Human Genome Structural Variation

1
Human Genome Structural Variation
Evan Eichler Howard Hughes Medical
Institute University of Washington
June 3rd, 2006, HGM, Helsinki
2
Structural Variation Disease

• Insertions, Deletions, Inversions, Duplications,
  Translocations
• Three different models of human genetic disease
  
• Rare Mendelian Diseaserare structural variants
  that segregate cause disease (eg. Parkinsons
  Alzheimer Disease)
• Recurrent Genomic Disordersrecurrent
  microdeletion or microduplication syndromes
  mediated by duplication (most de novo) (eg. PWS,
  CMT1A)
• Common Disease Susceptibilitycommon, copy-number
  polymorphisms that predispose to disease (Lpa
  coronary heart disease, FCGR3 lupus CC3L1
  AIDS susceptibility).

3
Genome-Wide Screens of Normal Variation
Large (100kb)
Small (7kb)

• 1503 variants, 115 Mb, 800 genes structurally
  variant
• Non-randomly distributed
• Environmental interaction/segmental duplications
• Normal individuals differ by thousands of events

Eichler (2006) Nat. Genet
http//humanparalogy.gs.washington.edu/structuralv
ariation
4
Structural Variation Disease Hypothesis

• Structural variants are common, more likely to
  be under selective constraint, more recurrent and
  associated with environmental interaction genes.
• We hypothesize that they will have a significant
  impact on rare and common disease in the
  population that can not be tracked by LD
  association mapping
• Approach Genomics Architecture ApproachTarget
  dynamic regions and then follow up with disease.
  

5
Objectives.

• Discovery of Novel Genomic Disorders

• Sequence-Based Resolution of Structural Variation

6
Model of Genomic Disease/Variation
A
B
C
TEL
A
B
C
TEL
Aberrant Recombination
GAMETES
Human Disease
Triplosensitive, Haploinsufficient and Imprinted
Genes

• Hypothesis Mechanism underlying Uncharacterized
  Mental Retardation?

7
Duplication Map of Human Genome

• 130 candidate regions (298 Mb)
• 23 associated with genetic disease
• Target patients array CGH

Bailey et al. (2002), Science2931003-1007
8
Array Comparative Genomic Hybridization
12 mm

• High-throughput detection of large-scale
  variation (gt50 kb),
• CNV (Copy-Number Variation)

9
Duplication Microarray Experimental Design
BACs
TEL
dist gt50 kblt5 Mb prop 95 identity, 10 kb

• 130 regions of the human genome
• 2178 BACs or on average 10-12 BACs per region
• Perform ArrayCGHreciprocal dye swap experiments
• Study Population
• Normal 269 samples 75 individual samples344
  total (284 unrelated 60 trios)
• Idiopathic Mental Retardation 291 probands
  (Flint cohort, negative for FraX normal
  karyotype)

10
Study Populations

• Normal unaffected (diversity panel and HapMap
  Samples). Target 800 samples, Completed 75
  269 samples344 total (284 unrelated 60 trios).

• Idiopathic Mental Retardation

11
Large-Scale CNPs among HapMap Samples
108 CNP novel BACs
149 CNP BACs previously identified
Altered CNV Frequency

• Total of 384 CNV BACs among HapMap DNAs (n263
  HapMap samples passed criteria)
• A total of 257 CNP BACs were identified (gt2
  individuals)
• 147 CNP observed previously in our study of
  sample of 55 or published by others
• 3.1 (9/257) were population-specific
• 127 singleton BAC observations

12
Validation using Nimblegen High-Density Oligo
Arrays
Log2 T/R
Deletion
Duplication

• n9, high density oligonucleotide array (385,000)
• Overall, validation for 194/257 (75.4 true
  positives).

Locke, Sharp, McCarroll et al., in press
13
Study Populations

• Normal unaffected (diversity panel and HapMap
  Samples). Target 800 samples, Completed 75
  269 samples344 totalIdentified additional 257
  CNPs.

• Idiopathic Mental Retardation
• Target 700 samples (300 samples Flint/Knight,
  400 CWRU samples) 291 complete

14
Evidence for Novel Recurrent Microdeletion
Syndromes from the Screening of 291 Mental
Retardation/MCA Cases
4 individuals with apparently identical
microdeletion never seen in gt300 normals
15
Refinement of the Breakpoints of 17q21.3
Microdeletion

• Customized oligonucleotide microarray (n11,000)

Seg Dups
IMR103 Father
1.0
0.0
IMR103 Mother
Log2 Relative Hyb intensity
-1.0
1.0
0.0
IMR103
-1.0
1.0
0.0
-1.0
(1.5 STDEV)
16
17q21.3 Microdeletion is Recurrent.
Normal
Normal
Affected IMR103
Affected IMR253
Affected IMR255
Affected IMR376
Normal
Normal

• -

• 4/291 patients , estimate 1 of mental
  retardation

17
Clinical Phenotypes of 17q21.3 Microdeletion
Patients
IMR253-moderate dev. delay, sparse eyebrows,
protruding tongue, low-set lop ears, large nose,
seizures (4 yrs), markedly blonde hair
IMR669-delayed speech, hypotonia, mild
dysmorphic features, seizures fits until 16
months, upslanting palpebral fissure prominent
philtrum, bulbous nose, fair
IMR376-severe learning difficulties, markedly
hypotonia, hypopigmented, Mongolian slant, pale
blue, almond shaped eyes, protruding tongue,
extensible joints
IMR669
18
(No Transcript)
19
1q21.1 Duplication-Mediated Microdeletion
IMR43
Father
Mother
2.4 Mb
20
15q24.1-24.2 Duplication-Mediated Microdeletion
IMR349
Father
Mother
4.7 Mb
21
Variation in IMR

• 291 IMR samples (Oxford Cohort) screened to date

• 23 (n31) novel sites of variation defined by gt2
  BACs

• 5 are seen in more
• than one unrelated patient

• 7/9 events are de novo
• Novel Genomic Disorder Candidates

22
Objectives.

• Discovery of Novel Genomic Disorders

• Sequence-Based Resolution of Structural Variation

23
Intermediate-Size Structural Variation (ISV) and
Inversions
Gene
Phenotype
Size
Locus
Freq.
Type
Dup
GSTT1
halothane/epoxide sensitivity
17kb/94
54.3 kb
22q11.2
20 -/-
Deletion
DEF3A-OR
heart disease susceptibility
5 Mb
8p23
26 -/
Inversion
400kb/98.9
EMD/FLN
none
219 kb
Xq28
33 -/
Inversion
48kb/99
HERC2
susceptibility to Angelman syndrome
3 Mb
15q11.2
4 -/
Inversion
gt300kb/99.8
IGVH26
immune response variation
Variable
14q32.3
4-15 /-
Deletion/Dup
91-97
toxin resistance, cancer susceptibility
GSTM1
18 kb
1p13.3
50 -/-
Deletion
24kb/95.6
CYP2D6
1-29
Antidepressant drug resistance
5 kb
22q13.1
Duplication
5.4kb/91-97
CYP21A2
Congenital adrenal hyperplasia
35 kb
6p21.3
1.6 /-
Duplication
0
CYP2A6
nicotine metabolism
7 kb
19q13.2
1.3 /-
Duplication
24kb/96.2
SMN2
SMA susceptibility
88.7/99.8
gt100 kb
5q13
50 /-
Duplication
Adapted from Buckland, Ann Med
24
Genome-wide Detection of Structural Variation
(gt8kb) by Fosmid End-Sequence Pairs

• Identified 295 potential candidates, Deletions,
  Insertions Inversions
• enriched among environmental interaction genes
  

Tuzun et al., 2005
25
Fine-Scale Structural Variation Map (build35
vs. Fosmids)

• 1.3 Discordant Fosmids
• Identify 295 clusters (2 or more)
• 246 supported by second haplotype
• 147 inserts, 93 deletions, 57 inverts
• 18 putative L1 events10 deletions
• and 8 insertions (6 kb insertion)
• 89 locate within gene regions.
• 138 unique regions of the genome
• 159 duplicated regions of the genome

Insertion(Fosmid)
Deletion
Inversions
Heterochromatic regions
Duplicated regions
26
Sequenced Structural Variation of APOBEC3B

• 24.5 kb deletion eliminates most of APOBEC3B but
  creates fusion gene
• Fusion APOBEC3A/3B lt1 frequency Africans, gt35
  Papua New Guineans

27
Sequenced Structural Variation of DEFA1
gt25 kb duplication DEFA1, additional exon 3
28
Sequenced Structural Variation of LCE1E/D
Oligo ArrayCGH
Log2 Relative Signal Intensity

• 9.2 kb deletion of LCE1D gene creates a fusion
  gene LCE1D/E
• confirmed in 3 unrelated individuals by
  oligonucleotide microarray technology

http//humanparalogy.gs.washington.edu/structuralv
ariation
29
Sequencing Genic Structural Variation
30
PCR Breakpoint Genotyping Assays for Structural
Variation

• Tested 11 structural variants (5 insertions, 4
  deletions, and 2 inversions)
• 7 successful assays (6 gt20 minor allele
  frequency)

31
Genotyping Illumina Golden-Gate Assays for
Binary Events
Newman et al., 2006
32

• A Human Genome Structural Variation Initiative
• 2 scientific meetings (2005)
• 2 working groups (AHG, MSWG (12/05)
• Coordinating Committee (1/06)
• NIH Council (2/06)
• Press Release (3/15/06)
• Goal Complete Characterization
• of Structural Variation in
• 48 HapMap Samples

CEPH
33
A Structural Variation Map of the Human Genome
(lt1 Mb)
ABC9 Japanese
ABC8 African
ABC7 African
G248 Hispanic
Insertion(Fosmid)
Deletion
Inversions
Heterochromatic regions
Duplicated regions
Putative Structural Variants from Four Individuals
34
(No Transcript)
35
Summary

• Genomic architecture approach systematically
  identify
• dynamic regions of structural variation
  gtgtgtgtphenotype

• Large-Scale Variation
• Normals Identified 257 CNPs using a targeted
• microarray to duplicated regions

• IMR Identified 31 sites (gt2 BACs) unique to
  patients
• (n291 probands) (5 are recurrent and 3 are
  confirmed de novo)
• Goal Discovery of Novel Genomic Disorders

• Fine-Scale Variation Developed an approach to
  map and
• sequence common fine-scale variation within the
  human
• 1000 differences gt 8 kb from 4 individuals.
• Goal Disease association with common
  disease/susceptibility

36
Acknowledgements
Nimblegen Rebecca Selzer Peggy Eis
MIT Steve McCarroll David Altshuler
Eichler Lab Andy Sharp Devin Locke Sierra
Hansen Sean McGrath Eray Tuzun Matthew
Johnson Zhaoshi Jiang Jon Bleyhl Tera Newman Jeff
Bailey Anne Morrison Lisa Pertz Ze Cheng Xinwei
She James Sprague
UCSF Dan Pinkel Donna Albertson
UWGSC Maynard Olson Rajinder Kaul Hillary
Hayden Eric Haugen
CWRU/UChicago Stuart Schwartz Laurie Christ
Agencourt Doug Smith
Oxford Jonathan Flint Samantha Knight
NHGRI Jim Mullikin AHG/MSWG
UW Mark Rieder Debbie Nickerson
37
Capturing Smaller Variants (lt8 kb)
a)
Select gt2 STDEV Clones
96 clones
Fingerprint MCD
Select 32 variant clones
Sequence
38
The Missing Human Genome
39
Finding Novel Human Sequence
40
Fosmid Pairs that fail to Map to build35

• 1573 fosmid paired-end sequences fail to map to
  build 35.
• 644 have 150 bp gtQ30 at either end and have gt100
  bp unique seq
• 565 of these have no hit to HTGS BAC sequence
• Four independent restriction enzymes (EcoR I,
  Hind III, Bgl II and Nsi I )
• 26 Contigs (constructed from Composite Mutual
  Overlap Statistic (CMOS) and 94 singletons
• Range in size from 208 kb to 38 kb (based on
  fingerprint data).
• Do they represent human sequence?

41
FISH Summary of Orphan Fosmids

• 119 contigs/singletons tested by FISH
• 22 contigs (1,296 kb) acro, 12 contigs (458 kb)
  pericentromeric,
• 48 contigs (2,577 kb) subtelomeric
• 32 interstitial euchromatin (9 corresponding to
  known gaps)

42
Hybridization
2
R921
1.5
1
0.5
0
-0.5
-1
-1.5
D3767
1.5
1-3
5
10
15
20
4-5
1
6
0.5
Log2 Hybridization Relative Intensity Test/Referen
ce
7-14
0
15
-0.5
16-20
-1
-1.5
1.5
0
5
10
15
20
R1080
1
0.5
0
-0.5
-1
-1.5
-2
0
5
10
15
20
BAC Probes
43
Genomic Variation
Forms of genetic variation.
Sequence

• Single base-pair changes point mutations
• Small insertions/deletions frameshift,
  microsatellite, minisatellite
• Mobile elementsretroelement insertions (300bp
  -10 kb in size)
• Large-scale genomic variation (gt10 kb)
• Large-scale Deletions
• Segmental Duplications
• Chromosomal variationtranslocations, inversions,
  fusions.

Cytogenetics