A schizophrenia risk gene, ZNF804A, influences neuroanatomical and neurocognitive phenotypes

Abstract

ZNF804A is one of the strongest candidate genes for schizophrenia (SZ), yet its function and role in disease pathophysiology are largely unknown. The only in vivo endophenotype study of the SZ-associated SNP (rs1344706) pointed towards effects on brain functional connectivity. We examined the relationship of this SNP to neuroanatomical and neurocognitive phenotypes that were assessed in healthy individuals. Volunteers with no history of psychiatric illness were assessed with structural magnetic resonance imaging (1.5T GE scanner, standard gradient-echo acquisition). Carriers of the minor allele were compared with homozygotes for the T (SZ-associated) allele on measures of total volume of the white matter (WM), gray matter (GM), and cerebrospinal fluid compartments, as well as on voxel-wise measurements of regional brain volumes. After examining the correlation between genotype-associated regions of interest and neurocognitive performance measures, the effects of rs1344706 genotype on a measure of visuomotor performance speed (trails A) were examined in an independent cohort of volunteers. Among healthy subjects, risk allele homozygotes showed larger total WM volumes than carriers of the other allele. Controlling for WM volumes, these same subjects showed reduced GM volumes in several regions comprising the 'default mode network,' including angular gyrus, parahippocampal gyrus, posterior cingulate, and medial orbitofrontal gyrus/gyrus rectus (FDR-corrected p<0.05). The risk allele dosage also predicted impairments on a timed visuomotor performance task (trails A). Results support a role of ZNF804A in phenotypes reflecting altered neural connectivity.

Figure 1

Covariate-corrected, segmented volumes for GM, WM, and CSF compartments in healthy subjects divided by rs1344706 genotype (G-allele carriers, n=21 vs T-allele homozygotes, n=18). Error bars represent twice the SE of the mean.

Figure 2

Covariate-corrected, segmented volumes for WM in healthy subjects divided by rs1344706 genotype (G-allele homozygotes, n=7; G/T heterozygotes, n=14; T-allele homozygotes, n=18). Error bars represent twice the SE of the mean.

Figure 3

Visualization of GM regions, which are significantly smaller in healthy volunteers homozygous for the T-allele at rs1344706 (n=18) compared with healthy G-allele carriers (n=21; FDR-corrected p<0.05; minimum of 100 contiguous voxels). All images depicted on the canonical MNI template provide in SPM5. Legends for region labels A–H are provided in the right-most column of Table 1.

Figure 4

Relationship between composite regional GM volume (expressed as a standardized metric) and performance on trail making test, part A (n=31 subjects with both measures available). Note that longer times indicate worse performance.

Figure 5

Relationship between rs1344706 genotype (G-allele homozygotes, n=22; G/T heterozygotes, n=67; T-allele homozygotes, n=47) and performance (time to completion) on trails A in healthy volunteers.