Chromatin organization in D' melanogaster

1
Chromatin organization in D. melanogaster

• Peter Kharchenko
• Park Lab
• Harvard Medical School

2
ENCODE and modENCODE
(Encyclopedia of DNA Elements)

• RNA
• Replication
• Transcription Factors
• Chromatin

• Common cell types,conditions, protocols

Nature June09
3
Goals of the chromatin group

• Genome profiles
• Histone modifications
• Chromatin binding proteins
• Chromatin modifying enzymes
• Insulator proteins
• Antibodies
• Validation
• Western, IF, RNAi, mass spec.
• Production
• 50 new Abs with SDI
• Community Resource
• www.modencode.org
• 9 month publication embargo

• Cell types
• S2, BG3, Kc
• Embryo (2-4hr, 14-16hr)
• Larvae, Heads
• Platforms
• Affymetrix 2.0 tiling array
• Solexa sequencing

4
modENCODE Fly chromatin group
Wet Work

• Gary Karpen LBNL and UC Berkeley
• Aki Minoda, Cameron Kennedy
• Sarah Elgin Washington University
• Nicole Riddle, Sarah Gadel, Sarah Marchetti
• Mitzi Kuroda Harvard Medical School
• Art Alekseyenko, Andrey Gorchakov
• Vince Pirrotta Rutgers University
• Yuri Schwartz, Daniela Linder-Basso, Greg
  Shanower
• Peter Park Harvard Medical School
• Michael Tolstorukov, Peter Kharchenko

Bioinformatics
5
Correlation of chromatin marks

• Cluster ofactive marks
• H3K36me3
• PcG
• Core histones
• H3K9me2and HP1

S2 cell data
6
Euchromatin and Heterochromatin

• Euchromatin
• Gene-rich
• Early replicating
• Recombines
• Chromosome arms

• Heterochromatin
• Repeat-rich
• Late replicating
• Lack of recombination
• Centromeres/telomeres

Transcriptional activators Hyper-acetylated
histone tail, H3K4 methylation
HP1 complex Hypo-acetylated histone tail, H3K9
methylation
7
Heterochromatin in D. melanogaster

• How do we define a region as heterochromatic ?
• Reporter Assays
• Biochemical Marks
• BAC mapping of chromatin

• Functions
• Centromere identity
• Genome stability
• Transposon silencing

HP1
H3K9me2
merge
red
variegating
8
Average chromatin states
BG3 cell data

• high HP1
• high H3K9me2, me3
• Depletion of H3K23ac
• Broad depletion of open chromatin marks
• Lack of H3K27me3
• Transcriptionally silent chromatin repressed
  through HP1 mechanism

9
Chromatin states in heterochromatin

• K-means clustering on PCA-normalized 500bp block
  enrichment matrix

BG3 cell data
10
Chromatin states of genes

• Reduced chromatin summary of each gene
• Similar pattern of active marks
• PcG silencing
• Mixed chromatin state

11
Transcribed heterochromatic genes

• Pronounced depletion of H3K9me2 / me3
• Depletion of HP1 downstream of TSS
• Almost absent in the S2 cells
• HP1 peak 800bp upstream

12
Heterochromatic boundaries
H3K9me2

• Cell-type specificity ofpericentromeric
  boundaries

shifts in the pericentromeric boundaries
13
Heterochromatic domains

• Base state outlined by Embryo patterns
• Cell-type specific extensions
• Large domains of enrichment in chromosome arms

14
Common euchromatic domains
H3K9me2

• Enriched for H3K9me2across S2, BG3 andfly
  samples
• Transcriptionallysilent
• 28Kbp median size
• Specific to chrX
• 86 on chrX
• Most are not observedin female cell line

chrX
15
Vignette 1 Insulator proteins

• What defines domainboundaries?
• Blocks of sequence
• Specific proteins
• Insulators
• Enhancer blocking
• Local remodeling
• Loop formation

Modified from Kuhn and Geyer. COCB, 2003,
15259265
16
Vignette 1 Insulator proteins

• CTCF, CP190, Zw5, Su(hw), mod2.2, Egg
• ChIP-seq
• Sequencespecificity
• Bindinggroups
• Orientation
• Focus onPcG silencing
• Most bindingoutside ofdomains

CP190
CTCF (standalone)
CP190 (standalone)
17
Insulators Combinatorial binding
Pc

• Functionalvalidation

18
Heterochromatic domains

• Base state outlined by Embryo patterns
• Cell-type specific extensions
• Large domains of enrichment in chromosome arms

19
S2-specific H3K9me2 domains

• Small (10Kbp) regions, with transcribed genes
• Divergentpromoters
• Low HP1

20
Heterochromatic domains

• Base state outlined by Embryo patterns
• Cell-type specific extensions
• Large domains of enrichment in chromosome arms

21
BG3-specific H3K9me2 domains
HP1

• Large (100Kbp)
• Transcriptionallysilent
• Sensory perceptionfunctions (P-value 6x10-19)
• At least someappear to be dueto transpositions

H3K9me2
HP1
H3K9me2
22
Vignette 2 Integrative analysis Chromatin
state, expression and ploidy
S2 chr2L

• CGH MacAlpine group
• RNA-seq - Celniker group
• Matching up with ploidy
• S2 and Bg3 cell lines
• CGH clusters
• Max-exon RNA-seqestimates
• Mean/median levelswithin clusters
• Expression differencesin non-overlappingregions
  of abnormalploidy
• Simple trending

Log2(S2 expr) Log2(Bg3 expr)
Bg3 CGH cluster
S2 CGH cluster
23
Chromatin state, expression and ploidy

• Predicting expression from chromatin state
  regression models

• Expression level of over-replicated genes is
  higher than that expected from the observed
  chromatin states
• Conversely the expression level of
  under-replicated genes is lower than expected
• Average bias is small (1.1-1.2 fold difference)
• Model limitations
• Enrichment differences

24
Summary points

• modENCODE project RNA, chromatin, chromosomal
  proteins for the same cell types
• Integrative analysis
• Heterochromatin not all of it is silent
• HP1-mediated repression outside of
  heterochromatic regions
• Beware of rearrangements and CNVs in cell lines

25
Acknowledgements

• Peter Park HMS
• Dan Day
• Michael Tolstorukov
• Mitzi Kuroda HMS
• Art Alekseyenko, Andrey Gorchakov

• Gary Karpen LBNL/UCal
• Aki Minoda, Cameron Kennedy
• Sarah Elgin WashU
• Nicole Riddle, Sarah Gadel, Sarah Marchetti
• Vince Pirrotta Rutgers
• Yuri Schwartz, Daniela Linder-Basso, Greg
  Shanower
• David MacAlpine - Duke
• ENCODE/modENCODE
• NHGRI, NIH

Looking for postdocs!
http//www.modencode.org
26
(No Transcript)
27
Histone modification profiles
Average Gene Profiles
Expression
Metagene
Gene Regions
Quintiles
Heterochromatic Genes
TSS
TTS
Gene Size
Domains
Pc
Chromosomes
28
Repetitive elements

• Estimate average enrichment of pre-defined repeat
  groups
• Histone methylations inmouse and human cell
  lines
• Distinct chromatin states associated with
  knownrepeat families
• Different mechanisms of ERV silencing (HP1 and
  PcG)

mES cells
manuscript in preparation
29
Repetitive elements

• Evaluate enrichment on a phylogenythat is
  optimal for the given dataset
• Similar repeat types
• Instances belonging to thesame repeat type
• Chromatin states
• modEncode
• Copy number variations
• tumor samples

H3K9me3 in mES
manuscript in preparation
30
Genome-wide clustering