Investigation of the effects of non-coding LDLR variants on hyperlipidaemia risk

Coding and non-coding variants of the LDLR gene have been reported to cause familial hyperlipidaemia (FH) and rare coding variants have been shown to increase risk of hyperlipidaemia in the general population. Hyperlipidaemia status had previously been assigned for 470,000 UK Biobank participants in a study of rare coding variants. Effects of variants obtained from whole genome sequencing in the five-prime untranslated region (5’UTR) and promoter region of LDLR were analysed using this phenotype. Variants previously reported to be implicated in FH and associated with altered LDLR expression were entered into logistic regression analysis. Other variants in these regions with minor allele frequency <= 0.01 were entered into a weighted burden test. For this analysis, insertion-deletion variants were weighted according to rarity while single nucleotide variants were additionally weighted using two pathogenicity predictors, CADD and GPN-MSA. The previously reported common variant, rs17248720C>T, was confirmed to slightly reduce hyperlipidaemia risk. Two previously reported rare variants which were associated with markedly reduced LDLR expression, c.-101T>C and c.-121T>C, had at most only moderate effects on hyperlipidaemia risk. In the weighted burden analysis, promoter variants annotated with CADD showed evidence for association with increased risk (p = 0.00097). However the effect size was small, with variants having CADD scores exceeding 0.5 having OR = 1.07 (0.99 – 1.15). Although rare non-coding variants do have detectable effects on hyperlipidaemia risk, these effects are much smaller than those detected for coding variants in the same sample. This research has been conducted using the UK Biobank Resource.