Identifying, Characterizing, and Modulating Regulatory Elements in Their Natural Context
Total Page:16
File Type:pdf, Size:1020Kb
Identifying, characterizing, and modulating regulatory elements in their natural context Gregory E. Crawford Center for Genomic and Computational Biology Department of Pediatrics The Human Genome What does the other 98% do? 45% repetitive DNA 53% Unique and segmental duplicated DNA 2% genes (25,000) Lots of genomic contexts to explore… Evolution Population Diseases Different tissues Environmental exposures Development Overview of talk • Regulatory elements in 200 diverse human cell types • Resource for understanding disease genes • Identifying non-coding variants that impact chromatin structure and gene expression • Validating the function of regulatory elements using CRISPR/Cas9 epigenome editing strategies • High-throughput screens DNase hypersensitive (HS) sites identify active gene regulatory elements DNase I HS sites Regions hypersensitive to DNase HS sites identify “open” Promoters regions of chromatin Enhancers Silencers Insulators Locus control regions Meiotic recombination hotspots High-throughput identification of regulatory elements DNase-seq Single base resolution ~100 million seQuences per cell line or tissue n=150 SeQuencing using Illumina (DNase-seQ) Boyle et al., Cell 2008 A single DNase experiment matches most ChIP-seq data from 50 factors Thurman et al., Nature, 2012 Generating a chromatin atlas From >200 cell types Complex Disease Type 2 Diabetes Cross species Cancer •Human Preterm birth •Chimpanzee Population differences Schizophrenia •Orangutan Lymphoblastoids •MacaQue Pushing the envelope From different individuals •Mouse •Difficult cell types •3 Europeans •Endogenous nuclease •3 African •Frozen tissues •70 humans •Small numbers of cells •Male vs. Female Stem Cells Diverse •Embryonic Stem Cells •Brain •iPS (induced pluripotent •Blood Environmental Exposure cells) •Cytokines •Skin Different blood cell types •HDAC inhibitors •Heart •B cells Differentiation •Chemotherapy •Liver •T cells •Myoblasts -> Myotubes •Hormones •Kidney •Activated B/T cells •Muscle differentiation •Microbiota •Muscle •Neutrophils •Mouse brain development •Fat 200 cell types (> 1 million DNase sites) 200 cell types (> 1 million DNase sites) What is this DHS doing? What TFs bind to this Element? What gene(s) does this Element regulate? Can this help us understand genes that cause rare or common diseases? Functional Validation Regulatory Elements Surrounding CFTR locus Ann Harris, Northwestern Yang et al., NAR 2016 Chromatin varies across individuals Identification of individual-specific open chromatin using lymphoblastoid cells from 6 individuals Approximately 5% of open chromatin regions display individual/population differences McDaniell et al., Science 2010 DNase sensitive quantitative trait loci (dsQTL) (Jonathan Pritchard, U. of Chicago) DNase site 1 G DNase-seq 2 G Performed on lymphoblastoid cells 3 G from 70 individuals 4 G T 5 T 6 T 7 70 T Degner et al., Nature 2012 DNase sensitive quantitative trait loci (dsQTL) (Jonathan Pritchard, U. of Chicago) DNase site ++++ 1 G DNase-seq 2 G ++++ Performed on lymphoblastoid cells 3 G ++++ from 70 individuals 4 G ++++ T 5 + T 6 + T + 7 70 T + Degner et al., Nature 2012 DNase sensitive quantitative trait loci (dsQTL) (Jonathan Pritchard, U. of Chicago) DNase site ++++ 1 G 2 G ++++ ~9000 dsQTL identified 3 G ++++ 55% of eQTL=dsQTL 4 G ++++ T Validated by 5 + ChIP-seQ T 6 + T + 7 70 T + Degner et al., Nature 2012 Chromatin QTL analyses • ~9000 chromatin QTLs identified in lymphoblastoid cell lines. • Direct mechanism for how non-coding variants leads to altered gene expression • Recently we have identified another 9000 chromatin QTLs in brain samples as part of a study for schizophrenia (these are common!) • Relevance to rare and common disease unknown. Possible explanation for rare disease modifiers that influences disease severity? Targeted epigenome modulation using CRISPR/Cas9 Genome and epigenome editing by CRISPR/Cas9 Pennisi Science 2013 Epigenetic modifier (EGEM) toolbox TABLE&1.&Domains(used(for(EGEMs Remodeler Class Modification method/Ab VP64 Scaffold multiple multiple Tet1 DNAmeth 5mC(A>(5hmC oxidative(bisulfite p300 HAT H3K27(A>(H3K27ac ChIP((Abcam(ab4729) Activate PRDM9 HMT H3K4(A>(H3K4me3 ChIP((Abcam(ab8580) JMJD2D HDM H3K9me3(A>(H3K9 ChIP((Abcam(ab8898) JMJD3 HDM H3K27me3(A>(H3K27 ChIP((Millipore(07A449) KRAB Scaffold multiple multiple DNMT3a DNAmeth C(A>(5mC bisulfite SMRT/NCoR HDAC H3K27ac(A>(H3K27 ChIP((Abcam(ab4729) Repress SUV39H1 HMT H3K9(A>(H3K9me3 ChIP((Abcam(ab8898) LSD HDM H3K4me3(A>(H3K4 ChIP((Abcam(ab8580) Ezh2 HMT H3K27(A>(H3K27me3 ChIP((Millipore(07A449) Tim Reddy and Charlie Gersbach Epigenetic modifier (EGEM) toolbox Tim Reddy and Charlie Gersbach Epigenome activation using TALE or CRISPR coupled to VP64 activator TALE/CRISPR-VP64 seQ - GFP (negative control) DNase Perez-Pinera Nature Methods 2013 Epigenome activation using CRISPR/Cas9 Epigenome silencing using dCas9-KRAB closes chromatin and induces H3K9me3 chr11 HS1 HS2 HS3 HS4 HS5 HBB HBD HBBP1 HBG1 HBG2 HBE1 K562 Dnase-seq dCas9-KRAB + Cr10 vs dCas9-KRAB only dCas9-KRAB + Cr10 vs dCas9-KRAB only DNase-seq RNA-seq K562 Dnase-seq HBB HBG1 chr11:5275809-5276109 (HBG2) HBD (Fold ChanGe) (Fold ChanGe) 2 HBG2 2 chr11:5305943-5306243 (HS3) HBBP1 og chr11:5279828-5280129 l og l chr11:5305806-5306106 (HS3) HBE1 chr11:5301764-5302064 (Cr10 target) chr11:5301930-5302230 (HS2) Mean Normalized DNase-seq SiGnal Mean Normalized RNA-seq SiGnal Thakore, Nature Methods 2015 Epigenome silencing using dCas9-KRAB closes chromatin and induces H3K9me3 chr11 HS1 HS2 HS3 HS4 HS5 HBB HBD HBBP1 HBG1 HBG2 HBE1 K562 Dnase-seq dCas9 ChIP-seQ dCas9 + Cr4 dCas9-KRAB + Cr4 dCas9 + Cr10 dCas9-KRAB + Cr10 5265000 5285000 5305000 HBG1 HBG2 HS2 K562 Dnase-seq HBB HBD HBBP1 HBE1 HS1 HS3 HS4 HS5 H3K9me3 ChIP-seQ dCas9 + Cr4 dCas9-KRAB + Cr4 dCas9 + Cr10 dCas9-KRAB + Cr10 Cr10 Cr4 Thakore, Nature Methods 2015 High-throughput epigenome screens using CRISPR/Cas9 High-throughput CRISPR epigenome screens 27 Klann et al., Submitted Globin CRISPR epigenome screens to find enhancers globin gene HS2 HS1 HS3 HS4 HS5 GFP tagged Klann et al., Submitted Functional assays to characterize non-coding elements and variants involved in disease In vivo Low High Throughput Throughput Luciferase reporter, Gel shifts, etc. In vitro Functional assays to characterize non-coding elements and variants involved in disease In vivo Model organisms Genome and epigenome editing (CRISPR/Cas9) Low High Throughput Throughput Massively parallel Luciferase reporter, reporter Gel shifts, etc. In vitro Functional assays to characterize non-coding elements and variants involved in disease In vivo Model organisms Genome and epigenome editing (CRISPR/Cas9) Low High Throughput Throughput Massively parallel Luciferase reporter, reporter Gel shifts, etc. In vitro Functional assays to characterize non-coding elements and variants involved in disease In vivo Model organisms Chromatin accessibility Human variation (dsQTL) Genome and Comparative epigenomics epigenome editing (primates and mice) (CRISPR/Cas9) Tissues from cases/controls for different diseases Low High Throughput Throughput Massively parallel Luciferase reporter, reporter Gel shifts, etc. In vitro Summary • Huge amounts of chromatin space to explore • Over a million DNase HS sites identified • Resource for understanding mechanism of genes that cause rare and common disease • Lots of common variants in population that influence gene expression levels – modifiers? • Epigenome editing can be used to validate regulatory elements and functional variants. • Use of epigenome editing for better understanding of rare disorders? Acknowledgments Duke Duke Collaborators UNC Crawford lab Tim Reddy Patrick Sullivan Chris Frank Chris Vockley Paola Giusti Alexias Safi Tony D’Ippolito Terry Furey Lingyun Song Charlie Gersbach Jeremy Simon Lee Edsall Lauren Polstein William Wier Linda Hong Ami Kabadi Karen Mohlke **Gurkan Yardimci Pratiksha Thakore **Nathan Sheffield Isaac Hilton Northwestern **Yoichiro Shibata Ann Harris Greg Wray Duke Sequencing Core Courtney Babbitt Fangfei Ye Raluca Gordon Funding Support Olivier Fedrigo Alex Hartemink NHGRI Wendy Parris Anne West NIGMS Allison Ashley-Koch NIMH **Former students Melanie Garrett.