IL130491A - Dna encoding rank polypeptide, the rank polypeptide and use thereof - Google Patents

Description

130491/2 130491 ?Φ I 453545 τηκ ia Ο ¾ΜΟ>Υ>™ RANK νυύ 'ηα ,RANK ooa»i ai> i nptirt N"ST DNA encoding RANK polypeptide, the RANK polypeptide and uses thereof Immunex Corporation C.118466-2 TITLE .

RECEPTOR ACTIVATOR OF NF-KAPPA B, RECEPTOR IS MEMBER OF TNF RECEPTOR SUPERFAMILY TECHNICAL FIELD OF THE INVENTION The present invention relates generally to the field of cytokine receptors, and more specifically to cytokine receptor/ligand pairs having immunoregulatory activity.

BACKGROUND OF THE INVENTION Efficient functioning of the immune system requires a fine balance between cell proliferation and differentiation and cell death, to ensure that the immune system is capable of reacting to foreign, but not self antigens. Integral to the process of regulating the immune and inflammatory resporise~are various members of the Tumor Necrosis Factor (TNF) Receptor/Nerve Growth Factor Receptor superfamily (Smith et al., Science 248: 1019; 1990). This family of receptors includes two different TNF receptors (Type I and Type Π; Smith et al., supra; and Schall et al., Cell 61:361, 1990), nerve growth factor receptor (Johnson et al., Cell 47:545, 1986), B cell antigen CD40 (Stamenkovic et al., EMBO J. 8: 1403, 1989), CD27 (Camerini et al., J. Immunol. 147:3165, 1991 ), CD30 (Durkop et al., Cell 68:421, 1992), T cell antigen OX40 (Mallett et al., EMBO J. 9: 1063, 1990), human Fas antigen (Itoh et al., Cell 66:233, 1991), murine 4-1BB receptor (Kwon et al., Proc. Natl. Acad. Sci. USA 86: 1963, 1989) and a receptor referred to as Apoptosis-Inducing Receptor (AIR; USSN 08/720,864, filed October 4, 1996).

CD40 is a receptor present on B lymphocytes, epithelial cells and some carcinoma cell lines that interacts with a ligand found on activated T cells, CD40L (USSN 08/249,189, filed May 24, 1994). The interaction of this ligand/receptor pair is essential for both the cellular and humoral immune response. Signal transduction via CD40 is mediated through the association of the cytoplasmic domain of this molecule with members of the TNF receptor-associated factors (TRAFs; Baker and Reddy, Oncogene 12: 1, 1996). It has recentiy been found that mice that are defective in TRAF3 expression due to a targeted disruption in the gene encoding TRAF3 appear normal at birth but develop progressive hypoglycemia and depletion of peripheral white cells, and die by about ten days of age (Xu et al., Immunity 5:407, 1996). The immune responses of chimeric mice reconstituted with TRAF3"^~ fetal liver cells resemble those of CD40-deficient mice, although TRAF3"/" B cells appear to be functionally normal.

The critical role of TRAF3 in signal transduction may be in its interaction with one of the other members of the TNF receptor superfamily, for example, CD30 or CD27, which are present on T cells. Alternatively, there may be other, as yet unidentified 1 SUBSTITUTE SHEET RULE 26 130491/2 members of this family of receptors that interact with TRAF3 and play an important role in postnatal development as well as in the development of a competent immune system. Identifying additional members of the TNF receptor superfamily would provide an additional means of regulating the immune and inflammatory response, as well as potentially providing further insight into post-natal development in mammals.

SUMMARY OF THE INVENTION The present invention provides an isolated DNA encoding for a RANK polypeptide or a fragment thereof selected from the group consisting of: (a) a DNA encoding a polypeptide comprising amino acids x to y of SEQ ID NO:6, wherein x is an amino terminal amino acid between amino acids 1 and 33, inclusive, of SEQ ID NO:6, and y is a carboxy terminal amino acid between amino acids 196 and 616, inclusive, of SEQ ID NO:6; (b) a DNA encoding a polypeptide comprising amino acids having at least 70% identity with amino acids x to y of SEQ ID NO:6, wherein x is an amino terminal amino acid between amino acids 1 and 33, inclusive, of SEQ ID NO:6, and y is a carboxy terminal amino acid between amino acids 196 and 213, inclusive, of SEQ ID NO:6; (c) a DNA encoding a polypeptide, the DNA being capable of hybridizing to a complement of the DNA of (a) or (b) under moderately stringent conditions or under stringent conditions, wherein moderately stringent conditions include prewashing with 5 x SSC, 0.5% SDS, 1.0 mM EDTA (pH 8.0) and hybridizing at 6 x SSC at 63°C overnight, and stringent conditions include hybridizing in 6 x SSC at 63°C overnight and washing in 3 x SSC at 55 °C, and (d) a DNA encoding for a fragment of a RANK polypeptide as defined in (a), (b), or (c); wherein said RANK polypeptide or said fragment has a biological activity exhibited by a native RANK molecule, with the amino acid sequence of SEQ ID NO:6.

RANK is a member of the TNF receptor superfamily and is a Type I transmembrane protein having 616 amino acid residues that interacts with TRAF3. 01184662M28-01 2 130491/1 Triggering of RANK by over-expression, co-expression of RANK and membrane bound RANK ligand (RANKL), and with addition of soluble RANKL or agonistic antibodies to RANK results in the upregulation of the transcription factor NF-KB, a ubiquitous transcription factor that is most extensively utilized in cells of the immune system.

Soluble forms of the receptor can be prepared and used to interfere with signal transduction through membrane-bound RANK, and hence upregulation of NF-KB; accordingly, pharmaceutical compositions comprising soluble forms of the novel receptor are also provided. Inhibition of NF-KB by RANK antagonists may be useful in ameliorating negative effects of an inflammatory response that result from triggering of RANK, for example in treating toxic shock or sepsis, graft-versus-host reactions, or acute inflammatory reactions. Soluble forms of the receptor will also be useful in vitro to screen for agonists or antagonists of RANK activity.

The cytoplasmic domain of RANK will be useful in developing assays for inhibitors of signal transduction, for example, for screening for molecules that inhibit interaction of RANK with TRAF2 or TRAF3.

The present invention also provides a recombinant expression vector comprising the DNA disclosed herein, a host cell transformed or transfected with the expression vector, a process for preparing a RANK polypeptide making use of the host cell, an isolated RANK polypeptide encoded by the DNA and uses of the RANK polypeptide.

Parts of the description which are no longer in ambit of the claims do not constitute part of the invention. 01184662M28-01 2a B5IEF DESCRIPTION OF THE DRAWINGS Figure ! demonstrates the influence of RANK.Fc and hRANKL on activated T cell growth. Human peripheral blood T cells were cultured as described in Example 12; viable T cell recovery was determined by triplicate trypan blue countings.

Figure 2 illustrates the ability of RANKL to induce human DC cluster formation.

Functionally mature dendritic cells (DC) were generated in vitro from CD34+ bone marrow (BM) progenitors and cultured as described in Example 13. CDla+ DC were cultured in a cytokine cocktail alone (Figure 2A), in cocktail plus CD40L (Figure 2B), RANKL (Figure 2C), or heat inactivated (ΔΗ) RANKL (Figure 2D), and then photographed using an inversion microscope.

Figure 3 demonstrates that RANKL enhances DC aUc-stimulatory capacity. Allogeneic T cells were incubated with varying numbers of irradiated DC cultured as described in Example 13. The cultures were pulsed with pH]-mymidine and the cells harvested onto glass fiber sheets for counting. Values represent the mean + standard deviation (SD) of triplicate cultures.

Figure 4 presents an alignment of human RANK with other TNFR family members in the region of structurally conserved extracellular cysteine-rich pseudorepeats. Predicted disulfide linkages (DS1-DS3) are indicated. RANK and CD40 contain identical amino acid substitutions (CAH, CAG) eliminating DS2 in the second pseudorepeat.

Figure 5 presents an alignment of human RANKL with other TNF family members.

DETAILED DESCRIPTION OF THE INVENTION A novel partial cDNA insert with a predicted open reading frame having some similarity to CD40 was identified in a database containing sequence information from cDNAs generated from human bone marrow-derived dendritic cells (DC). The insert was used to hybridize to colony blots generated from a DC cDNA library containing full-length cDNAs. Several colony hybridizations were performed, and two clones (SEQ ID NOs: l and 3) were isolated. SEQ ID NO:5 shows the nucleotide and amino acid sequence of a predicted full-length protein based on alignment of the overlapping sequences of SEQ ID NOs:l and 3.

RANK is a member of the TNF receptor superfamily; it most closely resembles CD40 in the extracellular region. Similar to CD40, RANK associates with TRAF2 and. TRAF3 (as determined by co-immunoprecipitation assays substantially as described by Rothe et al., Cell 83:1243, 1995). TRAFs are critically important in the regulation of the immune and inflammatory response. Through their association with various members of the TNF receptor superfamily, a signal is transduced to a cell. That signal results in the proliferation, differentiation or apoptosis of the cell, depending on which receptor(s) is/are triggered and which TRAF(s) associate with the receptor(s); different signals can be 3 WO 9S!2£ 2 PCT/US97/23866 transduced to ε cell vis coordination of various signaling events. Thus, a signal transduced through one member of this family may be proliferative, differentiative or apoptotic, depending on other signals being transduced to the cell, and/or the state of differentiation of the cell. Such exquisite regulation of this proliferative/apoptotic pathway is necessary to develop and maintain protection against pathogens; imbalances can result in autoimmune disease.

RANK is expressed on epithelial cells, some B cell lines, and on activated T cells. However, its expression on activated T cells is late, about four days after activation. This time course of expression coincides with the expression of Fas, a known agent of apoptosis. RANK may act as an anti-apoptotic signal, rescuing cells that express RANK from apoptosis as CD40 is known to do. Alternatively, RANK may confirm an apoptotic signal under the appropriate circumstances, again similar to CD40. RANK and its ligand are likely to play an integral role in regulation of the immune and inflammatory response.

Moreover, the post-natal lethality of mice having a targeted disruption of the TRAF3 gene demonstrates the importance of this molecule not only in the immune response but in development. The isolation of RANK, as a protein that associates with TRAF3, and its ligand will allow further definition of this signaling pathway, and development of diagnostic and therapeutic modalities for use in the area of autoimmune and/or inflammatory disease.

DNAs. Proteins and Analogs The present invention provides isolated RANK polypeptides and analogs (or muteins) thereof having an activity exhibited by the native molecule (i.e, RANK muteins that bind specifically to a RANK ligand expressed on cells or immobilized on a surface or to RANK-specific antibodies; soluble forms thereof that inhibit RANK ligand-induced signaling through RANK). Such proteins are substantially free of contaminating endogenous materials and, optionally, without associated native-pattern glycosylation. Derivatives of RANK within the scope of the invention also include various structural forms of the primary proteins which retain biological activity. Due to the presence of ionizable amino and carboxyl groups, for example, a RANK protein may be in the form of acidic or basic salts, or may be in neutral form. Individual amino acid residues may also be modified by oxidation or reduction. The primary amino acid structure may be modified by forming covalent or aggregative conjugates with other chemical moieties, such as glycosyl groups, lipids, phosphate, acetyl groups and the like, or by creating amino acid sequence mutants. Covalent derivatives are prepared by linking particular functional groups to amino acid side chains or at the N- or C-termini.

Derivatives of RANK may also be obtained by the action of cross-linking agents, such as M-maleimidobenzoyl succinimide ester and N-hydroxysuccinimide, at cysteine and 4 WO SS 2S42 PCT/US97 23866 l"Tsi~£ residues. The nventive proteins may also be covalently bound through reactive side groups to various insciubie substrates, such as cyanogen bromide-activated, bisoxirane-activaisd, carbonyidiirnidazole-activated or tosyl-activated agarose structures, or by adsorbing to polyolefin surfaces (with or without glutaraldehyde cross-linking). Once bound to a substrate, the proteins may be used to selectively bind (for purposes of assay or purification) antibodies raised against the proteins or against other proteins which are similar to RANK or RANKL, as well as other proteins that bind RANK or RANKL or homologs thereof.

Soluble forms of RANK are also within the scope of the invention. The nucleotide and predicted amino acid sequence of the RANK is shown in SEQ ID NOs: l through 6. Computer analysis indicated that the protein has an N-terminal signal peptide; the predicted cleavage site follows residue 24. Those skilled in the art will recognize that the actual cleavage site may be different than that predicted by computer analysis. Thus, the N-terminal amino acid of the cleaved peptide is expected to be within about five amino acids on either side of the predicted, preferred cleavage site following residue 24. Moreover a soluble form beginning with amino acid 33 was prepared; this soluble form bound RANKL. The signal peptide is predicted to be followed by a 188 amino acid extracellular domain, a 21 amino acid transmembrane domain, and a 383 amino acid cytoplasmic tail.

Soluble RANK comprises the signal peptide and the extracellular domain (residues 1 to 213 of SEQ ID NO:6) or a fragment thereof. Alternatively, a different signal peptide can be substituted for the native leader, beginning with residue 1 and continuing through a residue selected from the group consisting of amino acids 24 through 33 (inclusive) of SEQ ID NO:6. Moreover, fragments of the extracellular domain will also provide soluble forms of RANK. Fragments can be prepared using known techniques to isolate a desired portion of the extracellular region, and can be prepared, for example, by comparing the extracellular region with those of other members of the TNFR family and selecting forms similar to those prepared for other farnily members. Alternatively, unique restriction sites or PCR techniques that are known in the art can be used to prepare numerous truncated forms which can be expressed and analyzed for activity.

Fragments can be prepared using known techniques to isolate a desired portion of the extracellular region, and can be prepared, for example, by comparing the extracellular region with those of other members of the TNFR family (of which RANK is a member) and selecting forms similar to those prepared for other family members. Alternatively, unique restriction sites or PCR techniques that are known in the art can be used to prepare numerous truncated forms which can be expressed and analyzed for activity.

Other derivatives of the RANK proteins within the scope of this invention include covalent or aggregative conjugates of the proteins or their fragments with other proteins or polypeptides, such as by synthesis in recombinant culture as N-terminal or C-terrninal fusions. For example, the conjugated peptide may be a signal (or leader) polypeptide sequence εί the N-terminai region of the protein which co-translationalJy or post-translationally directs transfer of the protein' from its site of synthesis to its site of function inside or outside of the cell membrane or wall (e.g., the yeast a-factor leader).

Protein fusions can comprise peptides added to facilitate purification or identification of RANK proteins and homologs (e.g., poly-His). The amino acid sequence of the inventive proteins can also be linked to an identification peptide such as that described by Hopp et al., Bio/Technology 6:1204 (1988). Such a highly antigenic peptide provides an epitope reversibly bound by a specific monoclonal antibody, enabling rapid assay and facile purification of expressed recombinant protein. The sequence of Hopp et al. is also specifically cleaved by bovine mucosal enterokinase, allowing removal of the peptide from the purified protein. Fusion proteins capped with such peptides may also be resistant to intracellular degradation in E. coli.

Fusion proteins further comprise the amino acid sequence of a RANK linked to an immunoglobulin Fc region. An exemplary Fc region is a human IgG, having a nucleotide an amino acid sequence set forth in SEQ ID NO: 8. Fragments of an Fc region may also be used, as can Fc muteins. For example, certain residues within the hinge region of an Fc region are critical for high affinity binding to FcyRI. Canfield and Morrison (J. Exp. Med. 173: 1483; 1991) reported that Leu(234) and Leu(235)were critical to high affinity binding of IgG3 to FcyRI present on U937 cells. Similar results were obtained by Lund et al. (J. Immunol. 147:2657, 1991 ; Molecular Immunol. 29:53, 1991). Such mutations, alone or in combination, can be made in an IgG, Fc region to decrease the affinity of IgG, for FcR. Depending on the portion of the Fc region used, a fusion protein may be expressed as a dimer, through formation of interchain disulfide bonds. If the fusion proteins are made with both heavy and light chains of an antibody, it is possible to form a protein oligomer with as many as four RANK regions.

In another embodiment, RANK proteins further comprise an oligomerizing peptide such as a leucine zipper domain. Leucine zippers were originally identified in several DNA-binding proteins (Landschulz et al., Science 240: 1759, 1988). Leucine zipper domain is a term used to refer to a conserved peptide domain present in these (and other) proteins, which is responsible for dimerization of the proteins. The leucine zipper domain (also referred to herein as an oligomerizing, or oligomer-forming, domain) comprises a repetitive heptad repeat, with four or five leucine residues interspersed with other amino acids. Examples of leucine zipper domains are those found in the yeast transcription factor GCN4 and a heat-stable DNA-binding protein found in rat liver (C/EBP; Landschulz et al., Science 243: 1681 , 1989). Two nuclear transforming proteins, fos and jun, also exhibit leucine zipper domains, as does the gene product of the murine proto-oncogene, c-myc (Landschulz et al, Science 240: 1759, 1988). The products of the nuclear oncogenes fos 6 ζζή ίΐΰΐ cc~prise leucine zipper domains preferentially form a heterodimer (O'Shea et al., Science 245:546, 1989; Turner 'and Tjian, Science 243:1689, 1989). The leucine zipper domain is necessary for biological activity (DNA binding) in these proteins.

The fusogenic proteins of several different viruses, including paramyxovirus, coronavirus, measles virus and many retroviruses, also possess leucine zipper domains (Buckland and Wild, Nature 338:547,1989; Britton, Nature 353:394, 1991 ; Delwart and Mosialos, AIDS Research and Human Retroviruses 6:703, 1990). The leucine zipper domains in these fusogenic viral proteins are near the transmembrane region of the proteins; it has been suggested that the leucine zipper domains could contribute to the oiigomeric structure of the fusogenic proteins. Oligomerization of fusogenic viral proteins is involved in fusion pore formation (Spruce et al, Proc. Natl. Acad. Sci. U.S.A. 88:3523, 1991). Leucine zipper domains have also been recently reported to play a role in oligomerization of heat-shock transcription factors ( abindran et al., Science 259:230, 1993).

Leucine zipper domains fold as short, parallel coiled coils. (O'Shea et al., Science 254:539; 1991) The general architecture of the parallel coiled coil has been well characterized, with a "knobs-into-holes" packing as proposed by Crick in 1953 {Acta Crystallogr. 6:689). The dimer formed by a leucine zipper domain is stabilized by the heptad repeat, designated (abcdefg)n according to the notation of McLachlan and Stewart (J. Mol. Biol. 98:293; 1975), in which residues a and d are generally hydrophobic residues, with d being a leucine, which line up on the same face of a helix. Oppositely-charged residues commonly occur at positions g and e. Thus, in a parallel coiled coil formed from two helical leucine zipper domains, the "knobs" formed by the hydrophobic side chains of the first helix are packed into the "holes" formed between the side chains of the second helix.

The leucine residues at position d contribute large hydrophobic stabilization energies, and are important for dimer formation (Krystek et al., Int. J. Peptide Res. 38:229, 1991). Lovejoy et al. recently reported the synthesis of a triple-stranded a-helical bundle in which the helices run up-up-down (Science 259:1288, 1993). Their studies confirmed that hydrophobic stabilization energy provides the main driving force for the formation of coiled coils from helical monomers. These studies also indicate that electrostatic interactions contribute to the stoichiometry and geometry of coiled coils.

Several studies have indicated that conservative amino acids may be substituted for individual leucine residues with rninimal decrease in the ability to dimerize; multiple changes, however, usually result in loss of this ability (Landschulz et al., Science 243: 1681, 1989; Turner and Tjian, Science 243: 1689, 1989; Hu et al., Science 250: 1400, 1990). van Heekeren et al. reported that a number of different amino residues can be substituted for the leucine residues in the leucine zipper domain of GCN4, and further found that some GCN4 proteins containing two leucine substitutions were weakly active 7 130491/2 [N cL Acids Res. 20:3721 , 1992). Mutation of the first and second heptadic leucines of the leucine zipper domain of the measles virus fusion protein (MVF) did not affect syncytium formation (a measure of virally-induced cell fusion); however, mutation of all four leucine residues prevented fusion completely (Buckland et al., J. Gen. Virol. 73: 1703, 1992). None of the mutations affected the ability of MVF to form a tetramer.

Amino acid substitutions in the a and d residues of a synthetic peptide representing the GCN4 leucine zipper domain have been found to change the oligomerization properties of the leucine zipper domain (Alber, Sixth Symposium of the Protein Society, San Diego, CA). When all residues at position a are changed to isoleucine, the leucine zipper still forms a parallel dimer. When, in addition to this change, all leucine residues at position d are also changed to isoleucine, the resultant peptide spontaneously forms a trimeric parallel coiled coil in solution. Substituting all amino acids at position d with isoleucine and at position a with leucine results in a peptide that tetramerizes. Peptides containing these substitutions are still referred to as leucine zipper domains.

Also included within the scope of the invention are fragments or derivatives of the intracellular domain of RANK. Such fragments are prepared by any of the herein-mentioned techniques, and include peptides that are identical to the cytoplasmic domain of RANK as shown in SEQ ID NO:5, or of murine RANK as shown in SEP ID NO: 15. and those that comprise a portion of the cytoplasmic region. All techniques used in preparing soluble forms may also be used in preparing fragments or analogs of the cytoplasmic domain (i.e., RT-PCR techniques or use of selected restriction enzymes to prepare truncations). DNAs encoding all or a fragment of the intracytoplasmic domain will be useful in identifying other proteins that are associated with RANK signalling, for example using the immunoprecipitation techniques described herein, or another technique such as a yeast two-hybrid system (Rothe et al., supra).

The present invention also includes RANK with or without associated native-pattern glycosylation. Proteins expressed in yeast or mammalian expression systems, e.g., COS-7 cells, may be similar or slightly different in molecular weight and glycosylation pattern than the native molecules, depending upon the expression system. Expression of DNAs encoding the inventive proteins in bacteria such as E. coli provides non-glycosylated molecules. Functional mutant analogs of RANK protein having inactivated N-glycosylation sites can be produced by oligonucleotide synthesis and ligation or by site-specific mutagenesis techniques. These analog proteins can be produced in a homogeneous, reduced-carbohydrate form in good yield using yeast expression systems. N-glycosylation sites in eukaryotic proteins are characterized by the amino acid triplet Asn-A,-Z, where A, is any amino acid except Pro, and Z is Ser or Thr. In this sequence, asparagine provides a side chain amino group for covalent attachment of carbohydrate. Such a site can be eliminated by substituting another amino acid for Asn or for residue Z, 8 del ti g Asn or or inserting. £. non-Z amino acid between Ai and Z, or an amino acid other than Asn between Asn and A, .

RANK protein derivatives may also be obtained by mutations of the native RANK or subunits thereof. A RANK mutated protein, as referred to herein, is a polypeptide homologous to a native RANK protein, respectively, but which has an amino acid sequence different from the native protein because of one or a plurality of deletions, insertions or substitutions. The effect of any mutation made in a DNA encoding a mutated peptide may be easily determined by analyzing the ability of the mutated peptide to bind its counterstructure in a specific manner. Moreover, activity of RANK analogs, muteins or derivatives can be determined by any of the assays described herein (for example, inhibition of the ability of RANK to activate transcription).

Analogs of the inventive proteins may be constructed by, for example, making various substitutions of residues or -sequences or deleting terminal or internal residues or sequences not needed for biological activity. For example, cysteine residues can be deleted or replaced with other amino acids to prevent formation of incorrect intramolecular disulfide bridges upon renaturation. Other approaches to mutagenesis involve modification of adjacent dibasic amino acid residues to enhance expression in yeast systems in which KEX2 protease activity is present.

When a deletion or insertion strategy is adopted, the potential effect of the deletion or insertion on biological activity should be considered. Subunits of the inventive proteins may be constructed by deleting terminal or internal residues or sequences. Soluble forms of RANK can be readily prepared and tested for their ability to inhibit RANK-induced NF- B activation. Polypeptides corresponding to the cytoplasmic regions, and fragments thereof (for example, a death domain) can be prepared by similar techniques. Additional guidance as to the types of mutations that can be made is provided by a comparison of the sequence of RANK to proteins that have similar structures, as well as by perforaiing structural analysis of the inventive RANK proteins.

Generally, substitutions should be made conservatively; i.e., the most preferred substitute aniino acids are those which do not affect the biological activity of RANK (i.e., ability of the inventive proteins to bind antibodies to the corresponding native protein in substantially equivalent a manner, the ability to bind the counterstructure in substantially the same manner as the native protein, the ability to transduce a RANK signal, or ability to induce NF-κΒ activation upon overexpression in transient transfection systems, for example). Examples of conservative substitutions include substitution of amino acids outside of the binding domain(s) (either ligand/receptor or antibody binding areas for the extracellular domain, or regions that interact with other, intracellular proteins for the cytoplasmic domain), and substitution of amino acids that do not alter the secondary and/or 9 ίεΐϋετν s ruct re of the nsdvs protein. Additional examples include substituting one aliphatic residue for another, such as He, Val, Leu, or Ala for one another, or substitutions of one poiar residue for another, such as between Lys and Arg; Glu and Asp; or Gin and Asn. Other such conservative substitutions, for example, substitutions of entire regions having similar hydrophobicity characteristics, are well known.

Mutations in nucleotide sequences constructed for expression of analog proteins or fragments thereof must, of course, preserve the reading frame phase of the coding sequences and preferably will not create complementary regions that could hybridize to produce secondary mRNA structures such as loops or hairpins which would adversely affect translation of the mRNA.

Not all mutations in the nucleotide sequence which encodes a RANK protein or fragments thereof will be expressed in the final product, for example, nucleotide substitutions may be made to enhance expression, primarily to avoid secondary structure loops in the transcribed mRNA (see EPA 75,444A, incorporated herein by reference), or to provide codons that are more readily translated by the selected host, e.g., the well-known E. coli preference codons for E. coli expression.

Although a mutation site may be predetermined, it is not necessary that the nature of the mutation per se be predetermined. For example, in order to select for optimum characteristics of mutants, random mutagenesis may be conducted and the expressed mutated proteins screened for the desired activity. Mutations can be introduced at particular loci by synthesizing oligonucleotides containing a mutant sequence, flanked by restriction sites enabling ligation to fragments of the native sequence. Following ligation, the resulting reconstructed sequence encodes an analog having the desired amino acid insertion, substitution, or deletion.

Alternatively, oligonucleotide-directed site-specific mutagenesis procedures can be employed to provide an altered gene having particular codons altered according to the substitution, deletion, or insertion required. Exemplary methods of making the alterations set forth above are disclosed by Walder et al. {Gene 42:133, 1986); Bauer et al. (Gene 37:73, 1985); Craik (BioTechniques, January 1985, 12-19); Smith et al. (Genetic Engineering: Principles and Methods, Plenum Press, 1981); and U.S. Patent NOs. 4,518,584 and 4,737,462 disclose suitable techniques, and are incorporated by reference herein.

Other embodiments of the inventive proteins include RANK polypeptides encoded by DNAs capable of hybridizing to the DNA of SEQ ID NO: 6 under moderately stringent conditions (prewashing solution of 5 X SSC, 0.5% SDS, 1.0 mM EDTA (pH 8.0) and hybridization conditions of 50°C, 5 X SSC, overnight) to the DNA sequences encoding RANK, or more preferably under stringent conditions (for example, hybridization in 6 X SSC at 63°C overnight; washing in 3 X SSC at 55°C), and other sequences which are degenerate to these which encode the RANK. In one embodiment, RANK polypeptides £T£ £i le st about 70% identical in amino acid sequence to the amino acid sequence of native RANK protein as set forth in SEQ ID NO:5. In a preferred embodiment, RANK polypeptides are at least about 80% identical in amino acid sequence to the native form of RANK; most preferred polypeptides are those that are at least about 90% identical to native RANK.

Percent identity may be determined using a computer program, for example, the GAP computer program described by Devereux et al. (Nucl. Acids Res. 12:387, 1984) and available from the University of Wisconsin Genetics Computer Group (UWGCG). For fragments derived from the RANK protein, the identity is calculated based on that portion of the RANK protein that is present in the fragment The biological activity of RANK analogs or muteins can be determined by testing the ability of the analogs or muteins to inhibit activation of transcription, for example as described in the Examples herein. Alternatively, suitable assays, for example, an enzyme immunoassay or a dot blot, employing an antibody that binds native RANK, or a soluble form of RANKL, can be used to assess the activity of RANK analogs or muteins, as can assays that employ cells expressing RANKL. Suitable assays also include, for example, signal transduction assays and methods that evaluate the ability of the cytoplasmic region of RANK to associate with other intracellular proteins (i.e., TRAFs 2 and 3) involved in signal transduction will also be useful to assess the activity of RANK analogs or muteins. Such methods are well known in the art.

Fragments of the RANK nucleotide sequences are also useful. In one embodiment, such fragments comprise at least about 17 consecutive nucleotides, preferably at least about 25 nucleotides, more preferably at least 30 consecutive nucleotides, of the RANK DNA disclosed herein. DNA and RNA complements of such fragments are provided herein, along with both single-stranded and double-stranded forms- of the RANK DNA of SEQ ID NO:5, and those encoding the aforementioned polypeptides. A fragment of RANK DNA generally comprises at least about 17 nucleotides, preferably from about 17 to about 30 nucleotides. Such nucleic acid fragments (for example, a probe corresponding to the extracellular domain of RANK) are used as a probe or as primers in a polymerase chain reaction (PCR).

The probes also find use in detecting the presence of RANK nucleic acids in in vitro assays and in such procedures as Northern and Southern blots. Cell types expressing RANK can be identified as well. Such procedures are well known, and the skilled artisan can choose a probe of suitable length, depending on the particular intended application. For PCR, 5' and 3' primers corresponding to the termini of a desired RANK DNA sequence are employed to amplify that sequence, using conventional techniques. 1 1 FC /US97/23860 Other useful fragments' of the RANK nucleic acids are antisense or sense oligonucleotides comprising a single-stranded nucleic acid sequence (either RNA or DNA) capable of binding to target RANK mRNA (sense) or RANK DNA (antisense) sequences. The ability to create an antisense or a sense oligonucleotide, based upon a cDNA sequence for a given protein is described in, for example, Stein and Cohen, Cancer Res. 48:2659, 1988 and van der Krol et al., BioTechniques 6:958, 1988.

Uses of DNAs. Proteins and Analogs The RANK DNAs, proteins and analogs described herein will have numerous uses, including the preparation of pharmaceutical compositions. For example, soluble forms of RANK will be useful as antagonists of RANK-mediated NF-κΒ activation, as well as to inhibit transduction of a signal via RANK. RANK compositions (both protein and DNAs) will also be useful in development of both agonistic and antagonistic antibodies to RANK. The inventive DNAs are useful for the expression of recombinant proteins, and as probes for analysis (either quantitative or qualitative) of the presence or distribution of RANK transcripts.

The inventive proteins will also be useful in preparing kits that are used to detect soluble RANK or RANKL, or monitor RANK-related activity, for example, in patient specimens. RANK proteins will also find uses in monitoring RANK-related activity in other samples or compositions, as is necessary when screening for antagonists or mimetics of this activity (for example, peptides or small molecules that inhibit or mimic, respectively, the interaction). A variety of assay formats are useful in such kits, including (but not limited to) ELISA, dot blot, solid phase binding assays (such as those using a biosensor), rapid format assays and bioassays.

The purified RANK according to the invention will facilitate the discovery of inhibitors of RANK, and thus, inhibitors of an inflammatory response (via inhibition of NF-KB activation). The use of a purified RANK polypeptide in the screening for potential inhibitors is important and can virtually eliminate the possibility of interfering reactions with contaminants. Such a screening assay can utilize either the extracellular domain of RANK, the intracellular domain, or a fragment of either of these polypeptides. Detecting the inhibiting activity of a molecule would typically involve use of a soluble form of RANK derived from the extracellular domain in a screening assay to detect molecules capable of binding RANK and inhibiting binding of, for example, an agonistic antibody or RANKL, or using a polypeptide derived from the intracellular domain in an assay to detect inhibition of the interaction of RANK and other, intracellular proteins involved in signal transduction.

Moreover, in vitro systems can be used to ascertain the ability of molecules to antagonize or agonize RANK activity. Included in such methods are uses of RANK chimeras, for example, a chimera of the RANK intracellular domain and an extracellular 12 . WO _ PCT/US97/23866. do-n r. deri ed from £ protein aving a known ligand. The effects on signal transduction of various molecule can then be monitored by utilizing the known ligand to transduce a signal.

In addition, RANK polypeptides can also be used for structure-based design of RANK-inhibitors. Such structure-based design is also known as "rational drug design." The RANK polypeptides can be three-dimensionally analyzed by, for example, X-ray crystallography, nuclear magnetic resonance or homology modeling, all of which are well-known methods. The use of RANK structural information in molecular modeling software systems to assist in inhibitor design is also encompassed by the invention. Such computer-assisted modeling and drug design may utilize information such as chemical conformational analysis, electrostatic potential of the molecules, protein folding, etc. A particular method of the invention comprises analyzing the three dimensional structure of RANK for likely binding sites of substrates, synthesizing a new molecule that incorporates a predictive reactive site, and assaying the new molecule as described above.

Expression of Recombinant RANK The proteins of the present invention are preferably produced by recombinant DNA methods by inserting a DNA sequence encoding RANK protein or an analog thereof into a recombinant expression vector and expressing the DNA sequence in a recombinant expression system under conditions promoting expression. DNA sequences encoding the proteins provided by this invention can be assembled from cDNA fragments and short oligonucleotide linkers, or from a series of oligonucleotides, to provide a synthetic gene which is capable of being inserted in a recombinant expression vector and expressed in a recombinant transcriptional unit.

Recombinant expression vectors include synthetic or cDNA-derived DNA fragments encoding RANK, or homologs, muteins or bioequivalent analogs thereof, operably linked to suitable transcriptional or translational regulatory elements derived from mammalian, microbial, viral or insect genes. Such regulatory elements include a transcriptional promoter, an optional operator sequence to control transcription, a sequence encoding suitable mRNA ribosomal binding sites, and sequences which control the termination of transcription and translation, as described in detail below. The ability to replicate in a host, usually conferred by an origin of replication, and a selection gene to facilitate recognition of transformants may additionally be incorporated.

DNA regions are operably linked when they are functionally related to each other. For example, DNA for a signal peptide (secretory leader) is operably linked to DNA for a polypeptide if it is expressed as a precursor which participates in the secretion of the polypeptide; a promoter is operably linked to a coding sequence if it controls the transcription of the sequence; or a ribosome binding site is operably linked to a coding 13 sequence if it is positioned so as to permit translation. Generally, operably linked means contiguous and, in the case of secretory leaders, contiguous and in reading frame. DNA sequences encoding RANK, or homologs or analogs thereof which are to be expressed in a microorganism will preferably contain no introns that could prematurely terminate transcription of DNA into mRNA.

Useful expression vectors for bacterial use can comprise a selectable marker and bacterial origin of replication derived from commercially available plasmids comprising genetic elements of the well known cloning vector pBR322 (ATCC 37017). Such commercial vectors include, for example, pKK223-3 (Pharmacia Fine Chemicals, Uppsala, Sweden) and pGEMl (Promega Biotec, Madison, WI, USA). These pBR322 "backbone" sections are combined with an appropriate promoter and the structural sequence to be expressed. E. coli is typically transformed using derivatives of pBR322, a plasmid derived from an E. coli species (Bolivar et al., Gene 2:95, 1977). pBR322 contains genes for ampicillin and tetracycline resistance and thus provides simple means for identifying transformed cells .

Promoters commonly used in recombinant microbial expression vectors include the β-lactamase (penicillinase) and lactose promoter system (Chang et al., Nature 275:615, 1978; and Goeddel et al., Nature 281:544, 1979), the tryptophan (trp) promoter system (Goeddel et al., Nucl. Acids Res. 8:4051, 1980; and EPA 36,776) and tac promoter (Maniatis, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, p. 412, 1982). A particularly useful bacterial expression system employs the phage λ PL promoter and cI857ts thermolabile repressor. Plasmid vectors available from the American Type Culture Collection which incorporate derivatives of the λ PL promoter include plasmid pHUB2, resident in E. coli strain JMB9 (ATCC 37092) and pPLc28, resident in E. coli RR1 (ATCC 53082).

Suitable promoter sequences in yeast vectors" include the promoters for metallothionein, 3-phosphoglycerate kinase (Hitzeman et al., J. Biol. Chem. 255:2073, 1980) or other glycolytic enzymes (Hess et al., 7. Adv. Enzyme Reg. 7: 149, 1968; and Holland et al., Biochem. 77:4900, 1978), such as enolase, glyceraldehyde-3-phosphate dehydrogenase, hexokinase, pyruvate decarboxylase, phosphofructokinase, glucoses-phosphate isomerase, 3-phosphoglycerate mutase, pyruvate kinase, triosephosphate isomerase, phosphoglucose isomerase, and glucokinase. Suitable vectors and promoters for use in yeast expression are further described in R. Hitzeman et al., EPA 73,657.

Preferred yeast vectors can be assembled using DNA sequences from pBR322 for selection and replication in E. coli (Ampr gene and origin of replication) and yeast DNA sequences including a glucose-repressible ADH2 promoter and -factor secretion leader. The ADH2 promoter has been described by Russell et al. (J. Biol. Chem. 258:2614. 1982) and Beier et al. (Nature 300:124, 1982). The yeast a-factor leader, which directs secretion 14 . WO S®/2S42 _ PCT US97/23866. of heterologous proteins, c∑r. be inserted between the promoter and the structuril gene to be expressed. See, e.g., urjan et a!., Cell 30:933, 1982; and Bitter et al., Proc. Natl. Acad. Sci. USA 81:5330, 1984. The leader sequence may be modified to contain, near its 3' end, one or more useful restriction sites to facilitate fusion of the leader sequence to foreign genes..

The transcriptional and translational control sequences in expression vectors to be used in transforming vertebrate cells may be provided by viral sources. For example, commonly used promoters and enhancers are derived from Polyoma, Adenovirus 2, Simian Virus 40 (SV40), and human cytomegalovirus. DNA sequences derived from the SV40 viral genome, for example, SV40 origin, early and late promoter, enhancer, splice, and polyadenylation sites may be used to provide the other genetic elements required for expression of a heterologous DNA sequence. The early and late promoters are particularly useful because both are obtained easily from the virus as a fragment which also contains the SV40 viral origin of replication (Fiers et al., Nature 273: 1 13, 1978). Smaller or larger SV40 fragments may also be used, provided the approximately 250 bp sequence extending from the Hind III site toward the Bgll site located in the viral origin of replication is included. Further, viral genomic promoter, control and/or signal sequences may be utilized, provided such control sequences are compatible with the host cell chosen. Exemplary vectors can be constructed as disclosed by Okayama and Berg (Mol. Cell. Biol. 3:280, 1983).

A useful system for stable high level expression of mammalian receptor cDNAs in CI 27 murine mammary epithelial cells can be constructed substantially as described by Cosman et al. (Mol. Immunol. 23:935, 1986). A preferred eukaryotic vector for expression of RANK DNA is referred to as pDC406 (McMahan et al., EMBO J. 10:2821, 1991), and includes regulatory sequences derived from SV40, human immunodeficiency virus (HIV), and Epstein-Barr virus (EB V). Other preferred vectors include pDC409 and pDC410, which are derived from pDC406. pDC410 was derived from pDC406 by substituting the EBV origin of replication with sequences encoding the SV40 large T antigen. pDC409 differs from pDC406 in that a Bgl II restriction site outside of the multiple cloning site has been deleted, making the Bgl II site within the multiple cloning site unique.

A useful cell line that allows for episomal replication of expression vectors, such as pDC406 and pDC409, which contain the EBV origin of replication, is CV-1/EBNA (ATCC CRL 10478). The CV-1/EBNA cell line was derived by transfection of the CV-1 cell line with a gene encoding Epstein-Barr virus nuclear antigen- 1 (EBNA-1) and constitutively express EBNA-1 driven from human CMV immediate-early enhancer/promoter.

WO $£3842 PCT US97/23S66 Host Ceils Transformed host cells are ceils which have been transformed or transfected with expression vectors constructed using recombinant DNA techniques and which contain sequences encoding the proteins of the present invention. Transformed host cells may express the desired protein (RANK, or homologs or analogs thereof), but host cells transformed for purposes of cloning or amplifying the inventive DNA do not need to express the protein. Expressed proteins will preferably be secreted into the culture supernatant, depending on the DNA selected, but may be deposited in the cell membrane.

Suitable host cells for expression of proteins include prokaryotes, yeast or higher eukaryotic cells under the control of appropriate promoters. Prokaryotes include gram negative or gram positive organisms, for example E. coli or Bacillus spp. Higher eukaryotic cells include established cell lines of mammalian origin as described below. Cell-free translation systems could , also be employed to produce proteins using RNAs derived from the DNA constructs disclosed herein. Appropriate cloning and expression vectors for use with bacterial, fungal, yeast, and mammalian cellular hosts are described by Pouwels et al. (Cloning Vectors: A Laboratory Manual, Elsevier, New York, 1985), the relevant disclosure of which is hereby incorporated by reference.

Prokaryotic expression hosts may be used for expression of RANK, or homologs or analogs thereof that do not require extensive proteolytic and disulfide processing. Prokaryotic expression vectors generally comprise one or more phenotypic selectable markers, for example a gene encoding proteins conferring antibiotic resistance or supplying an autotrophic requirement, and an origin of replication recognized by the host to ensure amplification within the host. Suitable prokaryotic hosts for transformation include E. coli, Bacillus subtilis, Salmonella typhimurium, and various species within the genera Pseudomonas, Streptomyces, and Staphylococcus, although others may also be employed as a matter of choice.

Recombinant RANK may also be expressed in yeast hosts, preferably from the Saccharomyces species, such as S. cerevisiae. Yeast of other genera, such as Pichia or Kluyveromyces may also be employed. Yeast vectors will generally contain an origin of replication from the 2μ yeast plasmid or an autonomously replicating sequence (ARS), promoter, DNA encoding the protein, sequences for polyadenylation and transcription termination and a selection gene. Preferably, yeast vectors will include an origin of replication and selectable marker permitting transformation of both yeast and E. coli, e.g., the ampicillin resistance gene of E. coli and S. cerevisiae trpl gene, which provides a selection marker for a mutant strain of yeast lacking the ability to grow in tryptophan, and a promoter derived from a highly expressed yeast gene to induce transcription of a structural sequence downstream. The presence of the trpl lesion in the yeast host cell genome then 16 :iive environment for detecting transformation by growth in the absence of tryptophan.

Suitable yeast transformation protocols are known to those of skill in the art; an exemplary technique is described by Hinnen et al., Proc. Natl. Acad. Set USA 75: 1929, 1978, selecting for Trp+ transformants in a selective medium consisting of 0.67% yeast nitrogen base, 0.5% casamino acids, 2% glucose, 10 g/ml adenine and 20 μg ml uracil. Host strains transformed by vectors comprising the ADH2 promoter may be grown for expression in a rich medium consisting of 1% yeast extract, 2% peptone, and 1% glucose supplemented with 80 pg/ml adenine and 80 g/ml uracil. Derepression of the ADH2 promoter occurs upon exhaustion of medium glucose. Crude yeast supernatants are harvested by filtration and held at 4°C prior to further purification.

Various mammalian or insect cell culture systems can be employed to express recombinant protein. Baculovirus systems for production of heterologous proteins in insect cells are reviewed by Luckow and Summers, Bio/Technology 6:47 (1988). Examples of suitable mammalian host cell lines include the COS-7 lines of monkey kidney cells, described by Gluzman (Cell 25: 175, 1981), and other cell lines capable of expressing an appropriate vector including, for example, CV-1 EBNA (ATCC CRL 10478), L cells, CI 27, 3T3, Chinese hamster ovary (CHO), HeLa and BHK cell lines. Mammalian expression vectors may comprise nontranscribed elements such as an origin of replication, a suitable promoter and enhancer linked to the gene to be expressed, and other 5' or 3 ' flanking nontranscribed sequences, and 5' or 3' nontranslated sequences, such as necessary ribosome binding sites, a polyadenylation site, splice donor and acceptor sites, and transcriptional termination sequences.

Purification of Recombinant RANK Purified RANK, and homologs or analogs thereof are prepared by culturing suitable host/vector systems to express the recombinant translation products of the DNAs of the present invention, which are then purified from culture media or cell extracts. For example, supernatants from systems which secrete recombinant protein into culture media can be first concentrated using a commercially available protein concentration filter, for example, an Amicon or Millipore Pellicon ultrafiltration unit.

Following the concentration step, the concentrate can be applied to a suitable purification matrix. For example, a suitable affinity matrix can comprise a counter structure protein or lectin or antibody molecule bound to a suitable support. Alternatively, an anion exchange resin can be employed, for example, a matrix or substrate having pendant diethylaminoethyl (DEAE) groups. The matrices can be acrylamide, agarose, dextran, cellulose or other types commonly employed in protein purification. Alternatively, a cation exchange step can be employed. Suitable cation exchangers include various insoluble 17 matrices corr.prisir.g s lfopropyi or carboxymethyl groups. Sulfopropyl groups are preferred. Gel filtration chromatography also provides a means of purifying the inventive proteins.

Affinity chromatography is a particularly preferred method of purifying RANK and homologs thereof. For example, a RANK expressed as a fusion protein comprising an immunoglobulin Fc region can be purified using Protein A or Protein G affinity chromatography. Moreover, a RANK protein comprising an oligomerizing zipper domain may be purified on a resin comprising an antibody specific to the oligomerizing zipper domain. Monoclonal antibodies against the RANK protein may also be useful in affinity chromatography purification, by utilizing methods that are well-known in the art. A ligand may also be used to prepare an affinity matrix for affinity purification of RANK.

Finally, one or more reversed-phase high performance liquid chromatography (RP-HPLC) steps employing hydrophobic RP-HPLC media, e.g., silica gel having pendant methyl or other aliphatic groups, can be employed to further purify a RANK composition. Some or all of the foregoing purification steps, in various combinations, can also be employed to provide a homogeneous recombinant protein.

Recombinant protein produced in bacterial culture is usually isolated by initial extraction from cell pellets, followed by one or more concentration, salting-out, aqueous ion exchange or size exclusion chromatography steps. Finally, high- performance liquid chromatography (HPLC) can be employed for final purification steps. Microbial cells employed in expression of recombinant protein can be disrupted by any convenient method, including freeze-thaw cycling, sonication, mechanical disruption, or use of cell lysing agents.

Fermentation of yeast which express the inventive protein as a secreted protein greatly simplifies purification. Secreted recombinant protein resulting from a large-scale fermentation can be purified by methods analogous to those disclosed by Urdal et al. (J. Chromatog. 296: 11 \ , 1984). This reference describes two sequential, reversed-phase HPLC steps for purification of recombinant human GM-CSF on a preparative HPLC column.

Protein synthesized in recombinant culture is characterized by the presence of cell components, including proteins, in amounts and of a character which depend upon the purification steps taken to recover the inventive protein from the culture. These components ordinarily will be of yeast, prokaryotic or non-human higher eukaryptic origin and preferably are present in innocuous contaminant quantities, on the order of less than about 1 percent by weight. Further, recombinant cell culture enables the production of the inventive proteins free of other proteins which may be normally associated with the proteins as they are found in nature in the species of origin. 18 . G &S/2S424 . PCT OS97/23866 Uses and Adzninisira ion of RANK Compositions The present invention provides methods of using therapeutic compositions comprising an effective amount of a protein and a suitable diluent and carrier, and methods for regulating an immune or inflammatory response. The use of RANK in conjunction with soluble cytokine receptors or cytokines, or other immunoregulatory molecules is also contemplated.

For therapeutic use, purified protein is administered to a patient, preferably a human, for treatment in a manner appropriate to the indication. Thus, for example, RANK protein compositions administered to regulate immune function can be given by bolus injection, continuous infusion, sustained release from implants, or other suitable technique. Typically, a therapeutic agent will be adrninistered in the form of a composition comprising purified RANK, in conjunction with physiologically acceptable carriers, excipients or diluents. Such carriers will be nontoxic to recipients at the dosages and concentrations employed.

Ordinarily, the preparation of such protein compositions entails combining the inventive protein with buffers, antioxidants such as ascorbic acid, low molecular weight (less than about 10 residues) polypeptides, proteins, amino acids, carbohydrates including glucose, sucrose or dextrins, chelating agents such as EDTA, glutathione and other stabilizers and excipients. Neutral buffered saline or saline mixed with conspecific serum albumin are exemplary appropriate diluents. Preferably, product is formulated as a lyophilizate using appropriate excipient solutions (e.g., sucrose) as diluents. Appropriate dosages can be determined in trials. The amount and frequency of aclrninistration will depend, of course, on such factors as the nature and severity of the indication being treated, the desired response, the condition of the patient, and so forth.

Soluble forms of RANK and other RANK antagonists such as antagonistic monoclonal antibodies can be administered for the purpose of inhibiting RANK-induced induction of NF-κΒ activity. NF- Β is a transcription factor that is utilized extensively by cells of the immune system, and plays a role in the inflammatory response. Thus, inhibitors of RANK signalling will be useful in treating conditions in which signalling through RANK has given rise to negative consequences, for example, toxic or septic shock, or graft-versus-host reactions. They may also be useful in interfering with the role of NF-KB in cellular transformation. Tumor cells are more responsive to radiation when their NF- Β is blocked; thus, soluble RANK (or other antagonists of RANK signalling) will be useful as an adjunct therapy for disease characterized by neoplastic cells that express RANK. 19 WO 9S!2S42 PCT/US97/23866 The following examples are offered by way of illustration, and not by way of Those sidlled in the art will recognize that variations of the invention embodied in the examples can be made, especially in light of the teachings of the various references cited herein, the disclosures of which are incorporated by reference.

EXAMPLE 1 The example describes the identification and isolation of a DNA encoding a novel member of the TNF receptor superfamily. A partial cDNA insert with a predicted open reading frame having some similarity to CD40 (a cell-surface antigen present on the surface of both normal and neoplastic human B cells that has been shown to play an important role in B-cell proliferation and differentiation; Stamenkovic et al., EMBO J. 8:1403, 1989), was identified in a database containing sequence information from cDNAs generated from human bone marrow-derived dendritic cells (DC). The insert was excised from the vector by restriction endonuclease digestion, gel purified, labeled with 32P, and used to hybridize to colony blots generated from a DC cDNA library containing larger cDNA inserts using high stringency hybridization and washing techniques (hybridization in 5xSSC, 50% formamide at 42°C overnight, washing in 0.5xSSC at 63°C); other suitable high stringency conditions are disclosed in Sambrook et al. in Molecular Cloning: A Laboratory Manual. 2nd ed. (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY; 1989), 9.52-9.55. Initial experiments yielded a clone referred to as 9D-8A (SEQ ID NO:l); subsequent analysis indicated that this clone contained all but the extreme 5' end of a novel cDNA, with predicted intron sequence at the extreme 5' end (nucleotides 1-92 of SEQ ID NO:l). Additional colony hybridizations were performed, and a second clone was isolated. The second clone, referred to as 9D-15C (SEQ ID NO:3), contained the 5' end without intron interruption but not the full 3'end. SEQ ID NO:5 shows the nucleotide and amino acid sequence of a predicted full-length protein based on alignment of the overlapping sequences of SEQ ID NOs:l and 3.

The encoded protein was designated RANK, for receptor activator of NF-κΒ. The cDNA encodes a predicted Type 1 transmembrane protein having 616 amino acid residues, with a predicted 24 amino acid signal sequence (the computer predicted cleavage site is after Leu24), a 188 amino acid extracellular domain, a 21 amino acid transmembrane domain, and a 383 amino acid cytoplasmic tail. The extracellular region of RANK displayed significant amino acid homology (38.5% identity, 52.3% similarity) to CD40. A cloning vector (pBluescriptSK-) containing human RANK sequence, designated pBluescript:huRANK (in E. coli DH10B), was deposited with the American Type Culture Collection, Rockville, MD (ATCC) on December 20, 1996, under terms of the Budapest Treaty, and given accession number 98285. 130491/2 EXAMPLE 2 This example describes construction of a RANK DNA construct to express a RANK/Fc fusion protein. A soluble form of RANK fused to the Fc region of human IgG, was constructed in the mammalian expression vector pDC409 (WO 95/18819). This expression vector encodes the leader sequence of the Cytomegalovirus (CMV) open reading frame R27080 (SEQ ED NO: 9), followed by amino acids 33-213 of RANK, followed by a mutated form of the constant domain of human IgG, that exhibits reduced affinity for Fc receptors (SEQ ED NO:8; for the fusion protein, the Fc portion of the construct consisted of Arg3 through Lys232). An alternative expression vector encompassing amino acids 1-213 of RANK (using the native leader sequence) followed by the IgG, mutein was also prepared. Both expression vectors were found to induce high levels of expression of the RANK Fc fusion protein in transfected cells.

To obtain RANK/Fc protein, a RANK Fc expression plasmid is transfected into CV-l/EBNA cells, and supernatants are collected for about one week. The RANK/Fc fusion protein is purified by means well-known in the art for purification of Fc fusion proteins, for example, by protein A sepharose column chromatography according to manufacturer's recommendations (i.e., Pharmacia, Uppsala, Sweden). SDS-polyacrylamide gel electrophoresis analysis indicted that the purified RANK/Fc protein migrated with a molecular weight of ~55kDa in the presence of a reducing agent, and at a molecular weight of ~ 1 lOkDa in the absence of a reducing agent.

N-terminal amino acid sequencing of the purified protein made using the CMV R27080 leader showed 60% cleavage after Ala20, 20% cleavage after Pro22 and 20% cleavage after Arg28 (which is the Furin cleavage site; amino acid residues are relative to SEQ ED NO:9); N-terminal amino acid analysis of the fusion protein expressed with the native leader showed cleavage predominantly after Gln25 (80% after Gln25 and 20% after Arg23; amino acid residues are relative to SEQ ED NO:6, full-length RANK). Both fusion proteins were able to bind a ligand for RANK is a specific manner (i.e., they bound to the surface of various cell lines such as a murine thymoma cell line, EL ), indicating that the presence of additional amino acids at the N-terminus of RANK does not interfere with its ability to bind RANKL. Moreover, the construct comprising the CMV leader encoded RANK beginning at amino acid 33; thus, a RANK peptide having an N-terminus at an amino acid between Arg23 and Pro33, inclusive, is expected to be able to bind a ligand for RANK in a specific manner.

Other members of the TNF receptor superfamily have a region of amino acids between the transmembrane domain and the ligand binding domain that is referred to as a 'spacer' region, which is not necessary for ligand binding. In RANK, the amino acids between 196 and 213 are predicted to form such a spacer region. Accordingly, a soluble form of RANK that terminates with an amino acid in this region is expected to retain the 21 . O ?8/2S<-24 . PCT OS97/23866 ability to bind a ligand for RANK in a specific manner. Preferred C-terminal amino acids for soluble RANK peptides are selected from the group consisting of amino acids 213 and 196 of SEQ ID NO:6, although other amino acids in the spacer region may be utilized as a C- terminus.

EXAMPLE 3 This example illustrates the preparation of monoclonal antibodies against RANK. Preparations of purified recombinant RANK, for example, or transfected cells expressing high levels of RANK, are employed to generate monoclonal antibodies against RANK using conventional techniques, such as those disclosed in U.S. Patent 4,41 1,993. DNA encoding RANK can also be used as an immunogen, for example, as reviewed by Pardoll and Beckerleg in Immunity 3: 165, 1995. Such antibodies are likely to be useful in interfering with RANK-induced signaling (antagonistic or blocking antibodies) or in inducing a signal by cross-linking RANK (agonistic antibodies), as components of diagnostic or research assays for RANK or RANK activity, or in affinity purification of RANK.

To immunize rodents, RANK immunogen is emulsified in an adjuvant (such as complete or incomplete Freund's adjuvant, alum, or another adjuvant, such as Ribi adjuvant R700 (Ribi, Hamilton, MT), and injected in amounts ranging from 10-100 g subcutaneously into a selected rodent, for example, BALB/c mice or Lewis rats. DNA may be given intradermally (Raz et al., Proc. Natl. Acad. Set USA 91:9519, 1994) or intamuscularly (Wang et al., Proc. Natl. Acad. Sci. USA 90:4156, 1993); saline has been found to be a suitable diluent for DNA-based antigens. Ten days to three weeks days later, the immunized animals are boosted with additional immunogen and periodically boosted thereafter on a weekly, biweekly or every third week immunization schedule.

Serum samples are periodically taken by retro-orbital bleeding or tail-tip excision for testing by dot-blot assay (antibody sandwich), ELISA (enzyme-linked immunosorbent assay), irnmunoprecipitation, or other suitable assays, including FACS analysis. Following detection of an appropriate antibody titer, positive animals are given an intravenous injection of antigen in saline. Three to four days later, the animals are sacrificed, splenocytes harvested, and fused to a murine myeloma cell line (e.g., NS1 or preferably Ag 8.653 [ATCC CRL 1580]). Hybridoma cell lines generated by this procedure are plated in multiple microtiter plates in a selective medium (for example, one containing hypoxanthine, aminopterin, and thymidine, or HAT) to inhibit proliferation of non-fused cells, myeloma-myeloma hybrids, and splenocyte-splenocyte hybrids.

Hybridoma clones thus generated can be screened by ELISA for reactivity with RANK, for example, by adaptations of the techniques disclosed by Engvali et ai., Immunochem. &871 (1971) and in U.S. Patent 4,703,004. A preferred screening 22 lechdc e is the antibody a t re technique described by Beckman et al., J. Immunol. 144:4212 (1990). Positive clones are then injected into the peritoneal cavities of syngeneic rodents to produce ascites containing high concentrations (>1 mg/mi) of anti-RAN monoclonal antibody. The resulting monoclonal antibody can be purified by ammonium sulfate precipitation followed by gel exclusion chromatography. Alternatively, affinity chromatography based upon binding of antibody to protein A or protein G can also be used, as can affinity chromatography based upon binding to RANK protein.

Monoclonal antibodies were generated using RANK/Fc fusion protein as the immunogen. These reagents were screened to confirm reactivity against the RANK protein. Using the methods described herein to monitor the activity of the mAbs, both blocking (i.e., antibodies that bind RANK and inhibit binding of a ligand to RANK) and non-blocking (i.e., antibodies that bind RANK and do not inhibit ligand binding) were isolated.

EXAMPLE 4 This example illustrates the induction of NF-κΒ activity by RANK in 293/EBNA cells (cell line was derived by transfection of the 293 cell line with a gene encoding Epstein-Barr virus nuclear antigen- 1 (EBNA-1) that constitutively express EBNA-1 driven from human CMV immediate-early enhancer/promoter). Activation of NF-κΒ activity was measured in 293/EBNA cells essentially as described by Yao et al. (Immunity 3:811, 1995). Nuclear extracts were prepared and analyzed for NF-icB activity by a gel retardation assay using a 25 base pair oligonucleotide spanning the NF-κΒ binding sites. Two million cells were seeded into 10 cm dishes two days prior to DNA transfection and cultured in DMEM-F12 media containing 2.5% FBS (fetal bovine serum). DNA transfections were performed as described herein for the H_-8 promoter/reporter assays.

Nuclear extracts were prepared by solubilization of isolated nuclei with 400 mM NaCl (Yao et al., supra). Oligonucleotides containing an NF-κΒ binding site were annealed and endlabeled with 32P using T4 DNA polynucleotide kinase. Mobility shift reactions contained 10 g of nuclear extract, 4 g of poly(dl-dC) and 15,000 cpm labeled double-stranded oligonucleotide and incubated at room temperature for 20 minutes. Resulting protein-DNA complexes were resolved on a 6% native polyacrylamide gel in 0.25 X Tris-borate-EDTA buffer.

Overexpression of RANK resulted in induction of NF-κΒ activity as shown by an appropriate shift in the mobility of the radioactive probe on the gel. Similar results were observed when RANK was triggered by a ligand that binds RANK and transduces a signal to cells expressing the receptor (i.e., by co-transfecting cells with human RANK and murine RANKL DNA; see Example 7 below), and would be expected to occur when triggering is done with agonistic antibodies. 23 . WO SS/2S424 ·' . . . PCT/US97/23866 EXAMPLE 5 This example describes a gene promoter/reporter system based on the human Interteuidn-8 (IL-8 ) promoter used to analyze the activation of gene transcription in vivo. The induction of human IL-8 gene transcription by the cytokines Interleukin-1 (IL-1) or tumor necrosis factor-alpha (TNF-a) is known to be dependent upon intact NF-κΒ and NF-IL-6 transcription factor binding sites. Fusion of the cytokine-responsive IL-8 promoter with a cDNA encoding the murine 1L-4 receptor (mIL-4R) allows measurement of promoter activation by detection of the heterologous reporter protein (mIL-4R) on the cell surface of transfected cells.

Human kidney epithelial cells (293/EBNA) are transfected (via the DEAE DEXTRAN method) with plasmids encoding: 1). the reporter/promoter construct (referred to as pIL-8rep), and 2). the cDNA(s) of interest . DNA concentrations are always kept constant by the addition of empty vector DNA. The 293 EBNA cells are plated at a density of 2.5 x 10^ cells/ml (3 ml/ well) in a 6 well plate and incubated for two days prior to transfection. Two days after transfection, the mIL-4 receptor is detected by a radioimmunoassay (RIA) described below.

In one such experiment, the 293 EBNA cells were co-transfected with DNA encoding RANK and with DNA encoding RANKL (see Example 7 below). Co-expression of this receptor and its counterstructure by cells results in activation of the signaling process of RANK. For such co-transfection studies, the DNA concentration/well for the DEAE transfection were as follows: 40 ng of pIL-8rep [pBluescriptSK- vector (Stratagene)]; 0.4 ng CD40 (DNA encoding CD40, a control receptor; pCDM8 vector); 0.4 ng RANK (DNA encoding RANK; pDC409 vector), and either 1-50 ng CD40L (DNA encoding the ligand for CD40, which acts as a positive control when co-transfected with CD40 and as a negative control when co-transfected with RANK; in pDC304) or RANKL (DNA encoding a ligand for RANK; in pDC406). Similar experiments can be done using soluble RANKL or agonistic antibodies to RANK to trigger cells transfected with RAN For the irdL-4R-specific RIA, a monoclonal antibody reactive with mIL-4R is labeled with I2SI via a Chloramine T conjugation method; the resulting specific activity is typically 1.5 x 10^ cpm/nmol. After 48 hours, transfected cells are washed once with media (DMEM F12 5% FBS). Non-specific binding sites are blocked by the addition of pre-warmed binding media containing 5% non-fat dry milk and incubation at 37°C/5 C02 in a tissue culture incubator for one hour. The blocking media is decanted and binding buffer containing 125I anti-rnIL-4R (clone Ml; rat IgGl) is added to the cells and incubated with rocking at room temperature for 1 hour. After incubation of the cells with the radiolabeled antibody, cells are washed extensively with binding buffer (2X) and mice with 24 hcs haie-buf ezed saline (PBS). Ceils are iysea in 1 mi of 0.5M NaOH, and total radioactivity is measured with a gamma counter.

Using , this assay, 293 EBNA co-'transfected with DNAs encoding RANK demonstrated transcriptional activation, as shown by detection of muTL-4R on the cell surface. Overexpression of RANK resulted in transcription of muIL-4R, as did triggering of the RANK by RANKL. Similar results are observed when RANK is triggered by agonistic antibodies.

EXAMPLE 6 This example illustrates the association of RANK with TRAF proteins. Interaction of RANK with cytoplasmic TRAF proteins was demonstrated by ccnrnmunoprecipitation assays essentially as described by Hsu et al. (Cell 84:299; 1996). Briefly, 293/EBNA cells were co-transfected with plasnii& that direct the synthesis of RANK and epitope-tagged (FLAG®; SEQ ID NO:7) TRAF2 or TRAF3. Two days after transfection, surface proteins were labeled with biotin-ester, and cells were lysed in a buffer containing 0.5% NP-40. RANK and proteins associated with this receptor were immunoprecipitated with anti-RANK, washed extensively, resolved by electrophoretic separation on a 6-10% SDS polyacrylamide gel and electrophoretically transferred to a nitrocellulose membrane for Western blotting. The association of TRAF2 and TRAF3 proteins with RANK was visualized by probing the membrane with an antibody that specifically recognizes the FLAG® epitope. TRAFs 2 and 3 did not irnmunopreciptitate with anti-RANK in the absence of RANK expression.

EXAMPLE 7 This example describes isolation of a ligand for RANK, referred to as RANKL, by direct expression cloning. The ligand was cloned essentially as described in US Patent No. 5,961,974, filed May 24, 1994 (the relevant disclosure of which is incorporated by reference herein), for CD40L. Briefly, a library was prepared from a clone of a mouse thymoma cell line EL-4 (ATCC ΤΓΒ 39), called EL-40.5, derived by sorting five times with biotinylated CD40 Fc fusion protein in a FACS (fluorescence activated cell sorter). The cDNA library was made using standard methodology; the plasmid DNA was isolated and transfected into sub-confluent CV1-EBNA cells using a DEAE-dextran method. Transfectants were screened by slide autoradiography for expression of RANKL using a two-step binding method with RANK Fc fusion protein as prepared in Example 2 followed by radioiodinated goat anti-human IgG antibody.

A clone encoding a protein that specifically bound RANK was isolated and sequenced; the clone was referred to as 11H. An expression vector containing murine RANKL sequence, designated pDC406:muRANK-L (in E. coli DH10B), was deposited -~ith the A∑xs±xa Type Culture Collection, Rockvi!le, MD (ATCC) on December 20, 1996, under terms of the Budapest Treaty, and given accession number 98284. The nucleotide sequence and predicted amino acid sequence of this clone are illustrated in SEQ ID NO: 10. This clone did not contain an initiator methionine; additional, full-length clones were obtained from a 7B9 library (prepared substantially as described in US patent 5,599,905, issued February 4, 1997); the 5' region was found to be identical to that of human RANKL as shown in SEQ ID NO: 12, amino acids 1 through 22, except for substitution of a Gly for a Thr at residue 9.

This ligand is useful for assessing the ability of RANK to bind RANKL by a number of different assays. For example, transfected cells expressing RANKL can be used in a FACS assay (or similar assay) to evaluate the ability of soluble RANK to bind RANKL. Moreover, soluble forms of RANKL can be prepared and used in assays that are known in the art (i.e., ELISA ot BIAcore assays essentially as described in USSN 08/249,189, filed May 24, 1994). RANKL is also useful in affinity purification of RANK, and as a reagent in methods to measure the levels of RANK in a sample. Soluble RANKL is also useful in inducing NF-κΒ activation and thus protecting cells that express RANK from apoptosis.

EXAMPLE 8 This example describes the isolation of a human RANK ligand (RANKL) using a PCR-based technique. Murine RANK ligand-specific oligonucleotide primers were used in PCR reactions using human cell line-derived first strand cDNAs as templates. Primers corresponded to nucleotides 478-497 and to the complement of nucleotides 858-878 of murine RANK ligand (SEQ ID NO: 10). An amplified band approximately 400 bp in length from one reaction using the human epidermoid cell line KB (ATCC CCL-17) was gel purified, and its nucleotide sequence determined; the sequence was 85% identical to the corresponding region of murine RANK ligand, confirming that the fragment was from human RANKL.

To obtain full-length human RANKL cDNAs, two human RANKL-specific oligonucleotides derived from the KB PCR product nucleotide sequence were radiolabeled and used as hybridization probes to screen a human PBL cDNA library prepared in lambda gtlO (Stratagene, La Jolla, CA), substantially as described in US patent 5,599,905, issued February 4, 1997 . Several positive hybridizing plaques were identified and purified, their inserts subcloned into pBluescript SK" (Stratagene, La Jolla, CA), and their nucleotide sequence determined. One isolate, PBL3, was found to encode most of the predicted human RANKL, but appeared to be missing approximately 200 bp of 5' coding region. A second isolate, PBL5 was found to encode much of the predicted human RANKL, including the entire 5 end and an additional 200 bp of 5' untranslated sequence. 26 WO ?§/£§4£4 PCT/US97/23866 The 5: end of. PELS and ihs 3' end of PBL3 were iigated together to form a full iesgih cDNA encoding human RANKL. The nucleotide and predicted amino acid sequence of the fuil-iength human RANK iigand is shown in SEQ ID NO: 12. Human RANK ligand shares 83% nucleotide and 84% amino acid identity with murine RANK ligand. A plasmid vector containing human RANKL sequence, designated pBluescript:huRANK-L (in E. coli DH10B), was deposited with the American Type Culture Collection, Rockville, MD (ATCC) on March 11, 1997 under terms of the Budapest Treaty, and given accession number 98354.

Murine and human RANKL are Type 2 transmembrane proteins. Murine RANKL contains a predicted 48 amino acid intracellular domain, 21 amino acid transmembrane domain and 247 amino acid extracellular domain. Human RANKL contains a predicted 47 amino acid intracellular domain, 21 amino acid transmembrane domain and 249 amino acid extracellular domain. _____ EXAMPLE 9 This example describes the chromosomal mapping of human RANK using PCR-based mapping strategies. Initial human chromosomal assignments were made using RANK and RANKL-specific PCR primers and a BIOS Somatic Cell Hybrid PCRable DNA kit from BIOS Laboratories (New Haven, CT), following the manufacturer's instructions. RANK mapped to human chromosome 18; RANK ligand mapped to human chromosome 13. More detailed mapping was performed using a radiation hybrid mapping panel Genebridge 4 Radiation Hybrid Panel (Research Genetics, Huntsville, AL;. described in Walter, MA et al., Nature Genetics 7:22-28, 1994). Data from this analysis was then submitted electronically to the ΜΓΓ Radiation Hybrid Mapper (URL: http://www-genome.wi.mit.edu/cgi-bin/contig/rhmapper.pl) following the instructions contained therein. This analysis yielded specific genetic marker names which, when submitted electronically to the NCBI Entrez browser (URL: http://www3.ncbi.nlm.nih.gov/htbin-post/Entrez query?db=c&form=0), yielded the specific map locations. RANK mapped to chromosome 18q22.1, and RANKL mapped to chromosome 13ql4.

EXAMPLE 10 This example illustrates the preparation of monoclonal antibodies against RANKL. Preparations of purified recombinant RANKL, for example, or transfixed cells expressing high levels of RANKL, are employed to generate monoclonal antibodies against RANKL using conventional techniques, such as those disclosed in US Patent 4,411,993. DNA encoding RANKL can also be used as an immunogen, for example, as reviewed by Pardoll and Beckerleg in Immunity 3: 165, 1995. Such antibodies are likely to be useful in interfering with RANKL signaling (antagonistic or blocking antibodies), as components of 27 diagnostic or research assays for xAN L or ANKL activity, or in affinity purification of Ai'T L.

To imrnunize rodents, RANKL immunogen is emulsified in an adjuvant (such as complete or incomplete Freund's adjuvant, alum, or another adjuvant, such as Ribi adjuvant R700 (Ribi, Hamilton, MT), and injected in amounts ranging from 10-100 μg subcutaneously into a selected rodent, for example, BALB/c mice or Lewis rats. DNA may be given intradermally (Raz et al., Proc. Natl. Acad. Sci. USA 91:9519, 1994) or intamuscularly (Wang et al., Proc. Natl. Acad. Sci. USA 90:4156, 1993); saline has been found to be a suitable diluent for DNA-based antigens. Ten days to three weeks days later, the immunized animals are boosted with additional immunogen and periodically boosted thereafter on a weekly, biweekly or every third week immunization schedule.

Serum samples are periodically taken by retro-orbital bleeding or tail-tip excision for testing by dot-blot assay (antibody sandwich), ELISA (enzyme-linked immunosorbent assay), immunoprecipitation, or other suitable assays, including FACS analysis. Following detection of an appropriate antibody titer, positive animals are given an intravenous injection of antigen in saline. Three to four days later, the animals are sacrificed, splenocytes harvested, and fused to a murine myeloma cell line (e.g., NS1 or preferably Ag 8.653 [ATCC CRL 1580]). Hybridoma cell lines generated by this procedure are plated in multiple microtiter plates in a selective medium (for example, one containing hypoxanthine, aminopterin, and thymidine, or HAT) to inhibit proliferation of non-fused cells, myeloma-myeloma hybrids, and splenocyte-splenocyte hybrids.

Hybridoma clones thus generated can be screened by ELISA for reactivity with RANKL, for example, by adaptations of the techniques disclosed by Engvall et al., Immunochem. 8:871 (1971) and in US Patent 4,703,004. A preferred screening technique is the antibody capture technique described by Beckman et al., J. Immunol. 144:4212 (1990). Positive clones are then injected into the peritoneal cavities of syngeneic rodents to produce ascites containing high concentrations (>1 mg ml) of anti-RANK monoclonal antibody. The resulting monoclonal antibody can be purified by ammonium sulfate precipitation followed by gel exclusion chromatography. Alternatively, affinity chromatography based upon binding of antibody to protein A or protein G can also be used, as can affinity chromatography based upon binding to RANKL protein. Using the methods described herein to monitor the activity of the mAbs, both blocking (i.e., antibodies that bind RANKL and inhibit binding to RANK) and non-blocking (i.e., antibodies that bind RANKL and do not inhibit binding) are isolated. 28 EXAMPLE 11 This example demonstrates that RANK expression can be up-regulated. Human peripheral blood T cells were purified by flow cytometry sorting or by negative selection using antibody coated beads, and activated with anti-CD3 (OKT3, Dako) coated plates or phytohemagglutinin in the presence or absence of various cytokines, including Interleukin-4 (IL-4) .Transforming Growth Factor-β (TGF-β) and other commercially available cytokines ( ILl-a, IL-2, IL-3, IL-6, IL-7, IL-8, IL-10, IL-12, IL-15, IFN-γ, TNF- a). Expression of RANK was evaluated by FACS in a time course experiment for day 2 to day 8, using a mouse monoclonal antibody mAbl44 (prepared as described in Example 3), as shown in the table below. Results are expressed as '+' to '++++' referring to the relative increase in intensity of staining with anti-RANK. Double labeling experiments using both anti-RANK and anti-CD8 or anti-CD4 antibodies were also performed.

Table 1 : Upregulation of RANK by Cytokines Of the cytokines tested, IL-4 and TGF-β increased the level of RANK expression on both CD8+ cytotoxic and CD4+ helper T cells from day 4 to day 8. The combination of IL-4 and TGF-β acted synergistically to upregulate expression of this receptor on activated T cells. This particular combination of cytokines is secreted by suppresser T cells, and is believed to be important in the generation of tolerance (reviewed in Mitchison and Sieper, Z. Rheumatol. 54: 141, 1995), implicating the interaction of RANK in regulation of an immune response towards either tolerance or induction of an active immune response. 29 EXAMPLE 12 This example illustxates the influence of RANKLFc and hRANKL on activated T ceil growth. The addition of TGFB to anti-CD3 activated human peripheral blood T lymphocytes induces proliferation arrest and idtimately death of most lymphocytes within the first few days of culture. We tested the effect of RANK:RANKL interactions on TGFB-treated T cells by adding RANKLFc or soluble human RANKL to T cell cultures.

Human peripheral blood T cells (7 x 10s PBT) were cultured for six days on anti-CD3 (OKT3, and anti-Flag (Ml, 5μg/ml) coated 24 well plates in the presence of TGFB (lng/ml) and EL-4 (lOng/ml), with or without recombinant FLAG-tagged soluble hRANKL (^g/ml) or RANK.Fc (ΙΟμξ/πύ). Viable T cell recovery was determined by triplicate trypan blue countings.

The addition of RANK.Fc significantly reduced the number of viable T cells recovered after six days, whereas soluble RANKL gready increased the recovery of viable T cells (Figure 1). Thus, endogenous or exogenous RANKL enhances the number of viable T cells generated in the presence of TGFB. TGFB, along with IL-4, has been implicated in immune response regulation when secreted by the TH3/regulatory T cell subset. These T cells are believed to mediate bystander suppression of effector T cells. Accordingly, RANK and its ligand may act in an auto/paracrine fashion to influence T cell tolerance. Moreover, TGFB is known to play a role in the evasion of the immune system effected by certain pathogenic or opportunistic organisms. In addition to playing a role in the development of tolerance, RANK may also play a role in immune system evasion by pathogens.

EXAMPLE 13 This example illustrates the influence of the interaction of RANK on CDla+ dendritic cells (DC). Functionally mature dendritic cells (DC) were generated in vitro from CD34+ bone marrow (BM) progenitors. Briefly, human BM cells from normal healthy volunteers were density fractionated using Ficoll medium and CD34+ cells irnmunoaffinity isolated using an anti-CD34 matrix column (Ceprate, CellPro). The CD34+ BM cells were then cultured in human GM-CSF (20 ng/ml), human IL-4 (20 ng/ml), human TNF-a (20 ng/ml), human CHO-derived Flt3L (FL; 100 ng/ml) in Super McCoy's medium supplemented with 10% fetal calf serum in a fully humidified 37°C incubator (5% CO2) for 14 days. CDla+, HLA-DR+ DC were then sorted using a FACStar Plus™, and used for biological evaluation of RANK On human CDla÷ DC derived from CD34+ bone marrow cells, only a subset (20- 30%) of CDla+ DC expressed RANK at the cell surface as assessed by flow cytometric Q 9S!2&<24 . PCT/US97/23866 analysis. However, addition of CD40L to the DC cultures resulted in RANK surface expression on the majority of CDla÷ DC. CD40L has been shown to activate DC by enhancing in vitro cluster formation, inducing DC morphological changes and upregulating HLA-DR, CD54, CD58, CD80 and CD86 expression Addition of RANKL to DC cultures significantly increased the degree of DC aggregation and cluster formation above control cultures, similar to the effects seen with CD40L (Figure 2). Sorted human CDla+ DC were cultured in a cytokine cocktail (GM-CSF, IL-4, TNF-a and FL) (upper left panel), in cocktail plus CD40L ( g/ml) (upper right), in cocktail plus RANKL (^g/ml) (lower left), or in cocktail plus heat inactivated (ΔΗ) RANKL (^g/ml) (lower right) in 24-well flat bottomed culture plates in 1 ml culture media for 48-72 hours and then photographed using an inversion microscope. An increase in DC aggregation and cluster formation above control cultures was not evident when heat inactivated RANKL was used, indicating that this effect was dependent on biologically active protein. However, initial phenotypic analysis of adhesion molecule expression indicated that RANKL-induced clustering was not due to increased levels of CD2, CD1 l a, CD54 or CD58.

The addition of RANKL to CDla+ DC enhanced their allo-stimulatory capacity in a mixed lymphocyte reaction (MLR) by at least 3- to 10-fold, comparable to CD40L-cultured DC (Figure 3). Allogeneic T cells (lxlO5) were incubated with varying numbers of irradiated (2000 rad) DC cultured as indicated above for Figure 2 in 96-well round bottomed culture plates in 0.2 ml culture medium for four days. The cultures were pulsed with 0.5 mCi [3H] -thymidine for eight hours and the cells harvested onto glass fiber sheets for counting on a gas phase β counter. The background counts for either T cells or DC cultured alone were <100 cpm. Values represent the mean + SD of triplicate cultures. Heat inactivated RANKL had no effect. DC allo-stimulatory activity was not further enhanced when RANKL and CD40L were used in combination, possibly due to DC functional capacity having reached a maximal level with either cytokine alone. Neither RANKL nor CD40L enhanced the in vitro growth of DC over the three day culture period. Unlike CD40L, RANKL did not significantly increase the levels of HLA-DR expression nor the expression of CD80 or CD86.

RANKL can enhance DC cluster formation and functional capacity without modulating known molecules involved in cell adhesion (CD 18, CD54), antigen presentation (HLA-DR) or costimulation (CD86), all of which are regulated by CD40/CD40L signaling. The lack of an effect -on the expression of these molecules suggests that RANKL may regulate DC function via an alternate pathway(s) distinct from CD40/CD40L. Given that CD40L regulates RANK surface expression on in vitro-generated DC and that CD40L is upregulated on activated T cells during DC-T ceil 31 130491/2 interactions, RANK and its ligand may form an important part of the activation cascade that is induced during DC-mediated T cell expansion. Furthermore, culture of DC in RANKL results in decreased levels of CDlb/c expression, and increased levels of CD83. Both of these molecules are similarly modulated during DC maturation by CD40L (Caux et al. /. Exp. Med. 180:1263; 1994), indicating that RANKL induces DC maturation.

Dendritic cells are referred to as "professional" antigen presenting cells, and have a high capacity for sensitizing MHC-restricted T cells. There is growing interest in using dendritic cells ex vivo as tumor or infectious disease vaccine adjuvants (see, for example, Romani, et al., /. Exp. Med., 180:83, 1994). Therefore, an agent such as RANKL that induces DC maturation and enhances the ability of dendritic cells to stimulate an immune response is likely to be useful in immunotherapy of various diseases.

EXAMPLE 14 This example describes the isolation of the murine homolog of RANK, referred to as muRANK. MuRANK was isolated by a combination of cross-species PCR and colony hybridization. The conservation of Cys residues in the Cys-rich pseudorepeats of the extracellular domains of TNFR superfamily member proteins was exploited to design human RANK-based PCR primers to be used on murine first strand cDNAs from various sources. Both the sense upstream primer and the antisense downstream primer were designed to have their 3' ends terminate within Cys residues.

The upstream sense primer encoded nucleotides 272-295 of SEQ ID NO: 5 (region encoding amino acids 79-86); the downstream antisense primer encoded the complement of nucleotides 409-427 (region encoding amino acids 124-130). Standard PCR reactions were set up and run, using these primers and first strand cDNAs from various murine cell line or tissue sources. Thirty reaction cycles of 94°C for 30 seconds, 50°C for 30 seconds, and 72°C for 20 seconds were run. PCR products were anlyzed by electrophoresis, and specific bands were seen in several samples. The band from one sample was gel purified and DNA sequencing revealed that the sequence between the primers was approximately 85% identical to the corresponding human RANK nucleotide sequence.

A plasmid based cDNA library prepared from the murine fetal liver epithelium line FLE18 (one of the cell lines identified as positive in the PCR screen) was screened for full-length RANK cDNAs using murine RANK-specific oligonucleotide probes derived from the murine RANK sequence determined from sequencing the PCR product. Two cDNAs, one encoding the 5' end and one encoding the 3' end of full-length murine RANK (based on sequence comparison with the full-length human RANK) were recombined to generate a full-length murine RANK cDNA. The nucleotide and amino acid seqeunce of muRANK are shown in SEQ ID Nos:14 and 15. 32 130491/2 The cDNA encodes a predicted Type 1 transmembrane protein having 625 amino acid residues, with a predicted 30 amino acid signal sequence, a 184 amino acid extracellular domain, a 21 amino acid transmembrane domain, and a 390 amino acid cytoplasmic tail. The extracellular region of muRANK displayed significant amino acid homology (69.7% identity, 80.8% similarity) to huRANK. Those of skill in the art will recognize that the actual cleavage site can be different from that predicted by computer; accordingly, the N-terminal of RANK may be from amino acid 25 to amino acid 35.

Other members of the TNF receptor superfamily have a region of amino acids between the transmembrane domain and the ligand binding domain that is referred to as a 'spacer' region, which is not necessary for ligand binding. In muRANK, the amino acids between 197 and 214 are predicted to form such a spacer region. Accordingly, a soluble form of RANK that terminates with an amino acid in this region is expected to retain the ability to bind a ligand for RANK in a specific manner. Preferred C-terminal amino acids for soluble RANK peptides are selected from the group consisting of amino acids 214, and 197 of SEQ ID NO: 14, although other amino acids in the spacer region may be utilized as a C-terminus.

EXAMPLE 15 This example illustrates the preparation of several different soluble forms of RANK and RANKL. Standard techniques of restriction enzyme cutting and ligation, in combination with PCR-based isolation of fragments for which no convenient restriction sites existed, were used. When PCR was utilized, PCR products were sequenced to ascertain whether any mutations had been introduced; no such mutations were found.

In addition to the huRANK/Fc described in Example 2, another RANK/Fc fusion protein was prepared by ligating DNA encoding amino acids 1-213 of SEQ ID NO: 6, to DNA encoding amino acids 3-232 of the Fc mutein described previously (SEQ ID NO:8). A similar construct was prepared for murine RANK, ligating DNA encoding amino acids 1-213 of full-length murine RANK (SEQ ID NO:15) to DNA encoding amino acids 3-232 of the Fc mutein (SEQ ID NO:8).

A soluble, tagged, poly-His version of huRANKL was prepared by ligating DNA encoding the leader peptide from the immunoglobulin kappa chain (SEQ ID NO: 16) to DNA encoding a short version of the FLAG™ tag (SEQ ID NO: 17), followed by codons encoding Gly Ser, then a poly-His tag (SEQ ID NO: 18), followed by codons encoding Gly Thr Ser, and DNA encoding amino acids 138-317 of SEQ ID NO: 13. A soluble, poly-His tagged version of murine RANKL was prepared by ligating DNA encoding the CMV leader (SEQ ID NO:9) to codons encoding Arg Thr Ser, followed by DNA encoding poly-His (SEQ ID NO: 18) followed by DNA encoding amino acids 119-294 of SEQ ID NO: 11. 33 130491/2 A soluble, oligomeric form of huRANKL was prepared by ligating DNA encoding the CMV leader (SEQ ID NO:9) to a codon encoding Asp followed by DNA ending a trimer-former "leucine" zipper (SEQ ID NO: 19), then by codons encoding Thr Arg Ser followed by amino acids 138-317 of SEQ ID NO: 13.

These and other constructs are prepared by routine experimentation. The various DNAs are then inserted into a suitable expression vector, and expressed. Particularly preferred expression vectors are those which can be used in mammalian cells. For example, pDC409 and pDC304, described herein, are useful for transient expression. For stable transfection, the use of CHO cells is preferred; several useful vectors are described in USSN 08/785,150, now allowed, for example, one of the 2A5-3 λ-derived expression vectors discussed therein.

EXAMPLE 16 This example demonstrates that RANKL expression can be up-regulated on murine T cells. Cells were obtained from mesenteric lymph nodes of C57BL/6 mice, and activated with anti-CD3 coated plates, Concanavalin A (ConA) or phorbol myristate acetate in combination with ionomycin (anti-CD3: 500A2; Immunex Corporation, Seattle WA; ConA, PMA, ionomycin, Sigma, St. Louis, MO) substantially as described herein, and cultured from about 2 to 5 days. Expression of RANKL was evaluated in a three color analysis by FACS, using antibodies to the T cell markers CD4, CD8 and CD45RB, and RANK/Fc, prepared as described herein.

RANKL was not expressed on unstimulated murine T cells. T cells stimulated with either anti-CD3, ConA, or PMA/ionomycin, showed differential expression of RANKL: CD4+/CD45RBLo and CD4+/CD45RBHi cells were positive for RANKL, but CD8+ cells were not. RANKL was not observed on B cells, similar to results observed with human cells.

EXAMPLE 17 This example illustrates the effects of murine RANKL on cell proliferation and activation. Various cells or cell lines representative of cells that play a role in an immune response (murine spleen, thymus and lymphnode) were evaluated by culturing them under conditions promoting their viability, in the presence or absence of RANKL. RANKL did not stimulate any of the tested cells to proliferate. One cell line, a macrophage cell line referred to as RAW 264.7 (ATCC accession number TIB 71) exhibited some signs of activation.

RAW cells constitutively produce small amounts of TNF-a. Incubation with either human or murine RANKL enhanced production of TNF-a by these cells in a dose 34 - © £8/2§424 . PCT US97/23866' dspe dsr.t zmrs& . The results ware not due to contamination of RANKL preparations with endotoxin, since boiling RANKL for 10 minutes abrogated TNF-a production, whereas a similar treatment of purified endotoxin (LPS) did not affect the ability of the LPS to stimulate TNF-a production. Despite the fact that RANKL activated the macrophage cell line RAW T64.7 for TNF-a production, neither human RANKL nor murine RANKL stimulated nitric oxide production by these cells.

EXAMPLE 18 This example illustrates the effects of murine RANKL on growth and development of the thymus in fetal mice. Pregnant mice were injected with 1 mg of RANK Fc or vehicle control protein (murine serum albumin; MSA) on days 13, 16 and 19 of gestation. After birth, the neonates continued to be injected with RANK/Fc intraperitoneally (IP) on a daily basis, beginning at a dose of 1 g, and doubling the dose about every four days, for a final dosage of 4 g. Neonates were taken at days 1, 8 and 15 post birth, their thymuses and spleens harvested and examined for size, cellularity and phenotypic composition.

A slight reduction in thymic size at day 1 was observed in the neonates born to the female injected with RANK/Fc; a similar decrease in size was not observed in the control neonates. At day 8, thymic size and cellularity were reduced by about 50% in the RANK/Fc-treated animals as compared to MSA treated mice. Phenotypic analysis demonstrated that the relative proportions of different T cell populations in the thymus were the same in the RANK/Fc mice as the control mice, indicating that the decreased cellularity was due to a global depression in the number of thymic T cells as opposed to a decrease in a specific population(s). The RANK/Fc-treated neonates were not significantly different from the control neonates at day 15 with respect to either size, cellularity or phenotype of thymic cells. No significant differences were observed in spleen size, cellularity or composition at any of the time points evaluated. The difference in cellularity on day 8 and not on day 15 may suggest that RANK/Fc may assert its effect early in thymic development.

EXAMPLE 19 This example demonstrates that the C-terminal region of the cytoplasmic domain of RANK is important for binding of several different TRAF proteins. RANK contains at least two recognizable PXQX(X)T motifs that are likely TRAF docking sites. Accordingly, the importance of various regions of the cytoplasmic domain of RANK for TRAF binding was evaluated. A RANK/GST fusion protein was prepared substantially as described in Smith and Johnson, Gene 67:31 (1988), and used in the preparation of various truncations as described below.

Coinpsiisoa of the nucleotide sequence of murine and human RANK indicated that there re severs! conserved regions that could be important for TRAF binding. Accordingly, a PCR-based technique was developed to facilitate preparation of various C-terminal truncations that would retain the conserved regions. PCR primers were designed to introduce a stop codon and restriction enzyme site at selected points, yielding the truncations described in Table I below. Sequencing confirmed that no undesired mutations had been introduced in the constructs.

Radio-labeled (35S-Met, Cys) TRAF proteins were prepared by in vitro translation using a commercially available reticulocyte lysate kit according to manufacturer's instructions (Promega). Truncated GST fusion proteins were purified substantially as described in Smith and Johnson (supra). Briefly, E. coli were transfected with an expression vector encoding a fusion protein, and induced to express the protein. The bacteria were lysed, insoluble material removed, and the fusion protein isolated by precipitation with glutathione-coated beads (Sepahrose 4B, Pharmacia, Uppsala Sweden) The beads were washed, and incubated with various radiolabeled TRAF proteins.

After incubation and wash steps, the fusion protein TRAF complexes were removed from the beads by boiling in 0.1% SDS + β-mercaptoethanol, and loaded onto 12% SDS gels (Novex). The gels were subjected to autoradiography, and the presence or absence of radiolabeled material recorded. The results are shown in Table 2 below.

Table 2: Binding of Various TRAF Proteins to the Cytoplasmic Domain of RANK These results indicate that TRAFl, TRAF2, TRAF3, TRAF 5 and TRAF6 bind to the most distal portion of the RANK cytoplasmic domain (between amino-acid G544 and A616). TRAF6 also has a binding site between S339 and Y421. In this experiment, TRAF5 also bound the cytoplasmic domain of RANK. 36 SEQUENCE LISTING ) GENERAL INFORMATION: (i) APPLICANT: Immunex Corporation (ii) TITLE OF INVENTION: Receptor Activator of NF-kappaB (iii) NUMBER OF SEQUENCES: 19 (iv) CORRESPONDENCE ADDRESS: (A) ADDRESSEE: Immunex Corporation, Law Department (B) STREET: 51 University Street (C) CITY: Seattle (D) STATE: A (E) COUNTRY: USA (F) ZIP: 98101 (v) COMPUTER READABLE FORM: (A) MEDIU TYPE.;.. Floppy disk (B) COMPUTER: Apple Power Macintosh (C) OPERATING SYSTEM: Apple Operating System 7.5.5 (D) SOFTWARE: Microsoft Word for Power Macintosh 6.0 (vi) CURRENT APPLICATION DATA: .

(A) APPLICATION NUMBER:

Claims (31)

69 130491/6 CLAIMS:

1. An isolated DNA encoding for a RANK polypeptide or a fragment thereof selected from the group consisting of: (a) a DNA encoding a polypeptide comprising amino acids x to y of SEQ ID NO:6, wherein x is an amino terminal amino acid between amino acids 1 and 33, inclusive, of SEQ ID NO:6, and y is a carboxy terminal amino acid between amino acids 196 and 616, inclusive, of SEQ ID NO:6; (b) a DNA encoding a polypeptide comprising amino acids having at least 70% identity with amino acids x to y of SEQ ID NO:6, wherein x is an amino terminal amino acid between amino acids 1 and 33, inclusive, of SEQ ID NO:6, and y is a carboxy terminal amino acid between amino acids 196 and 213, inclusive, of SEQ ID NO:6; (c) a DNA encoding a polypeptide, the DNA being capable of hybridizing to a complement of the DNA of (a) or (b) under moderately stringent conditions or under stringent conditions, wherein moderately stringent conditions include prewashing with 5 x SSC, 0.5% SDS, 1.0 mM EDTA (pH 8.0) and hybridizing at 6 x SSC at 63°C overnight, and stringent conditions include hybridizing in 6 x SSC at 63°C overnight and washing in 3 x SSC at 55°C, and (d) a DNA encoding for a fragment of a RANK polypeptide as defined in (a), (b), or (c); wherein said RANK polypeptide or said fragment has a biological activity exhibited by a native RANK molecule, with the amino acid sequence of SEQ ID NO:6.

2. The isolated DNA of Claim 1, which encodes a RANK polypeptide that is at least 70% identical in amino acid sequence to the native form of RANK and has an activity exhibited by the native form of RANK with the amino acid sequence of SEQ ID NO:6.

3. The isolated DNA of Claim 1 or Claim 2, wherein the biological activity is selected from the group consisting ofbinding to TRAFs 1 2, 3, 5 or 6; specific binding to a soluble RANKL or a RANKL expressed on cells or immobilized on a surface; and binding to a RANK-specific antibody.

4. The isolated DNA of Claim 1 or Claim 2, which encodes a soluble RANK polypeptide. 01 184662M 12-01 70 130491/6

5. The isolated DNA of Claim 4, which further comprises a DNA encoding a polypeptide selected from the group consisting of an immunoglobulin Fc domain, an immunoglobulin Fc mutein, a FLAG™ tag, a peptide comprising at least about 6 His residues, a leucine zipper, and combinations thereof.

6. A recombinant expression vector comprising a DNA sequence according to any one of Claims f-5.

7. A host cell transformed or transfected with an expression vector according to Claim 6.

8. A process for preparing a RANK polypeptide, comprising culturing a host cell according to Claim 7 under conditions promoting expression and recovering the RANK polypeptide.

9. An isolated RANK polypeptide encoded by the DNA of Claim 1, the RANK polypeptide having a biological activity exhibited by a native RANK molecule with the amino acid sequence of SEQ ID NO: 6, and being selected from the group consisting of: (a) a polypeptide comprising amino acids 33 through 196 of SEQ ID NO:6; (b) a polypeptide comprising amino acids 1 through 213 or 33 through 213 of SEQ ID NO:6; (c) a polypeptide comprising amino acids 234 through 616 of SEQ ID NO:6; (d) a polypeptide comprising amino acids having at least 70% identity with amino acids 33 through 196 of SEQ ID NO:6; (e) a polypeptide comprising amino acids having at least 70% identity with amino acids 33 through 213 of SEQ ID NO:6; (f) a polypeptide encoded by a DNA capable of hybridizing to a complement of a DNA encoding the polypeptide of any one of (a) through (e) under moderately stringent conditions or under stringent conditions, wherein moderately stringent conditions include prewashing with 5 x SSC at 63 °C overnight, and stringent conditions include hybridizing in 6 x SSC at 63°C overnight and washing in 3 x SSC at 55°C, and the polypeptide has a biological activity exhibited by a native RANK molecule with the amino acid sequence of SEQ ID NO:6; and 01 184662M 12-01 71 130491/5 (g) a fragment of the polypeptide of any one of (a) through (f), which is biologically active.

10. The isolated RANK polypeptide of Claim 9, wherein the biological activity is selected from the group consisting of: inducing NF-κΒ activity; binding to TRAFs 1, 2, 3, 5 or 6; specific binding to a soluble RANKL or a RANKL expressed on cells or immobilized on a surface; and binding to a RANK-specific antibody.

11. The isolated RANK polypeptide of Claim 9 or Claim 10, comprising an amino acid sequence at least 80% identical or at least 90% identical to SEQ ID NO: 6.

12. The isolated RANK polypeptide of any one of Claims 9-11, which is a soluble RANK polypeptide.

13. The isolated RANK polypeptide of any one of Claims 9-12, which is a fusion polypeptide having a biological activity exhibited by a native RANK molecule with the amino acid sequence of SEQ ID NO:6.

14. The soluble RANK polypeptide of Claim 12, further comprising a polypeptide selected from the group consisting of an immunoglobulin Fc domain, an immunoglobulin Fc mutein, a FLAG™ tag, a peptide comprising at least about 6 His residues, a leucine zipper, and combinations thereof, having a biological activity exhibited by a native RANK molecule with the amino acid sequence of SEQ ID NO:6.

15. The soluble RANK polypeptide of Claim 14, which comprises the Fc portion of human IgGi consisting of amino acids 3 to 232 of SEQ ID NO:8.

16. A composition comprising an isolated RANK polypeptide as defined in any one of Claims 9-15 or as encoded by a DNA according to any one of Claims 1-5.

17. The composition of Claim 16, further comprising a physiologically acceptable carrier, excipient or diluent.

18. A RANK polypeptide as defined in any one of Claims 9-15, or as encoded by a DNA as defined in any one of Claims 1-5, or a composition as defined in Claim 16 or Claim 17, for use as a medicament.

19. The RANK polypeptide of Claim 18, which is a soluble polypeptide.

20. Use of a RANK polypeptide as defined in any one of Claims 9-15, or as encoded by a DNA according to any one of Claims 1-5, in the preparation of a medicament.

21. A soluble form of a RANK polypeptide which is derived from the extracellular domain thereof, encoded by a DNA according to any one of 01184662M 12-02 130491/7 Claims 1-5, for use as an agent in a screening assay to detect molecules capable of binding RANK.

22. A polypeptide according to Claim 21, wherein the detected molecules inhibit binding of an agonistic antibody or RANKL to said RANK polypeptide.

23. A RANK polypeptide which is derived from the intracellular domain of RANK, encoded by a DNA according to any one of Claims 1-5, for use as an agent in an assay to detect inhibition of the interaction of RANK and other intracellular proteins involved in a signal transduction.

24. A polypeptide according to Claim 23, wherein the assay is to detect inhibition of the association of RANK and TRAFs 1, 2, 3, 5 or 6.

25. A RANK polypeptide as defined in any one of Claims 9-13 or as encoded by a DNA according to any one of Claims 1-5, for use as an agent in a structure-based design of a RANK inhibitor.

26. A RANK polypeptide as defined in any one of Claims 9-13 or as encoded by a DNA according to any one of Claims 1-5, for use as an agent in a method for identifying antagonists or agonist of RANK.

27. A RANK polypeptide as defined in any one of Claims 9-13 or as encoded by a DNA according to any one of Claims 1-5, for use as an agent in a method for developing agonistic or antagonistic antibodies to RANK.

28. A RANK polypeptide as defined in any one of Claims 9-13 or as encoded by a DNA according to any one of Claims 1-5, for use as an agent in a screening method for screening for antagonists of RANK activity.

29. A kit comprising a soluble RANK polypeptide according to any one of Claims 12-15 having a biological activity exhibited by a native RANK molecule with the amino acid sequence of SEQ ID NO: 6, and instructions for use of said RANK polypeptide.

30. The kit of Claim 29, for use in detecting RANKL or for monitoring RANK-mediated activity in a patient.

31. The kit of Claim 29 or Claim 30, for use in an assay selected from the group consisting of ELISA, dot blot assay, solid phase binding assay, rapid format assay, and a bioassay. For the Applicants, REINHOLD COHN AND PARTNERS 72