Transcript Pippa Thomson - University of Edinburgh

Discovering Disease GenesThe Example of Schizophrenia
Pippa Thomson, Medical Genetics Section, Dept of Medical
Sciences, MMC, University of Edinburgh.
Importance of the illness
• Severe psychiatric
• Affects ~1% of the population
• One of the top 10 causes of disability worldwide
• Economic cost (23% Drug, 14% Hospital)
• Pharmacological rationale for treatment weak or
absent
• 1/3rd patients unresponsive or experience
unacceptable side effects
• Strong genetic component
– concordance rate between identical twins of 60%
Schizophrenia
Positive symptoms: visual &
auditory hallucinations,
delusions, incoherent speech
Schizophrenia
Negative symptoms:
withdrawal & isolation,
impaired attention & blunted
emotions
Altered brain structure &
function
High Heels Cause
Schizophrenia and 6 Other
Outlandish Medical Theories
2. High-heeled shoes cause schizophrenia. You have
to wonder where some medical theories originate. Why
did Swedish scientist Jarl Flensmark decide to study a
connection between heeled shoes and the incidence of
schizophrenia? The world may never know. But his initial
research seems sound, and he has connected certain
brain activity with stimulation of certain points on the feet.
The spread of schizophrenia around the globe has closely
followed the spread of availability of heeled shoes. Is it an
eerie coincidence or a real cause for concern? Look out,
men - this theory applies not only to stilettos, but to any
shoe with a heel.
remedicated.com
Relative risk of developing
Schizophrenia
MZ
~50%
Environment !
50
40
1st degree relatives
~10-15%
30
20
general
population
10
identical twin
non-identical twin
parent
brother or sister
grandparent
aunt or uncle
cousin
no affected relative
~1%
0
Benefits of gene identification
• Understand aetiology
• Improved drug development & testing
• Development of definitive diagnostic tests
• Understanding of interaction with nongenetic risk factors
• Insight into normal brain development &
function
Kraepelin, 1896
“As we do not know what causes the illness
there cannot be a rational treatment”
Magnitude of effect
Allelic architecture and
mapping strategy
Unlikely to exist
Family-based
linkage studies
Association studies in
populations
Fnct. Studies
Frequency in population
Locus Identification-problems
•
•
•
•
Uncertainty in diagnostic boundaries
Non-Mendelian inheritance
Variable age of onset
Genetic heterogeneity
– Many different genes can cause the illness
>1% risk world wide
>phenotypic variation
• Oligogenic/polygenic causation
– More than one mutant gene required to produce
phenotype
Locus identification- reducing
the problems
• Single large families
• Avoid bilineal descent
• rigorous interviews
• family history
Reduce genetic heterogeneity
Significant LOD score = gene of major effect
• Reduce uncertainty of diagnosis
– classify minor diagnoses as unaffected
– >1 category of affected phenotype
Linkage Analysis
• Marker analysis in multiply affected family or families
• Look for co-segregation of a particular allele with
phenotype
• Results expressed as a LOD score (Significant at > 3)
= log (likelihood of data, if locus & disease are linked)
---------------------------------------------------------------(likelihood of data, if locus & disease are not linked)
• Generally a large region is identified
A balanced t(1;11)(q42;q14)
translocation
1
11q14
der1
der11
11
1q42
11
1
t(1;11) co-segregates with
major mental illness
?
schizophrenia
recurrent major depression
bipolar affective disorder
translocation
increases risk by
50-fold
(1;11)(q42;q14) translocation
unaffected
minor diagnosis
100’s Individuals = 1% Schizophrenia
100’s Individuals = 100% Schizophrenia
• Genetic association studies seek to relate variation in
human DNA sequence with a disease or trait
• Association method provides greater power to detect
common genetic variants conferring susceptibility to
complex phenotypes
• Estimates population attributable risk (effect size)
• Controls should match cases and be a representative
sample of the population.
Case-control association
studies
T
G
A
C
• Comparison of frequencies of polymorphisms between
populations of cases and controls (usually a simple chi-square test
or logistic regression)
• Polymorphism studied can be directly responsible for the defect 
frequency of
cases >>> controls
• Polymorphism studied can be in linkage disequilibrium with the
mutation responsible for the disease  %T cases >> controls
• Association studies can be conducted for candidate genes, or
through a whole region or across the whole genome (WTCCC)
p
Mb
International HapMap project
SNPs are genotyped in parent-offspring trios, initially in CEPH trios.
This can be used to identify SNPs that co-segregate (i.e. are in linkage
disequilibrium) versus those that segregate independently.
A subset of SNPs can therefore be chosen that best represent the
genetic diversity in a region/gene, reducing the costs of genotyping.
Summary of genotyped SNPs:
Populations
CEU
Total Non-Redundant
3,204,709
CHB+JPT
3,244,897
YRI
3,150,433
http://www.hapmap.org/
Region of interest
HapMap genotyped SNPs
Known SNPs*
Known genes in the regions
Linkage Disequilibrium
(LD)
*http://www.ncbi.nlm.nih.gov/SNP/
Tagging SNP selection
Proportion of haplotype diversity explained : SNPs 1-23 - 97%
SNPs 24-46 - 98%
Genetic evidence implicating
DISC1 in psychiatric illness
LOD=1, TAIWAN (Hwu et al 2003)
SCZ
LOD=7.1, SCOTLAND
LOD=2, BRITAIN & ICELAND (Curtis et al 2003)
BPAD
SCZ & BPAD & MDD
TRANSLOCATION
D1S251
p=0.0044, p=0.0016
SCOTLAND
LOD=3.21, FINLAND (Ekelund et al 2001)
SCZ
SCZ, BPAD
D1S2709
HAPLOTYPE
1
2
3
4 5
6
78
9
10 11 12
13
DISC1
DISC2
HAPLOTYPE
SCZ & SCZAFF
p=0.00024, FINLAND (Hennah et al 2003)
rs6675281
SCZAFF
p=0.000027, North-America (Hodgkinson et al 2004)
>130 genes implicated
Table 1. Summary of current evidence supporting several of the
more promising genes implicated in schizophrenia, bipolar
disorder, and mixed bipolar-psychosis phenotypes
Craddock et al., SCZ Bulletin, 2006
DISC1 interactome
protein-protein interactions
Chris Carter, http://www.polygenicpathways.co.uk/disc11_vml.htm
Effects of altered DISC1 on
gene expression
• ENU generated mouse mutants
• Two independent lines with missense mutations in
DISC1 exon 2
– Q31L (Glutamine-Leucine)
> Q- hydrophillic; L – hydrophobic
– L100P (Leucine-Proline)
> Predicted to cause transition in polypeptide chain direction
• Normal levels of DISC1 protein in brain
• L100P line models schizophrenia; Q31L, depression
Clapcote et al., Neuron. 2007 May 3;54(3):387-402.
Effects of Altered DISC1 on
Behaviour (How do you know
if a mouse is schizophrenic?)
Phenotype
Anxiety (elevated plus-maze)
Horizontal activity
Vertical activity
Prepulse inhibition (PPI)
Acoustic startle response
Startle reactivity
Latent inhibition (LI)
Working memory (T maze)
Spatial learning and memory (Morris water maze)
Forced swim immobility (FST)
Sociability and social novelty
Sucrose consumption
Brain volume
PDE4B activity
31L/31L
=
=
=
<
=
=
<<
<
=
>
<
<
<
<<
Drug Treatment
PPI
31L/31L 100P/100P
=
+
+++
=
Clozapine
Bupropion
100P/100P
=
>>
>
<<
Schizophrenia
<
<
<<
<<
=
=
Depression
=
=
<<
=
Effects of altered DISC1 on
gene expression
•Samples collected and microarray study ongoing
–Mutated lines vs background strain (C57/BL6)
–47,000 transcripts
–Hippocampus
–Adult and embryonic stage- Microarray
–Confirmation/Investigation of changes
–Series of embryonic; postnatal and adult stages
–Drug treated adult mice
Detect disrupted pathways
Whole genome sequencing
Resequencing, SNP detection, genome comparisons,
gene expression, transcription factor studies, small RNA
analysis
Individual genomes – All SNPs in each individual
Currently :
Using the Illumina 1G to sequence genes in the DISC1 pathway.
Total sequence read 3.5megabases in 1200 individuals
•
•
•
•
Identify coding and non-coding polymorphisms
Mutation detection
Detection of variants in conserved regions
Detection of variants affecting binding of transcription factors
Psychiatric GeneticsUnanswered Questions
How many susceptibility genes are there?
What is their function? Is function conserved
across species? Can we relate gene
(dys)function to mental (dis)order? Do gene
variants predict risk, course, outcome and
response to treatment? Will gene discovery
lead to drug discovery? How do genes and
environment interact? How and when will the
patient benefit?
Acknowledgements
DISC1
Medical Genetics
Kirsty Millar
Shaun Mackie
Fumiaki Ogawa
Jennifer Chubb
Becky Carlyle
Nick Bradshaw
Sheila Christie
Prof David Porteous
Prof Douglas Blackwood
Walter Muir
Ben Pickard
Steve Clapcote
Kathy Evans
Sarah Brown
DISC1 Consortium
Wellcome Trust CRF
Illumina, San Diego
Cold Spring Harbor laboratories
William Hennah
Other collaborators