The three-dimensional folding of the -globin gene domain reveals - PowerPoint PPT Presentation

The three-dimensional folding of the α -globin gene domain reveals formation of chromatin globules. Davide Baù Structural Genomics Group Thursday, February 16, 12

� � � � � “Highlight 2011” The three-dimensional folding of the � -globin gene domain reveals formation of chromatin globules Davide Baù 1,4 , Amartya Sanyal 2,4 , Bryan R Lajoie 2,4 , Emidio Capriotti 1 , Meg Byron 3 , Jeanne B Lawrence 3 , Job Dekker 2 & Marc A Marti-Renom 1 Thursday, February 16, 12

Resolution Knowledge IDM INM DNA length 10 10 10 10 nt Volume 10 10 10 10 10 μ m Time 10 10 10 10 10 10 10 10 s Resolution 10 10 10 μ Adapted from: Langowski and Heermann. Semin Cell Dev Biol (2007) vol. 18 (5) pp. Thursday, February 16, 12

5C Technology Detecting up to millions of interactions in parallel http://my5C.umassmed.edu Dostie et al. Genome Res (2006) vol. 16 (10) pp. 1299-309 • 5C “copies” the 3C library into a 5C library containing only ligation junctions • Performed at high levels of multiplexing: •2,000 primers detect 1,000,000 unique interactions in 1 reaction Thursday, February 16, 12

Human α -globin Domain ENm008 genomic structure and environment ENCODE Consortium. Nature (2007) vol. 447 (7146) pp. 799-816 p13.3 13.2 12.3 p12.1 16p11.2 11.1 q11.2 q12.1 13 16q21 22.1 q23.1 0| 50000| 100000| 150000| 200000| 250000| 300000| 350000| 400000| 450000| 500000| LOC1001134368 RAB11FIP3 C16ORF35 ARHGDIG SNRNP25 RHBDF1 MRPL28 POLR3K LUC7L ITFG3 RGS11 PDIA2 AXIN1 TMEM8 DECR2 HB � 2 HB � 1 MPG HB � HB � HB � HS48 HS46 HS40 HS33 HS10 HS8 GM12878 CTCF K562 GM12878 diff RNA K562 GM12878 CTCF K562 GM12878 H3K4me3 K562 GM06990 DNaseI K562 The ENCODE data for ENm008 region was obtained from the UCSC Genome Browser tracks for: RefSeq annotated genes, Affymetrix/ CSHL expression data (Gingeras Group at Cold Spring Harbor), Duke/NHGRI DNaseI Hypersensitivity data (Crawford Group at Duke University), and Histone Modifications by Broad Institute ChIP-seq (Bernstein Group at Broad Institute of Harvard and MIT). Thursday, February 16, 12

� Human α -globin Domain ENm008 genomic structure and environment ENCODE Consortium. Nature (2007) vol. 447 (7146) pp. 799-816 p13.3 13.2 12.3 p12.1 16p11.2 11.1 q11.2 q12.1 13 16q21 22.1 q23.1 0| 50000| 100000| 150000| 200000| 250000| 300000| 350000| 400000| 450000| 500000| LOC1001134368 RAB11FIP3 C16ORF35 SNRNP25 ARHGDIG RHBDF1 MRPL28 POLR3K LUC7L ITFG3 RGS11 PDIA2 AXIN1 TMEM8 DECR2 HB � 2 HB � 1 MPG HB � HB � HB � HS48 HS46 HS40 HS33 HS10 HS8 GM12878 a b Reverse fragments Reverse fragments >1,000 >1,000 GM12878 cells 5C counts 750 K562 cells 5C counts 750 Forward fragments Forward fragments 500 500 250 250 0 0 GM12878 K562 Thursday, February 16, 12

Structure Determination Integrative Modeling Platform http://www.integrativemodeling.org Biomolecular structure determination 2D-NOESY data Chromosome structure determination 5C data Thursday, February 16, 12

Integrative modeling P1 P2 P1 P2 P1 P2 Thursday, February 16, 12

Representation Harmonic 2 ( ) 0 H i , j = k d i , j − d i , j i+1 Harmonic Lower Bound i i+2 $ 0 ; 2 ( ) 0 if d i , j ≤ d i , j lbH i , j = k d i , j − d i , j & % 0 ; & if d i , j > d i , j lbH i , j = 0 ' i+n Harmonic Upper Bound $ 0 ; 2 ( ) 0 if d i , j ≥ d i , j ubH i , j = k d i , j − d i , j & % 0 ; & if d i , j < d i , j ubH i , j = 0 ' Thursday, February 16, 12

Scoring GM1287 70 fragments 1,520 restraints Harmonic Upper Bound Harmonic Harmonic Lower Bound K562 70 fragments 1,049 restraints Thursday, February 16, 12

Optimization start CREATE PARTICLES 7.00E+06 ADD RESTRAINTS 6.00E+06 IMP Objective function SIMULATED ANEALING 5.00E+06 MONTE-CARLO 4.00E+06 LOCAL CONJUGATE GRADIENT 3.00E+06 5 steps 500 rounds 2.00E+06 1.00E+06 LOWEST OBJECTIVE FUNCTION 0.0E+00 0 50 100 150 200 250 300 350 400 450 500 Iteration end Thursday, February 16, 12

� � Not just one solution GM12878 K562 a b 2,800 915,980 500 244,000 Number of models in cluster Number of models in cluster Lowest IMP objective 2,700 Lowest IMP objective 400 function in cluster 914,555 function in cluster 241,500 2,600 300 2,500 913,130 237,000 200 2,400 911,705 234,500 100 2,300 910,280 2,200 0 231,000 0 1 2 3 4 5 0 50 100 150 200 250 300 350 400 Cluster number Cluster number Lowest IMP objective function in cluster 235,000 916,000 Lowest IMP objective function in cluster Cluster 10 234,450 914,500 168 models 234,747 OF Cluster 3 Cluster 4 2,282 models 2,270 models 915,890 OF 915,890 OF 233,800 913,000 Cluster 8 233,250 911,500 209 models 233,748 OF Cluster 7 226 models 233, 398 OF Cluster 3 Cluster 4 Cluster 5 Cluster 9 Cluster 6 232,600 275 models 910,000 265 models 256 models 205 models 228 models 233,110 OF 233,051 OF 233,026 OF 233,049 OF 232,819 OF Cluster 1 Cluster 2 483 models 314 models Cluster 2 232,698 OF Cluster 1 232,673 OF 2,668 models 2,780 models 910,400 OF 910,280 OF c Thursday, February 16, 12

� � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � Long-range interactions a b 100 nm 100 nm c c 1,000 800 Distance (nm) K562 600 400 GM12878 200 K562 0 GM12878 Fragment d e Thursday, February 16, 12

� � � � � � � � � � � � � � � � � � � � Chromatin globules Frequency contact map differences a b 100 nm 100 nm c Increased in GM12878 = Increased in K562 Thursday, February 16, 12

� � � � � � � � � � � � � � � � � � � � � � � � � � � � Genes location within the globules a b 100 nm 100 nm c b Promoters 2.5 2.5 Promoters Active genes Relative abundance Active genes Relative abundance CTCF sites 2.0 2.0 CTCF sites DNase I sites DNase I sites H3K4me3 sites 1.5 1.5 H3K4me3 sites Inactive genes Inactive genes 1.0 1.0 0.5 0.5 GM12878 K562 0 0 <50 <100 <150 <200 <250 <300 <350 <400 <50 <100 <150 <200 <250 <300 <350 <400 Increased in K562 Distance to center (nm) Distance to center (nm) 2.5 Thursday, February 16, 12

� � � � � � � � � � � � � � � � � � � � Model validation a b 100 nm 100 nm c GM12878 K562 500 GM12878 K562 400 Distance (nm) 300 200 100 0 FISH Models (2D) Thursday, February 16, 12

Summary 5C data results in consistent 3D models Thursday, February 16, 12

� � � � � � � � � � � � � � � � � � � � Summary Conformational changes correlate with gene expression a b 100 nm 100 nm LOC1001134368 c RAB11FIP3 C16ORF35 SNRNP25 ARHGDIG RHBDF1 MRPL28 POLR3K LUC7L ITFG3 RGS11 PDIA2 AXIN1 TMEM8 DECR2 HB � 2 HB � 1 MPG HB � HB � HB � HS48 HS46 HS40 HS33 HS10 HS8 GM12878 RNA diff K562 α -globin- Enhancer looping only in K562 Looping interaction In GM12878 cells Thursday, February 16, 12

The “chromatin globule” model a b Factory Eraf HBB PolII Münkel et al. JMB (1999) Osborne et al. Nat Genet (2004) Lieberman-Aiden et al. Science (2009) D. Baù et al. Nat Struct Mol Biol (2011) 18:107-14 A. Sanyal et al. Current Opinion in Cell Biology (2011) 23:325–33. Thursday, February 16, 12

OPEN POSITIONS IN THE LAB Starting early 2012 Acknowledgments Amartya Sanyal Emidio Capriotti Bryan R Lajoie Marc A. Marti-Renom Job Dekker Meg Byron Jeanne Lawrence Thursday, February 16, 12