NA-MIC National Alliance for Medical Image Computing http

NA-MIC National Alliance for Medical Image Computing http: //na-mic. org Core 3. 2 Activities (Toronto) University of Toronto—Neurogenetics Section, Center for Addiction and Mental Health James Kennedy MD Head Neurogenetics Section CAMH Aristotle Voineskos MD Chief Resident in Psychiatry CAMH Natalia Potapova MSc Neurogenetics Database Coordinator Claudia Rothe MD Research Fellow Neurogenetics CAMH Fabio Macciardi MD Ph. D Statistical Genetics Consultant Natalie Bulgin MSc Lab Coordinator Neurogenetics

Datasets Available • (1) Toronto genetic data on 300 schizophrenic patient and matched controls; • (2) Vancouver, 47 first episode schizophrenia patients with structural MRI scans, cognitive testing, and genetic; and • (3) Irvine 25 schizophrenic patients with f. MRI, PET, EEG and 100 k SNP genetic data. • How do we integrate the parts of Core 3. 2? National Alliance for Medical Image Computing http: //na-mic. org

Genetics of Schizophrenia: Neuroimaging Measures James L. Kennedy MD FRCPC Head, Neurogenetics Section, Director, Neuroscience Research Dept Centre for Addiction and Mental Health; I’Anson Professor of Psychiatry and Medical Science, University of Toronto

Genetics Basics I • • • Cell => nucleus => chromosome => gene Genome = all the chromosomes/genes Gene structure: coding, promoter, UTR Gene variant = allele Pair of alleles = genotype Mutation = rare variant <1% that disrupts gene function • Combination of variants = haplotype

Success of Molecular Genetics Single Gene Diseases (Mendelian) Complex Diseases • More than 600 disease genes discovered & screening now routine for many of these • Early onset Alzheimers presenilin genes pointing to new medications • Early onset breast cancer: diagnostic genetic tests • Cystic fibrosis: holdup in delivery of healthy gene • Huntingtons: toxic triplet repeat • Common Alzheimers: APOE gene predicts part of the risk • AIDS: the CCR 5 gene predicts progression of disease • Insulin dependent diabetes: HLA gene site is pointing to autoimmune mechanism • Coronary artery disease: multiple risk genes indicated • ADHD: dopamine D 4 and transporter genes associated

Molecular Genetic Approach Gene Expression Pharmacogenetics Variants Pharmacology Neurobiology Phenotype Psychophysiology Endophenotype Sub-pheno Neuroimaging

Will DNA Microarrays provide the important answers?

Microarrays: 1, 000 genetic tests in two hours

John Nash: 30 Years of Schizophrenia

Cytoarchitectural abnormalities Control Schizophrenia Comparison of hippocampal pyramids at the CA 1 and CA 2 interface between control and schizophrenic. Cresyl violet stain, original magnification X 250 Conrad et al. (1991) Arch Gen Psychiatry

BDNF Gene Structure Val-66 -met (GT)n repeat (promoter function? )

Transmission Disequilibrium Test for BDNF val 66/met in Schizophrenia * * TDT for val 66 c 2 = 4. 31; 1 df; p = 0. 03 Muglia et al, (2002) + S-TDT was used for the Canadian sample to incorporate the available affected sibs

Haplotype TDT: BDNF (GT)n repeat & val 66 met in schizophrenia * * HTDT for 170 -val 66 c 2 = 7. 11; 1 df; p = 0. 007 Muglia et al, (2002)

BDNF val 66 met: MRI functional brain imaging (Egan et al, Cell 2003) The red/yellow areas indicate brain regions (primarily hippocampus) that function differently between val/val (n=8) and val/met (n=5) subjects while performing a working memory task. Subjects with the met allele had more abnormal function.

Pezawas et al, 2005. Grey matter volume reduced in BDNF met carriers in hippocampus, R DLPC, and L frontal convexity.

BDNF and Hippocampal Volume (Szeszko et al, 2005) -- People with val/val homozygotes had higher hippocampal volume than val/met (f=8. 72, df=39, p=0. 005; sex and total brain volume as covariates). -- effect was larger in schiz patients than controls (partial eta sq = 0. 44 vs 0. 10)

NMDA as a final common path for Schizophrenia? BDNF (11 p) Neuregulin (8 p) Muglia et al, 2003 Syntaxin (7 q) Wong & Kennedy, in prep. Dystrobrevin (6 p) Glu Ca++ Straub et al, 2002 Irish families Chumakov et al, 2002 G 72 (13 q) Canadian & Russian families -also now +ve in Bipolar familes + D-amino acid oxidase (12 p) Current Medications Erb. B 4 Stefansson et al, 2002. Icelandic Scottish German Gly Dopamine D 2 D 1 calcineurin Tonegawa group, PNAS July 8/03

DRD 1, PET FDG, & Clozapine Response Genotype 2/2 BPRS = 30% Improvement Genotype 1/2 BPRS = 7% Genotype 1/2 Worsening (Potkin et al, 2002)

Common Serotonin Transporter Polymorphisms (17 q 11. 1 -q 12) 44 bp repeat (5 -HTTLPR) 5’ 17 bp repeat (VNTR) Exon 1 B Exon 2 3’ Promoter region Short (deleted)= C 14 Long (inserted) = C 16 Cells homozygous for the l allele have higher concentrations of 5 -HTT m. RNA and express two-fold 5 -HTT reuptake sites than s/s or s/l cells (Lesch et al, 1996); Similar differences found in 5 -HTT binding levels in human brain (Little et al, 1998; Lesch et al, 1998).

Serotonin Transporter Genetic Variation and the Response of the Human Amygdala Individuals carrying the s variant of the 5 HTTLPR exhibit an increased amygdala response to fearful stimuli when compared to those carrying the l variant. A. R. Hariri, V. S. Mattay, A. Tessitore, B. Kolachana, F. Fera, D. Goldman, M. F. Egan, D. R. Weinberger. Science, 297: 400 -403 (2002).

SNAP 25 Gene vs Grey Matter • Synaptic transmission activity partly determines extent of grey matter • SNAP 25 is a major component of synaptic vesicle release of neurotransmitters • SNAP 25 associated with schizophrenia (Potkin sample) and with treatment response (Muller et al, in press) • Role in grey matter volume?

EXTRACTING DATA FOR ANALYSIS Data are returned in a format suitable for association-type studies (m-link or casecontrol). Additional formats may be designed as needed (such as vertical haplotypes { 1 2 2 1 1 2 2 2 } ). Data may be transcribed and converted to document 2 1 1 1 2 2 1 1 formats supported by the analysis program (tab de-limited text, etc…). With access to source codes, or by invoking special features in downstream applications, the database can include automated running of analyses or transfer of data to other spreadsheets/databases.

Genetics and Grey Matter

Will the Brain Derived Neurotrophic Factor (BDNF) Gene Predict Grey Matter Volume? BDNF-1 SNP BDNF-2 Exon 11 Val-66 -met (GT)n repeat (function? m. RNA stability) BDNF-3 BDNF-4

BDNF val 66 met: MRI functional brain imaging (Egan et al, Cell 2003) The red/yellow areas indicate brain regions (primarily hippocampus) that function differently between val/val (n=8) and val/met (n=5) subjects while performing a working memory task. Subjects with the met allele had more abnormal function.

Haplotype TDT: BDNF (GT)n repeat & val 66 met in schizophrenia * * HTDT for 170 -val 66 c 2 = 7. 11; 1 df; p = 0. 007 Muglia et al, (2002)

SNAP 25 Genotype in Schizophrenia vs Controls Not for distribution Chi-sq = 9. 4; df=2; p=0. 009 Potkin sample

Genetics and White Matter

Myelin Oligodendrocyte Glycoprotein (MOG) Ü may function as: ð a cellular adhesion molecule ð a regulator of oligodendrocyte microtubule stability ð a mediator of interactions between myelin and the immune system, particularly as an activator of the classical complement cascade via activation of C 1 q (Johns and Bernard, 1997).

Location of MOG Gene in 6 p 21. 3 Region (MHC Region) DR DQ DO LMP/TAP DM DN DP C 4 A, C 4 B, C 2, factor B, 21 -OHase Class II NOTCH 4 Class III TNF HLA-G HLA-A HLA-C HLA-B MOG GABABR 1 Histone Family DTNBP 1 SCA 1 Class I telomere centromere (CA)n (TAAA)n ~ 2. 6 Mb Figure 2. Human MHC region and genes within the region.

Prefrontal f. MRI activity & myelin reduced in schizophrenia: Core 3. 1 Figure 3: 1 -4: Statistical parametric maps of the fractional anisotropy (FA) (left) and Magnetic Transfer Ratio (MTR) (myelin) (right) group comparison. Similar areas in yellow on both maps correspond to the location of both the internal capsule and prefrontal white matter, and indicate smaller values of FA and myelin in schizophrenia patients (n=14) compared with controls (n=15).

Will MOG gene variants predict white matter abnormalities? (CA) repeat Promoter region C 1334 T Start codon C 10991 T (TAAA) repeat Coding region (diagram not to scale)

MOG vs Total Brain White Matter • Sample: Dr. Honer UBC – 47 schiz, 24 cont • Phenotype: automated output from standard structural MRI – total grey and white matter • MRI=> 3 D SPGR: FOV 26 cm TE 11. 2 ms TR 2. 1 ms Matrix 256 Thickness 1. 5 mm Angle - perpendicular to AC-PC line Acquisition time - 6 minutes Not for distribution • C 1334 T marker genotype associated with white matter volume (P=0. 003) • Other MOG markers negative • All MOG markers negative for total grey matter volume

MAG (Myelin Associated Glycoprotein) • Reduced expression of MAG in postmortem studies of schizophrenic brain (frontal, cingulate, hippocampus) • Found in periaxonal membrane of oligodendrocyte and may be responsible for maintaining contact between myelin forming cells and the axon • Is a transmembrane adhesion molecule (CAM) present in both CNS and PNS myelin • Wan et al (Neuroscience Letters, 2005)

Myelin Associated Glycoprotein Gene (MAG) Yang, 2005

MAG SNPs • Two single nucoleotide polymorphisms examined located in intron 8 • Rs 720309 (T/A) • Rs 720308 (G/A) • Both SNPs significant in Chinese schizophrenia sample, and haplotype significant as well • Both SNPs associated with total white matter volume in schizophrenia cases, but not in controls

MAG rs 720309 (T/A) associated with White Matter Volume in Psychosis Cases mm 3 P = 0. 016

MAG rs 720308 (G/A) associated with White Matter Volume in Psychosis Cases mm 3 P = 0. 002

Rs 720308 and rs 720309 not associated with total white matter volume in controls (1, 2 = GA), (2, 2 = AA) (1, 1 = TT), (1, 2 = TA)

CNP SNP not associated with total white matter in early psychosis cases • 2’, 3’-cyclic nucleotide 3’-phosphodiesterase • Reduced expression reported in schiz brain • Knockout mice display schiz-like CNS path (reduced brain size, ventricle enlargement, corpus callosum atrophy) • Primary fxn may be in microtubule assembly, and cytoskeletal protein interaction • Region of interest assoc may be more fruitful (e. g. corpus callosum)

How can we move things forward collaboratively? • • • Region of interest data (grey matter) Region of interest data (white matter) f. MRI DTI Find ways to understand how genetic variation contributes to differences in these imaging measures

Imaging Measures to Combine Genetic Data with • Region of interest data – Grey matter (e. g. BDNF hippocampal volume) – White matter (e. g. CNP corpus callosum volume) – f. MRI measures (e. g. COMT gene with working memory as per f. MRI) All in normals and in cases

Imaging Measures to Combine Genetic Data with DTI • Fractional Anisotropy (a measure of the fraction of the magnitude of the tensor that can be ascribed to anisotropic diffusion) • e. g. MAG or MOG with FA in normals and in cases; can look at FA of cingulate bundle as it is known MAG is downregulated in cingulate cortex of schizophrenia cases

How can we adapt the current toolkit to integrate genetics ? • Incorporate our database • Make easy links/button clicks in Slicer such that genetic data can be added for each subject • Integrate statistical programs or links within slicer such that a simple ANOVA/ANCOVA can be carried out

Genetics Summary • SNAP 25 gene associated with schizophrenia in Potkin sample, and Toronto sample • BDNF gene candidate for grey matter vol and fxn • Serotonin transporter gene for amygdala function • DISC 1 gene for cortical thickness • Dopamine genes predict cortical & striatal fxn? • Newest data: MOG and MAG genes associated with total brain white matter (MOG hypothesized in grant app) • Relational database developed for organizing genetic + clinical + imaging data • Training available in genetics National Alliance for Medical Imaging and Computing