ABCG5 and ABCG8 genetic variants in familial hypercholesterolemia

Background: Familial hypercholesterolemia (FH) is a common inherited disease characterized by elevated low-density lipoprotein cholesterol (LDL-C) plasma levels and increased cardiovascular disease risk. Most patients carry a mutation in the low-density lipoprotein receptor gene (LDLR). Common and rare variants in the genes encoding adenosine triphosphate–binding cassette transporters G5 and G8 (ABCG5 and ABCG8) have been shown to affect LDL-C levels. Objective: The objective of this study was to investigate whether and to which extent heterozygous variants in ABCG5 and ABCG8 are associated with the hypercholesterolemic phenotype. Methods: We sequenced ABCG5 and ABCG8 in a cohort of 3031 clinical FH patients and compared the prevalence of variants with a European reference population (gnomAD). Clinical characteristics of carriers of putative pathogenic variants in ABCG5 and/or ABCG8 were compared with heterozygous carriers of mutations in LDLR. Furthermore, we assessed the segregation of one ABCG5 and two ABCG8 variants with plasma lipid and sterol levels in three kindreds. Results: The frequencies of (likely) pathogenic LDLR, APOB, PCSK9, ABCG5, and ABCG8 variants in our FH cohort were 11.42%, 2.84%, 0.69%, 1.48%, and 0.96%, respectively. We identified 191 ABCG5 and ABCG8 variants of which 53 were classified as pathogenic or likely pathogenic. Of these 53 variants, 51 were either absent from a reference population or more prevalent in our FH cohort than in the reference population. LDL-C levels were significantly lower in heterozygous carriers of a (likely) pathogenic ABCG5 or ABCG8 variant compared to LDLR mutation carriers (6.2 ± 1.7 vs 7.2 ± 1.7 mmol/L, P < .001). The combination of both an ABCG5 or ABCG8 variant and a LDLR variant was found not to be associated with significant higher LDL-C levels (7.8 ± 2.3 vs 7.2 ± 1.7 mmol/L, P = .259). Segregation analysis in three families (nine carriers, in addition to the index cases, and 16 noncarriers) did not show complete segregation of the ABCG5/G8 variants with high LDL-C levels, and LDL-C levels were not different (3.9 ± 1.3 vs 3.5 ± 0.6 mmol/L in carriers and noncarriers, respectively, P = .295), while plasma plant sterol levels were higher in carriers compared to noncarriers (cholestanol: 10.2 ± 1.7 vs 8.4 ± 1.6 μmol/L, P = .007; campesterol: 22.5 ± 10.1 vs 13.4 ± 3.5 μmol/L, P = .008; sitosterol: 17.0 ± 11.6 vs 8.2 ± 2.6 μmol/L, P = .024). Conclusions: 2.4% of subjects in our FH cohort carried putative pathogenic ABCG5 and ABCG8 variants but had lower LDL-C levels compared to FH patients who were heterozygous carriers of an LDLR variant. These results suggest a role for these genes in hypercholesterolemia in FH patients with less severely elevated LDL-C levels. We did not find evidence that these variants cause autosomal dominant FH