Transcript The Genetics of Addiction

Convergence of Genetic Findings for Nicotine
Dependence, Lung Cancer and COPD
Laura Jean Bierut, MD
Washington University
Financial Disclosure
• Patent on genetic variants that predict
addiction – “Markers of Addiction”.
• Consultant for Pfizer in 2008 for genetic
studies for smoking cessation.
• Funding of studies is through the National
Institutes of Health
Genetic Studies of Complex Diseases
A Retelling of the Emperor’s New Clothes
Laura Jean Bierut, MD
Washington University
Table of Contents
Chapter 1: What is the utility of linkage analysis in complex diseases?
Chapter 2: How to interpret all the previous genetic findings?
Chapter 3: What have we learned from Genome Wide Association
Studies of schizophrenia, bipolar disorder, depression,
alcoholism and autism? Do we have any findings?
Chapter 4: What is the best phenotype to study?
Chapter 5: What does gene environment interaction really mean?
Chapter 6: What is the power to detect gene environment interaction?
Chapter 7: Should we move into studying diverse populations?
Chapter 8: Don’t get me started
Chapter 9: The Happy Ending
Prologue
Model of Nicotine Dependence A many step process
Never Use
Initiation
First puff – First cigarette
Does everyone who
uses nicotine become
addicted?
Smoker
100 cigarettes lifetime
Nicotine Dependence
U.S. Population Screening and
Nicotine Dependence
No
Symptoms
3,051
Screened
53,742
50.9%
19.2%
Initiated
Smoking
27,372
58.0%
Smoked 100+
Cigarettes
15,881
35.2%
Some
Symptoms
5,596
44.3%
Collaborative Genetic Study of Nicotine Dependence
Nicotine
Dependence
7,028
Novel Gene in Dependence
• a5-a3-b4 nicotinic receptor gene cluster is
involved in the development of nicotine
dependence.
• How did we get there?
NICSNP Project
NICSNP is a large scale genome wide association study and
candidate gene study of nicotine dependence.
• Collaborative Genetic Study of Nicotine Dependence
Principal Investigator: Laura Jean Bierut (P01 CA 089392)
• The Genetics of Vulnerability to Nicotine Addiction
Principal Investigator: Pamela Madden (R01 DA 012854)
• Genes for Smoking in Related and Unrelated Individuals
Principal Investigator: Ovide Pomerleau (R01 DA 017640)
• Pharmacokinetics of Nicotine in Twins
Principal Investigator: Gary Swan (R01 DA 011170)
NIDA Phenotypic Repository
John Rice
Perlegen Sciences
Dennis Ballinger
Fagerström Test for Nicotine Dependence
Questions
Response Options
Score
Within 5 minutes
6-30 minutes
31-60 minutes
After 60 minutes
3
2
1
0
Yes
No
1
0
The first one in the morning
All others
1
0
4. How many cigarettes per day do you smoke?
10 or less
11-20
21-30
31 or more
0
1
2
3
5. Do you smoke more frequently during the first
hours after waking than during the rest of the
day?
Yes
No
1
0
6. Do you smoke if you are so ill that you are in bed
most of the day?
Yes
No
1
0
1. How soon after you wake up do you smoke your
first cigarette?
2. Do you find it difficult to refrain from smoking in
places where it is forbidden, e.g., in church, at the
library, in cinema, etc.?
3.
Which cigarette would you hate most to give up?
Heatherton TF, Kozlowski LT, Frecker RC, Fagerström KO. (1991). The Fagerstrom Test For Nicotine Dependence: A revision of the Fagerström Tolerance
Questionnaire. British Journal of Addiction 86:1119-1127.
Case and Control Phenotype Definition
• Case: Nicotine dependent defined by a
Fagerström Test for Nicotine Dependence
(FTND) > 4
• Control: Individual who has smoked 100 or
more cigarettes and never had any symptoms
of nicotine dependence (Lifetime FTND = 0).
Heatherton et al., 1991
Results from Candidate Gene Study
Saccone et al., 2007
Results from Candidate Gene Study
Saccone et al., 2007
SNPs highly correlated with rs16969968
Findings for Nicotine Dependence
rs16969968
Saccone et al., 2007
SNPs highly correlated with rs16969968
Findings for Nicotine Dependence
rs16969968
Saccone et al., 2007
rs1317286
Bierut et al., 2008
Berrettini et al., 2008
Sherva et al., 2008
Weiss et al., 2008
Stevens et al., 2008
rs1051730
Saccone et al., 2007
Thorgeirsson et al., 2008
Amos et al., 2008
Spitz et al., 2008
Results from Candidate Gene Study
The correlation between rs16969968 and rs578776 is < 0.2.
There are two distinct findings in the nicotinic gene cluster
associated with nicotine dependence.
Saccone et al., 2007
Genetic Association and the
Nicotinic Receptors - Chromosome 15
rs578776
Saccone et al., 2007
Genetic Association and the
Nicotinic Receptors - Chromosome 15
rs578776
Saccone et al., 2007
Bierut et al., 2008
Weiss et al., 2008
Stevens et al., 2008
rs6495308
Berrettini et al.,2008
Nature, 2008
Nature, 2008
Nature Genetics, 2008
PLOS Genetics, 2009
A Genome-Wide Association Study in Chronic
Obstructive Pulmonary Disease (COPD): Identification of
Two Major Susceptibility Loci
Sreekumar G. Pillai1*, Dongliang Ge2., Guohua Zhu1., Xiangyang Kong1., Kevin V. Shianna2, Anna C. Need2,
Sheng Feng2, Craig P. Hersh3, Per Bakke4, Amund Gulsvik4, Andreas Ruppert5, Karin C. Lødrup Carlsen6,
Allen Roses2,7, Wayne Anderson1, ICGN Investigators, Stephen I. Rennard8, David A. Lomas9, Edwin K.
Silverman3, David B. Goldstein2*
SNPs highly correlated with rs16969968
Findings for Nicotine Dependence, Lung Cancer, COPD
rs16969968
rs8034191
Amos et al., 2008
Hung et al., 2008
Liu et al., 2008
Pillai et al., 2009
Saccone et al., 2007
Bierut et al., 2008
Sherva et al., 2008
Weiss et al., 2008
Stevens et al., 2008
rs1317286
Berrettini et al., 2008
rs1051730
Saccone et al., 2007
Thorgeirsson et al., 2008
Amos et al., 2008
Hung et al., 2008
Thorgeirsson et al., 2008
Liu et al., 2008
Pillai et al., 2009
SNPs highly correlated with rs578776
Findings for Nicotine Dependence and Lung Cancer
rs578776
Saccone et al., 2007
Bierut et al., 2008
Weiss et al., 2008
Stevens et al., 2008
Hung et al., 2008
Liu et al., 2008
rs6495308
Berrettini et al.,2008
Genetic Association Data for
Nicotine Dependence and Lung Cancer
Prologue - The Smoke is Clearing
• There are at least two distinct genetic variants on
chromosome 15 associated with nicotine dependence and
smoking quantity.
• These same variants are associated with lung cancer and
COPD.
• Is the mechanism of action related to a change in protein
structure and expression?
• Big Question: Is the association with lung cancer and COPD
only an indirect effect through smoking or both an indirect
and direct effect?
Chapter 1
• What is the utility of linkage analysis in
complex diseases?
Linkage Analysis
Biologic Psychiatry
Genome search meta-analysis results for all independent genome scans on smoking behavior
(3404 families with 10,235 genotyped subjects). Significance levels corresponding to nominal
(p < 0.05), suggestive (p < 0.0085), and genome wide (p < 0.00042) significance are shown by the
horizontal lines.
Meta-analysis of 32 Genome-wide Linkage Studies of Schizophrenia
NYM Ng, DF Levinson, SV Faraone, BK Suarez, LE Delisi, T Arinami, B Riley, T Paunio, AE Pulver,
Irmansyah, PA Holmans, M Escamilla, DB Wildenauer, NM Williams, C Laurent, BJ Mowry, et al
Mol Psychiatry. 2009 Aug;14(8):774-85.
Meta-Analysis of 23 Type 2 Diabetes Linkage Studies from the
International Type 2 Diabetes Linkage Analysis Consortium
Weihua Guan, Anna Pluzhnokov, Nancy J. Cox, Michael Boehnke for the International
Type 2 Diabetes Linkage Analysis Consortium
Human Heredity 2008;66(1):35-49.
TCF7L2
Science 1996
Linkage analysis has little power to localize
genetic regions for complex diseases
• Linkage analysis is great to localize genetic
regions for Mendelian disorders such as rare
illnesses that are transmitted in families.
• There is very limited power for linkage analysis
to detect genetic regions that are associated
with complex illnesses.
Chapter 2
• How to interpret all the previous genetic
findings?
2005 – Time Zero
• 2005 was the start of new generation genetic
studies with genome wide association studies.
Complement Factor H Polymorphism in Age-Related Macular Degeneration
Robert J. Klein, Caroline Zeiss, Emily Y. Chew, Jen-Yue Tsai, Richard S. Sackler, Chad Haynes,
Alice K. Henning, John Paul SanGiovanni, Shrikant M. Mane, Susan T. Mayne, Michael B.
Bracken, Frederick L. Ferris, Jurg Ott, Colin Barnstable, Josephine Hoh
Science, 2005 April 15;308(5720):362-4
Genes reported associated with
diabetes mellitus type 2
140
120
100
80
60
When an association is strong and robust, it is quickly replicated
in various studies and across numerous populations.
40
20
0
1
18
35
52
69
86
103
120
137
154
171
188
205
222
239
256
273
290
307
324
341
358
375
392
409
426
443
460
477
494
511
528
545
562
579
596
613
Number of
Studies
625 genes reported as associated with diabetes mellitus type 2
since 2000. All the top genes were identified in genome wide
association studies.
Number of Genes
genes
Genes reported associated with
schizophrenia
200
180
160
140
120
100
80
60
40
20
0
1
20
39
58
77
96
115
134
153
172
191
210
229
248
267
286
305
324
343
362
381
400
419
438
457
476
495
514
533
552
571
590
609
628
647
666
685
704
723
742
761
Number of
Studies
777 genes reported as associated with schizophrenia
since 2000. None of the top genes were identified in
genome wide association studies.
Number of Genes
genes
Influence of Life Stress on
Depression: Moderation by a
Polymorphism in the 5-HTT Gene
Avshalom Caspi, Karen Sugden, Terrie E. Moffitt, Alan Taylor, Ian W. Craig, HonaLee
Harrington, Joseph McClay, Jonathan Mill,
Judy Martin, Antony Braithwaite, Richie Poulton
Interaction Between the Serotonin
Transporter Gene (5-HTTLPR),
Stressful Life Vents, and Risk of
Depression: A Meta-Analysis
Neil Risch; Richard Herrel; Thomas Lehner; Kung-Yee Liang;
Lindon Eaves; Josephine Hohn; Andrea Griem; Maria Kovacs; Jurg Ott; Kathleen ReisMerikangas
Logistic Regression Analyses of Risk of Depression for 14 Studies
Risch, N. et al. JAMA 2009;301:2462-2471.
2005 – Time Zero
• The new paradigm of genetic studies with
large scale genome wide association studies
has led to an explosion of genetic findings
related to illnesses.
• Findings for complex diseases prior to GWAS
studies are suspect.
Chapter 3
• What have we learned from Genome Wide
Association Studies of schizophrenia, bipolar
disorder, depression, alcoholism and autism?
• Do we have any findings?
Don’t let the p values fool you
• The number of genetic variants tested is in the range
of 500,000 to 1 million.
• P values at 10-5, 10-6 are common. A p value of 10-7 is
starting to be interesting.
Negative results are also a finding.
Genetic effects are modest
• Genetic risks for complex diseases are modest.
• A genetic risk (OR) of 1.3 is large.
• Most genetic risks are in the 1.1 to 1.2 range or
less.
This is true for most complex diseases in medicine.
Alcoholism, schizophrenia, bipolar disorder, lung
cancer, diabetes mellitus (type II).
What do modest genetic effects mean?
• Many genes are involved in disease, which is
consistent with genetic risk in the 1.1 range.
• If there are rare variants associated with
disease, they must be very strong for us to
detect them.
• No one gene will predict disease.
• Prediction of disease will remain difficult.
Chapter 4
• What is the best phenotype to study?
Best phenotype is one that is
associated with genetic variants
• P value ~ sample size and genetic risk.
• To improve the p value you can –
– Increase the sample size
– Increase the genetic effect
Does complex phenotyping help?
• Given that the genetic effect is modest, we will
need very large sample sizes to detect an effect.
(What is large? 50,000 individuals)
• If large sample sizes are needed, then the
phenotyping must be simple and standardized.
• If there are complex phenotypes with complex
measurements, then the genetic effect must be
very large to compensate for the smaller studied
population.
Chen et al., under review
Chapter 5
• What does gene environment interaction
really mean?
Gene Environment Interaction
• Genetic effect may differ in varying
environments.
• Common environmental variables include –
parental monitoring, peer smoking, childhood
sexual abuse, other childhood adversity.
Gene and Parental Monitoring Interplay
16.00
14.00
rs16969968=GG
(Reference)
rs16969968=GA
(OR=1.17)
rs16969968=AA
(OR=2.11***)
12.00
Odds Ratio
10.00
7.75
8.00
6.00
***
* p<.05, ** p<.01, *** P<.001 compared to the
reference group (GG & higher quartiles)
Average odds ratio for specific
genotype
4.00
***
**
2.02
1.71
2.00
***
1.81
1.17
1.00
0.00
higher quartiles
(n=673)
lowest quartile
(n=179)
higher quartiles
(n=664)
lowest quartile
(n=218)
parental monitoring
higher quartiles
(n=190)
lowest quartile
(n=58)
Chen et al., 2009
Predicted Probability of Nicotine
Dependence*
Gene and Peer Smoking Interplay
0.8
0.7
0.6
G/G
A/G
A/A
95% C.I.
0.5
0.4
0.3
rs16969968
0.2
0.1
0
1
2
3
4
5
6
7
8
Number of Smoking Peers
Johnson et al., in review
Gene Environment Conundrum
• Environmental risk reduction is universal.
• Common environmental variables – parental
monitoring, peer smoking, childhood sexual
abuse, other childhood adversity.
• Will we say “It’s ok not to monitor your child.
He won’t smoke.”
Chapter 6
• What is the power to detect gene
environment interaction?
Chapter 6
• What is the power to detect gene
environment interaction?
• Subtitle –
If you thought the power was poor to detect a
main effect, then wait till you test power to
identify an interaction.
Power Curves
Power and Effect Size by prevalence of comorbid disorder, given sample size N=2689
Prepared by Hong Xian
Sample Size Curves
Required Sample Size and Effect Size for 80% Power, by prevalence of comorbid disorder
Prepared by Hong Xian
Chapter 7
• Should we move into studying diverse
populations?
National Center for Biotechnology
Information (NCBI) database for Genotypes
and Phenotypes (dbGaP)
– 22 U.S. genetic studies (59,000 subjects)
– 6 largest studies (diabetes, lupus, macular
degeneration, age-related eye diseases)
• 33,000 subjects
• 180 African Americans (0.5%)
– 5,600 African Americans in 4 psychiatric studies
African American Subjects
Participate in Genetic Studies
African American
Subjects
Eligible Subjects
Donated blood for
genetic study of
Nicotine Dependence
706
European American
Subjects
2,473
504
1,415
71% of eligible
57% of eligible
p<0.0001 for difference in participation rates
between European Americans and African
Americans (χ2 test)
Hartz et al., in review
Differences in populations
• There are clearly differences between
populations in frequency of genetic variants
and prevalence of disease.
• Genetic variants act the same way in different
populations. (Ioannidis et al., 2004)
Diverse Populations
• The promise of personalized medicine and
genetic treatment is not there for minority
populations.
• Scientifically, diversity is good.
Chapter 8 – Don’t get me started
Three ways to validate a finding
• Replication
• Replication
• Replication
• Replication means same phenotype and same
variant in the same direction.
Nominal association of deletions at 1q21.1, 15q11.2 and 15q13.3
with schizophrenia and related psychoses in the phase I sample
Chromosome 1:
144.94-146.29 (Mb)
Chromosome 15:
20.31-20.78 (Mb)
Chromosome 15:
28.72-30.30 (Mb)
Locus
Cases
Controls
Cases
Controls
Cases
Controls
Iceland
Scotland
Germany
England
Italy
Finland
Total
OR
1 of 646
2 of 211
1 of 195
0 of 105
0 of 85
0 of 191
4 of 1,433
8 of 32,442
0 of 229
0 of 192
0 of 96
0 of 91
0 of 200
8 of 33,250
8.68 (1.02,
49.76)
0.024
4 of 646
2 of 211
3 of 195
1 of 105
0 of 85
0 of 191
10 of 1,433
58 of 32,442
0 of 229
0 of 192
0 of 96
0 of 91
1 of 200
59 of 33,250
3.90
(1.42, 9.37)
0.007
1 of 646
1 of 211
1 of 195
0 of 105
0 of 85
0 of 191
3 of 1,433
7 of 32,442
0 of 229
0 of 192
0 of 96
0 of 91
0 of 200
7 of 33,250
8.94 (0.79,
58.15)
0.040
P-value
Three deletions show nominal association with schizophrenia and related psychoses in the first sample of 1,433
patients and 33,250 controls. These deletions are large: the 1q21 deletion spans approximately 1.38 Mb, the one on
15q11.2 approximately 0.47Mb and the one on 15q13.3 approximately 1.57 Mb. P-values (uncorrected for the 66 tests)
are from the exact Cochran–Mantel–Haenszel test and are two-sided. Coordinates are based on Build 36 assembly of
the human genome. 95% confidence intervals are given within brackets.
From Stefansson et al. (2008) Large recurrent microdeletions associated with schizophrenia. Nature 455: 232-236.
And the rest
• There are no findings for psychiatric genetics.
• Deidentified genome wide association studies.
• Sample sizes less than 1,000 people for
genetic studies using disease status.
Chapter 9 – The Happy Ending
Mokdad et al., 2004
Mokdad et al., 2004
Mokdad et al., 2004
US Cigarette Use vs. Lung Cancer Deaths,
1900-2005
Cessation and RemissionThe Final Step
Initiation
Cigarette Use
Nicotine Dependence
Cessation and Remission
Phenotypic and genetic data are available to qualified investigators
through the NIDA Genetics Consortium and dbGaP.