Why imaging genetics?
Motivation – Why Imaging
Genetics?
Hao-Yang TAN MD, FRCP(C)
Seymour S Kety Fellow
Clinical Brain Disorders Branch
National Institute of Mental Health
National Institutes of Health
Bethesda, Maryland, USA
1. To translate candidate brain mechanisms of
cognition and disease from basic neurobiology
to the live working human brain
2. To link imaging strategies with human genetics/
genomics towards potential rational treatment
development.
3. Discovery of new genes and brain mechanisms
to guide hypothesis-generation
tanh@mail.nih.gov
Genetic dissection of dopaminergic (DA)
mechanisms in human prefrontal-striatal circuits and
its relationship with schizophrenia
Working memory component processes
1. Working memory (WM) dysfunction in
schizophrenia
2. Examine predictions from neurobiology of DA
function and their links with human working
memory processes via interacting DA-related
gene effects on prefrontal brain systems: COMT,
GRM3, AKT1.
Chee & Choo, J Neurosci 2004
Tan et al, Am J Psychiat 2005
Neuroimaging of working memory components in
schizophrenia
Working memory, DA and genetics of
schizophrenia
• WM is critically implicated in cognitive deficits of SCZ ;
DA is implicated in WM and SCZ
• Twin studies: WM performance is genetically heritable
(Toulopoulou et al. Arch Gen Psychiat 2007), as are WM
brain activation patterns (Koten et al. Science 2009)
Task*group interaction
F1,20=5.4, p<0.05
p<0.001
Task*group interaction
F1,20=5.9, p<0.05
SCZ sometimes utilize greater prefrontal brain
activation from a larger network of prefrontal
regions to perform the WM tasks at the same
level of accuracy (inefficiency). How is DA related
to WM dysfunction in SCZ?
Tan et al. Am J Psychiatry 2005
• WM brain activation is also inefficient in unaffected
siblings of SCZ – brain activity is a potential intermediate
phenotype representing genetic risk of SCZ (Callicott et
al Am J Psychiat 2003).
1
DA mechanisms of WM – Basic and computational models
Studying human cortical DA function: Genetic variation in
COMT and differential cortical DA bioavailability
CHROMOSOME 22
1
22q11.23
22q11.22
27kb
PROMOTER
PROMOTER
5´
COMT-MB START CODON
TRANSMEMBRANE SEGMENT
COMT-S START CODON
Seamans et al. Prog Neurobiol 2004
NlaIII
STOP CODON
PCR
210 BP
5´-CTCATCACCATCGAGATCAA
5´-GATGACCCTGGTGATAGTGG
NlaIII
NlaIII
NlaIII
NlaIII
Dopamine in cortical synapses is regulated by COMT (absence of DAT in
cortical synapses)
…CGTG…
Vijayrajhavan…Arnsten.
Nat Neurosci 2007
O’Reilly, Science 2006.
thermo-stable
ancestral allele
thermo-labile
human allele
SOURCE: NCBI GEN-BANK ACCESSION # Z26491
COMT Val>Met effects on working
memory component processes:
Mier et al Mol Psychiat 2009
COMT/ DA effects on WM component processes:
…CATG…
(Lotta et al Biochem 1995; Lachman et al Pharmacogenetics 1996; Chen et al Am J Hum Genet 2004; Yavich et al J
Neurosci 2007)
Meta-analysis COMT
Val>Met effects in imaging
of PFC WM function
G1947 → A1947
⇓
Catechol-O-methyltransferace hasCOMT-MB/S:
a coding
variation (Val158Met) :
..AGV
KD.. Reduced
..AGMKD...
- High activity variant
(Val):
cortical
dopamine
Val158/108 →
Met158/108
high-activity
low-activity
- Low activity variant
(Met): Increased cortical dopamine
Tan et al. Journal of Neuroscience (2007)
Mid-brain F-DOPA uptake (DA synthesis) and
cortical DA activation (via COMT Val-Met)
Val
• Active maintenance of information in
working memory engages DA effects in
DLPFC
• Manipulating/ rapidly updating new
information in working memory engages
DA effects via DLPFC and striatum
MM
MM
• Note no direct D1 or D2 measurements here
Tan et al. Journal of Neuroscience (2007)
Chen Am J Hum Genet 2004
Meyer-Lindenberg Nat Neurosci 2005
2
COMT Val>Met on WM prefrontal cortical activation with
amphetamine
What about other genes that impact
dopaminergic human brain
function?
• DA system acts in concert with NMDA/
glutamatergic (and GABAergic) brain systems
Egan PNAS 2001
• GRM3 variation impacts SCZ risk, glial glut
transporter expression, PFC processing
efficiency (Egan et al PNAS 2003)
• Combined deleterious COMT & GRM3
variation?
Shen et al Science 2008
Mattay PNAS 2004
Biologic epistasis of COMT and GRM3 in-vivo
Intracellular signaling of dopaminergic
processes
cAMP-DARPP32 signaling
BA9: 45 20 32.
p<0.05 FDR
COMT*GRM*load
F1,25=4.47,
p=0.045
Tan et al Proc Nat Acad Sci USA (2007)
Greengard, Science 2001; Beaulieu et al, Neuron 2005, Trends Pharmcol Sci 2007
AKT1 expression in humans:
Lymphoblasts
AKT1 knockouts
performed worse
on WM but not
non-WM tests
11/11 12/12
Frontal cortex
22/22
Harris et al 2005 PNAS
Lymphoblasts – NIMH imaging
genetics study
(n=13, 19)
Lai … Gogos, PNAS 2006
Emamian et al Nat Genet 2004
Tan et al. J Clin Invest 2008
3
Main effect of AKT1 rs1130233 A-carrier > GG on working
memory
Main effect of AKT1 rs1130233 A-carrier > GG on working
memory
t
Peak -38 30 36, z=3.47, p<0.0003
uncorrected; p<0.05 FWE small volume
corrected for prefrontal regions (BA 9,
46, 44, 45, 47) activated by main
2>0back task effect
t
Displayed at p<0.005 uncorrected, n=46
t
Displayed at p<0.005 uncorrected, n=46
Displayed at p<0.005 uncorrected, n=68
Replication sample with peak at -34 49 30,
z=2.76, p<0.003 uncorrected, p<0.05 FWE
corrected for 20mm box centered at -38 30 36
peak in sample 1
Tan et al. J Clin Invest 2008
COMT Val/Val COMT Met/Met
AKT1 effect
Tan et al. J Clin Invest 2008
COMT Val/Val COMT Met/Met
AKT1 effect
depends on
COMT Val158Met
background
AKT1 effect
AKT1 effect depends
on COMT Val158 Met
background
n=46, parameter estimates
extracted from same -38 30 36
peak for main AKT1 effect
shows an COMT*AKT1
interaction F(1,42) = 4.466, p =
0.041
VV/GG
VV/A carrier
MM/GG
MM/ A carrier
Parameter Estimates at -38 30 36
Parameter Estimates at -38 30 36
From 2nd sample, same peak
–34, 49, 30 mm; F(1,64) =
5.703, P = 0.020;
VV/GG
VV/A carrier
MM/GG
COMT/AKT1 Genotype
COMT/AKT1 Genotype
Error bars are +/- 1 standard error
Error bars are +/- 1 standard error
Tan et al. J Clin Invest (2008)
MM/ A carrier
Tan et al. J Clin Invest (2008)
Conclusions:
1. Active maintenance and manipulation processes in
working memory are critically implicated in
schizophrenia and heritable
2. Dopaminergic genes influence differing levels of human
working memory brain processes. Cortical DA
processes are engaged in active maintenance of
information, while striatal-cortical DA processes in the
rapid updating and manipulation of information.
- Integrating multiple samples for genetic association, imaging genetics, cell biology
experiments to elucidate the genetic brain mechanisms relevant to cognition in scz
4
Conclusions
Global burden of disease and cognitive deficits
3.
Imaging genetics can begin to
elucidate predictable interactions
between dopaminergic,
glutamatergic and intracellular
signaling genes, thus integrate
basic and human studies
4.
Systematic imaging studies of
genes together with biologic
evidence also provides tractable
paths to novel treatment
hypotheses for cognitive
dysfunction in schizophrenia eg
COMT, GRM3, AKT1, BDNF,
KCNH2 have existing/pipeline
drugs that could be relevant to
such imaging pharmacogenetic
experiments in the future
CJL Murray and AD Lopez, Editors, The global burden of disease and injury
series, volume 1: a comprehensive assessment of mortality and disability from
diseases, injuries, and risk factors in 1990 and projected to 2020, Harvard
University Press, Cambridge, MA, USA (1996).
CBDB, GCAP, NIMH
Daniel R Weinberger
Danny Q Chen
Joseph H Callicott
Venkata S Mattay
Andreas Meyer-Lindenberg
Kristin Nicodemus
Richard Straub
Bhaskar Kolachana
Yoshi Sei
Jose Apud/ CBDB Sibstudy Team
Funding
• APA-Astra-Zeneca Young Minds in Psychiatry Award
• NIH Fellowship Award
• Seymour S Kety Memorial Fellowship for
Translational Research
• NIMH Intramural Research Program
Preprint of review elaborating points in this presentation can be
found at: http://mysite.verizon.net/thy1/haoyangtan
tanh@mail.nih.gov
5