Objectives: Variants of the secreted glycoprotein, proprotein convertase subtilisin/kexin 9 (PCSK9), associate with both hypo- and hyper-cholesterolemic phenotypes. Herein, we carried out full exonic sequencing of PCSK9 documenting the frequency of single and multiple PCSK9 variations and their effects on serum lipoprotein and PCSK9 levels in Caucasian Canadians.
Methods: The 12 exons of PCSK9 were sequenced in 207 unrelated Caucasian Canadians. Minor allele frequencies of PCSK9 variants were compared amongst LDL cholesterol (LDLC) quintiles. Serum PCSK9 levels were measured by ELISA and lipoproteins by enzymatic methods. Comparisons were made with a Caucasian family cohort (n=51) and first generation African Canadians (n=31).
Results: In Caucasians, but not African Canadians, the c.61_63insCTG (denoted L10Ins) and A53V PCSK9 variations were linked and their frequency was significantly higher among Caucasian Canadians with LDLC levels in the <25th percentile. In both the unrelated and family Caucasian cohorts those carrying the L10A53V PCSK9 variant had significantly lower LDLC without reduction in plasma PCSK9. The I474V PCSK9 variant associated with significantly lower serum PCSK9 and LDLC. A novel PCSK9 variant was identified; E206K. We found that the frequency of multiple PCSK9 variations was higher in first generation African Canadians.
Conclusions: We showed that the L10A53V and I474V PCSK9 variants were significantly associated with lower LDLC levels in Caucasian Canadians but differed in their effect on serum PCSK9 concentrations, illuminating differences in their mechanism of inaction and indicating that that PCSK9 measurement alone may not always be a good indicator of PCSK9 function.Full exonic sequencing of PCSK9 pointed to factors that may contribute to L10Ins PCSK9 variant loss of function in Canadians of Caucasian but not those of African descent. These included; (1) its tight linkage with the A53V variant in Caucasians and/or (2) for both the L10 and I474V, the combined (and negating) effect of multiple, differing phenotypic PCSK9 variants within individuals of African ancestry for which combinations of PCSK9 variations and their overall frequency was higher. No population studies, to our knowledge, have addressed or accessed the effect of multiple PCSK9 variants on cholesterol profiles. Our results indicate that this should be considered.