Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Oct 24;21(4):1077-1088.
doi: 10.1016/j.celrep.2017.10.001.

Understanding Tissue-Specific Gene Regulation

Affiliations
Free PMC article

Understanding Tissue-Specific Gene Regulation

Abhijeet Rajendra Sonawane et al. Cell Rep. .
Free PMC article

Abstract

Although all human tissues carry out common processes, tissues are distinguished by gene expression patterns, implying that distinct regulatory programs control tissue specificity. In this study, we investigate gene expression and regulation across 38 tissues profiled in the Genotype-Tissue Expression project. We find that network edges (transcription factor to target gene connections) have higher tissue specificity than network nodes (genes) and that regulating nodes (transcription factors) are less likely to be expressed in a tissue-specific manner as compared to their targets (genes). Gene set enrichment analysis of network targeting also indicates that the regulation of tissue-specific function is largely independent of transcription factor expression. In addition, tissue-specific genes are not highly targeted in their corresponding tissue network. However, they do assume bottleneck positions due to variability in transcription factor targeting and the influence of non-canonical regulatory interactions. These results suggest that tissue specificity is driven by context-dependent regulatory paths, providing transcriptional control of tissue-specific processes.

Keywords: GTEx; gene expression; gene regulation; network biology; network medicine; regulatory networks; tissue specificity; transcription factors; transcriptional regulation; transcriptome.

Figures

Figure 1
Figure 1. Schematic Overview of Our Approach
We characterized tissue-specific gene regulation starting with GTEx gene expression data; the relative sample size of each of the 38 tissues in the expression data is shown in the color bar. We then used PANDA to integrate this information with protein-protein interaction (PPI) and transcription factor (TF) target information, producing 38 inferred gene regulatory networks, one for each tissue. We identified tissue-specific genes, transcription factors, and regulatory network edges, and we analyzed their properties within and across these networks.
Figure 2
Figure 2. Tissue Specificity of Network Elements
(A–C) Bar plots illustrating the number of edges (A), genes (B), and transcription factors (TFs) (C) that were identified as specific to each of the 38 GTEx tissues. The number of elements identified as specific in each tissue is shown to the right of each bar. The multiplicity of edges, genes, and TFs is indicated by the color of the bars. Note that an edge/gene/TF with a given multiplicity across all tissues (top bar plots) will appear in that number of tissue-specific bar plots (lower bar plots). See also Tables S1, S2, and S3 and Figures S1–S4.
Figure 3
Figure 3. Enrichment of Tissue-Specific Edges
(A–C) The log2 of the number of observed/expected edges of different multiplicities (0 = non-tissue specific) connected to tissue-specific (TS) genes (A) and tissue-specific transcription factors (TFs) (B) or overlapping with canonical regulatory interactions (C). See also the Supplemental Experimental Procedures and Table S4.
Figure 4
Figure 4. Tissue-Specific Targeting in Brain
(A) A hierarchical clustering (Euclidean distance, complete linkage) and heatmap depicting the gene set enrichment analysis results for tissue-specific targeting of all 644 transcription factors in the “brain other” gene regulatory network. FDR values for positive enrichment scores, indicating increased targeting, are shown in red; negative scores are in blue. FDR values greater than 0.25 appear in white. The top bar indicates whether a transcription factor was also identified as specific (black) to brain other or not (gray). (B) Heatmap for the 10 most (black) and 10 least (gray) tissue-specific transcription factors. AC, adenylate cyclase; act., activating pathway; reg., regulation; fam., family; MP, metabolic process. See also Table S5.
Figure 5
Figure 5. Tissue-Specific Targeting across All Tissues
(A) Heatmap depicting the results from gene set enrichment analyses on targeting profiles for all possible transcription factor (TF)-tissue pairs, grouped by the community assignment of the GO terms and TF-tissue pairs. (B) Word clouds summarizing the processes contained in each community. (C) An illustration of the tissues associated with each community. Edge width indicates the number of transcription factors identified as differentially targeting at least one GO term in the community in a particular tissue. For simplicity, we only illustrate connections to tissues that include five or more transcription factors. (D) Heatmap of transcription factors significantly associated with one of the nine largest communities; the grayscale gradient represents the probability that a transcription factor would be associated with a community by chance given a random shuffling of community assignments. See also Table S6 and Figures S5 and S6.
Figure 6
Figure 6. Centrality of Genes in Tissue-Specific Networks
(A) An example network illustrating the difference between high degree and betweenness. Transcription factors are shown as circles and target genes as squares. The color of each node indicates its centrality. An example node is shown with low degree but high betweenness. (B) Difference in the median centrality of tissue-specific (TS) genes compared to non-tissue-specific genes in each of the 38 networks. (C) Distribution of centrality values for all non-tissue-specific genes (black), genes specific in a particular tissue (red), and genes called tissue specific in some tissue, but not the tissue of interest (gray dashed line). See also Figure S7.
Figure 7
Figure 7. Centrality of eQTL-Associated Genes
(A and B) The median (dot) and interquartile range (error bar) across tissues of the percentile rank of genes based on their degree (A) or betweenness (B) centralities, as in Figure 6C, plotted separately for genes that have an eQTL association (1) in that tissue (QTLs-in-tissue), (2) with a GWAS variant in that tissue (GWAS-QTLs-in-tissue), (3) in another tissue (QTLs-in-other), or (4) with a GWAS variant in another tissue (GWAS-QTLs-in-other). For comparison, the median rank of tissue-specific and non-tissue-specific genes across these tissues is indicated by red and black lines, respectively. Note that this analysis is limited to 19 tissues and 29,155 genes. See also Table S7.

Similar articles

See all similar articles

Cited by 46 articles

See all "Cited by" articles

References

    1. Barabási AL, Oltvai ZN. Network biology: understanding the cell's functional organization. Nat Rev Genet. 2004;5:101–113. - PubMed
    1. Boyle EA, Li YI, Pritchard JK. An Expanded View of Complex Traits: From Polygenic to Omnigenic. Cell. 2017;169:1177–1186. - PMC - PubMed
    1. Clauset A, Newman MEJ, Moore C. Finding community structure in very large networks. Phys Rev E. 2004;70:066111. - PubMed
    1. Consortium, G.T.; GTEx Consortium. Human genomics The Genotype-Tissue Expression (GTEx) pilot analysis:multitissue gene regulation in humans. Science. 2015;348:648–660. - PMC - PubMed
    1. Dixit AB, Banerjee J, Srivastava A, Tripathi M, Sarkar C, Kakkar A, Jain M, Chandra PS. RNA-seq analysis of hippocampal tissues reveals novel candidate genes for drug refractory epilepsy in patients with MTLE-HS. Genomics. 2016;107:178–188. - PubMed

Substances

LinkOut - more resources

Feedback