Sabina Berretta - US grants

DESCRIPTION (provided by applicant): Chondroitin sulfate proteoglycans (CSPGs) are a main component of the brain extracellular matrix (ECM). Findings from our group, recently published in the Archives of General Psychiatry, point to substantial abnormalities affecting CSPGs in the amygdala and entorhinal cortex of subjects with schizophrenia, but not bipolar disorder. Marked increases of glial cells expressing CSPGs were not accompanied by astrocytosis, suggesting schizophrenia-specific anomalous regulation of CSPG expression. Concurrent reductions of perineuronal nets, CSPG-enriched ECM aggregates surrounding distinct neuronal populations, point to altered CSPG content in the ECM. CSPG functions, such as regulation of neuronal migration, stabilization of synaptic connectivity, maintenance of neuronal networks, neuronal microenvironment and volume transmission, bear direct relevance to the pathophysiology of schizophrenia. These functions, together with the magnitude of CSPG changes (419-1560%) in amygdala and entorhinal cortex of subjects with schizophrenia, point to a pivotal role for a disruption of glial-ECM-neuronal interactions in the pathogenesis of this disease. The main goal of the postmortem and in vitro studies proposed here is to test the hypothesis that molecular pathways regulating CSPGs may be altered in schizophrenia, causing CSPG accumulation in astrocytes and abnormal CSPG expression in extracellular matrix perineuronal aggregates. These studies are organized in four specific aims, designed to test interrelated aspects of this model. Specific Aim 1 will test the hypothesis that, in schizophrenia, increased CSPG-positive glial cells in medial temporal lobe regions correspond to astrocytes, as suggested by previous results. Specific Aim 2 will test the hypothesis that the molecular pathways regulating CSPG synthesis and secretion into the ECM are disrupted in the medial temporal lobe of subjects with schizophrenia. Specific Aim 3 will test the hypothesis that CSPG expression within the ECM may be reduced, resulting in CSPGs-impoverished perineuronal nets surrounding parvalbumin-expressing neurons. In vitro manipulations on cultured human astrocytes will be used in Specific Aim 4 to test dynamically the potential of growth factors and secretory carrier membrane proteins, investigated in Specific Aim 2, to cause CSPG accumulation in human astrocytes and decreased CSPG secretion into the extracellular space. In Specific Aims 1, 2 and 3, a group of subjects with bipolar disorder will be included to test whether CSPG abnormalities are specific to schizophrenia or represent a shared feature among major psychoses. The relevance of the proposed studies resides in their potential of uncovering an as yet unknown and distinctive aspect of the pathophysiology of SZ, affecting brain regions known to play an important role in this disease. We put forward that extracellular matrix/glial abnormalities may represent a unifying factor contributing to disturbances of neuronal migration, synaptic connectivity, and GABAergic, glutamatergic and dopaminergic neurotransmission in schizophrenia.

DESCRIPTION (provided by applicant): Recent findings from our group point to substantial abnormalities affecting the extracellular matrix (ECM) in medial temporal regions of subjects with schizophrenia, but not bipolar disorder. Our data points to anomalous expression of chondroitin sulfate proteoglycans (CSPGs), a main component of the ECM, in glial cells and ECM. CSPG functions during development and adulthod, such as regulation of neuronal migration, axon outgrowth, stabilization of synaptic connectivity, maintenance of neuronal networks and neuronal microenvironment, bear direct relevance to the pathophysiology of schizophrenia. Preliminary results suggest that similar abnormalities occur in the olfactory epithelium (OE) and olfactory bulb (OB), as well as peripherally in skin fibroblasts, of subjects with SZ. Investigations on ECM abnormalities in the olfactory system (OE and OB), as proposed here, offer compelling advantages. First, neurodevelopmental functions such as neuron diferentiation, migration in OE, and axon outgrowth toward OB, occur throughout life, alongside adult neural functions, making the olfactory system ideally suited for investigations on CSPGs. Second, growing evidence for olfactory deficits in SZ, particularly in association with negative symptoms, renders investigations on the olfactory system directly relevant to the clinical manifestations of this disease. Third, the OE is the only central nervous system structure easily accessible by biopsy, from which cel cultures can be developed. The main goal of these studies is to investigate the relationship between the pathophysiology of CSPG abnormalities in schizophrenia and clinical manifestations of this disease. The potential for CSPG abnormalities to represent a biological marker with pathophysiological relevance and specificity for schizophrenia will be assessed in the context of specific hypotheses on the mechanisms of ECM abnormalities and their association with core symptoms of this disease. The main hypothesis tested is that ECM abnormalities, due to a disruption of molecular pathways regulating CSPG synthesis and secreti0n, affect the olfactory system in subjects with SZ and can be detected peripherally. We postulate that such abnormalities may be associated with specific olfactory deficits and negative symptoms. To test this hypothesis, CSPG abnormalities will be assessed in OE and skin fibroblasts (biopsy/in vitro), as well as the OB (postmortem), from two cohorts of normal control, SZ and BD subjects. Human fibroblast and OE primary cultures obtained from skin and OE biopsies, respectively, will be used to test, in vitro, diagnosis effects on the molecular mechanisms regulating CSPG expression. Biopsy donors will be tested for olfactory functions and on psychiatric rating scales. BD subjects will be included to test whether CSPG abnormalities are specific to SZ or represent a shared feature among major psychoses.

? DESCRIPTION (provided by applicant): During postnatal brain development, the maturation of inhibitory neuronal circuits and formation of perineuronal nets (PNNs) around GABAergic interneurons result in a transition from juvenile, highly malleable, forms of plasticity to adult restricted modalities. Emerging evidence from the visual cortex points to a key role for the orthodenticle homeobox 2 (OTX2) protein in such critical developmental transitions. OTX2 internalization in neurons ensheated by PNNs induces their maturation and is necessary to open, then close, critical periods of plasticity. PNNs, specialized extracellular matrix structures surrounding distinct neuronal populations, regulate synaptic functions and plasticity, and sustain intracellular OTX2 levels in mature neurons. Converging evidence suggests that OTX2/PNN interactions may affect brain regions beyond the visual cortex, including the amygdala and prefrontal cortex (PFC). Of note, OTX2 is not produced within the adult brain. Results from rodents, and preliminary data in human, point to the choroid plexus (ChP) as a global source of OTX2, implying that altered OTX2 synthesis outside the brain parenchyma may have a profound impact on key neuronal functions. Together, these findings suggest the intriguing possibility that availability of ChP-derived OTX2 may modulate inhibitory neuronal circuits and adult forms of plasticity, a concept with far-reaching physiological, pathological and therapeutic implications. Notably, each element of this mechanism is of particular interest to the pathophysiology of schizophrenia (SZ): I) Involvement of the ChP, and the cerebrospinal fluid (CSF)/blood barrier, have been long suspected, although somewhat neglected in recent years. II) GABAergic neuron and PNN abnormalities have been reported in several brain regions, including the amygdala and PFC, brain regions involved in cognitive and emotion processing and in the pathophysiology of SZ. III) Preliminary evidence shows OTX2 decreases in the CSF, amygdala and PFC of subjects with SZ. We postulate that deficits of ChP-derived OTX2 and abnormalities affecting GABAergic interneurons and PNNs in SZ may be mechanistically linked. The proposed investigations employ a complementary, truly translational approach, combining human studies on postmortem amygdala, PFC, visual cortex, and CSF and in vitro studies on human ChP epithelial cells, with animal model approaches including conditional gene-targeting in mice, and whole-cell electrophysiology. These investigations will test the hypothesis that OTX2 originating from the ChP is pivotal to neuronal maturation and circuit plasticity in the PFC and amygdala. In particular, we postulate that OTX2 affects maturation and maintenance of PNNs surrounding GABAergic neurons and their functional and behavioral correlates. In SZ, we hypothesize that OTX2 deficits occur in the ChP and CSF as well as in the amygdala and PFC in association with PNN loss and GABAergic neuron abnormalities. The potential for systemic modulation of these mechanisms through the CP, tested in these studies, may broaden our understanding of brain plasticity and open novel therapeutic approaches to SZ.

? DESCRIPTION (provided by applicant): Disruption of thalamic connectivity figures prominently in hypotheses on the neural circuitry involved in schizophrenia. In contrast to robust evidence for such disruption from imaging studies, investigations on its underlying pathology lag far behind. Here, we focus on axonal pathways and myelinating cells within the mediodorsal nucleus (MD), a large thalamic nucleus interconnected with the frontal cortex through massive myelinated axon pathways and known to be involved in schizophrenia. We place our investigations in the context of emerging evidence for a role of chondroitin sulfate proteoglycans (CSPGs) in the pathophysiology of schizophrenia and in the regulation of brain connectivity and myelination. We have shown that the expression of CSPGs, main components of the extracellular matrix, is markedly altered in glial cells and perineuronal nets (PNNs; CSPG-enriched extracellular matrix structures surrounding distinct neuronal populations) in several brain regions of subjects with schizophrenia. The relevance of these findings to neural connectivity resides in the powerful role that CSPGs, and their interactions with oligodendrocyte progenitor cells (OPCs), play in axon guidance, fasciculation, myelination and impulse conduction. Preliminary results in human MD show axons enveloped by CSPG-enriched `axonal coats' and intimately associated with CSPG-expressing OPCs, as well as altered organization of myelinated fiber bundles in the MD of SZ subjects. Together, these considerations support the hypothesis that, in the MD of subjects with schizophrenia, altered CSPG expression in OPCs and axonal coats is associated with white matter/oligodendrocyte abnormalities and dysregulation of molecular pathways related to CSPGs and myelin biosynthesis and regulation. The proposed postmortem investigations will test this hypothesis using a combination of quantitative microscopy and proteomics/glycomics on MD samples from healthy control, schizophrenic and bipolar disorder subjects. Specific Aim 1 will elucidate the structure and composition of axonal coats, a novel extracellular matrix structure shown to surround axons in the human MD. Specific Aim 2 will use quantitative microscopy to test the hypothesis that, in the MD of subjects with schizophrenia, altered CSPG expression in OPCs and axonal coats is associated with disruption of myelination and oligodendrocyte reductions. Specific Aim 3 will use proteomics and glycomics analyses on the MD to test the hypothesis that CSPGs, myelin and molecular pathways related to their biosynthesis and regulation are disrupted in the MD of subjects with schizophrenia, thus setting microscopy studies in the context of focused hypotheses related to molecular mechanisms potentially responsible for abnormalities affecting CSPGs in axonal coats and OPCs. Specific Aim 4 will test the hypothesis that, in schizophrenia, CSPG/OPC/myelin abnormalities coexist with PNN decreases in a thalamic region, i.e. the reticular nucleus, which is particularly enriched in these ECM pericellular structures and plays a key role in gating prefrontal cortex-thalamus connectivity.

Project Summary. Alzheimer's Disease (AD) is increasingly conceived as a disorder of the synapse. Yet, despite decades of investigations, the relationship between synaptic pathology and well established ?-amyloid and tau pathology is not well understood. Emerging evidence strongly indicates that the extracellular matrix (ECM) may represent a critical element in this relationship, largely neglected thus far. The ECM, a molecular network representing 20% of the brain volume, plays a key role in neuroprotection, synaptic stabilization and plasticity, and interacts with ?-amyloid and tau. It organizes into aggregates such as perineuronal nets (PNNs), organized ECM perisynaptic structures enveloping neuronal populations and regulating synaptic stabilization and neuroprotection. Based on preliminary data showing a dramatic loss of PNN integrity and significant changes of ECM composition in AD, we put forth that the ECM may represent a key element of pathophysiological mechanisms involving synaptic, ?-amyloid and tau pathology in AD. However, little is known about expression and function of ECM and their relevance in Alzheimer Disease. Our central hypotheses for this supplemental application are that ECM abnormalities, mediated by altered expression of the transcription factor OTX2 and matrix metalloproteinases (MMPs), may represent a key link between synaptic disruption and ?-amyloid and tau pathology in AD, potentially contributing to cognitive decline in this disorder. We will test our hypotheses through two specific aims: Specific Aim 1 will expand our robust preliminary dataset on OTX2 and extracellular matrix regulation probing the dysregulation of OTX2, MMPs and ECM in AD. Specifically, this aim will test the hypothesis that a dysregulation of OTX2 and MMPs in the choroid plexus is part of an upstream signaling cascade responsible for a disintegration of PNN/ECM and subsequent synaptic loss in the brain parenchyma, specifically the amygdala. Specific Aim 2 will use machine-learning technologies to extract symptom dimensions based on the NIH RDoc definitions from tissue donors' records including a model of cognition that has been validated against neuropsychiatric testing. This process will be scaled up to develop a rich multi-dimensional phenotypic index and integrated with results from studies in Specific Aim 1. The large subject cohort to be used for these studies will allow us to further develop and validate this approach.

Transcriptional alterations of protein-coding and non-coding RNAs may be a common intermediate phenotype of both, genetic and environmental contribution to schizophrenia (SZ). Evidence from studies on micro RNAs and long non-coding RNAs suggest a profound influence of non-coding RNAs (ncRNAs) on brain function in health and disease. Beyond that, the function of the non-coding transcriptome remains for the most part unknown. Circular RNAs (circRNAs) are abundant non-coding RNAs that may have a regulatory role in transcription, translation and other cellular functions. Given their enrichment in the brain, their dynamic expression across development and in response to neuronal activation, it is conceivable that circRNAs may have a functional role in SZ. However, little is known about expression and function of circRNAs in normal human brain tissue and their relevance and function in SZ. Supported by preliminary data, our central hypotheses are that circRNAs show a region-, sex- and disease-specific expression profile in the brain and that circRNAs play a causal role in transcriptomic and proteomic changes in SZ. We will test these two hypotheses through two specific aims: Specific Aim 1 will expand our robust preliminary dataset on circLARP1B probing its region- and sex-specific expression profile. This aim will also test the hypothesis that circLARP1B is an upstream regulatory element of the RNA-binding protein LARP1B, which presumably controls SZ-specific molecular pathways through an interaction with distinct linear target mRNAs. Thus, we will expand our preliminary findings and investigate the molecular function of circLARP1B, a circRNA that shows a robust dysregulation in SZ. Specific Aim 2 will assess circRNA and linear RNA expression in the dorsolateral prefrontal cortex and anterior cingulate cortex in an extended sample of subjects with SZ and the Common Mind RNAseq database with ~1000 samples. Here, we will test the hypothesis that additional circRNAs from loci previously implicated in SZ are differentially expressed in regions known to play a key role in SZ. We will provide evidence for a functional role of circRNAs in pathways implicated in the pathophysiology of SZ. Thus, we will expand our preliminary findings and investigate the region- and sex-specific expression and function of circRNAs in pathways altered in SZ. The significance and potential impact of this proposal lays in the fact that knowledge on the expression and function of circRNAs in the healthy and diseased human brain is sparse. By generating data on region- and sex-specific circRNA expression and molecular function of circRNAs reproducibly dysregulated in SZ, we obtain preliminary data supporting an in-depth study of the molecular function of circRNAs in interaction with known genetic, epigenetic, transcriptional and proteomic alterations in SZ as part of a future R01 proposal. This will provide novel insights into the contribution of the non-coding transcriptome to the development of SZ, and may open new avenues for the cure of this devastating condition.

SUMMARY: PROJECT 5 (POSTMORTEM STUDIES OF PACAP-CRF IN HUMAN PTSD/BERRETTA) Compelling evidence indicates that corticotropic releasing factor (CRF) and pituitary adenylyl cyclase-activating polypeptide (PACAP), as well as their interactions together, make critical contributions to stress responses, anxiety, circadian rhythm regulation, and the pathophysiological mechanisms of post-traumatic stress disorder (PTSD). The underlying neural circuitry involved is poorly understood, but important clues point to the bed nucleus of the stria terminalis (BNST), amygdala (AMG), dorsal anterior cingulate gyrus (dACG), and the hypothalamus (HPTh) as critical regulators of these functions. Current evidence indicates that CRF and PACAP mechanisms that contribute to PTSD may be sex-specific, raising the possibility that the underlying brain changes and potential therapeutic targets may differ in males and females. Current and preliminary data suggest that PACAP signaling may directly affect CRF expressing cells and that circadian expression of PACAP and its cognate receptor PAC1R, and their subsequent regulation of CRF systems, may vary during the course of the day. Such variations may potentially contribute to disruptions of sleep/wake cycles associated with DSM-defined illnesses including PTSD, major depression, and anxiety disorders. Surprisingly, virtually no information is available on cell-level expression of CRF and PACAP signaling pathways in the human AMG, BNST, dACG and HPTh, their relationships to circadian rhythms, and the involvement of CRF/PACAP interactions in the neuropathology of PTSD. Our overarching hypothesis is that abnormalities affecting CRF/PACAP pathways in the BNST, AMG and dACG and HPTh contribute to the pathology of PTSD. In Aim 1, we address a critical gap of knowledge on the region-, sex- and circadian- specific expression and distribution of CRF, PACAP, and their receptors in healthy human brain. Our hypothesis is that protein and mRNA expression of CRF, PACAP, and their receptors are region- and sex-specific; in particular, we predict that PACAP receptors will show cell- and sex- specificity and expression in CRF-positive neurons, supporting the hypothesis that PACAP regulates these neurons in a sex dependent manner. In Aims 2 and 3, we examine whether?at the protein, gene expression, and cellular level?signaling pathways in the dACG, AMG, BNST and HPTh are altered in PTSD. Our hypothesis is that CRF and PACAP signaling pathways will be altered in subjects with PTSD, relative to healthy controls, with increased PACAP expression correlated with CRF signaling pathway changes, in a region-, sex-, and circadian rhythm-specific manner. We predict that increases of PACAP-positive cells and axons, reflecting increased PACAP expression locally and from hypothalamic inputs, will be accompanied by altered expression of PACAP receptors and down-stream signaling pathways in CRF-positive cells. Project 5 may identify CRF and PACAP systems and circuits as being fundamentally altered in stress-related illnesses such as PTSD, and is a key nexus of the Center that enhances, and is enhanced by, the other (preclinical, clinical) elements.

Bipolar disorder (BD) is characterized by profound affective dysregulation. Periods of aversive symptoms (depression, anxiety, decreased appetitive drive), alternate with mania (a state of enhanced appetitive drive for reward and pleasure). The clinical manifestation is heterogeneous, with diverse patterns of predominant symptoms, severity and duration. Notably, there are no current robust neurocircuit models to account for these clinical manifestations. Imaging and postmortem studies point to the amygdala, a nucleus embedded in circuits involved in threat and reward responses. Recent breakthroughs from our group and others are beginning to characterize molecularly identifiable, functionally divergent sets of amygdala neurons, which separately encode and regulate aversive and appetitive behaviors. Specifically, distinct neuronal types within the mouse amygdala promote aversive/fear responses (`FEAR-ON' neurons), vs. appetitive/reward responses (`APPT-ON' neurons). Our preliminary data using single-cell RNA sequencing show that analogous molecularly defined neuronal populations are present in human amygdala. Our overarching hypothesis is that neuronal populations impacting valence encoding and motivated behavior (FEAR-ON vs. APPT-ON neurons), are disrupted in BD, contributing to depression, anxiety and mania. What factors may regulate the functions of FEAR-ON and APPT-ON circuitry in health and disease states? An answer may lie within the distinctive molecular signatures of these neurons, consistent with their opposing functions. First, FEAR-ON and APPT-ON neurons express distinct molecular factors known to regulate fear/threat and reward processing within the amygdala, including anxiogenic (e.g. corticotropic releasing hormone [CRH]) and anxiolytic (e.g. neurotensin receptor 2 [NTSR2]) factors, respectively. Second, a well- validated distinguishing feature of amygdala FEAR-ON and APPT-ON neurons is their distinct expression pattern of Wnt/? catenin signaling molecules. This feature indicates that Wnt/? catenin pathways differentially regulate FEAR-ON and APPT-ON neurons. Pilot data also show altered expression of key molecules, including Wnt7a and CRH in the amygdala of people with BD. Our specific hypothesis is that cell-specific FEAR-ON and APPT- ON molecular factors modulating stress/anxiety and reward/appetitive behaviors are altered in BD, and that disruption of Wnt/? catenin pathways contributes to distinct abnormalities FEAR-ON and APPT-ON neurons. Human postmortem studies combining single-cell RNAseq, multiplex mRNA/protein cell labeling and quantitative analyses of RDoC clinical domains will test the hypothesis that quantifiable clinical `fingerprints' in BD are predictive of distinct patterns of molecular changes in FEAR-ON and APPT-ON neurons (Aims 1 and 2). Causal manipulation in mouse genetic models will mechanistically test the hypothesis that a disruption of Wnt signaling causally alters expression of reward- and stress- related molecules in circuits linking deep amygdala nuclei to the CE and nucleus accumbens (Aim 3).