Transcriptome analysis reveals dysregulation of innate immune response genes and neuronal activity-dependent genes in autism.
Gupta S, Ellis SE, Ashar FN, Moes A, Bader JS, Zhan J, West AB, Arking DE. · 2014
Read Transcriptome analysis reveals dysregulation of innate immune response genes and neuronal activity-dependent genes in autism. by Gupta S, Ellis SE, Ashar FN, Moes A, Bader JS, Zhan J, West AB, Arking DE. (2014) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc4270294.txt.
Introduction
Simone Gupta leads a team that read the gene activity of postmortem brain tissue from people with autism and from matched controls. Their study appeared in Nature Communications in 2014. Gene hunting in autism had found great variety, so the team looked instead for shared patterns in which genes are turned up or down.
Using RNA sequencing, they found two linked patterns. A set of genes tied to microglia, the brain's immune cells, ran in step with a set of genes tied to nerve cell activity, and both ran off course in autism. The work points to an interplay between immune response and nerve signaling as a core feature of the disorder.
Catalog record
| CREATOR | Gupta S, Ellis SE, Ashar FN, Moes A, Bader JS, Zhan J, West AB, Arking DE. |
|---|---|
| YEAR | 2014 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 50 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Transcriptome analysis reveals dysregulation of innate immune response genes and neuronal activity-dependent genes in autism By Simone Gupta, Shannon E. Ellis, Foram N. Ashar, Anna Moes, Joel S. Bader, Jianan Zhan, Andrew B. West, Dan E. Arking Published in: Nature Communications DOI: 10.1038/ncomms6748 ABSTRACT Recent studies of genomic variation associated with autism have suggested the existence of extreme heterogeneity. Large-scale transcriptomics should complement these results to identify core molecular pathways underlying autism. Here we report results from a large-scale RNA sequencing effort, utilizing region-matched autism and control brains to identify neuronal and microglial genes robustly dysregulated in autism cortical brain. Remarkably, we note that a gene expression module corresponding to M2-activation states in microglia is negatively correlated with a differentially expressed neuronal module, implicating dysregulated microglial responses in concert with altered neuronal activity-dependent genes in autism brains. These observations provide pathways and candidate genes that highlight the interplay between innate immunity and neuronal activity in the aetiology of autism. Autism is a neurodevelopmental disorder in which affected individuals show disruption of the normal course of social and communicative development, along with restrictive interests and repetitive behaviours. In the vast majority of cases, the cause of disease is not known. Despite a strong genetic component with heritability estimates between 50 and 90% (refs 1, 2, 3), the identification of genes important in disease susceptibility has progressed slowly, resulting in a limited understanding of the molecular basis of autism. Environmental factors underlying disease risk are likewise unclear. This slow progress can be attributed, in part, to both the extreme phenotypic heterogeneity and involvement of hundreds of loci in the disorder45. Recent studies to elucidate the molecular basis of autism have largely focused on genetic approaches, including both genome-wide association studies6 and whole-exome sequencing5789, to identify both inherited and de novo variation contributing to autism. A major mode of action for genetic variation is through altered gene expression, so direct analysis of gene expression in a disease-relevant tissue is a complementary approach to genetic studies. Despite the extreme genetic heterogeneity observed in autism, it is possible that common downstream mechanisms may be altered10. Thus, there has been an effort to use transcriptomics to identify and dissect molecular pathways that may be altered in autism spectrum disorder (ASD). Given low tissue sample availability in autism research, efforts have focused on assessing gene expression in lymphoblastoid cell lines or whole blood1112131415. However, given the core neurodevelopment phenotypes associated with autism, there is little doubt that direct assessment of gene expression in brains may be critical. Indeed, Voineagu et al. recently utilized co-expressed gene networks from RNA-Sequencing (RNA-Seq) carried out in 19 autism brains and 17 controls to identify a set of co-expressed neuronal genes enriched for known autism susceptibility genes as well as a set of co-expressed genes enriched for both immune genes and glial markers10. In the current study, we present results from the largest RNA sequencing of autism brains effort to date that allows for new insights into the aetiology of autism. We find clear differences in the transcriptome between control and ASD cortical brains. Using co-expression network analysis, we demonstrate that autism brains are specifically enriched for ‘activated’ M2 microglial and ‘immune response’ genes. Remarkably, the M2 microglial module is strongly negatively correlated with one of two differentially expressed neuronal modules, highlighting the interplay between innate immunity and neuronal activity in the aetiology of ASD. RESULTS Transcriptomes from 104 human brain cortical tissue samples were resolved using next-generation RNA sequencing technology at single-gene resolution and through co-expressing gene clusters or modules. Multiple cortical tissues corresponding to Brodmann Area 19 (BA19), Brodmann Area 10 (BA10) and Brodmann Area 44 (BA44) were sequenced in 62, 14 and 28 samples, respectively, resulting in a total of 57 (40 unique individuals) control and 47 (32 unique individuals) autism samples (Supplementary Data 1 and 2, and see ref. 16). Differential gene expression was estimated between the 57 controls and the 47 cases, with sample collection site, age, sex, brain region and independent surrogate variables (ISVs) as fixed effects in a linear mixed regression model (Supplementary Data 3a). In total, 13,262 genes with at least three reads per sample across 90% of the samples were tested, and two transcriptome-wide significant differentially expressed genes associated with autism were identified (Supplementary Data 4 and Fig. 1a). The most significant differentially expressed gene was Myelin And Lymphocyte Protein (MAL; P=2.16 × 10−7, Fig. 1b, and Supplementary Fig. 1a–c). MAL, along with other myelination genes, has previously been reported to show altered expression in patients with psychiatric disorders17. The second differentially expressed gene was C11orf30 (EMSY; P=3.29 × 10−7, Fig. 1c and Supplementary Fig. 1d–f). This gene has been implicated in chromatin modification, DNA repair and transcriptional regulation, and previous genome-wide association studies have linked C11orf30 to inflammatory and malignant diseases1819. We also performed analyses at the exon level, testing 21,310 exons for differential exon expression; however, no differences could be identified in autism cases and controls after correction for multiple testing (Supplementary Data 5). Next we asked whether the top differentially expressed genes from the single-gene analysis shared common pathways or functional categories. We tested for the enrichment of biological processes using Gene Ontology (GO) annotations and MSigDB curated gene sets. No gene-set enrichment with family-wise error rate (FWER) ≤0.05 was observed (Supplementary Data 6a–d). For common genetic variation, altered gene expression is a major mode of action2021. We therefore tested whether genes previously associated with autism through genetic analyses are enriched for altered gene expression. To identify genes underlying susceptibility to autism, we utilized a list of expertly curated genes developed by the Simons Foundation for Autism Research (SFARI)22. In addition to the SFARI list of genes, we also integrated genes associated with rare de novo variation (RDNV) and genes involved in intellectual disability (ID) compiled from four published whole-exome sequencing studies57823 and review articles24 (Supplementary Data 7). Overall, we find comparable expression of these genes in autism and control brain tissue (Supplementary Data 8a,b). Although these gene lists are not comprehensive and only reflect the current understanding of the genetic basis of autism, the lack of enrichment for genes known to harbour genetic signal for autism in altered gene expression suggests the potential for non-overlapping mechanisms between genetic and transcriptomic determinants of autism. However, we do find modest enrichment for a broader set of genes containing ‘brain-critical exons’, which have high gene expression in brain and low rare burden of rare mutations, and have been proposed to represent autism candidate genes25 (Supplementary Data 8a). In addition to the single-gene analyses, we applied weighted gene correlation network analysis18 (WGCNA) to identify discrete gene modules based on co-expression between genes. Considerable overlap was observed in the modules constructed separately from cases and controls, indicating that overall organization of transcript co-expression is conserved between autism and control brains (Supplementary Data 9). Therefore, we applied WGCNA to construct networks derived from the entire data set of 104 samples, adjusted for sequencing artefacts, age, sex, collection site and brain region, identifying 12 co-expressed modules (Supplementary Data 10 and Fig. 2a). We tested the association of each module, represented by its corresponding first principal component or module eigengene (ME), with case–control status using a linear