Transcript DNA and Gene Expression

Schizophrenia
Chapter 10
Psychopathology
• Sooner or later, half the population suffers a
serious run-in
– Reactive or endogenous
• Very expensive
• Historically, known to run in families
Schizophrenia
• Lifetime risk ~1%
• Long-term thought disorder
– Cognitive, emotional, motivational
•
•
•
•
Chronic and acute forms
Late adolescence to early adult onset
Earlier onset, the worse the prognosis
Very expensive (10% of homelessness)
Schizophrenia
• Disorganized thinking, hallucinations,
delusions, blunted affect, poverty of speech,
lack of motivation
• Subtypes
– Paranoid, Disorganized, Catatonic,
Undifferentiated
Family Risk
• Runs in families
• As r-value increases, so does probability of having
schizophrenia
• But not strictly hereditary
–
–
–
–
–
–
Cousins, 4%
Siblings, 9%
Dizygotic twins, 17%
Monozygotic twins, 48%
Parents of a schizophrenic, 6%
Offspring of schizophrenic parent, 13%
Diversity
• Risk is heritable, but subtypes (e.g., catatonic vs.
paranoid) are not
– So, is this one disorder or several?
• More severe forms are more heritable
• Type-I (hallucinations) have better prognosis;
type-II (withdrawal, blunted affect) is more severe
and more heritable
• Some symptom dimensions (e.g., disorganization)
more heritable than others
Diathesis-Stress Model
• Individual may have genes for
schizophrenia, but only phenotypically
express them after particular stressful life
event
• Genes create predisposition
Identification of Genes
•
•
•
•
Very active area of research
Numerous candidates genes proposed
Highly polygenic
So far, primary candidate genes proposed on
chromosomes 2, 5, 6, 8, 13, 22
• Results very difficult to interpret, though
NOTCH4
•
•
•
•
On chromosome 6p (also containing some HLAs)
Member of a type 1 transmembrane protein family
Receptor for membrane bound ligands
Roles in variety of developmental processes by
controlling cell fate decisions, including postmitotic differentiation of cortical neurons
• Evolutionarily conserved intercellular signaling
pathway (e.g., homolog to human gene in
Drosophila)
NOTCH4
• Wei & Hemmings (2000)
• Association of schizophrenia with 4 SNPs in the
NOTCH4 gene
• Highly significant association results for 3 of the
SNPs in a sample of 80 British parent-offspring trios
Follow-up Studies
• Only two with Caucasians (Prasad et al., 2004;
Skol et al., 2003) showed weak to modest
associations for 2 of the SNP markers, and
• Six studies with Caucasians (e.g., Anttila et al.,
2003; Luo et al. 2004) found no associations;
• Two studies of African samples found associations
(Luo et al. 2004; Skol et al. 2003)
• But, these studies focused only on a small portion
of the gene (5´ promoter region and first exon);
additional sites throughout gene need study
Liu et al. (2007)
• Scanned entire genomic region of NOTCH4
gene using 14 SNPs, 7 of which were
validated (minor allele frequency over 10%)
• Used relatively large family sample of
schizophrenia (218 families with at least
two affected siblings)
• Asian sample
Results
• Evidence of association with distal genomic
region of NOTCH4
• Failed to find association with 4 SNPs from
5´ region
• Weak to modest association found for a few
other previously reported SNPs throughout
the gene
Why Such Poor Replication
• Clinical and genetic heterogeneity of schiz.
• Ethnicity
– Caucasian and African samples show some evidence for
proximal region of NOTCH4
– Asian samples don’t, but show evidence for distal
region of NOTCH4
• Could be two or more disease-underlying variants
at NOTCH4 locus (see Zhang et al. 2004)
SNPs and Ethnicity
• Each SNP has its own genetic ancestral
heritage
• Frequencies of each variant may differ
between ethnic populations
• Remember, synonymous vs. nonsynonymous SNPs
COMT
• Catechol-O-methyl transferase
• Enzyme degrading dopamine, epinephrine,
norepinephrine
• A non-synonymous SNP (substituting valine for
methionine) affects cognitive tasks (set shifting,
set inhibition, abstract thought) by reducing
dopamine at four times the regular rate
• Neurons with mutation need higher levels of
activation to release normal levels of dopamine
post-synaptically
Interesting Interaction
• Caspi et al. (2004)
• Cannabis use linked to twofold increase in
late onset schizophrenia
• Majority of young users do not develop any
psychosis
• Suggests some individuals vulnerable to its
effect
• Gene-environment interaction
Candidate Gene
• COMT gene on chromosome 22q11; region
implicated in genome scans for schizophrenia
• Dopaminergic function disturbances
implicated in schizophrenia (e.g., Kapur 2003)
• Valine (V) for methionine (M) substitution
• Genotypes: M/M (“wildtype”) has lowest
COMT activity, V/V highest, M/V
intermediate (co-dominant alleles)
% with schizophrenia
disorder at age 26
Outcome
20
15
10
5
0
M/M
V/M
V/V
COMT genotype
No adolescent cannabis use
Adolescent cannabis use
And so…
• Evidence that adolescent, but not adult, cannabis
use is associated with schizophrenia through V-M
polymorphism
• May be limited to a sensitive period of brain
development
• Study is not identifying a major cause of
schizophrenia
• Supports COMT V-M functional polymorphism as
having a role in psychosis, but perhaps only in
context of exposure to environmental pathogens
Estrogen
• Sex differences in schizophrenia
• Affected females have better course of disease
than males
• Interestingly, women show a second peak-of-onset
post-menopause (not seen in men)
• Clinical observations of affected women show
increased symptoms when estrogen levels are low
(pre-menstrual, post-partum, post-menopause) and
remission of symptoms when estrogen high (e.g.,
pregnancy)
Martorell et al. (2008)
• Study to evaluate hypothesis that estrogen
receptor genes ESR1 and ESR2 are
involved in schizophrenia onset
• Also genes APOE and COMT, both
regulated by estrogen receptors could also
be involved
Method
• Analyzed SNPs
– 26 in ESR1, 14 in ESR2, 7 in APOE, 12 in
COMT
• Allele frequencies evaluated in 585
schizophrenics and 615 controls
• Both male and female subjects, but grouped
and sex-separated analysis performed
Findings
• Analysis failed to find any significant associations
between each of the candidate genes and the
diagnosis of schizophrenia
• A little unexpected, as various earlier animal
models indicate role of estrogen genes in
dopamine function
– Also, Shifman et al. (2002) found association between
3 COMT SNPs and schizophrenia in Ashkenazi Jews;
ethnic variability?
• Worth considering previous study’s findings that
exposure to environmental toxins may be relevant
factor
Zinkstok et al. (2008)
• Used MRIs to examine gray and white matter
density and volume differences between genotype
groups for COMT and another gene, PRODH
• 51 schizophrenic patients scanned and genotyped
• Hypothesize COMT and PRODH polymorphisms
may result in structural and functional brain
abnormalities
Findings
• A SNP in the promoter region of COMT is
associated with a gray matter increase in the right
superior temporal gurus
• Two nonsynonymous SNPs in PRODH gene
associated with reductions in white matter density
reductions in frontal lobes
• Interestingly, evidence for COMT and PRODH
epistasis
– Only patients with a COMT valine allele and one or
two mutated PRODH alleles showed increased white
matter density in left inferior frontal lobes
So Where Are We?
• Lots of suggestions for candidate genes and SNP
effects
• Inconsistent findings across many studies
• Very complicated
• Individual effects by genes most likely small and
highly variable due to variability of genotype,
environmental interactions, epistasis
• So far, results not terribly conclusive
• Is there another approach?
Endophenotype
• Psychiatric concept; biomarker
–
–
–
–
–
Associated with illness
Heritable
Stable over time
Within families, endophenotype and illness co-segregate
Found in relatives of affected individual at higher rate than
in general population
• Divide behavioural symptoms into more stable
phenotypes with clear genetic connection
• Overt symptom is a psychosis; underlying phenotype
are endophenotype. e.g., sensory gating and decline
in working memory (the endophenotypes)
Braff, Schork & Gottesman (2007)
• Argue for endophenotype strategy to understand
genetic architecture of schizophrenia
• The ultimate endophenotypes are perturbed levels
of specific proteins or gene expression
• But this level of analysis not yet possible for
schizophrenia
• So, use neurophysiological and neurocognitive
measures
– Ones that reflect more elementary aspects of the
biology than clinical features themselves
Gur et al. (2007)
• Examines computerized neurocognitive measures
as candidate endophenotypic markers
• Multiplex multigenerational families
– Beneficial because power to detect genes for
quantitative traits through linkage analysis increases
with family size (better than sibpairs)
– 349 European Americans from 35 multiplex
multigenerational families (58 patients, 291 relatives)
and 154 psychiatrically healthy European Americans
(control comparator group)
Approach
• Heterogeneity of schizophrenia at phenotypic and
genotypic levels
• Endophenotypes genetically simpler than disease
endpoints; an advantage (big one!)
• Can measure such quantitative parameters in family
members where clinical diagnosis may be absent or
difficult to gain
• If neurocognitive deficits are associated with genetic
liability, they should increase with relation to
affected individuals
Measures
• Diagnostic assessment
• Computerized neurocognitive “scan”
–
–
–
–
–
–
–
–
Abstraction and mental flexibility
Attention
Verbal memory
Face memory
Spatial memory
Spatial processing
Sensorimotor dexterity
Emotional processing
0.5
Accuracy
0.0
-0.5
-1.0
**
***
**
***
Mean z score
-1.5
***
***
***
***
***
***
***
-2.0
0.5
Speed
0.0
-0.5
*
**
-1.0
***
***
-1.5
-2.0
Abstraction/
flexibility
Attention
Comparison subjects (N=154)
Verbal
memory
Face
memory
Spatial
memory
Spatial
processing
Relatives of schizophrenic patient (N=291)
Sensorimotor
Emotion
identification
Schizophrenia patients (N=58)
*, **, ** p<0.5, p<0.01, p<0.001, respectively, vs. comparison subjects
Interpretation
• Schizophrenics show deficits on various
neurocognitive tasks
• Relatives (both 1st and 2nd degree) without
schizophrenia also show impairments in specific tasks
– Extended family tested; can analyze based on shared genes,
but won’t have shared environment
– Advantage of the multigenerational design.
– Parent-sib and co-sib studies can’t separate out shared
environment from certain genetic effects
• Supports neurocognitive measures as endophenotypic
markers of vulnerability to schizophrenia
• Target endophenotypic markers for gene identification
Hall et al. (2007)
• Event-related potentials as candidate endophenotypes
– P300 (reduced amplitude and prolonged latency)
– P50 (reduction of inhibitory response)
– Mismatch negativity (reduced amplitude)
• Objectives
– Estimate heritabilities for range of event-related potential
indices
– Quantify strength of relationship of each index with
schizophrenia
– Examine genetic and environmental overlap with the
illness
ERPs
• Reliably measured using electroencephalography (EEG)
• Deflections in voltage in brain activity after a stimulus
presentation
• Reflect “higher” cognitive processes
– Memory, expectation, attention, etc.
• Positive (P), negative (N)
– N400: negative voltage deflection 400ms after stimulus
– P300: positive deflection, 300ms delay (this one responds to
unexpected/novel/cognitively salient stimuli across
senory/perceptual domains)
Twin Design
•
•
•
•
16 MZ twins concordant for schizophrenia
9 MZ twins discordant for schizophrenia
45 normal MZ twins
32 normal DZ twins
Results
Correlation with Genetic
Factors in Schizophrenia
0.6
In MZ twins
concordant or
discordant for
schizophrenia (N =
50) and healthy
comparison twins
(N=154)
0.4
0.2
0
-0.2
-0.4
-0.6
Mismatch
P300
Negativity Amplitude
Amplitude
P300
Latency
P50
Suppression
Ratio
Breakdown
Type of Correlation
Mismatch
Negativity
Amplitude
Phenotypic correlation
0.31
-0.35
0.35
0.40
0.24
-0.35
0.18
0.40
Due to shared environmental factors 0.01
0.05
0.13
0.00
Due to unique environmental factors 0.06
-0.05
0.03
-0.01
Due to additive genetic factors
P300
Amplitude
P300
Latency
P50
Suppression
Ratio
P50
• Attractive ERP endophenotype
– Little environmental variance (except measurement
error)
– Highest phenotypic correlation with schizophrenia
– Almost all the correlation explained by genetic factors
• Measure of sensory inhibition in brain;
individual’s ability to filter out repetitive stimuli to
avoid information overload
• Linked to alpha-7 nicotinic receptor gene
(CHRNA7)