Joseph Gogos - US grants

DESCRIPTION (provided by applicant): We have developed two extremely powerful genetic techniques that permit us for the first time to manipulate the survival and synaptic transmission of specific subsets of olfactory neurons in a temporally and spatially regulated manner. Our ability to kill or synaptically silence olfactory neurons during embryonic development, early postnatal life as well as adulthood provides us with unprecedented tools and opportunities to dissect the olfactory system as thoroughly as has been accomplished for other sensory systems. The goal of this proposal is to use these two novel and powerful techniques to investigate how precise connections between the periphery and the brain are established and continually maintained to preserve the integrity of the olfactory sensory maps, to examine whether there is modularity in the assembly of these maps and to probe the role of activity-dependent plasticity. Specifically, we wish to determine whether a) the olfactory bulb map is re-established after carefully controlled genetic ablation of all olfactory neurons. b) there is neuronal competition during the formation of the olfactory bulb map. c) synaptic contact between olfactory axons and mitral cells is necessary for the formation and differentiation of the mitral cell dendritic field d) primary olfactory input is required for the formation and maintenance of the olfactory cortex map. e) there is activity dependent competition for targets during the formation and maintenance of the olfactory cortex map. How order emerges in the development of the olfactory maps, while interesting in its own right, may apply broadly to other parts of the brain and provide a general understanding of the development of topographic patterns in brain. Moreover, understanding the mechanisms that underlie the matching patterns of connections from one brain structure to another is fundamental to our understanding of the pathology of neural diseases and may provide valuable insight into the fidelity of synapse formation when stem cells are used to cure neurological disorders.

DESCRIPTION (provided by applicant): Recent genetic studies provide strong support for the gene Disrupted-in-Schizophrenia-1 (DISC1) in mental illness. Additional data obtained primarily by in vitro studies suggested that DISC1 may play a key role in neurodevelopment and cell signaling by interacting with other proteins, including nuclear distribution E-like (NUDE-L) protein and phosphodiesterase 4B. Despite recent advances in the study of DISC1 function, several critical issues remain unanswered. We propose to utilize a unique and reliable mouse model that we have established in our laboratory, which closely mimics the translocation observed in an affected family while preserving endogenous proteins levels, in order to facilitate a better understanding of the physiological contribution of DISC1 in normal and abnormal brain development and function in the context of psychiatric disorders. Our main goals are to a) facilitate a better understanding of the contribution of DISC1 in neurodevelopment, as well as in synaptic transmission and plasticity and b) test whether existing in vitro results can be validated in vivo in the mouse model we have established. Our analysis promises to provide valuable insights into the ways DISC1 increases the risk of psychiatric disorders and at the same time provide a well-characterized animal model that can be used to test further hypotheses and facilitate future drug development efforts.

[unreadable] DESCRIPTION (provided by investigator): The field of schizophrenia genetics has reached a point where several strong susceptibility genes have been proposed, but understanding of the functional significance of the implicated alleles or haplotypes and how they interact with each other, is still lacking. Using gene targeting and transgenesis approaches, as well as available mouse lines, we have established a cohort of 6 mouse models of strong candidate schizophrenia susceptibility genes, which we propose to use to elucidate the structure of the genetic networks that modulate the genetic risk of the disease. Analysis of the expression profile in the hippocampus and prefrontal cortex of these 6 mutant lines and the 5 double-mutant lines we propose to generate, along with cross-species comparison of gene expression patterns, will allow us to obtain an unbiased evaluation of the transcriptional programs affected by impaired function or expression of these genes, reflecting downstream effects of the mutation, or adaptive and compensatory changes. It will also allow us to assess the relative similarities or differences among gene networks affected in different mouse models and therefore facilitate the identification of key signaling pathways that modulate the disease risk. By deconstructing the genetic component of schizophrenia using reliable animal models under conditions that are not confounded by the effects of the treatment or the disease itself, our approach offers unique advantages over traditional expression profiling in diseased brains. Another strong point of this proposal is the fact that all these mouse models are available within our laboratory and will be examined under identical controlled conditions using the same methodologies. Given the heterogeneity of the disease and variability in experimental conditions across sites, this is a unique opportunity to study accurately the effect of each of these genes and combinations of them. The importance and timely nature of this proposal cannot be overemphasized. Our ability to analyze transcriptional profiles of entire genomes for any one mutation is likely to transform the traditional view of a simple disease gene or genetic pathway, into a more complex concept of genetic networks and at the same time provide a wealth of useful drug targets. [unreadable] [unreadable]

DESCRIPTION (provided by applicant): Challenge Area: 14 --Stem Cells Challenge Topic: 14-MH-101*, "Developing iPS cells for mental disorders" The recent description of somatic cell reprogramming to an embryonic stem (ES) cell-like phenotype, termed induced pluripotent stem (iPS) cell technology, presents an exciting venue toward cell-based models of disease mutations. This technology affords for the first time the potential to analyze neuronal cells grown from individuals carrying a given genetic lesion and offers a unique way of direct assessment of pathology. Recent studies have established an important role for rare copy number variants (CNVs, genomic deletions and duplications) in the etiology of schizophrenia and autism. Rare, but recurrent CNVs affecting the CNTNAP2 gene, a member of the neurexin family, have been described in both disorders. Given the high penetrance and clearly delineated genomic structure, which offers clear clues regarding the underlying functional deficit, this CNV is ideal to model at the cellular level, as well as at the level of the organism. We propose to generate iPS cells and iPS cell- derived forebrain neurons from schizophrenic patients carrying variable size CNTNAP2 gene deletions and their unaffected relatives and undertake an initial characterization of their basic morphological and electrophysiological properties. A unique aspect of our proposal is that the proposed comparisons will use as a reference point data acquired both in vitro and in vivo from hippocampal and cortical neurons from an animal model genetically engineered to carry a Cntnap2 gene deletion. Cell lines generated from individuals with schizophrenia carrying specific well-defined structural mutations offer an unprecedented opportunity to recapitulate pathologic human neural tissue formation in vitro and provide a unique platform for studies aimed at both providing valuable insights into the disease mechanisms, and the potential discovery of new compounds to treat this devastating disorder. PUBLIC HEALTH RELEVANCE: There is considerable promise in generating induced pluripotent stem (iPS) cell lines from patients afflicted with central nervous system diseases, including psychiatric disorders. We propose to generate neurons from patients with schizophrenia carrying CNTNAP2 deletions, a genetic risk factor for the disease, study their properties and compare them to brain neurons of a knock/out mouse model of the same gene. This is an unprecedented opportunity for the field to combine data from human neurons carrying a bona fide genetic risk factor for schizophrenia and an established mouse model of the same genetic lesion.

Description (provided by applicant): This proposal represents a move from genomics to biology designed to identify the patterns of gene expression induced by disease-associated mutations and help frame our understanding and elucidate the structure of the still elusive genetic interactions underlying disease risk ("Schizophrenia interactome"). Recent studies have unequivocally demonstrated an important contribution of rare structural mutations to the genetic architecture of numerous psychiatric disorders, including schizophrenia. Determining how such mutations act in concert with modifiers to cause and influence the clinical phenotype is an important question that remains to be addressed. We contend that common genetic variation (including single nucleotide polymorphisms, SNPs) plays a key role in determining the penetrance or the expressivity of rare mutations. This hypothesis has not been directly tested and given the difficulties and uncertainties associated with testing common variation in patient populations it may be impossible to test it unequivocally using human genetic approaches. However, availability of animal models could offer important insights into how rare and common variation interact to affect key neurobiological processes and the gene expression networks underlying such processes. To accomplish this goal, we propose to utilize four "key" mouse lines generated in our lab: i. Two lines that faithfully model two rare mutations that unequivocally predispose to schizophrenia: a truncation of the DISC1 (short for Disrupted-In-Schizophrenia 1) gene and a microdeletion on chromosome 22q11.2;ii. Two lines that faithfully model two alleles of a common variant (BDNF Val66Met) associated with a number of psychiatric diseases and related traits, which undoubtedly modulates neurotrophic action in the developing brain. We propose to analyze the transcriptional profile in the hippocampus and prefrontal cortex across critical developmental periods to obtain an unbiased evaluation of the transcriptional programs affected by the combined effect of rare and common disease-associated variation, reflecting downstream effects of the mutation and/or adaptive/compensatory changes. Our work promises to advance our current genetic knowledge, identify novel disease-related genes or genetic pathways that could be tested in future human genetic studies of SCZ, as well as provide targets for novel pharmacotherapy approaches

DESCRIPTION (provided by applicant): The identification of rare mutations that result in predisposition to SCZ with relatively high penetrance has led to the development of mouse models of proven etiologic relevance. DISC1 is a susceptibility gene identified through a rare genetic lesion, a balanced chromosomal translocation segregating with SCZ and mood disorders in a large pedigree. We used a disease-focused knock-in approach to introduce a truncating lesion in the murine Disc1 orthologue designed to model the effects of this translocation. During the first round of this grant we showed that Disc1 mutant mice display specific and robust deficiencies in spatial working memory tests. We also uncovered widespread cytoarchitectural alterations in the dentate gyrus during neonatal and adult neurogenesis, which include errors in axonal pathfinding and are accompanied by changes in neural activity and short-term plasticity. We also showed that dysregulation of cAMP levels contributes to the structural connectivity deficits. Finally we provided evidence that mutant mice have altered functional connectivity of prefrontal areas with temporal lobe structures. Building on these findings, here we propose to complete our analysis on the effect of the Disc1 mutation on hippocampus, extend our structural and functional analysis to prefrontal cortex and finally analyze the effect of the modeled mutation on the communication between these two areas. In addition to our previous results from Disc1 mutant mice our proposed research is dictated by parallel analysis of other models of rare mutations, which affords the opportunity to compare results, identify key common pathways and enable development of a comprehensive and integrative model of schizophrenia pathogenesis and pathophysiology. This knowledge will facilitate discovery of novel treatments and biomarkers.

DESCRIPTION (provided by applicant): Patients with schizophrenia typically suffer from severe and disabling cognitive function, including disturbances in executive function and working memory. To clarify the neurobiology underlying these disturbances, we have studied cognitive function in mouse lines engineered to model a microdeletion on chromosome 22, an etiologically relevant mutation unequivocally associated with susceptibility to schizophrenia. Patients with schizophrenia, as well as subjects with these mutations, have pronounced disturbances in cognitive tasks that depend on the hippocampus and prefrontal cortex. Mice carrying the microdeletion perform poorly in tests of spatial working memory. We have recently shown that deficits in functional connectivity between the hippocampus and prefrontal cortex contribute to this spatial working memory dysfunction in these mice. Building on these findings, we propose to (1) examine the molecular basis of these effects by studying working memory and hippocampal-prefrontal connectivity in mice carrying mutations of single genes within the microdeletion region; (2) study the role of the ventral hippocampus in the behavioral and physiological phenotypes in the mutants, and (3) study the role of the thalamus in these phenotypes. The proposed experiments serve both basic and translational goals. Understanding of the neurobiological mechanisms of working memory in the mouse is an important step in determining the relevance of such models to cognitive tasks studied in humans. Exploring these mechanisms in mice carrying schizophrenia-predisposing mutations uses this understanding to identify the behaviorally relevant neural consequences of these mutations. The end goal of this work is to develop an integrative model of schizophrenia pathogenesis and pathophysiology that demonstrates how these genetic lesions alter neural cells, circuits and systems to disrupt cognitive function.

DESCRIPTION (provided by applicant): MicroRNAs (miRNAs) are a class of small non-coding regulatory RNAs. Abnormalities in miRNA expression and miRNA-mediated gene regulation have been observed in a variety of human diseases, including psychiatric and neurodevelopmental disorders. In most cases, miRNAs appear to be components of both the genetic architecture of these complex phenotypes and integral parts of the biological pathways that mediate the effects of the primary genetic deficits and could serve as novel therapeutic targets. Some of the strongest evidence for a direct pathogenic link between psychiatric disorders, cognitive dysfunction and miRNAs is provided by studies on the mouse model of a well- established genetic risk factor, the 22q11.2 microdeletion (Df(16)A+/- mice). Analysis of this mouse model provided compelling evidence that the 22q11.2 microdeletion results in abnormal processing of brain miRNAs. Our recent work has identified two major components of the 22q11.2-associated miRNA dysregulation, as well as a major downstream target of the miRNA dysregulation. Despite substantial progress, the extent to which miRNA dysregulation contributes to the cellular, synaptic and behavioral phenotypes associated with the 22q11.2 microdeletion in vivo remains to be determined and additional key downstream targets remain to be identified. This is the focus of this grant proposal. Understanding how miRNA-dependent gene regulation disrupted by a structural mutation with unequivocal causal links to schizophrenia and cognitive dysfunction contributes to the emergence of the psychiatric and cognitive phenotypes associated with this genomic imbalance will provide important mechanistic insights and can guide analysis of miRNA contribution to other psychiatric, neurodevelopmental and cognitive disorders. PUBLIC HEALTH RELEVANCE: This proposal is inherently translational in nature, aimed at elucidating the neurobiological substrates of psychiatric disease. It is aimed at identifying specific patterns of abnormal gene expression caused by loci predisposing to schizophrenia and cognitive dysfunction and link them to the cellular, synaptic and cognitive processes they impact. Identifying such patterns would set the stage for a novel approach to therapies aimed at reversing the underlying pathophysiology and restoring normal function.

DESCRIPTION (provided by applicant): Hyperprolinemia has been shown to correlate with the risk of psychosis, schizophrenia, schizoaffective disorder, and seizures in a number of human diseases. One cause of hyperprolinemia is loss-of-function mutations of the PRODH gene, which is involved in L-proline degradation. The PRODH gene maps in the 22q11.2 locus, and is heterozygously deleted in the 22q11.2 microdeletion syndrome, which is associated with high-risk for neurodevelopmental abnormalities and psychosis. One hypothesis explaining the effects of elevated L-proline within the CNS is that L-proline may act as a neuroactive small molecule that interferes with the normal function of other neurotransmitter systems within the brain. In preliminary studies we have found that L-proline is a GABA-mimetic metabolite capable of activating GABA-A receptor ion channels. Based on chemical structural database searches we have identified two additional proline-like metabolites that are known to accumulate in two other human neuropsychiatric diseases: -aminolevulinic acid, which accumulates in acute intermittent porphyria, a disease associated with psychosis and seizures, and L-pipecolic acid, which accumulates in pyridoxine (vitamin B6) dependent epilepsy, a disease associated with seizures. Similar to L-proline, we have found that both of these metabolites are also GABA-mimetic and capable of activating GABA-A receptors. We hypothesize that accumulation of these GABA-mimetic metabolites within the CNS may disrupt normal GABA-ergic synaptic transmission in these diseases with overlapping neuropsychiatric symptomatology. To test this hypothesis we have proposed the following Specific Aims: (1) to determine whether these metabolites interfere with normal GABA processing, handling, and detection by the components of the GABA-system and (2) to determine the impact of L-proline accumulation upon GABA-ergic synaptic transmission and network properties within the medial prefrontal cortex. The results of our proposed studies will provide important initial insights into the role of accumulated metabolites in GABA-ergic dysfunction in these clinically relevant human diseases with overlapping neurodevelopmental and neuropsychiatric dysfunction. These results will also lay the foundation for guiding future studies targeted at the development of pharmacological rescue strategies for these diseases.

PROJECT SUMMARY Collective data from recent whole exome sequencing studies in schizophrenia confirmed a prominent enrichment of gene-disruptive de novo loss-of-function mutations and led to the identification of the contribution of SETD1A, which encodes for a histone methyltransferase. Notably, SETD1A mutations confer a large increase in disease risk, which provides a good starting point for disease modeling. Unambiguous identification of SETD1A as a SCZ risk gene emphasizes the important role that neural gene regulation plays in the genetic architecture of schizophrenia, consistent with accumulating evidence supporting an important role of regulatory common and rare variants in neuropsychiatric disease risk. This finding is also consistent with several lines of evidence suggesting that histone methylation is more broadly relevant to SCZ including the recent observation that histone methylation showed the strongest statistical enrichment among 4,939 biological pathways in GWAS data of psychiatric disorders. The fact that both common and rare risk variants aggregate in this particular biological pathway highlights its importance for the etiology of schizophrenia. However it is not clear at this stage how to translate a ubiquitous molecular process such as chromatin modification into a mechanistic and disease- specific insight. In this regard, the SETD1A finding provides a handle, a starting point from which to build a model and test hypotheses. The goal of this proposal is to address the critical question of how chromatin regulation deficits play a role in the pathogenesis of SCZ by (i) investigating the developmental requirement of Setd1a on cognitive and synaptic function in mice and the nature of the neural circuits affected by its deficiency (ii) identifying direct neuronal targets of Setd1a in the prefrontal cortex and (iii) generating and analyzing human SETD1A-deficient cortical neurons. The ultimate goal of the proposed studies of chromatin regulation in mental illness is to understand when/where/how genetic vulnerabilities affect gene expression in the brain and shape brain circuitry and function. The proposed studies will also reveal a host of schizophrenia candidate genes and promise important advances in our understanding, diagnosis, and treatment of debilitating psychiatric disorders, such as schizophrenia.

Project Summary Patients with schizophrenia, as well as subjects with 22q11.2 deletion, which raises the risk of schizophrenia 30-fold, have pronounced disturbances in cognitive tasks that depend on the hippocampus and prefrontal cortex, including working memory. To clarify the neurobiology underlying these disturbances, we have studied working memory in a mouse model of the 22q11.2 deletion (Df(16)A+/? mice). In the first iteration of this grant, we demonstrated that the deletion results in deficits in axon branching, working memory and neural synchrony in the hippocampal-prefrontal circuit. We further demonstrated that these deficits were at least partially due to haploinsufficiency of the Zdhhc8 gene, which results in the axonal mislocalization of key signaling proteins, including AKT, a negative regulator of the kinase Gsk3. Moreover, consistent with the hypothesis that Zdhhc8 deficiency may affect working memory and hippocampal-prefrontal HPC-PFC synchrony via hyperactivity of Gsk3 signaling, we found that developmental Gsk3 inhibition reversed deficits in spatial working memory task acquisition, neural synchrony and prefrontal representations of goal information in Df(16)A+/? mice. Building on these findings, this competitive renewal proposes to (1) Determine when during the lifespan (postnatal, adolescent and adult time periods) Gsk3 antagonism is most effective at reversing physiological and behavioral deficits (2) Test whether isoform-specific Gsk3 (? or ?) antagonists can reverse working memory- related phenotypes in Df(16)A+/? mice, using recently developed investigational drugs more suitable for future clinical trials, aimed specifically at either isoform and (3) confirm the causal role of hippocampal-prefrontal synchrony in working memory deficits by manipulating synchrony physiologically using state-of-the-art approaches established in the course of this grant and measuring the resultant effects on working memory behavior in wild-type and 22q11.2 model mice. The proposed experiments serve both basic and translational goals. From the basic perspective, they will clarify the causal relationship between a specific circuit (the hippocampal-prefrontal circuit) and a well-characterized behavior (spatial working memory) while simultaneously clarifying the molecular and physiological substrates for plasticity within the system, both during development and in the adult. From the translational perspective, they will help guide the development of therapies aimed at manipulating this system, including pharmacological and/or brain stimulation treatments, and help define whether such treatments might be efficacious in affected adults, or should be targeted to earlier time points in highly susceptible individuals.

PROJECT SUMMARY The heterogeneity of genetic etiology and the corresponding neural complexity of schizophrenia have rendered the task of understanding disease pathophysiology and developing new improved treatments rather inauspicious. In light of this complexity there is need to identify convergent molecular and neural substrates that can serve as entry points to prevent or reverse disease progression. Along the same lines, identification of mutations or variants that confer protection against disease by disabling protein function via loss-of-function (LoF) effects, akin to those of a therapeutic agent, hold great promise for devising therapeutic schemes to restore or prevent some or all of disease symptoms. During the first iteration of this grant, we characterized the microRNA dysregulation in a model of the 22q11.2 deletion, one of the strongest genetic risk factor for schizophrenia [Df(16)A+/- mice]. We found that postnatal brain upregulation of Mirta22/Emc10, an inhibitor of neuronal maturation, represents the major transcriptional effect of the 22q11.2-associated microRNA dysregulation. Mice where the Df16(A) deficiency is combined with a LoF Mirta22 allele show a profound rescue of core SCZ-related deficits such as sensorimotor gating deficits, working and social memory deficits, as well as several of the underlying synaptic and cellular deficits. Thus several key disease alterations observed in Df(16)A+/? mice can be attributed to the abnormally sustained inhibitory influence of elevated Mirta22 levels. Building on these findings, this competitive renewal aims to elucidate further the nature of neural substrates underlying the protective influences of Mirta22 LoF mutations, compare the effects of normalizing Mirta22 levels during neonatal, adolescent and adult time periods using conditional genetic manipulations in mouse models (including the use of new therapeutic modalities of translatable value) and determine the relevance of our mouse results in human disease neurons. Determining when during the lifespan Mirta22 normalization is most effective at reversing disease phenotypes will be crucial for determining its potential use as a therapeutic target. !

Schizophrenia is a debilitating psychiatric disorder that effects 1% of the population, with an additional 2-3% developing a schizoaffective disorder. SCZ patients exhibit a spectrum of cognitive deficits including defective episodic memory, present prior to the onset of psychosis and frequently expressed in relatives of affected individuals. Episodic memory formation is dictated in part by spatially tuned (place cell) activity of principal cells in the hippocampus. The biological mechanisms driving this learning capacity in the healthy hippocampus remain largely unknown, let alone their disruption in schizophrenia, leaving large gaps in our knowledge that need to be addressed. Using in vivo functional imaging in mouse dorsal hippocampal area CA1 during head-fixed during learning behaviors, we recently uncovered specific alterations in in vivo physiological properties of CA1 pyramidal cells in the Df(16)A+/? transgenic mouse model of 22q11.2 deletion syndrome, the largest known genetic risk to develop SCZ. Df(16)A+/? CA1 place cells exhibit reduced long-term stability, impaired context- related and lack of reward-related reorganization. A novel form of synaptic plasticity, termed behavioral time- scale synaptic plasticity (BTSP), has been found to drive rapid formation of spatially selective firing fields in CA1 pyramidal cells; notably, our preliminary studies suggest that this form of plasticity is dysregulated in Df(16)A+/? mice. We thus hypothesize that BTSP, a major form of plasticity that drives place cell-recruitment during learning, is disrupted by SCZ risk mutations. These findings at the neuronal population level provide entry points for dissecting the underlying cellular, molecular and microcircuit dysfunctions caused by schizophrenia risk mutations. To gain these mechanistic insights we will unite the complementary expertise of the Losonczy lab and the Gogos lab in etiologically valid genetic mouse models of neuropsychiatric disorders to carry out multiscale dissection of microcircuit, cellular and molecular pathophysiology of schizophrenia-related memory deficits in the adult mouse hippocampal CA1 circuitry. Aim 1 is aimed at assessing altered synaptic plasticity in CA1 pyramidal cells during episodic learning in Df(16)A+/? mice. Aim 2 deals with dissecting inhibitory microcircuit dynamics during episodic learning, while Aim 3 is focused at dissecting altered excitatory and neuromodulatory input dynamics to CA1 during episodic learning in Df(16)A+/? mice. Taken together, Aims 1-3 provide a tractable path to a deeper, mechanistic understanding of hippocampus-related cognitive memory deficits in schizophrenia.

PROJECT SUMMARY There is now unequivocal evidence that the behavioral and cognitive phenotypes associated with psychiatric disorders are mediated by perturbations to specific brain circuits, i.e. sets of strongly anatomically and functionally connected brain structures. However, there are currently no unbiased computational approaches to implicate disease-related circuits, in a brain-wide fashion and at a high spatial resolution, and then to connect abnormalities in these circuits to specific patient phenotypes. The main goal of the proposal is to develop and optimize a computational approach which will make it possible, for the first time and at an unprecedented resolution, to discover functional brain circuits involved in mental disorders. The proposed approach is based on synergistic analyses of genetics data, ultra-high-resolution expression and brain-wide connectome data ? available for the same mouse strain, and in a common coordinate system. An important virtue of the approach is that it is based exclusively on genome- and brain-wide data and therefore is not biased towards any prior hypothesis about disorders' etiology. We specifically propose Aim 1. Identify brain circuits and associated cell types primarily affected by genetic insults in autism spectrum disorder (ASD) and schizophrenia (SCZ). We will develop data-driven computational approaches to identify genetic biases towards anatomically connected functional brain circuits. Aim 2. Experimentally test the identified circuits in several mouse models of ASD and SCZ. Functional circuits identified by the computational approach will be tested using two independent mouse models of ASD and two models of SCZ. The dynamics of the circuits will be explored using multi-site photometric imaging. Aim 3. Correlate mutation biases towards brain regions, circuits, and cell types with specific ASD phenotypes. Using extensive and deep phenotypic human data together with genetic data from the same patient cohorts, we will correlate mutation biases towards brain cell types and circuits with multiple specific ASD phenotypes.