Sorting an LDL receptor with bound PCSK9 to intracellular degradation

ReviewSorting an LDL receptor with bound PCSK9to intracellular degradation Trond P.Leren*Unit for Cardiac and Cardiovascular Genetics,Department of Medical Genetics,Oslo University Hospital,Oslo,Norwaya r t i c l e i n f o Article history:Received18June2014Received in revised form28July2014Accepted19August2014 Available online2September2014Keywords:DegradationCathepsinEndosomeLDL receptorPCSK9a b s t r a c tObjective:This article reviews the mechanism by which the low density lipoprotein receptor(LDLR)that has bound proprotein convertase subtilisin/kexin type9(PCSK9),is rerouted to intracellular degradation instead of being recycled.Methods:A search of relevant published literature has been conducted.Results:PCSK9binds to the LDLR at the cell surface.It is the catalytic domain of PCSK9that binds to the epidermal growth factor repeat A of the LDLR.The LDLR:PCSK9complex is internalized through clathrin-mediated endocytosis.Due to an additional electrostatic interaction at acidic pH between the C-terminal domain of PCSK9and the ligand-binding domain of the LDLR,PCSK9remains bound to the LDLR in the sorting endosome.As a consequence,the LDLR fails to adopt a closed conformation and is degraded instead of being recycled.The mechanism for the failure of the LDLR to recycle appears to involve ectodomain cleavage of the extended LDLR by a cysteine cathepsin in the sorting endosome.The cleaved LDLR ectodomain will be confined to the vesicular part of the sorting endosome for degradation in the endosomal/lysosomal tract.Conclusion:Ectodomain cleavage of an LDLR with bound PCSK9in the sorting endosome disrupts the normal recycling of the LDLR.©2014Elsevier Ireland Ltd.All rights reserved.Contents1.LDL receptor-mediated endocytosis (76)2.Proprotein convertase subtilisin/kexin type9 (77)3.PCSK9-mediated degradation of the LDLR (77)4.Role of the cytoplasmic domain of the LDLR (77)5.Role of the ectodomain of the LDLR (78)6.Role of the C-terminal domain of PCSK9 (78)7.Ectodomain cleavage (79)Conflict of interest (80)References (80)1.LDL receptor-mediated endocytosisApproximately70%of plasma cholesterol is carried in low density lipoprotein(LDL)[1],and high levels of LDL cholesterol are a major risk factor for coronary heart disease.LDL is cleared from plasma by binding to the cell-surface LDL receptor(LDLR)and is internalized by LDLR-mediated endocytosis[2](Fig.1).The importance of the LDLR in cholesterol metabolism is illustrated by the severe hypercholesterolaemia in subjects who lack LDLRs[2].The LDLR is a Type1transmembrane protein and hasfive do-mains[3e5](Fig.1).The amino terminal ligand-binding domain consists of seven repeats of approximately40amino acids each.The epidermal growth factor(EGF)precursor homology domain of approximately400amino acids consists of two40amino acid re-peats,EGF-A and EGF-B,a280amino acid b-propeller and a third40 amino acid repeat,EGF-C.The O-linked sugar domain consists of58*Unit for Cardiac and Cardiovascular Genetics,Department of Medical Genetics, Oslo University Hospital,P.O.Box4956Nydalen,NO-0424Oslo,Norway.Tel.:þ47 23026389;fax:þ4722118699.E-mail addresses:trond.leren@rikshospitalet.no,TrondPaul.Leren@ rikshospitalet.no.Contents lists available at ScienceDirectAtherosclerosisjo urn al homepag e:/locate/at herosclerosis/10.1016/j.atherosclerosis.2014.08.0380021-9150/©2014Elsevier Ireland Ltd.All rights reserved.Atherosclerosis237(2014)76e81amino acids.The transmembrane domain has 22amino acids and the cytoplasmic domain which contains the motifs required for concentrating the LDLR in clathrin-coated pits,has 50amino acids.LDL which binds to the ligand-binding domain at the cell sur-face,is released from the LDLR at the acidic pH of the sorting en-dosome [6](Fig.1).Released LDL is con fined to the vesicular part of the sorting endosome which matures to become a late endosome [6].LDL is finally degraded in the lysosome.Concomitant with the release of LDL in the sorting endosome,the LDLR folds back on itself to obtain a closed conformation whereby ligand-binding repeats 4and 5interact with the b -propeller [7](Fig.1).The LDLR then re-cycles back to the cell surface.2.Proprotein convertase subtilisin/kexin type 9Proprotein convertase subtilisin/kexin type 9(PCSK9)has been shown to play a major role in regulating plasma LDL cholesterol levels through its ability to mediate intracellular degradation of the LDLR [8e 10].PCSK9is a zymogen consisting of 692amino acids of which residues 1e 30constitute the signal peptide [8].The remainder of the protein is divided into three domains.The prodomain consists of residues 31e 152,the catalytic domain consists of residues 153e 454and the cysteine-and histidine-rich C-terminal domain consists of residues 455e 692[11,12].After intramolecular auto-catalytic cleavage in the endoplasmic reticulum,the cleaved pro-domain is non-covalently bound to the catalytic domain of the mature PCSK9to block enzymatic activity and to act as a chaperone to promote proper folding and exit out of the endoplasmic reticu-lum [8].In contrast to other proprotein convertases,PCSK9does not undergo a second cleavage to release the prodomain [8].PCSK9is therefore secreted as an enzymatic inactive protein.3.PCSK9-mediated degradation of the LDLRAt the cell surface,the catalytic domain of PCSK9binds to the EGF-A repeat of the LDLR [9,13](Fig.2).The residues of the LDLR and PCSK9involved in this interaction have been well characterized[9,13,14].Following binding of PCSK9to the LDLR at the cell surface,the LDLR:PCSK9complex is internalized through clathrin-coated pits [9,15].Whereas,LDL is released from the LDLR at the acidic milieu of the sorting endosome,the af finity of PCSK9to bind to the LDLR increases 120e 150fold at acidic pH [11,12,16,17].This makes PCSK9remain bound to the LDLR in the sorting endosome and prevents the LDLR from adopting a closed conformation at acidic pH [18](Fig.2).There is a 1:1stoichiometry between the LDLR and PCSK9.Thus,PCSK9does not exhibit a general effect on the LDLRs within a clathrin-coated pit or within a sorting endosome,but rather acts as a tag to promote degradation of an individual LDLR.The question then is how an LDLR with bound PCSK9is identi fied in the sorting endosome and rerouted to intracellular degradation.4.Role of the cytoplasmic domain of the LDLRFor several receptors that are rerouted to intracellular degra-dation after having bound their ligands,such as the EGF receptor,ligand-binding introduces a conformational change of the cyto-plasmic domain which leads to ubiquitination and activation of the endosomal sorting complex required for traf ficking (ESCRT)ma-chinery [19].This machinery translocates these receptors to the vesicular part of the sorting endosome for transport down the endosomal/lysosomal tract.However,because an LDLR lacking the cytoplasmic domain also undergoes PCSK9-mediated degradation [20],modi fications of the cytoplasmic domain or an interaction between the cytoplasmic domain and ubiquitin or adaptor proteins,are not involved in rerouting the LDLR with bound PCSK9to intracellular degradation.An alternative mechanism for intracellular degradation could involve an interaction within the endosomal lumen between a co-receptor on one hand and PCSK9or the LDLR:PCSK9complex on the other hand.This could lead to ubiquitination of the cytoplasmic domain of the co-receptor and activation of the ESCRT machinery.However,data from Wang et al.[15]showing that PCSK9is able to degrade the LDLR when ubiquitination is inhibited or the ESCRT machinery is inactivated,exclude such a mechanism.Wang et al.[15]have also shown that the LDLR is degraded when proteasomal activity is inhibited or when the autophagocytic pathway is inac-tivated.Thus,PCSK9-mediated degradation of the LDLR doesnotFig.1.The normal function of the LDLR.The ligand-binding domain of the LDLR with its seven ligand-binding repeats,is shown in yellow.The EGF-A,EGF-B and EGF-C repeats of the EGF precursor homology domain are shown in blue and the b -propel-ler is shown in red.The O-linked sugar domain,the transmembrane domain and the cytoplasmic domain are indicated by a solid black line.The cell membrane is shown as a horizontal structure.At the cell surface the ectodomain of the LDLR is extended from the cell membrane and the LDLR binds LDL through its ligand-binding domain (left).At the acidic pH of the sorting endosome,LDL is released from the LDLR and the LDLR adopts a closed conformation whereby ligand-binding repeats 4and 5interact with the b -propeller (right).The figure is not drawn toscale.Fig.2.PCSK9prevents the LDLR from adopting a closed conformation in the sorting endosome.PCSK9consisting of a prodomain (P),a catalytic domain (CAT)and a C-terminal domain (CTD)binds to the EGF-A repeat of the LDLR at the cell surface through its catalytic domain.At the acidic milieu of the sorting endosome,the C-ter-minal domain of PCSK9obtains an increased positive charge and electrostatically at-tracts the negatively charged ligand-binding domain of the LDLR.This prevents the LDLR from adopting a closed conformation at acidic pH.The figure is not drawn to scale.T.P.Leren /Atherosclerosis 237(2014)76e 8177involve commonly used pathways for lysosomal or proteasomal degradation.5.Role of the ectodomain of the LDLRBecause the cytoplasmic domain of the LDLR is not involved in the intracellular degradation[20],it is possible that elements of the ectodomain are involved in targeting an LDLR with bound PCSK9to intracellular degradation.Regarding the ligand-binding domain of the LDLR,it is not required for the LDLR to bind and internalize PCSK9[16,18,21],although an LDLR lacking the ligand-binding domain appears to internalize PCSK9less efficiently than the full-length LDLR[21].However,the ligand-binding domain plays a critical role in rerouting the LDLR to intracellular degradation.From a series of deletion experiments,Zhang et al.[18]have shown that at least three of the seven ligand-binding repeats are required for an LDLR to undergo intracellular degradation.From their data it appears that the more ligand-binding repeats that are present,the more efficiently is the LDLR degraded.Moreover,no single ligand-binding repeat is required[18].Rather,it is the number of repeats that is important.Thesefindings make it unlikely that specific residues of the ligand-binding domain are engaged in a key and lock interaction with another protein such as an endosomal co-receptor or PCSK9.Rather,there may be some general features of the ligand-binding repeats or the length of the ligand-binding domain,that are important.A common feature of the ligand-binding repeats is their negative charge due to clusters of acidic residues which coordinate one calcium ion[5].Studies of cells expressing a mutant LDLR lacking the ligand-binding domain have shown that this mutant LDLR is able to internalize substantial amounts of PCSK9[18,21].Together with the observation that an LDLR lacking the ligand-binding domain is not degraded,but instead recycles back to the cell membrane[18,21], indicate that internalized PCSK9is released from an LDLR lacking the ligand-binding domain in the sorting endosomes.Thus,the ligand-binding domain could be required for PCSK9to remain bound to the LDLR at the acidic milieu and low calcium concen-tration of the sorting endosome.To study this,Tveten et al.[21]performed studies to determine the ability of LDLRs with varying number of ligand-binding repeats to bind PCSK9at acidic conditions.CHO T-REx cells stably trans-fected with the various plasmids were incubated with the gain-of-function mutant D374Y-PCSK9at4 C to study the amount of PCSK9 bound to the cells as a function of pH of the culture medium. Whereas,PCSK9remained stably bound to the wild-type LDLR when pH was raised from5.3to pH6.9,PCSK9was released from an LDLR lacking the ligand-binding domain within few minutes at pH 6.1or higher[21].Moreover,an LDLR with three ligand-binding repeats bound less PCSK9than the wild-type LDLR at acidic con-ditions,but bound more PCSK9than an LDLR with two ligand-binding repeats,particularly at pH 6.5and below.These data indicate that the negatively charged ligand-binding domain of the LDLR is required for PCSK9to remain bound to the LDLR during the early phase of endosomal acidification when pH gradually de-creases from pH7.4to pH~6.1.The effect of varying number of ligand-binding repeats on the amount of PCSK9bound at these mildly acidic conditions,are in agreement with the data of Zhang et al.[18]regarding the requirement for a certain number of ligand-binding repeats for PCSK9to promote intracellular degradation of the LDLR.6.Role of the C-terminal domain of PCSK9The C-terminal domain of PCSK9consists of three modules where each module is built from six b-strands connected with three disulfide bridges[11,12].Even though the C-terminal domain of PCSK9is not required for PCSK9to bind and be internalized by the LDLR,the C-terminal domain is required for PCSK9to disrupt recycling of the LDLR[11,18,22,23].Whereas,some studies have failed to observe an interaction between the C-terminal domain and the extracellular part of the LDLR at pH7.4or at pH6.0[11,18,22],Yamamoto et al.[24]have shown that the C-terminal domain of PCSK9does bind to the ligand-binding domain of the LDLR,particularly at low pH.In that study,Yamamoto et al.[24]attached a soluble LDLR to an affinity chromatography column and studied the amount of wild-type or mutant PCSK9bound to the column at pH7.3or pH5.4.One of their findings was that when the C-terminal domain was passed over the column,significant amounts bound to the LDLR at pH5.4,but only small amounts bound at pH7.3.Further experiments showed that it was the ligand-binding domain of the LDLR that bound the C-ter-minal domain of PCSK9at acidic pH[24].The C-terminal domain of PCSK9is less evolutionary conserved than the catalytic domain[14].Apart from the18cysteines involved in forming disulfide bonds,only12of the remaining residues are highly conserved.In the study of Holla et al.[23]a total of14 conserved residues of the C-terminal domain of PCSK9were mutated to study the activity of these mutant PCSK9s towards the LDLR.Flow cytometric measurements of the amount of LDL inter-nalized in transiently transfected HepG2cells were used as a measure of the number of LDLRs in these experiments.Of the nine PCSK9mutants that were secreted,all had an activity towards the LDLR similar to that of wild-type PCSK9[23].Thus,thisfinding makes it unlikely that residues of the C-terminal domain are engaged in specific protein e protein interactions with residues of the LDLR or within other proteins.Another characteristics of the C-terminal domain is the large number of histidines which have been proposed to contribute to PCSK9's increased affinity to bind to the LDLR at low pH[11,12]. Changing the pH from pH7.4to pH5.4,makes these histidines obtain a net increased positive charge of8e10units[23].Seven of the14histidines in the C-terminal domain of PCSK9are surface-exposed.To study the role of these histidines on the activity of PCSK9towards the LDLR,Holla et al.[23]transfected HepG2cells with PCSK9plasmids where these seven histidines had been mutated to neutral alanines(7H>A-PCSK9),positively charged ly-sines(7H>K-PCSK9)or negatively charged glutamic acid(7H>E-PCSK9).From studies where the amount of LDL internalized by the transfected cells was used as a measure of the number of LDLRs, they showed that the relative activities of the mutant PCSK9s to-wards the LDLR,was:7H>K-PCSK9>wild-type PCSK9>7H>A-PCSK9>7H>E-PCSK9[23].Thus,the activity of the mutant PCSK9s correlated with the net charge of their C-terminal domains.The more positively charged the C-terminal domain,the more potent was the activity towards the LDLR.Holla et al.[23]also showed that replacement of the C-terminal domain of PCSK9with a similarly sized,but unrelated protein containing seven histidines,made the chimeric PCSK9maintain an activity towards the LDLR similarly to that of wild-type PCSK9.Thesefindings indicate that the role of the C-terminal domain of PCSK9for degradation of the LDLR,is medi-ated by its size and its positive charge at acidic pH,and not through specific protein e protein interactions.The notion that the C-terminal domain of PCSK9interacts with the negatively charged ligand-binding domain of PCSK9,was sup-ported by studies by Tveten et al.[21].They showed that that a mutant PCSK9lacking the C-terminal domain,was released from the LDLR at mildly acidic conditions in a fashion similar to the release of wild-type PCSK9from an LDLR lacking the ligand-binding domain.However,when the mutant PCSK9lacking the C-terminal domain was fused to the unrelated protein containingT.P.Leren/Atherosclerosis237(2014)76e81 78seven histidines mentioned above,the binding of the chimeric PCSK9to the LDLR at acidic pH was similar to that of wild-type PCSK9[21].Together,these data indicate that the absence of the ligand-binding domain of the LDLR or the absence of the C-terminal domain of PCSK9,make PCSK9unable to remain bound to the LDLR at the mildly acidic conditions of the sorting endosome.These findings explain why PCSK9in both instances is unable to reroute the LDLR to intracellular degradation.The notion that the ligand-binding domain of the LDLR must interact with the C-terminal domain of PCSK9for the LDLR to un-dergo intracellular degradation,is consistent with thefindings of Ni et al.[22].They found that an antigen-binding fragment against the C-terminal domain reduced the activity of PCSK9towards the LDLR by50%,whereas mutating the six residues recognized by the antigen-binding fragment,had little effect on the activity of PCSK9 [22].A likely explanation for these observations is that the binding of the antigen-binding fragment prevented the interaction be-tween the C-terminal domain of PCSK9and the ligand-binding domain of the LDLR.Also a monoclonal antibody against Module 1of the C-terminal domain of PCSK9,has been shown to interfere with the activity of PCSK9towards the LDLR,without affecting the interaction between PCSK9and the EGF-A repeat of the LDLR[25]. In a study where one or two of the three repeat modules of the C-terminal domain were deleted,Saavedra et al.[26]showed that Module2of the C-terminal domain was required for PCSK9to be active,and that increased basicity of an exposed loop within this module,increased the activity towards the LDLR.In a situation where individual residues of the ligand-binding domain of the LDLR or the C-terminal domain of PCSK9are not involved in the interaction between the two proteins,it is likely that the negatively charged ligand-binding domain of the LDLR is electrostatically attracted to the C-terminal of PCSK9at acidic pH.It is assumed that this interaction is the mechanism by which an LDLR with bound PCSK9fails to adopt a closed conformation at acidic pH [18](Fig.2).Regarding the role of other elements of the ectodomain of the LDLR for intracellular degradation,Zhang et al.[18]have shown that EGF-B of the EGF precursor homology domain and the O-linked sugar domain are not required for the LDLR to be degraded. However,PCSK9fails to promote degradation of an LDLR lacking the b-propeller of the EGF precursor homology domain.The reason for this is not clear,but a crystallographic study has shown that there is a minor interface between of the prodomain of PCSK9and the b-propeller of the LDLR at neutral pH[16].7.Ectodomain cleavageIn a situation where commonly used mechanisms for intracel-lular degradation of receptors such as ubiquitination have been excluded,Tveten et al.[27]explored the possibility that there could be an LDLR-specific mechanism for rerouting the LDLR with bound PCSK9to intracellular degradation.Such a mechanism could then also be responsible for rerouting Class5mutant LDLRs that have bound LDL,to intracellular degradation.Class5mutant LDLRs are typically caused by mutations in the EGF precursor homology domain and these mutant LDLRs fail to release LDL in the sorting endosome[4].As a consequence,Class5mutant LDLRs fail to obtain a closed conformation at acidic pH[4].Similarly to an LDLR un-dergoing PCSK9-mediated degradation,the cytoplasmic domain is not required for rerouting Class5LDLRs to intracellular degradation [28].It was therefore hypothesized that the failure to adopt a closed conformation in these instances,could make the ectodomain vulnerable for cleavage in the sorting endosome.Such a cleavage would release the ectodomain of the LDLR into the lumen of the sorting endosome where it predominantly would be confined to the vesicular part of the sorting endosome for further transport down the endosomal/lysosomal tract.From studies of regulated intramembrane proteolysis,ectodo-main cleavage which occurs within thefirst30residues of the ectodomain,will activate g-secretase cleavage within the trans-membrane domain[29].Following g-secretase cleavage,the released cytoplasmic domain is rapidly degraded which makes it difficult to detect by methods such as Western blot analysis[29,30]. However,when g-secretase cleavage is inhibited by the g-secretase inhibitor N-(N-(3,5-difluorophenacetyl)-L-alanyl)-S-phenylglycine t-butyl ester(DAPT),these cytoplasmic domains remain membrane-bound and become more stable and easier to detect.The mainfinding was that when purified PCSK9was added to HepG2cells in the presence of DAPT,a17kDa C-terminal fragment of the LDLR was detected by Western blot analysis[27].Because the predicted molecular weight of the72amino acids cytoplasmic and transmembrane domains of the LDLR is only~8kDa,cleavage of the LDLR to generate the17kDa C-terminal fragment must have occurred within the ectodomain.This notion is also supported by the observation that the17kDa C-terminal fragment was only observed when g-secretase cleavage within the transmembrane domain was inhibited by DAPT.Moreover,as the amount of the160kDa mature LDLR decreased with time in the presence of PCSK9,the amount of the17kDa C-terminal fragment increased in a proportionate fashion[27].The 17kDa C-terminal fragment was not generated when PCSK9was added to cells expressing mutant LDLRs that do not undergo PCSK9-mediated degradation.Furthermore,the17kDa C-terminal fragment was not generated in the presence of chemical reagents such as ammonium chloride or the cysteine cathepsin inhibitor E64d,which prevent PCSK9-mediated degradation of the LDLR [27].Thus,the17kDa C-terminal fragment of the LDLR was generated in an acidic intracellular compartment from an LDLR undergoing PCSK9-mediated degradation(Fig.3).After g-secretase cleavage of17kDa C-terminal fragment,the cytoplasmic domain will be released into cytosol.Whether,the released cytoplasmic domain has any signalling functions,remains to be determined.Such an ectodomain cleavage of the160kDa mature LDLR must have generated a complementary143kDa N-terminal fragment. However,no143kDa N-terminal fragment has been detected by Western blot analysis using antibodies against the N-terminal domain of the LDLR[27].A likely explanation for this failure is that the cleaved143kDa ectodomain released into the endosomal lumen,is rapidly acted upon by endosomal proteases.As a conse-quence,the ectodomain may be rapidly degraded or the antigenic epitopes may be destroyed.One factor that could contribute to rapid endosomal proteolytic degradation,is unfolding of the ligand-binding repeats at the low calcium concentration and acidic milieu of the sorting endosome[31,32].However,further studies should be performed to try to identify the143kDa N-terminal fragment within the endosomal lumen.Generation of the17kDa C-terminal fragment could have been caused by a metalloproteinase or a disintegrin and metal-loproteinase(ADAM)which are commonly involved in shedding of ectodomains of transmembrane proteins[33],or by endosomal proteases[34].However,a panel of metalloproteinase inhibitors did not prevent generation of the17kDa C-terminal fragment[27]. Thefinding that the irreversible cysteine cathepsin inhibitor E64d which prevents PCSK9-mediated degradation of the LDLR[15,27], also prevented generation of the17kDa C-terminal fragment,in-dicates that ectodomain cleavage was caused by an endosomal cysteine cathepsin which predominantly act as endopeptidases [34].Even though most cathepsins in the endosomal/lysosomal tract become more active as the pH decreases,many cathepsins areT.P.Leren/Atherosclerosis237(2014)76e8179active at a pH of ~6.1[35].Moreover,it has been shown that rapid proteolysis of internalized receptors may take place in pre-lysosomal compartments of the endosomal/lysosomal tract [36e 39].In contrast to the ectodomain of the LDLR which seems to escape immunological detection early in the endosomal/lysosomal tract,PCSK9is clearly detectable intracellularly.It is therefore assumed that PCSK9is degraded further down the endosomal/lysosomal tract such as in late endosomes or in lysosomes.The 17kDa C-terminal fragment generated during intracellular degradation of an LDLR that has bound PCSK9,was also generated from a Class 5mutant LDLR that has bound LDL [27].Thus,this finding supports the notion that the failure of an LDLR to fold back on itself to adopt a closed conformation at acidic pH [4,18],makes structures of the LDLR close to the membrane accessible to pro-teolytic cleavage.Further support for this model,comes from data (Tveten K &Leren TP,unpublished data)showing that the 17kDa C-terminal fragment is also generated when cells are incubated with polyvalent,polyclonal anti-LDLR antibodies which cause intracel-lular degradation of the LDLR [40].It is assumed that these poly-clonal antibodies which cause aggregation of the LDLRs [40]prevent the LDLR from adopting a closed conformation at acidic pH.A similar need to adopt a pH-induced conformational change at acidic pH,has been shown for the EGF receptor and asialoglyco-protein receptor to be resistant to proteolysis by the cysteine peptidase papain [41].The notion that the cleaved ectodomain escapes detection because it is rapidly degraded,is apparently in con flict with studies where LDLRs subjected to PCSK9-mediated degradation have been found to co-localize with markers of late endosomes and lysosomes by confocal laser-scanning microscopy or subcel-lular fractionation [9,15,42e 44].However,data from such co-localization experiments should be interpreted with caution because markers such as cathepsin D and mannose-6-phosphate receptor which were used as markers for lysosomes and late endosomes,respectively,are not speci fic for these organelles.Rather,after their transport from the trans-Golgi network to the endosomal/lysosomal tract,they are found in various amounts also in organelles early in endosomal/lysosomal tract such as in early endosomes [34,45e 47].The data presented in this review indicate that an LDLR that fails to adopt a closed conformation at endosomal pH,is rapidly cleaved by an endosomal cysteine cathepsin to release the ectodomain into the endosomal lumen.Targeting intracellular cleavage of an LDLR that fails to adopt a closed conformation in the sorting endosome,could therefore be a novel strategy to lower plasma LDL cholesterol levels.However,unless this protease is speci fic for the LDLR,tar-geting the protease would likely cause major side effects.Con flict of interest None.References[1]Herbert PN,Assmann G,Gotto AM,Fredrickson DS.Familial lipoprotein de fi-ciency:abetalipoproteinemia,hypolipoproteinemia and Tangier disease.In:Stanbury JB,Wyngaarden JB,Fredrickson DS,Goldstein JL,Brown MS,editors.The metabolic basis of inherited disease.McGraw-Hill;1983.p.589e 621.[2]Goldstein JL,Hobbs HH,Brown MS.Familial hypercholesterolemia.In:Scriver CR,Beaudet AL,Sly WS,Valle D,editors.The metabolic &molecular basis of inherited disease.McGraw-Hill;2001.p.2863e 914.[3]Südhof TC,Goldstein JL,Brown MS,Russell DW.The LDL receptor gene:amosaic of exons shared with different proteins.Science 1985;228:815e 22.[4]Hobbs HH,Russell DW,Brown MS,Goldstein JL.The LDL receptor locus infamilial hypercholesterolemia:mutational analysis of a membrane protein.Annu Rev Genet 1990;24:133e 70.[5]Yamamoto T,Davis CG,Brown MS,Schneider WJ,Casey ML,Goldstein JL,et al.The human LDL receptor:a cysteine-rich protein with multiple Alu sequences in its mRNA.Cell 1984;39:27e 38.[6]Brown MS,Anderson RG,Goldstein JL.Recycling receptors:the round-tripitinerary of migrant membrane proteins.Cell 1983;32:663e 7.[7]Rudenko G,Henry L,Henderson K,Ichtchenko K,Brown MS,Goldstein JL,et al.Structure of the LDL receptor extracellular domain at endosomal pH.Science 2002;298:2353e 8.[8]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 U S A 2003;100:928e 33.[9]Zhang DW,Lagace TA,Garuti R,Zhao Z,McDonald M,Horton JD,et al.Bindingof proprotein convertase subtilisin/kexin type 9to epidermal growth factor-like repeat A of low density lipoprotein receptor decreases receptor recy-cling and increases degradation.J Biol Chem 2007;282:18602e 12.[10]Mousavi SA,Berge KE,Leren TP.The unique role of proprotein convertasesubtilisin/kexin 9in cholesterol homeostasis.J Intern Med 2009;266:507e 19.Fig.3.Ectodomain cleavage of an LDLR with bound PCSK9in the sorting endosome.An LDLR with bound PCSK9fails to adopt a closed conformation in the sorting endosome and undergoes ectodomain cleavage by a cysteine cathepsin.As a consequence,the 143kDa ectodomain is released into the lumen of the vesicular part of the sorting endosome,which matures down the endosomal/lysosomal tract to become a late endosome for subsequent fusion with a lysosome.The remaining 17kDa C-terminal fragment is subjected to g -secretase cleavage within the transmembrane domain to release the cytoplasmic domain into cytosol.The figure is not drawn to scale.T.P.Leren /Atherosclerosis 237(2014)76e 8180。