Steven A. Goldman - US grants

nervous system transplantation; nervous system regeneration;

Clinical neurology has long operated on the premise that the adult central nervous system has limited capacity for self-repair following damage. However, several neural systems have now been described which exhibit neuronal production in adulthood. Among these, the songbird vocal control nucleus, HVc, is the best characterized. Its high rate of neuronal production, well-describing anatomy and physiology, functional dependence on circulating gonadal steroids, and association with a well-defined behavioral endpoint - song-have made this an appealing system within which to study neuronal generation in adulthood. I have recently developed a preparation by which explants of the adult songbird HVc may be maintained in long-term culture, under conditions which permit sustained neuronal production, migration, and differentiation in vitro. Using this system, I propose to study factors regulating adult neuronal production and migration in vitro, by addressing the following issues: 1. By what humoral means is the neurogenic activity of the ventricular zone precursor population constrained? What factors influence the in vitro rate of neuronal mitogenesis by adult avian ventricular zone precursor cells? This study will pursue the hypotheses that fetal serum harbors anti-mitotic factors which suppress neuronal mitogenesis in vitro, and that serum-stimulated astrocytes and ependymal cells may release neuronal differentiation agents with anti-mitogenic activity. 2. What is the ontogeny of the ventricular zone precursor cells? To what extent do these cells remain pluripotential? To what differentiated cell types do they give rise? This work will address the postulate that newly generated neurons and guide cells are derived from a pluripotential neuroepithelial precursor cell. By retroviral-insertion of the lacZ marker gene into ventricular zone progenitor cells in vitro, with complementary video analysis, the ontogeny and fate of the neuronal and guide cell precursors will be defined. 3. By what cellular and molecular adhesive mechanisms are newly produced neurons able to migrate into adult avian brain tissue? These experiments will test the hypothesis that the guide fiber network of the adult avian brain is dynamic in its cytoarchitecture, and that its geometry may be influenced by the regulated turnover and fiber extension of its constituent ependymal cells. Complementary experiments will examine the idea that specific neuroependymal adhesive molecules determine both the migration efficacy and fate of newly generated neurons within the adult avian forebrain. The species-specific, regionally-limited nature of adult neurogenesis may reflect specific molecular constraints upon neuronal mitogenesis and migration in non-neurogenic species and brain regions. The circumvention of these constraints by directed pharmacotherapy may thus provide a strategy for brain repair in otherwise non-neurogenic species. This work is intended to provide the data base and rationale by which potential therapeutic modalities such as induced neurogenesis and site-directed neuronal migration may be developed, in the belief that these techniques will become powerful options for the reconstitution of the damaged adult brain.

Among adult mammals, brain neurogenesis is highly restricted, both spatially and phylogenetically, and has not previously been found in primates. In contrast, neurogenesis is widespread and robust in the adult songbird forebrain, which continues to generate neurons from mitotic ventricular zone (VZ) precursor cells. We previously established a preparation by which neurogenesis could be studied in cultures of the adult avian forebrain. The percentage of neurons generated in these cultures varied as an inverse function of the serum level, indicating that serum might harbor or induce factors which are anti-mitotic for VZ precursors. On this basis, we proposed that the lack of neuronal production by non-neurogenic adult brain might result not from an absence of suitable precursor cells, but rather from their tonic inhibition by serum-born or locally-derived agents. Consistent with this idea, recent reports have demonstrated that neuronal precursor cells are present in cultures derIved from adult rodent brain. Since brain neurogenesis in mammalian embryogeny is VZ-based, just as it is in the adult songbird, we hypothesized that the adult human VZ might continue to harbor neuroepithelial stem cells; these cells may remain neurogenic in selected groups such as the songbirds, yet become vestigial in mammals. We further postulated that the adult human forebrain might also harbor such cells, which although non-neurogenic in vivo, retain the capacity for neurogenesis in vitro, once removed from local tissue influences. In preliminary studies, we indeed obtained evidence of newly produced neurons in cultures of adult human temporal lobe. We now propose to further study this phenomenon, by examining the mitotic capability, lineage potential of, and regulatory constraints upon, these precursor cells of the adult mammalian forebrain. Although the emphasis of this proposal is upon adult human forebrain, these experiments will include study of both rat and human brain cells: Rat tissue will be used as a more accessible alternative to human brain for experiments requiring many matched samples run in parallel, such as comparisons of the effects of defined growth factors upon VZ cell proliferation and differentiation. In this proposal, we will ask, and hope to answer, the following questions: 1. Can mitotic precursor cells with neuronal potential be identified in explant cultures of the adult human forebrain ventricular zone? 2. At what regulatory level is neurogenesis suppressed in the adult mammalian brain in vivo? What humoral signals are operative in the temporal and spatial restriction of adult neurogenesis? 3. Can typically quiescent glial phenotypes, particularly oligodendrocytes, also be generated anew from resident precursor cells in the adult human brain? 4. Are individual adult human forebrain ventricular zone precursor cells multipotential? Does the cell type generated by the ventricular zone stem cell depend upon its ambient environment? The establishment of adult human neurogenesis in vitro may permit us to induce this process in vivo, whether from endogenous or introduced progenitors. In pursuing this work, we hope to develop an operational rationale by which induced neurogenesis may become a viable option in the repair of structurally-damaged brain.

DESCRIPTION: (Taken from the applicant's abstract) This proposal seeks to address the mechanisms, cellular targets and target proccesses by which gonadal hormones act to recruit new neurons into the adult brain, using the higher vocal center (HVc) and the adjacent neostriatum in adult avian (canary, zebra finch) as a model. Given that estrogen is known to promote neuronal recruitment in the adult, the applicant posulates that this recruitment might be the result of estrogen acting on other paracrine agents to promote survival of post-mitotic neurons during or after migration, promote neuronal departure from the ventricular zone, direct the production and lineage commitment of VZ daughter cells, and/or enhance the recruitment of new neurons into the parenchyma, the latter by modulating the calcium response of migrating cells. Four aims are presented: 1). To identify estrogen-regulated agents, whether they directly promote neuronal survival, and if so, at what stages during generation, migration, and maturation. 2). To study whether estrogen or its induced paracrine agents, promote neuronal recruitment or departure from the adult ventricular zone. Based on the applicant's finding that this departure is preceded by the down-regulation of N-cadherin by the new migrants, the experiments will address whether the estrogen-associated agents play a role in down-regulation of N-cadherin. 3). To study whether neuronal mitogenesis and later committment is regulated by non-steroidal gonadal peptides (TGF-b, activin, inhibin), shown previously by the applicant to suppress neurogenesis. This will be analyzed by two approaches: a). testing these factors, including other estrogen-induced agests identified in Aim 1, in suspension cultures of microcarrier-born VZ aggreages, allowing low serum, long-term maintenance of adult VZ cells. b). observing effects of these agents on single precursors transfected with a construct under the control of the immature neuronal promoter Talpha1, with a green fluorescent protein reporter gene, in turn faciltiating cell-sorting, to isolate precursor cells derived from fetal and adult VZ. 4. To determine whether estrogen or its induced agents influence the recruitment of new neurons into the adult brain by modulating the NgCAM-triggered calcium signal, which is temporally-restricted to new bipolar migrants.

DESCRIPTION (Abstract reproduced verbatim): Neuronal precursor cells are distributed throughout the ventricular subependyma, or subventricular zone (SVZ), of the adult vertebrate forebrain. We previously found that adult precursors are more widespread, both spatially and phylogenetically, than the limited instances of adult neurogenesis in vivo might indicate. On this basis, we postulated that neuronal recruitment into the adult brain is restricted not by the distribution of neural precursors, but rather by a failure of the mature brain environment to provide permissive support of neuronal differentiation and survival. The focus of this proposal is to define the distribution, lineage potential and humoral control of the progenitor pools, in the adult human forebrain. Our aim is to elicit neuronal production in the damaged brain, by inducing endogenous precursors to resume neurogenesis, and by providing for the differentiation and survival of their neuronal progeny. To wit,1. What is the distribution of neuronal and uncommitted neural precursors in the adult human SVZ? 2. Can the adult human subependyma be deconstructed into its constituent progenitor phenotvpes? By sorting dissociates of human SVZ after transfection with P/Talphal:hGFP, can neuronal progenitors be harvested specifically, and in high yield? Can their mitotic potential be retained after neuronal commitment? Can their transmitter phenotypes be modulated? 3. Is there a distinct phenotype of neural stem cell, as rigorously defined as a self-replicating multipotential progenitor, in the adult human VZ? Using the early neural regulatory sequences for nestin a musashi proteins, coupled to GFP as reporters of phenotype, can such uncommitted progenitors be selected from adult human brain tissue? Can these cells be directed towards neurogenesis by humoral neurotrophins? 4. Is there a distinct class of hippocampal progenitor? Are adult human hippocampal progenitors multipotential or neuronally-committed? Is their native lineage potential more restricted than that of SVZ progenitors? We intend to learn enough about the biology and selective harvest of adult human neuronal progenitors to establish an operational basis for their therapeutic use. Our goal is to utilize these cells for neuronal replacement within the damaged brain, whether by activating endogenous progenitors, or implanting exogenous cells. If this work is successful, we may soon need to consider how best to train newly generated networks of replacement neurons and glia to assume lost functions, in diseases as diverse as stroke, trauma, and the neurodegenerations.

Demyelination in the setting of oligodendrocytic loss is a significant contributor to neurological dysfunction in a variety of subcortical pathologies, including posthypoxic leukoencephalopathy, capsular stroke, and head trauma, as well as in the inflammatory, hereditary and degenerative leukoencephalopathies. We have found that a distinct and separate pool of mitotically-competent oligodendrocyte (oligo) progenitor cells resides in the adult human capsular white matter. This glial progenitor does not appear to be uncommon; it may comprise as many as 4 percent of the cells of the mature white matter. Although the existence of an analogous cell-type has been postulated in rats, its presence in humans had remained controversial, in part because of the difficulty in specifically identifying or obtaining these cells. We therefore developed a means of isolating these cells from adult human patients, by transfecting dissociates of resected white matter with plasmids that selectively identify oligo progenitors, by their expression of fluorescent transgenes controlled by early oligodendrocyte promoters. Specifically, we use the early promoter (P2) for cyclic nucleotide phosphodiesterase (CNP), which is preferentially active in mitotic oligo progenitors, to direct expression of green fluorescence protein (GFP) to these cells. This approach has allowed us not only to identify live oligo progenitors in vitro, but to enrich them by fluorescence-activated cell sorting (FACS). As a result, we can use FACS based upon CNP2:hGFP expression to purify oligodendrocyte progenitor cells from the adult human white matter, in high-yield, and with retained mitotic potential, differentiation competence and in vitro survival. In this application, I will capitalize upon our acquisition of these purified adult human oligodendrocyte progenitor cells, with experiments designed to assess their capacity for structural remyelination. To this end, I plan to define several therapeutically-relevant aspects of the biology of these cells, including 1) their lineage potential and capacity for myelination, and the humoral control thereof, and 2) the humoral regulation of their clonal expansion, and 3) their capacity for oligodendrocytic maturation and myelination upon engraftment. To these ends, I plan to transplant P/CNP2:hGFP-defined, FACS-purified populations of adult human oligo progenitors into a rodent model of focal demyelination. These latter studies would serve as a prelude to preclinical analysis of this cell type's response to implantation in nonhuman primates, in models of both chemical and radiation-induced demyelination. This application is intended to develop oligodendrocyte precursor implants, as well as induced oligoneogenesis from endogenous progenitors, as feasible therapeutic options for the structural repair of demyelinated brain.

DESCRIPTION (provided by applicant): The vocal control nucleus HVC of adult songbirds generates new neurons throughout life from nearby ventricular zone progenitor cells. The recruitment and survival of these neurons is modulated by the gonadal steroids testosterone and estradiol. Yet endothelial cell division is the first cellular response noted in the adult HVC after testosterone administration, and precedes androgen-associated gliogenesis and neuronal recruitment by over a week. We have found that testosterone rapidly induces the production of both vascular endothelial growth factor (VEGF) and its receptor VEGF-R2/KDR in HVC. This leads to endothelial division, which anticipates the regionally-restricted expansion of the HVC microvasculature. The activated endothelial cells then produce the neurotrophin BDNF, which has been shown to support neuronal recruitment from the mammalian as well as the avian ventricular zone, and whose induction is associated with the recruitment of new neurons to HVC. Gonadal steroid-induced endothelial and matrix-released cytokines may thereby contribute importantly to neuronal recruitment in the adult brain. In this competitive renewal application, we postulate that gonadal steroid-induced angiogenesis may be critical to neuronal addition to the adult HVC, and that angiogenesis may provide necessary permissive conditions for neuronal recruitment into adult brain parenchyma. To better understand the permissive conditions for the integration of new neurons into adult brain, we propose to ask the following: 1) Is the gonadal steroid-associated induction of VEGF and its receptor necessary for testosterone-induced angiogenesis in the adult HVC? 2) Is steroid-induced angiogenesis necessary for testosterone-mediated neuronal recruitment? Does the suppression of angiogenesis abrogate neuronal addition to HVC? 3) Is local angiogenesis sufficient to direct neuronal recruitment from a neurogenic epithelium? Do new neurons migrate selectively to HVC because of its high BDNF levels after gonadal steroid activation? 4) Do steroid-activated matrix metalloproteinases contribute importantly to neuronal addition? Does MMP inhibition suppress recruitment? These experiments ask if neurogenesis in the adult HVC depends upon antecedent local angiogenesis. They seek to define those necessary and sufficient conditions for neuronal recruitment that may be provided by steroid-activated endothelial cells. By so doing, they extend our understanding of the permissive conditions for neurogenesis in the adult brain, and may inform us as to how to induce neuronal recruitment to, and acceptance by, otherwise non-neurogenic regions of the adult brain.

DESCRIPTION (provided by applicant): In the first 4 years of this grant, we assessed the cell biology and transplantation characteristics of adult human oligodendrocyte progenitor cells. We established means for their specific isolation, using CNP2:GFP and A2B5 based fluorescence-activated cell sorting (FACS), and then assessed their lineage potential. We discovered that when removed to low density culture, the progenitors were multipotential, and gave rise to neurons as well as to gila. Thus, absent autocrine and paracrine influences on their differentiation, adult WMPCs were not restricted to oligodendrocytic fate. To assess the role of the tissue environment in regulating progenitor fate, we then investigated gene expression by adult human WMPCs. We focused on identifying those transcripts that are differentially expressed by the WMPC, relative to its local tissue environment. This process enabled us to predict ligand-receptor interactions that maintain the progenitor state, as well as those that determine whether a given cell develops into an astrocyte, oligodendrocyte, or neuron. This analysis identified a set of parallel pathways that together appear to regulate the maintenance, mobilization and differentiated fate of parenchymal progenitor cells. In particular, we noted a complex interaction of: 1) receptor tyrosine phosphatase beta/zeta signaling, as regulated by its ligands pleiotrophin and NrCAM; 2) a parallel avenue of syndecan3 regulated signaling through CASK and tbr1; 3) FGFR3-dependent signaling, likely mediated through syndecan3 proteolysis and release of CASK; and 4) a neuralin and BAMBI-suppression of BMP signals. It appears that the net output of these pathways biases adult progenitors to either self-renewal or differentiation. In this application, we propose to use a combination of protein delivery, adenoviral overexpression and lentiviral RNAi knock-down to evaluate the individual elements of these candidate systems. By this means, we intend to better define the niche for gliogenesis in the adult human white matter, and by so doing to establish both necessary and sufficient genetic targets for directing the phenotypes generated by resident progenitor cells.

[unreadable] DESCRIPTION (provided by applicant): We have found that adenoviral overexpression of BDNF in the adult rodent ventricular system induces the recruitment of new striatal neurons from the progenitor cell pool of the forebrain subependyma. The new striatal neurons project to the globus pallidus and adopt a DARPP32/GABAergic/calbindin+ phenotype, characteristic of medium spiny projection neurons. This is the major neostriatal phenotype lost in Huntington's Disease (HD); as such, the induced regeneration of these cells may be a feasible strategy for moderating disease progression. We also noted that the numbers of new neurons recruited to the striatum in response to BDNF could be greatly potentiated by suppressing gliogenesis, using adenoviral overexpression of noggin, a soluble antagonist of the pro-gliogenic bone morphogenetic proteins (BMPs). On this basis, this proposal asks if induced striatal neuronal recruitment might offer therapeutic benefit in murine models of Huntington's Disease. We have already noted that the R6-2 mouse, a transgenic model of Huntington's disease, indeed harbors competent striatal progenitor ceils, that these give rise to striatal neurons in response to intraventricular AdBDNF, and this process is potentiated by AdNoggin. On this basis, we now ask: 1) Does spontaneous neuronal recruitment occur in HD mice, in response to striatal neuronal loss? Are new neurons recruited in response to striatal apoptosis? What cellular and molecular signals elicit neuronal recruitment? 2) What is the lifespan and connectivity of AdBDNF/Noggin-induced striatal neurons in R6-2 nice? As what transmitter and functional phenotypes do these neurons integrate? 3) Does AdBDNF/AdNoggin-induced neuronal addition prolong the survival of R6-2 mice? Does it slow their motor deterioration? To what extent is AdBDNF/Noggin's effect due to neuronal addition, as opposed to AdBDNF-dependent neuroprotection'? Can the survival of BDNF/noggin-treated R6-2 mice be further improved by histone deacetylase inhibition, as a potentially synergistic neuroprotective strategy? 4) Is sustained BDNF and noggin expression required to maintain neuronal recruitment at levels sufficient to compensate for loss due to Huntington's disease? With what vector systems might this best be accomplished? Does the striatal progenitor pool deplete with sustained stimulation? 5) Can noggin and BDNF be used to induce meaningful levels of neuronal recruitment in the adult primate? Can AdBDNF/AdNoggin-treatment mediate the replacement of striatal neurons lost following treatment with ibotenic acid (IA)? If successful, these experiments should provide both a conceptual and operational foundation for the evaluation of induced striatal neurogenesis as a therapeutic strategy in Huntington's Disease, while providing new insight into the mechanistic bases for both compensatory and induced neurogenesis in the adult mammalian brain. [unreadable] [unreadable]

[unreadable] DESCRIPTION (provided by applicant): We have found that adenoviral overexpression of BDNF in the adult rodent ventricular system induces the recruitment of new striatal neurons from the progenitor cell pool of the forebrain subependyma. The new striatal neurons project to the globus pallidus and adopt a DARPP32/GABAergic/calbindin+ phenotype, characteristic of medium spiny projection neurons. This is the major neostriatal phenotype lost in Huntington's Disease (HD); as such, the induced regeneration of these cells may be a feasible strategy for moderating disease progression. We also noted that the numbers of new neurons recruited to the striatum in response to BDNF could be greatly potentiated by suppressing gliogenesis, using adenoviral overexpression of noggin, a soluble antagonist of the pro-gliogenic bone morphogenetic proteins (BMPs). On this basis, this proposal asks if induced striatal neuronal recruitment might offer therapeutic benefit in murine models of Huntington's Disease. We have already noted that the R6-2 mouse, a transgenic model of Huntington's disease, indeed harbors competent striatal progenitor ceils, that these give rise to striatal neurons in response to intraventricular AdBDNF, and this process is potentiated by AdNoggin. On this basis, we now ask: 1) Does spontaneous neuronal recruitment occur in HD mice, in response to striatal neuronal loss? Are new neurons recruited in response to striatal apoptosis? What cellular and molecular signals elicit neuronal recruitment? 2) What is the lifespan and connectivity of AdBDNF/Noggin-induced striatal neurons in R6-2 nice? As what transmitter and functional phenotypes do these neurons integrate? 3) Does AdBDNF/AdNoggin-induced neuronal addition prolong the survival of R6-2 mice? Does it slow their motor deterioration? To what extent is AdBDNF/Noggin's effect due to neuronal addition, as opposed to AdBDNF-dependent neuroprotection'? Can the survival of BDNF/noggin-treated R6-2 mice be further improved by histone deacetylase inhibition, as a potentially synergistic neuroprotective strategy? 4) Is sustained BDNF and noggin expression required to maintain neuronal recruitment at levels sufficient to compensate for loss due to Huntington's disease? With what vector systems might this best be accomplished? Does the striatal progenitor pool deplete with sustained stimulation? 5) Can noggin and BDNF be used to induce meaningful levels of neuronal recruitment in the adult primate? Can AdBDNF/AdNoggin-treatment mediate the replacement of striatal neurons lost following treatment with ibotenic acid (IA)? If successful, these experiments should provide both a conceptual and operational foundation for the evaluation of induced striatal neurogenesis as a therapeutic strategy in Huntington's Disease, while providing new insight into the mechanistic bases for both compensatory and induced neurogenesis in the adult mammalian brain. [unreadable] [unreadable]

There is an abundant pool of glial progenitor cells (GPCs) dispersed throughout adult human brain tissue. This proposal seeks to define the niche for astrogliogenesis in the human white matter, with an emphasis on defining the molecular basis for reactive astrocytosis. Absent autocrine and paracrine influences, adult GPCs are not restricted to any given neural lineage. Rather, the environment regulates their differentiated fate, suggesting that an ischemic environment might specifically direct GPCs to astrocytic fate in reactive astrocytosis. On this basis, we investigated the gene expression patterns of adult human GPCs derived from normal brain tissue. We identified a set of parallel and interacting ligand-receptor interactions, as well as their cognate signaling pathways, that may to determine whether a given GPC remains undifferentiated, or instead develops into an astrocyte, oligodendrocyte, or neuron. We now propose to define those pathways involved in both normal and post-ischemic astrocytosis from adult human glial progenitors. By this means, we expect to identify genetic and pharmacological targets by which to suppress reactive astrocytosis following ischemic injury, while sustaining our ability to mobilize progenitors to desired lineages. To this end, we shall focus on several pathways that appear differentially expressed in isolated GPCs. These include receptor tyrosine phosphatase-p/^ (RTPZ), which may play a central role in modulating li-catenin trafficking, its chondroitin sulfate proteoglycan ligands, and two interacting receptor systems, the FGFR3 tyrosine kinase, and the BMP4-dependent serine/threonine kinases. Specifically, weask: 1. Does the inhibition of RTPZ signaling suppress astrocytosis in vitro? Is this effect mediated by impeding B-catenin translocation? Can RTPZ inhibition suppress glial scar formation in vivo? What are the functional effects of suppressing post-ischemic gliosis through RTPZ inhibition? 2. Can reactive astrocytosis and glial scar formation after stroke be prevented by inhibiting BMP4-signaled astrocyte induction? What BMP inhibitors are best for this purpose? 3. How does the RTPZ pathway interact with FGFR3 and BMP signaling to instruct astrocytic fate? 4. Does the expression of CSPGs by GPCs produce an environment biased to astrocytosis, through CSPG-dependent activation of RTPZ? Can astrocytosis be suppressed through CSPG inactivation? Can concurrent RTPZ suppression further attenuate gliosis? What are the functional consequences of these strategies, in particular after MCA occlusion? The implications of this work are profound, not only in regards to stroke and trauma, but more broadly with regards to diabetic and hypertensive encephalopathies, which share an hypoxia-triggered disruption of the normal niches for cell genesis in the brain. In each of these disorders, reactive astrocytosis culminates in the sclerotic pathology that typifies terminally non-regenerative brain and spinal cord. Our goal is too prevent this fate, and by so doing preserve the cellular plasticity and regenerative capabilities of the healthy CNS. PHS 398 (Rev. 09/04) Page 145 Form Page 2 POT NS050315 Principal investigator (Last, First, Middle): Nedergaard, Maiken/Goldman, Steven Projects MODULATION OF POST-ISCHEMIC ASTROGLIOSIS BY HUMAN GLIAL PROGENITOR CELLS I. RESPONSE TO REVIEWERS I would like to thank the referees for their helpful critiques of my proposal, Modulation of post-ischemic astrogliosis by human glial progenitor cells. By way of review, this application proposes to assess the molecular basis for reactive astrocytosis, with particular attention to signaling pathways we have identified in adult human glial progenitor cells (GPCs), that appear to be important to GPC mobilization and astrocyte differentiation therefrom. In brief, the application was criticized as being insufficiently detailed in regards to some of the proposed methodologies and experimental designs, and too expansive in terms of the number of experiments proposed. In addition, some concern was voiced as to the proposed use of RNAi reagents whose validation we had not made explicit. More general concerns were also stated as to insufficient integration across the individual projects of this program project, which prompted us to describe in greater detail the interactions among our group. In response to the referees' specific concerns: Critique 1 Aim 1 proposes to test the effects of RTPZ modulation after MCA occlusion. It was criticized as providing insufficient definition of the stroke bed or its boundaries, and therefore insufficiently described histological endpoints for assessing treatment-associated effects. To this end, we have added a paragraph to the Methods section which better defines the histological criteria and scoring procedures that we are using. Aim 2 proposes to assess the role of BMP signaling and antagonists thereof in modulating post- ischemic astrocytic fate. It was criticized for insufficient delineation of the numbers of animals necessary of each experiment. The referee also noted that we provided no justification for the extensive in vivo testing proposed. We have now added clear estimates of our anticipated experimental sample sizes for all 4 aims of this application, with numbers based on data that we previously obtained using noggin to suppress astrocyte production in the adult VZ(Chmielnicki et al., 2004b) (Appendix 8). The BMP inhibitors assessed - neuralin, BAMBI, DAN, gremlin and noggin - have different BMP ligand specificities and binding efficiencies, and different local bioavailabilities, due to variable degrees of heparin binding. As a result, we intend to test each of them in vivo, both by adenoviral over-expression and lentiRNAi knock-down. I have to assume, until we learn otherwise, that the results of one BMP inhibitor will not necessarily predict the results using another. In regards to reagent availability, we have already constructed all of the adenoviral over-expression constructs, and expect to have all of the lentiviral shRNAis constructed within the next few months; to date, we have made lentiRNAis for neuralin and BAMBI, besides those made and validated for RTPZ and its interactants, syndecanS and CASK, and are now generating similar knock-down vectors for gremlin and DAN. I have added a table to the last page of the Methods that includes all of the vectors needed in this application, and the preparation status and current availability of each. Aim 3 addresses the interaction of RTPZ and BMP signaling with that effected through FGFR3, and tests the hypothesis that FGFR3 blocks astrocytic production and differentiation by inhibiting RTPZ and BMP signals, thereby potentiating the oligodendrocytic lineage bias imparted by RTPZ inhibition). The aim was criticized fro being too expansive, in that it includes many discrete experiments, with unbiased cell counts and formal stereological assessment and reconstruction being required for each. Yet the required throughput for cell counting, stereology and imaging is well within the capabilities of my group. We have published several unrelated studies (Appendices 6, 8 and 9; also(Louissaint et al., 2002)), that employ the same methods proposed here of scoring BrdU-tagged cells double or triple labeled with other markers, then quantifying the labeling indices as a function of treatment, time point and region. Prior papers from my lab (Chmielnicki et al., 2004b; Windrem et al., 2004) (Appendices 6 and 8) have been individually comparable in effort to what is proposed here in Aim 2. We have two BioQuant imaging systems in the lab, each run full-time by technicians under the direction of Martha Windrem and Abdel Benraiss - both investigators who moved with mefromNew York to Rochester - each of whom has considerable experience in these methods. My estimates of what we're capable of doing are thus predicated on past experience, running similar types of samples, with the same image acquisition and analysis systems, used by the same group of experienced investigators. The reviewer also requests a table presenting the individual treatment groups and assessment endpoints, and expresses concern as to the sufficiency of our sample sizes, as well as the detail of our experimental designs, especially in regards to the in vivo experiments within each aim. Accordingly, I have 146

DESCRIPTION (provided by applicant): Disorders of myelin include the hereditary leukodystrophies and cerebral palsies, as well as adult vascular, traumatic and inflammatory demyelination syndromes. To address this large and diverse group of disease, we established a cell-therapeutic approach to central remyelination, by which transplants of isolated human glial progenitor cells (GPCs) are delivered intracerebrally to neonatal recipients, which are then allowed to mature to adulthood. When the recipients are hypomyelinated mutants, such as the shiverer mouse, the transplanted cells mature largely as myelinating oligodendrocytes, and can rescue both the neurological phenotype and lifespan of the treated animals. Remarkably though, large numbers of human progenitors integrate into the recipient brains, wherein they effectively out-compete mouse progenitors, yielding mice with a substantially humanized white matter, and a major contingent of human glial progenitors - and ultimately astrocytes - in the gray matter as well. The resultant human glial-chimeric mouse brains provide us a variety of hitherto unavailable opportunities for studying human glial cells and their progenitors in vivo, including their responses to injury and disease processes that cannot be adequately modeled in vitro. In the proposed experiments, we will use these mice to assess the effects of toxic demyelination on human GPCs in vivo, so as to identify their molecular responses to injury-induced mobilization and oligodendrocytic differentiation during compensatory remyelination. In particular, we will use phenotype-specific cell sorting and gene expression analysis, to define the responses of these xenografted human GPCs to demyelination in vivo. These data, the first ever obtained specifically from human GPCs during demyelination and remyelination in vivo, should afford us fundamental new insights into the signaling events associated with remyelination, and their potentially targetable points of regulatory control. To achieve that end, we propose the following Aims: In Aim 1, we will treat glial-chimeric mice with cuprizone so as to better understand the responses of human GPCs to demyelination, assessing their mobilization, differentiation, responses to repetitive induction, as well as their thresholds for mitotic senescence and the regulatory control thereof. In Aim 2, we will examine the expression patterns of human GPCs in vivo, sorting them from chimeric shiverer mice both at baseline and in response to demyelination, so as to define those genes and pathways differentially regulated during GPC mobilization and remyelination. In Aim 3 we will compare the responses of co-resident human and mouse GPCs to cuprizone demyelination, so as to identify those shared pathways likely to be high-value targets in drug development, as well as those species-specific pathways whose investigation in mice might not predict human therapeutic outcome. Together, these experiments promise to inform our efforts to define new strategies for treating demyelinating brain or spinal cord injury. In addition, the databases to be generated in the course of this work, as freely available resources to the field should prove catalytic in advancing our understanding of remyelination in vivo. PUBLIC HEALTH RELEVANCE: In the course of developing new human cell therapeutics for treating brain disease, we established mice in which a substantial proportion of all CNS glial cells were of human origin, especially in the white matter, the region affected in human myelin disease. In this application, we propose to use perinatal transplants of human glial progenitor cells to establish mice with largely humanized white matter, and to then subject these mice to experimental demyelination, yielding pathology similar to that of disorders such as subcortical stroke and multiple sclerosis. We will then use cell sorting techniques to separate the human glial progenitor cells from the affected mouse brains, followed by gene expression analyses to identify which genes respond to the demyelinating insult, and by what time course. By this means, we expect to define those signaling pathways involved in remyelination by normal human glial progenitor cells in vivo, knowledge that should provide us great insight into how to induce and regulate this process therapeutically.

DESCRIPTION (provided by applicant): Human evolution has been accompanied by a diversification of astrocytic phenotype and function, that has contributed to species-specific aspects of both human brain function and disease. As such, the development of human astrocytic complexity has paralleled the appearance in evolution of psychiatric disorders unique to humans - the schizophrenias in particular. Yet despite this correlative suggestion that astrocytic pathology might contribute to the disordered thought of schizophrenia, the role of astroglial pathology in its pathogenesis has been difficult to study, in part because of the lack o animal models of human glial pathophysiology. We propose to overcome this limitation, using a new model of human glial-chimeric mouse brains that we have developed, paired with our newly-developed protocols for efficiently and reliably generating astrocytes from patient-derived human induced pluripotential cells (hiPSCs). Using mice neonatally engrafted with glial progenitor cells (GPCs) derived from hiPSCs generated from schizophrenic patients, we will assess the specific contributions of schizophrenic patient-derived astrocytes to disease pathogenesis. In these human glial chimeric mouse brains, the vast majority of resident glia are replaced by human GPC-derived astrocytes and their progenitors, allowing human glial physiology, gene expression, and effects on neural function to be assessed in live adult mice. By pairing this chimerization approach with protocols that we have developed for both generating and purifying GPCs and astrocytes from hiPSCs, and using hiPSC lines produced from patients with juvenile-onset schizophrenia, we will produce mice whose resident glia are largely derived from patients with schizophrenia. In Aim 1, we will assess the relative effects of these schizophrenia-associated astrocytes upon glial syncytial transmission within the cortices of the chimeric mice. In Aim 2, we will next assess the synaptic plasticity of the chimeric mice, as well as the effects of schizophrenia-derived glial chimerization upon their behavioral phenotype and responses to pharmacological stressors. In Aim 3, we will sort engrafted astroglia from the brains into which they have integrated, so as to assess the gene expression patterns of schizophrenic iPSC-derived astrocytes, relative to those of normal hiPSC-derived glia. By means of this multimodal approach, we hope to define the disease-specific effects, gene expression patterns, and paracrine toxicities of schizophrenic hiPSC-derived astrocytes relative to normal hiPSC-derived glia. These diverse lines of investigation should provide us great insight into the species- and cell type-specific roles of human astrocytes in the pathogenesis of schizophrenia. At the same time, by providing a new human glial chimeric model system, new cellular reagents in the form of schizophrenic patient-derived astrocytes, and new gene expression databases covering schizophrenic hiPSC-derived astrocytes, this project should allow us to make available to the field a broad and exciting new set of tools, capabilities and databases. Together, these should greatly accelerate our understanding of human glial dysfunction in the pathogenesis of schizophrenia.

DESCRIPTION (provided by applicant): Schizophrenia is perhaps the most human-specific psychopathology. It is one of the most severe - and common - psychiatric disorders, a cause of tremendous disability, distress and expense, and yet little remains clear about its etiology. That said, the specificity of schizophrenia to humans, and its frequent association in lesser forms with creativity, suggest its association with specifically human neural competencies. Paradoxically though, gene association studies have highlighted the association of oligodendrocytic and white matter genes with schizophrenia, rather than the neuronal genes that one might have expected of a pathology so uniquely human. Similarly, recent radiographic studies have highlighted the early appearance of hypomyelination in schizophrenics, often long before disease onset. These observations suggest that the aberrant features of neuronal organization and neuritic structure noted in schizophrenics might derive from glial progenitor and/or oligodendrocytic pathology, with progenitor dysregulation, dysmyelination and failed oligodendrocytic support of neurons yielding a developmentally-disrupted network structure. In this application, we will focus on the roles of OPCs and oligodendroglia in the genesis of schizophrenia, using a novel set of technologies by which we can produce chimeric mice whose glial populations have been largely replaced by hiPSC-derived glia, themselves derived from patients with juvenile-onset schizophrenia. By using congenitally hypomyelinated shiverer mice as hosts, we can generate mice whose white matter glia and myelin are almost entirely of human origin, with complete replacement of host OPCs and oligodendroglia by human cells. To establish the feasibility of this proposal, we established human glial chimeras in myelin-deficient shiverer mice, using schizophrenia-derived hiPSC OPCs. We found that the resultant human glial chimeric mice - which typically myelinate well following neonatal human OPC transplants - instead phenocopied the hypomyelination so characteristic of schizophrenic patients, with aberrant OPC dispersal and migration accompanying overt capsular and callosal dysmyelination. On that basis, we now propose to study the anatomy and transcriptional architecture of the white matter of these mice, focusing on the distribution of their engrafted OPCs relative to normal control hiPSC OPCs, as well as on oligodendrocytic differentiation and myelin formation by schizophrenia-derived hiPSC OPCs. We will then investigate their gene expression patterns, in terms of both mRNA and miRNA, the latter so as to define any higher-order transcriptional dysregulation that may be experienced by the schizophrenia-derived hiPSC OPCs. By providing a new human glial chimeric model system, new cellular reagents in the form of schizophrenic patient-derived OPCs, and new phenotype-specific and disease-associated gene expression databases, this project will make available a broad and exciting new set of tools, capabilities and databases. Together, these studies should provide us great insight into the role of glial progenitor cells and their derived myelin in the pathogenesis of schizophrenia.

Project Summary/Abstract This project will develop and implement a training program in neurotherapeutics discovery and development for faculty members and advanced postdoctoral fellows, centered around a 3½-day short course that will provide the trainees with the various knowledge elements required to discover and advance a neurotherapeutic agent to IND. Following the short course, the training program will continue for a two-year period in which students will have individualized mentoring and assessment. The training, which is designed to be applicable to diverse diseases of the nervous system, will equip students with a broad understanding of the various component steps in the neurotherapeutics drug discovery and development process. Students will learn how to identify a good drug discovery target; how to construct an assay; the elements of medicinal chemistry; how to conduct animal efficacy testing; the principles of ADME studies, safety testing, and formulation; the principles of experimental medicine and biomarkers; the steps required to prepare an IND document and the principles for interacting with the FDA; the principles of intellectual property as they relate to neurotherapeutics discovery and development; and how to seek funding for academic drug discovery research. They will also receive training in responsible conduct of research. Students will be equipped with the skills to develop and coordinate an entire drug discovery and development effort, and to work collaboratively with experts in each of the component areas. The training will combine didactic lectures with active engagement activities in which the students will be challenged to utilize the lecture material to work through their own drug discovery project plan with the guidance of the area experts. The 3½-day short course (followed by two-years of mentorship and assessment) will be offered annually, a total of five times.

Abstract In this application, we will test a longstanding but effectively untested hypothesis in myelin biology, that white matter recovery after sustained or recurrent demyelination might be constrained by the finite mitotic competence of the human glial progenitor cell pool. Such a mobilization-dependent depletion of mitotically- competent progenitors might lead to mitotic senescence, and hence to the eventual failure of remyelination noted in progressive multiple sclerosis. In addition, any such depletion of competent progenitor cells might also be expected to limit the utility of differentiation-based approaches towards induced remyelination. We thus propose to assess the responses of human glial progenitor cells (hGPCs) to demyelination in vivo, by defining the single cell RNA expression patterns of hGPCs, both at baseline and in response to sustained cuprizone demyelination in vivo. To this end, we will use mice neonatally chimerized with human GPCs, a novel model we have developed in which mouse oligodendrocytes and astrocytes are largely replaced by their human counterparts in vivo. Using these human glial chimeras, we will ask the following questions: 1) What is the phenotypic and transcriptional heterogeneity among single human glial progenitor cells in vivo, in the otherwise undisturbed adult glial chimeric brain? Are all GPCs multi-lineage competent? Are some more restricted than others to astrocytic or oligodendrocytic fate? Are some in cell cycle while others are more quiescent? How do these phenotypic distributions change with age? 2) How heterogeneous are the transcriptional responses of resident human GPCs to demyelination in vivo? In response to cuprizone-mediated demyelination, what differentially regulated pathways distinguish quiescent, initially mobilized, and actively remyelinating hGPCs? These experiments will combine the use of human glial chimeras engrafted with genetically tagged GFP+ hGPCs, with later single cell RNA-seq of both the post-demyelination white matter, and of pooled hGPC isolates after post-demyelination FACS, to define the transcriptional events associated with human GPC mobilization and remyelination in vivo. 3) Does the efficiency of remyelination by hGPCs fall with sustained demyelination? Are human GPCs capable of self-renewal during sustained demyelination, or is remyelination delimited by their mitotic senescence? These experiments will assess both methylation state and telomeric length of GPCs in vivo, both before and after sustained cuprizone exposure, so as to define the effects of sustained demyelination on these hallmarks of cellular aging. In addition, we will assess the transcriptional concomitants to methylation state-defined aging, by RNA-seq of the same cells as a function of time after demyelination. By this means, we intend to define both the transcriptional hallmarks of mitotic exhaustion by hGPCs, and the epigenetic correlates to that process, and by doing so to identify therapeutic targets by which to delay or control the demyelination-associated depletion of mitotically-competent hGPCs.

Abstract Glial progenitor cells (GPCs) pervade the adult human brain, and can give rise to new oligodendrocytes and astrocytes in response to myelin loss; yet they may fail to do so in chronic neuroinflammatory and age- related white matter diseases. Our goal is to identify the transcriptional and epigenetic basis for age-related glial progenitor failure, with the goal of identifying the repressive transcription factors and epigenetic states that restrict progenitor cell expansion and differentiation with age. By targeting these repressive networks, we hope to restore the functional viability of human GPCs, and by so doing prevent the myelin loss that characterizes both aging and those neurodegenerative and inflammatory disorders associated with white matter disease. By so doing, we hope to preserve not only the differentiation competence of the cells, but also their self-renewal, so that myelin- ogenesis may be induced from hGPCs without the progenitor depletion to be expected of strategies designed to trigger terminal oligodendrocytic differentiation. Achieving this in human GPCs, which differ substantially in their biology from mouse, and doing so in vivo, has proven a significant challenge to the field. To this end, we will ask: 1. To what extent is the aging of human ESC-derived GPCs cell-intrinsic and linked to prior cell division, both in vitro and in vivo? What are the transcriptional and epigenetic concomitants to hGPC aging in vivo, and which of these restrict hGPC expansion and differentiation? How do hGPCs, extracted back from neonatally- transplanted human chimeric mouse brains, change in their DNA methylation patterns, their ATAC-Seq-defined patterns of chromatin accessibility, and their consequent RNA expression, over the 2-year lifespan of a mouse? 2. To what extent are the effects of aging on hGPCs a function of the aged brain environment, rather than cell autonomous? In order to define the relationship of hGPC cell age to expansion and myelination competence - and the extent to which the age of the host influences hGPC fate ? these experiments will include a set of reciprocal, heterochronic transplants, grafting aged cells into neonates, and new hGPCs into aged brains. 3. In aged GPCs, can genetic knock-down of those repressors implicated in the progression to adult hGPC phenotype restore the transcriptional signature, as well as the expansion and differentiation competence in vivo, of younger hGPCs? Is suppression of MAX, potentially together with a core set of other over-expressed repressors, sufficient to restore MYC-dependent mitotic expansion and host colonization by aged hGPCs? With this work, we expect to establish a granular understanding of the relative roles of cell-intrinsic, expansion- dependent senescence and host context in regulating the proliferation and remyelination competence of human GPCs. Furthermore, if the introduction of young hGPCs into an aged environment allows the selective colonization of the host white matter by those younger hGPCs, the implications may be profound, as disorders as varied as progressive multiple sclerosis and the neurodegenerative disorders might then become potential targets of cell replacement strategies based on the competitive advantages of young over aged glial progenitors.

Abstract Synaptic failure is an important feature of HIV infection of the brain, and a likely key contributor to HIV- associated neurocognitive disorders (HAND). Yet current animal models have proven of limited utility in defining the mechanisms of this process, in part because of the species-specific nature of HIV infection, but also because of the greater complexity of human astrocytes relative to those of mice. To address this issue, we will utilize mice chimeric for both human microglia and human astrocytes, to assess the effects of HIV infection on central neurons. To that end, we will engraft mice with both human glial progenitor cells (hGPCs) and microglia, each derived from embryonic stem cells (hESCs). We have established the methods of generating these human glial chimeras, by the neonatal implantation of hGPCs, which outcompete and ultimately replace the host mouse GPCs, yielding adult chimeras broadly colonized with human astroglia1-5. This process is especially robust in the neostriatum, allowing the glial humanization of regions critically involved in striatal reward and addiction circuits. We have recently extended this approach to include chimerization with hESC-derived microglia, paired with the use of CSF1R null mice lacking host microglia, crossed to NSG SGM3 mice to allow the stable xenograft of hGPCs. The mice are thus chimeric for hGPC-derived astrocytes as well as microglia, in a T- and B-cell deficient background that allows the effects of glial HIV infection on neurons to be isolated, following intracerebral inoculation with HIV-infected microglia. These chimeric human astroglial-microglial (CHAM) mice are especially attractive, since they incorporate the hominid-specific features of human astroglia, which are themselves key components of central synapses. Using this model, we will test the postulate that astrocytes become both structurally and functionally impaired by microglial HIV infection, resulting in the loss of synaptic engagement by affected astrocytes, with consequent dendritic involution and network disruption. By infecting CHAM mice with HIV, and using rabies viral-EGFP to trace striatal dendrites, we will assess the effects of astrocytic HIV infection on the dendritic architecture and synaptic structure of resident medium spiny neurons. In parallel, we will study the effects in CHAM mice of HIV infection complicated by methamphetamine use ? a common and disabling comorbidity that suppresses dopaminergic input to the striatum ? focusing on the structural and transcriptional responses of human glia to the combination of infection and addiction, as well as on the behavioral effects of that combination. To that end, we will use single cell RNA-Seq to assess the changes in gene expression by human astrocytes and their partnered mouse neurons caused by HIV infection, both alone and together with chronic methamphetamine use, to identify those changes that contribute to the striatal synaptic disruption and behavioral pathology of these mice. Our goal is to test the hypothesis that the HIV-infected striatum, by virtue of astrocytic fiber disengagement from dopaminergic synapses in particular, is especially vulnerable to the effects of amphetamine abuse, while defining the transcriptional basis for the glial pathology underlying that vulnerability.