Genome-Wide Co-Expression Analysis in Multiple Tissues

University of Cape Town, South Africa
PLoS ONE (Impact Factor: 3.23). 12/2008; 3(12):e4033. DOI: 10.1371/journal.pone.0004033
Source: PubMed


Expression quantitative trait loci (eQTLs) represent genetic control points of gene expression, and can be categorized as cis- and trans-acting, reflecting local and distant regulation of gene expression respectively. Although there is evidence of co-regulation within clusters of trans-eQTLs, the extent of co-expression patterns and their relationship with the genotypes at eQTLs are not fully understood. We have mapped thousands of cis- and trans-eQTLs in four tissues (fat, kidney, adrenal and left ventricle) in a large panel of rat recombinant inbred (RI) strains. Here we investigate the genome-wide correlation structure in expression levels of eQTL transcripts and underlying genotypes to elucidate the nature of co-regulation within cis- and trans-eQTL datasets. Across the four tissues, we consistently found statistically significant correlations of cis-regulated gene expression to be rare (<0.9% of all pairs tested). Most (>80%) of the observed significant correlations of cis-regulated gene expression are explained by correlation of the underlying genotypes. In comparison, co-expression of trans-regulated gene expression is more common, with significant correlation ranging from 2.9%-14.9% of all pairs of trans-eQTL transcripts. We observed a total of 81 trans-eQTL clusters (hot-spots), defined as consisting of > or =10 eQTLs linked to a common region, with very high levels of correlation between trans-regulated transcripts (77.2-90.2%). Moreover, functional analysis of large trans-eQTL clusters (> or =30 eQTLs) revealed significant functional enrichment among genes comprising 80% of the large clusters. The results of this genome-wide co-expression study show the effects of the eQTL genotypes on the observed patterns of correlation, and suggest that functional relatedness between genes underlying trans-eQTLs is reflected in the degree of co-expression observed in trans-eQTL clusters. Our results demonstrate the power of an integrative, systematic approach to the analysis of a large gene expression dataset to uncover underlying structure, and inform future eQTL studies.

Download full-text


Available from: Nick Dickens, Mar 17, 2014
  • Source
    • "In other cases, when the expression level of a gene associates with a locus clearly distinct from that of the gene itself, it is defined as a distal eQTL (also called " trans-eQTL " ). In the case of trans-eQTLs, it has been observed that a single genetic locus can show linkage to the abundance of the mRNA transcript of several genes across the genome, forming so-called trans-eQTL " hotspots " or " bands " (Grieve et al. 2008; Wu et al. 2008). One example would, for instance, be that of a cis-eQTL regulating the expression of a transcription factor, which in turns regulates the expression of many other genes belonging to a hotspot of trans-eQTLs. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Although gene co-expression domains have been reported in most eukaryotic organisms, data available to date suggested that co-expression rarely concerned more than doublets or triplets of adjacent genes in mammals. Using expression data from hearts of mice from the panel of AxB/BxA recombinant inbred mice, we detected (according to window sizes) 42-53 loci linked to the expression levels of clusters of 3 or more neighboring genes. These loci thus formed "cis-expression QTL clusters", because their position matched that of the genes whose expression was linked to the loci. Compared to matching control regions, genes contained within cis-eQTL clusters showed much higher levels of co-expression. Corresponding regions also showed greater abundance of polymorphic elements corresponding to SINE retrotransposons. The latter showed enrichment for the motifs of binding sites for various transcription factors, with binding sites for the chromatin organizing CCCTC-binding factor showing highest levels of enrichment in polymorphic SINEs. Similar cis-eQTL clusters were also detected when using data obtained with several tissues from BxD recombinant inbred mice. In addition to strengthening the evidence for gene expression domains in mammalian genomes, our data suggest a possible mechanism whereby non-coding polymorphisms could affect the coordinate expression of several neighboring genes.
    G3-Genes Genomes Genetics 03/2013; 3(4). DOI:10.1534/g3.113.005546 · 3.20 Impact Factor
  • Source
    • "A total of 5,818 overlapping QTLs (spanning the same genomic region) were found in leaf and tuber, with only 401 features (6.9%) exhibiting trans-acting transcriptional control (Table 1c). It has previously been suggested that trans-acting QTLs are more often tissue specific [5,28]. In this study, we find a similar result when considering only those QTLs that were uniquely found in one of the two tissues (leaf or tuber) for which the percentages of trans-acting QTLs are elevated to 34.8% for leaf and 48.8% for tuber, respectively. "
    [Show abstract] [Hide abstract]
    ABSTRACT: With the completion of genome sequences belonging to some of the major crop plants, new challenges arise to utilize this data for crop improvement and increased food security. The field of genetical genomics has the potential to identify genes displaying heritable differential expression associated to important phenotypic traits. Here we describe the identification of expression QTLs (eQTLs) in two different potato tissues of a segregating potato population and query the potato genome sequence to differentiate between cis- and trans-acting eQTLs in relation to gene subfunctionalization. Leaf and tuber samples were analysed and screened for the presence of conserved and tissue dependent eQTLs. Expression QTLs present in both tissues are predominantly cis-acting whilst for tissue specific QTLs, the percentage of trans-acting QTLs increases. Tissue dependent eQTLs were assigned to functional classes and visualized in metabolic pathways. We identified a potential regulatory network on chromosome 10 involving genes crucial for maintaining circadian rhythms and controlling clock output genes. In addition, we show that the type of genetic material screened and sampling strategy applied, can have a high impact on the output of genetical genomics studies. Identification of tissue dependent regulatory networks based on mapped differential expression not only gives us insight in tissue dependent gene subfunctionalization but brings new insights into key biological processes and delivers targets for future haplotyping and genetic marker development.
    BMC Plant Biology 02/2012; 12(1):17. DOI:10.1186/1471-2229-12-17 · 3.81 Impact Factor
  • Source
    • "Recent advances in techniques to genetically modify the rat [26-30] enables detailed analyses of rat physiology at molecular levels. Furthermore, well-established genetic systems, such as congenic, consomic and recombinant inbred lines are versatile tools for studying the effect of genetic variation on quantitative traits such as blood pressure [31] or gene expression [32] (corresponding to quantitative trait loci (QTLs) and expression (e)QTLs, respectively). Comprehensive small RNA inventories and profiles are instrumental in such genetical genomics and systems biological approaches, as they serve as a resource for annotation of the genome. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Digital gene expression (DGE) profiling has become an established tool to study RNA expression. Here, we provide an in-depth analysis of small RNA DGE profiles from two different rat strains (BN-Lx and SHR) from six different rat tissues (spleen, liver, brain, testis, heart, kidney). We describe the expression patterns of known and novel micro (mi)RNAs and piwi-interacting (pi)RNAs. We confirmed the expression of 588 known miRNAs (54 in antisense orientation) and identified 56 miRNAs homologous to known human or mouse miRNAs, as well as 45 new rat miRNAs. Furthermore, we confirmed specific A to I editing in brain for mir-376a/b/c and identified mir-377 as a novel editing target. In accordance with earlier findings, we observed a highly tissue-specific expression pattern for all tissues analyzed. The brain was found to express the highest number of tissue-specific miRNAs, followed by testis. Notably, our experiments also revealed robust strain-specific differential miRNA expression in the liver that is caused by genetic variation between the strains. Finally, we identified two types of germline-specific piRNAs in testis, mapping either to transposons or in strand-specific clusters. Taken together, the small RNA compendium described here advances the annotation of small RNAs in the rat genome. Strain and tissue-specific expression patterns furthermore provide a strong basis for studying the role of small RNAs in regulatory networks as well as biological process like physiology and neurobiology that are extensively studied in this model system.
    BMC Genomics 04/2010; 11(1):249. DOI:10.1186/1471-2164-11-249 · 3.99 Impact Factor
Show more