Transcript The Genetics of Pain

Using functional genomics to
understand pain
Professor Rodney J. Scott
Director of Genetics, Hunter Area
Pathology Service
• Max MB and Stewart WE (2008) The
Molecular epidemiology of pain:A new
discipline for drug discovery. Nature
Reviews 7:637-658
• Watkins LR and Maier SE (2003) Glia: a
novel drug discovery target for clinical
pain. Nature Reviews 2:975-985
Definition of a Genetic Influence
•
Inherited predispositions to disease
Dominant disease effect (with modification)
•
Inherited traits that only become apparent when the
appropriate environment is present. Disease does not
develop UNLESS environmental exposure occurs
•
Gene-Gene interactions important in dictating disease
response and outcomes
Environment
•
Host
metabolism (e.g. estrogen metabolism)
normal functioning (e.g. neurotransmitters)
aging
•
External
infection
carcinogen exposure
life style
Genotype
Phenotype
Environment
Most diseases are multifactorial: Genetic predispositions,
environmental factors, result in genetic changes that play a
role in disease development throughout life
Stroke
Cardiac diseases
Asthma
Infectious diseases
Obesity
Hyperlipidaemia
Cancer
Blood pressure
Diabetes
Neuropathies
Psychiatric conditions
Pain
Molecular Epidemiology of Pain
• Cost estimated to be ~ $ 1 Trillion annually in US
• Common pain treatments have changed little over 30
years
• Should be a tractable area for drug development –
unique chemical mediators
• Traditional research has focused on ~ 200 molecules
(<1% of the genome)
• Several novel drug classes that relieve pain in animal
models have failed in clinical evaluation
• Despite intensive efforts, clinical pain control has
remained a puzzlingly elusive target.
• Genomic studies in humans might compensate for the
limitations of animal studies
Cell Types & Pain
• Neurons have been the centre of attention
• Glia Cells (micro glia cells and astrocytes) originally
considered to support neuronal activity, should also be
considered as they release a wide variety of molecules
on stimulation
• Glia dynamically modulate the function of neurons under
both physiological and pathological conditions
• Garrison et al. reported that peripheral nerve damage
that created exaggerated nociceptive responses
(neuropathic ‘pain’ behaviours) also activated spinal cord
glia
New view of pathological
pain
Garrison et al demonstrated that astrocytes in spinal cord were activated (as
reflected by immunohistochemistry for the astrocyte-specific activation marker,
glial fibrillary acidic protein) in response to sciatic nerve damage.
Glia
• Glia cells are stimulated by cytokines and
chemokines
• IL1, TNF & IL6 are pro-inflammatory
cytokines that activate glia
• IL10 is an anti-inflammatory cytokine that
blocks the action of IL1 and IL6
• All 4 cytokines harbour functional
polymorphisms that alter their activity
Genomic Studies
• There is heterogeneity in the sensation of pain –
suggestive of variance in the molecular mechanisms
underlying the sensation
• Variance in pain sensation is likely to be a result of
genetic differences between individuals
• Genomic studies on pain might reveal more than that
identified in animal models
– Under- or over-expression of target proteins in different pain
phenotypes can be evaluated before clinical trials.
• Genome wide studies of other disease have revealed
novel genes in a wide variety of disorders
– Macular degeneration, Crohn’s Disease, Cancer, Diabetes, etc…
Genomic Studies into Pain
• Heritability of pain
– Animal studies clearly indicate difference in
pain thresholds – different mouse strains
indicate 30% – 76% (median 46%) heritability
– Human studies (Twin studies)
• Monozygotic twins should be similar w.r.t. pain
thresholds; Dizygotic twins will demonstrate
normal sibling variance (even though they have a
shared environment).
• Variability in pain processing
Genomic Studies into Pain
• Case-Control Studies of familial aggregation
– Suitable for common pain conditions (back pain or
migraine) – large number of cases
– Identify cases from clinical records
– Control subjects from the general population
• Compare heritability
– Frequency of pain in one population (this includes
family members) compared to the other
Genomic Studies into Pain
• Candidate Gene Studies
– Many potential modifiers of pain sensation
– Many cell types involved
– Functional polymorphisms exist in genes
involved in pain sensation
• Promoter gene polymorphism
inappropriate
protein expression or loss of control fidelity
• Polymorphisms in exons
change of function
• Alter miRNA controlling species
Association of COMT haplotypes with experimental pain sensitivity and
with rate of metabolism of catecholamines
(HPS = high pain sensitivity, APS = average pain sensitivity and LPS = low pain sensitivity)
Gene Expression in
Rats after nerve injury
Pain protective effects of the GCH1
haplotype in patients after discectomy
Association of GCH1 haplotypes with human chronic spinal nerve root pain,
experimental pain and synthesis of biopterin (essential for NO production
Genomic Studies into Pain
(the identification of new
biomarkers)
Biomarker Discovery
• Define the question you wish to ask
• There are many biomarkers both known
and unknown
• Determine the number of samples you
need to answer the question (power
calculation)
• Ensure that there are sufficient sample
sets for replication studies
Definition of a Biomarker
•
•
•
•
•
Quantitative Trait
Assess the risk of disease
Better define the disease
Define response to treatment
Assess the risk of recurrence
Definition of a Biomarker ctd.
• Quantitative trait:
– inheritance of a phenotypic characteristic that
varies in degree and can be attributed to the
interactions between two or more genes and
their environment e.g. height, blood pressure,
pain
Definitions:
Environment: The sum total of all conditions and elements
which make up the surroundings and influence the
development and actions of an individual.
Gene: a segment of a DNA molecule that contains all the
information required for synthesis of a product (protein or
RNA molecule), the deficiency of which may only become
apparent after exposure to an appropriate environment.
The Human Genome
ATCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGAGCCTAGGTGGCGAAATT
GCAGAGTTTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGAGCCTAGGT
GGCGAAATTGCAGAGTTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGA
TGCCTAGGTGGCGAAATTGCAGAGTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGAACCTTAGA
GCCTAGGTGGCGAAATTGCAGAGTTTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGAGCCTAGGTGGCGA
AATTGCAGAGTTCGGACCTTAGAGCCTAGGTGGTCGAAATTGCAGAGTTGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTC
GGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTTCGGACCTTAGAGCC
TAGGTGGCGAAATTGCAGAGTTGACCTTAGAGCCTAGGTGGCGAAATTGCAGATGTTCGGACCTTAGAGCCTAGGTGGCGAAAT
TGCAGAGTTGACCTTAGAGCCTAGGTGGCGAACCCATTGCAGAGTTCGGACCTTAGAGCCTA
GGTGGCGAAATTGCAGAGTTGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTTCGGACCT
TAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTCGGACCTTAGAGCCTAGGT
GGCGAAATTGCAGAGTTGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGA
GTTGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGATT
GCAGAGTTGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTTCGGACCTTAGAGCAGAGCCTAGGTGGCGAAATTGCAGAGTT
TTAGAGCCTAGGTGGCGGCCTAGGTGGCGAAATTGCAGAGTTTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGCCTTAGAGCCT
AGGTGGCGAAATTATCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGAGCCTAGGT
GGCGAAATTGCAGAGTTTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAG
AGCCTAGGTGGCGAAATTGCAGAGTTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTT
GACCTTAGATGCCTAGGTGGCGAAATTGCAGAGTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAG
TTGAACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGAG
CCTAGGTGGCGAAATTGCAGAGTTCGGACCTTAGAGCCTAGGTGGTCGAAATTGCAGAGTTGACCTTAGAGCCTAGGTGGCGAA
ATTGCAGAGTTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTTCG
GACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGAGCCTAGGTGGCGAAATTGCAGATGTTCGGACCTTAGAGCCT
AGGTGGCGAAATTGCAGAGTTGACCTTAGAGCCTAGGTGGCGA
ACCCATTGCAGAGTTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGAGC
CTAGGTGGCGAAATTGCAGAGTTTCGGACCTTAGAGCCTAGGTGGCGAAATTGCAGAGTTGACCTTAGAGC
Genetic Variability
99,9% identical
0,1% different
> 10 Million Sequence variants
Different Types of DNA Sequence
Variation
 Differences in the copy number of repeated sequences
(Tandemly repeated DNA):
Satellite DNA (large arrays)
Minisatellite markers (VNTR)
Microsatellite Markers: di-, tri-, tetra- nucleotide markers
 Interspersed repetitive DNA (retrotransposons):
LINEs Long Interspersed Nuclear Elements (Alu)
SINEs Short Interspersed Nuclear Elements (L1)
transposable elements
 Small Insertions and Deletions (Copy Number Variations or CNVs)
 Single nucleotide polymorphisms (SNPs) markers:
90% human genome variations are single nucleotide changes
Sequence variations
Mutation
– Rare sequence variants
Presence in the population:
<1%
– Great influence on disease
development
Polymorphisms
– Common sequence variants
Present in the population:
>1%
– Weak of no influence on disease
development
Single Nucleotide Polymorphism
(SNP)
Consequences of Mutation
Genetic Variability:
adaptation, evolution
Deleterious mutation:
Inborn error -- embryonic lethal.
Germ cells -- hereditary disease
Somatic cells -- malfunctioning of genes
malignancies, atherosclerosis, etc.
What is a SNP?
Example order of bases in a section
of DNA on a chromosome:
...C C A T T G A C...
…G G T A A C T G...
...C C G T T G A C...
…G G C A A C T G...
Some people have a different
base at a given location
Classification of SNP by
location
• Coding region:
Synonymous: mutation does not change amino acid.
Non-synonymous: mutation changes amino acid seq.
i.e. rare mutations that cause Mendelian diseases
with allele frequencies below 1%.
• Non-coding region:
5’ and 3’ UTR’s
Introns
Intervening Space
GENOME-WIDE SNP ARRAYS
Affymetrix Scanner
Illumina Bead Station
SNP Array
SNP arrays have developed very rapidly:
100K
2004
320K
2005
550K
2006
1000K
2007
2.4M
2008
5M
2010
Costs $
Size: 50K
Year: 2002
Time (years)
Offer the advantage of very dense coverage with over 5,000,000
data points across the genome.
Genome-wide association study publications
Annual number of publications
600
500
400
300
200
100
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
Years
The Trend & Applications of SNP markers
Diseased
Families
Linkage Studies
Diseased Families
Isolate Populations
Case - Control cohorts
Caes – Case studies
Candidate Regions
Positional
Cloning
Candidate
Genes
Functional
Studies
Single Gene Disease
Drug Targets/Diagnostics
Association studies
Candidate Genes or genome-wide
searches
Complex Diseases
Drug Targets
Diagnostic/Prognostic/Risk
Markers
Prevention
Pharmacogenetics
Subject Ascertainment and Design
Case-Control Design – most common
Case selection –
minimise phenotypic heterogeneity
Can be constrained by financial or operational
constraints
or
Focus on extreme and/or familial cases
This could actually reduce power of detection
Control Selection
• Optimal selection remains controversial
• Some concerns over-stated (common
controls do work WTCCC and Iceland)
• Need to be aware of:
» Selection bias
» Misclassification bias
» Population stratification
Other Case-Control Design
Issues
• Sample Size – the more the better
• Population stratification and cryptic
relatedness
(potentially inflate type 1 errors) Matching of
cases of controls is essential
• Merits of Family-based and case-control
association methods (good in reducing
stratification problems but comes at a cost of
reduced power).
• Use of historical controls (storage and data
acquisition problems)
Mapping Disease Genes
microsatellites
disease gene
chromosome
genes
SNPs
• Look for genetic linkage of disease to marker
• Microsatellite markers are too widely spaced to
identify individual genes.
• There are common SNPs in every gene.
• Need to identify those that are informative
Linkage Disequilibrium (LD)
• LD occurs when a marker (microsatellite, SNP
etc.) segregates with a disease gene.
• Extremely important for mapping studies.
• The likelihood of LD is a function of the
distance between a marker and the gene of
interest (ie. the closer a marker is to a disease
gene the less likely a recombination event will
occur).
Marker Selection and Assay
Design
• Marker selection: less of a problem due to
increased marker density arrays and the
dominance of companies involved in array
production
• DNA pooling: reduces costs but decreased
power and less accuracy - becoming an
irrelevant approach
• Robustness of the genotyping data: HWE
often used but imprecise for QC purposes. It
can be indicative of an association if modestly
present in cases
No association
Quantile-quantile
plots
Signal Intensity
plots
Manhattan Plot
Well defined
Stratification of
relatedness
Suggestive excess
Compelling evidence
of strong association of strong association
Allele calling error Genotype call failure Biased estimates
Validation and Replication
• Importance of replication
Technical Validation – re-analysis of
original samples – make sure the
genotypes called are as called
Replication uses a NEW sample set
that ideally includes both cases and
controls
Optimal GWAS Design
• GWAS on large sample set of Cases and
Controls
• Replication of the findings using a second
assay platform (assay validation)
• Replication of the study using a second
set of cases and controls (do not need a
2nd GWAS study)
SNP success stories
Coronary Artery Disease
Colon Cancer
Multiple Sclerosis
Breast Cancer
Celiac Disease
Autism
Prostate Cancer
Osteoporosis
Systemic Lupus Erythromatosis
Schizophrenia
rheumatoid arthritis
A GWAS for Pain
• Must have a reliable measurement of pain
(do not rely of self reporting)
• Identify non-genetic causes of pain
(personality and mood disorders)
• Response to treatment
• Age of onset
• Nature of onset
Summary
• Genetic studies required to identify new genetic
factors
• Case control studies for new biomarkers on risk
• Case case studies for pharmacological response
studies
• Gene expression studies to understand
mechanisms of pain
• Proteomics required to understand the effect of
genetic variation
Manhattan Plot from Turnball et al.
Relative Frequency of Disease
Caused Purely By Either Genetic
Or Environmental Influence
• Extreme environmental causes of
disease are very rare
• Genetic predispositions to any given
disease are also rare