Transcript Document

Genetics of Alcohol Response
Addiction Medicine
State of the Art Conference
October, 2003
Ray White
Alcoholism
• Many definitions
– Precision important for specific studies
– Working definition:
• alcohol craving has become encompassing drive
• Individual is losing, or has lost, job, family, health
• Economic and personal costs astronomical
– 10s of thousands of traffic deaths each year
– Affect not only individual, but family, bystanders…
• Cultural context
Genetics and
Molecular Etiology
• Cell and molecular biology describe molecular pathology
of disease state
• Often cannot distinguish cause from effect or
consequence of disease
– Important to understand chain of causality in order to
intervene
– Genetics can define components that are “sufficient” to cause
disease
• APC gene
• Genetics can identify etiologic components not seen by
cell, molecular biology analyses
– APC gene
– Molecular mechanism
– Therapeutic targets
Alcoholism Has
Genetic Component
Twin Pair Concordance Rates
Twin Group
#Pairs
Prevalence
%
S.E.
Concordance
%
S.E.
MZF
932
5.3 0.6
29.5
6.1
DZF
534
6.4 0.7
16.7
6.1
MZM
396
21.2 1.7
56.1
4.6
DZM
231
24.2 2.0
33.3
5.7
DZUS
592 M
26.1 1.7
59.5
8.1
14.0
2.8
F
Heath et al, Psych Med 27:1381 (1997)
6.4
0.9
Genes for Alcoholism
How do we find them?
• Positional cloning with families
–
–
Global scan for disease susceptibility genes (+)
All genes in region become candidates
*
M1
*
M1
*
*
*
Ascertain
family cluster
*
M1
*
M1
*
M1
*
M1
*
M1
*
M1
*
M1
Expand into pedigree
Analyze by linkage
*
M1
Positional Cloning with
Families
*
M1
*
M1
*
M1
*
M1
*
M
*
* 1
M1 M1
*
* *
M1 M1 M1
13
13
12
12
11
11
11
11
12
12
21
21
22
22
23
23
24
24
25
25
• Global scan for disease susceptibility genes (+)
• All genes in region become candidates
L5.63
L5.63
M1
L5.79,71,63
L5.79,71,63
MCC
MCC
500
500 kb
kb
5 mb
DP1
DP1
SRP14
100
100 kb
kb SRP19
30-40
30-40 mb
mb
CHROMOSOME 5
260
260 kb
kb
APC
L5.48
L5.48
*
*
Genes for Alcoholism
Positional Cloning with Families
Reich et al.
Amer. J. Med
Gen 81:207
(1998)
Genes for Alcoholism
COGA Replica Set
Faroud et al,
Alcohol: Clin
and Exp Res
24:933 (2000)
Genes for Alcoholism
Linkage Mapping in Families:
Conclusions and Caveats
• Significant linkage findings strongly support genetic
component to alcoholism
• Broad peaks
– Many genes implicated
– 20 to 40 genes in most regions
– More data does not solve problem
• Replication uncertain
– Peaks lower, breadth maintained
• Only rarely has led to gene identification in complex
diseases –
– But often in rare, syndromic diseases
Genes for Alcoholism
Associations in Populations
SNP1
Dvariant
SNP2
SNP3
High frequency marker SNPs (alleles @ 10% - 50%)
• Experiment
– Genotype SNPs in cases and controls
• Expected outcome
– Some SNPs show increased frequency among cases
– Association of SNP haplotypes identifies chromosome
region carrying mutation (Dvariant) causing disease
susceptibility
Genes for Alcoholism
Functional Candidates:
Associations in Populations
• 5 variants – Individuals with family history;
% Alcoholism Diagnosis as function of genotype
– 5-HT Transporter,
• LL–57% SL–21% SS-12.5% ; p=0.04
– 5-HT2A T102C,
– 5-HT2A TYR,
– 5-HT2C CYS-SER,
– GABAAa6
• Pro/Pro-23% Pro/Ser-71% ; p=0.02
Schuckit, M. A., C. Mazzanti, et
al. (1999). Biol Psychiatry
45(5): 647-51
Genes for Alcoholism
Association in Populations:
Conclusions and Caveats
• Common SNP markers may miss variants that
are low frequency (1% - 2%)
• Most interesting variants may be in the low
frequency class
Individual Response to
Alcohol Challenge
• Broad range of response among individuals
• Not explained by pharmacokinetics
– Response largely independent of blood alcohol levels
– Response ranges from extreme sensitivity to “hollow
leg”
• Response measured by questionnaire (euphoric feeling…)
and by body sway index
Level of Response Heritable
Schuckit, M.A., et al., A genome-wide
search for genes that relate to a low level
of response to alcohol. Alcohol Clin Exp
Res, 2001. 25(3): p. 323-9.
Key Finding: Level of Response
Predicts Risk of Alcoholism
• Entering college students with family history of alcoholism
– No strong personal alcohol history
– Tested with questionnaire and physiological measures
– Followed for extended period
• Low response seen in
–
40% of children of alcoholics
J Stud Alcohol. 1998 Sep;59(5):485-94.
Biological, psychological and
environmental predictors of the
alcoholism risk: a longitudinal study.
Schuckit MA
– 10% of family history negative controls
• Individuals with a low response to alcohol
– 4-fold more likely to become alcoholic
– Drink more, hang out with heavy drinking groups
• May explain much of inherited alcoholism risk
Level of Response Mapping
Schuckit, M.A., et al., A genome-wide
search for genes that relate to a low level
of response to alcohol. Alcohol Clin Exp
Res, 2001. 25(3): p. 323-9.
Multipoint Sib-Pair Linkage - Chromosomes 1, 21
Sibs from Lower Third of FIRST 5 (SRE)
Chromosome 1
Schuckit, M.A., et al., A genome-wide
search for genes that relate to a low level
of response to alcohol. Alcohol Clin Exp
Res, 2001. 25(3): p. 323-9.
Chromosome 21
Mapping Studies: Alcoholism
vs Response Level
COGA Alcoholism Study
Chromosome 1
COGA Level of Response
Chromosome 1
Mapping Response Level Genes II
Wilhelmsen,
K.C., et al.,
Alcohol Clin
Exp Res, 2003.
27(7): p. 10417.
Positional Cloning:
Not Working in Complex Disease
• Too many candidates in linkage
• Too much heterogeneity
– Locus heterogeneity often lethal for linkage and
association approaches
– Allelic heterogeneity often kills association
• Association needs common variants
• New approaches needed - Candidate Genes
– Sequencing for rare variants
– Isoallelic cohorts for phenotypic characterization
Excellent Candidate Genes
for Alcohol Response
• Positional Candidates – in chromosomal
locations that are implicated in disease
• Functional Candidates – in pathways that are
implicated in disease process; e.g. (cAMP/PKA)
• Model system Candidates – found in animal
models to impact alcohol response
• Best candidates meet all three criteria
Genes for Alcoholism
Functional Candidates
Slide: Anastasia Constantinescu
Genes for Alcoholism
Functional Candidates
A2R
AC Gs
cAMP
Ca
Ca RII
Nuclear
CREB
CRE
PKA
envelope
*
Ca
cAMP inducible genes
A Mouse Model:
PKA RIIb Knockout
Consumption of ethanol by
mutant mice lacking the RII
subunit of PKA (RII
/ ) and
wild-type control mice (RII
+/+) maintained on a
129 SvJ × C57BL/6 hybrid
background. a, Consumption
(grams per kilogram) of a 20%
ethanol solution. b,
Consumption (grams per
kilogram per day) at each
ethanol solution (8-d average).
c, Ethanol preference ratios
(volume of ethanol
consumed/total fluid
consumed) as a measure of
relative ethanol preference. All
values reported as
mean ± SEM. ANOVAs
indicated that the
RII
/ mice drank
significantly more ethanol
than RII +/+ mice.
RII
/ versus RII +/+,
*p < 0.05.
Thiele, Todd E., Willis, Brandon, Stadler, Julia, Reynolds, James G.,
Bernstein, Ilene L., McKnight, G. Stanley
J. Neurosci. 2000 20: 75-
Measures of acute sensitivity to
the sedative effects of ethanol,
consumption of nonalcoholic
tastants, and plasma ethanol
levels (mean ± SEM). a, Time to
regain the righting reflex
(minutes) after injection of
ethanol (4.0 gm/kg; i.p.). b,
Consumption (milliliters per
kilograms per day) of solutions
containing either sucrose (Suc)
or quinine (Qui). c, Plasma
ethanol concentration
(milligrams per deciliter) either
1 or 3 hr after ethanol injection
(4.0 gm/kg; i.p). ANOVAs
indicated that RII / mice
recovered from ethanol-induced
sedation significantly sooner
than RII +/+ mice. On the other
hand, RII / and RII +/+ mice
did not differ significantly in
consumption of nonalcoholic
tastants or plasma ethanol
levels. RII / versus RII +/+,
*p < 0.05
Chromosome 7 Mapping
PKA RIIb
Candidate Genes II:
Ethanol Consumption by
PKCe -/- Mice
Reduced responding for ethanol-reinforced lever presses in PKC /
mice compared with PKC +/+ mice. (A) Total number of ethanolreinforced lever presses in a 16-h period, averaged across 8 weeks
of testing. PKC / mice (open bars) demonstrated a significantly
lower total number of lever presses than PKC +/+ mice (filled bars;
t = 2.8, P< 0.05). (B) Total number of ethanol-reinforced lever
presses following two different durations of ethanol deprivation. PKC
/ mice demonstrated a significant reduction in total number of lever
presses following 104 h ethanol deprivation compared with PKC +/+
mice [F (1,23)genotype = 7.4, P< 0.05]. In addition, only PKC +/+ mice
demonstrated a significant reduction in total number of lever
presses following 104 vs. 32 h ethanol deprivation [F
(1,23)duration = 21.6, P< 0.01]. (C) Number of ethanol reinforcers per
/ mice following 104 h ethanol
bout were reduced in PKC
deprivation compared with wild-type controls [F (1,23)genotype = 5.0, P<
0.05], and decreased as a function of duration of ethanol
deprivation period [F (1,23)duration = 9.0, P< 0.05] only in PKC +/+ mice.
(Inset) Number of ethanol self-administration bouts did not differ
among genotypes, but number of bouts decreased as a function of
duration of ethanol deprivation period in both genotypes [F
(1,23)duration = 27.3, P< 0.01]. (D) Rate of lever pressing for ethanol
reinforcement was significantly reduced in PKC / mice compared
with wild-types [F (1,23)genotype = 21.9, P< 0.01]. Similarly, rate of lever
pressing decreased as a function of duration of ethanol deprivation
period [F (1,23)duration = 8.4, P< 0.05] only in PKC +/+ mice. Data are
expressed as mean ± SEM. *Significantly different from 104 h
deprivation period. Significantly different from wild-type controls
Choi, D. S., D. Wang, et al.
(2002). J Neurosci 22(22):
9905-11
Genetics of Genes
Candidates
• Goal is discovery of phenotypes associated with
variants in interesting genes
• Candidates emerge from
– functional analysis – biochemical pathways
– Model organisms
• Problem: Since most interesting variants <1%-2%
frequency, usually see only one example
– Risky to draw conclusions from one example
– Need an isoallelic cohort -analogous to mouse
knockouts
– 25 – 50 subjects who carry same rare allele
Genetics of Genes: Bottom-Up
Driven by Gene Activity –
Phenotype is the Object of Discovery
• Functional candidates
– Many pathways offer exciting
candidates
– Ultimately a “Genetics of the Genes”
Positional Cloning:Top-Down
Phenotype-driven Pathway to
Novel Candidate Genes
• Mapping reduces the complexity >100
fold
– Primary mapping
– Secondary mapping by association with
variants identified in affected sample sets
– Functional candidates within mapped region
• Functional candidates
– Many pathways now offer exciting candidates
– Ultimately a “Genetics of Genes”
Next Generation Positional Cloning:
Gene Mapping in Superfamilies
• Central Problem: Most allele carriers for complex
disease susceptibility alleles do not display
phenotype
– Low, moderate penetrance alleles need bigger window to
ascertain families
– Cannot extend families by following footprints of disease
• Superfamilies provide bigger window, reveal
footprints of low, moderate penetrance alleles
– > 1,000 family members
• Superfamilies display multiple aspects of phenotype
– Expect, and are seeing, extensive overlap among
cancers
Superfamilies
Utah Population Databases

50,000 Families, each founded by a
pioneer couple

Many kinships >1,000

Pedigrees recorded in computerized
database
Advantages
 Reduces heterogeneity
 Founder logic – one locus, one
allele
 Reveals dominant alleles with
reduced penetrance
Superfamilies Provide
Identity-By-Descent Mapping
• One gene, one variant brought in by founding
pioneer couple
• Power for high resolution mapping
• Pairwise testing shows IBD segment 10mb – 20mb
Co-Aggregation of Cancers
in Superfamilies
Co-aggregating Disease
Outcome
Disease
Breas
t
Breast
Colon Lun
g
Melanom Ovar
a
y
Pancrea
s
Prostate
2.4
1.8
0.7
3.5
0.2
2.7
1.2
0.3
0.4
2.2
2.7
1.1
0.7
1.1
3.5
0.9
1.4
4.2
4.0
2.4
Colorectal 3.0
Lung
3.5
1.8
Melanom
a
1.7
0.5
1.3
Ovarian
10.5
0.9
1.8
1.4
Pancreas
0.5
3.7
1.5
1.5
4.3
Prostate
2.2
1.8
1.3
1.4
0.6
Odds ratios for SF at “high-risk” for Outcome Disease given that they
are classified as “high-risk” for Co-aggregating Disease. If OR is
significantly different from 1.0, number is in light mustard-biege.
8.3
0.8
Affected Individuals within Kinship
Identical-by-Descent at Susceptibility Locus
• Will see association through genome scan with
STR markers
A Portion of a
Very Large Utah Family
Conserved
Chromosomal
Segment Size
D5S1393
112.2 mb
D5S1346
118.3 mb
D5S2501
122.4 mb
D5S2051
123.4 mb
D5S421
APC
D5S656
APC
Chromosome 5
D5S2065
D5S659
D5S1720
124.3 mb
124.8-125.1 mb
125.7 mb
126.7 mb
127.2 mb
127.5 mb
D5S494
D5S639
130.8 mb
131.0 mb
ATA24E05
131.8 mb
D5S592
132.7 mb
D5S615
139.0 mb
D5S1391
141.3 mb