Chromosome Conformation Capture, Looping, and Mechanisms of Bglobin Regulation

1
Chromosome Conformation Capture, Looping, and
Mechanisms of B-globin Regulation

• C. Titus Brown
• Transcr Reg J-club
• Jul 11, 2008

2
My research?

• http//ged.msu.edu/
• Combine computational and experimental
  approaches, esp. regulatory genomics, to build
  gene regulatory networks.

3
(No Transcript)
4
(No Transcript)
5
Introducing the B-globin locus
Six DNAse hypersensitive regions (HS1-6) a.k.a
Locus Control Region, or LCR Two embryonic
genes (in black) Two adult genes (in red) Bounded
by olfactory receptor genes (OR)
6
CCC
Corrects for - PCR efficiency - ligation
efficiency
7
Linear conformation (non-expressing cells)
8
HS regions/LCR are associated with adult genes
during expression
9
HS regions/LCR are associated with adult genes
during expression, in reverse
10
Inactive genes arent associated with LCR
11
Its a party! All HS sites associated with each
other.
12
Conclusions, thus far

• When B-globin adult genes are activated in liver
  tissue, LCR and other hypersensitive regions are
  looped into proximity with the active genes.
• No looping occurs with the inactive genes.
• All HS regions are in proximity to each other,
  too (i.e. in the same cell).

13
Further questions

• How is the looping mediated?
• Is this looping regulated co-incident with gene
  activation?
• What do I mean by gene activation -- promoter
  occupancy, or elongation?

14
Experimental setup

• GATA-1 and FOG-1 are required for B-major
  activation at both LCR and promoter.
• Use GATA-ER (inducible) and various mutations to
  characterize presence of factors, looping, and
  function.

15
Answers

• How is the looping mediated?

No new protein synthesis reqd
Interaction with FOG-1 reqd
16

• How is the looping mediated? (2)

LCR is not required for GATA-1 presence. Nor
is the LCR required for FOG-1 presence.
17
Answers

• Is this looping regulated co-incident with gene
  activation?
• Yes (using GATA-1 presence as a proxy for gene
  activation)
• but levels dont correlate.

18
Answers

• What do I mean by gene activation -- promoter
  occupancy, or elongation?

(both)
19
Conclusions, so far

• GATA-1 and FOG-1 bind to HS (not shown) and
  promoter, and then cause looping.
• Consonant with a view where looping is an
  architectural modification necessary to bring
  trans-factors in cis.

20
4C (CCC on chip)
21
Interactions with B-globin HS2 (chr 7, 8, and 14)
22
Active inactive loci still show looping
23
Active inactive loci still show looping
24
Whole-chromosome map of interactions
Interactions have tissue-specific
patterns. Interacting regions avg 150-200kb in
size.
25
Active/active, inactive/inactvive
Liver (active) Brain (inactive)
26
Housekeeping gene
Stereotyped pattern in both tissues (i.e.
previous result is specific to genes, not tissue.)
27
Summary points

• Active and inactive genes both display long-range
  intra- and inter-chromosomal contacts (inactivity
  is not boring).
• Even boring genes show this (not shown).
• No evidence for intrachromosomal clustering of
  tissue-specific genes.
• Some evidence for spatial clustering of highly
  expressed genes (active gene-dense areas).
• 4C results largely correlated with FISH-detected
  interactions (not shown).

28
Summary points, simplified

• Its not a simple situation.
• More generally, its not clear what all of these
  intra- and inter-chromosomal contacts really
  mean Gene expression factories? Gene regulation
  complexes? How specific, how necessary?

29
Question/comment/thought 1

• Early chromatin structure is thought to be very
  different than adult chromatin structure. What
  happens if you look in an early embryo?

30
Question/comment/thought 2

• Thorough high-throughput sequencing (454 or
  paired-end Solexa) would be able to sequence all
  interacting regions, with no need for inverse
  PCR. Volunteers?

31
Question/comment/thought 3

• It would be interesting to correlate whole-genome
  contact information with (evolutionary) synteny
  and syntenyorder across (say) the vertebrates.
• Note that it is thought that syntenic ordered
  blocks are conserved because of the need to
  preserve long range interactions but does
  distance really matter in the nucleus?

32
Question/comment/thought 4

• What do medium-range (10-200kb) enhancer/promoter
  contacts look like?? What if we look at
  microstructure?
• Can you use 3C to validate known enhancers?
• Can you use modified 3C to fish out or enrich for
  likely distal enhancers? perhaps in combination
  with conservation?