Biology subjects

Roussos, P.

Publications and source records attributed to Roussos, P..

At least 37 records · Page 2Linked to original sources

bioRxiv2022

Multi-omic profiling of the developing human cerebral cortex at the single cell level

The cellular complexity of the human brain is established via dynamic changes in gene expression throughout development that is mediated, in part, by the spatiotemporal activity of cis-regulatory elements. We simultaneously profiled gene expression and chromatin accessibility in 45,549 cortical nuclei across 6 broad developmental time-points from fetus to adult. We identified cell-type specific domains in which chromatin accessibility is highly correlated with gene expression. Differentiation pseudotime trajectory analysis indicates that chromatin accessibility at cis-regulatory elements precedes transcription and that dynamic changes in chromatin structure play a critical role in neuronal lineage commitment. In addition, we mapped cell-type and temporally specific genetic loci implicated in neuropsychiatric traits, including schizophrenia and bipolar disorder. Together, our results describe the complex regulation of cell composition at critical stages in lineage determination, serve as a developmental blueprint of the human brain and shed light on the impact of spatiotemporal alterations in gene expression on neuropsychiatric disease. One-Sentence SummarySimultaneous profiling of gene expression and chromatin accessibility in single nuclei from 6 developmental time-points sheds light on cell fate determination in the human cerebral cortex and on the molecular basis of neuropsychiatric disease.

neuroscience↗

bioRxiv2022

Transcriptome and chromatin accessibility landscapes across 25 distinct human brain regions expand the susceptibility gene set for neuropsychiatric disorders

Brain region- and cell-specific transcriptomic and epigenomic molecular features are associated with heritability for neuropsychiatric traits, but a systematic view, considering cortical and subcortical regions, is lacking. Here, we provide an atlas of chromatin accessibility and gene expression in neuronal and non-neuronal nuclei across 25 distinct human cortical and subcortical brain regions from 6 neurotypical controls. We identified extensive gene expression and chromatin accessibility differences across brain regions, including variation in alternative promoter-isoform usage and enhancer-promoter interactions. Genes with distinct promoter-isoform usage across brain regions are strongly enriched for neuropsychiatric disease risk variants. Using an integrative approach, we characterized the function of the brain region-specific chromatin co-accessibility and gene-coexpression modules that are robustly associated with genetic risk for neuropsychiatric disorders. In addition, we identified a novel set of genes that is enriched for disease risk variants but is independent of cell-type specific gene expression and known susceptibility pathways. Our results provide a valuable resource for studying molecular regulation across multiple regions of the human brain and suggest a unique contribution of epigenetic modifications from subcortical areas to neuropsychiatric disorders.

neuroscience↗

bioRxiv2022

DeepGAMI: Deep biologically guided auxiliary learning for multimodal integration and imputation to improve phenotype prediction

Genotype-phenotype association is found in many biological systems, such as brain-related diseases and behavioral traits. Despite the recent improvement in the prediction of phenotypes from genotypes, they can be further improved and explainability of these predictions remains challenging, primarily due to complex underlying molecular and cellular mechanisms. Emerging multimodal data enables studying such mechanisms at different scales from genotype to phenotypes involving intermediate phenotypes like gene expression. However, due to the black-box nature of many machine learning techniques, it is challenging to integrate these multi-modalities and interpret the biological insights in prediction, especially when some modality is missing. Biological knowledge has recently been incorporated into machine learning modeling to help understand the reasoning behind the choices made by these models. To this end, we developed DeepGAMI, an interpretable deep learning model to improve genotype-phenotype prediction from multimodal data. DeepGAMI uses prior biological knowledge to define the neural network architecture. Notably, it embeds an auxiliary-learning layer for cross-modal imputation while training the model from multimodal data. Using this pre-trained layer, we can impute latent features of additional modalities and thus enable predicting phenotypes from a single modality only. Finally, the model uses integrated gradient to prioritize multimodal features and links for phenotypes. We applied DeepGAMI to multiple emerging multimodal datasets: (1) population-level genotype and bulk-tissue gene expression data for predicting schizophrenia, (2) population-level genotype and gene expression data for predicting clinical phenotypes in Alzheimers Disease, (3) gene expression and electrophysiological data of single neuronal cells in the mouse visual cortex, and (4) cell-type gene expression and genotype data for predicting schizophrenia. We found that DeepGAMI outperforms existing state-of-the-art methods and provides a profound understanding of gene regulatory mechanisms from genotype to phenotype, especially at cellular resolution. DeepGAMI is an open-source tool and is available at https://github.com/daifengwanglab/DeepGAMI.

bioinformatics↗

bioRxiv2022

Convergent impact of schizophrenia risk genes

Genetic studies of schizophrenia reveal a complex polygenic risk architecture comprised of hundreds of risk variants; most are common in the population at-large, non-coding, and act by genetically regulating the expression of one or more gene targets ("eGenes"). It remains unclear how genetic variants predicted to confer individually small effects combine to yield substantial clinical impacts in aggregate. Here, we demonstrate that eGenes have shared downstream transcriptomic effects ("convergence") that may underlie unexpected interactions ("non-additive effects") observed when eGenes are manipulated in combination. We apply a pooled CRISPR approach to perturb schizophrenia eGenes in human induced pluripotent stem cell-derived glutamatergic neurons. The strength and specificity of convergence increased between functionally similar eGenes. Predicting that convergence might impact additive relationships between risk loci when inherited together, we use an arrayed approach to explore bidirectional combinatorial perturbations of a partially overlapping set of fifteen schizophrenia eGenes. When specifically considering groups of synaptic or epigenetic eGenes, combinatorial eGene perturbations yield changes that are smaller than predicted by summing individual eGene effects ("sub-additive effects"). Moreover, convergent and non-additive downstream transcriptomic effects overlap, suggesting that functional redundancy of eGenes may be a major mechanism underlying non-additivity. Combinatorial perturbations result in outcomes that are not yet well-predicted by single eGene perturbations alone, indicating that the effects of polygenic risk cannot necessarily be extrapolated from experiments testing one risk gene at a time.

neuroscience↗

bioRxiv2021

From compartments to gene loops: Functions of the 3D genome in the human brain

The 3D genome plays a key role in the regulation of gene expression. However, little is known about the spatiotemporal organization of chromatin during human brain development. We investigated the 3D genome in human fetal cortical plate and in adult prefrontal cortical neurons and glia. We found that neurons have weaker compartments than glia that emerge during fetal development. Furthermore, neurons form loop domains whereas glia form compartment domains. We show through CRISPRi on CNTNAP2 that transcription is coupled to loop domain insulation. Gene regulation during neural development involves increased use of enhancer-promoter and repressor-promoter loops. Finally, transcription is associated with gene loops. Altogether, we provide novel insights into the relationship between gene expression and different scales of chromatin organization in the human brain.

genomics↗

bioRxiv2021

Transcriptome-wide association study in UK Biobank Europeans identifies associations with blood cell traits.

Previous genome-wide association studies (GWAS) of hematological traits have identified over 10,000 distinct trait-specific risk loci, but the underlying causal mechanisms at these loci remain incompletely characterized. We performed a transcriptome-wide association study (TWAS) of 29 hematological traits in 399,835 UK Biobank (UKB) participants of European ancestry using gene expression prediction models trained from whole blood RNA-seq data in 922 individuals. We discovered 557 TWAS signals associated with hematological traits distinct from previously discovered GWAS variants, including 10 completely novel gene-trait pairs corresponding to 9 unique genes. Among the 557 associations, 301 were available for replication in a cohort of 141,286 participants of European ancestry from the Million Veteran Program (MVP). Of these 301 associations, 199 replicated at a nominal threshold ( = 0.05) and 108 replicated at a strict Bonferroni adjusted threshold ( = 0.05/301). Using our TWAS results, we systematically assigned 4,261 out of 16,900 previously identified hematological trait GWAS variants to putative target genes. Compared to coloc, our TWAS results show reduced specificity and increased sensitivity to assign variants to target genes.

genetics↗

bioRxiv2021

A comprehensive series of temporal transcription factors in the fly visual system

The brain consists of thousands of different neuronal types that are generated through multiple divisions of neuronal stem cells. These stem cells have the capacity to generate different neuronal types at different stages of their development. In Drosophila, this temporal patterning is driven by the successive expression of temporal transcription factors (tTFs). While a number of tTFs are known in different animals and across various parts of the nervous system, these have been mostly identified by informed guesses and antibody availability. We used single-cell mRNA sequencing to identify the complete series of tTFs that specify most Drosophila medulla neurons in the optic lobe. We tested the genetic interactions among these tTFs. While we verify the general principle that tTFs regulate the progression of the series by activating the next tTFs in the series and repressing the previous ones, we also identify more complex regulations. Two of the tTFs, Eyeless and Dichaete, act as hubs integrating the input of several upstream tTFs before allowing the series to progress and in turn regulating the expression of several downstream tTFs. Moreover, we show that tTFs not only specify neuronal identity by controlling the expression of cell type-specific genes. Finally, we describe the very first steps of neuronal differentiation and find that terminal differentiation genes, such as neurotransmitter-related genes, are present as transcripts, but not as proteins, in immature larval neurons days before they are being used in functioning neurons; we show that these mechanisms are conserved in humans. Our results offer a comprehensive description of a temporal series of tTFs in a neuronal system, offering mechanistic insights into the regulation of the progression of the series and the regulation of neuronal diversity. This represents a proof-of-principle for the use of single-cell mRNA sequencing for the comparison of temporal patterning across phyla that can lead to an understanding of how the human brain develops and how it has evolved.

developmental biology↗

bioRxiv2021

Acetylated Chromatin Domains Link Chromosomal Organization to Cell- and Circuit-level Dysfunction in Schizophrenia and Bipolar Disorder

To explore modular organization of chromosomes in schizophrenia (SCZ) and bipolar disorder (BD), we applied population-scale correlational structuring of 739 histone H3-lysine 27 acetylation and H3-lysine 4 trimethylation profiles, generated from the prefrontal cortex (PFC) of 568 cases and controls. Neuronal histone acetylomes and methylomes assembled as thousands of cis-regulatory domains (CRDs), revealing fine-grained, kilo-to megabase scale chromatin organization at higher resolution but firmly integrated into Hi-C chromosomal conformations. Large clusters of domains that were hyperacetylated in disease shared spatial positioning within the nucleus, predominantly regulating PFC projection neuron function and excitatory neurotransmission. Hypoacetylated domains were linked to inhibitory interneuron- and myelination-relevant genes. Chromosomal modular architecture is affected in SCZ and BD, with hyperacetylated domains showing unexpectedly strong convergences defined by cell type, nuclear topography, genetic risk, and active chromatin state across a wide developmental window.

genomics↗

bioRxiv2021

Transcribed enhancers in the human brain identify novel disease risk mechanisms

Identification of risk variants for neuropsychiatric diseases within enhancers underscores the importance of understanding the population-level variation of enhancers in the human brain. Besides regulating tissue- and cell-type-specific transcription of target genes, enhancers themselves can be transcribed. We expanded the catalog of known human brain transcribed enhancers by an order of magnitude by generating and jointly analyzing large-scale cell-type-specific transcriptome and regulome data. Examination of the transcriptome in 1,382 brain samples in two independent cohorts identified robust expression of transcribed enhancers. We explored gene-enhancer coordination and found that enhancer-linked genes are strongly implicated in neuropsychiatric disease. We identified significant expression quantitative trait loci (eQTL) for 25,958 enhancers which mediate 6.8% of schizophrenia heritability, mostly independent from standard gene eQTL. Inclusion of enhancer eQTL in transcriptome-wide association studies enhanced functional interpretation of disease loci. Overall, our study characterizes the enhancer-gene regulome and genetic mechanisms in the human cortex in both healthy and disease states.

neuroscience↗

bioRxiv2021

Disruption of nuclear architecture as a cause of COVID-19 induced anosmia

Olfaction relies on a coordinated partnership between odorant flow and neuronal communication. Disruption in our ability to detect odors, or anosmia, has emerged as a hallmark symptom of infection with SARS-CoV-2, yet the mechanism behind this abrupt sensory deficit remains elusive. Here, using molecular evaluation of human olfactory epithelium (OE) from subjects succumbing to COVID-19 and a hamster model of SARS-CoV-2 infection, we discovered widespread downregulation of olfactory receptors (ORs) as well as key components of their signaling pathway. OR downregulation likely represents a non-cell autonomous effect, since SARS-CoV-2 detection in OSNs is extremely rare both in human and hamster OEs. A likely explanation for the reduction of OR transcription is the striking reorganization of nuclear architecture observed in the OSN lineage, which disrupts multi-chromosomal compartments regulating OR expression in humans and hamsters. Our experiments uncover a novel molecular mechanism by which a virus with a very selective tropism can elicit persistent transcriptional changes in cells that evade it, contributing to the severity of COVID-19.

neuroscience↗

bioRxiv2021

The three-dimensional landscape of chromatin accessibility in Alzheimer's disease

Much is still unknown about the neurobiology of Alzheimers disease (AD). To better understand AD, we generated 636 ATAC-seq libraries from cases and controls to construct detailed genomewide chromatin accessibility maps of neurons and non-neurons from two AD-affected brain regions, the entorhinal cortex and superior temporal gyrus. By analyzing a total of 19.6 billion read pairs, we expanded the known repertoire of regulatory sequences in the human brain. Multi-omic data integration associated global patterns of chromatin accessibility with gene expression and identified cell-specific enhancer-promoter interactions. Using inter-individual variation in chromatin accessibility, we define cis-regulatory domains capturing the 3D structure of the genome. Multifaceted analyses uncovered disease associated perturbations impacting chromatin accessibility, transcription factor regulatory networks and the 3D genome, and implicated transcriptional dysregulation in AD. Overall, we applied a systematic approach to understand the role of the 3D genome in AD and to illuminate novel disease biology that can advance diagnosis and therapy.

neuroscience↗

bioRxiv2021

The landscape of human brain immune response in patients with severe COVID-19

In coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, the relationship between brain tropism, neuroinflammation and host immune response has not been well characterized. We analyzed 68,557 single-nucleus transcriptomes from three brain regions (dorsolateral prefrontal cortex, medulla oblongata and choroid plexus) and identified an increased proportion of stromal cells and monocytes in the choroid plexus of COVID-19 patients. Differential gene expression, pseudo-temporal trajectory and gene regulatory network analyses revealed microglial transcriptome perturbations, mediating a range of biological processes, including cellular activation, mobility and phagocytosis. Quantification of viral spike S1 protein and SARS-CoV-2 transcripts did not support the notion of brain tropism. Overall, our findings suggest extensive neuroinflammation in patients with acute COVID-19. One Sentence SummarySingle-nucleus transcriptome analysis suggests extensive neuroinflammation in human brain tissue of patients with acute coronavirus disease 2019.

neuroscience↗

bioRxiv2020

PCQC: Selecting optimal principal components for identifying clusters with highly imbalanced class sizes in single-cell RNA-seq data

SummaryPrior to identifying clusters in single cell gene expression experiments, selecting the top principal components is a critical step for filtering out noise in the data set. Identifying these top principal components typically focuses on the total variance explained, and principal components that explain small clusters from rare populations will not necessarily capture a large percentage of variance in the data. We present a computationally efficient alternative for identifying the optimal principal components based on the tails of the distribution of variance explained for each observation. We then evaluate the efficacy of our approach in three different single cell RNA-sequencing data sets and find that our method matches, or outperforms, other selection criteria that are typically employed in the literature. Availability and implementationpcqc is written in Python and available at github.com/RoussosLab/pcqc

bioinformatics↗

bioRxiv2020

Sex differences in the human brain transcriptome of cases with schizophrenia

While schizophrenia differs between males and females in age of onset, symptomatology and the course of the disease, the molecular mechanisms underlying these differences remain uncharacterized. In order to address questions about the sex-specific effects of schizophrenia, we performed a large-scale transcriptome analysis of RNA-seq data from 437 controls and 341 cases from two distinct cohorts from the CommonMind Consortium. Analysis across the cohorts identifies a reproducible gene expression signature of schizophrenia that is highly concordant with previous work. Differential expression across sex is reproducible across cohorts and identifies X- and Y-linked genes, as well as those involved in dosage compensation. Intriguingly, the sex expression signature is also enriched for genes involved in neurexin family protein binding and synaptic organization. Differential expression analysis testing a sex-by-diagnosis interaction effect did not identify any genome-wide signature after multiple testing corrections. Gene coexpression network analysis was performed to reduce dimensionality and elucidate interactions among genes. We found enrichment of co-expression modules for sex-by-diagnosis differential expression signatures, which were highly reproducible across the two cohorts and involve a number of diverse pathways, including neural nucleus development, neuron projection morphogenesis, and regulation of neural precursor cell proliferation. Overall, our results indicate that the effect size of sex differences in schizophrenia gene expression signatures is small and underscore the challenge of identifying robust sex-by-diagnosis signatures, which will require future analyses in larger cohorts.

genomics↗

bioRxiv2020

Common variants contribute to intrinsic functional architecture of human brain

The human brain remains active in the absence of explicit tasks and forms networks of correlated activity. Resting-state functional magnetic resonance imaging (rsfMRI) measures brain activity at rest, which has been linked with both cognitive and clinical outcomes. The genetic variants influencing human brain function are largely unknown. Here we utilized rsfMRI from 44,190 individuals of multiple ancestries (37,339 in the UK Biobank) to discover and validate the common genetic variants influencing intrinsic brain activity. We identified hundreds of novel genetic loci associated with intrinsic functional signatures (P < 2.8 x 10-11), including associations to the central executive, default mode, and salience networks involved in the triple network model of psychopathology. A number of intrinsic brain activity associated loci colocalized with brain disorder GWAS (e.g., Alzheimers disease, Parkinsons disease, schizophrenia) and cognition, such as 19q13.32, 17q21.31, and 2p16.1. Particularly, we detected a colocalization between one (rs429358) of the two variants in the APOE {varepsilon}4 locus and function of the default mode, central executive, attention, and visual networks. Genetic correlation analysis demonstrated shared genetic influences between brain function and brain structure in the same regions. We also detected significant genetic correlations with 26 other complex traits, such as ADHD, major depressive disorder, schizophrenia, intelligence, education, sleep, subjective well-being, and neuroticism. Common variants associated with intrinsic brain activity were enriched within regulatory element in brain tissues.

genetics↗

bioRxiv2020

Integrative multi-omics analyses identify cell-type disease genes and regulatory networks across schizophrenia and Alzheimer's disease

Understanding cell-type-specific gene regulatory mechanisms from genetic variants to diseases remains challenging. To address this, we developed an open-source computational pipeline, scGRNom, to predict the cell-type disease genes and regulatory networks from multi-omics data, including cell-type chromatin interactions, epigenomics, and single-cell transcriptomics. With applications to Schizophrenia and Alzheimers Disease, our predicted cell-type regulatory networks link transcription factors and enhancers to disease genes for excitatory and inhibitory neurons, microglia, and oligodendrocytes. The enrichments of cell-type disease genes reveal cross-disease and disease-specific functions and pathways. Finally, machine learning analysis found that cell-type disease genes shared by diseases have improved clinical phenotype predictions.

neuroscience↗

bioRxiv2020

Transcriptional and epigenetic characterization of early striosomes identifies Foxf2 and Olig2 as factors required for development of striatal compartmentation and neuronal phenotypic differentiation

The basal ganglia, best known for processing information required for multiple aspects of movement, is also part of a network which regulates reward and cognition. The major output nucleus of the basal ganglia is the striatum, and its functions are dependent on neuronal compartmentation, including striosomes and matrix, which are selectively affected in disease. Striatal projection neurons are GABAergic medium spiny neurons (MSNs), all of which share basic molecular signatures but are subtyped by selective expression of receptors, neuropeptides, and other gene families. Neurogenesis of the striosome and matrix occurs in separate waves, but the factors regulating terminal neuronal differentiation following migration are largely unidentified. We performed RNA- and ATAC-seq on sorted murine striosome and matrix cells at postnatal day 3. Focusing on the striosomal compartment, we validated the localization and role of transcription factors and their regulator(s), previously not known to be associated with striatal development, including Irx1, Foxf2, Olig2 and Stat1/2. In addition, we validated the enhancer function of a striosome-specific open chromatin region located 15Kb downstream of the Olig2 gene. These data and data bases provide novel tools to dissect and manipulate the networks regulating MSN compartmentation and differentiation and thus provide new approaches to establishing MSN subtypes from human iPSCs for disease modeling and drug discovery.

neuroscience↗

bioRxiv2020

Identifying Nootropic Drug Targets via Large-Scale Cognitive GWAS and Transcriptomics

BackgroundCognitive traits demonstrate significant genetic correlations with many psychiatric disorders and other health-related traits. Many neuropsychiatric and neurodegenerative disorders are marked by cognitive deficits. Therefore, genome-wide association studies (GWAS) of general cognitive ability might suggest potential targets for nootropic drug repurposing. Our previous effort to identify "druggable genes" (i.e., GWAS-identified genes that produce proteins targeted by known small molecules) was modestly powered due to the small cognitive GWAS sample available at the time. Since then, two large cognitive GWAS meta-analyses have reported 148 and 205 genome-wide significant loci, respectively. Additionally, large-scale gene expression databases, derived from post-mortem human brain, have recently been made available for GWAS annotation. Here, we 1) reconcile results from these two cognitive GWAS meta-analyses to further enhance power for locus discovery; 2) employ several complementary transcriptomic methods to identify genes in these loci with variants that are credibly associated with cognition; and 3) further annotate the resulting genes to identify "druggable" targets. MethodsGWAS summary statistics were harmonized and jointly analysed using Multi-Trait Analysis of GWAS [MTAG], which is optimized for handling sample overlaps. Downstream gene identification was carried out using MAGMA, S-PrediXcan/S-TissueXcan Transcriptomic Wide Analysis, and eQTL mapping, as well as more recently developed methods that integrate GWAS and eQTL data via Summary-statistics Mendelian Randomization [SMR] and linkage methods [HEIDI], Available brain-specific eQTL databases included GTEXv7, BrainEAC, CommonMind, ROSMAP, and PsychENCODE. Intersecting credible genes were then annotated against multiple chemoinformatic databases [DGIdb, KI, and a published review on "druggability"]. ResultsUsing our meta-analytic data set (N = 373,617) we identified 241 independent cognition-associated loci (29 novel), and 76 genes were identified by 2 or more methods of gene identification. 26 genes were associated with general cognitive ability via SMR, 16 genes via STissueXcan/S-PrediXcan, 47 genes via eQTL mapping, and 68 genes via MAGMA pathway analysis. The use of the HEIDI test permitted the exclusion of candidate genes that may have been artifactually associated to cognition due to linkage, rather than direct causal or indirect pleiotropic effects. Actin and chromatin binding gene sets were identified as novel pathways that could be targeted via drug repurposing. Leveraging on our various transcriptome and pathway analyses, as well as available chemoinformatic databases, we identified 16 putative genes that may suggest drug targets with nootropic properties. DiscussionResults converged on several categories of significant drug targets, including serotonergic and glutamatergic genes, voltage-gated ion channel genes, carbonic anhydrase genes, and phosphodiesterase genes. The current results represent the first efforts to apply a multi-method approach to integrate gene expression and SNP level data to identify credible actionable genes for general cognitive ability.

genomics↗