Transcript Document

Physiological Genomics from
Rats to Human
Monika Stoll, Ph.D
Director, Genetic Epidemiology of vascular disorders
Leibniz-Institute for Arteriosclerosis Research, Münster
Genome-oriented Medicine
Genetic Variation influences
-
disease susceptibility
disease progression
therapeutic response
unwanted drug effects
The use of genetic variation for diagnostic
purposes and targeted treatment
“Heterogeneity “ of complex diseases
“polygenic with genetic Heterogeneity”
Gene+
Gene
-
Epistasis
Gene+
Gene
-
Gene
complex
phenotype
Gene+
others
Salt intake
Psychosocial
Stress
+
Diet
“Environmental factors”
Gene-environment interactions and CVD
Genetic factors
Environment
Risk factors
Trait
Phenotype
Diet, Smoking, Stress
Hypertension, Diabetes, Obesity,
Age, Lipids, Genetic Background
Atherosclerosis
Myocardial
infarction
Stroke
Peripheral
vascular
disease
Complex Diseases do not have a
clear phenotype but may or may not
share some features
Example: metabolic syndrome (syndrome X)
atherosclerosis
dislipidemia
hypertension
vascular
disease
Insulin
resistance
obesity
hyperglycemia
Genetics of Multifactorial Diseases
Difficulties
Human Linkage Analysis
Family studies/ Sib-Pair
Analysis:
large number of patients
(2,500 sibpairs)
Modest resolution
Multiple Genes:
Interaction, Epistasis
Lack of Power
Difficulties
Disease Etiology
Polygenic:
modest effects of single genes
Incomplete penetrance
Age-of-onset
Environmental component
Genetic Heterogeneity
High Complexity
Genetics of Multifactorial Diseases
Solutions
Association studies
Solutions
Reduction of complexity
Large scale association studies
Animal models e.g. rat
Transmission Disequilibrium Tests
Sib - TDT
Association studies on quantitative
traits
Controlled genetic background
Controlled environment
Controlled experimental setting
Large number of progenies
Increased statistical power
High density typing necessary
Decreased heterogeneity
Provide candidate regions
Comparative Maps
Positional candidate loci for high density genotyping
Comparative Genomics with Biology
Human
Genes and Genetic Manipulation
relevant to human disease
Mouse
Genes, Physiology and Pharmacology
relevant to human disease
Rat
Ability to avoid many biological barriers
unique to one species
Why ‚Comparative Genomics‘?
Take advantage of the wealth of genome information
from the various Genome Projects
Genomic regions are evolutionary conserved between mammalian species
(Synteny)
Sequence is highly conserved between species
(Homology)
The genomic sequence of human, rat and mouse genomes are available
QTLs/Genes identified in rodent models are predictive for human loci
Rodent models can help to elucidate the function of novel disease genes
e.g. implicated by human linkage studies or expression profiling
Strategies for ‚comparative genomics‘
• Map ‚novel‘ genes identified e.g. in expression profiling and anchor
on existing comparative maps (www.rgd.edu/VCMap)
• Sequence positional candidate genes in mouse, rat and human to
identify conserved mutations and/or regulatory elements
• Predict potential target regions for human linkage studies based on
model organisms
• Characterize candidate genes from human studies in representative
experimental model (inbred strains, congenics, transgenics,
conditional knock-outs)
Experimentelles Modell
Monogene Erkrankung
Geschwisterpaar-Untersuchungen:
Bestätigung Kandidatengen-Locus
Assoziationsstudien:
Identifizierung von Kandidatengen-Polymorphismen
(polygene) komplexe Erkrankung
Cross design
SHR-SP
Backcross
SHR or WKY
F1
F2
SHR or WKY
SHR-SP
x
F1
F2
Human Chromosome Regions Implicated in
Hypertension via a Cross-Species Comparison
Blood Pressure Phenotypes
27 independent blood pressure phenotypes
• Baseline Blood Pressure
• Maximal Response
• MAP, DBP, SBP, PP
• MAP, DBP, SBP, PP after salt-load
• Drug Challenges
• Delta BPs
Rat Models for Genetic Hypertension
Spontaneously Hypertensive Rat (SHR)
High blood pressure
Cardiovascular disease
Genetically Hypertensive Rat (GH)
Hypertension, cardiac hypertrophy
Vascular disease, not salt-sensitive
SHR x WKY
SHR x DNY
SHR x BN
GH x BN
Dahl Salt-Sensitive Rat (SS)
Salt-sensitive hypertension
Hyperlipidemia, insulin resistance
SS x BN
Lyon Hypertensive Rat (LH)
Mild hypertension, hyperlipidemia
LH x LN
Fawn-hooded Hypertensive Rat (FHH)
Systolic hypertension
Renal failure
FHH x ACI
Linkage Analysis for Blood Pressure QTLs
Independent total genome scans in 7 intercrosses representing
a model for genetic hypertension
200-300 SSLP markers
10-20 cM spacing
57- 390 animals
Linkage analysis using MAPMAKER/QTL computer package
LOD score >2.8 suggestive
LOD score >4.3 significant
Integration of QTLs on integrated map based on genotyping
information from crosses used for linkage analysis
Analysis of QTL Clustering
5
4.5
4
3.5
3
Reihe1
2.5
2
1.5
1
0.5
0
LOD score
QTL #1
QTL #2
Reihe1
3
Reihe2
2
Reihe3
1
QTL cluster
1M
i t5
D
LS
N
1M
i
D t1
1M
i t2
1
D
1M
i t4
0
D
LOD score
4
1M
i t5
D
LS
N
1M
i t1
1M
i t2
1
D
1M
i t4
QTL #3
5
Drop of 1.6 LOD
units =
95% confidence
interval
R
D
1M
i t5
1M
i t1
1M
i t2
1
D
1M
i t4
D
LS
N
1M
i t5
D
D
D
1M
i t1
1M
i t2
1
D
1M
i t4
0
D
1
LS
N
2
D
3
D
4
3.5
3
2.5
2
Reihe1
1.5
1
0.5
0
LOD score
LOD score
5
Establishment of Syntenic Regions in Human Genome
Identification of syntenic regions and evolutionary
breakpoints using comparative maps between rat, mouse and
human
Definition of positional candidate regions in human genome
based on QTLs identified in rat models of hypertension
Designation of ‘first priority’ and ‘second priority’ regions
first priority region
second priority region
based on QTLs from
multiple rat crosses
based on QTLs from
single rat cross
QTLs identified in Rat
68 blood pressure QTLs total
LOD score > 4.3
LOD score 2.8-4.3
LOD score 2.5-2.8
13
44
11
13 QTL clusters total
7 QTL clusters 2 or more crosses
6 QTL clusters within one cross
10 single QTLs
First priority regions
Second priority regions
Baseline BP
Max. response
MAP, DBP, SBP, PP
Salt MAP, DBP, SBP, PP
Drug challenge
7
Delta BP
2
7
19
22
11
Coverage of rat genome in cM
500 cM (31%)
Syntenic Regions in Human
36 syntenic regions total
Classification
23 ‘first priority’ regions
13 ‘second priority’ regions
Coverage of human genome in cM
~800 cM (~24%)
Confidence level
highest: 7 regions (14 QTLs)
high: 20 regions (38 QTLs)
moderate: 5 regions (10 QTLs)
conversion incomplete or
impossible 6 QTLs
Identification of Syntenic Regions and Evolutionary
Breakpoints
Identify homologous genes mapped
in rat, mouse and/or human
Preliminary comparative maps of
genes in common on the genetic
maps of rat and mouse
RATMAP server
http://ratmap.gen.gu.se
Oxford Maps
http://www.well.ox.ac.uk
MIT Maps
http://www.genome.wi.mit.edu/rat/
Preliminary comparative maps of
genes in common on the genetic
maps of mouse and human
Framework comparative maps
RATMAP server
Mouse Genome Database
http://www.informatixs.jax.org
UniGene
http://www.ncbi.nlm.nih.gov/
UniGene/index.html
Genome Database
http://gdbwww.gdb.org
VC-MAP : Bioinformatics-‘Tool‘ for comparative maps
Stoll et al., Genome Res. 10: 473
– 482, 2000
http://www.genome.org/cgi/content
/full/10/4/473
Free access
Kwitek et al. Genome Res. 11:
1935 – 1943, 2001
http://www.genome.org/cgi/content
/full/11/11/1935
Free access
www.rgd.mcw.edu
Comparative Mapping
Human chr. 22 and its homologies to rat chr. 11, 20, 6, 14 and 7
5
LOD score
4
Series1
3
Series2
2
Series3
19.0 cM
Comparative mapping of BP QTLs
D HS - LS SBP
D HS - LS MAP
HS basaler DBP
HS aktiver MAP
Tag 2 DBP
TPM Alpha2
HS Prot Excr
HDL
D1Mit5
LSN
D1Mit4
D1Mit21
0
D1Mit1
1
10
13
1.5
D18Rat85
18
D18Mgh3
D18Rat9
GJA1, D18Mit16
D18Mit8
5q
MBP
MC5R
FECH
ADRB2
DRD1
PDGFRB
16.2
6.7
2.5
6.7
2.7
2.5
1.6
3.4
Ratte Chr. 18 7.6
D18Rat57
D18Rat18
D18Mit5
D18Mgh9
D18Mgh7
D18Mit3
D18Mit14, D18Mgh8
D18Mit1
D18Mit12
GRL1
FGF1
EGR1
Humane Homologie
Predicted susceptibility loci in the human genome
Mouse
Rat
39,40,41,42
30,31,32,33,38
34,35,36,37
20,21,22,23,
24,25,26
45,46,47,48
51,52,53,54
27,28,29
13,14,15,16
17,18,19
Krushkal et al.
20,21,22,23,
24,25,26
13,14,15,16
17,18,19
51,52,53,54
Mansfield et
al.
Chr.1
Chr.2
Chr.3
Chr.4
Stoll et al. Genome Res. 10: 473 – 482, 2000
http://www.genome.org/cgi/content/full/10/4/473
Free access
Conclusion
The regions in the human genome implicated for
hypertension may be useful as primary targets
1. Large scale testing in human populations
Association studies
TDT, Sib-TDT
Linkage studies
2. High density mapping
Targeted genome scans
Single Nucleotide Polymorphisms (SNPs)
Genetic studies in
human populations
Is there a genetic component ?
Mendelian Disease:
Exhibits Mendelian mode of inheritance
Complex Disease:
Appears to cluster in families
Family, twin, adoption studies show greater risk to
relatives of affecteds than the population incedence
Segregation analysis can provide estimates of genetic
and environmental contribution to disease
Where is the gene ?
Linkage analysis:
Cosegregation of mapped marker with the disease
Fine mapping to narrow the region
In Complex Disease:
Requires a defined genetic model
Requires classifying people as affects and unaffecteds
Allele sharing methods (sib pairs etc.)
Population association studies
Genetic Methods
(ca. 400 Mikrosatellites, 10cM)
Traditional
LOD score (MLS score)
Genomwide linkage
Chromosome 10 linkage
2.5
2
1.5
1
*
0.5
0
Fine mapping
(Saturation with Mikrosatellites, 1cM)
Association and Linkage Disequilibrium
(SNPs, 3-50kB, Transmission Disequilibrium, LD, Haplotype analysis)
Association in Case/Control Design
2
2
2
2
1
2
1
1
(SNPs, Haplotype Case/Controls, ethnically divergent populations)
Linkage analysis
Linkage Disequilibrium
Linkage analysis
Non-parametric linkage studies
1/2 3/4
1/3
2/4
1/2 3/4
1/2 3/4
1/3
1/3
2/4
Looking at a marker
Association in between families
Extended families
Affected relative pairs
Discordant pairs
1/4
1/2 3/4
1/3
Affected sib pairs
Problem: late onset of CAD
2/4
Non-Parametric Linkage Analysis
Chromosome
LOD= log10 [L()/L(1/2)]
= log10 [Prob. Linkage/Prob. No Linkage]
m1
Disease gene
m2
m3
See Figure 1 from Broekel et al.
Nature Genetics 30, 210 - 214 (2002)
http://www.nature.com/ng/journal/v30/n2/full/ng827.html
Free access
m4
Several examples for hypertension linkage in
human study populations
How to get from linkage to
the causative gene variant ?
What is Linkage Disequilibrium ?
Linkage - property of the relative position of loci, not their alleles.
Linkage is the cosegregation of a disease or trait with a
specific genomic region in multiple families (it can involve
any allele at the marker locus in a given family)
Association - property of alleles: a specific allele of a gene or marker
is found with a disease or trait in a population
Linkage Disequilibrium – the presence of linkage AND association
Cosegregation of a specific allele with the disease in a
significant number of families
Why do we care about Linkage Disequilibrium ?
It is a tool for fine mapping
Affected sib pair analysis may not be sensitive enough to detect
minor genes
Association test may be sensitive but the association detected may
not be due to linkage disequilibrium. It could be caused by
population stratification (confounding due to race, admixture,
heterogeneity in the population for some other reason)
How do you analyze for Linkage Disequilibrium ?
Transmission Disequilibrium Test (TDT):
TDT tests for equal numbers of transmissions of specific alleles
and all others from heterozygous parents to an affected offspring
GENEHUNTER: Transmitted vs. Untransmitted alleles
TRANSMIT: Expected vs. Observed alleles
TDT test is McNemar‘s Chi-square test = (b-c)2/(b+c)
Allele 1
Allele 2
Trans Untrans
211
138
138
211
Chi-square= 15.27
p=0.000093
Limitations: locus heterogeneity, allelic heterogeneity, need for
specific polymorphisms, can only detect linkage in the presence of
association, need to be very close to disease gene
What‘s all that Fuzz about Haplotypes ?
Linkage Disequilibrium decays with time (No. of recombinations)
2
m2
X2
m1
2
1
LD
 = (1-)t
1
1
X1
2
1
2
2
X
1
X1
t
2
1
2
1
2
2
2
2
2
1
2
1
2
1
1
Size of Haplotype blocks depends on population history
L. Kruglyak (1998): need 1 SNP/3kb for genomewide association
D. Reich (2001): haplotype block size in Caucasians 60-120kb due to
bottle neck in population history 50,000 years ago
haplotype block size in Africans 10-30 kb
M. Daly (2001): haplotype block structure in human genome
2003:
haplotype structure varies. Blocks of long range LD
interspersed with recombination hot spots
 Human Haplotype Map – will be finished in 2005
Hierachical Linkage Disequilibrium Mapping
See figures from Stoll et al.
Nature Genetics 36 (5): 476-480, 2004
http://www.nature.com/ng/journal/v36/n5/index.html
Subscription access only
ALOX5AP is a susceptibility gene for MI and stroke
See figure from
Helgadottir A. et al.
296 multiplex icelandic families (713 individuals)
Linkage on 13q12-13
LOD score: 2.86
Nature Genetics 36 (3): 233-239 (2004)
http://www.nature.com/ng/journal/v36/n3/index
.html
Subscription required
14 additional microsatellites
LOD score 2.48 (p=0.0036) at D13S289
Haplotype based case-control association using
150 microsatellites
Haplotype with association to MI (p=0.00004)
Gene within haplotype ALOX5AP
144 SNPs identified by resequencing 97 individuals
2 haplotype blocks in strong LD
Association testing in case/control study design
ALOX5AP is a susceptibility gene for MI and stroke
See Table 1 from
Helgadottir A. et al.
Nature Genetics 36 (3): 233-239 (2004)
http://www.nature.com/ng/journal/v36/n3/index.html
Subscription required
See Table 2 from
Helgadottir A. et al.
Nature Genetics 36 (3): 233-239 (2004)
http://www.nature.com/ng/journal/v36/n3/index.html
Subscription required
Conclusion
Success stories for Comparative Genomics
Obesity:
Discovery of Leptin as the human homologue of the mouse (ob) mutant
Leptin receptor and db/db mice (diabetes and obesity phenotype)
Melanocortin-4 receptor and severe obesity in mice and man
Diabetes:
Cd36 as a susceptibility factor for insuline resistance in the SHR rat
Cblb (ubiquitin-protein ligase) as susceptibility factor for Type I Diabetes
Atherosclerosis:
APOAI/CIII/AIV gene cluster and lipid metabolism in mice and man
Hypertension:
Predictive power of QTLs from rodents for human hypertension
Total Genome Scan
Phenotype
Candidate
Gene Approach
Positional
Cloning
Congenics
Consomics
ENU-Mutagenesis
Case-control Studies
Gene
Transgenics
Knock-outs
Knock-ins