Brain Cell Type Specific Gene Expression and Co-expression Network Architectures.

McKenzie AT, Wang M, Hauberg ME, Fullard JF, Kozlenkov A, Keenan A, Hurd YL, Dracheva S, Casaccia P, Roussos P, Zhang B. · 2018

Read Brain Cell Type Specific Gene Expression and Co-expression Network Architectures. by McKenzie AT, Wang M, Hauberg ME, Fullard JF, Kozlenkov A, Keenan A, Hurd YL, Dracheva S, Casaccia P, Roussos P, Zhang B. (2018) 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-pmc5995803.txt.

Introduction

McKenzie and co-authors build consensus gene signatures for brain cell types from several data sets. Their study appeared in Scientific Reports. Knowing which genes mark each brain cell type aids many studies. The team compared five human and mouse transcriptome data sets. They defined specificity, enrichment, and absolute expression measures. Consensus signatures were well conserved across the sources. Top markers tracked cell shares in bulk brain samples.

New marker genes were checked against an independent chromatin data set. Network analysis of single-cell data found robust cell modules. The authors built an R package, BRETIGEA, for the task. It estimates cell proportions from bulk expression data.

Catalog record

CREATORMcKenzie AT, Wang M, Hauberg ME, Fullard JF, Kozlenkov A, Keenan A, Hurd YL, Dracheva S, Casaccia P, Roussos P, Zhang B.
YEAR2018
COLLECTIONPAPERS
FORMATTXT
FILE SIZE77 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC5995803

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Brain Cell Type Specific Gene Expression and Co-expression Network Architectures By Andrew T McKenzie, Minghui Wang, Mads E Hauberg, John F Fullard, Alexey Kozlenkov, Alexandra Keenan, Yasmin L Hurd, Stella Dracheva, Patrizia Casaccia, Panos Roussos, Bin Zhang Published in: Scientific Reports DOI: 10.1038/s41598-018-27293-5 ABSTRACT Abstract Elucidating brain cell type specific gene expression patterns is critical towards a better understanding of how cell-cell communications may influence brain functions and dysfunctions. We set out to compare and contrast five human and murine cell type-specific transcriptome-wide RNA expression data sets that were generated within the past several years. We defined three measures of brain cell type-relative expression including specificity, enrichment, and absolute expression and identified corresponding consensus brain cell “signatures,” which were well conserved across data sets. We validated that the relative expression of top cell type markers are associated with proxies for cell type proportions in bulk RNA expression data from postmortem human brain samples. We further validated novel marker genes using an orthogonal ATAC-seq dataset. We performed multiscale coexpression network analysis of the single cell data sets and identified robust cell-specific gene modules. To facilitate the use of the cell type-specific genes for cell type proportion estimation and deconvolution from bulk brain gene expression data, we developed an R package, BRETIGEA. In summary, we identified a set of novel brain cell consensus signatures and robust networks from the integration of multiple datasets and therefore transcend limitations related to technical issues characteristic of each individual study. Subject terms: Network models, Network models INTRODUCTION Interactions among multiple cell types orchestrate the structures and functions of all animal tissues, including the mammalian brain. Distinct cell types in the brain play different and specialized roles in electrical signaling1,2, metabolic coupling3, axonal ensheathing4, regulation of blood flow5, and immune surveillance6,7. These cell types belong to distinct lineages and are developmentally specified through an integrated transcriptional and epigenetic control of cell differentiation and gene expression8,9. A conclusive number of distinct cell types in the mammalian brain cannot be provided without a certain level of uncertainty related to the goals of any given analysis, and is profoundly affected by the sensitivity and specificity of the technology used for cell classification. In bulk brain tissue, gene expression experiments have highlighted cell type composition based on the expression value of markers for five major cell types: neurons, astrocytes, oligodendrocytes, microglia, and endothelial cells10. However, within the neuronal population, depending on the source, it has been reported that approximately 50–250 neuronal sub-cell types11–13 exist. Similarly, within other lineages, many other cell types have been classified as separate entities, including oligodendrocyte precursor cells (also known as NG2 cells), ependymal cells, smooth muscle cells, and pericytes14. Over the past few years, a series of comprehensive RNA-seq experiments in different brain cell types have been published in humans15,16 and mice17–20. Some of these experiments have profiled gene expression of cell populations isolated through immunopanning procedures15,17. Immunopanning involves immunoprecipitation of particular cell types in cell culture plates, based on selection for an antibody adsorbed to the plate surface21. As such, the analysis of currently available data has to take into consideration the limitation of the cell-type isolation procedures, which often included a series of positive and negative selections with pre-defined cell type-specific markers. Others studies have performed RNA profiling of single cells with microfluidics devices and used clustering methods to identify cell types from the resulting RNA expression profiles16,18,19. The devices used for single cell RNA sequencing (scRNA-seq) often select cells based on size or via encapsulation in a droplet22 and involve the creation of a cDNA library from the transcriptome from a theoretical maximum of one cell. Single cell experiments capture a wider range of cell types than in immunopanning, which reduces bias but acts to increase the variance of the resulting cell type signatures, thus requiring larger sample sizes for analysis. This larger sample size in scRNA-seq, in turn, allows investigators to interrogate the correlation space through network analysis of the interactions among genes23,24. However, to the best of our knowledge, when these methods have been applied to brain scRNA-seq data, they have not used a multiscale approach that allows for identification of overlapping gene modules as well as individual gene-gene interactions, as can be performed by MEGENA (Multiscale Embedded Gene Co-expression Network Analysis)25. Previous studies have analyzed brain cell type-specific expression signatures using microarray or RNA-seq in mice26,27. However, the existing studies have been mainly based on individual datasets, and are, therefore, subject to systematic noise, including sampling bias due to sample collection or preparation technique, as well as stochastic gene expression. As an increasing number of RNA-seq cell type-specific transcriptomic experiments have become available for both human and mouse, it is desirable to conduct a comprehensive meta-analysis of brain cell type gene signatures. In this manuscript, we first systematically evaluate cell type-specific RNA expression patterns identified in five of these RNA-seq studies15–19. The six cell types that we set out to compare are: astrocytes, endothelial cells, microglia, neurons, oligodendrocytes, and oligodendrocyte precursor cells (OPCs). We defined three criteria of ascription of cell type-associated gene expression: specificity, which measures whether a gene is expressed in only one cell type; enrichment, which measures whether a gene tends to have higher expression in one cell type compared to all others; and absolute expression, which measures whether a gene tends to have high expression in a given cell type, irrespective of its expression levels in other cell types. We showed that cell type enrichment patterns in the brain have high overall conservation in pairwise and global comparisons based on all three measures. Next, we interrogated marker genes most specific to each of the cell types, and found that they were significantly associated with text mining results in PubMed (https://www.ncbi.nlm.nih.gov/pubmed) searches, and that their aggregate expression patterns were well-correlated with immunohistochemistry-level data from the same brain samples. As a resource to the community, we release these gene signatures and functions for leveraging them in an R package, BRETIGEA (BRain cEll Type specIfic Gene Expression Analysis), which we validated on a postmortem brain gene expression data set. Finally, we perform network analysis on the scRNA-seq data sets using the multiscale network modeling approach MEGENA, showing that there is conservation of modules both within and across cell types. Taken together, our results establish fundamental cell type-associated gene expression patterns in the brain that can be used for many applications, including cell type enrichment analysis of differential expression signatures, deconvolution of relative cell type proportions from bulk RNA-seq data, and facilitating the interpretation of network analyses in scRNA-seq experiments. RESULTS Identification and comparison of cell type-associated expression across multiple data sets The goal of this study is to define and explore cell type-specific gene expression patterns via a meta-analysis of brain cell type RNA expression data from five studies with publically available RNA-sequencing reads15–19 (Table 1). Towards this end, we downloaded and reprocessed the raw sequencing reads with a unified pipeline, and used three separate metrics to calculate cell type-relative expression specificity, enrichment, and absolute expression data from each data set. We defined cell type specificity as the minimum fold change in expression between the cell type of interest and each of the other cells; cell type enrichment, as

Read the complete text