Genetic analysis of the PCSK9 locus in psychological, psychiatric, metabolic and cardiovascular traits in UK Biobank

Abstract

The association between severe mental illness (SMI) and cardiovascular and metabolic disease (CMD) is poorly understood. PCSK9 is expressed in systems critical to both SMI and CMD and influences lipid homeostasis and brain function. We systematically investigated relationships between genetic variation within the PCSK9 locus and risk for both CMD and SMI. UK Biobank recruited ~500,000 volunteers and assessed a wide range of SMI and CMD phenotypes. We used genetic data from white British ancestry individuals of UK Biobank. Genetic association analyses were conducted in PLINK, with statistical significance defined by the number of independent SNPs. Conditional analyses and linkage disequilibrium assessed the independence of SNPs and the presence of multiple signals. Two genetic risk scores of lipid-lowering alleles were calculated and used as proxies for putative lipid-lowering effects of PCSK9. PCSK9 variants were associated with central adiposity, venous thrombosis embolism, systolic blood pressure, mood instability, and neuroticism (all p < 1.16 × 10^-4). No secondary signals were identified. Conditional analyses and high linkage disequilibrium (r² = 0.98) indicated that mood instability and central obesity may share a genetic signal. Genetic risk scores suggested that the lipid-lowering effects of PCSK9 may be causal for greater mood instability and higher neuroticism. This is the first study to implicate the PCSK9 locus in mood-disorder symptoms and related traits, as well as the shared pathology of SMI and CMD. PCSK9 effects on mood may occur via lipid-lowering mechanisms. Further work is needed to understand whether repurposing PCSK9-targeting therapies might improve SMI symptoms and prevent CMD.

Fig. 1. Tissue expression patterns of genes of interest.

Gene expression of A PCSK9, B DHCR24, and C USP24, by tissue (GTEx analysis release V8). Expression values are shown in transcription per million (TPM) calculated from a gene model with isoforms collapsed to a single gene. Box plots are shown as median and 25th and 75th percentiles; points are displayed as outliers if they are above or below 1.5 times the interquartile range.

Fig. 2. Regional plots of main results in white British ancestry individuals.

A SBPadj, (B) WHRadjBMI, (C) VTE, (D) stroke, (E) mood instability and (F) neuroticism score. Locuszoom regional plots of each significant SNP identified for a trait in the UKB. SNPs are aligned on the X-axis by their position on chromosome 9, and by their association with the trait on the Y-axis (P values are on a − log10). Significance was set at p < 1.16 × 10⁻⁴ or −log10 P > 3.93 (approximated by the red horizontal line). SNPs are colour coded by their estimates of pairwise LD (r2) with lead SNP.

Fig. 3. Linkage disequilibrium (LD) between lead SNPs.

Haploview LD plot showing the lead SNPs in the PCSK9 locus associated with cardiometabolic or psychiatric traits, and the SNPs included in PGS4 (*) and PGS7 (#). Each box provides estimated statistics of % frequency of coinheritance (r² values). Darker shaded boxes represent stronger LD (r² colours).

Fig. 4

Potential PCSK9 mechanisms proposed pathways of how genetic variation in the PCSK9 locus might influence cardiovascular and metabolic traits.