Inference of cell type content from human brain transcriptomic datasets illuminates the effects of age, manner of death, dissection, and psychiatric diagnosis

Resource: A curated database of brain-related functional gene sets (Brain.GMT)

Transcriptional profiling has become a common tool for investigating the nervous system. During analysis, differential expression results are often compared to functional ontology databases, which contain curated gene sets representing well-studied pathways. This dependence can cause neuroscience studies to be interpreted in terms of functional pathways documented in better studied tissues (e.g., liver) and topics (e.g., cancer), and systematically emphasizes well-studied genes, leaving other findings in the obscurity of the brain "ignorome". To address this issue, we compiled a curated database of 918 gene sets related to nervous system function, tissue, and cell types ("Brain.GMT") that can be used within common analysis pipelines (GSEA, limma, edgeR) to interpret results from three species (rat, mouse, human). Brain.GMT includes brain-related gene sets curated from the Molecular Signatures Database (MSigDB) and extracted from public databases (GeneWeaver, Gemma, DropViz, BrainInABlender, HippoSeq) and published studies containing differential expression results. Although Brain.GMT is still undergoing development and currently only represents a fraction of available brain gene sets, "brain ignorome" genes are already better represented than in traditional Gene Ontology databases. Moreover, Brain.GMT substantially improves the quantity and quality of gene sets identified as enriched with differential expression in neuroscience studies, enhancing interpretation. •We compiled a curated database of 918 gene sets related to nervous system function, tissue, and cell types ("Brain.GMT").•Brain.GMT can be used within common analysis pipelines (GSEA, limma, edgeR) to interpret neuroscience transcriptional profiling results from three species (rat, mouse, human).•Although Brain.GMT is still undergoing development, it substantially improved the interpretation of differential expression results within our initial use cases.

Expression of WNT Signaling Genes in the Dorsolateral Prefrontal Cortex in Schizophrenia

Gene expression alterations in postmortem schizophrenia tissue are well-documented and are influenced by genetic, medication, and epigenetic factors. The Wingless/Integrated (WNT) signaling pathway, critical for cell growth and development, is involved in various cellular processes including neurodevelopment and synaptic plasticity. Despite its importance, WNT signaling remains understudied in schizophrenia, a disorder characterized by metabolic and bioenergetic defects in cortical regions. In this study, we examined the gene expression of 10 key WNT signaling pathway transcripts: IQGAP1, CTNNβ1, GSK3β, FOXO1, LRP6, MGEA5, TCF4, βTRC, PPP1Cβ, and DVL2 in the dorsolateral prefrontal cortex (DLPFC) using postmortem tissue from schizophrenia subjects (n = 20, 10 males, 10 females) compared to age, pH, and postmortem interval (PMI)-matched controls (n = 20, 10 males, 10 females). Employing the R-shiny application Kaleidoscope, we conducted in silico "lookup" studies from published transcriptomic datasets to examine cell- and region-level expression of these WNT genes. In addition, we investigated the impact of antipsychotics on the mRNA expression of the WNT genes of interest in rodent brain transcriptomic datasets. Our findings revealed no significant changes in region-level WNT transcript expression; however, analyses of previously published cell-level datasets indicated alterations in WNT transcript expression and antipsychotic-specific modulation of certain genes. These results suggest that WNT signaling transcripts may be variably expressed at the cellular level and influenced by antipsychotic treatment, providing novel insights into the role of WNT signaling in the pathophysiology of schizophrenia.

Resource: A Curated Database of Brain-Related Functional Gene Sets (Brain.GMT)

Transcriptional profiling has become a common tool for investigating the nervous system. During analysis, differential expression results are often compared to functional ontology databases, which contain curated gene sets representing well-studied pathways. This dependence can cause neuroscience studies to be interpreted in terms of functional pathways documented in better studied tissues (e.g., liver) and topics (e.g., cancer), and systematically emphasizes well-studied genes, leaving other findings in the obscurity of the brain "ignorome". To address this issue, we compiled a curated database of 918 gene sets related to nervous system function, tissue, and cell types ("Brain.GMT") that can be used within common analysis pipelines (GSEA, limma, edgeR) to interpret results from three species (rat, mouse, human). Brain.GMT includes brain-related gene sets curated from the Molecular Signatures Database (MSigDB) and extracted from public databases (GeneWeaver, Gemma, DropViz, BrainInABlender, HippoSeq) and published studies containing differential expression results. Although Brain.GMT is still undergoing development and currently only represents a fraction of available brain gene sets, "brain ignorome" genes are already better represented than in traditional Gene Ontology databases. Moreover, Brain.GMT substantially improves the quantity and quality of gene sets identified as enriched with differential expression in neuroscience studies, enhancing interpretation.

The role of glial cells in mental illness: a systematic review on astroglia and microglia as potential players in schizophrenia and its cognitive and emotional aspects

Schizophrenia is a complex and severe mental disorder that affects approximately 1% of the global population. It is characterized by a wide range of symptoms, including delusions, hallucinations, disorganized speech and behavior, and cognitive impairment. Recent research has suggested that the immune system dysregulation may play a significant role in the pathogenesis of schizophrenia, and glial cells, such as astroglia and microglia known to be involved in neuroinflammation and immune regulation, have emerged as potential players in this process. The aim of this systematic review is to summarize the glial hallmarks of schizophrenia, choosing as cellular candidate the astroglia and microglia, and focusing also on disease-associated psychological (cognitive and emotional) changes. We conducted a systematic review following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. We searched PubMed, Scopus, and Web of Science for articles that investigated the differences in astroglia and microglia in patients with schizophrenia, published in the last 5 years. The present systematic review indicates that changes in the density, morphology, and functioning of astroglia and microglia may be involved in the development of schizophrenia. The glial alterations may contribute to the pathogenesis of schizophrenia by dysregulating neurotransmission and immune responses, worsening cognitive capabilities. The complex interplay of astroglial and microglial activation, genetic/epigenetic variations, and cognitive assessments underscores the intricate relationship between biological mechanisms, symptomatology, and cognitive functioning in schizophrenia.

Cryopreservation of brain cell structure: a review

Cryopreservation, the preservation of tissues at subzero temperatures, is a mainstay of brain banking that allows for the storage of brain tissue without the use of chemical fixatives. This is particularly important for molecular studies that are incompatible with tissue fixation. However, brain tissue is vulnerable to various forms of damage during the cryopreservation process, in particular due to the phase transition of water from a liquid to a solid state with the formation of ice crystals, which can disrupt cellular morphology. There is a critical need to characterize the effects of cryopreservation on brain cell structure at the microscopic level. In this review, we conducted a comprehensive literature search, identifying 97 studies that yielded 146 distinct observations of the effects of cryopreservation on neurohistology. We classified the reviewed studies into three main categories: cryofixation, freezing, and cryopreservation with cryoprotectants. Cryofixation techniques enable vitrification and excellent ultrastructural preservation of thin tissue samples but are limited in terms of the depth of tissue that can be preserved without ice artifacts. Freezing methods, particularly when applied to brain slices, can achieve rapid cooling rates that result in minimal ice artifacts detectable by light microscopy. Cryoprotectant-based approaches have the potential to reduce ice damage and achieve vitrification. For thin tissue samples, immersion in cryoprotectants has been found to be effective for structural preservation. However, for larger samples or the entire brain, perfusion of cryoprotectants is necessary to perform rapid distribution, and this has a more limited evidence base. In conclusion, while current cryopreservation methods can provide sufficient quality for some downstream applications, there is a need for improved techniques that enable the cryopreservation of larger brain tissue samples while maintaining excellent structural preservation.

Challenges and opportunities to computationally deconvolve heterogeneous tissue with varying cell sizes using single-cell RNA-sequencing datasets

Deconvolution of cell mixtures in "bulk" transcriptomic samples from homogenate human tissue is important for understanding disease pathologies. However, several experimental and computational challenges impede transcriptomics-based deconvolution approaches using single-cell/nucleus RNA-seq reference atlases. Cells from the brain and blood have substantially different sizes, total mRNA, and transcriptional activities, and existing approaches may quantify total mRNA instead of cell type proportions. Further, standards are lacking for the use of cell reference atlases and integrative analyses of single-cell and spatial transcriptomics data. We discuss how to approach these key challenges with orthogonal "gold standard" datasets for evaluating deconvolution methods.

The Prolonged Terminal Phase of Human Life Induces Survival Response in the Skin Transcriptome

Human death marks the end of organismal life under conditions such that the components of the human body continue to be alive. Such postmortem cellular survival depends on the nature (Hardy scale of slow-fast death) of human death. Slow and expected death typically results from terminal illnesses and includes a prolonged terminal phase of life. As such organismal death process unfolds, do cells of the human body adapt for postmortem cellular survival? Organs with low energy cost-of-living, such as the skin, are better suited for postmortem cellular survival. In this work, the effect of different durations of terminal phase of human life on postmortem changes in cellular gene expression was investigated using RNA sequencing data of 701 human skin samples from the Genotype-Tissue Expression (GTEx) database. Longer terminal phase (slow-death) was associated with a more robust induction of survival pathways (PI3K-Akt signaling) in postmortem skin. Such cellular survival response was associated with the upregulation of embryonic developmental transcription factors such as FOXO1 , FOXO3 , ATF4 and CEBPD . Upregulation of PI3K-Akt signaling was independent of sex or duration of death-related tissue ischemia. Analysis of single nucleus RNA-seq of post-mortem skin tissue specifically identified the dermal fibroblast compartment to be most resilient as marked by adaptive induction of PI3K-Akt signaling. In addition, slow death also induced angiogenic pathways in the dermal endothelial cell compartment of postmortem human skin. In contrast, specific pathways supporting functional properties of the skin as an organ were downregulated following slow death. Such pathways included melanogenesis and those representing the skin extracellular matrix (collagen expression and metabolism). Efforts to understand the significance of death as a biological variable (DABV) in influencing the transcriptomic composition of surviving component tissues has far-reaching implications including rigorous interpretation of experimental data collected from the dead and mechanisms involved in transplant-tissue obtained from dead donors.

Challenges and opportunities to computationally deconvolve heterogeneous tissue with varying cell sizes using single cell RNA-sequencing datasets

Deconvolution of cell mixtures in "bulk" transcriptomic samples from homogenate human tissue is important for understanding the pathologies of diseases. However, several experimental and computational challenges remain in developing and implementing transcriptomics-based deconvolution approaches, especially those using a single cell/nuclei RNA-seq reference atlas, which are becoming rapidly available across many tissues. Notably, deconvolution algorithms are frequently developed using samples from tissues with similar cell sizes. However, brain tissue or immune cell populations have cell types with substantially different cell sizes, total mRNA expression, and transcriptional activity. When existing deconvolution approaches are applied to these tissues, these systematic differences in cell sizes and transcriptomic activity confound accurate cell proportion estimates and instead may quantify total mRNA content. Furthermore, there is a lack of standard reference atlases and computational approaches to facilitate integrative analyses, including not only bulk and single cell/nuclei RNA-seq data, but also new data modalities from spatial -omic or imaging approaches. New multi-assay datasets need to be collected with orthogonal data types generated from the same tissue block and the same individual, to serve as a "gold standard" for evaluating new and existing deconvolution methods. Below, we discuss these key challenges and how they can be addressed with the acquisition of new datasets and approaches to analysis.

Neurotransmission-related gene expression in the frontal pole is altered in subjects with bipolar disorder and schizophrenia

The frontal pole (Brodmann area 10, BA10) is the largest cytoarchitectonic region of the human cortex, performing complex integrative functions. BA10 undergoes intensive adolescent grey matter pruning prior to the age of onset for bipolar disorder (BP) and schizophrenia (SCHIZ), and its dysfunction is likely to underly aspects of their shared symptomology. In this study, we investigated the role of BA10 neurotransmission-related gene expression in BP and SCHIZ. We performed qPCR to measure the expression of 115 neurotransmission-related targets in control, BP, and SCHIZ postmortem samples (n = 72). We chose this method for its high sensitivity to detect low-level expression. We then strengthened our findings by performing a meta-analysis of publicly released BA10 microarray data (n = 101) and identified sources of convergence with our qPCR results. To improve interpretation, we leveraged the unusually large database of clinical metadata accompanying our samples to explore the relationship between BA10 gene expression, therapeutics, substances of abuse, and symptom profiles, and validated these findings with publicly available datasets. Using these convergent sources of evidence, we identified 20 neurotransmission-related genes that were differentially expressed in BP and SCHIZ in BA10. These results included a large diagnosis-related decrease in two important therapeutic targets with low levels of expression, HTR2B and DRD4, as well as other findings related to dopaminergic, GABAergic and astrocytic function. We also observed that therapeutics may produce a differential expression that opposes diagnosis effects. In contrast, substances of abuse showed similar effects on BA10 gene expression as BP and SCHIZ, potentially amplifying diagnosis-related dysregulation.

Simulated annealing aided genetic algorithm for gene selection from microarray data

In recent times, microarray gene expression datasets have gained significant popularity due to their usefulness to identify different types of cancer directly through bio-markers. These datasets possess a high gene-to-sample ratio and high dimensionality, with only a few genes functioning as bio-markers. Consequently, a significant amount of data is redundant, and it is essential to filter out important genes carefully. In this paper, we propose the Simulated Annealing aided Genetic Algorithm (SAGA), a meta-heuristic approach to identify informative genes from high-dimensional datasets. SAGA utilizes a two-way mutation-based Simulated Annealing (SA) as well as Genetic Algorithm (GA) to ensure a good trade-off between exploitation and exploration of the search space, respectively. The naive version of GA often gets stuck in a local optimum and depends on the initial population, leading to premature convergence. To address this, we have blended a clustering-based population generation with SA to distribute the initial population of GA over the entire feature space. To further enhance the performance, we reduce the initial search space by a score-based filter approach called the Mutually Informed Correlation Coefficient (MICC). The proposed method is evaluated on 6 microarray and 6 omics datasets. Comparison of SAGA with contemporary algorithms has shown that SAGA performs much better than its peers. Our code is available at https://github.com/shyammarjit/SAGA.

Genome-wide landscape of RNA-binding protein (RBP) networks as potential molecular regulators of psychiatric co-morbidities: a computational analysis

Background

Psychiatric disorders are a major burden on global health. These illnesses manifest as co-morbid conditions, further complicating the treatment. There is a limited understanding of the molecular and regulatory basis of psychiatric co-morbidities. The existing research in this regard has largely focused on epigenetic modulators, non-coding RNAs, and transcription factors. RNA-binding proteins (RBPs) functioning as multi-protein complexes are now known to be predominant controllers of multiple gene regulatory processes. However, their involvement in gene expression dysregulation in psychiatric co-morbidities is yet to be understood.

Results

Ten RBPs (QKI, ELAVL2, EIF2S1, SRSF3, IGF2BP2, EIF4B, SNRNP70, FMR1, DAZAP1, and MBNL1) were identified to be associated with psychiatric disorders such as schizophrenia, major depression, and bipolar disorders. Analysis of transcriptomic changes in response to individual depletion of these RBPs showed the potential influence of a large number of RBPs driving differential gene expression, suggesting functional cross-talk giving rise to multi-protein networks. Subsequent transcriptome analysis of post-mortem human brain samples from diseased and control individuals also suggested the involvement of ~ 100 RBPs influencing gene expression changes. These RBPs were found to regulate various processes including transcript splicing, mRNA transport, localization, stability, and translation. They were also found to form an extensive interactive network. Further, hnRNP, SRSF, and PCBP family RBPs, Matrin3, U2AF2, KHDRBS1, PTBP1, and also PABPN1 were found to be the hub proteins of the RBP network.

Conclusions

Extensive RBP networks involving a few hub proteins could result in transcriptome-wide dysregulation of post-transcriptional modifications, potentially driving multiple psychiatric disorders. Understanding the functional involvement of RBP networks in psychiatric disorders would provide insights into the molecular basis of psychiatric co-morbidities.

PTSD, major depression, and advanced transcriptomic age in brain tissue

BACKGROUND

Psychiatric disorders have been associated with advanced epigenetic age in DNA methylation, yet this relationship has not been studied in the brain transcriptome. We examined transcriptomic age using an RNA-based algorithm recently developed by Ren and Kuan ("RNAAgeCalc") and the associations between posttraumatic stress disorder (PTSD), major depressive disorder (MDD), and alcohol use disorder with age-adjusted RNA age ("RNA age residuals") in three brain regions: dorsolateral prefrontal cortex, ventromedial prefrontal cortex (vmPFC), and motor cortex.

METHODS

RNA sequencing was used to measure gene expression in postmortem brain tissue from the VA National PTSD Brain Bank (n = 94; 59% male).

RESULTS

Linear models revealed that diagnoses of PTSD and/or MDD were positively associated with RNA age residuals in vmPFC only (p-adj = 0.012). Three genes in the RNAAgeCalc algorithm (KCNJ16, HYAL2, and CEBPB) were also differentially expressed in association with PTSD/MDD in vmPFC (p-adj = 6.45E-05 to 0.02). Enrichment analysis revealed that inflammatory and immune-related pathways were overrepresented (p-adj < 0.05) among the 43 genes in RNAAgeCalc that were also at least nominally associated with PTSD/MDD in vmPFC relative to the 448 RNAAgeCalc genes. Endothelial and mural cells were negatively associated with RNA age residuals in vmPFC (both p-adj = 0.028) and with PTSD/MDD (both p-adj = 0.017).

CONCLUSIONS

Results highlight the importance of inflammation and immune system dysregulation in the link between psychopathology and accelerated cellular aging and raise the possibility that blood-brain barrier degradation may play an important role in stress-related accelerated brain aging.

Intestinal flora induces depression by mediating the dysregulation of cerebral cortex gene expression and regulating the metabolism of stroke patients

Post-stroke depression (PSD) is a common cerebrovascular complication characterized by complex pathogenesis and poor treatment effects. Here, we tested the influence of differentially expressed genes (DEGs), non-targeted metabolites, and intestinal microbes on the occurrence and development of PSD. We acquired gene expression profiles for stroke patients, depression patients, and healthy controls from the Gene Expression Omnibus database. After screening for DEGs using differential expression analysis, we identified common DEGs in stroke and depression patients that were considered to form the molecular basis of PSD. Functional enrichment analysis of DEGs also revealed that the majority of biological functions were closely related to metabolism, immunity, the nervous system, and microorganisms, and we also collected blood and stool samples from healthy controls, stroke patients, and PSD patients and performed 16S rDNA sequencing and untargeted metabolomics. After evaluating the quality of the sequencing data, we compared the diversity of the metabolites and intestinal flora within and between groups. Metabolic pathway enrichment analysis was used to identify metabolic pathways that were significantly involved in stroke and PSD, and a global metabolic network was constructed to explore the pathogenesis of PSD. Additionally, we constructed a global regulatory network based on 16S rDNA sequencing, non-targeted metabolomics, and transcriptomics to explore the pathogenesis of PSD through correlation analysis. Our results suggest that intestinal flora associates the dysregulation of cerebral cortex gene expression and could potentially promote the occurrence of depression by affecting the metabolism of stroke patients. Our findings may be helpful in identifying new targets for the prevention and treatment of PSD.

A novel multi-class classification model for schizophrenia, bipolar disorder and healthy controls using comprehensive transcriptomic data

Two common psychiatric disorders, schizophrenia (SCZ) and bipolar disorder (BP), confer lifelong disability and collectively affect 2% of the world population. Because the diagnosis of psychiatry is based only on symptoms, developing more effective methods for the diagnosis of psychiatric disorders is a major international public health priority. Furthermore, SCZ and BP overlap considerably in terms of symptoms and risk genes. Therefore, the clarity of the underlying etiology and pathology remains lacking for these two disorders. Although many studies have been conducted, a classification model with higher accuracy and consistency was found to still be necessary for accurate diagnoses of SCZ and BP. In this study, a comprehensive dataset was combined from five independent transcriptomic studies. This dataset comprised 120 patients with SCZ, 101 patients with BP, and 149 healthy subjects. The partial least squares discriminant analysis (PLS-DA) method was applied to identify the gene signature among multiple groups, and 341 differentially expressed genes (DEGs) were identified. Then, the disease relevance of these DEGs was systematically performed, including (α) the great disease relevance of the identified signature, (β) the hub genes of the protein-protein interaction network playing a key role in psychiatric disorders, and (γ) gene ontology terms and enriched pathways playing a key role in psychiatric disorders. Finally, a popular multi-class classifier, support vector machine (SVM), was applied to construct a novel multi-class classification model using the identified signature for SCZ and BP. Using the independent test sets, the classification capacity of this multi-class model was assessed, which showed this model had a strong classification ability.

Identification of potential blood biomarkers associated with suicide in major depressive disorder

Suicides have increased to over 48,000 deaths yearly in the United States. Major depressive disorder (MDD) is the most common diagnosis among suicides, and identifying those at the highest risk for suicide is a pressing challenge. The objective of this study is to identify changes in gene expression associated with suicide in brain and blood for the development of biomarkers for suicide. Blood and brain were available for 45 subjects (53 blood samples and 69 dorsolateral prefrontal cortex (DLPFC) samples in total). Samples were collected from MDD patients who died by suicide (MDD-S), MDDs who died by other means (MDD-NS) and non-psychiatric controls. We analyzed gene expression using RNA and the NanoString platform. In blood, we identified 14 genes which significantly differentiated MDD-S versus MDD-NS. The top six genes differentially expressed in blood were: PER3, MTPAP, SLC25A26, CD19, SOX9, and GAR1. Additionally, four genes showed significant changes in brain and blood between MDD-S and MDD-NS; SOX9 was decreased and PER3 was increased in MDD-S in both tissues, while CD19 and TERF1 were increased in blood but decreased in DLPFC. To our knowledge, this is the first study to analyze matched blood and brain samples in a well-defined population of MDDs demonstrating significant differences in gene expression associated with completed suicide. Our results strongly suggest that blood gene expression is highly informative to understand molecular changes in suicide. Developing a suicide biomarker signature in blood could help health care professionals to identify subjects at high risk for suicide.

Comprehensive evaluation of deconvolution methods for human brain gene expression

Transcriptome deconvolution aims to estimate the cellular composition of an RNA sample from its gene expression data, which in turn can be used to correct for composition differences across samples. The human brain is unique in its transcriptomic diversity, and comprises a complex mixture of cell-types, including transcriptionally similar subtypes of neurons. Here, we carry out a comprehensive evaluation of deconvolution methods for human brain transcriptome data, and assess the tissue-specificity of our key observations by comparison with human pancreas and heart. We evaluate eight transcriptome deconvolution approaches and nine cell-type signatures, testing the accuracy of deconvolution using in silico mixtures of single-cell RNA-seq data, RNA mixtures, as well as nearly 2000 human brain samples. Our results identify the main factors that drive deconvolution accuracy for brain data, and highlight the importance of biological factors influencing cell-type signatures, such as brain region and in vitro cell culturing.

Transcriptome deconvolution aims to estimate cellular composition based on gene expression data. Here the authors evaluate deconvolution methods for human brain transcriptome and conclude that partial deconvolution algorithms work best, but that appropriate cell-type signatures are also important.

A review and meta-analysis of gene expression profiles in suicide

Suicide claims over 800,000 deaths worldwide, making it a serious public health problem. The etiopathophysiology of suicide remains unclear and is highly complex, and postmortem gene expression studies can offer insights into the molecular biological mechanism underlying suicide. In the current study, we conducted a meta-analysis of postmortem brain gene expression in relation to suicide. We identified five gene expression datasets for postmortem orbitofrontal, prefrontal, or dorsolateral prefrontal cortical brain regions from the Gene Expression Omnibus repository. After quality control, the total sample size was 380 (141 suicide deaths and 239 deaths from other causes). We performed the analyses using two meta-analytic approaches. We further performed pathway and cell-set enrichment analyses. We found reduced expression of the KCNJ2 (Potassium Inwardly Rectifying Channel Subfamily J Member 2), A2M (Alpha-2-Macroglobulin), AGT (Angiotensinogen), PMP2 (Peripheral Myelin Protein 2), and VEZF1 (Vascular Endothelial Zinc Finger 1) genes (FDR p<0.05). Our findings support the involvement of astrocytes, stress response, immune system, and microglia in suicide. These findings will require further validation in additional large datasets.

Deficient Autophagy in Microglia Aggravates Repeated Social Defeat Stress-Induced Social Avoidance

Major depressive disorder (MDD) is associated with repeated exposure to environmental stress. Autophagy is activated under various stress conditions that are associated with several diseases in the brain. This study was aimed at elucidating the autophagy signaling changes in the prefrontal cortex (PFC) under repeated social defeat (RSD) to investigate the involvement of microglial autophagy in RSD-induced behavioral changes. We found that RSD stress, an animal model of MDD, significantly induced initial autophagic signals followed by increased transcription of autophagy-related genes (Atg6, Atg7, and Atg12) in the PFC. Similarly, significantly increased transcripts of ATGs (Atg6, Atg7, Atg12, and Atg5) were confirmed in the postmortem PFC of patients with MDD. The protein levels of the prefrontal cortical LC3B were significantly increased, whereas p62 was significantly decreased in the resilient but not in susceptible mice and patients with MDD. This indicates that enhanced autophagic flux may alleviate stress-induced depression. Furthermore, we identified that FKBP5, an early-stage autophagy regulator, was significantly increased in the PFC of resilient mice at the transcript and protein levels. In addition, the resilient mice exhibited enhanced autophagic flux in the prefrontal cortical microglia, and the autophagic deficiency in microglia aggravated RSD-induced social avoidance, indicating that microglial autophagy involves stress-induced behavioral changes.

Integrated co-expression network analysis uncovers novel tissue-specific genes in major depressive disorder and bipolar disorder

Tissue-specific gene expression has been found to be associated with multiple complex diseases including cancer, metabolic disease, aging, etc. However, few studies of brain-tissue-specific gene expression patterns have been reported, especially in psychiatric disorders. In this study, we performed joint analysis on large-scale transcriptome multi-tissue data to investigate tissue-specific expression patterns in major depressive disorder (MDD) and bipolar disorder (BP). We established the strategies of identifying tissues-specific modules, annotated pathways for elucidating biological functions of tissues, and tissue-specific genes based on weighted gene co-expression network analysis (WGCNA) and robust rank aggregation (RRA) with transcriptional profiling data from different human tissues and genome wide association study (GWAS) data, which have been expanded into overlapping tissue-specific modules and genes sharing with MDD and BP. Nine tissue-specific modules were identified and distributed across the four tissues in the MDD and six modules in the BP. In general, the annotated biological functions of differentially expressed genes (DEGs) in blood were mainly involved in MDD and BP progression through immune response, while those in the brain were in neuron and neuroendocrine response. Tissue-specific genes of the prefrontal cortex (PFC) in MDD-, such as IGFBP2 and HTR1A, were involved in disease-related functions, such as response to glucocorticoid, taste transduction, and tissue-specific genes of PFC in BP-, such as CHRM5 and LTB4R2, were involved in neuroactive ligand-receptor interaction. We also found PFC tissue-specific genes including SST and CRHBP were shared in MDD-BP, SST was enriched in neuroactive ligand-receptor interaction, and CRHBP shown was related to the regulation of hormone secretion and hormone transport.

Microglial glutaminase 1 deficiency mitigates neuroinflammation associated depression

Glutaminase 1 (GLS1) has recently been reported to be expressed in microglia and plays a crucial role in neuroinflamation. Significantly increased level of GLS1 mRNA expression together with neuroinflammation pathway were observed in postmortem prefrontal cortex from depressed patients. To find out the function of microglial GLS1 in depression and neuroinflammation, we generated transgenic mice (GLS1 cKO), postnatally losing GLS1 in microglia, to detect changes in the lipopolysaccharide (LPS)-induced depression model. LPS-induced anxiety/depression-like behavior was attenuated in GLS1 cKO mice, paralleled by a significant reduction in pro-inflammatory cytokines and an abnormal microglia morphological phenotype in the prefrontal cortex. Reduced neuroinflammation by GLS1 deficient microglia was a result of less reactive astrocytes, as GLS1 deficiency enhanced miR-666-3p and miR-7115-3p levels in extracellular vesicles released from microglia, thus suppressing astrocyte activation via inhibiting Serpina3n expression. Together, our data reveal a novel mechanism of GLS1 in neuroinflammation and targeting GLS1 in microglia may be a novel strategy to alleviate neuroinflammation-related depression and other disease.

Classification for psychiatric disorders including schizophrenia, bipolar disorder, and major depressive disorder using machine learning

Schizophrenia (SCZ), bipolar disorder (BP), and major depressive disorder (MDD) are the most common psychiatric disorders. Because there were lots of overlaps among these disorders from genetic epidemiology and molecular genetics, it is hard to realize the diagnoses of these psychiatric disorders. Currently, plenty of studies have been conducted for contributing to the diagnoses of these diseases. However, constructing a classification model with superior performance for differentiating SCZ, BP, and MDD samples is still a great challenge. In this study, the transcriptomic data was applied for discovering key genes and constructing a classification model. In this dataset, there were 268 samples including four groups (67 SCZ patients, 40 BP patients, 57 MDD patients, and 104 healthy controls), which were applied for constructing a classification model. First, 269 probes of differentially expressed genes (DEGs) among four sample groups were identified by the feature selection method. Second, these DEGs were validated by the literature review including disease relevance with the psychiatric disorders of these DEGs, the hub genes in the PPI (protein–protein interaction) network, and GO (gene ontology) terms and pathways. Third, a classification model was constructed using the identified DEGs by machine learning method to classify different groups. The ROC (receiver operator characteristic) curve and AUC (area under the curve) value were used to assess the classification capacity of the model. In summary, this classification model might provide clues for the diagnoses of these psychiatric disorders.

Allele-specific DNA methylation maps in monozygotic twins discordant for psychiatric disorders reveal that disease-associated switching at the EIPR1 regulatory loci modulates neural function

The non-Mendelian features of phenotypic variations within monozygotic twins are likely complicated by environmental modifiers of genetic effects that have yet to be elucidated. Here, we performed methylome and genome analyses of blood DNA from psychiatric disorder-discordant monozygotic twins to study how allele-specific methylation (ASM) mediates phenotypic variations. We identified that thousands of genetic variants with ASM imbalances exhibit phenotypic variation-associated switching at regulatory loci. These ASMs have plausible causal associations with psychiatric disorders through effects on interactions between transcription factors, DNA methylations, and other epigenomic markers and then contribute to dysregulated gene expression, which eventually increases disease susceptibility. Moreover, we also experimentally validated the model that the rs4854158 alternative C allele at an ASM switching regulatory locus of EIPR1 encoding endosome-associated recycling protein-interacting protein 1, is associated with demethylation and higher RNA expression and shows lower TF binding affinities in unaffected controls. An epigenetic ASM switching induces C allele hypermethylation and then recruits repressive Polycomb repressive complex 2 (PRC2), reinforces trimethylation of lysine 27 on histone 3 and inhibits its transcriptional activity, thus leading to downregulation of EIPR1 in schizophrenia. Moreover, disruption of rs4854158 induces gain of EIPR1 function and promotes neural development and vesicle trafficking. Our study provides a powerful framework for identifying regulatory risk variants and contributes to our understanding of the interplay between genetic and epigenetic variants in mediating psychiatric disorder susceptibility.

Maternal whole blood mRNA signatures identify women at risk of early preeclampsia: a longitudinal study

PURPOSE

To determine whether previously established mRNA signatures are predictive of early preeclampsia when evaluated by maternal cellular transcriptome analysis in samples collected before clinical manifestation.

MATERIALS AND METHODS

We profiled gene expression at exon-level resolution in whole blood samples collected longitudinally from 49 women with normal pregnancy (controls) and 13 with early preeclampsia (delivery <34 weeks of gestation). After preprocessing and removal of gestational age-related trends in gene expression, data were converted into Z-scores based on the mean and standard deviation among controls for six gestational-age intervals. The average Z-scores of mRNAs in each previously established signature considered herein were compared between cases and controls at 9-11, 11-17, 17-22, 22-28, 28-32, and 32-34 weeks of gestation.Results: (1) Average expression of the 16-gene untargeted cellular mRNA signature was higher in women diagnosed with early preeclampsia at 32-34 weeks of gestation, yet more importantly, also prior to diagnosis at 28-32 weeks and 22-28 weeks of gestation, compared to controls (all, p < .05). (2) A combination of four genes from this signature, including a long non-protein coding RNA [H19 imprinted maternally expressed transcript (H19)], fibronectin 1 (FN1), tubulin beta-6 class V (TUBB6), and formyl peptide receptor 3 (FPR3) had a sensitivity of 0.85 (0.55-0.98) and a specificity of 0.92 (0.8-0.98) for prediction of early preeclampsia at 22-28 weeks of gestation. (3) H19, FN1, and TUBB6 were increased in women with early preeclampsia as early as 11-17 weeks of gestation (all, p < .05). (4) After diagnosis at 32-34 weeks, but also prior to diagnosis at 11-17 weeks, women destined to have early preeclampsia showed a coordinated increase in whole blood expression of several single-cell placental signatures, including the 20-gene signature of extravillous trophoblast (all, p < .05). (5) A combination of three mRNAs from the extravillous trophoblast signature (MMP11, SLC6A2, and IL18BP) predicted early preeclampsia at 11-17 weeks of gestation with a sensitivity of 0.83 (0.52-0.98) and specificity of 0.94 (0.79-0.99).

CONCLUSIONS

Circulating early transcriptomic markers for preeclampsia can be found either by untargeted profiling of the cellular transcriptome or by focusing on placental cell-specific mRNAs. The untargeted cellular mRNA signature was consistently increased in early preeclampsia after 22 weeks of gestation, and individual mRNAs of this signature were significantly increased as early as 11-17 weeks of gestation. Several single-cell placental signatures predicted future development of the disease at 11-17 weeks and were also increased in women already diagnosed at 32-34 weeks of gestation.

Transcriptomics of the depressed and PTSD brain

Stress is the response of an organism to demands for change, yet excessive or chronic stress contributes to nearly all psychiatric disorders. The advent of high-throughput transcriptomic methods such as single cell RNA sequencing poses new opportunities to understand the neurobiology of stress, yet substantial barriers to understanding stress remain. Stress adaptation is an organismal survival mechanism conserved across all organisms, yet there is an infinity of potential stressful experiences. Unraveling shared and separate transcriptional programs for adapting to stressful experience remains a challenge, despite methodological and analytic advances. Here we review the state of the field focusing on the technologies used to study the transcriptome for the stress neurobiologist, and also attempt to identify central questions about the heterogeneity of stress for those applying transcriptomic approaches. We further explore how postmortem transcriptome studies aided by preclinical animal models are converging on common molecular pathways for adaptation to aversive experience. Finally, we discuss approaches to integrate large genomic datasets with human neuroimaging and other datasets.

Gene expression in the dorsolateral and ventromedial prefrontal cortices implicates immune-related gene networks in PTSD

Studies evaluating neuroimaging, genetically predicted gene expression, and pre-clinical genetic models of PTSD, have identified PTSD-related abnormalities in the prefrontal cortex (PFC) of the brain, particularly in dorsolateral and ventromedial PFC (dlPFC and vmPFC). In this study, RNA sequencing was used to examine gene expression in the dlPFC and vmPFC using tissue from the VA National PTSD Brain Bank in donors with histories of PTSD with or without depression (dlPFC n = 38, vmPFC n = 35), depression cases without PTSD (n = 32), and psychopathology-free controls (dlPFC n = 24, vmPFC n = 20). Analyses compared PTSD cases to controls. Follow-up analyses contrasted depression cases to controls. Twenty-one genes were differentially expressed in PTSD after strict multiple testing correction. PTSD-associated genes with roles in learning and memory (FOS, NR4A1), immune regulation (CFH, KPNA1) and myelination (MBP, MOBP, ERMN) were identified. PTSD-associated genes partially overlapped depression-associated genes. Co-expression network analyses identified PTSD-associated networks enriched for immune-related genes across the two brain regions. However, the immune-related genes and association patterns were distinct. The immune gene IL1B was significantly associated with PTSD in candidate-gene analysis and was an upstream regulator of PTSD-associated genes in both regions. There was evidence of replication of dlPFC associations in an independent cohort from a recent study, and a strong correlation between the dlPFC PTSD effect sizes for significant genes in the two studies (r = 0.66, p < 2.2 × 10⁻¹⁶). In conclusion, this study identified several novel PTSD-associated genes and brain region specific PTSD-associated immune-related networks.

Physiological and Pathological Ageing of Astrocytes in the Human Brain

Ageing is the greatest risk factor for dementia, although physiological ageing by itself does not lead to cognitive decline. In addition to ageing, APOE ε4 is genetically the strongest risk factor for Alzheimer's disease and is highly expressed in astrocytes. There are indications that human astrocytes change with age and upon expression of APOE4. As these glial cells maintain water and ion homeostasis in the brain and regulate neuronal transmission, it is likely that age- and APOE4-related changes in astrocytes have a major impact on brain functioning and play a role in age-related diseases. In this review, we will discuss the molecular and morphological changes of human astrocytes in ageing and the contribution of APOE4. We conclude this review with a discussion on technical issues, innovations, and future perspectives on how to gain more knowledge on astrocytes in the human ageing brain.

Gene expression correlates of advanced epigenetic age and psychopathology in postmortem cortical tissue

Psychiatric stress has been associated with accelerated epigenetic aging (i.e., when estimates of cellular age based on DNA methylation exceed chronological age) in both blood and brain tissue. Little is known about the downstream biological effects of accelerated epigenetic age on gene expression. In this study we examined associations between DNA methylation-derived estimates of cellular age that range from decelerated to accelerated relative to chronological age (“DNAm age residuals”) and transcriptome-wide gene expression. This was examined using tissue from three post-mortem cortical regions (ventromedial and dorsolateral prefrontal cortex and motor cortex, n = 97) from the VA National PTSD Brain Bank. In addition, we examined how posttraumatic stress disorder (PTSD) and alcohol-use disorders (AUD) moderated the association between DNAm age residuals and gene expression. Transcriptome-wide results across brain regions, psychiatric diagnoses, and cohorts (full sample and male and female subsets) revealed experiment-wide differential expression of 11 genes in association with PTSD or AUD in interaction with DNAm age residuals. This included the inflammation-related genes IL1B, RCOR2, and GCNT1. Candidate gene class analyses and gene network enrichment analyses further supported differential expression of inflammation/immune gene networks as well as glucocorticoid, circadian, and oxidative stress-related genes. Gene co-expression network modules suggested enrichment of myelination related processes and oligodendrocyte enrichment in association with DNAm age residuals in the presence of psychopathology. Collectively, results suggest that psychiatric stress accentuates the association between advanced epigenetic age and expression of inflammation genes in the brain. This highlights the role of inflammatory processes in the pathophysiology of accelerated cellular aging and suggests that inflammatory pathways may link accelerated cellular aging to premature disease onset and neurodegeneration, particularly in stressed populations. This suggests that anti-inflammatory interventions may be an important direction to pursue in evaluating ways to prevent or delay cellular aging and increase resilience to diseases of aging.

Pairwise interactions in gene expression determine a hierarchical transcriptional profile in the human brain

The orchestrated expression of thousands of genes gives rise to the complexity of the human brain. However, the structures governing these myriad gene-gene interactions remain unclear. By analyzing transcription data from more than 2000 sites in six human brains, we found that pairwise interactions between genes, without considering any higher-order interactions, are sufficient to predict the transcriptional pattern of the genome for individual brain regions and the transcriptional profile of the entire brain consisting of more than 200 areas. These findings suggest a quadratic complexity of transcriptional patterns in the human brain, which is much simpler than expected. In addition, using a pairwise interaction model, we revealed that the strength of gene-gene interactions in the human brain gives rise to the nearly maximal number of transcriptional clusters, which may account for the functional and structural richness of the brain.

A likelihood-based deconvolution of bulk gene expression data using single-cell references

Direct comparison of bulk gene expression profiles is complicated by distinct cell type mixtures in each sample which obscure whether observed differences are actually due to changes in expression levels themselves or simply due to differing cell type compositions. Single-cell technology has made it possible to measure gene expression in individual cells, achieving higher resolution at the expense of increased noise. If carefully incorporated, such single-cell data can be used to deconvolve bulk samples to yield accurate estimates of the true cell type proportions, thus enabling one to disentangle the effects of differential expression and cell type mixtures. Here, we propose a generative model and a likelihood-based inference method that uses asymptotic statistical theory and a novel optimization procedure to perform deconvolution of bulk RNA-seq data to produce accurate cell type proportion estimates. We demonstrate the effectiveness of our method, called RNA-Sieve, across a diverse array of scenarios involving real data and discuss extensions made uniquely possible by our probabilistic framework, including a demonstration of well-calibrated confidence intervals.

Sex-dependent monoamine oxidase isoforms expression patterns during human brain ageing

Human behavior is influenced by both genetic and environmental factors. Monoamine oxidase A (MAOA) is among the most investigated genetic determinants of violent behaviors, while the monoamine oxidase B (MAOB) is explored in Parkinson's disease. We collected twenty-four post-mortem brain tissue datasets of 3871 and 1820 non-demented males and females, respectively, who died from causes not attributable to neurodegenerative diseases. The gene expressions of MAOA and MAOB (MAO genes) were analyzed in these subjects, who were further stratified according to age into eleven groups ranging from late Infancy (5-9 months) to centenarians (>100 years). MAO genes were differently expressed in brains during the entire life span. In particular, maximal and minimal expression levels were found in early life and around the teen years. Females tended to have higher MAO gene levels throughout their lives than those found in age-matched males, even when expressions were separately measured in different brain regions. We demonstrated the existence of age- and sex- related variations in the MAO transcript levels in defined brain regions. More in-depth protein studies are needed to confirm our preliminary results obtained only on messenger RNAs in order to establish the role played by MAO genes in human development.

Downregulation by CNNM2 of ATP5MD expression in the 10q24.32 schizophrenia-associated locus involved in impaired ATP production and neurodevelopment

Genome-wide association studies (GWAS) have accelerated the discovery of numerous genetic variants associated with schizophrenia. However, most risk variants show a small effect size (odds ratio (OR) <1.2), suggesting that more functional risk variants remain to be identified. Here, we employed region-based multi-marker analysis of genomic annotation (MAGMA) to identify additional risk loci containing variants with large OR value from Psychiatry Genomics Consortium (PGC2) schizophrenia GWAS data and then employed summary-data-based mendelian randomization (SMR) to prioritize schizophrenia susceptibility genes. The top-ranked susceptibility gene ATP5MD, encoding an ATP synthase membrane subunit, is observed to be downregulated in schizophrenia by the risk allele of CNNM2-rs1926032 in the schizophrenia-associated 10q24.32 locus. The Atp5md knockout (KO) in mice was associated with abnormal startle reflex and gait, and ATP5MD knockdown (KD) in human induced pluripotent stem cell-derived neurons disrupted the neural development and mitochondrial respiration and ATP production. Moreover, CNNM2-rs1926032 KO could induce downregulation of ATP5MD expression and disruptions of mitochondrial respiration and ATP production. This study constitutes an important mechanistic component that links schizophrenia-associated CNNM2 regions to disruption in energy adenosine system modulation and neuronal function by long-distance chromatin domain downregulation of ATP5MD. This pathogenic mechanism provides therapeutic implications for schizophrenia.

CLEP: A Hybrid Data- and Knowledge- Driven Framework for Generating Patient Representations

Summary

As machine learning and artificial intelligence increasingly attain a larger number of applications in the biomedical domain, at their core, their utility depends on the data used to train them. Due to the complexity and high dimensionality of biomedical data, there is a need for approaches that combine prior knowledge around known biological interactions with patient data. Here, we present CLinical Embedding of Patients (CLEP), a novel approach that generates new patient representations by leveraging both prior knowledge and patient-level data. First, given a patient-level dataset and a knowledge graph containing relations across features that can be mapped to the dataset, CLEP incorporates patients into the knowledge graph as new nodes connected to their most characteristic features. Next, CLEP employs knowledge graph embedding models to generate new patient representations that can ultimately be used for a variety of downstream tasks, ranging from clustering to classification. We demonstrate how using new patient representations generated by CLEP significantly improves performance in classifying between patients and healthy controls for a variety of machine learning models, as compared to the use of the original transcriptomics data. Furthermore, we also show how incorporating patients into a knowledge graph can foster the interpretation and identification of biological features characteristic of a specific disease or patient subgroup. Finally, we released CLEP as an open source Python package together with examples and documentation.

Availability and implementation

CLEP is available to the bioinformatics community as an open source Python package at https://github.com/hybrid-kg/clep under the Apache 2.0 License.

Supplementary information

Supplementary data are available at Bioinformatics online.

A meta-study on transcription factor networks in the suicidal brain

There is evidence supporting the presence of brain gene expression differences between suicides and non-suicides. Such differences have been implicated in suicide pathophysiology. However, regulatory factors underlying these gene expression differences have not been fully understood. Therefore, the identification of differences in regulatory mechanisms, i.e., transcriptional factors between suicides and non-suicides is crucial for the understanding of suicide neurobiology. In this study, we conducted a transcription factor network meta-study with freely available data from the prefrontal cortex of suicides and non-suicides with different mental disorders, including major depression disorder, bipolar disorder and schizophrenia, as well as healthy controls. Disorder-specific characteristics of suicides and non-suicides transcription factor networks were detected, i.e., the presence of immune response genes in both suicides and non-suicides with major depression disorder networks. Also, we found the presence of ESR1, which has been implicated to give resilience to social stress, in the non-suicides network but not in the suicides with major depression network. Suicides and non-suicides with bipolar disorder shared only three genes in common: FOS, CRY1 and PER2. In addition, we found a higher number of genes involved in immune response in the non-suicides with bipolar disorder compared to the suicides with bipolar disorder network. The suicides and non-suicides with schizophrenia networks exhibited clear differences, including the presence of circadian cycle genes in the suicides with schizophrenia network and their absence in the non-suicides with schizophrenia network. The results of this study provide insight on the regulatory mechanisms underpinning transcriptional changes in the suicidal brain.

Agonal Factors Distort Gene-Expression Patterns in Human Postmortem Brains

Agonal factors, the conditions that occur just prior to death, can impact the molecular quality of postmortem brains, influencing gene expression results. Our study used gene expression data of 262 samples from ROSMAP with the detailed terminal state recorded for each donor, such as fever, infection, and unconsciousness. Fever and infection were the primary contributors to brain gene expression changes, brain cell-type-specific gene expression, and cell proportion changes. Furthermore, we also found that previous studies of gene expression in postmortem brains were confounded by agonal factors. Therefore, correction for agonal factors is important in the step of data preprocessing. Our analyses revealed fever and infection contributing to gene expression changes in postmortem brains and emphasized the necessity of study designs that document and account for agonal factors.

Selective time-dependent changes in activity and cell-specific gene expression in human postmortem brain

As a means to understand human neuropsychiatric disorders from human brain samples, we compared the transcription patterns and histological features of postmortem brain to fresh human neocortex isolated immediately following surgical removal. Compared to a number of neuropsychiatric disease-associated postmortem transcriptomes, the fresh human brain transcriptome had an entirely unique transcriptional pattern. To understand this difference, we measured genome-wide transcription as a function of time after fresh tissue removal to mimic the postmortem interval. Within a few hours, a selective reduction in the number of neuronal activity-dependent transcripts occurred with relative preservation of housekeeping genes commonly used as a reference for RNA normalization. Gene clustering indicated a rapid reduction in neuronal gene expression with a reciprocal time-dependent increase in astroglial and microglial gene expression that continued to increase for at least 24 h after tissue resection. Predicted transcriptional changes were confirmed histologically on the same tissue demonstrating that while neurons were degenerating, glial cells underwent an outgrowth of their processes. The rapid loss of neuronal genes and reciprocal expression of glial genes highlights highly dynamic transcriptional and cellular changes that occur during the postmortem interval. Understanding these time-dependent changes in gene expression in post mortem brain samples is critical for the interpretation of research studies on human brain disorders.

Genetic Liability for Internalizing Versus Externalizing Behavior Manifests in the Developing and Adult Hippocampus: Insight From a Meta-analysis of Transcriptional Profiling Studies in a Selectively Bred Rat Model

BACKGROUND

For more than 16 years, we have selectively bred rats for either high or low levels of exploratory activity within a novel environment. These bred high-responder (bHR) and bred low-responder (bLR) rats model temperamental extremes, exhibiting large differences in internalizing and externalizing behaviors relevant to mood and substance use disorders.

METHODS

We characterized persistent differences in gene expression related to bHR/bLR phenotype across development and adulthood in the hippocampus, a region critical for emotional regulation, by meta-analyzing 8 transcriptional profiling datasets (microarray and RNA sequencing) spanning 43 generations of selective breeding (postnatal day 7: n = 22; postnatal day 14: n = 49; postnatal day 21: n = 21; adult: n = 46; all male). We cross-referenced expression differences with exome sequencing within our colony to pinpoint candidates likely to mediate the effect of selective breeding on behavioral phenotype. The results were compared with hippocampal profiling from other bred rat models.

RESULTS

Genetic and transcriptional profiling results converged to implicate multiple candidate genes, including two previously associated with metabolism and mood: Trhr and Ucp2. Results also highlighted bHR/bLR functional differences in the hippocampus, including a network essential for neurodevelopmental programming, proliferation, and differentiation, centering on Bmp4 and Mki67. Finally, we observed differential expression related to microglial activation, which is important for synaptic pruning, including 2 genes within implicated chromosomal regions: C1qa and Mfge8.

CONCLUSIONS

These candidate genes and functional pathways may direct bHR/bLR rats along divergent developmental trajectories and promote a widely different reactivity to the environment.

Unraveling Targetable Systemic and Cell-Type-Specific Molecular Phenotypes of Alzheimer's and Parkinson's Brains With Digital Cytometry

Alzheimer's disease (AD) and Parkinson's disease (PD) are the two most common neurodegenerative disorders worldwide, with age being their major risk factor. The increasing worldwide life expectancy, together with the scarcity of available treatment choices, makes it thus pressing to find the molecular basis of AD and PD so that the causing mechanisms can be targeted. To study these mechanisms, gene expression profiles have been compared between diseased and control brain tissues. However, this approach is limited by mRNA expression profiles derived for brain tissues highly reflecting their degeneration in cellular composition but not necessarily disease-related molecular states. We therefore propose to account for cell type composition when comparing transcriptomes of healthy and diseased brain samples, so that the loss of neurons can be decoupled from pathology-associated molecular effects. This approach allowed us to identify genes and pathways putatively altered systemically and in a cell-type-dependent manner in AD and PD brains. Moreover, using chemical perturbagen data, we computationally identified candidate small molecules for specifically targeting the profiled AD/PD-associated molecular alterations. Our approach therefore not only brings new insights into the disease-specific and common molecular etiologies of AD and PD but also, in these realms, foster the discovery of more specific targets for functional and therapeutic exploration.

Robust partial reference-free cell composition estimation from tissue expression

MOTIVATION

In the analysis of high-throughput omics data from tissue samples, estimating and accounting for cell composition have been recognized as important steps. High cost, intensive labor requirements and technical limitations hinder the cell composition quantification using cell-sorting or single-cell technologies. Computational methods for cell composition estimation are available, but they are either limited by the availability of a reference panel or suffer from low accuracy.

RESULTS

We introduce TOols for the Analysis of heterogeneouS Tissues TOAST/-P and TOAST/+P, two partial reference-free algorithms for estimating cell composition of heterogeneous tissues based on their gene expression profiles. TOAST/-P and TOAST/+P incorporate additional biological information, including cell-type-specific markers and prior knowledge of compositions, in the estimation procedure. Extensive simulation studies and real data analyses demonstrate that the proposed methods provide more accurate and robust cell composition estimation than existing methods.

AVAILABILITY AND IMPLEMENTATION

The proposed methods TOAST/-P and TOAST/+P are implemented as part of the R/Bioconductor package TOAST at https://bioconductor.org/packages/TOAST.

CONTACT

ziyi.li@emory.edu or hao.wu@emory.edu.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Spatial cell type composition in normal and Alzheimers human brains is revealed using integrated mouse and human single cell RNA sequencing

Single-cell RNA sequencing (scRNA-seq) resolves heterogenous cell populations in tissues and helps to reveal single-cell level function and dynamics. In neuroscience, the rarity of brain tissue is the bottleneck for such study. Evidence shows that, mouse and human share similar cell type gene markers. We hypothesized that the scRNA-seq data of mouse brain tissue can be used to complete human data to infer cell type composition in human samples. Here, we supplement cell type information of human scRNA-seq data, with mouse. The resulted data were used to infer the spatial cellular composition of 3702 human brain samples from Allen Human Brain Atlas. We then mapped the cell types back to corresponding brain regions. Most cell types were localized to the correct regions. We also compare the mapping results to those derived from neuronal nuclei locations. They were consistent after accounting for changes in neural connectivity between regions. Furthermore, we applied this approach on Alzheimer's brain data and successfully captured cell pattern changes in AD brains. We believe this integrative approach can solve the sample rarity issue in the neuroscience.

Postsynaptic damage and microglial activation in AD patients could be linked CXCR4/CXCL12 expression levels

Alzheimer's disease (AD) is one of the most common forms of dementia with still unknown pathogenesis. Several cytokines and chemokines are involved in the pathogenesis of AD. Among the chemokines, the CXCR4/CXCL12 complex has been shown to play an important role in the pathogenetic development of AD. We investigated the expression levels of CXCR4 / CXCL12 in fifteen brain regions of healthy non-demented subjects (NDHC) (2139 sample) and AD patients (1170 sample) stratified according to sex and age. Furthermore, we correlated their expressions with the Neurogranin (NRGN) and CHI3L1 levels, two inflamm-aging markers. We highlighted that CXCR4 gene expression levels were age-correlated in the brain of NDHC subjects and that AD nullified this correlation. A similar trend, but diametrically opposite was observed for CXCL12. Its expression was decreased during the aging in both sexes, and in the brains of AD patients, it underwent an inversion of the trend, only and exclusively in females. Brains of AD patients expressed high CXCR4 and CHI3L1, and low CXCL12 and Neurogranin levels compared to NDHC subjects. Both CXCR4 and CXCL12 correlated significantly with CHI3L1 and Neurogranin expression levels, regardless of disease. Furthermore, we showed a selective modulation of CXCL12 and CXCR4 only in specific brain regions. Taken together our results demonstrate that CXCL12 and CXCR4 are linked to Neurogranin and CHI3L1 expression levels and the relationship between postsynaptic damage and microglial activation in AD could be shown using all these genes. Further confirmations are needed to demonstrate the close link between these genes.

CHI3L2 Expression Levels Are Correlated with AIF1, PECAM1, and CALB1 in the Brains of Alzheimer's Disease Patients

Alzheimer's disease (AD) represents one of the main forms of dementia that afflicts our society. The expression of several genes has been associated with disease development. Despite this, the number of genes known to be capable of discriminating between AD patients according to sex remains deficient. In our study, we performed a transcriptomes meta-analysis on a large court of brains of healthy control subjects (n = 2139) (NDHC) and brains of AD patients (n = 1170). Our aim was to verify the brain expression levels of CHI3L2 and its correlation with genes associated with microglia-mediated neuroinflammation (IBA1), alteration of the blood-brain barrier (PECAM1), and neuronal damage (CALB1). We showed that the CHI3L2, IBA1, PECAM1, and CALB1 expression levels were modulated in the brains of patients with AD compared to NDHC subjects. Furthermore, both in NDHC and in AD patient's brains, the CHI3L2 expression levels were directly correlated with IBA1 and PECAM1 and inversely with CALB1. Additionally, the expression levels of CHI3L2, PECAM1, and CALB1 but not of IBA1 were sex-depended. By stratifying the samples according to age and sex, correlation differences emerged between the expression levels of CHI3L2, IBA1, PECAM1, and CALB1 and the age of NDHC subjects and AD patients. CHI3L2 represents a promising gene potentially involved in the key processes underlying Alzheimer's disease. Its expression in the brains of sex-conditioned AD patients opens up new possible sex therapeutic strategies aimed at controlling imbalance in disease progression.

Transcriptomic comparison of human and mouse brain microvessels

The brain vasculature maintains brain homeostasis by tightly regulating ionic, molecular, and cellular transport between the blood and the brain parenchyma. These blood-brain barrier (BBB) properties are impediments to brain drug delivery, and brain vascular dysfunction accompanies many neurological disorders. The molecular constituents of brain microvascular endothelial cells (BMECs) and pericytes, which share a basement membrane and comprise the microvessel structure, remain incompletely characterized, particularly in humans. To improve the molecular database of these cell types, we performed RNA sequencing on brain microvessel preparations isolated from snap-frozen human and mouse tissues by laser capture microdissection (LCM). The resulting transcriptome datasets from LCM microvessels were enriched in known brain endothelial and pericyte markers, and global comparison identified previously unknown microvessel-enriched genes. We used these datasets to identify mouse-human species differences in microvessel-associated gene expression that may have relevance to BBB regulation and drug delivery. Further, by comparison of human LCM microvessel data with existing human BMEC transcriptomic datasets, we identified novel putative markers of human brain pericytes. Together, these data improve the molecular definition of BMECs and brain pericytes, and are a resource for rational development of new brain-penetrant therapeutics and for advancing understanding of brain vascular function and dysfunction.

Meta-analysis of human prefrontal cortex reveals activation of GFAP and decline of synaptic transmission in the aging brain

Despite ongoing research efforts, mechanisms of brain aging are still enigmatic and need to be elucidated for a better understanding of age-associated cognitive decline. The aim of this study is to investigate aging in the prefrontal cortex region of human brain in a meta-analysis of transcriptome datasets. We analyzed 591 gene expression datasets pertaining to female and male human prefrontal cortex biopsies of distinct ages. We used hierarchical clustering and principal component analysis (PCA) to determine the influence of sex and age on global transcriptome levels. In sex-specific analysis we identified genes correlating with age and differentially expressed between groups of young, middle-aged and aged. Pathways and gene ontologies (GOs) over-represented in the resulting gene sets were calculated. Potential causal relationships between genes and between GOs were explored employing the Granger test of gene expression time series over the range of ages. The most outstanding results were the age-related decline of synaptic transmission and activated expression of glial fibrillary acidic protein (GFAP) in both sexes. We found an antagonistic relationship between calcium/calmodulin dependent protein kinase IV (CAMK4) and GFAP which may include regulatory mechanisms involving cAMP responsive element binding protein (CREB) and mitogen-activated protein kinase (MAPK, alias ERK). Common to both sexes was a decline in synaptic transmission, neurogenesis and an increased base-level of inflammatory and immune-related processes. Furthermore, we detected differences in dendritic spine morphogenesis, catecholamine signaling and cellular responses to external stimuli, particularly to metal (Zinc and cadmium) ions which were higher in female brains.

Amniotic fluid cell-free transcriptome: a glimpse into fetal development and placental cellular dynamics during normal pregnancy

Background

The amniotic fluid (AF) cell-free transcriptome is modulated by physiologic and pathologic processes during pregnancy. AF gene expression changes with advancing gestation reflect fetal development and organ maturation; yet, defining normal expression and splicing patterns for biomarker discovery in obstetrics requires larger heterogeneous cohorts, evaluation of potential confounding factors, and novel analytical approaches.

Methods

Women with a normal pregnancy who had an AF sample collected during midtrimester (n = 30) or at term gestation (n = 68) were included. Expression profiling at exon level resolution was performed using Human Transcriptome Arrays. Differential expression was based on moderated t-test adjusted p < 0.05 and fold change > 1.25; for differential splicing, a splicing index > 2 and adjusted p < 0.05 were required. Functional profiling was used to interpret differentially expressed or spliced genes. The expression of tissue-specific and cell-type specific signatures defined by single-cell genomics was quantified and correlated with covariates. In-silico validation studies were performed using publicly available datasets.

Results

1) 64,071 genes were detected in AF, with 11% of the coding and 6% of the non-coding genes being differentially expressed between midtrimester and term gestation. Expression changes were highly correlated with those previously reported (R > 0.79, p < 0.001) and featured increased expression of genes specific to the trachea, salivary glands, and lung and decreased expression of genes specific to the cardiac myocytes, uterus, and fetal liver, among others. 2) Single-cell RNA-seq signatures of the cytotrophoblast, Hofbauer cells, erythrocytes, monocytes, T and B cells, among others, showed complex patterns of modulation with gestation (adjusted p < 0.05). 3) In 17% of the genes detected, we found differential splicing with advancing gestation in genes related to brain development processes and immunity pathways, including some that were missed based on differential expression analysis alone.

Conclusions

This represents the largest AF transcriptomics study in normal pregnancy, reporting for the first time that single-cell genomic signatures can be tracked in the AF and display complex patterns of expression during gestation. We also demonstrate a role for alternative splicing in tissue-identity acquisition, organ development, and immune processes. The results herein may have implications for the development of fetal testing to assess placental function and fetal organ maturity.

EPIC: A Tool to Estimate the Proportions of Different Cell Types from Bulk Gene Expression Data

Gene expression profiling is nowadays routinely performed on clinically relevant samples (e.g., from tumor specimens). Such measurements are often obtained from bulk samples containing a mixture of cell types. Knowledge of the proportions of these cell types is crucial as they are key determinants of the disease evolution and response to treatment. Moreover, heterogeneity in cell type proportions across samples is an important confounding factor in downstream analyses.Many tools have been developed to estimate the proportion of the different cell types from bulk gene expression data. Here, we provide guidelines and examples on how to use these tools, with a special focus on our recent computational method EPIC (Estimating the Proportions of Immune and Cancer cells). EPIC includes RNA-seq-based gene expression reference profiles from immune cells and other nonmalignant cell types found in tumors. EPIC can additionally manage user-defined gene expression reference profiles. Some unique features of EPIC include the ability to account for an uncharacterized cell type, the introduction of a renormalization step to account for different mRNA content in each cell type, and the use of single-cell RNA-seq data to derive biologically relevant reference gene expression profiles. EPIC is available as a web application ( http://epic.gfellerlab.org ) and as an R-package ( https://github.com/GfellerLab/EPIC ).

Pathway and network embedding methods for prioritizing psychiatric drugs

One in five Americans experience mental illness, and roughly 75% of psychiatric prescriptions do not successfully treat the patient's condition. Extensive evidence implicates genetic factors and signaling disruption in the pathophysiology of these diseases. Changes in transcription often underlie this molecular pathway dysregulation; individual patient transcriptional data can improve the efficacy of diagnosis and treatment. Recent large-scale genomic studies have uncovered shared genetic modules across multiple psychiatric disorders - providing an opportunity for an integrated multi-disease approach for diagnosis. Moreover, network-based models informed by gene expression can represent pathological biological mechanisms and suggest new genes for diagnosis and treatment. Here, we use patient gene expression data from multiple studies to classify psychiatric diseases, integrate knowledge from expert-curated databases and publicly available experimental data to create augmented disease-specific gene sets, and use these to recommend disease-relevant drugs. From Gene Expression Omnibus, we extract expression data from 145 cases of schizophrenia, 82 cases of bipolar disorder, 190 cases of major depressive disorder, and 307 shared controls. We use pathway-based approaches to predict psychiatric disease diagnosis with a random forest model (78% accuracy) and derive important features to augment available drug and disease signatures. Using protein-protein-interaction networks and embedding-based methods, we build a pipeline to prioritize treatments for psychiatric diseases that achieves a 3.4-fold improvement over a background model. Thus, we demonstrate that gene-expression-derived pathway features can diagnose psychiatric diseases and that molecular insights derived from this classification task can inform treatment prioritization for psychiatric diseases.

Gene Regulatory Network of Dorsolateral Prefrontal Cortex: a Master Regulator Analysis of Major Psychiatric Disorders

Despite the strong genetic component of psychiatric disorders, traditional genetic studies have failed to find individual genes of large effect size. Thus, alternative methods, using bioinformatics, have been proposed to solve these biological puzzles. Of these, here we employ systems biology-based approaches to identify potential master regulators (MRs) of bipolar disorder (BD), schizophrenia (SZ), and major depressive disorder (MDD), their association with biological processes and their capacity to differentiate disorders' phenotypes. High-throughput gene expression data was used to reconstruct standard human dorsolateral prefrontal cortex regulatory transcriptional network, which was then queried for regulatory units and MRs associated with the psychiatric disorders of interest. Furthermore, the activity status (active or repressed) of MR candidates was obtained and used in cluster analysis to characterize disease phenotypes. Finally, we explored the biological processes modulated by the MRs using functional enrichment analysis. Thirty-one, thirty-four, and fifteen MR candidates were identified in BD, SZ, and MDD, respectively. The activity state of these MRs grouped the illnesses in three clusters: MDD only, mostly BD, and a third one with BD and SZ. While BD and SZ share several biological processes related to ion transport and homeostasis, synapse, and immune function, SZ showed peculiar enrichment of processes related to cytoskeleton and neuronal structure. Meanwhile, MDD presented mostly processes related to glial development and fatty acid metabolism. Our findings suggest notable differences in functional enrichment between MDD and BD/SZ. Furthermore, similarities between BD and SZ may impose particular challenges in attempts to discriminate these pathologies based solely on their transcriptional profiles. Nevertheless, we believe that systems-oriented approaches are promising strategies to unravel the pathophysiology peculiarities underlying mental illnesses and reveal therapeutic targets.

Phosphoinositides: Regulators of Nervous System Function in Health and Disease

Phosphoinositides, the seven phosphorylated derivatives of phosphatidylinositol have emerged as regulators of key sub-cellular processes such as membrane transport, cytoskeletal function and plasma membrane signaling in eukaryotic cells. All of these processes are also present in the cells that constitute the nervous system of animals and in this setting too, these are likely to tune key aspects of cell biology in relation to the unique structure and function of neurons. Phosphoinositides metabolism and function are mediated by enzymes and proteins that are conserved in evolution, and analysis of knockouts of these in animal models implicate this signaling system in neural function. Most recently, with the advent of human genome analysis, mutations in genes encoding components of the phosphoinositide signaling pathway have been implicated in human diseases although the cell biological basis of disease phenotypes in many cases remains unclear. In this review we evaluate existing evidence for the involvement of phosphoinositide signaling in human nervous system diseases and discuss ways of enhancing our understanding of the role of this pathway in the human nervous system's function in health and disease.

Central and Peripheral Changes in FOS Expression in Schizophrenia Based on Genome-Wide Gene Expression

Schizophrenia is a chronic, debilitating neuropsychiatric disorder. Multiple transcriptomic gene expression profiling analysis has been used to identify schizophrenia-associated genes, unravel disease-associated biomarkers, and predict clinical outcomes. We aimed to identify gene expression regulation, underlying pathways, and their roles in schizophrenia pathogenesis. We searched the Gene Expression Omnibus (GEO) database for microarray studies of fibroblasts, lymphoblasts, and post-mortem brains of schizophrenia patients. Our analysis demonstrated high FOS expression in non-neural peripheral samples and low FOS expression in brain tissues of schizophrenia patients compared with healthy controls. FOS exhibited predictive value for schizophrenia patients in these datasets. Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis revealed that “amphetamine addiction” was among the top 10 significantly enriched KEGG pathways. FOS and FOSB, which are implicated in the amphetamine addiction pathway, were up-regulated in schizophrenia fibroblast samples. Protein–protein interaction (PPI) network analysis revealed that proteins closely interacting with FOS-encoded protein were also involved in the amphetamine addiction pathway. Pearson correlation test indicated that FOS showed positive correlation with genes in the amphetamine pathway. The results revealed that FOS was acceptable as a biomarker for schizophrenia and may be involved in schizophrenia pathogenesis.

Cross-species systems analysis identifies gene networks differentially altered by sleep loss and depression

To understand the transcriptomic organization underlying sleep and affective function, we studied a population of (C57BL/6J × 129S1/SvImJ) F2 mice by measuring 283 affective and sleep phenotypes and profiling gene expression across four brain regions. We identified converging molecular bases for sleep and affective phenotypes at both the single-gene and gene-network levels. Using publicly available transcriptomic datasets collected from sleep-deprived mice and patients with major depressive disorder (MDD), we identified three cortical gene networks altered by the sleep/wake state and depression. The network-level actions of sleep loss and depression were opposite to each other, providing a mechanistic basis for the sleep disruptions commonly observed in depression, as well as the reported acute antidepressant effects of sleep deprivation. We highlight one particular network composed of circadian rhythm regulators and neuronal activity-dependent immediate-early genes. The key upstream driver of this network, Arc, may act as a nexus linking sleep and depression. Our data provide mechanistic insights into the role of sleep in affective function and MDD.