Understanding Genetics of Schizophrenia

Carlos N. Pato, M.D. , Ph.D.

Professor and Chair of Psychiatry

Center for Genomic Psychiatry

Keck School of Medicine

University of Southern California

Lifetime Risk for Schizophrenia

Schizophrenia

• 0.5-1.0% General Population

• 10-15% If a parent or sibling (including dizygotic twin) is schizophrenic

• 40% If both parents are schizophrenic

• 45-75% If monozygotic twin is schizophrenic – (Same risk to children- Fisher)

Genes (DNA) do not read the DSM-IV.

Phenotype Definition

A

C

H

D G

FE

J

B

Genetic Strategies

• What is linkage ?– What is a LOD score or a NPL ?

• What is an association ?– How to understand statistical significance.

• Candidate genes

• Genome wide scans (genomic mapping)

• Gene expression arrays

Schizophrenia on Chromosome 5q

Region rich in candidate genes

• Glutamate receptor-GRIA1• GABA cluster-GABRA1, GABRA2

• Serotonin Receptor-HTR4• Glycine Receptor-GLRA1

• Glucocorticoid receptor-NR3C1• Adrenergic receptor-ADRB2

• Neuregulin-NRG2• Kinase-CAMK2A

AA BB AB

Human Mapping Assay

A rapid, reliable and cost-effective assay for simultaneously genotyping many thousands of SNPs distributed across the genome

Generic complexity reduction scheme

Hybridization-based allele discrimination

Accuracy >99%Accuracy >99%

Transcript probes with the highest diagnostic utility for distinguising BP, SCZ, and control subjects

Title Symbol LocationCDP-diacylglycerol--inositol 3-phosphatidyltransferase (phosphatidylinositol synthase) CDIPT 16p12.1chemokine (C-X-C motif) ligand 3 CXCL3 4q21chemokine (C-C motif) receptor 1 CCR1 3p21split hand/foot malformation (ectrodactyly) type 3 SHFM3 10q24KIAA0082 KIAA0082 6p21.2

The expression patterns of the 35 most predictive genes correctly classified all BP and control subjects and 27 of 33 SCZ subjects.

Genome-wide Survey of CNVsNature (2008)

• 3380 patients with schizophrenia and 3139 ancestrally-matched controls

• identified three regions• large (>500kb) deletions increase disease risk• Deletions easier to detect because of their size

and replicability compared to single point mutations (single SNP)

Genome-wide Survey of CNVsNature (2008)

• On chromosome 22q11.2 - identified deletions in ~0.3% of schizophrenia patients (P =0.00056 versus controls)

• Odds ratio = 21.6

Genome-wide Survey of CNVsChromosome 22 deletion

Genome-wide Survey of CNVsNature (2008)

• On chromosome 15q13.2- identified deletions in ~0.3% of schizophrenia patients (P =0.00056 versus controls)

• Odds ratio= 17.9

Genome-wide Survey of CNVsChromosome 15

Genome-wide Survey of CNVsNature (2008)

• On chromosome 1q21.1 - identified deletions in ~0.3% of schizophrenia patients (P =0. 0.024 versus controls)

• Odds ratio= 6.6

Genome-wide Survey of CNVsChromosome 1

Genome-wide Survey of CNVsNature (2008)

• In the same issue of Nature, a parallel paper by DeCode showed the same results

• Extremely strong evidence for these relatively rare mutations

• Proves the necessity for extremely large studies

• Common polygenic variation contributes to risk

• Rare variants likely to contribute to risk

• major histocompatibility complex strongly replicated association with schizophrenia

MHC and Common VariantsNature, 2009

Genomic Psychiatry Cohort

• We have established the goal of studying 30,000 patients and 30,000 controls

• Schizophrenia and Bipolar disorder

• We have begun to bring together the funding for this large-scale program

• The NIMH launched this program with an initial $25 million dollars in grants to USC and MGH/Broad

Future Directions

• Gene identification

• Gene expression

• Proteomics

• Treatment development• At risk studies- with a focus on development of

pre-clinical diagnosis + treatment