New Frontier for the Genetic Study of Common Disease: Comprehensive Association Analysis

1
New Frontier for the Genetic Study of Common
Disease Comprehensive Association Analysis

• Jianjun Liu, PhD
• Genome Institute of Singapore

2
Outlines

• Infrastructures for genetic study at GIS
• Overview of the Human Genetics Program at GIS
• Our strategy for Genetic Study induction of
  several projects

3
SNP Genotyping Platforms At GIS

• Available, but rarely used
• ABI3730 (SNPlex, high thpt. OLA assays)
• ABI7900HT (Taqman and related assays)
• Why so many platforms? (throughput, price, assay
  conversion, quantitative sensitivity)

4
(No Transcript)
5
(No Transcript)
6
(No Transcript)
7
Resources For Supporting Genotyping Analysis

• Computational Support
• LIMS and (multiple) Databases
• Data Storage
• Manpower Support
• Industrial operation central facility
• A dedicated team for production

8
Support for Statistic Analysis

• Current support
• Access to multiple commercial software packages
  for single SNP analysis Varia, Helixtree and
  Exampler
• In-house algorithm development for haplotype
  analysis
• New recruitment

9
Additional Resources

• Singapore Tissue Network
• 2 Autopure automated DNA extractor
• Cell culturing capacity
• Sample collection, extraction, archiving and
  distribution with thorough QC
• Secure TTP delinking
• Rigorous attention to ethical standards

10
Human Genetics Program at GIS

• Infectious Diseases (Martin Hibberd) pneumonia,
  dengue
• Autoimmune Disorders (Mark Seielstad) RA, TB,
  Psoriasis and Vaccine response
• Cancer Genetics (JJ Liu) breast cancer, and
  other cancers through collaboration with
  Singapore Cancer Syndicate
• Neurological and Psychiatric Disorders (JJ Liu)
  Parkinsons disease, schizophrenia and stroke

11
Case I Genetic Risk Factors for Breast
CancerHypothesis-driven Study
(Balmain, Gray and Ponder, Nature Genetics, 2003)
Funded by Susan G Komen Foundation and National
Cancer Institute, NIH
12
Large Population-based Sample
1500 breast cancers
1500 controls
13
Comprehensive Analysis of Candidate Gene
Pathway-based Analysis
Estrogen
Working Hypothesis Polymorphisms in
estrogen-related pathways can influence breast
cancer risk by modifying either exposure to
estrogen or cellular response to the exposure.
14
Comprehensive Coverage of Candidate Gene
Pathway-based Analysis
Estrogen
15
ER-mediated Expression Response Machinery
ER Binding Sites
ER Co-factors
16
Comprehensive Coverage of Candidate Gene
Pathway-based Analysis
Estrogen
17
Identification of Candidate ER Direct Target
Genes By In-vitro Expression Profile Analysis
E2ICI
E2
E2CHX
Time
Up-regulated
Down-regulated
18
Identification of Candidate ER-associated Genes
By In-vivo Expression Profile Analysis
4 breast cancer cohorts on Affymetrix U133AB
platform
Uppsala (260)
Oxford (227)
Stockholm (159)
Spore (100)
ER
ER-
ER
ER-
ER
ER-
ER
ER-
19
Comprehensive Coverage of Candidate Gene
Pathway-based Analysis
Estrogen
20
Breast Cancer Association Consortium (BCAC)
20 Groups Around the World
21
BCAC 9 SNPs and breast cancer risk (Nature
Genetics, in press)
Gene SNP cases P-trend odds
ratio (95 CI) ADH1B 3UTRagtg 11,391 0.54
0.99 (0.95, 1.03) CASP8 D302H 16,423 1.1x10-7
0.88 (0.84, 0.92) CDKN1A S31R 18,290 0.28 1.03
(0.98, 1.09) ICAM5 V301I 17,687 0.78 1.00
(0.97, 1.03) IGFBP3 -202cgta 13,101 0.046 0.97
(0.94, 1.00) SOD2 V16A 16,273 0.31 0.98 (0.96,
1.01) TGFB1 L10P 12,493 0.00005 1.07 (1.04,
1.11) ATM S49C 15,905 0.08 1.13 (0.99,
1.30) NUMA1 A794G 14,642 0.52 1.03 (0.94, 1.14)
22
(No Transcript)
23
(No Transcript)
24
What Can We Learn From BCAC?

• Large-scale collaboration is feasible
• A very large sample is needed for identifying
  risk alleles with moderate effect

25
Case II Genetic Biomarker For Prognosis
Whole-genome Approach

• 804 breast cancer cases chosen based on
• Complete information on
• Tumour size
• Histological grade
• Lymph node involvement
• Survival (200 fatal cases after 10 yrs
  follow-up)
• Sufficient DNA concentration
• Aim to uncover genetic variants related to breast
  cancer prognosis

Nottingham Prognostic Index
26
Why use whole genome approach to study prognosis?

• Few studies are able to study prognosis due to
  lack of follow-up information
• Comprehensive follow-up information is a unique
  strength of our study
• Lesser is known about genes affecting prognosis
  compared to breast cancer incidence
• Using whole genome scan to study prognosis may
  uncover more new knowledge
• More clinically relevant

27
Case III Genetic Analysis of TF Binding SitesA
Functional Approach

• Will SNPs within motifs and surrounding sequences
  influence binding and expression regulation?
• Will SNPs within TF binding region influence
  disease susceptibility?

28
SNP Analysis of p53 Binding Sites

• Identification of SNPs within p53 binding sites
• 1376 putative p53 binding sites with motif
• SNP identification by mining the dbSNP database
• Functional Analysis impact on
• Binding Activity
• Gene Expression Regulation
• Protein Expression

29
A Motif SNP (rs1860746) in PRKAG2
Binding Site Location internal intron of PRKAG2
30
Genotype Influences p53 Binding Target Gene
Expression Regulation
CHIP Analysis
Real-time PCR Analysis
31
Validation Analysis Genotype Influences Binding
Activity
Motifs with G allele (both homo hetero) have
strong binding activity than motifs only with T
allele (analysis of 8 cell lines)
TT
GT
GG
32
Validation Analysis Genotype influences the
down-regulation of target gene expression
rs1860746_PRKAG2

• 13 cell lines with different
• genotypes at rs1860746
• 10 hrs drug treatment
• vs. no drug treatment
• Triplicate real-time PCR
• analysis for each cell line

33
Genotypes Down-Regulation Effect At Protein Level
AICAR/5FU (hrs)
0
8
24
48
0
8
24
48
p53
Total AMPK
Phospho AMPK (Thr172)
Actin
T/T (Mutant Motif)
G/G (wild-type motif)
34
At GIS, We have

• A strong capacity for large-scale genetic study
• An active program in the genetic study of common
  disorders
• A diverse approach for genetic analysis
• Pathway analysis
• Whole-genome
• Functional analysis