Genome-wide copy number variation in schizophrenia

Authors:

T. Lencz1,2,3, C. Lambert4, T. Vance Morgan5, J. M. Kane1,2,3, R. Kucherlapati5,6, A. K. Malhotra1,2,3

1The Zucker Hillside Hospital, Glen Oaks, NY, United States, 2Albert Einstein College of Medicine, Bronx, NY, United States, 3The Feinstein Institute for Medical Research, Manhasset, NY, United States, 4Golden Helix, Inc., Bozeman, MT, United States, 5Harvard Partners Center for Genetics and Genomics, Cambridge, MA, United States, 6Harvard Medical School, Boston, MA, United States.

Abstract/Background:

Recently, considerable attention has focused on the surprisingly common degree of structural variation in the human genome. While excess rates of psychiatric illness have been reported in cytogenetically-defined chromosomal alterations, whole-genome microarray technology has not yet been utilized to identify sub-microscopic structural variation underlying psychiatric phenotypes. We report result of a novel approach to examining genome-wide structural variation in a case-control schizophrenia (SZ) cohort.

Methods:

SZ cases (65F/113M) and controls (63F/81M) were examined. All subjects self-identified as Caucasian non-Hispanic; testing of 210 ancestry informative markers revealed no population stratification. Single nucleotide polymorphisms (SNPs) were assayed using the Affymetrix 500K array. Quality control procedures yielded mean call rates of 97%, reliability (concordance across repeated samples) > 99%, and 439,511 high-quality SNPs available for analysis. A novel algorithm was implemented to identify and quantify extended runs of homozygosity (ROHs), which can signify chromosomal microdeletions and/or ancestral autozygosity.

Results:

A total of 339 “common” ROHs (observed in >3% of subjects) were identified across the 22 autosomes. In genome-wide analysis, SZ patients had significantly more ROHs than controls (32±12 vs.28±13, p=0.009). Frequency of nine specific ROHs significantly differed between patients and controls at a nominal p<.01; all demonstrate greater frequency in cases. Two prominent ROHs overlapped genes for DTNBP1 and DISC1 binding partners.

Conclusions:

To our knowledge, this is the first study of structural variation in schizophrenia using genome-wide SNP data. Results suggest that extended homozygosity, and possibly microdeletions are both more frequent in schizophrenia and specific to critical risk loci.