Detection and Characterization of
Gene Conversion in Mouse and
Human Recombination Hotspots
Yves Clément
IMPRS Colloquium
25/11/2011
Why is meiotic recombination
important? (1)
1 diploid cell
Recombination
4 haploid cells
(sperm, egg)
Why is meiotic recombination
important? (2)
A
B
a
b
A
b
a
B
Meiotic recombination
5’
3’
paternal
3’
chromosome
DNA repair and
5’
synthesis
maternal
chromosome
Gene conversion
Double-strand break
5’ to 3’ resection
Synthesis of new DNA
Invasion of DNA with
sequence similarity
Holliday junction
formation
[Duret & Galtier 2009]
Questions
• What is the length distribution of gene
conversion tracts?
• Can we detect other characteristics of meiotic
recombination?
GC-Biased Gene Conversion (gBGC)
Fixation bias
G
T
mismatch
A
T
W
G
C
S
Biased mismatch repair
[Meunier & Duret, 2004]
[Duret & Arndt, 2008]
[Smagulova et al, 2011]
Mus m. musculus
DSB hotspot
Mus m. castaneus
[Smagulova et al, 2011]
Methods
DSB “hotspots”
middle point
100 bp
Mus m. musuclus
Mus m. castaneus
Mus spretus
[Smagulova et al, 2011]
[Keane et al, 2011]
Inferring substitution rates
• Estimation of nucleotide substitution frequencies
in lineages from multiple alignments:
C
• on each branch:
M1
C
T
Mus m. musculus
M2
M4
M3
C
C
Mus m. castaneus Mus spretus
– 12 XY substitutions
– 2 CpG methylation
deamination processes:
CpG  CpA / TpG
– GC* (equilibrium or
“future” GC-content)
[Duret & Arndt, 2008]
≈ 1.5 kbp
Control (1)
DSB “coldspots”
DSB “hotspots”
Mus m. musculus
Control (2)
DSB “hotspots”
Mus m. musculus
Mus m. castaneus
Gene conversion (1)
• Centered on DSB hotspots middle points.
• Affects region of approximately 1.5 kbp
• Gene conversion tracts have variable length
Base composition skews
[Smagulova et al, 2011]
Strand specific mutations
G-T
C-G
G-C
C-G
G-T
C-A
G-C
C-G
G-C
T-G
A-C
T-G
GA ≠= CT
5’
AAAAA
3’
TTTTT
a
5’
5’
TTTTT
a
AAAAA
3’
AAAAA
TTTTT
a
3’
AAAAA
TTTTT
a
3’
5’
5’
GA
CT
3’
3’
CT
GA
5’
5’
GA
CT
3’
5’
GA
CT
3’
Gene conversion (2)
• No evidence of strand specific mutations
caused by meiotic recombination.
?
✔
DSB Hotspots
 DSB Hotspots
✔ Recombination Hotspots
PRDM9
genome
markers
CCNCCNTNNCCNC
PRDM9
middle points
Gene conversion in human
• Gene conversion centered around PRDM9
binding sites:
– DSB occur in very close proximity to PRDM9
binding sites.
• PRDM9 binding sites directly affected by gene
conversion
– “hotspot paradox”
Acknowledgements
•
•
•
•
•
Peter F Arndt
Brian Cusack
Barbara Wilhelm
Laurent Duret
Martin Vingron
• Thank you for your attention!
Recombination hotspots
recombination rate
marker
1 Hotspot
Recombination >= 10 cM/Mb  Hotspot
genome
5’
3’
CCNCCNTNNCCNC
3’
CCNCCNTNNCCNC
CCNCCNTNNCCNC
5’
CCNCCNTNNCCNC
5’
3’
3’
5’
CCNCCNTNNCCNC
5’
3’
100bp
0
Substitution patterns…
Inferring substitution rates
• Estimation of nucleotide substitution frequencies
in lineages from multiple alignments:
C
• on each branch:
M1
C
T
human
M2
M4
M3
C
chimp
C
gorilla
– 12 XY substitutions
– 2 CpG methylation
deamination processes: CpG 
CpA / TpG
– GC* (equilibrium or “future”
GC-content)
[Duret & Arndt, 2008]
≈ 1.5 kbp
Gene conversion tract regulation?
Loose Regulation
GC*
GC*
Tight regulation
middle points
middle points