See editorial by Paquette and Baass, pages 1259e1260 of this issue.

Proprotein convertase subtilisin/kexin type-9 (PCSK9) is a ligand of low-density lipoprotein (LDL) receptor (LDLR) that promotes LDLR degradation in late endosomes/lysosomes. In human plasma, 30–40% of PCSK9 is bound to LDL particles; however, the physiological significance of this interaction remains unknown. LDL binding in vitro requires a disordered N-terminal region in PCSK9's prodomain. Here, we report that peptides corresponding to a predicted amphipathic α-helix in the prodomain N terminus adopt helical structure in a membrane-mimetic environment. This effect was greatly enhanced by an R46L substitution representing an atheroprotective PCSK9 loss-of-function mutation. A helix-disrupting proline substitution within the putative α-helical motif in full-length PCSK9 lowered LDL binding affinity >5-fold. Modeling studies suggested that the transient α-helix aligns multiple polar residues to interact with positively charged residues in the C-terminal domain. Gain-of-function PCSK9 mutations associated with familial hypercholesterolemia (FH) and clustered at the predicted interdomain interface (R469W, R496W, and F515L) inhibited LDL binding, which was completely abolished in the case of the R496W variant. These findings shed light on allosteric conformational changes in PCSK9 required for high-affinity binding to LDL particles. Moreover, the initial identification of FH-associated mutations that diminish PCSK9's ability to bind LDL reported here supports the notion that PCSK9-LDL association in the circulation inhibits PCSK9 activity.

Keywords: proprotein convertase subtilisin/kexin type 9 (PCSK9), low-density lipoprotein (LDL), cholesterol regulation, intrinsically disordered protein, conformational change, computer modeling SUMMARYProprotein convertase subtilisin/kexin type-9 (PCSK9) is a ligand of low-density lipoprotein receptor (LDLR) that promotes LDLR degradation in late endosomes/lysosomes. In human plasma, 30-40% of PCSK9 is bound to LDL particles; however, the physiological significance of this interaction remains unknown. LDL binding in vitro requires a disordered N-terminal region in PCSK9's prodomain. Here we report that peptides corresponding to a predicted amphipathic ahelix in the prodomain N-terminus adopted helical structure in a membrane-mimetic environment; this effect was greatly enhanced by an R46L substitution representing an atheroprotective PCSK9 loss-of-function mutation. A helix-disrupting proline substitution within the putative a-helical motif in full-length PCSK9 lowered LDL binding affinity >5-fold. Modeling studies suggested the transient a-helix aligns multiple polar residues to interact with positivecharged residues in the C-terminal domain. Gain-of-function PCSK9 mutations associated with familial hypercholesterolemia (FH) and clustered at the predicted interdomain interface (R469W, R496W, F515L) inhibited LDL binding, which was abolished for the R496W variant. These studies inform on allosteric conformational changes in PCSK9 required for high-affinity binding to LDL particles. Moreover, we report the initial identification of FH-associated mutations that diminish the ability of PCSK9 to bind LDL, supporting that LDL association in the circulation inhibits PCSK9 activity. PCSK9 mutations in hypercholesterolemia

No abstract

Rationale: We have previously shown that a substantial proportion of plasma PCSK9 (30-40%) is associated with LDL particles in normolipidemic subjects. Cellular assays show that LDL-bound PCSK9 is less active for binding to cell surface LDLRs. Therefore, the ability of circulating PCSK9 to direct LDLR degradation in liver could be regulated by plasma LDL levels. In addition, LDL subspecies may have altered abilities in binding PCSK9. We have mapped the LDL binding region to a short stretch of amino acids (aa 31-52) in the PCSK9 prodomain. It is unknown whether a common loss-of-function PCSK9 mutation (R46L) within this region affects LDL binding. Objective: To determine whether plasma PCSK9 distribution (LDL-bound versus unbound) is affected in hypercholesterolemic subjects. To further characterize the interaction of PCSK9 and LDL, we investigated the interaction of PCSK9 with two subspecies of LDL - large, buoyant LDL (LBLDL; d=1.019-1.044 g/ml) and small, dense LDL (SDLDL; d=1.044-1.063 g/ml). Additionally, we investigated the effect of the R46L PCSK9 mutation on the LDL binding affinity of PCSK9. Methods and Results: We used flotation ultracentrifugation in Optiprep density gradients to fractionate human plasma samples followed by immunoprecipitation and western blot to quantify PCSK9 distribution in LDL and non-LDL fractions. In a pilot study, the proportion of total plasma PCSK9 in the LDL fraction was increased from 38±5% to 57±3% (N=6) in hypercholesterolemic subjects (LDL>4.9mM, TG<2.3mM) versus normal controls (LDL<3 mM, TG<2.3mM). Saturation binding assays showed that SDLDL bound PCSK9 with lower affinity (Kd = 361.9 nM) than LDLDL (Kd = 263.9 nM). Competition binding assays determined that recombinant purified PCSK9-R46L secreted from HEK293 cells did not bind to isolated LDL with significantly altered affinity compared to wild-type PCSK9. Conclusion: Our preliminary results indicate that plasma PCSK9 distribution is altered in hypercholesterolemia, with an increased proportion of total PCSK9 bound to LDL particles. Our in vitro results suggest that circulating small, dense LDL may bind more poorly to PCSK9 than larger LDL subspecies.