Paola Arlotta - US grants

DESCRIPTION (provided by applicant): The goal of this proposal is to determine the molecular signals that instruct the fate specification and the lineage-specific development of corticospinal motor neurons (CSMN). These neurons are a clinically relevant population that, in humans, selectively dies in neurodegenerative diseases, including Amyotrophic Lateral Sclerosis (ALS), Hereditary Spastic Paraplegia (HSP), and Primary Lateral Sclerosis (PLS). They are also the cells permanently injured and responsible for paralysis in spinal cord injury (SCI). In the nervous system, studies aimed at investigating the molecular controls over birth, survival and connectivity of individual neuron types have been notoriously difficult, owing to the astonishing cellular heterogeneity of the tissue, combined with the inability to distinguish and purify one neuron type in isolation from others. In my postdoctoral work, I addressed this issue directly in the cortex, and have identified and begun to functionally characterize the first series of genes that in a combinatorial fashion uniquely identify CSMN as this neuron type develops [1]. Most relevant to the present proposal, we discovered that the transcription factor Fezf2 is a "master gene" that is both necessary for the birth of CSMN (i.e. CSMN are absent from the cortex of Fezf2-/- mice), and is at least in part sufficient to instruct the fate-specification of cortical progenitors to CSMN (i.e. elevated levels of Fezf2 can induce a "fate-switch" in progenitors destined to form upper layer neurons towards forming CSMN and deep layer neurons) [2]. Here, I build on this prior work and on new data from my own laboratory to directly investigate the central questions of this proposal: (1): What are the molecular signals that instruct the fate-specification and early development of cortical progenitors into CSMN? (Aim 1 and Aim 2) (2): Do postmitotic neurons of a different cortical type maintain the ability to generate CSMN in response to Fezf2 or, rather, are neuron lineage-specification decisions made and only modulated at the progenitor stage? (Aim 3) We present prior published work and substantial new data that support the feasibility of these experiments, and the direct relevance of the results to the development of novel therapeutic strategies to replace CSMN in neurodegenerative and traumatic diseases of the corticospinal circuitry. PUBLIC HEALTH RELEVANCE: Different neurodegenerative diseases of the CNS are typically characterized by the progressive death of specific neuron types. Corticospinal motor neuron (CSMN) degeneration and injury is a key component of motor neuron disease (including ALS), and of spinal cord injury. Here we propose to determine the molecular signals that instruct the birth of this clinically relevant neuron type, and to investigate the extent to which CSMN can be regenerated for therapeutic application.

DESCRIPTION (provided by applicant): We propose to develop and apply a paradigm-shifting technological platform that uses a series of Light- Inducible Transcriptional Effectors (LITEs) to orchestrate the temporal regulation of multiple genes in both individual cells in vitro and in the intact organism. The technology we propose to develop will be very broadly applicable and has the potential to radically transform the scale and rate of discovery across different biomedical fields. Application of this novel technology will enable high throughput discovery of the upstream transcriptional regulatory elements of any endogenous gene, as well as temporally precise modulation of gene expression in the native genome. Precise temporal and spatial patterns of gene expression are observed in different tissue and cell types, and are orchestrated and maintained by complex transcriptionally regulated circuits involving multiple genes. Due to the lack of integrated control and readout technologies that enable simultaneous perturbation and "fast" tuning of multiple genes, our ability to causally link transcriptional network dynamics with physiology and development remains at the infancy stage. The LITEs platform will enable modification and regulation of gene expression on the time-scale of hours using a non-invasive, light-mediated inductive strategy, thereby enabling a new generation of interactive genetic studies currently inaccessible with conventional techniques. Our proposal consists of two main components: 1) Novel technology development of light sensitive designer zinc finger (ZF) transcriptional modulators (LITEs-ZF). Genomic, synthetic biology, and protein engineering approaches will be used to develop a suite of novel light-inducible transcriptional regulators targeted at specific genes in the native genome of mammalian cells. Since ZF DNA binding domains can be engineered to target any DNA sequence, ZF-LITEs are applicable to a broad range of biological research studies in a variety of different organisms and cells types. 2) Application of the technology toward high-throughput in vitro and in vivo interrogation of transcriptional network dynamics in the central nervous system. In vitro application will be aimed at identification of upstream effectors of gene expression critical to the differentiation of corticospinal motor neurons (CSMN), a clinically relevant neuronal population that degenerates in amyotrophic lateral sclerosis (ALS) and is injured in spinal cord injury. In vivo application will be focused on directing the regeneration of CSMN by mimicking specific temporal sequences of CSMN-specific developmental cues within the adult brain. Given the broad applicability of this technology, the impact of this proposed work will be far reaching and will radically transform existing experimental approaches for studying gene interactions in all fields of life science and medicine. PUBLIC HEALTH RELEVANCE: Understanding the transcriptional networks that drive cellular development and repair has landscape- shifting impacts in the field of regenerative medicine. Here we develop a non-invasive, optical technology to enable genome-wide tuning of gene expression dynamics toward regeneration of corticospinal motor neurons.

DESCRIPTION (provided by applicant): Genetic targeting of cortical pyramidal neuron subtypes in the mouse Abstract A key obstacle to studying the development, organization, and function of neural circuits in the cerebral cortex is the stunning diversity of neuron types and a lack of comprehensive knowledge about their basic biology. The problem of neuronal diversity and identity in the cortex is fundamental, transcending developmental and systems neuroscience, and lies at the heart of defining the biology of cognition and psychiatric disorders. Glutamatergic pyramidal neurons (PyNs) constitute ~80% of cortical neurons, are endowed with large capacity for information coding, storage, plasticity, and carry the output of cortical computation. PyNs consist of diverse subtypes based on their specific laminar locations, axonal projection patterns, and gene expression profiles. Subsets of PyNs form multiple and hierarchical subnetworks of information processing, with distinct output channels to cortical and subcortical targets that subserve sensory, motor, cognitive and emotional functions. Importantly, PyN subtypes are differentially affected in various neuropsychiatric and neurodegenerative disorders. However, the severe lack of specific and effective genetic tools for studying PyNs has hampered progress in understanding cortical circuits. We propose to build a comprehensive genetic tool set for major PyN subtypes in the mouse through a joint project led by Dr. Josh Huang at Cold Spring Harbor Laboratory and Dr. Paola Arlotta at Harvard University with key collaboration from Dr. Hongkui Zeng at the Allen Institute for Brain Science. We have discovered a set of specific and combinatorial markers that distinguish major PyNs subtypes. We will use intersection, subtraction, and inducible strategies to target PyN subtypes. Aim 1 will generate ~25 knockin Cre and Flp driver lines that target major subclasses and lineages of pyramidal neurons. Aim 2 will generate multiple intersectional and subtractive reporters that are broadly useful for labeling distinct neuronal subtypes. Aim 3 will characterize the specificity of drivers, and organize and display data and resource in public databases. We will combine anatomical tracing with mRNA in situ and immunocytochemistry to determine the specificity of genetic targeting. We will use high- throughput and high- resolution pipeline at the Allen Institut to characterize PyN subtype labeling and axon projections. We will deposit all tools and reagents in major repositories (JAX) that are publically accessible (Allen Brain Atlas website). Genetic targeting will provide entry points to cortical circuits by integrating the full range of modern technologies and facilitate a systematic and comprehensive analysis of PyNs, from cell fate specification to circuit integration, connectivity, and function in cortical processing and behavior. These genetic tools will further provide sensitive probes to pathogenic mechanisms in models of brain disorders including autism, schizophrenia, bipolar and ALS, epilepsy, Alzheimer's, dementia, and may yield cell and molecular targets for therapeutic intervention. Although this grant is not hypothesis-driven as a traditional R01, it is certainly hypothesis-enabling for many future R01s to elucidate the fundamental role of the cerebral cortex in the larger framework of CNS architecture and function. 1

DESCRIPTION (provided by applicant): The goal of this proposal is to understand the molecular mechanisms by which distinct subtypes of excitatory projection neurons of the neocortex govern the laminar distribution of their interneuron partners, and to define whether acquisition of projection neuron subtype-specific identity is necessary for proper connectivity with local interneurons. The work aims at understanding the contribution of different classes of projection neurons to the establishment of balanced cortical microcircuitry. High-level neocortical function including cognition, sensory perception and motor function relies on the coordinated assembly of a local microcircuitry among an astonishing diversity of excitatory projection neurons and inhibitory interneurons. Indeed, disgenesis and/or disfunction of the local microcircuitry is associated with epilepsy, psychiatric disease and neurodevelopmental disorders [1-3]. The developmental events governing the integration of projection neurons and interneurons into balanced circuitry are poorly understood. We have reported on the central role played by projection neurons in governing this process and the precision by which different subtypes of projection neurons uniquely and differentially determine the laminar distribution of distinct classes of cortical interneurons [4]. We found that absence of subcerebral projection neurons from the neocortex of Fezf2 null-mutant mice and their replacement by commissural projection neurons cause abnormal lamination of interneurons and altered GABAergic inhibition. In agreement, experimental generation of either subcerebral projection neurons or callosal neurons in proximity to the cortex is sufficient to recruit cortical interneurons to these ectopic locations, with class-specificity. The data demonstrate that individual populations of projection neurons cell-extrinsically control the laminar fate of specific interneuron classes with striking precision. This suggests the existence of a molecular code that governs the specific interaction between projection neuron and interneuron partners during assembly of the local circuitry. Here, we build on this published work, as well as our recent demonstration that the identity of postmitotic projection neurons can be reprogrammed from one subtype into another in vivo [5] to answer the following questions: 1) Is there a molecular code enabling subtype-specific interactions among classes of projection neurons and interneurons to guide interneuron lamination? What are the molecules involved? (Aim 1) 2) Are codes of cadherin family members involved in establishing proper interneuron lamination? (Aim 2) 3) Is the acquisition of projection neuron subtype-specific identity necessary for the establishment of balanced circuitry/connectivity with interneuron partners? Does the inhibitory input by specific classes of interneurons change upon a change in projection neuron class-specific identity? (Aim 3) We present substantial published and pilot data supporting the significance and feasibility of this work.

DESCRIPTION (provided by applicant): Here, we propose to elucidate the role of candidate lincRNAs during mammalian brain development using a unique resource of 20 mouse lincRNA knockout models that we have recently generated. We specifically focused on candidates we hypothesized, through previous studies, to play a role in neuronal development. Indeed 15 of the 20 candidates exhibit cell-type and brain specific expression patterns. More importantly, we have already observed lethality phenotypes and gross anatomical abnormalities during brain development in selected mutants. We aim to explore the molecular and physiological underpinnings of the observed phenotypes as well as of others that we may encounter through this work. We propose to take a comprehensive and systematic approach to address the following questions: What are the spatial temporal dynamics of lincRNA expression in vivo during neuronal development and what gene pathways are regulated? (Aim 1). We will address these questions by leveraging a knock-in lacZ reporter to monitor lincRNA expression patterns and dynamics throughout brain development. We will address the second question through next generation RNA-sequencing studies between wild-type and knockout strains to identify specific genes and pathways regulated by lincRNAs in vivo. What are the physiological and molecular underpinnings of lethality and brain development phenotypes observed in selected lincRNA mutants? (Aim 2). Here, we will address this question through a battery of in vivo experiments to define perturbations to normal neuronal cell fate specification in selected mutants. We will focus first on three lincRNAs that have already shown brain specific phenotypes resulting in lethality and abnormalities of neurogenesis and one expressed in neuronal precursor cells. If time and resources allow we will expand these phenotypic studies to additional strains. How are lincRNAs working on a molecular and mechanistic level to ensure proper neuronal cell fate specification and viability? (Aim 3). Here, we use multifaceted and novel experimental protocols to define the specific regions of lincRNAs required to recover phenotypes. Moreover, we will identify lincRNA protein partners and how they interact to modulate cell fate decisions. The proposed research is driven by our vision, which seeks to unify experimental and computational genomics to push new frontiers in understanding the physiological and molecular regulatory roles of lincRNAs during neuronal development. Collectively, these studies will present the first phenotyping of lincRNA functional roles during neuronal development and will explore their biochemical mechanisms in vivo.

Genetic studies have identified many specific loci with significant associations to psychiatric disorders. However, unless we can develop useful approaches for systematically turning genetic information into neurobiological insights about brain disorders, there is a danger that costly and hard-won genetic findings will not be exploitable to understand pathophysiology and generate important therapeutic hypotheses. The goal of our collaborative, interdisciplinary effort is to develop powerful, generalizable approaches for discovering how risk variants for psychiatric disorders shape neurobiological processes at multiple levels of analysis, and to identify the processes whose dysregulation underlies disease. To do this, we propose to develop new experimental and inferential systems to bridge a longstanding gap between human genetics and experimental biology. We aim to identify biological causes and effects that span the genetic, molecular, and cellular levels of the nervous system. Our interdisciplinary team will develop new experimental systems that measure genetic influences across levels of analysis (RNA, proteins, and cellular function including physiology) in precise, scalable, well- controlled ways. We will make use of emerging cellular systems including three-dimensional cortical spheroids and organoids, and radically novel ?population in a dish? experimental systems that collect data on cells from hundreds of donors in a shared environment, inferring donor identity at the time of phenotypic readout. The analysis of such systems in turn requires sophisticated inferential strategies and new ideas from computer science. We propose to develop and widely share experimental and computational resources, including cell lines, methods, datasets, and analytic tools. The successful completion of this work will identify key neurobiological processes for multiple psychiatric disorders, and fortify many other scientists in making such connections in their own work. We hope in so doing to create a new kind of interdisciplinary science that ? by combining the strengths of data-driven, unbiased human genetics with the power of emerging experimental systems ? transforms the rate at which human- genetic leads lead to insights about disease mechanisms.

Project summary: Modern genomic sequencing technologies have allowed the field to identify important genetic polymorphisms associated with neurodevelopmental and neuropsychiatric disorders such as schizophrenia (SCZ) and autism spectrum disorder (ASD). However, we still have a limited understanding of the cellular and gene-expression defects associated with genetic mutation and variation in these pathologies. Finding answers to these key questions is made difficult by the complexity of these diseases (which affect multiple cell types in distinct brain regions), the lack of single, ideal experimental models for these specifically ?human? pathologies, and the need to investigate phenotypic abnormalities across many genetic backgrounds. Rodent models have important limitations due to the inherent differences in the development, architecture and function of their brains compared to humans; it is increasingly clear that work in rodents must be integrated with the use of primate models, including models of the human brain. Studies using endogenous human brain tissue are complicated by practical and ethical concerns of tissue availability, expansion and manipulation. However, recent progress has enabled the development of cellular models of the human developing brain via the generation of 3D brain organoids, which we propose can complement animal model systems to model basic aspects of human brain development and pathology. Although reductionist in nature, 3D human brain organoids are amenable to genetic engineering and high- throughput analysis, making them advantageous platforms for investigating a spectrum of genetic mutations. These models can provide a valuable platform to link mutations in disease-associated genes with specific abnormalities in human neurons and circuits, as well as to help identify molecular targets for intervention. The CHD8 gene is one of the most commonly mutated genes in sporadic ASD, producing an ASD subtype frequently associated with macrocephaly. Although it has been demonstrated that CHD8 regulates many other ASD risk genes, limited information is available on the cellular and molecular defects across different cell types in CHD8 mutant human tissue. We have recently established an optimized culture system that is able to develop healthy human brain organoids for up to 13 months, producing unusually mature organoids containing diverse cell types that molecularly resemble their endogenous counterparts, and mature neurons that develop dendritic spines and participate in spontaneously active networks (Quadrato et al., Nature, in press). We will use this protocol to characterize the expression profile of ASD risk genes in individual human brain cell types within organoids using high-throughput single-cell sequencing. In addition, we have created human brain organoids from pluripotent stem cells engineered to carry a heterozygous null mutation in CHD8, which we show recapitulate some of the phenotypic changes seen in patients. We will use this model to investigate the molecular and cellular defects resulting from CHD8 mutation at the single-cell level.

The hippocampus is a region of the brain that continues to produce new dentate granule cells (GCs) throughout life. Development of adult-born GCs and their integration into preexisting circuits is modulated by environment and by electrical activity of the local circuits, and is altered in the aging brain. To understand how GC integration occurs and how it is modulated by activity and aging, it is crucial to dissect the precise molecular mechanisms underlying these processes. The technology required for addressing these fundamental problems is unavailable in Argentina, but routinely applied in the Arlotta lab at Harvard. Joining efforts to address this specific problem is a natural next step. The overarching goal of this proposal is to exploit the strategies for transcriptional profiling and bioinformatic analysis in brain development validated by the Arlotta lab and the Schinder lab's expertise in functional characterization of adult-born GCs to build an experimental pipeline for the discovery of new molecules controlling circuit plasticity in the adult and aging brain, that includes assays for testing the roles of individual proteins to an unprecedented level of molecular and functional detail. In Aim 1, we propose to generate a pipeline to reveal transcription factors, epigenetic regulators, or effector genes controlling the developmental transitions along GC maturation and integration. The proposed experimental pipeline is similar for all three Aims: (i) FACS-purify birth-dated adult-born GCs at different stages; (ii) transcriptionally profile each population using two complementary forms of RNA sequencing; (iii) bioinformatically identify transcription factors or epigenetic regulators that may control stage progression; and (iv) functionally test candidate molecules through in vivo knock-down or overexpression. Using this same approach, we will then investigate how stage-specific transcriptome dynamics are altered in GCs from the aging hippocampus, to identify and functionally test changes in regulatory molecules that may be responsible for their protracted development (Aim 2). Finally, we will seek to identify molecular mediators of activity-mediated acceleration in GC development in the adult and the aging brain. (Aim 3). This grant will further the aims of the Fogarty International Center in expanding the technical capacities of Dr Schinder's lab, including availability of equipment, training of Argentinian graduate students and postdoctoral fellows, and building collegial networks between members of the Leloir and Harvard research communities. This work will significantly expand the capacity of Leloir to apply state-of-the-art transcriptomic profiling (currently limited, throughout Argentina) to solve new problems related to brain function and disease, locally supported by a bioinformatician who is co-investigator in the project. The workflow proposed will therefore not only enable a new generation of molecular studies in the Schinder lab by implementing the most advanced molecular tools and technologies, but also transfer expertise and know-how through personnel training such that these technologies and approaches will become available for routine use at the Leloir Institute at large.

Modeling ASD-linked mutations in 3D human brain organoids Neuropsychiatric diseases are very complex, and we still have a limited understanding of the abnormalities associated with genetic mutation in these pathologies. This is in part complicated by the lack of single, ideal experimental models for these overtly ?human? diseases, and the need to investigate abnormalities in diverse genetic backgrounds. Rodent models have proven valuable to highlight phenotypic abnormalities associated with autism spectrum disorder (ASD)-linked mutations. However, rodents differ in the development, architecture and function of their brain compared to humans making discoveries difficult to relate to patients. We share the vision of this consortium that integrated investigation of multiple experimental models, including models of the human brain, is needed to progress understanding of ASD. Studies using human brain tissue are complicated by practical and ethical concerns of tissue availability, expansion and manipulation. However, recent progress has enabled the development of cellular models of the human developing brain via the generation of 3D brain organoids, which we propose can complement rodent and non-human primate systems to model basic aspects of human pathology. Although reductionist in nature, 3D brain organoids are amenable to high-throughput genetic engineering and can provide a valuable platform to link mutations in disease-associated genes with specific abnormalities in human neurons and circuits, as well as help identify molecular targets. Here, we will use a protocol that we recently established to generate long-term cultures of human brain organoids engineered to carry the same mutations in the SHANK3 and MECP2 genes investigated in rodents and marmosets by the other members of the consortium. We will pioneer extensive molecular, morphological and electrophysiological analysis of mutant and control organoids to understand whether these mutations induce defects in human neurons and networks similar to those observed in mice, and to generate a transcriptional map of molecular changes that informs mechanistic understanding. In addition, we will optimize our recent Method for Analyzing RNA following Intracellular Sorting (MARIS) to molecularly profile specific subclasses of cortical neurons from rodent and marmoset brain and brain organoids. This will provide the first inter-species comparison of disease-relevant mutant and control neurons in three model systems to highlight molecular abnormalities and pinpoint cell type-specific defects.

Experimental models of the human developing brain are needed to investigate human-specific aspects of brain development, evolution, and neurological disease. Progress in the field has been hampered by the lack of models, considering that the endogenous developing human brain cannot be directly investigated; animal models often fail to recapitulate human disorders and cannot feasibly be used to study complex polygenic states spanning many genes. While reductionist in nature, stem-cell derived 3D human brain organoids offer a first-of- its-kind opportunity to study processes of human brain formation and wiring that are otherwise not accessible. However, there is an unmet need for organoid models that are cellularly complete and reproducible and for methodology to decode the establishment, connectivity and dynamics of neural circuits in organoids, at scale and with high fidelity. If we could map the activity and connectivity of organoids at scale, both to understand circuit function/dysfunction and to guide further development of organoids, we could close the loop on organoid design and application. Towards this goal, we have developed many molecular and imaging tools for high-throughput analysis of neural activity and connectivity, which we propose to apply to new, next-generation organoid models. Here, we propose a collaborative approach among four groups (Arlotta - brain organoids and human brain development; Boyden - circuit physiology and neural imaging technology; Lewis - material science and bioengineering and Insoo Hyun- bioethics) to pioneer a robust organoid system that combines the development of vascularized brain organoids incorporating more complete cell diversity and maturation with advanced high-throughput functional molecular and imaging tools to enable interrogation of circuit activity, connectivity, and molecular changes in cells participating in physiologically relevant circuits. We will build on a highly reproducible brain organoid model that we recently developed to promote the generation of cell types that are currently absent in organoids but needed for circuit maturation, refinement, and functionality. This work is intended to generate more advanced organoid models designed to promote maturation and robust network activity. In parallel, we will develop a pipeline to record neural activity from intact organoids using all- optical-electrophysiology techniques at scale, and optimize epitope-based barcoding and expansion microscopy to enable molecularly-annotated connectomics of brain organoids. The work proposed here will enable the use of human organoid models to study human circuit formation, plasticity, and function, analyses that are currently hampered by the lack of technologies and assays for high-throughput measurements of circuit physiology and connectivity in organoids. Beyond the work proposed here, these methods will directly enable investigation into how disease states alter information processing in the brain; for example, linking mutations in disease-associated genes with specific abnormalities in human neurons and circuits to inform the identification of molecular targets for therapeutic intervention.