Structure-based development of a receptor activatorof nuclear factor-κB ligand (RANKL) inhibitor peptideand molecular basis for osteopetrosisHai Minh Taa, Giang Thi Tuyet Nguyena, Hye Mi Jinb, Jongkeun Choia, Hyejin Parka, Nacksung Kimb,Hye-Yeon Hwanga, and Kyeong Kyu Kima,1

aDepartment of Molecular Cell Biology, Samsung Biomedical Research Institute, Sungkyunkwan University School of Medicine, Suwon 440-746, Korea;and bDepartment of Pharmacology, Chonnam National University Medical School, Gwangju 501-746, Korea

Edited* by Sung-Hou Kim, University of California, Berkeley, CA, and approved September 28, 2010 (received for review August 6, 2010)

The receptor activator of nuclear factor-κB (RANK) and its ligandRANKL, which belong to the tumor necrosis factor (TNF) recep-tor-ligand family, mediate osteoclastogenesis. The crystal structureof the RANKL ectodomain (eRANKL) in complex with the RANKectodomain (eRANK) combined with biochemical assays of RANKmutants indicated that three RANK loops (Loop1, Loop2, andLoop3) bind to the interface of a trimeric eRANKL. Loop3 isparticularly notable in that it is structurally distinctive from otherTNF-family receptors and forms extensive contacts with RANKL.The disulfide bond (C125-C127) at the tip of Loop3 is importantfor determining the unique topology of Loop3, and dockingE126 close to RANKL, which was supported by the inability ofC127A or E126A mutants of RANK to bind to RANKL. Inhibitoryactivity of RANK mutants, which contain loops of osteoprotegerin(OPG), a soluble decoy receptor to RANKL, confirmed that OPGshares the similar binding mode with RANK and OPG. Loop3 playsa key role in RANKL binding. Peptide inhibitors designed to mimicLoop3 blocked the RANKL-induced differentiation of osteoclastprecursors, suggesting that they could be developed as therapeuticagents for the treatment of osteoporosis and bone-relateddiseases. Furthermore, some of the RANK mutations associatedwith autosomal recessive osteopetrosis (ARO) resulted in reducedRANKL-binding activity and failure to induce osteoclastogenesis.These results, together with structural interpretation of eRANK-eRANKL interaction, provided molecular understanding for patho-genesis of ARO.

Bone is a dynamic organ that is maintained by a balancebetween bone resorption by osteoclasts and bone formation

by osteoblasts. The interaction between receptor activator ofnuclear factor-κB ligand (RANKL) on osteoblast/stromal cellsand the RANK receptor on osteoclast precursors results in thematuration of osteoclasts and subsequent bone resorption (1–4).Osteoprotegerin (OPG) functions as a soluble decoy receptor toRANKL and competes with RANK for RANKL binding. Accord-ingly, OPG has been shown to be an effective inhibitor of matura-tion and activation of osteoclasts in vitro and in vivo (5, 6). Theratio between RANKL and OPG elegantly regulates the orienta-tion of bone metabolism to either bone formation or resorption;therefore, dysregulation of this ratio causes an imbalance be-tween bone formation and resorption and results in bone diseasessuch as osteoporosis, rheumatoid arthritis, and osteolytic bonemetastasis (7–10). For the same reasons, mutations in RANK,OPG, or RANKL are associated with genetic skeletal abnormal-ities such as autosomal recessive osteopetrosis (ARO) (11, 12).Because of the critical roles of RANKL/OPG/RANK proteins inbone metabolism, their interaction and RANK signaling are con-sidered promising targets for the control of bone metabolic dis-eases (7). Consequently, RANK-Fc, Fc-OPG, and anti-RANKLantibodies have been developed as therapeutics for osteoporosis(13–19). Alternatively, peptide mimics of OPG (OP3-4 peptide)(20, 21) and the tumor necrosis factor (TNF) receptor (WP9QY

peptide) (22) were also developed and showed inhibitory activityagainst the RANKL-induced osteoclastogenesis.

The RANKL-RANK complex belongs to the TNF ligand–receptor superfamily, whose members share a similar bindingmode despite low sequence homology: The receptors bind to agroove at the junction of monomers in the trimeric ligand that isformed by edge-to-face packing of monomeric subunits (23–27).However, the key structural features in the binding interface thatcontrol the biological specificity of a particular ligand–receptorpair have not been defined. For example, the binding modebetween RANKL and RANK is not yet clearly understood,although the crystal structure of RANKL was extensively charac-terized (28, 29).

We sought to identify structural determinants that govern thespecific ligand–receptor recognition of RANKL-RANK and,thus, to provide a molecular foundation for further investigationof bone-related diseases and development of previously unde-scribed pharmaceuticals. In this study, based on crystal structureof the ectodomain of mouse RANKL (eRANKL) complexed withthe ectodomain of RANK (eRANK) at 2.5-Å resolution and thebiochemical and functional characterization of eRANK mutants,we identified the key structural determinants governing therecognition specificity of eRANK and developed potential inhi-bitors of RANK-RANKL interaction through structure-basedapproaches. Furthermore we were able to explain the molecularbasis for mutations associated with ARO.

ResultsOverall Structure of the eRANK-eRANKL Complex. The complex, withapproximate dimensions of 60 Å × 70 Å × 100 Å, contains threeeRANK molecules with four full cysteine-rich domains (CRDs)inserted into three crevices on the subunit interfaces of a trimericeRANKL (Fig. 1A and Fig. S1A). Binding of eRANK induceslocal conformational changes in eRANKL near the AA″, CD,and EF loops, thereby disordering the N terminus of strand D,and the C terminus of strand E (Fig. S1B). eRANK contains fourCRDs (Fig. 1B and Figs. S2 and S3) and shows some structuralfeatures distinct from other canonical receptors of the TNFfamily (23–27). Each CRD typically has six conserved Cys resi-dues that form three disulfide pairs, but the disulfide bondbetween the third and fifth Cys residues is missing in CRD2,

Author contributions: H.M.T. and K.K.K. designed research; H.M.T., G.T.T.N., H.M.J., J.C.,and H.P. performed research; H.M.T., N.K., H.-Y.H., and K.K.K. analyzed data; andH.M.T., N.K., H.-Y.H., and K.K.K. wrote the paper.

The authors declare no conflict of interest.

*This Direct Submission article had a prearranged editor.

Data deposition: The atomic coordinates and structure factors have been deposited in theProtein Data Bank, www.pdb.org (PDB ID code 3NZY).1To whom correspondence should be addressed. E-mail: kkim@med.skku.ac.kr.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1011686107/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1011686107 PNAS ∣ November 23, 2010 ∣ vol. 107 ∣ no. 47 ∣ 20281–20286

BIOCH

EMISTR

Y

Dow

nloa

ded

by g

uest

on

June

9, 2

020

Dow

nloa

ded

by g

uest

on

June

9, 2

020

Dow

nloa

ded

by g

uest

on

June

9, 2

020

CRD3, and CRD4 of RANK (Fig. S3). Because disulfide bondsare essential for the overall fold of CRDs, the absence of a dis-ulfide bond is likely to cause conformational diversity in eRANK.In fact, the structure of CRD3 of eRANK was the most divergentwhen compared with other TNF receptors (Fig. S4). However,CRD2 shows little structural heterogeneity (Fig. S4), probablybecause the hydrogen bonds of His90-Asn106, His90-Arg111,and Lys91-Arg111 functionally replace the missing disulfide bond(Fig. S3). Another feature of CRD3 is the presence of a nonca-nonical disulfide bond, Cys125-Cys127 (Fig. S3), which appearsto play a role in RANKL binding and recognition, whereas otherdisulfide bonds mainly stabilize the overall fold.

Receptor-Ligand Interface. In the eRANK-eRANKL complex,Loop1 and Loop2 in CRD2 and Loop3 in CRD3 fit into pocketsof RANKL with perfect geometric and electrostatic complemen-tarity, accounting for the recognition specificity (Fig. 1C andFig. S5A). And 7680 Å2 of the solvent-accessible area that is15.4% of the total surface area of uncomplexed molecules isburied in the complex. The wide electrostatic network formedby charged residues appeared to be pivotal for specific interaction(Fig. S5A). The detailed atomic interactions between each loopand RANKL are described in Table S1 and Fig. S5.

Among the three RANKL-binding loops, Loop3 consistingof residues 119–130, provides the largest binding surface area(Fig. 1D). Glu126 in Loop3 is particularly notable due to its ex-tensive interaction with RANKL. Glu126 not only interacts withLys180 of RANKL, but forms bidentate water-mediated hydro-gen bonds with Lys256, Asp301, and Asp303 on one side, andAsn253 and Gln291 on the other. It is also in van der Waalscontact with Tyr240. It is unique that RANK has a noncanonicaldisulfide bridge (Cys125-Cys127) at the tip of Loop3, which

most likely determines the shape of the loop and helps to orientGlu126 for extensive interactions with RANKL (Fig. 1D).The C127A replacement abolished eRANKL binding activity(Fig. S6), indicating the importance of this noncanonical disulfidebond for RANKL binding.

To confirm the role of each RANKL-binding loop, keyresidues that are involved in RANKL binding through ionic andhydrogen bonds (D85 in Loop1, K97 in Loop2, and E126 inLoop3) were replaced with Ala. Isothermal titration calorimetry(ITC) showed that the wild-type eRANK has strong affinity foreRANKL with Kd ¼ 230 nM; however, no interaction wasobserved for E126A eRANK (Fig. 2A). E126A eRANK wasnot able to suppress eRANKL-induced osteoclast differentiation,whereas wild-type eRANK completely inhibited osteoclastogen-esis (Fig. 2B). Loop2 contributes only 15% of the total buriedsurface area, but appears to exert a substantial effect on RANKLbinding, because the K97A mutation completely abrogatedeRANKL binding (Fig. 2A) and osteoclastogenesis inhibition(Fig. 2B). D85, a key residue in Loop1, did not appear to be asimportant as K97 in Loop2 or E126 in Loop3, because D85AeRANK retained marginal binding affinity (Kd ¼ 24.2 μM) andinhibitory effects (Fig. 2).

Binding Mode of OPG. OPG is known as a decoy receptor toRANKL, and its binding affinity for RANKL is comparable tothat of RANK (5, 6). Therefore, we hypothesized that OPGand RANK might have similar binding modes and modeledthe OPG-RANKL complex (SI Text) to verify this hypothesis.In the modeling, OPG Loop3 was shown to fit into RANKL pock-et (Fig. S7), and it is possible that E116 in Loop3 substitutes forE126 of RANK (Fig. 3A). We also created chimeric RANKs,which contain OPG loops (Fig. S2). The substitution of Loop3

Fig. 1. Overall structure of eRANK-eRANKL com-plex. (A) Ribbondiagramof eRANK-eRANKL complex.Three eRANKs (pale cyan, pale green, and light blue)bind to a trimeric eRANKL (light orange, green cyan,and lime). The residues involved in ligand–receptorinteraction, depicted as ball-and-stickmodels, are red(Loop1), green (Loop2), and blue (Loop3). N andC ter-mini and β-strands of eRANKL are labeled. (B) Overallstructure of eRANK in ribbon diagram. The four CRDsare pink, forest, marine, andmagenta, and conserveddisulfide bonds are in yellow ball-and-stick models.The disulfide bond between Cys125 and Cys127 isin a red ball-and-stick model. (C) Surface presenta-tions of the binding interfaces of eRANK monomer(Left) and eRANKL dimer (Right). The eRANK residuesin binding loops are red (Loop1), green, (Loop2), andblue (Loop3). Counterpart residues in RANKL are inthe same color scheme. (D) Stereo presentation ofthe binding mode of Loop3 to RANKL. eRANKL andeRANK are green and pink, respectively. The residuesinvolved in ligand–receptor binding are drawn asball-and-stick models and labeled. Blue dashed linesrepresent hydrogen or ionic bonds, and importantstrands or loops are labeled. Cys125-Cys127 andwaters are depicted as yellow and red ball-and-stickmodels, respectively.

20282 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1011686107 Ta et al.

Dow

nloa

ded

by g

uest

on

June

9, 2

020

(eRANK-Loop3OPG) or all three loops (eRANK-Loop123OPG)did not result in substantial changes in the inhibitory activityof eRANK (Fig. 3B), suggesting that binding loops of RANKand OPG may function in similar modes, and be functionallyexchangeable.

Peptide Inhibitors Derived from RANK Loops. The crystal structureof the eRANK-eRANKL complex and the biochemical character-ization of RANK mutants proposed that Loop2 and Loop3are the main determinants of the receptor–ligand interaction.To confirm this notion, synthetic peptides harboring the keyresidues in Loop2 and Loop3 were designed (Table 1), and theirinhibitory effects on the eRANKL-induced osteoclastogenesiswere examined. A Cys or Tyr residue was introduced for cycliza-tion, so that the peptides have overall shapes similar to the bind-ing loops in the complex. For example, Val100, which was locatedmost proximal to Cys93 in the crystal structure of the complex,was replaced with Cys in order for the formation of a disulfidelinkage with Cys93 (L2-1 and L2-2). Similarly, Trp121 (L3-2)and His120 (L3-3) were replaced with Cys, in consideration oftheir proximity to Cys128. Ala95 was changed to Gly in L2-2to endow conformational flexibility (Table 1). L3-1 was expectedto maintain the β-loop because it contains all residues of Loop3,so was not cyclized. However, Cys128 in L3-1 was changed to Ser

to prevent undesired disulfide linkage with Cys125 or Cys127(Table 1). Of the peptide mimics, L3-3 showed dose-dependentinhibitory activity, whereas the effects of L2-1 and L2-2 did notappear to be significant, implying that Loop2 is not as importantas Loop3 (Fig. 4A) or that the folding of Loop2-mimic peptideswas not similar to the native structure of Loop2. L3-1 and L3-2might have lost the native folding necessary for effective RANKLbinding, because of a loose or distorted conformation. In con-trast, the restraint imposed on L3-3 may stabilize the conforma-tion so that it is similar to Loop3 in the eRANK-eRANKLcomplex.

The inhibitory activity of L3-3 was characterized by the celldifferentiation assay in a wide concentration range. OP3-4, a well-known OPG-mimic inhibitor of the osteoclastogenesis (20, 21),served as a positive control. L3-3 effectively inhibited theeRANKL-induced osteoclastogenesis, even more effectivelythan OP3-4 at high concentrations (Fig. 4B). This was furtherconfirmed by ITC analysis of L3-3 and OP3-4 (Fig. S6B), whichshowed that L3-3 has 16-fold higher binding affinity to eRANKL(Kd ¼ 17.3 μM) than OP3-4 (Kd ¼ 284.4 μM). This result de-monstrated that L3-3 is a potent inhibitor of RANKL and couldbe developed as a potential therapeutic agent for the treatmentof bone-related diseases.

Fig. 2. Functional significance of residues involved in RANKL binding. (A) ITC analyses of eRANKL binding to the wild-type (eRANK-WT) and mutant eRANKs(D85A, K97A, and E126A). The raw data (Upper) and the best fit (Lower) are shown, with the dissociation constants (Kd ). (B) Inhibitory effects of wild-type andmutant eRANKs on the eRANKL-induced osteoclast differentiation were estimated by the number of TRAP-positive multinucleated cells. Concentrations ofeRANKs are indicated.

Ta et al. PNAS ∣ November 23, 2010 ∣ vol. 107 ∣ no. 47 ∣ 20283

BIOCH

EMISTR

Y

Dow

nloa

ded

by g

uest

on

June

9, 2

020

Molecular Basis for the Pathogenesis of ARO. The proposed bindingmodes of eRANK and eRANKL provided a molecular basis forthe pathogenesis of ARO, a human inherited bone disease inwhich the differentiation of osteoclasts is impaired by mutationsin RANKL or RANK (11, 12). Three mutations of RANKL areassociated with ARO: the deletion of β-strand A and half of theAA″ loop; the substitution mutation M199K; and the frameshiftcausing the deletion of β-strands G, H, and F (11). Deletion of theAA″ loop could abolish the interaction with RANK, because thisloop forms an extensive interaction network with Loop1 andLoop3 of eRANK (Fig. 1D, Fig. S5B, and Table S1). The frame-shift mutation and the resultant loss of strand F, which is impor-tant for the trimerization of RANKL, could significantly affectthe conformation and binding activity. Of seven mutationsidentified in human RANK, four (G53R, R129C, R170G, andC175R) are located in the extracellular domain (12). In the

mouse eRANK, Arg130, the equivalent of Arg129 of humanRANK, forms ionic and hydrogen bonds with Glu225 andAsn266 of RANKL (Table S1). Therefore, the R129C substitu-tion is expected to impair the affinity for RANKL and, accord-ingly, the ability to induce the differentiation of osteoclasts.

On the other hand, M199 of human RANKL, and G53, R170,and C175 of human RANK, which correspond to mouse M198,G54, K171, and C176, do not seem to directly affect the forma-tion of the complex, because they are not on the RANK-RANKLinterface. To understand how these mutations cause ARO,recombinant proteins were expressed in a bacterial expressionsystem and characterized. The M198K mutant was completely in-soluble under the expression conditions of wild-type eRANKL,suggesting that this mutation might result in instability or misfold-ing of the protein, although we cannot rule out the possibility ofthe soluble expression in animal cells. In the osteoclastogenicanalysis, the G54R, K171G, and C176R mutants all showedsignificant decreases in inhibitory activity (Fig. 4C). Consistentwith this observation, eRANKL binding was not detected forG54R and K171G by ITC under current experimental conditions(Fig. S6A). These results suggest that mutant proteins may stillretain very weak binding affinity to RANK and thus affect in vitroosteoclastogeneis only at the higher concentration. The low so-lubility of the C176R mutant prevented ITC measurement, butconsidering the close correlation between the binding affinityand osteoclastogenesis, C176R seems to have reduced RANKL-binding activity. Conformational changes caused by these muta-tions were proposed to diminish the RANKL-binding affinity,consequently impairing bone resorption. This was confirmed

Table 1. Peptide design for the development of RANKL inhibitors

The original loop sequences Peptide sequences

Loop2 93- CDAGKALV-100 L2-1 YC DAGKAL CYL2-2 YC DGGKAL CY

Loop3 118-GYHWNSDCECCRRN-131 L3-1 GYHWNSDCECSRRNL3-2 YC NSDCEC CY RRL3-3 YC WNSDCEC CY RR

OPG 113-LEIEFCLKHR-122 OP3-4 YC EIEF CY LIR

Residues involved in RANKL interaction are highlighted and underlined.Cys and Tyr residues introduced for cyclization of peptides are alsohighlighted.

Fig. 3. Themodeling of OPG and inhibitory activities of chimeric RANKs. (A) Ribbon diagrams of the ligand-binding domains (CRD2 and½ CRD3) of RANK, andthe enlarged view of Loop3 (yellow) superposed onto DR5 (magenta) and OPG (orange). Cys125-Cys127 in RANK, and Glu residues in RANK (Glu126), DR5(Glu151), and OPG (Glu116) are shown in ball-and-stick models and labeled. (B) Functional analyses of wild-type (eRANK-wt) and chimeric eRANKs (Loop3OPG

and Loop123OPG). Their inhibitory effects on the eRANKL-induced osteoclast differentiation weremonitored by countingmultinucleated TRAP-expressing cells.Concentrations of eRANKs are indicated.

20284 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1011686107 Ta et al.

Dow

nloa

ded

by g

uest

on

June

9, 2

020

by CD analyses of G54R and K171G, which showed that theirspectra did not overlap with that of wild type and indicated thatthe secondary structures were altered (Fig. S6C). Also, the C176Rsubstitution is likely to change the conformation of RANK sub-stantially, because Cys176 is a key residue to determine the fold-ing of CRD4 by the formation of the disulfide bond.

DiscussionRANK has the overall fold common to all TNF-family receptors(Fig. S4), but it also has distinctive conformational features andligand-binding mode that confer the binding specificity forRANKL. The most striking difference between RANK and otherTNF-family receptors is the presence of an additional disulfidebond (Cys125-Cys127) in Loop3 that restrains the conformationand determines its unique topology (Figs. 1 and 3A). In eDR5,the loop corresponding to the RANK Loop3 has a loose confor-mation because the conformational restraint is alleviated(Fig. S7). RANKL has a smaller and deeper binding pocket,whereas that of TNF-related apoptosis-inducing ligand (TRAIL)is larger and shallower (Fig. S7). The overlapping of eRANKonto the eDR5-eTRAIL complex, or eDR5 onto the eRANK-eRANKL complex, shows severe steric clashes in the Loop3 bind-ing region. Therefore, it is thought that the structural heteroge-

neity in Loop3 is key to the recognition specificity for eRANKLor eTRAIL, even though the central ligand-binding domains ofRANK and DR5 are very well conserved in structural aspects(Fig. 3A and Fig. S7).

By analogy, the structural comparison of Loop3 explains thedual specificity of OPG for both RANKL and TRAIL. OPGLoop3 is two and four residues shorter than those in RANKand DR5, respectively (Fig. S2). Therefore, it has more compactand shorter shape. Furthermore, E116, located on the tip ofOPG Loop3, overlaps well with E126 in RANK and is proximalto E151 in DR5. Considering the structural flexibility of the loop,it is expected that OPG Loop3 can fit into pockets of bothRANKL and TRAIL without steric hindrance, and the major in-teractions between Loop3 and the binding pocket are preserved.

Loop1 of RANK seems to be less important than those in otherTNF-family receptors. In the TNFR1-TNFβ or DR5-TRAIL com-plex, the hydrophobic interaction between Loop1 of the receptorand the DE loop of the ligand is the main force for binding, withthe conserved tyrosine residue in the DE loop penetrating thehydrophobic groove formed by hydrophobic residues in Loop1(23–25). However, Ile248 of eRANKL, which is equivalent tothe Tyr residue in TNFR or DR5, is thought to be only marginalfor receptor binding, because the I248D mutant maintainedbinding activity to RANK (28). This was corroborated in thestructures presented here, because the electron densities of theIle248 residues in the DE loop (residues 245–250) were very poorin both free (29) and ligand-bound forms (current study).

Recombinant proteins that inhibit RANK-RANKL inter-action, such as RANK-Fc, Fc-OPG, and anti-RANKL antibodies,have been developed as therapeutic agents for osteoporosis(13–19). Denosumab, a humanized monoclonal antibody againstRANKL, is currently in phase III trials for osteoporosis treatment(16), and in phase II trials for the treatment of rheumatoid ar-thritis and osteolytic bone metastases (18, 19). AMGN-0007, arecombinant Fc-OPG, is in phase I clinical trials for the treatmentof osteolytic bone metastasis (14, 15). However, the use of largemacromolecules for therapeutic intervention can be hindered bydrawbacks including low stability, poor bioavailability, high cost,and difficulties in administration. Therefore, small peptides orpeptidomimetics with high specificity to RANKL can be an alter-native to overcome disadvantages of using macromolecules. Forthis purpose, small peptides, WP9QYand OP3-4, were developedand their inhibitory activity for bone resorption was confirmed invivo (20–22). However, further improvement of their efficacy isnecessary for the application for therapeutic purposes.

The structural interpretation and functional studies ofeRANK mutants indicated that the disulfide-restrained Loop3is the major determinant for specific RANK-RANKL recogni-tion. This enabled us to design previously undescribed inhibitorpeptides that mimic the structure and interaction of Loop3. Thecyclic peptide L3-3 strongly bound to eRANKL and blockedRANKL-induced osteoclast differentiation more efficiently thanOP3-4 under our experimental conditions (Fig. 4B and Fig. S6B).An OPG-RANKL model and cell differentiation assay usingchimeric RANKs containing OPG loops suggested similarity inthe binding mode of RANK and OPG. Accordingly, it was ex-pected that Loop3 in OPG is also involved in RANKL bindingin the analogous way, which was also supported by the fact thatOP3-4, a peptide comprising essential binding residues of Loop3,was proven as a potent RANKL inhibitor in vitro and in animalmodels (20, 21). The enhanced inhibitory activity of L3-3 mightbe due to additional binding residues (Table 1). Further modifi-cation of L3-3 in parallel comparison with OP3-4 will allow thedevelopment of more potent peptides or peptide-derived chemi-cal agents that can be applied for therapeutic purposes. In thisaspect, current study provides a framework for rational designof RANKL inhibitors.

Fig. 4. Inhibitory activities of synthetic peptides and ARO-associated eRANKmutants. (A) Osteoclast precursors were treated with 10 ng∕ml eRANKL, andvarying concentrations (10 to 50 μM) of peptides (L2-1, L2-2, L3-1, L3-2, andL3-3), and their differentiation was monitored by TRAP assay. (B) Inhibitoryactivities of L3-3 and OP3-4 were compared by TRAP assay under the sameconditions as in Fig. 4A over a wide concentration range (1 to 200 μM).(C) Concentration-dependent inhibitory activities of wild-type and mutanteRANKs (G54R, K171G, and C176R) were compared by TRAP assay. eRANKLat the concentration of 100 ng∕mL was used for cell differentiation.

Ta et al. PNAS ∣ November 23, 2010 ∣ vol. 107 ∣ no. 47 ∣ 20285

BIOCH

EMISTR

Y

Dow

nloa

ded

by g

uest

on

June

9, 2

020

Structure analysis of RANK-RANKL combined with mutantstudies enabled us to explain the molecular basis for the inheritedbone disease, ARO. The crystal structure revealed that some ofthe mutations found in ARO patients are on the RANK-RANKLinterface and are likely to directly interfere with the binding.Current study proved that some ARO-associated mutations ofRANK cause defective folding and affect the binding to RANKL,although they are not on the binding interface.

Materials and MethodsStructure Determination. The crystal structure of the eRANK-eRANKL complexwas determined by molecular replacement using the crystal structure ofRANKL as a search probe (PDB ID code 1IQA; SI Text). The final model refinedat 2.5-Å resolution with R and Rfree values of 22.6% and 24.9%, respectively,comprises residues 35–198 of eRANK and residues 161–315 of eRANKL. Poly-Ala peptides were modeled at residues 35–46 in eRANK and residues 245–250in eRANKL due to weak electron densities in these regions. The refinementstatistics are summarized in Table S2.

Site-Directed and Loop Replacement Mutagenesis. Point mutations of eRANKwere introduced using a Quick Change mutagenesis kit (Stratagene).Chimeric RANK constructs were created using PCR in which Loop1 (residues73–89), Loop2 (residues 94–100), or Loop3 (residues 119–130) was replaced bycorresponding regions of OPG, Loop1 (residues 67–82), Loop2 (residues88–94), and Loop3 (residues 111–120). Mutant and chimeric proteins were

expressed and purified as described for the wild-type protein. The elutionprofile of Superdex 200 gel filtration chromatography and the size distribu-tion profile determined by dynamic light scattering were used to verify prop-er folding and assembly of mutant proteins in comparison to the wild type.

Osteoclast Differentiation Assay. The biological activities of eRANKL, eRANK,and inhibitory peptides were estimated using an osteoclast differentiationassay. Mouse osteoclast precursors were extracted and purified from thetibia and femur of 5- to 8-wk-old male mice. Osteoclast precursors werecultured in 96-well plates (2 × 104 cells∕well) in α-MEM supplemented with10% FBS and 30 ng∕mL macrophage colony stimulating factor. Varying con-centrations of eRANKL and eRANK were added, and the cells were incubatedat 37 °C in a humidified atmosphere containing 5% CO2. Differentiation ofosteoclast precursors was measured in vitro by quantitation of tartrate-resis-tant acid phosphatase (TRAP) activity. TRAP staining assay was performedusing Naphthol AS phosphate substrate in 100 mM sodium acetate (pH5.2), and the reaction product was quantified by spectroscopic measurementat 405 nm. TRAP-positive multinucleated cells were counted under the micro-scope, and each experiment was repeated three times for statistical analysis.

ACKNOWLEDGMENTS. This work was supported by the 21C Frontier FunctionalProteomics Program (FPR08B2-270), Korea Healthcare Technology R&DProject (A092006), Ubiquitome Research Program (M105 33010001-05N3301-00100), National Research Laboratory Program (NRL-2006-02287), and KoreaResearch Foundation Grant (KRF-2008-220-C00040).

1. Lacey DL, et al. (1998) Osteoprotegerin ligand is a cytokine that regulates osteoclastdifferentiation and activation. Cell 93:165–176.

2. Yasuda H, et al. (1998) Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc Natl AcadSci USA 95:3597–3602.

3. Boyle WJ, Simonet WS, Lacey DL (2003) Osteoclast differentiation and activation.Nature 423:337–342.

4. Teitelbaum SL, Ross FP (2003) Genetic regulation of osteoclast development andfunction. Nat Rev Genet 4:638–649.

5. Simonet WS, et al. (1997) Osteoprotegerin: A novel secreted protein involved in theregulation of bone density. Cell 89:309–319.

6. Yasuda H, et al. (1998) Identity of osteoclastogenesis inhibitory factor (OCIF) andosteoprotegerin (OPG): A mechanism by which OPG/OCIF inhibits osteoclastogenesisin vitro. Endocrinology 139:1329–1337.

7. Tanaka S, Nakamura K, Takahasi N, Suda T (2005) Role of RANKL in physiological andpathological bone resorption and therapeutics targeting the RANKL-RANK signalingsystem. Immunol Rev 208:30–49.

8. Vega D, Maalouf NM, Sakhaee K (2007) CLINICAL Review #: The role of receptoractivator of nuclear factor-kappaB (RANK)/RANK ligand/osteoprotegerin: clinicalimplications. J Clin Endocrinol Metab 92:4514–21.

9. Theoleyre S, et al. (2004) The molecular triad OPG/RANK/RANKL: Involvement inthe orchestration of pathophysiological bone remodeling. Cytokine Growth F R15:457–475.

10. Wright HL, McCarthy HS, Middleton J, Marshall MJ (2009) RANK, RANKL andosteoprotegerin in bone biology and disease. Curr Rev Musculoskelet Med 2:56–64.

11. Sobacchi C, et al. (2007) Osteoclast-poor human osteopetrosis due to mutations in thegene encoding RANKL. Nat Genet 39:960–962.

12. Guerrini MM, et al. (2008) Human osteoclast-poor osteopetrosis with hypogammaglo-bulinemia due to TNFRSF11A (RANK) mutations. Am J Hum Genet 83:64–76.

13. Feeley BT, et al. (2006) Mixed metastatic lung cancer lesions in bone are inhibited bynoggin overexpression and Rank:Fc administration. J Bone Miner Res 21:1571–1580.

14. Body JJ, et al. (2003) A phase I study of AMGN-0007, a recombinant osteoprotegerinconstruct, in patients with multiple myeloma or breast carcinoma related bonemetastases. Cancer 97:887–892.

15. Kostenuik PJ (2005) Osteoprotegerin and RANKL regulate bone resorption, density,geometry and strength. Curr Opin Pharmacol 5:618–625.

16. Cummings SR, et al. (2009) Denosumab for prevention of fractures in postmenopausalwomen with osteoporosis. New Engl J Med 361:756–765.

17. Roman-Gonzalez A, Ackerman KE (2009) Pharmacotherapy of bone loss in postmeno-pausal women: Focus on denosumab. Clin Med Ther 1:1131–1143.

18. Lipton A, et al. (2007) Randomized active-controlled phase II study of denosumabefficacy and safety in patients with breast cancer related bonemetastases. J Clin Oncol25:4431–4437.

19. Cohen SB, et al. (2008) Denosumab treatment effects on structural damage,bone mineral density, and bone turnover in rheumatoid arthritis: A twelve-month,multicenter, randomized, double-blind, placebo-controlled, phase II clinical trial.Arthritis Rheum 58:1299–1309.

20. Cheng X, et al. (2004) Disabling of receptor activator of nuclear factor-kappaB (RANK)receptor complex by novel osteoprotegerin-like peptidomimetics restores bone loss invivo. J Biol Chem 279:8269–8277.

21. Heath DJ, et al. (2007) An osteoprotegerin-like peptidomimetic inhibits osteoclasticbone resorption and osteolytic bone disease in myeloma. Cancer Res 67:202–208.

22. Aoki K, et al. (2006) A TNF receptor loop peptide mimic blocks RANK ligand-inducedsignaling, bone resorption, and bone loss. J Clin Invest 116:1525–1534.

23. Banner DW, et al. (1993) Crystal structure of the soluble human 55 kd TNF receptor-human TNFβ complex: Implications for TNF receptor activation. Cell 73:431–445.

24. Mongkolsapaya J, et al. (1999) Structure of the TRAIL-DR5 complex reveals mechan-isms conferring specificity in apoptotic initiation. Nat Struct Biol 6:1048–1053.

25. Hymowitz SG, et al. (1999) Triggering cell death: The crystal structure of Apo2L/TRAILin a complex with death receptor 5. Mol Cell 4:563–71.

26. Cha SS, et al. (2000) Crystal structure of TRAIL-DR5 complex identifies a criticalrole of the unique frame insertion in conferring recognition specificity. J Biol Chem275:31171–31177.

27. Zhang G (2004) Tumor necrosis factor family ligand-receptor binding. Curr Opin StructBiol 14:154–160.

28. Lam J, Nelson CA, Ross FP, Teitelbaum SL, Fremont DH (2001) Crystal structure ofthe TRANCE/RANKL cytokine reveals determinants of receptor-ligand specificity. J ClinInvest 108:971–979.

29. Ito S, et al. (2002) Crystal structure of the extracellular domain of mouse RANK ligandat 2.2-A resolution. J Biol Chem 277:6631–6636.

20286 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1011686107 Ta et al.

Dow

nloa

ded

by g

uest

on

June

9, 2

020

Correction and Retraction

CORRECTION

BIOCHEMISTRYCorrection for “Structure-based development of a receptor ac-tivator of nuclear factor-κB ligand (RANKL) inhibitor peptideand molecular basis for osteopetrosis,” by Hai Minh Ta, GiangThi Tuyet Nguyen, Hye Mi Jin, Jongkeun Choi, Hyejin Park,Nacksung Kim, Hye-Yeon Hwang, and Kyeong Kyu Kim, whichappeared in issue 47, November 23, 2010, of Proc Natl Acad SciUSA (107:20281–20286; first published November 8, 2010;10.1073/pnas.1011686107).The authors note that an incorrect atomic coordinate file

(corresponding to an intermediate refinement result) has nowbeen replaced with a corrected file in the Protein Data Bank(PDB ID code 3QBQ).Additionally, the authors omitted a reference to an article by

Liu et al. The complete reference appears below. These errorsdo not affect the conclusions of the article.

Liu C, et al. (2010) Structural and functional insights of RANKL–RANK interaction andsignaling. J Immunol 184:6910–6919.

www.pnas.org/cgi/doi/10.1073/pnas.110660910810022

RETRACTION

GENETICSRetraction for “A prion of yeast metacaspase homolog (Mca1p)detected by a genetic screen,” by Julie Nemecek, Toru Nakayashikiand Reed B. Wickner, which appeared in issue 6, February 10,2009, of Proc Natl Acad Sci USA (106:1892–1896; first publishedJanuary 27, 2009; 10.1073/pnas.0812470106).The authors wish to note the following: “In our efforts to

extend our published paper, we now find that we cannot repro-duce certain results reported there. Specifically, all transfor-mants of strain 4827 carrying p20MCA (lacking p1116) areAde+, making selection of prion-containing cells impossible.Also, strains denoted mca1Δ have a normal MCA1 gene. Otherefforts by R.B.W. and T.N. to demonstrate a prion of Mca1phave also been unsuccessful. We therefore retract the paper.”

Julie NemecekToru NakayashikiReed B. Wickner

www.pnas.org/cgi/doi/10.1073/pnas.1107490108

10022 | PNAS | June 14, 2011 | vol. 108 | no. 24 www.pnas.org