Dorothy P. Schafer, Ph.D. - US grants

DESCRIPTION (provided by applicant): During development, neural circuitry undergoes a remodeling process in which excess synapses are eliminated or pruned and the remaining synapses are strengthened. While it is clear that developmental synaptic remodeling is an activity-dependent process whereby weaker or less-active synapses are selectively eliminated, the precise molecular mechanisms have not been elucidated. Recently, our laboratory discovered that components of the classical complement cascade (C1q and C3) were necessary for synapse elimination in the developing and diseased central nervous system (CNS)(Stevens et al., 2007). In addition, C1q protein was specifically expressed in neurons and localized to synapses at ages and brain regions consistent with active synaptic pruning. One of the major questions arising from these findings is by what mechanism is complement facilitating synapse removal. The complement cascade is traditionally associated with the innate immune system in which complement components coat or opsonizes debris (e.g. pathogens, cell corpses, etc.) for removal. A canonical pathway for complement-opsonized debris removal is through phagocytosis by cells that express receptors for complement proteins Similar to the innate immune system, we suggest that complement is acting in the developing nervous system to tag weak synapses for removal by activated microglia, the primary phagocytic cell in the CNS. To characterize the role of microglia in synaptic pruning (Aim 1), we will test the capacity of microglia to phagocytose synaptic endings (subaim 1a) using various imaging techniques (e.g in vivo live imaging, etc.). In addition, we will deterimine if microglia-mediated synaptic removal is a complement-dependent process (subaim 1b) by assessing pruning deficits in complement receptor KO mice and microglia physiology in C3 and C1q KO mice. The final component of this proposal (Aim 2) is to test the pervasive nature of microglia-mediated synapse remodeling throughout the developing CNS. Previous and proposed work has been done in the developing retinogeniculate system;therefore, I will test whether microglia-mediated synaptic remodeling also occurs in another brain region, the cerebellum (subaim 2a). Furthermore, I will determine if this is a complement-dependent process (subaim 2b). In addition to normal development, deficits in this pruning process have been implicated in a broad range of developmental (e.g. autism) and psychiatric (e.g. schizophrenia) disorders (Innocenti et al., 2003;Pardo et al., 2005;Vargas et al., 2005;Woo and Crowell, 2005). Furthermore, there is evidence that synapse removal occurs during early stages of neurodegenerative disease (e.g. Alzheimer's)(Selkoe, 2002). Therefore, studying basic mechanisms of normal development provides both a basic understanding of a biological mechanism but also a framework for studying mechanisms of CNS disease and development of therapeutic strategies. PUBLIC HEALTH RELEVANCE: Following birth, in a process called synaptic pruning, many connections between neurons and their nervous system targets are removed leaving behind a highly refined circuitry. Abnormalities in this pruning process are associated with developmental (e.g. autism) and psychiatric (e.g. schizophrenia) disorders as well as early stages of neurodegenerative disease (e.g. Alzheimer's disease). Thus, in addition to elucidating basic mechanisms of nervous system development, understanding synaptic pruning has implications for development of therapeutic strategies targeted toward treatment and/or prevention of nervous system diseases. .

During the first year of my K99, I received an offer and accepted a positon as an Assistant Professor in the Neurobiology Department at the University of Massachusetts Medical School. As a result, I am applying for transition to the ROO phase where I will be investigating microglia-specific mechanisms driving activity-dependent synaptic remodeling in the developing brain. The ultimate goal is to apply these mechanisms to neurodevelopmental and neuropsychiatric disorders. Immature synapses form a crude wiring diagram that must remodel during development to achieve the precise connectivity characteristic of the mature nervous system. While It is clear that neural activity drives synaptic remodeling, the underlying mechanisms are not fully understood. We recently identified that microglia participate in synaptic remodeling by engulfing synaptic elements in an activity-dependent manner. However, it is unknown whether microglia engulf intact synapses or whether this is a non-cell autonomous event. In addition, it is known that microglia change their motility and interactions with synapse in response to neural activity but the underlying molecular mechanisms and functional consequences of these responses are unknown. As a result the goals of this proposal are to: 1) Aimi: Test the hypothesis that microglia are actively sensing, interacting with, and phagocytosing intact synapses in response to changes in neural activity. The laboratory is uniquely positioned to address this question with expertise in imaging microglia by 2-photon In vivo live Imaging, a skill acquired during the K99 phase. 2) Aim 2: Dissect the molecular mechanisms underlying microglia responses to changes in neural activity. A number of candidate microglia-specific molecules and cytokines have been identified and validated, including IL-12. Thus, activity-dependent microglia motility and interactions with synapses will be assessed In mice deficient in genes of interest. 3) Aim 3: Determine the functional significance of activity-dependent microglial responses by testing the hypothesis that these responses regulate the development of synaptic circuit structure and function in mice with deficiencies in genes previously identified and validated.

The goal of this proposal is to identify novel, activity-dependent mechanisms by which microglia interact with neural circuits and regulate plasticity of synapses. Neuronal activity is a critical regulator of synaptic architecture. Both in the developing brain and during adulthood, changes in activity drives the formation of new synapses, elimination of less active synapses, and maintenance of more active synapses. Despite over 50 years of research, the mechanisms by which neuronal activity drives changes in synaptic connectivity remain to be fully deciphered. Unexpectedly, we discovered that microglia, resident central nervous system (CNS) macrophages, are regulated by neuronal activity and engulf and prune away less active synapses in the healthy developing visual system. Further studies have identified similar microglial synaptic pruning functions in other brain regions as well as functions in synaptogenesis and neural transmission. These data compel us to consider microglia as key cellular regulators of synaptic connectivity and evoke the paradigm shifting possibility that disruptions in microglial function result in synaptic defects (density, transmission, etc.) in a wide range of neurological disorders ranging from autism to Alzheimer?s disease. Here, we propose that changes in neuronal activity elicit activity-dependent genetic programs in microglia necessary for regulating synaptic connectivity. We will focus our analyses on microglia-mediated synaptic pruning. In support of this hypothesis, we recently silenced excitatory neurons in the adult brain using designer receptors exclusively activated by designer drugs (DREADDS) followed by microglia-specific translating ribosomal affinity purification and RNA sequencing (TRAP-seq). Sequencing results revealed rapid induction of phagocytic pathways in microglia following acute neuronal silencing. We will now use our newly developed in vitro assays to molecularly dissect the function of these phagocytosis-related genes in microglia-dependent synaptic engulfment (Aim 1). To identify new molecules relevant to activity-dependent synaptic pruning during brain development, we will also perform microglia-specific TRAP-seq in a paradigm we recently identified to robustly upregulate microglial synaptic engulfment in response to changes in sensory experience in vivo?whisker deprivation-induced synapse elimination in the mouse barrel cortex (Aim 2). We will again use our in vitro assays to assess microglial synaptic engulfment and validate gene hits. Using our combined expertise in cell- specific TRAP-seq (A. Schaefer lab) and microglial synaptic pruning (D. Schafer lab), we have a unique and powerful advantage to identify novel mechanisms regulating activity-dependent microglia-synapse interactions necessary for modulating synaptic connectivity. Long term, we will collaborate to explore these mechanisms in vivo and determine whether mechanisms are dysregulated in neuropsychiatric disorders with known defects in microglial inflammatory state and synaptic connectivity.

The goal of this proposal is to determine how microglia and sensory experience integrate to remodel synapses into precise, functional brain maps. Trillions of synapses form highly precise topographic maps in the brain representing each part of the body. These maps are shaped and maintained by sensory experience (vision, touch, etc.), including elimination of less active synapses and formation and maintenance of other synapses. Despite over 50 years of research, the underlying mechanisms by which experience dictates removal or maintenance of specific synapses still remains an open question. We made the initial exciting and unexpected observation that microglia, the resident CNS macrophages, engulfed and eliminated a subset of less active synapses in the developing retinogeniculate system. Further, reducing microglia-mediated engulfment of synapses by 50% (complement receptor 3 KO) resulted in sustained increases in retinogeniculate synapse number. This work established a new way of thinking about synaptic remodeling and inspired several important new questions: Is microglia-mediated synaptic remodeling necessary for achieving functional circuits? Do microglia remodel synapses in the adult brain? How does neural activity regulate microglia-mediated synaptic remodeling? The retinogeniculate system was limiting for addressing these questions. We required a robust system for studying synaptic remodeling that involved plasticity of synapses throughout life, tractable assays for measuring function, and a topographic arrangment with high spatial and temporal resolution. The mouse barrel cortex fit all these critera and will enable us to test the hypothesis that experience regulates microglia to shape developing and mature syanpses into functional brain circuits. Our new preliminary data show for the first time that microglia engulf excitatory thalamocortical (TC) synapses in the developing barrel cortex and following sensory deprivation (whisker removal) in the neonate. Further, mice deficient in microglia (colony stimulating factor 1 receptor KO; CSF1R KO) have defects in the development of approriate barrel architecture and TC input elimination following whisker deprivation is completely blocked in mice deficient in a microglia-specific chemokine receptor (fractalkine receptor KO, CX3CR1 KO). We will now use a combination of high resolution static and functional imaging and molecular biology in the mouse barrel cortex to: 1) Determine whether microglia sculpt developing cortical circuits into functional brain maps (Aim 1). 2) Determine whether microglia regulate experience-dependent plasticity of cortical maps in the neonate and adult (Aim 2). 3) Identify how microglia remodel synapses in response to changes in neural activity (Aim 3). Answers will uncover new mechanisms regulating how sensory experience regulates the development of structural and functional brain maps, will identify new ways to achieve plasticity in the adult brain, and will provide new mechanistic insight into how synapses remodel in multiple contexts (development, learning and memory, disease, etc.).

Schafer, Dorothy P. Project Summary Multiple sclerosis (MS) is an inflammatory demyelinating disease of the central nervous system (CNS), which has a profound, currently intractable, neurodegenerative component--a large, unmet clinical need. In many neurodegenerative diseases, one of the earliest degenerative events is synapse dysfunction and loss. There is also synapse loss in MS, but the underlying molecular mechanism(s) remains an open question. The overall hypothesis of this proposal is that complement-dependent signaling underlies synapse loss in demyelinating disease in a subset of vulnerable neurons. This is largely based on our initial findings in the developing retinogeniculate circuit demonstrating that classical complement cascade proteins C1q and C3 localize to synapses and that phagocytic microglia engulf and eliminate synapses via the C3 receptor, complement receptor 3 (CR3). Strikingly, we have new evidence that a subset of retinogeniculate synapses are also engulfed by microglia, leading to synapse loss, in MS and in multiple MS-relevant animal models of demyelinating disease (e.g. non-human primate and mouse experimental autoimmune encephalomyelitis (EAE) models). We further identified that this synapse loss can occur early prior to demyelination, axon degeneration, or cell death, but is coincident with peripheral immune cell infiltration, reactive microgliosis, and increased levels of complement C1q and C3. However, unlike development, C3, but not C1q, is localized to synapses. Finally, inhibiting C3 specifically at retinogeniculate synapses in mouse EAE prevents microglial synapse engulfment, synapse loss, and visual dysfunction. These experiments establish C3 and microglia as key regulators of synapse loss in MS-relevant demyelinating disease and open up several new questions that we will explore: 1) What cells produce complement necessary for synapse elimination in demyelinating disease (Aim 1)? 2) Does microglial complement receptor CR3 regulate synapse loss in demyelinating disease (Aim 2)? 3) Which RGCs are most vulnerable to complement-mediated synapse elimination and later degeneration (Aim 3)? To address these questions, we will continue to use the retinogeniculate circuit. This is a highly tractable and powerful system for studying synaptic changes and it is highly relevant to MS, where inflammation of the optic nerve (i.e. optic neuritis) occurs in upwards of 50% of patients and results in prolonged, often permanent, visual dysfunction. We will now use a combination of cell-specific molecular genetics and high-resolution imaging of retinogeniculate synapses in the mouse EAE model to molecularly dissect synapse loss in inflammatory demyelinating disease. Results could uncover novel targets aimed at slowing or preventing neurodegeneration in MS, which could be broadly applicable to other neurodegenerative disease with synapse loss and neuroinflammation (Alzheimer?s disease, frontotemporal dementia, etc.).