase will be extended to test this definitively. Interestingly, mice heterozygous for glutaminase appear to be already in a sensitized state, as they show an exaggerated response to stimulants. Using tissue-specific rescue and deletion approaches, the final aim is to identify the crucial glutamatergic circuits underlying the sensitized phenotype. Finally, the role of glutamatergic cotransmission by dopamine neurons in the development of sensitization will be tested in mice lacking glutaminase in their dopamine neurons. This integrated approach should help to elucidate the crucial glutamatergic circuits underlying sensitization, and offer new targets for the pharmacological reduction of sensitization and thus of drug dependence.

DESCRIPTION(From applicant's abstract): Repeated psychostimulant exposure causes psychostimulant sensitization, an increasing motor response to repeated psychostimulant administration that is thought to be an animal model of drug dependence. Psychostimulants - as well as most drugs of abuse - appear to have their principal action in the mesoaccumbens dopamine system. Sensitization involves pre- and postsynaptic adaptations in dopamine as well as glutamate synaptic transmission in the mesoaccumbens systems. So, it is striking that the dopamine neurons that form the mesoaccumbens system appear to corelease glutamate. This raises the hypothesis that glutamate synaptic transmission by dopamine neurons is crucial to psychostimulant sensitization and provides a novel therapeutic target for the pharmacotherapy of addiction. To address this hypothesis, transgenic techniques will be used to generate three lines of mutant mice lacking glutaminase - the enzyme principally responsible for the synthesis of neurotransmitter glutamate. The first line will be a constitutive knockout; this should serve to confirm that glutaminase is the principal source of neurotransmitter glutamate. The second line of mice will involve the rescue of glutaminase expression in forebrain neurons of the constitutive knockout mice; this will yield a monoaminergic neuron-selective glutaminase knockout, and provide an initial way to evaluate the role of monoaminergic neuron glutamate cotransmission in sensitization. In the third line of mice, the dopamine transporter promotor will be used to drive the selective elimination of glutaminase expression in dopamine neurons. In the three lines of mice, classical histology, western blotting, in situ hybridization, and immunocytochemistry will be done to evaluate the impact of the knockouts on the cellular organization of the brain, the cellular selectivity of the glutaminase knockout, and the effects on cellular glutamate levels. Dopamine neuron cultures prepared from the mutant mice will be used to show that eliminating glutaminase expression blocks glutamatergic synaptic transmission. Behavioral testing of the mutant mice will then be done to evaluate changes in their motor activity, appetitive behavior, propensity to manifest psychostimulant sensitization, and vulnerability to self-administration. If psychostimulant sensitization and self-administration are indeed blocked by eliminating dopamine neuron glutamate cotransmission, then these synapses would become a novel target for therapeutic intervention in the treatment of drug dependence.

DESCRIPTION (provided by applicant): Dysregulated glutamatergic neurotransmission has been strongly implicated in the psychopathology of schizophrenia. Recent studies have highlighted the therapeutic promise of presynaptic reductions in glutamate transmission. We have shown that neurons from glutaminase (GLS1) knockout mice show an activity-dependent presynaptic reduction in glutamatergic synaptic transmission. While GLS1 knockout mice die shortly after birth, GLS1 haploinsufficient mice with one functional GLS1 allele (GLS1 hets) are remarkably normal. Strikingly, functional imaging reveals that the mice have focal hypometabolism in the hippocampus, mainly involving the CA1 subregion and the subiculum that is the exact inverse of recent imaging findings in patients with schizophrenia. Moreover, when challenged with pro-psychotic drugs, GLS1 het mice manifest behavioral and neurochemical phenotypes consistent with schizophrenia resilience. Thus, reducing glutaminase activity appears to have therapeutic potential for schizophrenia. To translate this discovery to clinical application, we propose testing the hypothesis that GLS1 het mice are in fact resilient to a range of pro-schizophrenic insults. Using tissue-specific GLS1 deletions, we will ask whether the hippocampal hypometabolism arises from the reduction in GLS1 in the hippocampus, and whether this modulation is sufficient to produce the resilience phenotype. To begin to understand the implications of the resilience phenotype for the pathophysiology of schizophrenia, we will identify the synaptic alterations in the hippocampus that underlie the hypoactivity profile. To test the therapeutic potential of GLS1 inhibition directly, we will induce GLS1 haploinsufficiency in adult mice, doing what we term genetic-pharmacotherapy, to investigate the acute and chronic effects of the intervention. We will induce GLS1 haploinsufficiency earlier in development to explore potential benefits of early intervention and neurodevelopmental contributions. Finally, we will do high-throughput screening to identify small-molecule GLS1 inhibitors with nanomolar efficacy as drug candidates. GLS1 inhibition has therapeutic potential not only for schizophrenia, but also for stroke, and other neurodegenerative disorders involving excitotoxicity, so a CNS-active GLS1 inhibitor will likely have broad therapeutic promise. In summary, the planned preclinical studies together with identification of drug candidates should provide the basis for movement of GLS1 inhibition towards clinical trials as a novel pharmacotherapy for schizophrenia.

? DESCRIPTION (provided by applicant): Many brain neurons release a mix of transmitters, but it has been challenging to identify their synapses based on transmitter status. The application of proximity ligation assay (PLA) technology in optogenetic mice offers a way to address these issues comprehensively, enabling visualization of all synapses of an identified population of neurons and their transmitter status. PLA is a hybrid immunochemical and in situ hybridization approach in which two selected epitopes, which are within about 20 nm of each other, generate a discrete fluorescence signal. Synaptic vesicles at the active zone are a distinctive functional element of synapses, which are within 20 nm of the plasmalemma when poised for release. With PLA, we have recently visualized dopamine receptor oligomerization in striatal neurons and addressed the developmental trajectory of dopamine receptor colocalization. In this project, we will use PLA to visualize the proximity of synaptic markers in dopamine neurons to determine key measures of dopamine neuron synaptic function. The specific aims are to: <1> Visualize dopamine neuron synaptic release sites specifically, then visualize dopamine neuron release sites based on the transmitter released. This will be done in optogenetic mice conditionally expressing the exogenous membrane, axon-targeted protein ChR2-EYFP in dopamine neurons by detecting proximity of ChR2-EYFP in the plasmalemma to key synaptic vesicle proteins. <2> Correlate PLA measurements of dopamine neuron connectivity in target areas with functional connectivity to determine the functional readout of the PLA measurements. <3> Demonstrate the ability of PLA to identify synaptic release sites in tissue from wild type mice using proximity of native plasmalemma and vesicular membrane proteins. Stereology will be used for systematic image acquisition and mapping. The quantitative information obtained with PLA will enable asking how does synaptic connectivity change, with regional specificity, over development, and in disease models? PLA will further enable addressing questions such as, where in the brain do drugs impact most, over what time, and for how long? Once prototyped in the dopamine neurons, the PLA approaches developed in this project could be extended to the other major modulatory neuron groups. This will lay the groundwork for post-mortem studies in clinical material, and enable asking, questions such as, which synaptic connections are most affected in disease states, do treatments impact connectivity, are treatments inducing compensatory changes or reversing pathological changes?