- Email: [email protected]
Identification and characterization of new gain-of-function mutations in the PCSK9 gene responsible for autosomal dominant hypercholesterolemia
Atherosclerosis 223 (2012) 394e400
Contents lists available at SciVerse ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Identification and characterization of new gain-of-function mutations in the PCSK9 gene responsible for autosomal dominant hypercholesterolemia Marianne Abifadel a, b, *,1, Maryse Guerin c, d, e,1, Suzanne Benjannet f, Jean-Pierre Rabès a, g, h, Wilfried Le Goff c, d, e, Zélie Julia c, d, e, Josée Hamelin f, Valérie Carreau c, d, i, Mathilde Varret a, Eric Bruckert c, d, e, i, Laurent Tosolini a, Olivier Meilhac a, Philippe Couvert c, d, e, j, Dominique Bonnefont-Rousselot k, l, John Chapman c, d, e, Alain Carrié c, d, e, j, Jean-Baptiste Michel a, Annik Prat f, Nabil G. Seidah f, Catherine Boileau a, g, h a
Institut National de la Santé et de la Recherche Médicale, Inserm UMR698, Hemostasis, Bio-Engineering and Cardiovascular Remodelling, Hôpital Bichat-Claude Bernard, 46 Rue Henri Huchard, 75877 Paris Cedex 18, France b Laboratoire de Biochimie, Faculté de Pharmacie et Pôle Technologie-Santé, Université Saint-Joseph, Beirut, Lebanon c INSERM UMRS939, Hôpital de la Pitié, Paris, France d Université Pierre et Marie Curie-Paris 6, Paris, France e Institute of Cardiometabolism and Nutrition, Hôpital de la Pitié, Paris, France f Laboratory of Biochemical Neuroendocrinology, Clinical Research Institute of Montreal, Montreal, Quebec H2W 1R7, Canada g Laboratoire de Biochimie et de Génétique Moléculaire, Hôpital Ambroise-Paré, Assistance Publique-Hôpitaux de Paris (AP-HP), Paris, France h Université Versailles-Saint-Quentin-en-Yvelines, UFR Médicale Paris-Ile-de-France Ouest, Boulogne, France i Department of Endocrinology, AP-HP, Hôpital de la Pitié, Paris, France j Unit of Molecular and Oncologic Endocrinology, AP-HP, Hôpital de la Pitié-Salpêtrière, Paris, France k Departement of Metabolic Biochemistry, Groupe hospitalier Pitié-Saplêtrière (AP-HP), Hôpital de la Pitié, Paris, France l Department of Experimental, Metabolic and Clinical Biology EA4466, Faculty of Pharmacy, Paris Descartes University, Sorbonne Paris Cité, Paris, France
a r t i c l e i n f o
a b s t r a c t
Article history: Received 18 November 2011 Received in revised form 13 April 2012 Accepted 19 April 2012 Available online 17 May 2012
Background: The identification of mutations in PCSK9 (proprotein convertase subtilisin kexin9) in autosomal dominant hypercholesterolemia (ADH), has revealed the existence of a new player in cholesterol homeostasis. PCSK9 has been shown to enhance the degradation of the LDL receptor (LDLR) at the cell surface. Gain-of-function mutations of PCSK9 induce ADH and are very rare, but their identification is crucial in studying PCSK9’s role in hypercholesterolemia, its detailed trafficking pathway and its impact on the LDLR. Methods: In order to identify new mutations and understand the exact mechanisms of action of mutated PCSK9, PCSK9 was sequenced in 75 ADH patients with no mutations in the LDLR or APOB genes. Functional analyses in cell culture were conducted and the impact of novel PCSK9 mutations on the quantitative and qualitative features of lipoprotein particles and on the HDL-mediated cellular cholesterol efflux was studied. Results: Among these 75 ADH probands with no mutations in the LDLR or APOB genes, four gain-offunction mutations of PCSK9 were identified, of which two were novel: the p.Leu108Arg and the p.Asp35Tyr substitutions. In vitro studies of their consequences on the activity of PCSK9 on cell surface levels of LDLR showed that the p.Leu108Arg mutation clearly results in a gain-of-function, while the p.Asp35Tyr mutation created a novel Tyr-sulfation site, which may enhance the intracellular activity of PCSK9. Conclusion: These data further contribute to the characterization of PCSK9 mutations and to better understanding of the impact on cholesterol metabolism of this new therapeutic target. Ó 2012 Elsevier Ireland Ltd. All rights reserved.
Keywords: Familial hypercholesterolemia PCSK9 LDL-C Mutations
1. Introduction * Corresponding author. INSERM U 698, Centre hospitalo-universitaire Xavier Bichat Secteur Claude Bernard, 46, rue Henri Huchard, 75877 Paris Cedex 18, France. Tel.: þ33 1 40 25 75 21; fax: þ33 1 40 25 86 02. E-mail address: [email protected] (M. Abifadel). 1 These two authors contributed equally to this work. 0021-9150/$ e see front matter Ó 2012 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2012.04.006
Hypercholesterolemia is a major risk factor for coronary heart disease. Autosomal dominant hypercholesterolemia (ADH OMIM144400), is characterized by an elevation in the low density
M. Abifadel et al. / Atherosclerosis 223 (2012) 394e400
lipoproteins cholesterol (LDL-C) levels and has been classically associated with mutations in the LDLR gene encoding the low density lipoprotein receptor [1] or APOB gene encoding it’s ligand apolipoprotein B [2]. Our discovery in 2003 of the third gene implicated in ADH, PCSK9 (proprotein convertase subtilisin kexin9), by identifying its first mutants in French families with hypercholesterolemia, has revealed a potent new key regulator of plasma LDL-C levels [3]. Since then PCSK9 has been extensively studied and now constitutes, one of the most interesting targets to silence in the treatment of hypercholesterolemia [4]. Particularly expressed in the liver, gut, kidney and nervous system [5], PCSK9 is the ninth member of the subtilisin family of kexin-like proprotein convertases, with a signal peptide (amino acids 1e30), a pro-domain (amino acids 31e152) and a catalytic domain (containing the triad of active site residues Asp186, His226 and Ser386), followed by a 243-amino-acid cysteine-rich and histidine-rich C-terminal region [5,6]. PCSK9 is synthesized as an inactive zymogen of 74-kDa (proPCSK9) that undergoes intramolecular autocatalytic cleavage in the endoplasmic reticulum (ER), which produces the 60-kDa mature and active form of PCSK9. Autocatalytic cleavage of the zymogen in the endoplasmic reticulum (ER) [5] is essential for its transport from this compartment and for its secretion [6,7]. The cleaved prosegment of 14 kDa remains associated with the catalytic domain, permitting the mature protein to move out of the ER into the secretory pathway and assisting in the proper folding and the inhibition of the catalytic activity of the enzyme [5,6]. The secreted 60-kDa mature form is further cleaved at the motif RFHR218Y into a 53-kDa inactivated fragment by other proprotein convertases, particularly Furin [8,9]. PCSK9 binds to the LDL receptor intracellularly and at cell surface and promotes its intracellular degradation [10e12]. Gain-of-function mutations of PCSK9 are associated with familial hypercholesterolemia and its complications and are very rare [13]. However nonsense mutations of PCSK9 are associated with hypocholesterolemia due to a loss-of-function mechanism. These mutations are much more frequent in individuals of African descent and are associated with a reduction of LDL-C and a considerably reduced risk of coronary heart disease (CHD) [14]. Only a few gain-of-function mutations have been reported in ADH families, notably p.Ser127Arg, p.Asp129Gly, p.Arg215His, p.Phe216Leu, p.Arg218Ser, p.Asp374Tyr, p.Asp374His [reviewed in 13]. Their identification was crucial in studying PCSK9’s role in hypercholesterolemia and its impact on the LDLR. Nevertheless, the mechanism of action of these mutants in ADH seems to be heterogeneous. The p.Asp374Tyr increases 10- to 25-fold the affinity of PCSK9 for the LDLR and enhances receptor degradation [10]. The natural gain-of-function mutations p.Arg218Ser, p.Phe216Leu results in total (p.Arg218Ser) or partial (p.Phe216Leu) loss of the Furin processing of PCSK9 and act by increasing the stability of PCSK9 [8,9]. The p.Ser127Arg variant interferes with autocatalytic cleavage, which is crucial for secretion from the cell; but its exact mechanism in hypercholesterolemia has not yet been elucidated [7], and may involve an enhanced degradation of the LDLR by both the intracellular and extracellular pathways [12]. Thus the identification of novel ADH mutations in PCSK9 and new families carrying PCSK9 mutations, the study of their genetic, cellular, biochemical impact on the metabolism of lipoproteins are very important in order to understand PCSK9’s mechanism of action and the details of its role in pathophysiology. Herein we report families with ADH carrying PCSK9 mutations and identify two new gain-of-function mutations of the gene: the p.Leu108Arg and p.Asp35Tyr substitutions. These mutations have been studied on a genetic basis and their impact on the function of PCSK9 was detailed in cell-based assays. Equally, we evaluated the impact of these PCSK9 mutations on both the quantitative and
395
qualitative features of lipoprotein particles and on the cholesterol ester transfer protein (CETP) activity. Finally, because it is still not certain that the atherogenic effect of PCSK9 is due solely to its LDLelevating action [15], the impact of PCSK9 mutations on the HDLmediated cellular cholesterol efflux has also been investigated for the first time. 2. Materials and methods 2.1. Patients and blood samples The 75 French probands with autosomal dominant hypercholesterolemia were recruited through the Lipid clinic of Hôpital de La Pitié Salpêtrière in Paris. Inclusion criteria [16], blood sampling and ethical issues are detailed in Supplemental Data. 2.2. Genetic analysis Details of the studies of the LDLR, APOB and PCSK9 genes and the sequencing strategy are given in Supplemental Data [3]. The nomenclature and In silico prediction of effect of molecular events are developed also in the same section. 2.3. Biosynthetic analysis of the WT, p.Leu108Arg and p.Asp35Tyr mutants The generation of the single point mutants of wild type (WT) Cterminally V5-tagged human PCSK9 was achieved as previously described [8]. We also generated a double mutant D35Y þ Y38F which eliminates the Tyr38 sulfation. Transfections were done with 3 105 HEK293, HuH7 or HepG2 cells using Effectene (Qiagen). For biosynthetic analyses we pulse labeled the cells with either 35S(Met þ Cys) for 3 h or Na35 2 SO4 for 2 h and the products were immunoprecipitated with a V5 monoclonal antibody and separated by SDS-PAGE as previously described [6,8]. The details of the transfections, culture, immunoprecipitation and blotting are given in Supplemental Data. A sensitive Elisa assay for human PCSK9 was achieved as previously described [17]. 2.4. Lipoprotein fractionation Plasma lipoproteins were isolated from plasma by density gradient ultracentrifugation detailed in Supplemental Methods [18]. 2.5. Determination of endogenous CETP activity and free cholesterol efflux assays Determination of endogenous cholesterol ester (CE) transfer from HDL to apolipoprotein B-containing lipoproteins [18] and free cholesterol efflux assays [19] were studied as detailed in Supplemental Data. 3. Results We presently identified, among 75 ADH probands recruited in the Lipid Clinic of La Pitié Salpêtrière hospital, for whom the APOBp.Arg3527Gln mutation and mutations in the LDLR gene had been excluded, four point mutations in the PCSK9 gene, of which two were novel. These PCSK9 mutations segregated with ADH in each family and were not found in normocholesterolemic controls. The first mutation corresponding to a substitution, p.Leu108Arg (c.323T > G), in exon 2 of PCSK9 was identified in a Black family (Family A) originating from the Republic of Mauritius (Fig. 1). The 53 y.o proband (A.II.4) was ascertained at 41 years old with
396
M. Abifadel et al. / Atherosclerosis 223 (2012) 394e400
Fig. 1. Pedigree and genetic analysis of families A and B. Birth date, Age (in years) at given lipid measurement, total cholesterol (TC), low density lipoprotein cholesterol (LDL-C), High density lipoprotein (HDL-C) and triglycerides (TG) levels are given in mmol/l (with untreated values for affected individuals). The probands are indicated by an arrow. (A) Sequence analysis in family A. The proband and affected members are heterozygous with respect to the p.Leu108Arg mutation. (B) Sequence analysis in family B. The proband and his son are heterozygous with respect to p.Asp35Tyr substitution in exon 1 of PCSK9.
9.44 mmol/l of TC; 7.83 mmol/l of LDL-C; with tendon xanthomas. More recently he developed type II diabetes with glucose levels of 7.89 mmol/l. Intima media thickness (IMT) measurements performed for AII.4 patient showed an increase of both common carotid and common femoral arteries with a right femoral plaque. He was recently hospitalized with an acute coronary syndrome and a right coronary artery lesion. The proband’s family has been recruited. One of his three children (A.III.3) suffers from
hypercholesterolemia and carries also the p.Arg108Leu mutation. The proband’s sister (A.II.8) has severe hypercholesterolemia with 10.39 mmol/l of total cholesterol and 8.61 mmol/l of LDL-cholesterol at 47 y.o, before any lipid lowering therapy, and an increase of both carotid and femoral IMT with microcalcifications and a slight femoral plaque. She has a 20 y.o son (A.III.8) who is hypercholesterolemic and carries also the mutation. Her brother suffered myocardial infarction (A.II.7) at 45 y.o and her husband (who is also
M. Abifadel et al. / Atherosclerosis 223 (2012) 394e400
her cousin) died from myocardial infarction (A.II.9). The p.Arg108Leu mutation of PCSK9 was found in all affected patients of the family whose DNA was available (A. II.4; II.8; III.3; III.8). The second mutation corresponds to a substitution p.Asp35Tyr (c.103G > T) in exon 1 of the PCSK9 gene. It was identified in the Family B proband (B.I.1) with 8.14 mmol/l of TC and 6.05 mmol/l of LDL-C at 55 y.o. His echodoppler at 62 y.o. showed slight left femoral plaque and right femoral calcification with no carotid lesion. One of his two sons (B.II.1) is hypercholesterolemic with 7.76 mmol/l of total cholesterol and 6.10 mmol/l of LDL-C and was found to carry the p.Asp35Tyr mutation in PCSK9. It is noteworthy that two previously reported mutants of PCSK9, p.S127R [3,13] and p.L21tri [13], have also been identified in 2 of the 75 non LDLR/non APOB probands. Thus, mutations in PCSK9 have been found in 5.3% of the ADH probands, with no mutation in the LDLR or the APOB genes, recruited in this study. Biosynthetic analyses showed upon overexpression of either the WT or the p.Asp35Tyr and p.Leu108Arg mutants in HEK293 (or HepG2) cells, that all forms of PCSK9 were equally well autocatalytically cleaved and secreted as a complex with their prosegment (Fig. 2A; supplemental Fig. 1A). Furthermore, all forms were similarly susceptible to Furin cleavage at Arg218Y, resulting in comparative levels of PCSK9-DN218 [8,9] in the media of HEK293 cells. Using an ELISA assay we estimated the levels of secreted PCSK9 and showed that WT, p.Asp35Tyr and p.Leu108Arg mutants were secreted from HEK293 cells comparatively well, giving 24 h levels of w1600, 1400 or 1900 ng/ml respectively. Human hepatic HuH7 cells were then incubated for 4 h with equal amounts of each protein (730 ng/ml) and the levels of cell surface LDLR were measured by FACS [20]. The data show that the reduction in LDLR afforded by the p.Asp35Tyr mutant is significantly slightly higher than with WT PCSK9, but that the p.Leu108Arg mutant exhibited a marked w2-fold enhanced degrading activity towards the LDLR, resulting in a clear and significant gain-of-function in this assay (Fig. 2B). In an effort to rationalize the mechanism behind the presumably gain-of-function mutation p.Asp35Tyr, we postulated that this
397
may result in a possible new Tyr-sulfation site, similar to the WT Tyr38 sulfation [8]. Indeed, upon incubation of HEK293 cells overexpressing either WT or p.Asp35Tyr PCSK9 with Na35 2 SO4, we observed a large increase in the level of sulfated prosegment in the cells and media consistent with a double sulfation at Tyr35 (Asp35Tyr) and Tyr38 (WT) in the p.Asp35Tyr mutant (Fig. 3A). As a negative control we showed that the single mutant p.Tyr38Phe in the WT sequence is not sulfated. To prove that the Asp35Tyr did not cause an over sulfation of Tyr38, we obtained the double mutant p.Asp35Tyr þ p.Tyr38Phe. Repeating the above Na35 2 SO4 incorporation in either HepG2 cells (Fig. S1B) or HEK293 cells revealed that indeed preventing sulfation of Tyr38 in the mutant Tyr38Phe, sulfation still occurred independently at Tyr35 (Fig. 3B), confirming that the latter is a bona fide sulfation site. Finally, repeating the incubation of HuH7 cells with 420 ng/ml of each of these above PCSK9 constructs (estimated by ELISA), we confirmed the results of Fig. 1. The data also revealed that: (i) the gain-offunction of the p.Leu108Arg mutant is less than that of the p.Asp374Tyr mutant, (ii) lack of Tyr38 sulfation does not affect the activity of PCSK9 on LDLR, and (iii) sufation at Tyr35 results in lower secretion (as also observed for p.Asp374Tyr) but slightly higher activity (Figure S2). Plasma lipid and apolipoprotein analyses showed, by comparison with control values obtained from healthy normolipidemic subjects, that patients carrying the p.Leu108Arg (AII.4, AII.8, AIII.3 and AIII.8) or the p.Asp135Tyr (BI.1 and BII.1) mutations in the PCSK9 gene displayed a marked elevation in their mean plasma LDL-C concentration (6.45 1.55 mmol/l and 2.84 0.52 mmol/l in ADH patients and normolipidemic controls respectively, p < 0.0001). Despite the fact that no major difference was observed in plasma HDL-C between PCSK9 mutants and control subjects the relative proportion of HDL-C was significantly reduced in ADH patients as compared with normolipidemic subjects (15.7 4.0% and 31.2 3.4%, in ADH patients and in control subjects respectively, p < 0.0001). Their response to lipid lowering therapy is detailed in Supplemental Data and Supplemental Table 1.
Fig. 2. Biosynthetic analyses and Extracellular activity of PCSK9 and its mutants. (A) HEK293 cells were transiently transfected with cDNAs coding for WT, p.Asp35Tyr (D35Y) or p.Leu108Arg (L108R) mutants all tagged with V5 at their C-terminus. The next day, the cells were incubated with [35S]Met/Cys for 3 h and the cell lysates and media immunoprecipitated with an anti-V5 mAb. The immunoprecipitates were then resolved by SDS-PAGE and the radiolabeled proteins identified by autoradiography, as reported [9]. The migration positions of proPCSK9, PCSK9 and the Furin cleaved product PCSK9-DN218 are emphasized, as well as the associated inhibitory prosegment. (B) Extracellular activity of PCSK9 and its mutants. HEK293 cells were transiently transfected with cDNAs coding for WT, p.Asp35Tyr or p.Leu108Arg mutants tagged with V5 at their C-terminus. The next day the cells were washed and incubated in RPMI þ 5% LPDS þ 1% sodium pyruvate, and the secreted proteins collected for 24 h, and analyzed by a sensitive ELISA assay for human PCSK9 [17] and the triplicate expressed in ng/ml. 730 ng/ml of each protein construct were then incubated with HuH7 cells for 4 h and the cell surface LDLR levels estimated by FACS analysis in triplicate, as reported [17].
398
M. Abifadel et al. / Atherosclerosis 223 (2012) 394e400
Fig. 3. Tyr Sulfation. HEK293 cells were transiently transfected with cDNAs coding for WT, the mutants (A,B) Asp35Tyr (D35Y), Tyr38Phe (Y38F), or (B) the double mutant Asp35Tyr þ Tyr38Phe (D35Y þ Y38F) or an empty vector (pIRES) all tagged with V5 at their C-terminus. The next day, the cells were incubated with Na35 2 SO4 for 2 h and the cell lysates and media immunoprecipitated with an anti-V5 mAb. The immunoprecipitates were then resolved by SDS-PAGE and the radiolabeled proteins identified by autoradiography, as reported [6,8]. Notice how the p.Asp35Tyr mutant exhibits a much higher level of secreted Tyr-sulfated prosegment in the medium. The double mutant shows that the absence of Tyr-sulfation at Tyr38 did not prevent sulfation at Tyr35.
The distribution of cholesterol among density gradient subfractions obtained from patients carrying PCSK9 mutations and normolipidemic controls as well as plasma concentrations of major lipoprotein fractions are presented in Fig. S3 and in Supplemental Table 2, respectively and detailed in Supplemental Data. The mean weight percent chemical composition of major lipoprotein fractions isolated from the plasma of ADH patients carrying PCSK9 mutations and control subjects are presented in Supplemental Table 3. No significant difference in the relative proportion of individual lipids in either apoB- or apoAI-containing lipoprotein particles was detected between ADH patients and normolipidemic control subjects. As shown in Supplemental Figure 4, endogenous plasma CETP activity, expressed as percentage of total CE transferred from HDL to apoB-containing lipoproteins was significantly increased in all ADH patients carrying the p.Leu108Arg (AII.4, AII.8, AIII.3 and AIII.8) mutation in the PCSK9 gene. By contrast, the BI.1 patient carrying the p.Asp135Tyr mutation in the PCSK9 gene (BI.1 and BII.1) displayed an endogenous plasma CETP activity similar to that observed in control subjects. It is relevant to note that a strong positive correlation between CETP activity and plasma levels of LDL-C exits (r ¼ 0.85; p ¼ 0.0003) indicating that the elevation of endogenous plasma CETP activity presently observed in some ADH patients carrying PCSK9 mutations results primarily from the increased levels of apoB-containing lipoprotein cholesteryl ester acceptors. Finally, we evaluated the overall capacity of total plasma of ADH patients carrying PCSK9 mutations to promote cellular free cholesterol efflux from cholesterol loaded macrophage [19]. As shown in Supplemental Data and Figure S5A, no significant difference was observed in the plasma efflux capacity of patients carrying the p.Leu108Arg or p.Asp135Tyr mutation in the PCSK9 gene as compared to control subjects. Plasma efflux capacity via the ABCA1 pathway was similar for both ADH patients carrying PCSK9 mutations and controls (Fig. S5B). In addition, we did not observe any significant differences in the capacity of either larger HDL2 or smaller HDL3 particles to promote cellular efflux via SR-BI (Fig. S5CD) or ABCG1 (Fig. S5EF) between these ADH patients and control subjects (Supplemental data).
4. Discussion Here we report, two non LDLR/non APOB families carrying two distinct new molecular defects located on the PCSK9 gene corresponding to distinct amino acid substitutions within the prosegment of PCSK9: p.Asp35Tyr and p.Leu108Arg. These mutations segregated with ADH in each family and were not found in normocholesterolemic controls. In these subjects, familial hypercholesterolemia leads to premature atherosclerosis with coronary lesions and cardiovascular complications. The aspartate at position 35 is highly conserved between most species (primates, Mus musculus, Rattus norvegicus) but not in Tetraodon nigroviridis, nor in Zebrafish, where it is replaced by a tyrosine. While Leucine at position 108 is well conserved in primates, it is not conserved in M. musculus and R. norvegicus. The in silico estimation of the impact of each variant on the protein using the SIFT tools showed that the p.Leu108Arg was well tolerated while the p.Asp35Tyr was not. However, using Polyphen tools, the p.Leu108Arg seems to be possibly damaging (PSIC score difference 1.752) and the p.Asp35Tyr probably damaging (score difference 2.090). Thus, theoretical predictions have limitations and we wished to define experimentally the impact of these mutations on proPCSK9 zymogen activation into PCSK9 in the endoplasmic reticulum, the subsequent secretion of the latter out of the cell and its ability to enhance the degradation of the cell surface LDLR. Biosynthetic analyses showed that the WT or the p.Leu108Arg and p.Asp35Tyr mutants were equally well autocatalytically cleaved and secreted as a complex with their prosegment and were similarly susceptible to Furin cleavage at Arg218Y. Furthermore the WT, p.Asp35Tyr and p.Leu108Arg mutants were secreted from HEK293 cells comparatively well. This suggested that these mutations did not result in gross structural alterations or folding of proPCSK9. However, when incubated with HuH7 cells the p.Leu108Arg variant of PCSK9 showed a w2-fold enhanced degrading activity towards the LDLR by FACS measurements compared to WT PCSK9 (Fig. S2). This was a clear evidence that in this assay that measures the extracellular activity of PCSK9 on the LDLR, the p.Leu108Arg mutation results in a gain-of-function, likely due to the positive charge of Arg. Comparatively, the reduction in LDLR afforded by WT PCSK9 was slightly less than that seen with
M. Abifadel et al. / Atherosclerosis 223 (2012) 394e400
the p.Asp35Tyr mutant, suggesting that the latter exhibits a mild gain-of-function activity in this assay. In an effort to rationalize the mechanism behind the gain-of-function mutation p.Asp35Tyr, we noticed that the resulting Tyr35 is surrounded by acidic residues and could potentially be a novel Tyr-sulfation site [21]. Radioactive sulfate incorporation experiments in either HEK293 or HepG2 cells showed that indeed this mutation resulted in a new Tyr-sulfation site (Fig. 3, S1B) similar to the wild type Tyr38 sulfation [8]. Furthermore, the double mutant p.Asp35Tyr þ p.Asp38Phe is still sulfated (Fig. S1), confirming that sulfation occurs at Tyr35 independently from that at Tyr38, and that sulfation at Tyr38 does not change the activity of PCSK9 towards the LDLR (Fig. S2). Even though our recent data demonstrated that the negative charge in the N-terminal portion of the prosegment acts as an inhibitor to PCSK9 function on the LDL receptor [19], replacement of Asp35 with a negatively charged sulfated Tyr (Tyr-SO 3 ) may mask this phenomenon and result in a gain-of-function by modifying the secondary structure of this disordered portion of the prosegment [22]. A similar observation that loss of acidic residues in the Nterminal part of the prosegment is associated with hypercholesterolemia, was also reported in heterozygote or homozygote Japanese subjects exhibiting the p.Glu32Lys mutation [23], which we also could not correlate with a marked enhanced extracellular activity of PCSK9 (Seidah NG, unpublished observation). It is therefore plausible that the p.Asp35Tyr or p.Glu32Lys mutations may rather enhance the intracellular activity of PCSK9 [12], an effect that would be missed in the extracellular activity measured here (Fig. 2). In an effort to rationalize the structural consequence of the p.Leu108Arg mutation on the PCSK9-LDLR complex, we analyzed the recently made publicly available 3D model of PCSK9 in complex with the complete ectodomain of the LDLR (http://www.pdb.org/ pdb/explore/explore.do?structureId¼3M0C). Amazingly, the data show that Leu108 of the prosegment of PCSK9 makes a direct hydrophobic interaction with Leu626 within a loop structure of the b-propeller domain of the LDLR (Fig. S5eS8). The nearby Arg385 of the LDLR (Fig. S7,8) might thus be repelled by the Arg substitution of Leu108 in the prosegment of the PCSK9, and the latter mutant may possibly electrostatically bind the nearby Glu605 of the LDLR, thereby favoring a tighter complex of the PCSK9 with the LDLR. Thus, we believe that the p.Leu108Arg substitution is a gain-offunction due to the re-orientation of the interaction of Leu626 in the LDLR with Leu108 of the prosegment of PCSK9 towards a more stabilizing interaction within the PCSK9-LDLR complex via an ionic interaction of the resulting mutant Arg108 of PCSK9 with the neighboring Glu605 within the b-propeller domain of the LDLR (Fig. S6eS8). It is noteworthy that the details of the beta propeller/ pro-domain interface have been very recently reported by Lo Surdo et al. and showed the close proximity of residues associated with familial hypercholesterolemia: S127, D129 and now L108 [24]. In the present study, we observed that ADH caused by mutation on the PSCK9 gene, was not associated with abnormal physicochemical and biological properties of HDL particles, and in particular in their capacity to promote cellular free cholesterol efflux (Supplemental data and Fig. S5). It is relevant to consider that in this way ADH caused by mutations in the PSCK9 gene differs from those involving mutations in the LDLR gene (discussed in Supplemental data). As compared to normolipidemic subjects, we presently observed that CETP-mediated CE transfer from HDL to apoB-containing lipoproteins is enhanced in ADH patients carrying the PCSK9 mutations as a result of increased plasma levels of cholesteryl ester acceptors (Fig. S4). This observation is in good agreement with previous studies showing an accelerated CE transfer in familial hypercholesterolemia as a result of increased plasma levels in both CETP mass and apoB-containing cholesteryl ester acceptors [25]. In addition, we presently observed that elevated CETP activity is
399
associated with a predominance of denser LDL subfractions in the plasma from ADH patients carrying PCSK9 mutations, highlighting the pro-atherogenic role of CETP in the formation and accumulation of atherogenic small dense LDL particles. It is noteworthy that most of the ADH patients, in this study, were under lipid lowering therapy which could limit the interpretation of the results. We thus cannot exclude the possibility that pharmacological therapies used to reduce plasma cholesterol levels in hypercholesterolemic patients might at least in part contribute to normalize some of the abnormal quantitative and qualitative features of HDL particles, in particular their capacity to promote cellular free cholesterol efflux. In conclusion, in this work we identified two new gain-of-function mutations of PCSK9 (p.Leu108Arg and p.Asp35Tyr) in ADH families with no mutations in the LDLR or APOB genes. We studied in vitro, the impact of these mutations on PCSK9 processing and on its activity on cell surface LDLR levels, and showed that the p.Leu108Arg mutation clearly results in a gain-of-function, while the p.Asp35Tyr mutation created a novel Tyr-sulfation site, which may enhance the intracellular activity of PCSK9. Furthermore, we evaluated the impact of these mutations on both quantitative and qualitative features of lipoprotein particles and on the HDL-mediated cellular cholesterol efflux. These data further contribute to the characterization of PCSK9 mutations and allow us to better understand the impact on cholesterol metabolism of this new therapeutic target in hypercholesterolemia. Acknowledgments We are indebted to the family members for their cooperation. This work was supported by grants from ANR (ANR-05-PCOD-017, ANR-06-MRAR-038, ANR-08-GENO-002-01), PHRC (AOM06024), Institut National de la Santé et de la Recherche Médicale and Conseil de la Recherche de l’Université Saint-Joseph (Beirut, Lebanon : to MA). This work was supported in part (to NGS & AP) by CIHR grants # CTP 82946, MOP 102741 and a Canadian Chair # 216684 and a Strauss Foundation (to NGS). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.atherosclerosis.2012.04.006. References [1] Goldstein JL, Brown MS. Familial hypercholesterolemia: pathogenesis of a receptor disease. Johns Hopkins Med J 1978;143:8e16. [2] Innerarity TL, Weisgraber KH, Arnold KS, Mahley RW, Krauss RM, Vega GL, et al. Familial defective apolipoprotein B-100: low density lipoproteins with abnormal receptor binding. Proc Natl Acad Sci USA 1987;84:6919e23. [3] Abifadel M, Varret M, Rabes JP, Allard D, Ouguerram K, Devillers M, et al. Mutations in PCSK9 cause autosomal dominant hypercholesterolemia. Nat Genet 2003;34:154e6. [4] Abifadel M, Pakradouni J, Collin M, Samson-Bouma ME, Varret M, Rabès JP, et al. Strategies for proprotein convertase subtilisin kexin 9 modulation: a perspective on recent patents. Expert Opin Ther Pat 2010;20:1547e71. [5] Seidah NG, Benjannet S, Wickham L, Marcinkiewicz J, Jasmin SB, Stifani S, et al. The secretory proprotein convertase neural apoptosis-regulated convertase 1 (NARC-1): liver regeneration and neuronal differentiation. Proc Natl Acad Sci USA 2003;100:928e33. [6] Benjannet S, Rhainds D, Essalmani R, Mayne J, Wickham L, Jin W, et al. NARC-1/PCSK9 and its natural mutants: zymogen cleavage and effects on the low density lipoprotein (LDL) receptor and LDL cholesterol. J Biol Chem 2004; 279:48865e75. [7] McNutt MC, Lagace TA, Horton JD. Catalytic activity is not required for secreted PCSK9 to reduce low density lipoprotein receptors in HepG2 cells. J Biol Chem 2007;282:20799e803. [8] Benjannet S, Rhainds D, Hamelin J, Nassoury N, Seidah NG. The proprotein convertase (PC) PCSK9 is inactivated by Furin and/or PC5/6A: functional consequences of natural mutations and post-translational modifications. J Biol Chem 2006;281:30561e72.
400
M. Abifadel et al. / Atherosclerosis 223 (2012) 394e400
[9] Essalmani R, Susan-Resiga D, Chamberland A, Abifadel M, Creemers JW, Boileau C, et al. In vivo evidence that furin from hepatocytes inactivates PCSK9. J Biol Chem 2011;286:4257e63. [10] Lagace TA, Curtis DE, Garuti R, McNutt MC, Park SW, Prather HB, et al. Secreted PCSK9 decreases the number of LDL receptors in hepatocytes and in livers of parabiotic mice. J Clin Invest 2006;116:2995e3005. [11] Cameron J, Holla OL, Ranheim T, Kulseth MA, Berge KE, Leren TP. Effect of mutations in the PCSK9 gene on the cell surface LDL receptors. Hum Mol Genet 2006;15:1551e8. [12] Poirier S, Mayer G, Poupon V, McPherson PS, Desjardins R, Ly K, et al. Dissection of the endogenous cellular pathways of PCSK9-induced low density lipoprotein receptor degradation: evidence for an intracellular route. J Biol Chem 2009;284:28856e64. [13] Abifadel M, Rabès JP, Devillers M, Munnich A, Erlich D, Junien C, et al. Mutations and polymorphisms in the proprotein convertase subtilisin kexin 9 (PCSK9) gene in cholesterol metabolism and disease. Hum Mut 2009;30:520e9. [14] Cohen JC, Boerwinkle E, Mosley Jr TH, Hobbs HH. Sequence variations in PCSK9, low LDL, and protection against coronary heart disease. N Engl J Med 2006;354:1264e72. [15] Brown MS, Goldstein JL. Biomedicine. Lowering LDLenot only how low, but how long? Science 2006;311:1721e3. [16] Siest G, Visvikis S, Herbeth B, Gueguen R, Vincent-Viry M, Sass C, et al. Objectives, design and recruitment of a familial and longitudinal cohort for studying gene-environment interactions in the field of cardiovascular risk: the Stanislas cohort. Clin Chem Lab Med 1998;36:35e42. [17] Dubuc G, Tremblay M, Pare G, Jacques H, Hamelin J, Benjannet S, et al. A new method for measurement of total plasma PCSK9: clinical applications. J Lipid Res 2010;51:140e9.
[18] Catalano G, Julia Z, Frisdal E, Vedie B, Fournier N, Le Goff W, et al. Torcetrapib differentially modulates the biological activities of HDL2 and HDL3 particles in the reverse cholesterol transport pathway. Arterioscler Thromb Vasc Biol 2009;29:268e75. [19] Wang N, Lan D, Chen W, Matsuura F, Tall AR. ATP-binding cassette transporters G1 and G4 mediate cellular cholesterol efflux to high-density lipoproteins. Proc Natl Acad Sci USA 2004;101:9774e9. [20] Catalano G, Julia Z, Frisdal E, Vedie B, Fournier N, Le Goff W, et al. Effects of the prosegment and pH on the activity of PCSK9: evidence for additional processing events. J Biol Chem 2010;285:40965e78. [21] Beisswanger R, Corbeil D, Vannier C, Thiele C, Dohrmann U, Kellner R, et al. Existence of distinct tyrosylprotein sulfotransferase genes: molecular characterization of tyrosylprotein sulfotransferase-2. Proc Natl Acad Sci USA 1998; 95:11134e9. [22] Cunningham D, Danley DE, Geoghegan KF, Griffor MC, Hawkins JL, Subashi TA, et al. Structural and biophysical studies of PCSK9 and its mutants linked to familial hypercholesterolemia. Nat Struct Mol Biol 2007;14:413e9. [23] Miyake Y, Kimura R, Kokubo Y, Okayama A, Tomoike H, Yamamura T, et al. Genetic variants in PCSK9 in the Japanese population: rare genetic variants in PCSK9 might collectively contribute to plasma LDL cholesterol levels in the general population. Atherosclerosis 2008;196:29e36. [24] Surdo PL, Bottomley MJ, Calzetta A, Settembre EC, Cirillo A, Pandit S, et al. Mechanistic implications for LDL receptor degradation from the PCSK9/LDLR structure at neutral pH. EMBO Rep 2011;12:1300e5. [25] De Grooth GJ, Smilde TJ, Van Wissen S, Klerkx AH, Zwinderman AH, Fruchart JC, et al. The relationship between cholesteryl ester transfer protein levels and risk factor profile in patients with familial hypercholesterolemia. Atherosclerosis 2004;173:261e7.
Contents lists available at SciVerse ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Identification and characterization of new gain-of-function mutations in the PCSK9 gene responsible for autosomal dominant hypercholesterolemia Marianne Abifadel a, b, *,1, Maryse Guerin c, d, e,1, Suzanne Benjannet f, Jean-Pierre Rabès a, g, h, Wilfried Le Goff c, d, e, Zélie Julia c, d, e, Josée Hamelin f, Valérie Carreau c, d, i, Mathilde Varret a, Eric Bruckert c, d, e, i, Laurent Tosolini a, Olivier Meilhac a, Philippe Couvert c, d, e, j, Dominique Bonnefont-Rousselot k, l, John Chapman c, d, e, Alain Carrié c, d, e, j, Jean-Baptiste Michel a, Annik Prat f, Nabil G. Seidah f, Catherine Boileau a, g, h a
Institut National de la Santé et de la Recherche Médicale, Inserm UMR698, Hemostasis, Bio-Engineering and Cardiovascular Remodelling, Hôpital Bichat-Claude Bernard, 46 Rue Henri Huchard, 75877 Paris Cedex 18, France b Laboratoire de Biochimie, Faculté de Pharmacie et Pôle Technologie-Santé, Université Saint-Joseph, Beirut, Lebanon c INSERM UMRS939, Hôpital de la Pitié, Paris, France d Université Pierre et Marie Curie-Paris 6, Paris, France e Institute of Cardiometabolism and Nutrition, Hôpital de la Pitié, Paris, France f Laboratory of Biochemical Neuroendocrinology, Clinical Research Institute of Montreal, Montreal, Quebec H2W 1R7, Canada g Laboratoire de Biochimie et de Génétique Moléculaire, Hôpital Ambroise-Paré, Assistance Publique-Hôpitaux de Paris (AP-HP), Paris, France h Université Versailles-Saint-Quentin-en-Yvelines, UFR Médicale Paris-Ile-de-France Ouest, Boulogne, France i Department of Endocrinology, AP-HP, Hôpital de la Pitié, Paris, France j Unit of Molecular and Oncologic Endocrinology, AP-HP, Hôpital de la Pitié-Salpêtrière, Paris, France k Departement of Metabolic Biochemistry, Groupe hospitalier Pitié-Saplêtrière (AP-HP), Hôpital de la Pitié, Paris, France l Department of Experimental, Metabolic and Clinical Biology EA4466, Faculty of Pharmacy, Paris Descartes University, Sorbonne Paris Cité, Paris, France
a r t i c l e i n f o
a b s t r a c t
Article history: Received 18 November 2011 Received in revised form 13 April 2012 Accepted 19 April 2012 Available online 17 May 2012
Background: The identification of mutations in PCSK9 (proprotein convertase subtilisin kexin9) in autosomal dominant hypercholesterolemia (ADH), has revealed the existence of a new player in cholesterol homeostasis. PCSK9 has been shown to enhance the degradation of the LDL receptor (LDLR) at the cell surface. Gain-of-function mutations of PCSK9 induce ADH and are very rare, but their identification is crucial in studying PCSK9’s role in hypercholesterolemia, its detailed trafficking pathway and its impact on the LDLR. Methods: In order to identify new mutations and understand the exact mechanisms of action of mutated PCSK9, PCSK9 was sequenced in 75 ADH patients with no mutations in the LDLR or APOB genes. Functional analyses in cell culture were conducted and the impact of novel PCSK9 mutations on the quantitative and qualitative features of lipoprotein particles and on the HDL-mediated cellular cholesterol efflux was studied. Results: Among these 75 ADH probands with no mutations in the LDLR or APOB genes, four gain-offunction mutations of PCSK9 were identified, of which two were novel: the p.Leu108Arg and the p.Asp35Tyr substitutions. In vitro studies of their consequences on the activity of PCSK9 on cell surface levels of LDLR showed that the p.Leu108Arg mutation clearly results in a gain-of-function, while the p.Asp35Tyr mutation created a novel Tyr-sulfation site, which may enhance the intracellular activity of PCSK9. Conclusion: These data further contribute to the characterization of PCSK9 mutations and to better understanding of the impact on cholesterol metabolism of this new therapeutic target. Ó 2012 Elsevier Ireland Ltd. All rights reserved.
Keywords: Familial hypercholesterolemia PCSK9 LDL-C Mutations
1. Introduction * Corresponding author. INSERM U 698, Centre hospitalo-universitaire Xavier Bichat Secteur Claude Bernard, 46, rue Henri Huchard, 75877 Paris Cedex 18, France. Tel.: þ33 1 40 25 75 21; fax: þ33 1 40 25 86 02. E-mail address: [email protected] (M. Abifadel). 1 These two authors contributed equally to this work. 0021-9150/$ e see front matter Ó 2012 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2012.04.006
Hypercholesterolemia is a major risk factor for coronary heart disease. Autosomal dominant hypercholesterolemia (ADH OMIM144400), is characterized by an elevation in the low density
M. Abifadel et al. / Atherosclerosis 223 (2012) 394e400
lipoproteins cholesterol (LDL-C) levels and has been classically associated with mutations in the LDLR gene encoding the low density lipoprotein receptor [1] or APOB gene encoding it’s ligand apolipoprotein B [2]. Our discovery in 2003 of the third gene implicated in ADH, PCSK9 (proprotein convertase subtilisin kexin9), by identifying its first mutants in French families with hypercholesterolemia, has revealed a potent new key regulator of plasma LDL-C levels [3]. Since then PCSK9 has been extensively studied and now constitutes, one of the most interesting targets to silence in the treatment of hypercholesterolemia [4]. Particularly expressed in the liver, gut, kidney and nervous system [5], PCSK9 is the ninth member of the subtilisin family of kexin-like proprotein convertases, with a signal peptide (amino acids 1e30), a pro-domain (amino acids 31e152) and a catalytic domain (containing the triad of active site residues Asp186, His226 and Ser386), followed by a 243-amino-acid cysteine-rich and histidine-rich C-terminal region [5,6]. PCSK9 is synthesized as an inactive zymogen of 74-kDa (proPCSK9) that undergoes intramolecular autocatalytic cleavage in the endoplasmic reticulum (ER), which produces the 60-kDa mature and active form of PCSK9. Autocatalytic cleavage of the zymogen in the endoplasmic reticulum (ER) [5] is essential for its transport from this compartment and for its secretion [6,7]. The cleaved prosegment of 14 kDa remains associated with the catalytic domain, permitting the mature protein to move out of the ER into the secretory pathway and assisting in the proper folding and the inhibition of the catalytic activity of the enzyme [5,6]. The secreted 60-kDa mature form is further cleaved at the motif RFHR218Y into a 53-kDa inactivated fragment by other proprotein convertases, particularly Furin [8,9]. PCSK9 binds to the LDL receptor intracellularly and at cell surface and promotes its intracellular degradation [10e12]. Gain-of-function mutations of PCSK9 are associated with familial hypercholesterolemia and its complications and are very rare [13]. However nonsense mutations of PCSK9 are associated with hypocholesterolemia due to a loss-of-function mechanism. These mutations are much more frequent in individuals of African descent and are associated with a reduction of LDL-C and a considerably reduced risk of coronary heart disease (CHD) [14]. Only a few gain-of-function mutations have been reported in ADH families, notably p.Ser127Arg, p.Asp129Gly, p.Arg215His, p.Phe216Leu, p.Arg218Ser, p.Asp374Tyr, p.Asp374His [reviewed in 13]. Their identification was crucial in studying PCSK9’s role in hypercholesterolemia and its impact on the LDLR. Nevertheless, the mechanism of action of these mutants in ADH seems to be heterogeneous. The p.Asp374Tyr increases 10- to 25-fold the affinity of PCSK9 for the LDLR and enhances receptor degradation [10]. The natural gain-of-function mutations p.Arg218Ser, p.Phe216Leu results in total (p.Arg218Ser) or partial (p.Phe216Leu) loss of the Furin processing of PCSK9 and act by increasing the stability of PCSK9 [8,9]. The p.Ser127Arg variant interferes with autocatalytic cleavage, which is crucial for secretion from the cell; but its exact mechanism in hypercholesterolemia has not yet been elucidated [7], and may involve an enhanced degradation of the LDLR by both the intracellular and extracellular pathways [12]. Thus the identification of novel ADH mutations in PCSK9 and new families carrying PCSK9 mutations, the study of their genetic, cellular, biochemical impact on the metabolism of lipoproteins are very important in order to understand PCSK9’s mechanism of action and the details of its role in pathophysiology. Herein we report families with ADH carrying PCSK9 mutations and identify two new gain-of-function mutations of the gene: the p.Leu108Arg and p.Asp35Tyr substitutions. These mutations have been studied on a genetic basis and their impact on the function of PCSK9 was detailed in cell-based assays. Equally, we evaluated the impact of these PCSK9 mutations on both the quantitative and
395
qualitative features of lipoprotein particles and on the cholesterol ester transfer protein (CETP) activity. Finally, because it is still not certain that the atherogenic effect of PCSK9 is due solely to its LDLelevating action [15], the impact of PCSK9 mutations on the HDLmediated cellular cholesterol efflux has also been investigated for the first time. 2. Materials and methods 2.1. Patients and blood samples The 75 French probands with autosomal dominant hypercholesterolemia were recruited through the Lipid clinic of Hôpital de La Pitié Salpêtrière in Paris. Inclusion criteria [16], blood sampling and ethical issues are detailed in Supplemental Data. 2.2. Genetic analysis Details of the studies of the LDLR, APOB and PCSK9 genes and the sequencing strategy are given in Supplemental Data [3]. The nomenclature and In silico prediction of effect of molecular events are developed also in the same section. 2.3. Biosynthetic analysis of the WT, p.Leu108Arg and p.Asp35Tyr mutants The generation of the single point mutants of wild type (WT) Cterminally V5-tagged human PCSK9 was achieved as previously described [8]. We also generated a double mutant D35Y þ Y38F which eliminates the Tyr38 sulfation. Transfections were done with 3 105 HEK293, HuH7 or HepG2 cells using Effectene (Qiagen). For biosynthetic analyses we pulse labeled the cells with either 35S(Met þ Cys) for 3 h or Na35 2 SO4 for 2 h and the products were immunoprecipitated with a V5 monoclonal antibody and separated by SDS-PAGE as previously described [6,8]. The details of the transfections, culture, immunoprecipitation and blotting are given in Supplemental Data. A sensitive Elisa assay for human PCSK9 was achieved as previously described [17]. 2.4. Lipoprotein fractionation Plasma lipoproteins were isolated from plasma by density gradient ultracentrifugation detailed in Supplemental Methods [18]. 2.5. Determination of endogenous CETP activity and free cholesterol efflux assays Determination of endogenous cholesterol ester (CE) transfer from HDL to apolipoprotein B-containing lipoproteins [18] and free cholesterol efflux assays [19] were studied as detailed in Supplemental Data. 3. Results We presently identified, among 75 ADH probands recruited in the Lipid Clinic of La Pitié Salpêtrière hospital, for whom the APOBp.Arg3527Gln mutation and mutations in the LDLR gene had been excluded, four point mutations in the PCSK9 gene, of which two were novel. These PCSK9 mutations segregated with ADH in each family and were not found in normocholesterolemic controls. The first mutation corresponding to a substitution, p.Leu108Arg (c.323T > G), in exon 2 of PCSK9 was identified in a Black family (Family A) originating from the Republic of Mauritius (Fig. 1). The 53 y.o proband (A.II.4) was ascertained at 41 years old with
396
M. Abifadel et al. / Atherosclerosis 223 (2012) 394e400
Fig. 1. Pedigree and genetic analysis of families A and B. Birth date, Age (in years) at given lipid measurement, total cholesterol (TC), low density lipoprotein cholesterol (LDL-C), High density lipoprotein (HDL-C) and triglycerides (TG) levels are given in mmol/l (with untreated values for affected individuals). The probands are indicated by an arrow. (A) Sequence analysis in family A. The proband and affected members are heterozygous with respect to the p.Leu108Arg mutation. (B) Sequence analysis in family B. The proband and his son are heterozygous with respect to p.Asp35Tyr substitution in exon 1 of PCSK9.
9.44 mmol/l of TC; 7.83 mmol/l of LDL-C; with tendon xanthomas. More recently he developed type II diabetes with glucose levels of 7.89 mmol/l. Intima media thickness (IMT) measurements performed for AII.4 patient showed an increase of both common carotid and common femoral arteries with a right femoral plaque. He was recently hospitalized with an acute coronary syndrome and a right coronary artery lesion. The proband’s family has been recruited. One of his three children (A.III.3) suffers from
hypercholesterolemia and carries also the p.Arg108Leu mutation. The proband’s sister (A.II.8) has severe hypercholesterolemia with 10.39 mmol/l of total cholesterol and 8.61 mmol/l of LDL-cholesterol at 47 y.o, before any lipid lowering therapy, and an increase of both carotid and femoral IMT with microcalcifications and a slight femoral plaque. She has a 20 y.o son (A.III.8) who is hypercholesterolemic and carries also the mutation. Her brother suffered myocardial infarction (A.II.7) at 45 y.o and her husband (who is also
M. Abifadel et al. / Atherosclerosis 223 (2012) 394e400
her cousin) died from myocardial infarction (A.II.9). The p.Arg108Leu mutation of PCSK9 was found in all affected patients of the family whose DNA was available (A. II.4; II.8; III.3; III.8). The second mutation corresponds to a substitution p.Asp35Tyr (c.103G > T) in exon 1 of the PCSK9 gene. It was identified in the Family B proband (B.I.1) with 8.14 mmol/l of TC and 6.05 mmol/l of LDL-C at 55 y.o. His echodoppler at 62 y.o. showed slight left femoral plaque and right femoral calcification with no carotid lesion. One of his two sons (B.II.1) is hypercholesterolemic with 7.76 mmol/l of total cholesterol and 6.10 mmol/l of LDL-C and was found to carry the p.Asp35Tyr mutation in PCSK9. It is noteworthy that two previously reported mutants of PCSK9, p.S127R [3,13] and p.L21tri [13], have also been identified in 2 of the 75 non LDLR/non APOB probands. Thus, mutations in PCSK9 have been found in 5.3% of the ADH probands, with no mutation in the LDLR or the APOB genes, recruited in this study. Biosynthetic analyses showed upon overexpression of either the WT or the p.Asp35Tyr and p.Leu108Arg mutants in HEK293 (or HepG2) cells, that all forms of PCSK9 were equally well autocatalytically cleaved and secreted as a complex with their prosegment (Fig. 2A; supplemental Fig. 1A). Furthermore, all forms were similarly susceptible to Furin cleavage at Arg218Y, resulting in comparative levels of PCSK9-DN218 [8,9] in the media of HEK293 cells. Using an ELISA assay we estimated the levels of secreted PCSK9 and showed that WT, p.Asp35Tyr and p.Leu108Arg mutants were secreted from HEK293 cells comparatively well, giving 24 h levels of w1600, 1400 or 1900 ng/ml respectively. Human hepatic HuH7 cells were then incubated for 4 h with equal amounts of each protein (730 ng/ml) and the levels of cell surface LDLR were measured by FACS [20]. The data show that the reduction in LDLR afforded by the p.Asp35Tyr mutant is significantly slightly higher than with WT PCSK9, but that the p.Leu108Arg mutant exhibited a marked w2-fold enhanced degrading activity towards the LDLR, resulting in a clear and significant gain-of-function in this assay (Fig. 2B). In an effort to rationalize the mechanism behind the presumably gain-of-function mutation p.Asp35Tyr, we postulated that this
397
may result in a possible new Tyr-sulfation site, similar to the WT Tyr38 sulfation [8]. Indeed, upon incubation of HEK293 cells overexpressing either WT or p.Asp35Tyr PCSK9 with Na35 2 SO4, we observed a large increase in the level of sulfated prosegment in the cells and media consistent with a double sulfation at Tyr35 (Asp35Tyr) and Tyr38 (WT) in the p.Asp35Tyr mutant (Fig. 3A). As a negative control we showed that the single mutant p.Tyr38Phe in the WT sequence is not sulfated. To prove that the Asp35Tyr did not cause an over sulfation of Tyr38, we obtained the double mutant p.Asp35Tyr þ p.Tyr38Phe. Repeating the above Na35 2 SO4 incorporation in either HepG2 cells (Fig. S1B) or HEK293 cells revealed that indeed preventing sulfation of Tyr38 in the mutant Tyr38Phe, sulfation still occurred independently at Tyr35 (Fig. 3B), confirming that the latter is a bona fide sulfation site. Finally, repeating the incubation of HuH7 cells with 420 ng/ml of each of these above PCSK9 constructs (estimated by ELISA), we confirmed the results of Fig. 1. The data also revealed that: (i) the gain-offunction of the p.Leu108Arg mutant is less than that of the p.Asp374Tyr mutant, (ii) lack of Tyr38 sulfation does not affect the activity of PCSK9 on LDLR, and (iii) sufation at Tyr35 results in lower secretion (as also observed for p.Asp374Tyr) but slightly higher activity (Figure S2). Plasma lipid and apolipoprotein analyses showed, by comparison with control values obtained from healthy normolipidemic subjects, that patients carrying the p.Leu108Arg (AII.4, AII.8, AIII.3 and AIII.8) or the p.Asp135Tyr (BI.1 and BII.1) mutations in the PCSK9 gene displayed a marked elevation in their mean plasma LDL-C concentration (6.45 1.55 mmol/l and 2.84 0.52 mmol/l in ADH patients and normolipidemic controls respectively, p < 0.0001). Despite the fact that no major difference was observed in plasma HDL-C between PCSK9 mutants and control subjects the relative proportion of HDL-C was significantly reduced in ADH patients as compared with normolipidemic subjects (15.7 4.0% and 31.2 3.4%, in ADH patients and in control subjects respectively, p < 0.0001). Their response to lipid lowering therapy is detailed in Supplemental Data and Supplemental Table 1.
Fig. 2. Biosynthetic analyses and Extracellular activity of PCSK9 and its mutants. (A) HEK293 cells were transiently transfected with cDNAs coding for WT, p.Asp35Tyr (D35Y) or p.Leu108Arg (L108R) mutants all tagged with V5 at their C-terminus. The next day, the cells were incubated with [35S]Met/Cys for 3 h and the cell lysates and media immunoprecipitated with an anti-V5 mAb. The immunoprecipitates were then resolved by SDS-PAGE and the radiolabeled proteins identified by autoradiography, as reported [9]. The migration positions of proPCSK9, PCSK9 and the Furin cleaved product PCSK9-DN218 are emphasized, as well as the associated inhibitory prosegment. (B) Extracellular activity of PCSK9 and its mutants. HEK293 cells were transiently transfected with cDNAs coding for WT, p.Asp35Tyr or p.Leu108Arg mutants tagged with V5 at their C-terminus. The next day the cells were washed and incubated in RPMI þ 5% LPDS þ 1% sodium pyruvate, and the secreted proteins collected for 24 h, and analyzed by a sensitive ELISA assay for human PCSK9 [17] and the triplicate expressed in ng/ml. 730 ng/ml of each protein construct were then incubated with HuH7 cells for 4 h and the cell surface LDLR levels estimated by FACS analysis in triplicate, as reported [17].
398
M. Abifadel et al. / Atherosclerosis 223 (2012) 394e400
Fig. 3. Tyr Sulfation. HEK293 cells were transiently transfected with cDNAs coding for WT, the mutants (A,B) Asp35Tyr (D35Y), Tyr38Phe (Y38F), or (B) the double mutant Asp35Tyr þ Tyr38Phe (D35Y þ Y38F) or an empty vector (pIRES) all tagged with V5 at their C-terminus. The next day, the cells were incubated with Na35 2 SO4 for 2 h and the cell lysates and media immunoprecipitated with an anti-V5 mAb. The immunoprecipitates were then resolved by SDS-PAGE and the radiolabeled proteins identified by autoradiography, as reported [6,8]. Notice how the p.Asp35Tyr mutant exhibits a much higher level of secreted Tyr-sulfated prosegment in the medium. The double mutant shows that the absence of Tyr-sulfation at Tyr38 did not prevent sulfation at Tyr35.
The distribution of cholesterol among density gradient subfractions obtained from patients carrying PCSK9 mutations and normolipidemic controls as well as plasma concentrations of major lipoprotein fractions are presented in Fig. S3 and in Supplemental Table 2, respectively and detailed in Supplemental Data. The mean weight percent chemical composition of major lipoprotein fractions isolated from the plasma of ADH patients carrying PCSK9 mutations and control subjects are presented in Supplemental Table 3. No significant difference in the relative proportion of individual lipids in either apoB- or apoAI-containing lipoprotein particles was detected between ADH patients and normolipidemic control subjects. As shown in Supplemental Figure 4, endogenous plasma CETP activity, expressed as percentage of total CE transferred from HDL to apoB-containing lipoproteins was significantly increased in all ADH patients carrying the p.Leu108Arg (AII.4, AII.8, AIII.3 and AIII.8) mutation in the PCSK9 gene. By contrast, the BI.1 patient carrying the p.Asp135Tyr mutation in the PCSK9 gene (BI.1 and BII.1) displayed an endogenous plasma CETP activity similar to that observed in control subjects. It is relevant to note that a strong positive correlation between CETP activity and plasma levels of LDL-C exits (r ¼ 0.85; p ¼ 0.0003) indicating that the elevation of endogenous plasma CETP activity presently observed in some ADH patients carrying PCSK9 mutations results primarily from the increased levels of apoB-containing lipoprotein cholesteryl ester acceptors. Finally, we evaluated the overall capacity of total plasma of ADH patients carrying PCSK9 mutations to promote cellular free cholesterol efflux from cholesterol loaded macrophage [19]. As shown in Supplemental Data and Figure S5A, no significant difference was observed in the plasma efflux capacity of patients carrying the p.Leu108Arg or p.Asp135Tyr mutation in the PCSK9 gene as compared to control subjects. Plasma efflux capacity via the ABCA1 pathway was similar for both ADH patients carrying PCSK9 mutations and controls (Fig. S5B). In addition, we did not observe any significant differences in the capacity of either larger HDL2 or smaller HDL3 particles to promote cellular efflux via SR-BI (Fig. S5CD) or ABCG1 (Fig. S5EF) between these ADH patients and control subjects (Supplemental data).
4. Discussion Here we report, two non LDLR/non APOB families carrying two distinct new molecular defects located on the PCSK9 gene corresponding to distinct amino acid substitutions within the prosegment of PCSK9: p.Asp35Tyr and p.Leu108Arg. These mutations segregated with ADH in each family and were not found in normocholesterolemic controls. In these subjects, familial hypercholesterolemia leads to premature atherosclerosis with coronary lesions and cardiovascular complications. The aspartate at position 35 is highly conserved between most species (primates, Mus musculus, Rattus norvegicus) but not in Tetraodon nigroviridis, nor in Zebrafish, where it is replaced by a tyrosine. While Leucine at position 108 is well conserved in primates, it is not conserved in M. musculus and R. norvegicus. The in silico estimation of the impact of each variant on the protein using the SIFT tools showed that the p.Leu108Arg was well tolerated while the p.Asp35Tyr was not. However, using Polyphen tools, the p.Leu108Arg seems to be possibly damaging (PSIC score difference 1.752) and the p.Asp35Tyr probably damaging (score difference 2.090). Thus, theoretical predictions have limitations and we wished to define experimentally the impact of these mutations on proPCSK9 zymogen activation into PCSK9 in the endoplasmic reticulum, the subsequent secretion of the latter out of the cell and its ability to enhance the degradation of the cell surface LDLR. Biosynthetic analyses showed that the WT or the p.Leu108Arg and p.Asp35Tyr mutants were equally well autocatalytically cleaved and secreted as a complex with their prosegment and were similarly susceptible to Furin cleavage at Arg218Y. Furthermore the WT, p.Asp35Tyr and p.Leu108Arg mutants were secreted from HEK293 cells comparatively well. This suggested that these mutations did not result in gross structural alterations or folding of proPCSK9. However, when incubated with HuH7 cells the p.Leu108Arg variant of PCSK9 showed a w2-fold enhanced degrading activity towards the LDLR by FACS measurements compared to WT PCSK9 (Fig. S2). This was a clear evidence that in this assay that measures the extracellular activity of PCSK9 on the LDLR, the p.Leu108Arg mutation results in a gain-of-function, likely due to the positive charge of Arg. Comparatively, the reduction in LDLR afforded by WT PCSK9 was slightly less than that seen with
M. Abifadel et al. / Atherosclerosis 223 (2012) 394e400
the p.Asp35Tyr mutant, suggesting that the latter exhibits a mild gain-of-function activity in this assay. In an effort to rationalize the mechanism behind the gain-of-function mutation p.Asp35Tyr, we noticed that the resulting Tyr35 is surrounded by acidic residues and could potentially be a novel Tyr-sulfation site [21]. Radioactive sulfate incorporation experiments in either HEK293 or HepG2 cells showed that indeed this mutation resulted in a new Tyr-sulfation site (Fig. 3, S1B) similar to the wild type Tyr38 sulfation [8]. Furthermore, the double mutant p.Asp35Tyr þ p.Asp38Phe is still sulfated (Fig. S1), confirming that sulfation occurs at Tyr35 independently from that at Tyr38, and that sulfation at Tyr38 does not change the activity of PCSK9 towards the LDLR (Fig. S2). Even though our recent data demonstrated that the negative charge in the N-terminal portion of the prosegment acts as an inhibitor to PCSK9 function on the LDL receptor [19], replacement of Asp35 with a negatively charged sulfated Tyr (Tyr-SO 3 ) may mask this phenomenon and result in a gain-of-function by modifying the secondary structure of this disordered portion of the prosegment [22]. A similar observation that loss of acidic residues in the Nterminal part of the prosegment is associated with hypercholesterolemia, was also reported in heterozygote or homozygote Japanese subjects exhibiting the p.Glu32Lys mutation [23], which we also could not correlate with a marked enhanced extracellular activity of PCSK9 (Seidah NG, unpublished observation). It is therefore plausible that the p.Asp35Tyr or p.Glu32Lys mutations may rather enhance the intracellular activity of PCSK9 [12], an effect that would be missed in the extracellular activity measured here (Fig. 2). In an effort to rationalize the structural consequence of the p.Leu108Arg mutation on the PCSK9-LDLR complex, we analyzed the recently made publicly available 3D model of PCSK9 in complex with the complete ectodomain of the LDLR (http://www.pdb.org/ pdb/explore/explore.do?structureId¼3M0C). Amazingly, the data show that Leu108 of the prosegment of PCSK9 makes a direct hydrophobic interaction with Leu626 within a loop structure of the b-propeller domain of the LDLR (Fig. S5eS8). The nearby Arg385 of the LDLR (Fig. S7,8) might thus be repelled by the Arg substitution of Leu108 in the prosegment of the PCSK9, and the latter mutant may possibly electrostatically bind the nearby Glu605 of the LDLR, thereby favoring a tighter complex of the PCSK9 with the LDLR. Thus, we believe that the p.Leu108Arg substitution is a gain-offunction due to the re-orientation of the interaction of Leu626 in the LDLR with Leu108 of the prosegment of PCSK9 towards a more stabilizing interaction within the PCSK9-LDLR complex via an ionic interaction of the resulting mutant Arg108 of PCSK9 with the neighboring Glu605 within the b-propeller domain of the LDLR (Fig. S6eS8). It is noteworthy that the details of the beta propeller/ pro-domain interface have been very recently reported by Lo Surdo et al. and showed the close proximity of residues associated with familial hypercholesterolemia: S127, D129 and now L108 [24]. In the present study, we observed that ADH caused by mutation on the PSCK9 gene, was not associated with abnormal physicochemical and biological properties of HDL particles, and in particular in their capacity to promote cellular free cholesterol efflux (Supplemental data and Fig. S5). It is relevant to consider that in this way ADH caused by mutations in the PSCK9 gene differs from those involving mutations in the LDLR gene (discussed in Supplemental data). As compared to normolipidemic subjects, we presently observed that CETP-mediated CE transfer from HDL to apoB-containing lipoproteins is enhanced in ADH patients carrying the PCSK9 mutations as a result of increased plasma levels of cholesteryl ester acceptors (Fig. S4). This observation is in good agreement with previous studies showing an accelerated CE transfer in familial hypercholesterolemia as a result of increased plasma levels in both CETP mass and apoB-containing cholesteryl ester acceptors [25]. In addition, we presently observed that elevated CETP activity is
399
associated with a predominance of denser LDL subfractions in the plasma from ADH patients carrying PCSK9 mutations, highlighting the pro-atherogenic role of CETP in the formation and accumulation of atherogenic small dense LDL particles. It is noteworthy that most of the ADH patients, in this study, were under lipid lowering therapy which could limit the interpretation of the results. We thus cannot exclude the possibility that pharmacological therapies used to reduce plasma cholesterol levels in hypercholesterolemic patients might at least in part contribute to normalize some of the abnormal quantitative and qualitative features of HDL particles, in particular their capacity to promote cellular free cholesterol efflux. In conclusion, in this work we identified two new gain-of-function mutations of PCSK9 (p.Leu108Arg and p.Asp35Tyr) in ADH families with no mutations in the LDLR or APOB genes. We studied in vitro, the impact of these mutations on PCSK9 processing and on its activity on cell surface LDLR levels, and showed that the p.Leu108Arg mutation clearly results in a gain-of-function, while the p.Asp35Tyr mutation created a novel Tyr-sulfation site, which may enhance the intracellular activity of PCSK9. Furthermore, we evaluated the impact of these mutations on both quantitative and qualitative features of lipoprotein particles and on the HDL-mediated cellular cholesterol efflux. These data further contribute to the characterization of PCSK9 mutations and allow us to better understand the impact on cholesterol metabolism of this new therapeutic target in hypercholesterolemia. Acknowledgments We are indebted to the family members for their cooperation. This work was supported by grants from ANR (ANR-05-PCOD-017, ANR-06-MRAR-038, ANR-08-GENO-002-01), PHRC (AOM06024), Institut National de la Santé et de la Recherche Médicale and Conseil de la Recherche de l’Université Saint-Joseph (Beirut, Lebanon : to MA). This work was supported in part (to NGS & AP) by CIHR grants # CTP 82946, MOP 102741 and a Canadian Chair # 216684 and a Strauss Foundation (to NGS). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.atherosclerosis.2012.04.006. References [1] Goldstein JL, Brown MS. Familial hypercholesterolemia: pathogenesis of a receptor disease. Johns Hopkins Med J 1978;143:8e16. [2] Innerarity TL, Weisgraber KH, Arnold KS, Mahley RW, Krauss RM, Vega GL, et al. Familial defective apolipoprotein B-100: low density lipoproteins with abnormal receptor binding. Proc Natl Acad Sci USA 1987;84:6919e23. [3] Abifadel M, Varret M, Rabes JP, Allard D, Ouguerram K, Devillers M, et al. Mutations in PCSK9 cause autosomal dominant hypercholesterolemia. Nat Genet 2003;34:154e6. [4] Abifadel M, Pakradouni J, Collin M, Samson-Bouma ME, Varret M, Rabès JP, et al. Strategies for proprotein convertase subtilisin kexin 9 modulation: a perspective on recent patents. Expert Opin Ther Pat 2010;20:1547e71. [5] Seidah NG, Benjannet S, Wickham L, Marcinkiewicz J, Jasmin SB, Stifani S, et al. The secretory proprotein convertase neural apoptosis-regulated convertase 1 (NARC-1): liver regeneration and neuronal differentiation. Proc Natl Acad Sci USA 2003;100:928e33. [6] Benjannet S, Rhainds D, Essalmani R, Mayne J, Wickham L, Jin W, et al. NARC-1/PCSK9 and its natural mutants: zymogen cleavage and effects on the low density lipoprotein (LDL) receptor and LDL cholesterol. J Biol Chem 2004; 279:48865e75. [7] McNutt MC, Lagace TA, Horton JD. Catalytic activity is not required for secreted PCSK9 to reduce low density lipoprotein receptors in HepG2 cells. J Biol Chem 2007;282:20799e803. [8] Benjannet S, Rhainds D, Hamelin J, Nassoury N, Seidah NG. The proprotein convertase (PC) PCSK9 is inactivated by Furin and/or PC5/6A: functional consequences of natural mutations and post-translational modifications. J Biol Chem 2006;281:30561e72.
400
M. Abifadel et al. / Atherosclerosis 223 (2012) 394e400
[9] Essalmani R, Susan-Resiga D, Chamberland A, Abifadel M, Creemers JW, Boileau C, et al. In vivo evidence that furin from hepatocytes inactivates PCSK9. J Biol Chem 2011;286:4257e63. [10] Lagace TA, Curtis DE, Garuti R, McNutt MC, Park SW, Prather HB, et al. Secreted PCSK9 decreases the number of LDL receptors in hepatocytes and in livers of parabiotic mice. J Clin Invest 2006;116:2995e3005. [11] Cameron J, Holla OL, Ranheim T, Kulseth MA, Berge KE, Leren TP. Effect of mutations in the PCSK9 gene on the cell surface LDL receptors. Hum Mol Genet 2006;15:1551e8. [12] Poirier S, Mayer G, Poupon V, McPherson PS, Desjardins R, Ly K, et al. Dissection of the endogenous cellular pathways of PCSK9-induced low density lipoprotein receptor degradation: evidence for an intracellular route. J Biol Chem 2009;284:28856e64. [13] Abifadel M, Rabès JP, Devillers M, Munnich A, Erlich D, Junien C, et al. Mutations and polymorphisms in the proprotein convertase subtilisin kexin 9 (PCSK9) gene in cholesterol metabolism and disease. Hum Mut 2009;30:520e9. [14] Cohen JC, Boerwinkle E, Mosley Jr TH, Hobbs HH. Sequence variations in PCSK9, low LDL, and protection against coronary heart disease. N Engl J Med 2006;354:1264e72. [15] Brown MS, Goldstein JL. Biomedicine. Lowering LDLenot only how low, but how long? Science 2006;311:1721e3. [16] Siest G, Visvikis S, Herbeth B, Gueguen R, Vincent-Viry M, Sass C, et al. Objectives, design and recruitment of a familial and longitudinal cohort for studying gene-environment interactions in the field of cardiovascular risk: the Stanislas cohort. Clin Chem Lab Med 1998;36:35e42. [17] Dubuc G, Tremblay M, Pare G, Jacques H, Hamelin J, Benjannet S, et al. A new method for measurement of total plasma PCSK9: clinical applications. J Lipid Res 2010;51:140e9.
[18] Catalano G, Julia Z, Frisdal E, Vedie B, Fournier N, Le Goff W, et al. Torcetrapib differentially modulates the biological activities of HDL2 and HDL3 particles in the reverse cholesterol transport pathway. Arterioscler Thromb Vasc Biol 2009;29:268e75. [19] Wang N, Lan D, Chen W, Matsuura F, Tall AR. ATP-binding cassette transporters G1 and G4 mediate cellular cholesterol efflux to high-density lipoproteins. Proc Natl Acad Sci USA 2004;101:9774e9. [20] Catalano G, Julia Z, Frisdal E, Vedie B, Fournier N, Le Goff W, et al. Effects of the prosegment and pH on the activity of PCSK9: evidence for additional processing events. J Biol Chem 2010;285:40965e78. [21] Beisswanger R, Corbeil D, Vannier C, Thiele C, Dohrmann U, Kellner R, et al. Existence of distinct tyrosylprotein sulfotransferase genes: molecular characterization of tyrosylprotein sulfotransferase-2. Proc Natl Acad Sci USA 1998; 95:11134e9. [22] Cunningham D, Danley DE, Geoghegan KF, Griffor MC, Hawkins JL, Subashi TA, et al. Structural and biophysical studies of PCSK9 and its mutants linked to familial hypercholesterolemia. Nat Struct Mol Biol 2007;14:413e9. [23] Miyake Y, Kimura R, Kokubo Y, Okayama A, Tomoike H, Yamamura T, et al. Genetic variants in PCSK9 in the Japanese population: rare genetic variants in PCSK9 might collectively contribute to plasma LDL cholesterol levels in the general population. Atherosclerosis 2008;196:29e36. [24] Surdo PL, Bottomley MJ, Calzetta A, Settembre EC, Cirillo A, Pandit S, et al. Mechanistic implications for LDL receptor degradation from the PCSK9/LDLR structure at neutral pH. EMBO Rep 2011;12:1300e5. [25] De Grooth GJ, Smilde TJ, Van Wissen S, Klerkx AH, Zwinderman AH, Fruchart JC, et al. The relationship between cholesteryl ester transfer protein levels and risk factor profile in patients with familial hypercholesterolemia. Atherosclerosis 2004;173:261e7.