Anthony J. Koleske - US grants

The long-term goal of this research is to understand how the Abl and Arg nonreceptor tyrosine kinases regulate the differentiation, morphogenesis, and function of neurons in mice. Our first major aim is to understand how Abl and Arg regulate the organization of actin filaments in cells of the developing neural tube. Abl and Arg are the only nonreceptor tyrosine kinases that contain actin-binding domains, but the precise role of these kinases in regulating the actin cytoskeleton is unclear. We have found that actin filaments are disordered in abl-/-arg-/- neuroepithelial cells. As a consequence, these embryos suffer from serious neural tube defects. We will determine which of the proteins that regulate actin filament formation or localization in neuroepithelial cells mediate control of the actin cytoskeleton by Abl and Arg. Our second major aim is to understand the roles for Abl and Arg in neuronal differentiation and morphogenesis. We have shown that the number of early hindbrain neurons is severely reduced in abl-/-arg-/-embryos, but it is unclear which stages of neuronal development require Abl and Arg. We will determine whether abl- /-arg-/-neuronal precursor cells proliferate, migrate, and differentiate normally. In Drosphila, D-Abl acts within neurons to control axon outgrowth and fasciculation, but it is not known which axon guidance pathways are regulated by D-Abl. We will measure the extension rate of neurites from abl-/-arg-/-neurons on a select group of extracellular matrix molecules to determine which growth cone responses require Abl and Arg. Our third major aim is to understand the function of Arg in the mature brain. Our observations that Arg is concentrated in synapses and the arg-/-mice exhibit behavioral abnormalities suggest that Arg performs an important function in mature neurons. In order to obtain clues to this function, we will first pinpoint the localization of Arg within synapses. Next, we will identify regulators and substrates of Arg in fractioned brain extracts to determine which cellular processes Arg regulates. Finally, following up on a clue to Arg function obtained from behavioral studies, we will perform physiological tests to understand why arg-/-cochlear hair cells do not function properly. The information gained from these studies will allow us to pose sharper mechanistic questions aimed at understanding the biochemistry of signal transduction by Abl and Arg.

DESCRIPTION (provided by applicant): The Abl and Arg nonreceptor tyrosine kinases regulate neuronal migration and morphogenesis in the developing mouse brain. Abl and Arg relay information from cell surface receptors to promote cytoskeletal rearrangements. The goal of this proposal is to understand the molecular mechanisms by which Abl and Arg regulate cytoskeletal structure and function in developing neurons. Our first aim is to elucidatetheroles for Abl and Arg in neuronal development. Our morphometric analysis of dye-filled neurons has shown that dendrite arbors are reduced in arg-/- hippocampal neurons relative to wild type. Neurons are more tightly packed in the cerebral cortices of brain-specific-abl-/-arg-/- double knockout mice. This phenotype may reflect a severe deficit in dendrite arborization. We also find that a subset of cerebellar granular neurons fail to migrate normally in the brain-specific-abl-/-arg-/- mice. We will study neuronal differentiation in knockout mice and in explanted neurons from these mice to examine how mutations in genes encoding components of Abl- and Arg-signaling pathways affect dendrite morphogenesis and cerebellar granular neuron migration. Our second aim is to understand how Arg regulates cytoskeletal structure in developing neurons. Arg overexpression leads to increased membrane ruffling and decreased cell motility in fibroblasts and increased neurite branching in cultured cortical neurons. We have identified two distinct pathways by which Arg can regulate cytoskeletal structure: 1) the Arg C-terminal half can bundle F-actin and crosslink F-actin bundles to microtubules; and 2) Arg can phosphorylate and activate the p190 Rho GTPase-activating protein (p190RhoGAP) to inhibit the cytoskeletal regulator Rho. We will express Arg and/or p190RhoGAP mutants in cultured fibroblasts to dissect how these two pathways contribute to the regulation of cytoskeletal structure and function. Our third aim is to determine how Abl and Arg kinase activities are regulated in developing neurons. We have shown that in vitro phosphorylation of Abl and Arg at several sites can regulate their kinase activities, but it is unclear how Abl and Arg become phosphorylated in developing neurons. Progress in this area has been limited because the cell surface receptors that regulate Abl and Arg kinase activity in developing neurons have not been identified. Genetic and biochemical experiments have identified a small number of candidate receptors (the Robo receptors, TrkB, integrin ?5) that are likely to regulate Abl and Arg kinase activity in developing neurons. We will determine whether these receptors regulate Abl and Arg kinase activity and examine whether Abl and Arg mediate migratory or morphogenetic signals from these receptors.

[unreadable] DESCRIPTION (provided by applicant): Protein tyrosine kinases are essential mediators of axon and dendrite guidance cues. In most cases, the protein targets that mediate the downstream actions of these kinases remain to be identified. The goal of our research is to adapt a biochemical technique, the "bump-hole" approach, to identify tyrosine kinase substrates in the developing brain. The Abl and Arg nonreceptor tyrosine kinase are required for cortical dendrite branch formation in the developing mouse brain. As a proof of concept, we propose to identify and functionally characterize substrates of Abl and Arg in the developing brain. Our first aim is to identify Abl and Arg substrates in developing neurons. We have synthesized a chemically altered form of ATP that contains a benzyl group "bump" on the adenosine ring of ATP (bumped ATP) and engineered a corresponding "hole" in Abl and Arg by removing a bulky amino acid residue from the ATP-binding pocket. Importantly, these altered specificity- (as-) forms of Abl and Arg can utilize the bumped ATP, whereas kinases present in brain extracts cannot. We will selectively label Abl and Arg substrates in vitro by incubating brain extracts with as-Abl or as-Arg in the presence of gamma32P-labeled bumped ATP. We will purify labeled substrates using biochemical fractionation and anti-phosphotyrosine affinity resins and identify them by mass spectrometry. We will use in vitro and cell-based phosphorylation assays to confirm that the identified candidates are bona fide Abl/Arg substrates. Our second aim is to determine whether the substrates regulate axon or dendrite morphogenesis. We will examine the localization of each substrate in cultured wild type and Arg-deficient cortical neurons and determine how this localization is influenced by integrin-mediated adhesion or elevated Arg kinase activity. We will test whether overexpression of the substrate alone or in combination with Arg affects axon or dendrite branching in cultured cortical neurons. We will determine how RNAi-mediated reduction of the regulator influences axon and dendrite branching following adhesion or Arg overexpression. These studies should prove the utility of using the bump-hole approach to identify tyrosine kinase targets in developing neurons. Identifying these substrate will expand the "toolkit" available to dissect the molecular mechanisms of axon and dendrite morphogenesis. [unreadable] [unreadable]

[unreadable] DESCRIPTION (provided by applicant): Protein tyrosine kinases are essential mediators of axon and dendrite guidance cues. In most cases, the protein targets that mediate the downstream actions of these kinases remain to be identified. The goal of our research is to adapt a biochemical technique, the "bump-hole" approach, to identify tyrosine kinase substrates in the developing brain. The Abl and Arg nonreceptor tyrosine kinase are required for cortical dendrite branch formation in the developing mouse brain. As a proof of concept, we propose to identify and functionally characterize substrates of Abl and Arg in the developing brain. Our first aim is to identify Abl and Arg substrates in developing neurons. We have synthesized a chemically altered form of ATP that contains a benzyl group "bump" on the adenosine ring of ATP (bumped ATP) and engineered a corresponding "hole" in Abl and Arg by removing a bulky amino acid residue from the ATP-binding pocket. Importantly, these altered specificity- (as-) forms of Abl and Arg can utilize the bumped ATP, whereas kinases present in brain extracts cannot. We will selectively label Abl and Arg substrates in vitro by incubating brain extracts with as-Abl or as-Arg in the presence of gamma32P-labeled bumped ATP. We will purify labeled substrates using biochemical fractionation and anti-phosphotyrosine affinity resins and identify them by mass spectrometry. We will use in vitro and cell-based phosphorylation assays to confirm that the identified candidates are bona fide Abl/Arg substrates. Our second aim is to determine whether the substrates regulate axon or dendrite morphogenesis. We will examine the localization of each substrate in cultured wild type and Arg-deficient cortical neurons and determine how this localization is influenced by integrin-mediated adhesion or elevated Arg kinase activity. We will test whether overexpression of the substrate alone or in combination with Arg affects axon or dendrite branching in cultured cortical neurons. We will determine how RNAi-mediated reduction of the regulator influences axon and dendrite branching following adhesion or Arg overexpression. These studies should prove the utility of using the bump-hole approach to identify tyrosine kinase targets in developing neurons. Identifying these substrate will expand the "toolkit" available to dissect the molecular mechanisms of axon and dendrite morphogenesis. [unreadable] [unreadable]

DESCRIPTION (provided by applicant): Neural circuits do not develop properly in neurodevelopmental disorders and degrade prematurely in neurodegenerative disorders. We have shown that dendritic spine morphogenesis and later synapse and dendrite stability requires the Abl-related gene (Arg) nonreceptor tyrosine kinase, which acts downstream of integrin adhesion receptors to mediate changes in cytoskeletal structure. We seek to understand how integrins signal through Arg and its effectors to control the formation and maintenance of neural circuitry. Our first aim is to elucidate the roles for Arg in dendritic spine morphology, synapse stability, and dendrite maintenance. Dendrites and synapses develop normally through postnatal day 21 in the arg-/- mouse cortex and hippocampus, but dendritic spines do not mature properly, leading to synapse and dendrite loss and behavioral deficits by postnatal day 42. We will use electrophysiology and analyze three-dimensional reconstructions of synapses and dendrites in arg-/- mice to examine how the loss of Arg-signaling pathways compromises the formation, stability, and function of synapses and dendrites. We will also use a conditionally inactivatable arg allele and an inducible arg transgene to determine when Arg signaling is required for proper synapse development and to protect against synapse loss and dendritic degeneration. Our second aim is to understand how integrins activate Arg to regulate synapse and dendrite stability. Our work has shown that integrins act through Arg to mediate changes in cytoskeletal structure, but we do not understand how Arg is recruited to integrin heterodimers to achieve kinase activation in vivo. We will test the hypothesis, supported by preliminary data, that integrin 21 or 23 cytoplasmic tails bind directly to Arg to mediate kinase activation. Integrins containing 21 or 23 subunits regulate synapse formation and dendrite stability in vivo, but it is unclear which specific integrin heterodimers act through Arg to regulate synapse and dendrite maintenance. We will monitor Arg signaling pathways and analyze synapse and dendrite structure in integrin 21 and 23 mutant mice to determine which integrins act through Arg to control dendritic spine morphogenesis and synapse/dendrite stability. Our third aim is to determine how Arg signals through its effectors to regulate synapse and dendrite stability. Arg is required for proper dendritic spine morphogenesis in vivo, but we do not understand how Arg coordinates the cytoskeletal changes required for these processes. Our biochemical studies have identified several substrates (p190RhoGAP, cortactin, myosin IIB) through which Arg acts to promote changes in cytoskeletal structure. We will examine how integrin signaling through Arg affects the distribution of these Arg substrates in cultured cortical neurons. We will also test whether RNAi knockdown of the substrates affects dendritic spine structure and synapse and dendrite stability in established hippocampal neuronal cultures. PUBLIC HEALTH RELEVANCE: Neural circuits do not develop properly in neurodevelopmental disorders, such as mental retardation, and degrade prematurely in neurodegenerative disorders, such as Alzheimer's Disease. Defects in synapse function and/or reductions in synapse number lead to the loss of dendrite segments and degeneration of neural circuits. We will study a biochemical pathway that regulates synapse morphogenesis and function and protects against degeneration of neural circuits in the brain.

DESCRIPTION (provided by applicant): Complications from metastasis are the leading cause of mortality from breast cancer. Invasive cancer cells use F-actin-rich protrusions called invadopodia to degrade extracellular matrix (ECM) barriers to migration. We have shown that integrin ¿1-mediated adhesion stimulates the Arg nonreceptor tyrosine kinase to interact with the actin polymerization regulators cortactin, N-WASp, and Vav2 at nascent sites of F-actin-mediated cell edge protrusion in non-cancerous cells. Each of these proteins localizes to invadopodia and is required for invadopodial function. Indeed, we find that Arg-mediated cortactin phosphorylation triggers actin polymerization within human breast cancer cell invadopodia, leading to their stabilization and acquisition of matrix-degrading activity. We will elucidate the mechanisms by which the integrin ¿1:Arg:cortactin:N-WASp:Vav2 axis controls invadopodia function during breast cancer invasion and metastasis and screen for inhibitors of key interactions between these regulators as lead compounds for drug development. Our first aim is to understand how Arg is localized and regulated within invadopodia. Arg uses distinct domains to bind directly to F-actin, microtubules, and integrin ¿1. Our preliminary work strongly suggests that these interactions regulate localization and activation of this key regulator at invadopodia. We will use RNAi knockdown of Arg and integrin ¿1, rescue with interaction-defective Arg and integrin ¿1 mutants, and use quantitative immunofluorescence and matrix degradation assays to determine which of these interactions mediates Arg localization to and activity within invadopodia in invasive human breast cancer cells. Our second aim is to identify the interactions most critical for invadopodia function and screen for inhibitors of these interactions. In addition to binding cortactin, Arg uses a distinct domain to bind and activate N-WASp. Arg-mediated cortactin phosphorylation also promotes its binding to Vav2, a regulator of actin polymerization. We hypothesize that Arg coordinates the activation and assembly of cortactin, N-WASp, and Vav2 within invadopodia to trigger Arp2/3 complex-mediated actin polymerization. We will use a knockdown/complementation approach similar to Aim 1 to identify which interactions are most critical for invadopodial function. We will also perform high throughput small molecule screens to identify compounds that disrupt key interactions between these proteins and test their ability to block breast cancer cell invasiveness. Our third aim is to test how disruption of key invadopodial actin regulators affect breast cancer invasion and metastasis. Invadopodia mediate penetration of matrix barriers in vitro, but whether and how they mediate breast cancer invasiveness in vivo has not been rigorously tested. We will use scid mouse xenograft and MMTV-polyoma middle T breast cancer models to determine how disrupting these invadopodial regulators affects breast cancer cell invasion and metastasis in vivo.

DESCRIPTION (provided by applicant): Integrin adhesion receptor signaling through the Abl and Arg nonreceptor tyrosine kinases activates several major cytoskeletal effectors to coordinate changes in cytoskeletal structure. Abl/Arg-mediated signaling events are essential for nervous and immune system development and function, while inappropriately elevated Abl/Arg signaling is associated with several cancers. Despite these critical physiological roles, the mechanisms by which integrins activate Abl and Arg are poorly understood. Based on extensive preliminary evidence, we hypothesize a multistage model for integrin ¿1 activation of Arg kinase activity that we will test in three Aims. Our first aim is to identify and characterize he interfaces that mediate initial Arg recruitment to the integrin ¿1 tail. We find that a sequence in the integrin ¿1 tail not known to interact with other integrin-binding proteins binds directly to te isolated Arg kinase domain and activates Arg kinase activity. We will identify the interfaces on integrin ¿1 and Arg that mediate these interactions and use mutagenesis and binding assays to determine the contributions of specific residues to this interaction. We have expressed, purified and obtained crystals of the Arg kinase domain. We will now use X-ray crystallography to determine the structure of the Arg kinase domain in complex with integrin ¿1 tail peptides. Our second aim is to investigate the dual roles that the Arg SH2 domain plays in both retainment and reinforcement of Arg kinase activation. Arg can phosphorylate integrin ¿1 tail directly, thereby creating a binding site for the Arg SH2 domain, which we hypothesize retains Arg and promotes optimal kinase activation. We will use biochemistry and X-ray crystallography to elucidate an atomic resolution understanding of how the Arg SH2 domain and kinase domains bind coordinately to the phosphorylated integrin ¿1 tail. As part of this work, we will test the hypothesis that coordinated binding to the phosphorylated integrin ¿1 tail promotes cyclin-like Arg SH2 domain binding to the Arg kinase domain N-terminal lobe to reinforce kinase activation. Our third aim is to investigate the role of the recruitment, retainment, and reinforcement model in regulating Arg signaling and its control of dendritic spine and dendrite stability in vivo. Interin signaling through Arg and its substrate p190RhoGAP negatively regulates the RhoA GTPase, a major antagonist of synapse, dendritic spine, and dendrite stability. Genetic studies suggest that integrin ¿1 interacts functionally with Arg to mediate dendritic spine and dendrite stabilization i the adolescent mouse brain. We will generate integrin ¿1 and Arg mutants deficient in interactions that mediate Arg kinase recruitment, retainment, or reinforcement, and reconstitute them into integrin ¿1- or Arg-deficient fibroblasts and neurons. We will employ biochemical, FRET, and cell-based assays to determine how these interfaces contribute to integrin ¿1: Arg interactions and Arg- mediated signaling in fibroblasts and neurons, and control of dendritic spine and dendrite stability in neurons.

ABSTRACT Complications from metastasis are the leading cause of mortality from breast cancer. Invasive cancer cells use F-actin-rich protrusions called invadopodia to degrade extracellular matrix (ECM) barriers to migration. We have shown that integrin -mediated adhesion stimulates the Arg nonreceptor tyrosine kinase to interact with the actin polymerization regulators cortactin, N-WASp, and Vav2 at nascent sites of F-actin-mediated cell edge protrusion in non-cancerous cells. Each of these proteins localizes to invadopodia and is required for invadopodial function. Indeed, we find that Arg-mediated cortactin phosphorylation triggers actin polymerization within human breast cancer cell invadopodia, leading to their stabilization and acquisition of matrix-degrading activity. We will elucidate the mechanisms by which the integrin :Arg:cortactin:N-WASp:Vav2 axis controls invadopodia function during breast cancer invasion and metastasis and screen for inhibitors of key interactions between these regulators as lead compounds for drug development. Our first aim is to understand how Arg is localized and regulated within invadopodia. Arg uses distinct domains to bind directly to F-actin, microtubules, and integrin . Our preliminary work strongly suggests that these interactions regulate localization and activation of this key regulator at invadopodia. We will use RNAi knockdown of Arg and integrin , rescue with interaction-defective Arg and integrin  mutants, and use quantitative immunofluorescence and matrix degradation assays to determine which of these interactions mediates Arg localization to and activity within invadopodia in invasive human breast cancer cells. Our second aim is to identify the interactions most critical for invadopodia function and screen for inhibitors of these interactions. In addition to binding cortactin, Arg uses a distinct domain to bind and activate N-WASp. Arg-mediated cortactin phosphorylation also promotes its binding to Vav2, a regulator of actin polymerization. We hypothesize that Arg coordinates the activation and assembly of cortactin, N-WASp, and Vav2 within invadopodia to trigger Arp2/3 complex-mediated actin polymerization. We will use a knockdown/complementation approach similar to Aim 1 to identify which interactions are most critical for invadopodial function. We will also perform high throughput small molecule screens to identify compounds that disrupt key interactions between these proteins and test their ability to block breast cancer cell invasiveness. Our third aim is to test how disruption of key invadopodial actin regulators affect breast cancer invasion and metastasis. Invadopodia mediate penetration of matrix barriers in vitro, but whether and how they mediate breast cancer invasiveness in vivo has not been rigorously tested. We will use scid mouse xenograft and MMTV-polyoma middle T breast cancer models to determine how disrupting these invadopodial regulators affects breast cancer cell invasion and metastasis in vivo.

DESCRIPTION (provided by applicant): Dendritic spines and their associated synapses become prematurely destabilized in psychiatric and neurodegenerative diseases. Proper control of the actin cytoskeleton is critical for the long-term structural stability of dendritic spines, bt currently little is known about the molecules and mechanisms that confer long-term structural stability on spines and the field remains understudied. We discovered that loss of integrin ?3?1 signaling through the Abl2/Arg nonreceptor tyrosine kinase causes widespread dendrite arbor loss and dendritic spine destabilization. Even though Arg inhibits the RhoA GTPase to stabilize dendrite arbors, this mechanism does not impact spine stability, raising the fundamental question of how Arg stabilizes spines. We provide evidence that Arg directly binds and stabilizes actin filaments and also regulates the binding and actions of the actin regulators cortactin and Arp2/3 complex on actin filaments. We also find that Arg-mediated recruitment of cortactin to dendritic spines is crucial for spine stability. Our proposal will test the highly innovative hypothesis that Arg interacts physically and functionally with actin filaments and actin regulatory proteins to directly regulate actin dynamics and thereby stabilize dendritic spines. Our first aim will elucidate how Arg:cortactin interactions control actin dynamics. We find that Arg binding to actin filaments stabilizes them from depolymerization. Arg binding also recruits the actin-binding protein cortactin, which stabilizes actin filaments and increases actin branch formation by Arp2/3 complex. We will use total internal reflection microscopy to observe single filaments and to measure how Arg and cortactin affect actin filament stability, Arp2/3 complex-mediated branch formation, and cofilin-mediated actin filament severing. We will use mutants of these proteins that do not interact with each other or with actin filaments to identify which protein:protein interaction interfaces are critical for effects on actin dynamics. These studies will reveal how Ar and cortactin affect actin filament stability, branching, and turnover. Our second aim will determine how Arg and cortactin modulate spine stability via effects on actin dynamics. We find that knockdown of Arg in neurons results in the loss of cortactin from spines and triggers their destabilization. We hypothesize this destabilization is due to the disruption of normal actin dynamics in spines. Knockdown of Arg or cortactin in established hippocampal neuron cultures compromises dendritic spine stability. These deficits can be quantitatively rescued by re-expression of shRNA-resistant versions of Arg or cortactin, respectively. Employing our collection of Arg and cortactin mutants, we will test how mutational disruption of key interaction interfaces in these proteins affects dendritic spine shape and stability. We will use fluorescence recovery after photobleaching (FRAP) of GFP-actin in spines to reveal how manipulations of Arg and cortactin function affect actin dynamics in spines and determine how this relates to the effects of these proteins on actin biochemistry and spine stability.

DESCRIPTION (provided by applicant): The development, plasticity, and stability of dendrites and dendritic spines are defective in autism, mental retardation, stroke, and psychiatric diseases. Mutations or reduced levels of heterotrimeric laminin extracellular matrix proteins are associated with these human brain disorders. We provide evidence that neuron-specific ablation of the laminin alpha5 subunit in mice increases spine densities, destabilizes dendrite branches, and compromises normal synaptic transmission and animal behavior. We propose to elucidate the mechanisms by which laminin alpha5 and a new putative laminin alpha5 receptor we have discovered regulate dendrite and dendritic spine development and function. We will use complementary in vivo imaging, electrophysiological, biochemical, and genetic approaches to achieve the following aims: Aim 1. Determine how laminin alpha5 regulates development, plasticity, and function of dendrites, dendritic spines, and synapses. Our data strongly suggest that laminin alpha5 controls dendrite branch and dendritic spine dynamics. We will use transcranial two-photon microscopy of dendrites in the somatosensory cortex, alone and in combination with sensory input manipulation, to reveal how the loss of laminin alpha5 impacts branch and spine dynamics during development and activity-driven plasticity. We will also use electron microscopy and whole cell recording to test the hypothesis that laminin alpha5 regulates synaptic transmission by controlling the structure, transmission properties, and plasticity of individual synapses. Aim 2. Elucidate the composition, origin, and timing of function of alpha5-containing laminins in dendrite and spine development. We do not know which laminin beta and gamma chains partner with laminin alpha5, where they are produced, or when they act. We will use biochemical and genetic knockout approaches to identify laminin beta and gamma chains that associate with laminin alpha5 in neurons to regulate dendrite and spine development. We will also inactivate laminin alpha5 in specific cell types using inducible Cre transgenes to determine where and when laminin alpha5 is required to regulate dendrite and dendritic spine development. Aim 3. Characterize SIRPalpha function in laminin alpha5-mediated dendrite and dendritic spine development. We have shown that the integrin alpha3beta1 receptor for laminin alpha5 mediates dendrite branch stability, but our genetic analysis indicates that other receptors are essential to mediate the effects of laminin alpha5 on dendritic spine development. Our data strongly suggest that the Signal Regulatory Protein alpha (SIRPalpha) transmembrane receptor serves as a novel laminin alpha5 receptor in the control of spine development. We will use cell adhesion assays and in vitro binding assays with purified proteins to identify which domains in SIRPalpha and alpha5-laminins mediate these interactions. We will test how excitatory neuron-specific ablation of SIRPalpha function alone or in combination with integrin alpha3beta1 affects dendrite and spine development and synaptic function and plasticity.

ABSTRACT Proper control of the actin cytoskeleton is critical for long-term stability of spines, which are destabilized prematurely in psychiatric and neurological disorders. Dendritic spines contain at least two distinct pools of filamentous- (F-) actin, a small stable pool that turns over slowly and resides within the central core of the spine, and a larger dynamic pool that extends from the central core to the spine periphery. What mechanisms control these distinct F-actin pools, how the pools interface with the neurotransmission machinery, and how they contribute to dendritic spine plasticity and stability are fundamental unresolved questions in the field. Our lab discovered that disruption of the Abl2/Arg nonreceptor tyrosine kinase causes widespread postnatal dendritic spine destabilization and synapse loss. In addition to being a kinase, Arg binds F-actin and cortactin, and Arg and cortactin synergize to stabilize F-actin and activate F-actin branch nucleation by the Arp2/3 complex. We hypothesize that Arg recruits cortactin to spines, promotes its binding to F-actin, and together Arg and cortactin maintain the stable F-actin pool to stabilize spines. We will test this hypothesis in three Aims: Our first aim will define the molecular basis for cortactin bindng to F-actin. We hypothesize that cortactin's ability to bind F-actin is essential for it to regulate F-actin dynamics and mediate dendritic spine stability, but we completely lack a high-resolution understanding of how cortactin binds F-actin. We find that the cortactin repeats (CR) domain is natively unfolded in solution, but we can obtain CR:F-actin complexes suitable for high resolution structure determination using cryo-EM. We will also use CR domain truncations in tandem with hydrogen-deuterium exchange mass spectrometry to map residues at the cortactin:F-actin binding interface. Our second aim will elucidate how Arg and cortactin interact to control actin filament dynamics. The mechanisms by which Arg and cortactin maintain the stable pool of F-actin in dendritic spines are unknown. We find that Arg and cortactin interact to control the stability of F-actin and new actin branch nucleation in vitro. We will use measurements of protein:protein interactions, total internal reflection microscopy-based single actin filament assays, and structure determination via cryo-EM to understand how Arg and cortactin interact with each other and the Arp2/3 complex to regulate F-actin stability and actin branch nucleation. Our third aim will elucidate the role of the stable actin pool in dendritic spine infrastructure and stability. Our preliminary data suggest that the stable F-actin pool may stabilize spines both by acting as a central organizer of spine infrastructure and by attenuating NMDA receptor (NMDAR) activity. We will use a knockdown/ complementation strategy with Arg or cortactin mutants in cultured neurons to reveal whether their actin regulatory and/or other functions are required to maintain the spine's stable F-actin pool, regulate NMDARs, and spine stability. We will also use live cell and super-resolution microscopy to measure how disruption of the stable F-actin pool impacts the organization of key actin regulators and subcompartments within the spine.

Immature excitatory synapses in the perinatal brain contain high release probability (Pr) presynaptic terminals coupled to postsynaptic specializations with GluN2B subunit-containing NMDA receptors (hi-Pr, hi-GluN2B synapses). For over two decades, we have known that these immature synapses mature in an activity- dependent manner to low-Pr, low-GluN2B synapses, but the mechanisms that coordinate this transition, why it occurs, and how it contributes to circuit plasticity and stability remain controversial and are fundamental unanswered questions. Addressing these issues will identify basic mechanisms that control synapse development and that may be disrupted in neurodevelopmental and psychiatric disorders. Disruption of the Arg/Abl2 kinase in mice yields a population of hi-Pr, hi-GluN2B synapses that persist into early adulthood. The persistence of these immature synapses drives a >40% net loss of hippocampal synapses between postnatal day (P) 21 and P42, and impairs synaptic plasticity and behavior. Building on these findings, we will identify new regulators of synapse maturation, and determine how they regulate synaptic plasticity and stability. In Aim 1, we will identify the cell surface receptors that activate Arg to coordinate the maturation from hi-Pr, hi- GluN2B synapses to low-Pr, low-GluN2B synapses. We provide preliminary data that integrin ?3?1 adhesion receptor and platelet-derived growth factor receptor ? (PDGFR?) act upstream of Arg to control synapse function and stability. We will use selective gene inactivation in the pre- and postsynaptic neurons along with genetic epistasis and rescue experiments to address how and where these receptors interact with Arg and each other to regulate Pr and postsynaptic GluN2B levels. In Aim 2, we will elucidate how Arg mediates GluN2B downregulation at the synapse. Arg-mediated signaling is critical to downregulate GluN2B during maturation. We identified the SHP2 tyrosine phosphatase and the NMDAR-associated protein BRAG1, both mutated in intellectual disability, as likely functional links between Arg and developmental GluN2B downregulation. We will use biochemical, cell-based, and genetic approaches to test how Arg interacts with SHP2 and BRAG1 to downregulate GluN2B function. In Aim 3, we will characterize how immature and mature synapses differentially contribute to plasticity and stability. We will use patterned glutamate uncaging at single synapses to test whether hi-Pr, hi-GluN2B and low-Pr, low-GluN2B synapses have altered ability to undergo long-term potentiation (LTP) and long-term depression (LTD) in arg?/? mice. We will use in vivo imaging to examine how enlarged dendritic spines at hi- Pr, hi-GluN2B cortical synapses in arg?/? mice differ from normal spines in their plasticity and stability. Our studies will elucidate the mechanisms by which receptors act through Arg and its downstream targets to control Pr and NMDAR composition during synapse maturation. Disruption of these mechanisms may underlie the defects in synapse development, plasticity, and stability in intellectual disability and other brain disorders.

ABSTRACT Dendritic arbors and dendritic spines and their associated synapses become destabilized prematurely in neurological disorders. The Abl2/Arg nonreceptor tyrosine kinase is essential for neuronal stability. Disruption of laminin a5/integrin a3b1 signaling through Abl2 causes significant dendrite and spine loss in the late postnatal mouse brain, accompanied by progressive defects in behavioral flexibility, learning, and memory. The mechanisms by which Abl2 stabilizes dendrites and dendritic spines are fundamental, yet unresolved questions. Genetic studies in Drosophila show that abl interacts functionally with MTs to control neurite outgrowth and axon pathfinding, but the underlying mechanism is unknown. We report the unexpected finding that the Abl2 C- terminal half (Abl2-557-C), which lacks the kinase domain, binds MTs and tubulin dimers and increases the growth velocity (vg), reduces shortening rate, and decreases catastrophe frequency (fcat) of MT plus ends in vitro. Disruption of Abl2 reduces MT plus end elongation rates in fibroblast cells, which can be restored by re- expression of Abl2 or Abl-557-C at physiological levels. We will elucidate the mechanism by which Abl2 regulates MTs in vitro and determine whether and how it contributes to Abl2-mediated dendrite and dendritic spine stability. Our first aim will elucidate how Abl2 regulates MT elongation. To understand how Abl2 regulates MT plus-end dynamics, we need to know where Abl2 and Abl2-557-C bind MTs and how this relates to regulation of discrete MT behaviors. We will use TIRFM to measure single and bulk Abl2-GFP molecule binding to growing rhodamine- labeled MTs to measure the Kd, kon, and koff of single Abl2/Abl2 mutant-GFP molecules to the MT lattice vs. MT plus tip, and use these and other measurements (vg and fcat) to computationally model the effects of Abl2 on MT plus-end dynamics. We will use fluorescence anisotropy to identify the tubulin dimer binding region in Abl2 and TIRFM-based assays to probe how it impacts MT dynamics in vitro. Finally, to test if this is a general function of Abl kinases, we will study whether and how vertebrate Abl1 and Drosophila Abl control MTs. Our second aim will determine how Abl2 controls MT dynamics and dendrite stability in neurons. We will measure MT plus-end dynamics in Abl2-deficient cultured hippocampal neurons and rescue them with WT Abl2 and our set of biochemically-characterized Abl2 mutants to reveal which Abl2 functions are required for normal MT dynamics in axons and dendrites. In a subset of experiments, we will perform two-color TIRFM imaging of the MT plus-end marker GFP-MACF43 and Abl2/Abl2 mutant-mCherry to address how growing MTs interact with Abl2 in real time. We will use Abl2-deficient neurons reconstituted with Abl2 or Abl2 mutants with discrete effects on MT plus-end dynamics to determine how these functions contribute to dendritic branch and dendritic spine stability in neurons.

Abstract Heterozygous loss-of-function (LOF) or damaging variants in the TRIO gene are associated with increased risk for schizophrenia and autism spectrum disorders. However, the functional role of TRIO in neuronal biology and circuit function are not well understood, which limits the advance of therapies for these disorders. TRIO acts downstream of cell surface receptors to control axon and dendrite pathfinding, synapse development, and synaptic transmission. Deletion of a single TRIO allele in mouse cortical excitatory neurons drives reductions in cortical neuropil and defects in dendrite and synapse development and function, yielding social and motor deficits and increased anxiety and compulsivity. However, the links between specific TRIO mutations and subsequent consequences for cortical function are unknown. Here, we will integrate a broad array of highly complementary, interdisciplinary approaches including genetics, biochemistry and proteomics, optogenetic analysis of synaptic function, and multimodal in vivo imaging of cortical network dynamics to address this question. Our first aim will identify the biochemical mechanisms by which TRIO regulates cortical neuron development. We identified several new candidate TRIO signaling partners (PDE4A5, L1CAM, and the LGI1/ADAM22/ADAM23 complex) and will elucidate how they interact with TRIO to regulate cortical neuron dendritic arbor, dendritic spine, and synapse development. We also generated CRISPR mice heterozygous for three disorder-related TRIO variants - K1431M (autism), K1918X (schizophrenia), M2145T (bipolar disorder) - that differentially impact TRIO?s biochemical activities and yield different anatomical and behavioral phenotypes. We will use mass spectrometry-based comparative proteomics to discover new signaling partners differentially impacted by these discrete TRIO alleles. Our second aim will determine how different TRIO variants impact neuronal connectivity and synaptic function. We will assess the consequences of our TRIO CRISPR variants for cortical neuron development by measuring how they impact axon, dendrite, and synapse development, synaptic transmission and plasticity. We will also use viral Cre-mediated sparse TRIO disruption and whole cell recordings to test which deficits reflect cell- autonomous versus network level effects. Our third aim will test how alterations in TRIO impact the functional organization of cortical networks in vivo, taking advantage of our recently developed strategies for combining single cell and mesoscopic imaging of GCaMP6-labeled neurons to measure circuit organization in awake, behaving mice. Our overall goal is to understand how altered TRIO function impacts neuronal function at the cellular, synaptic, and network levels, providing a broad framework for understanding how genetic dysregulation drives changes in behavior.