Limits...
High-throughput identification of long-range regulatory elements and their target promoters in the human genome.

Hwang YC, Zheng Q, Gregory BD, Wang LS - Nucleic Acids Res. (2013)

Bottom Line: We observed that these promoter-interacting hotspots significantly overlap with known enhancer-associated histone modifications and DNase I hypersensitive sites.Thus, we defined thousands of candidate enhancer elements by incorporating these features, and found that they have a significant propensity to be bound by p300, an enhancer binding transcription factor.In total, our study presents a novel high-throughput workflow for confident, genome-wide discovery of enhancer-target promoter pairs, which will significantly improve our understanding of these regulatory interactions.

View Article: PubMed Central - PubMed

Affiliation: Genomics and Computational Biology Graduate Program, University of Pennsylvania, Philadelphia, PA, USA.

ABSTRACT
Enhancer elements are essential for tissue-specific gene regulation during mammalian development. Although these regulatory elements are often distant from their target genes, they affect gene expression by recruiting transcription factors to specific promoter regions. Because of this long-range action, the annotation of enhancer element-target promoter pairs remains elusive. Here, we developed a novel analysis methodology that takes advantage of Hi-C data to comprehensively identify these interactions throughout the human genome. To do this, we used a geometric distribution-based model to identify DNA-DNA interaction hotspots that contact gene promoters with high confidence. We observed that these promoter-interacting hotspots significantly overlap with known enhancer-associated histone modifications and DNase I hypersensitive sites. Thus, we defined thousands of candidate enhancer elements by incorporating these features, and found that they have a significant propensity to be bound by p300, an enhancer binding transcription factor. Furthermore, we revealed that their target genes are significantly bound by RNA Polymerase II and demonstrate tissue-specific expression. Finally, we uncovered that these elements are generally found within 1 Mb of their targets, and often regulate multiple genes. In total, our study presents a novel high-throughput workflow for confident, genome-wide discovery of enhancer-target promoter pairs, which will significantly improve our understanding of these regulatory interactions.

Show MeSH

Related in: MedlinePlus

Genome-wide enhancer element identification workflow.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3643598&req=5

gkt188-F1: Genome-wide enhancer element identification workflow.

Mentions: We built an analysis workflow that extracts high-quality DNA interacting sites from Hi-C datasets. Figure 1 shows the overall workflow for identifying these DNA interacting hotspots, which we analyzed further to identify putative enhancer elements and their promoter partners. All three samples from the original Hi-C study (16) were used in our analyses [cell line GM06990 with REs HindIII and NcoI, as well as cell line K562 with HindIII (referred to as GM/HindIII, GM/NcoI and K562/HindIII, respectively)]. The original Hi-C study used a 1 Mb window size to uncover the 3D organization of human nuclear chromosomes. However, this resolution is far too coarse for studying regulatory elements, which requires single nucleotide resolution. To improve resolution for our purposes of identifying DNA interacting hotspots, we applied our genomic distribution-based analysis for identification of these specific genomic regions (24). Briefly, our algorithm first identifies clusters of Hi-C reads that are closer to each other than what the background geometric distribution dictates. We then labeled the resulting clusters as hotspots if their lengths on the chromosomes are longer than 99% of all clusters. We found that a hotspot is on average ∼1 kb in length, and between 107 059 and 185 042 total hotspots were identified in each of the three samples.Figure 1.


High-throughput identification of long-range regulatory elements and their target promoters in the human genome.

Hwang YC, Zheng Q, Gregory BD, Wang LS - Nucleic Acids Res. (2013)

Genome-wide enhancer element identification workflow.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3643598&req=5

gkt188-F1: Genome-wide enhancer element identification workflow.
Mentions: We built an analysis workflow that extracts high-quality DNA interacting sites from Hi-C datasets. Figure 1 shows the overall workflow for identifying these DNA interacting hotspots, which we analyzed further to identify putative enhancer elements and their promoter partners. All three samples from the original Hi-C study (16) were used in our analyses [cell line GM06990 with REs HindIII and NcoI, as well as cell line K562 with HindIII (referred to as GM/HindIII, GM/NcoI and K562/HindIII, respectively)]. The original Hi-C study used a 1 Mb window size to uncover the 3D organization of human nuclear chromosomes. However, this resolution is far too coarse for studying regulatory elements, which requires single nucleotide resolution. To improve resolution for our purposes of identifying DNA interacting hotspots, we applied our genomic distribution-based analysis for identification of these specific genomic regions (24). Briefly, our algorithm first identifies clusters of Hi-C reads that are closer to each other than what the background geometric distribution dictates. We then labeled the resulting clusters as hotspots if their lengths on the chromosomes are longer than 99% of all clusters. We found that a hotspot is on average ∼1 kb in length, and between 107 059 and 185 042 total hotspots were identified in each of the three samples.Figure 1.

Bottom Line: We observed that these promoter-interacting hotspots significantly overlap with known enhancer-associated histone modifications and DNase I hypersensitive sites.Thus, we defined thousands of candidate enhancer elements by incorporating these features, and found that they have a significant propensity to be bound by p300, an enhancer binding transcription factor.In total, our study presents a novel high-throughput workflow for confident, genome-wide discovery of enhancer-target promoter pairs, which will significantly improve our understanding of these regulatory interactions.

View Article: PubMed Central - PubMed

Affiliation: Genomics and Computational Biology Graduate Program, University of Pennsylvania, Philadelphia, PA, USA.

ABSTRACT
Enhancer elements are essential for tissue-specific gene regulation during mammalian development. Although these regulatory elements are often distant from their target genes, they affect gene expression by recruiting transcription factors to specific promoter regions. Because of this long-range action, the annotation of enhancer element-target promoter pairs remains elusive. Here, we developed a novel analysis methodology that takes advantage of Hi-C data to comprehensively identify these interactions throughout the human genome. To do this, we used a geometric distribution-based model to identify DNA-DNA interaction hotspots that contact gene promoters with high confidence. We observed that these promoter-interacting hotspots significantly overlap with known enhancer-associated histone modifications and DNase I hypersensitive sites. Thus, we defined thousands of candidate enhancer elements by incorporating these features, and found that they have a significant propensity to be bound by p300, an enhancer binding transcription factor. Furthermore, we revealed that their target genes are significantly bound by RNA Polymerase II and demonstrate tissue-specific expression. Finally, we uncovered that these elements are generally found within 1 Mb of their targets, and often regulate multiple genes. In total, our study presents a novel high-throughput workflow for confident, genome-wide discovery of enhancer-target promoter pairs, which will significantly improve our understanding of these regulatory interactions.

Show MeSH
Related in: MedlinePlus