Neurodevelopmental signature of a transcriptome-based polygenic risk score for depression

Abstract

Disentangling the molecular underpinnings of major depressive disorder (MDD) is necessary for identifying new treatment and prevention targets. The functional impact of depression-related transcriptomic changes on the brain remains relatively unexplored. We recently developed a novel transcriptome-based polygenic risk score (tPRS) composed of genes transcriptionally altered in MDD. Here, we sought to investigate effects of tPRS on brain structure in a developmental cohort (Adolescent Brain Cognitive Development study; n = 5124; 2387 female) at baseline (9-10 years) and 2-year follow-up (11-12 years). We tested associations between tPRS and Freesurfer-derived measures of cortical thickness, cortical surface area, and subcortical volume. Across the whole sample, higher tPRS was significantly associated with thicker left posterior cingulate cortex at both baseline and 2-year follow-up. In females only, tPRS was associated with lower right hippocampal volume at baseline and 2-year follow-up, and lower right pallidal volume at baseline. Furthermore, regional subcortical volume significantly mediated an indirect effect of tPRS on depressive symptoms in females at both timepoints. Conversely, tPRS did not have significant effects on cortical surface area. These findings suggest the existence of a sex-specific neurodevelopmental signature associated with shifts towards a more depression-like brain transcriptome, and highlight novel pathways of developmentally mediated MDD risk.

Keywords: ABCD study; Cortical thickness; Depression; Development; MDD; Neuroimaging; Polygenic risk score; Subcortical volume; Surface area; Transcriptomics.

Figure 1.. tPRS-associations with regional cortical thickness and cortical surface area

Both (a) and (b) were parcellated based on the Desikan-Killiany (DK) atlas. For (a), the plotted sample excludes two subjects with extreme values (CT < −6), exclusion of which did not substantially alter the plotted relationships (baseline with extreme values: r = 0.052, p < 0.001; 2-year follow-up with extreme values: r = 0.055, p = 0.001).

Figure 2.. Indirect effects of tPRS on CBCL withdrawn/depressed (‘depressive’) symptoms, as mediated by covariate-adjusted regional subcortical volume, in the female subsample (n=2319)

Associations with tPRS-by-sex interactions on covariate-adjusted subcortical volume (a) were identified Subcortical volume was parcellated based on the Desikan-Killiany (DK) atlas. Regression coefficients are unstandardized, and 95% confidence intervals are in brackets. HPC.R = right hippocampus; PAL.R = right pallidum; SCV = subcortical volume.