Gremlin-1 inhibitor for the treatment of a bone fracture or bone defect

Abstract

The present invention relates to methods for the treatment of a bone fracture or bone defect. The invention discloses the effective use of an anti-gremlin-1 antibody to accelerate the healing and bridging of bone tissue in segmental gap defects; and demonstrates that inhibitors of gremlin-1 activity may provide improved therapies for treating or preventing fracture non-union.

Claims

1. A method for the treatment of a bone fracture or bone defect comprising administering a therapeutically effective amount of an anti-gremlin-1 antibody or functionally active fragment, variant or derivative thereof comprising a heavy chain variable region comprising SEQ ID NO: 3 or 4 for heavy chain CDR (HCDR) 1, SEQ ID NO: 5 for HCDR2, SEQ ID NO: 6 for HCDR3 and a light chain variable region comprising SEQ ID NO: 7 for light chain CDR (LCDR) 1, SEQ ID NO: 8 for LCDR2 and SEQ ID NO: 9 for LCDR3, wherein the bone fracture or bone defect is a delayed-union fracture, a non-union fracture, a bone defect with a loss of bone, or a bone disease selected from the group consisting of osteoporosis, osteogenesis imperfecta, and Paget's disease of bone.

2. The method according to claim 1, wherein the anti-gremlin-1 antibody or functionally active fragment, variant or derivative thereof comprises a heavy chain variable region comprising SEQ ID NO: 3 for heavy chain CDR (HCDR) 1, SEQ ID NO: 5 for HCDR2, SEQ ID NO: 6 for HCDR3 and a light chain variable region comprising SEQ ID NO: 7 for light chain CDR (LCDR) 1, SEQ ID NO: 8 for LCDR2 and SEQ ID NO: 9 for LCDR3.

3. The method according to claim 1, wherein the anti-gremlin-1 antibody or functionally active fragment, variant or derivative thereof comprises a heavy chain variable region comprising SEQ ID NO: 4 for heavy chain CDR (HCDR) 1, SEQ ID NO: 5 for HCDR2, SEQ ID NO: 6 for HCDR3 and a light chain variable region comprising SEQ ID NO: 7 for light chain CDR (LCDR) 1, SEQ ID NO: 8 for LCDR2 and SEQ ID NO: 9 for LCDR3.

4. The method according to claim 1, wherein the antibody comprises a heavy chain variable region (HCVR) of SEQ ID NO: 10 and/or a light chain variable region (LCVR) of SEQ ID NO: 11.

5. The method according to claim 1, wherein the heavy chain variable region comprises a sequence having at least 95% identity to the sequence of SEQ ID NO: 10 and the light chain variable region comprises a sequence having at least 95% identity to the sequence of SEQ ID NO: 11.

6. The method according to claim 1, wherein the functionally active fragment is a Fab, Fab', F(ab')2, Fv or scFv.

7. The method according to claim 1, wherein the bone fracture or bone defect is a delayed-union fracture.

8. The method according to claim 1, wherein the bone fracture or bone defect is a non-union fracture.

9. The method according to claim 1, wherein the bone fracture or bone defect is a bone defect with a loss of bone.

10. The method according to claim 1, wherein the bone fracture or bone defect is osteoporosis.

11. The method according to claim 1, wherein the bone fracture or bone defect is osteogenesis imperfecta.

12. The method according to claim 1, wherein the bone fracture or bone defect is Paget's disease of bone.

Description

FIGURES

(1) FIG. 1 shows percentage restoration of signal for the immunisation derived antibodies in the HEK-ID1 reporter gene assay.

(2) FIG. 2 shows percentage restoration of signal for library derived antibodies in the HEK-ID1 reporter gene assay.

(3) FIG. 3 shows results for the HEK-ID1 reporter gene assay with titrations of human Gremlin (FIG. 3A) and mouse Gremlin (FIG. 3B) and the effect of antibody 7326 (shown as antibody PB376) in restoring signalling of BMP.

(4) FIG. 4 shows a structural model of the Gremlin-Fab complex, with the possible BMP binding regions and the Fab epitope highlighted.

(5) FIG. 5 shows examination of the area devoid of callus/bone tissue during fracture repair in acquired X-Ray images. The area within the defect, which was devoid of tissue, was quantified using definiens image analysis and subsequently compared between control and anti-gremlin 1 treated group. Results are presented as the mean±SD of 10 rats/group. *P<0.05; **P<0.01; ***P<0.001 as measured by Mann-Whitney U test.

(6) FIG. 6 shows examination of LMB (low mineral bone; newly formed bone) and HMB (high mineral bone; mature bone) within a 3 mm femoral bone defect. Panel A: 3D μCt analysis of femoral bone defect region to detect newly formed bone or mature bone. Percentage of bone volume/tissue volume was measured and compared all subjects (total) between control vs. treatments with anti-gremlin 1. Comparisons between control versus anti-gremlin 1 treated group in animals separated as low responders (LR- incomplete bridging) and high responders (HR complete bridging) were also performed. Results are presented as the mean ±SD. *P<0.05; **P<0.01; ***P<0.001 as measured by Mann-Whitney U test. Panel B: representative μCt illustrating the 3D bone volume renderings of LR and HR groups in control and after anti-gremlin 1 treatment.

(7) FIG. 7 shows histomorphometric analysis of femoral bone defect. Percentage of bone volume/tissue volume (BV/TV ((Yip)), trabecular number (Tb.N) and trabecular separation (Tb.Sp) was compared between control versus anti-gremlin 1 treated group. Results are presented as the mean±SD of 10 rats/group. *P<0.05; **P<0.01; ***P<0.001 as measured by Mann-Whitney U test.

(8) FIG. 8 shows correlation of 3D μCT analysis and 2D histomorphometry analysis of total BV/TV %. Correlations were performed in both groups on LMB (low mineral bone; newly formed bone) and HMB (high mineral bone; mature bone) within a 3 mm femoral bone defect and compared to 2D histomorphometry score data (n=20). The Pearson's score indicates significant correlation of BV/TV % between 3D μCT analysis and 2D histomorphometry analysis.

(9) The following Examples illustrate the invention.

EXAMPLES

Example 1—Protein Expression, Purification, Refolding and Structure Determination

(10) Protein Expression and Inclusion Body Preparation

(11) A truncated human Gremlin-1 coding sequence (SEQ ID NO: 20), optimised for expression in E. coli, was cloned into a modified pET32a vector (Merck Millipore) using BamHI/Xhol, generating a vector encoding the Gremlin sequence with an N-terminal 6His-TEV tag (pET-hGremlin1).

(12) Expressed sequence:

(13) MGSSHHHHHHSSGENLYFQGSAMPGEEVLESSQEALHVTERKYLKRDWCKTQPLKQTIHEEGCNSRTIINRFCYGQCNSFYIPRHIRKEEGSFQSCSFCKPKKFTTMMVTLNCPELQPPTKKKRVTRVKQCRCISIDLD; SEQ ID NO: 2 (with non-Gremlin residues of the 6His-TEV tag shown in italics). Sequence numbering based on UniProt 060565 & SEQ ID NO: 1.

(14) The pET-hGremlin1 plasmid DNA was used to transform BL21(DE3) cells. A single ampicillin resistant colony was picked from a LB/Amp agar plate and used to inoculate a 100 ml starter culture of LB/Amp. After shaking (200 rpm) for 16 hr at 37° C., 25 ml of the starter culture was used to inoculate 500 mL of 2×TY/Amp media. The culture was shaken (250 rpm) at 37° C. until an OD.sub.600 of 3 was achieved. Subsequently, the culture was supplemented with 20 mL of a MOPS+glycerol feed mix (1M MOPS pH 7.4, 40% glycerol, 0.5% MgSO.sub.4, 0.42% MgCl.sub.2), induced with 300 μM IPTG and further incubated at 17° C., 180 rpm for 16 hours. Cells were harvested in a centrifuge (4,000 g for 20 minutes at 4° C.).

(15) Cell pellets were resuspended in Lysis Buffer (PBS pH 7.4, 0.35 mg/ml lysozyme, 10 μg/ml DNase and 3 mM MgCl.sub.2) at 4° C. and the insoluble fraction was harvested by centrifugation at 3,500 g for 30 minutes at 4° C. Pelleted inclusion bodies were washed three times by resuspending in wash buffer (50 mM Tris, 500 mM NaCl, 0.5% Triton X-100, pH 8.0), followed by centrifugation at 21,000 g for 15 minutes. An additional two washes were performed using wash buffer without Triton X-100.

(16) Solubilisation

(17) Inclusion bodies were resuspended in denaturing buffer (8 M Urea, 100 mM Tris, 1 mM EDTA, 10 mM Na.sub.2S.sub.4O.sub.6 and 100 mM Na.sub.2SO.sub.3, pH 8.5), stirred for 16 hrs at room-temperature and clarified by centrifugation at 21,000 g for 15 minutes.

(18) Pre-Refolding Purification

(19) The solubilized inclusion bodies were loaded onto a Sephacryl S-200 26/60 column (120 mL) equilibrated in 8 M Urea, 50 mM MES, 200 mM NaCl, 1 mM EDTA, pH 6.0. Fractions containing Gremlin-1 protein were diluted with 6 M Urea, 20 mM MES, pH 6.0 and loaded onto HiTrap SP HP cation exchange columns and eluted with a 1 M NaCl gradient over 10 column volumes (10 CVs). Fractions containing purified, denatured hGremlin-1 protein were pooled.

(20) Refolding

(21) Denatured purified Gremlin-1 protein was added drop-wise to re-folding buffer (50 mM Tris, pH 8.5, 150 mM NaCl, 5 mM GSH and 5 mM GSSG, 0.5 mM Cysteine, 5 mM EDTA, 0.5 M Arginine) to a final concentration of 0.1 mg/ml and incubated at 4° C. with constant stirring for 5 days. After 5 days the Gremlin-1 protein was dialysed against 20 mM HEPES, 100 mM NaCl, pH 7.5.

(22) Following dialysis protein was applied to heparin HiTrap column and eluted using a gradient of 0-100% heparin elution buffer (20 mM HEPES, 1 M NaCl, pH 7.5) over 20 CV. Correctly folded protein eluted at 1 M NaCl whereas any misfolded protein eluted at lower salt concentrations.

(23) Protein eluting at 1 M NaCl was concentrated and purified further on a S75 26/60 column equilibrated with 20 mM Hepes, pH 7.5, 1 M NaCl.

(24) Protein was characterised by SDS PAGE (shift in gel), demonstrated to have the expected molecular weight and correct arrangement of disulphide bonds using liquid chromatography mass spectrometry (LC-MS) and to be active in a cell assay (ID1 reporter assay).

(25) Gremlin 1 Structure Determination

(26) Gremlin 1 protein crystals were grown using the hanging-drop method by mixing a solution of Gremlin 1 at 6.6 mg/ml and 0.1 M citric acid at pH 4, 1 M lithium chloride and 27% polyethylene glycol (PEG) 6000 in a 1:1 ratio. Before data collection, crystals were cryo-protected by adding 20% glycerol to the crystallization buffer. Diffraction data were collected at the Diamond Light Source and were processed using XDS (Kabsch, Wolfgang (2010) Acta Crystallographica Section D 66, 125-132). Diffraction data statistics are summarized in the table below:

(27) TABLE-US-00002 TABLE 2 Diffraction data statistics Diffraction Statistics Wavelength (Å) 0.97949 Space group C2 Cell dimensions a = 84.55 Å, b = 107.22 A, c = 77.09 Å; α = 90.00°, β = 120.43°, γ = 90.00° Resolution range* (Å) 26.19-2.72 (2.79-2.72) Completeness (%) 98.5 (99.0) Multiplicity 3.4 (3.4) I/sigma 9.6 (2.0) Rmerge 0.095 (0.622) Refinement Statistics Resolution Range (Å) 26.19-2.72 R.sub.cryst 0.24 R.sub.free 0.29 R.m.s.d. bonds (Å)** 0.013 R.m.s.d. angles (°) 1.782 *values in parenthesis correspond to the highest resolution shell **r.m.s.d root mean square deviation

(28) Gremlin-1 structure was solved by molecular replacement using Phaser (McCoy et al, J Appl Cryst (2007), 40, 658-674) and a Gremlin-1 model available from proprietary Gremlin-1/Fab complex coordinates. The resultant model of Gremlin-1 contained four copies of Gremlin 1 monomer organised as two dimers. Model corrections were made with Coot (Emsley et al Acta Crystallographica Section D: Biological Crystallography 66 (4), 486-501) and coordinates were refined using Refmac (Murshudov et al REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallographica Section D: Biological Crystallography. 2011;67(Pt 4):355-367). Final coordinates were validated with Molprobity (Chen et al. (2010) MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallographica D66:12-21). A summary of model refinement statistics is shown in Table 2 above.

Example 2—BMP Binding Residues on Gremlin-1

(29) As discussed above, Gremlin-1 belongs to the bone morphogenic protein (BMP) antagonist protein family within a sub-group known as the DAN family. Within the DAN family, Gremlin-1 shares greatest homology with Gremlin-2 (PRDC).

(30) The 2.7 A human Gremlin-1 structure resolved in Example 1 shares many features in common with the published mouse Gremlin-2 structure (Nolan et al (2013), Structure, 21, 1417-1429). The overall fold is very similar, with two copies of Gremlin-1 forming an antiparallel, non-covalent dimer, arranged in an arch. Each monomer adopts the characteristic finger-wrist-finger arrangement with a cystine-knot motif towards the wrist end, opposite the fingers. Sequence identity between the proteins is 52% rising to 67% within the sequence visible in the two structures. The most highly conserved region lies in the extensive dimer interface where all the key contact residues are 100% conserved.

(31) Residues involved in BMP's 2, 4 & 7 binding to mouse Gremlin-2 (PRDC) and DAN (NBL1) have been identified using mutagenesis (Nolan et al (2013), Structure, 21, 1417-1429 and Nolan et al (2014) J. Biol. Chem. 290, 4759-4771). The predicted BMP binding epitope encompasses a hydrophobic patch spanning across both monomers on the convex surface of the dimer. Six residues were identified by mutagenesis; Trp72, Phe96, Tyr98, Phe104, Tyr105 & Phe117 and are 100% conserved in human Gremlin-1 (numbering based on the mouse Gremlin-2 sequence). The degree of homology extends to the positioning of the side chains which adopt an identical conformation in both proteins.

(32) The amino acid numbering used in the Gremlin PDB file matches the numbering in the published mouse Gremlin-2 structure based on a structural alignment. This enables like for like comparison of amino acids when describing the structures. However, for clarity the key residues identified as playing a role in BMP binding are shown below with numbering based on the PDB file and UniProt file of SEQ ID NO: 1 in brackets:

(33) Trp72(93), Phe96(117), Tyr98(119), Phe104(125), Tyr105(126) & Phe117(138).

(34) In both mouse Gremlin-2 and human Gremlin-1 the hydrophobic BMP binding epitope is partially buried by an alpha helix formed by the N-terminal residues of each protein. A model of BMP binding has been proposed whereby the N-terminus can flex, exposing the full BMP binding interface (Nolan et al (2013), Structure, 21, 1417-1429). In the present analysis, the N-terminal residues were removed from the human Gremlin-1 and mouse Gremlin-2 structures before rendering a surface to reveal the similarity of the BMP binding faces on each protein.

(35) The literature only describes mutagenesis of six resides that have an effect on BMP binding. It is possible that the actual BMP epitope covers a larger surface area, encompassing neighbouring amino acids. By highlighting all residues, within 6 Å of those mutated, on the surface of Gremlin-1, a larger region of Gremlin-1 is revealed that could potentially be targeted by a therapeutic. This more extensive region encompasses the following amino acids of human Gremlin-1: Asp92-Leu99 Arg116-His130 Ser137-Ser142 Cys176-Cys178 (Numbering based on SEQ ID NO: 1)

(36) By combining published information with the crystal structure information of human Gremlin-1, regions of human Gremlin-1 that offer themselves as a potential route for therapeutic intervention blocking its interaction with BMP's have been identified.

Example 3—Hek Id1 Reporter Gene Assay

(37) Background

(38) The Hek Id1 reporter gene assay uses Clone 12 Hek293-Id1 reporter cells. This cell line was stably transfected with Id1 transcription factor. Id1 is a transcription factor in the BMP signalling pathway. Gremlin is known to bind BMPs prevent binding to their receptors reducing the luciferase signal from the reporter gene. Therefore, using this reporter assay, it is possible to screen anti-Gremlin antibodies and see if there are any that block the interaction of Gremlin with BMPs. A restoration of the luciferase signal is seen in these cells if there is a blocking of this interaction.

(39) Method

(40) Clone 12 cells were cultured in DMEM containing 10% FCS, 1×L-Glutamine & 1×NEAA. Cells are also grown in the presence of Hygromycin B (200 μg/ml) to ensure cells do not lose Idi gene expression. Cells were assayed in DMEM containing 0.5% FCS, 1×L-Glutamine & 1×NEAA. Hygromycin B is not needed for the short time that the cells are in the assay.

(41) The cells were washed in PBS, lifted off using cell dissociation buffer, spun and counted before being seeded at 5×104/well in 70 μl (Density of 7.14×10.sup.5/ml). Plates used were white, opaque Poly-D-Lysine coated 96-well sterile. Cells go in incubator for about 3-4 hours to settle down. BMP heterodimers were reconstituted to 200 μg/ml in 4 mM HCL. BMP was diluted to 10 μg/ml in assay media using a glass vial to give a new working stock.

(42) In a polypropylene plate, Gremlin-1 was diluted 1:2 for an 8 point dose response curve with a top final dose of 1 μg/ml.

(43) An additional volume of 20 μl media was added per well and plates were incubated at 37° C. for 45 mins.

(44) BMP prepared at 100× was added to all wells except wells containing cells only. All wells are made up to 60 μl with assay medium and incubated for a further 45 mins at 37° C.

(45) Post incubation, 30 μl of sample was transferred per well of assay plate and incubated for 20-24 hours before measuring luminescence signal.

(46) Cell Steady Glo was thawed in advance at room temperature. Assay plates were cooled to room temperature for about 10-15 mins before adding the reagent. Luciferase signal was detected by addition of cell steady glo reagent (100 μl) for 20 minutes on shaker at room temperature and measuring luminescence using cell titre glo protocol on Synergy 2.

(47) The maximum signal was generated from wells containing BMP and the minimum signal was generated from the wells containing cells only.

(48) Results

(49) Gremlin-1 full length and truncated forms were tested in the Hek-Id1 reporter gene assay to confirm the blocking activity against BMP4/7.

(50) The percentage of inhibition from dose response assays was calculated based on the maximum and minimum signals in the assay and the data fitted using 4 parameter logistical fit. The IC.sub.50 was calculated based on the inflexion point of the curve.

(51) TABLE-US-00003 TABLE 3 Potency results for full length Gremlin-1 and truncated Gremlin-1 in the Hek-Id1 reporter gene assay. Hek-Id1 Reporter Geometric 95% CI (or range gene assay N mean (nM) where N = <4) Gremlin 1 Full 2 1.6 1.3-1.9 length Gremlin 1 2 1.7 1.1-2.5 truncated

(52) Conclusion

(53) Gremlin 1 was able to inhibit the BMP 4/7 signalling in the Hek-Id1 reporter gene assay.

Example 4—Production of Anti-Gremlin-1 Antibodies

(54) Anti-Gremlin-1 antibodies were derived by immunisation using purified gremlin-1 as described in Example 1, and by library panning. The library was generated in-house as a naive human library with the V-regions amplified from blood donations.

(55) Immunisation yielded 26 distinct antibodies binding Gremlin-1 from the first round of immunisation. These antibodies were scaled up and purified for testing in screening assays.

(56) 25 human and mouse cross-reactive antibodies from the library were panned using recombinant human Gremlin from R&D Systems. 10 antibodies were selected for scale up and purified as scFvs for testing in the screening assays.

Example 5—Screening of Anti-Gremlin-1 Antibodies

(57) Antibodies were screened using the Hek-Id1 reporter gene assay described in Example 3 and by measuring SMAD phosphorylation. SMAD1, 5 and 8 are phosphorylated upon BMP signalling. Inhibitors of Gremlin-1 therefore increase SMAD phosphorylation.

(58) SMAD phosphorylation assays were conducted on A549 cells or on human lung fibroblasts. Phosphorylation levels were determined using MSD.

(59) Results

(60) In the Hek-Id1 reporter gene assay, there were no apparent hits with the immunisation derived antibodies (with a 10 fold excess of antibody tested against a BMP4/7 heterodimer). Results are shown in FIG. 1.

(61) In contrast, a number of library derived antibodies were capable of restoring signal in the Hek-Id1 reporter gene assay (50-fold excess of antibodies with a 50% gremlin dose) (FIG. 2). Of these, Ab2416 and Ab2417 contained high levels of endotoxin. Ab7326 maintained blocking ability at a 10-fold excess and 80% inhibition Gremlin-1 concentration.

(62) Additional results are presented in FIGS. 3A (human gremlin) and 3B (mouse Gremlin). These Figures show titrations of Ab7326 (labelled as PB376) up to 15 nM. Ab7326 was shown to restore signalling of BMP when blocked by either human (IC.sub.50 of 1.3 nM) or mouse (IC.sub.50 of 0.2 nM Gremlin). The antibody functions both as a human and mouse IgG1.

(63) Sequences of the mouse and human full length IgG1 are presented below. In order to synthesise the mouse and human full length IgG1 proteins, the Ab7326 variable regions derived from the library were re-cloned into vectors comprising the appropriate antibody constant domains.

(64) Because Ab7326 came from a naïve human library, where Abs are cloned as scFvs, in order to re-clone the 7326 variable regions as full length Abs or Fabs, it was necessary to PCR amplify the VH and VK using pools of primers/degenerate primers. The amplified PCR products were then digested and cloned simultaneously into mouse and human vectors. As the VH and VK were amplified by pools of primers/degenerate primers, two variant forms of the products were obtained, differing by a single amino acid residue derived from subtly different primers annealing during the PCR process.

(65) The two variant forms of heavy chain variable region differed by a single amino acid at position 6, and the two variant forms of the light chain variable region differed by a single amino acid at position 7, as shown below: Heavy chain variable region variant 1 has glutamic acid (E) at position 6. Heavy chain variable region variant 2 has glutamine (Q) at position 6. Light chain variable region variant 1 has serine (S) at position 7. Light chain variable region variant 2 has threonine (T) at position 7.

(66) TABLE-US-00004 Mouse full length IgG1 - heavy chain variant 1 (SEQ ID NO: 14) QVQLVESGAE VKKPGATVKI SCKVSGYTFT DYYMHWVQQA PGKGLEWMGL VDPEDGETIY AEKFQGRVTI TADTSTDTAY MELSSLRSED TAVYYCATDA RGSGSYYPNH FDYWGQGTLV TVSSAKTTPP SVYPLAPGSA AQTNSMVTLG CLVKGYFPEP VTVTWNSGSL SSGVHTFPAV LQSDLYTLSS SVTVPSSTWP SETVTCNVAH PASSTKVDKK IVPRDCGCKP CICTVPEVSS VFIFPPKPKD VLTITLTPKV TCVVVDISKD DPEVQFSWFV DDVEVHTAQT QPREEQFNST FRSVSELPIM HQDWLNGKEF KCRVNSAAFP APIEKTISKT KGRPKAPQVY TIPPPKEQMA KDKVSLTCMI TDFFPEDITV EWQWNGQPAE NYKNTQPIMD TDGSYFVYSK LNVQKSNWEA GNTFTCSVLH EGLHNHHTEK SLSHSPGK Mouse full length IgG1 - light chain variant 1 (SEQ ID NO: 15) DIVMTQSPDS LAVSLGERAT INCKSSQSVL YSSNNKNYLA WYQQKPGQPP KLLIYWASTR ESGVPDRFSG SGSGTDFTLT INSLQAEDVA VYFCQQYYDT PTFGQGTRLE IKRTDAAPTV SIFPPSSEQL TSGGASVVCF LNNFYPKDIN VKWKIDGSER QNGVLNSWTD QDSKDSTYSM SSTLTLTKDE YERHNSYTCE ATHKTSTSPI VKSFNRNEC Mouse full length IgG1 - heavy chain variant 2 (SEQ ID NO: 28) QVQLVQSGAE VKKPGATVKI SCKVSGYTFT DYYMHWVQQA PGKGLEWMGL VDPEDGETIY AEKFQGRVTI TADTSTDTAY MELSSLRSED TAVYYCATDA RGSGSYYPNH FDYWGQGTLV TVSSAKTTPP SVYPLAPGSA AQTNSMVTLG CLVKGYFPEP VTVTWNSGSL SSGVHTFPAV LQSDLYTLSS SVTVPSSTWP SETVTCNVAH PASSTKVDKK IVPRDCGCKP CICTVPEVSS VFIFPPKPKD VLTITLTPKV TCVVVDISKD DPEVQFSWFV DDVEVHTAQT QPREEQFNST FRSVSELPIM HQDWLNGKEF KCRVNSAAFP APIEKTISKT KGRPKAPQVY TIPPPKEQMA KDKVSLTCMI TDFFPEDITV EWQWNGQPAE NYKNTQPIMD TDGSYFVYSK LNVQKSNWEA GNTFTCSVLH EGLHNHHTEK SLSHSPGK Mouse full length IgG1 - light chain variant 2 (SEQ ID NO: 29) DIVMTQTPDS LAVSLGERAT INCKSSQSVL YSSNNKNYLA WYQQKPGQPP KLLIYWASTR ESGVPDRFSG SGSGTDFTLT INSLQAEDVA VYFCQQYYDT PTFGQGTRLE IKRTDAAPTV SIFPPSSEQL TSGGASVVCF LNNFYPKDIN VKWKIDGSER QNGVLNSWTD QDSKDSTYSM SSTLTLTKDE YERHNSYTCE ATHKTSTSPI VKSFNRNEC Human full length IgG1 - heavy chain variant 1 (SEQ ID NO: 30) QVQLVESGAE VKKPGATVKI SCKVSGYTFT DYYMHWVQQA PGKGLEWMGL VDPEDGETIY AEKFQGRVTI TADTSTDTAY MELSSLRSED TAVYYCATDA RGSGSYYPNH FDYWGQGTLV TVSSASTKGP SVFPLAPSSK STSGGTAALG CLVKDYFPEP VTVSWNSGAL TSGVHTFPAV LQSSGLYSLS SVVTVPSSSL GTQTYICNVN HKPSNTKVDK KVEPKSCDKT HTCPPCPAPE LLGGPSVFLF PPKPKDTLMI SRTPEVTCVV VDVSHEDPEV KFNWYVDGVE VHNAKTKPRE EQYNSTYRVV SVLTVLHQDW LNGKEYKCKV SNKALPAPIE KTISKAKGQP REPQVYTLPP SRDELTKNQV SLTCLVKGFY PSDIAVEWES NGQPENNYKT TPPVLDSDGS FFLYSKLTVD KSRWQQGNVF SCSVMHEALH NHYTQKSLSL SPGK Human full length IgG1 - light chain variant 1 (SEQ ID NO: 31) DIVMTQSPDS LAVSLGERAT INCKSSQSVL YSSNNKNYLA WYQQKPGQPP KLLIYWASTR ESGVPDRFSG SGSGTDFTLT INSLQAEDVA VYFCQQYYDT PTFGQGTRLE IKRTVAAPSV FIFPPSDEQL KSGTASVVCL LNNFYPREAK VQWKVDNALQ SGNSQESVTE QDSKDSTYSL SSTLTLSKAD YEKHKVYACE VTHQGLSSPV TKSFNRGEC Human full length IgG1 - heavy chain variant 2 (SEQ ID NO: 16) QVQLVQSGAE VKKPGATVKI SCKVSGYTFT DYYMHWVQQA PGKGLEWMGL VDPEDGETIY AEKFQGRVTI TADTSTDTAY MELSSLRSED TAVYYCATDA RGSGSYYPNH FDYWGQGTLV TVSSASTKGP SVFPLAPSSK STSGGTAALG CLVKDYFPEP VTVSWNSGAL TSGVHTFPAV LQSSGLYSLS SVVTVPSSSL GTQTYICNVN HKPSNTKVDK KVEPKSCDKT HTCPPCPAPE LLGGPSVFLF PPKPKDTLMI SRTPEVTCVV VDVSHEDPEV KFNWYVDGVE VHNAKTKPRE EQYNSTYRVV SVLTVLHQDW LNGKEYKCKV SNKALPAPIE KTISKAKGQP REPQVYTLPP SRDELTKNQV SLTCLVKGFY PSDIAVEWES NGQPENNYKT TPPVLDSDGS FFLYSKLTVD KSRWQQGNVF SCSVMHEALH NHYTQKSLSL SPGK Human full length IgG1 - light chain variant 2 (SEQ ID NO: 17) DIVMTQTPDS LAVSLGERAT INCKSSQSVL YSSNNKNYLA WYQQKPGQPP KLLIYWASTR ESGVPDRFSG SGSGTDFTLT INSLQAEDVA VYFCQQYYDT PTFGQGTRLE IKRTVAAPSV FIFPPSDEQL KSGTASVVCL LNNFYPREAK VQWKVDNALQ SGNSQESVTE QDSKDSTYSL SSTLTLSKAD YEKHKVYACE VTHQGLSSPV TKSFNRGEC

(67) Antibody CDRs were determined using the Kabat method (highlighted in bold in the above sequences). Additional HCDR1 residues using the Chothia definition are in italics. Constant region sequences are underlined.

(68) Restoration of p-SMAD signalling with anti-Gremlin 1 antibodies is shown in Table 4 below.

(69) TABLE-US-00005 TABLE 4 Restoration of p-SMAD signalling 2417 2418 2419 2481 2482 2483 2484 7326 8427 BMP 2 109.1% +/− 58.2% +/− 32.6% +/− 40.4% +/− 35.3% +/− 43.1% +/− 104.0% +/− 107.2% +/− 51.3% +/− 50 ng/ml 2.8% 1.9% 1.4% 0.6% 0.8% 2.1% 2.7% 3.5% 1.4% BMP 4 109.6% +/− 71.3% +/− 31.7% +/− 60.1% +/− 54.4% +/− 72.5% +/− 105.2% +/− 110.0% +/− 78.2% +/− 25 ng/ml 3.0% 3.1% 1.2% 2.2% 1.3% 2.1% 3.3% 3.8% 2.5% BMP 7 111.5% +/− 99.5% +/− 53.8% +/− 64.4% +/− 52.3% +/− 66.2% +/− 105.2% +/− 108.0% +/− 72.6% +/− 200 ng/ml 3.8% 3.2% 3.4% 1.3% 1.1% 1.2% 4.3% 3.2% 2.5% BMP-2/7 119.3% +/− 78.6% +/− 50.8% +/− 53.7% +/− 47.6% +/− 56.1% +/− 120.4% +/− 128.5% +/− 62.8% +/− 50 ng/ml 2.6% 3.6% 2.7% 3.1% 1.5% 2.5% 4.4% 2.9% 2.5% BMP4/7 113.7% +/− 78.0% +/− 61.4% +/− 48.3% +/− 41.7% +/− 50.8% +/− 112.4% +/− 127.0% +/− 63.3% +/− 50 ng/ml 3.1% 4.0% 4.0% 2.1% 1.7% 1.7% 2.5% 3.1% 2.1%

(70) Results are shown as a percentage of SMAD phosphorylation by BMP alone (control BMP). Experiments were performed using lung fibroblasts from idiopathic pulmonary fibrosis patients. rhGremlin-1 and the anti-Gremlin-1 antibodies were preincubated for 45 minutes at room temperature. rhGremlin-1 and the anti-Gremlin-1 antibodies were then added with BMP to the cells for 30 minutes.

(71) Table 5 then shows further results in the SMAD phosphorylation assay, where displacement of BMP-2 or BMP4/7 from Gremlin 1-BMP complexes by anti-Gremlin-1 antibodies was investigated. Experiments were again performed using lung fibroblasts from idiopathic pulmonary fibrosis patients. rhBMP-2 or rhBMP 4/7 were preincubated with rhGremlin-1 for 1 hour at room temperature. The BMP-2-or BMP4/7-Gremlin-1 complexes were incubated with different concentrations of the anti-Gremlin-1 antibodies overnight at 4° C. Antibody concentrations represent the final concentration on the plate.

(72) TABLE-US-00006 TABLE 5 Displacement of BMP-2 or BMP4/7 from Gremlin 1-BMP complexes by anti-Gremlin-1 antibodies 81.3 μg/ml 40.6 μg/ml 20.3 μg/ml 10.2 μg/ml 5.1 μg/ml 2.55 μg/ml 1.27 μg/ml 0.63 μg/ml 2484 BMP 2 100.3% +/− 98.8% +/− 97.0% +/− 93.5% +/− 86.4% +/− 79.9% +/− 66.5% +/− 54.8% +/− 50 ng/ml 3.5% 2.7% 2.9% 2.6% 2.0% 1.9% 2.8% 0.3% 2484 BMP4/7 136.4% +/− 133.2% +/− 121.4% +/− 108.1% +/− 86.6% +/− 74.7% +/− 65.8% +/− 60.7% +/− 50 ng/ml 4.2% 1.0% 1.4% 4.9% 4.4% 2.2% 0.6% 1.5% 7326 BMP 2 103.7% +/− 101.5% +/− 99.4% +/− 103.8% +/− 100.3% +/− 103.2% +/− 102.8% +/− 97.0% +/− 50 ng/ml 1.1% 2.4% 3.8% 2.4% 2.2% 4.3% 2.8% 2.9% 7326 BMP4/7 133.7% +/− 132.3% +/− 130.3% +/− 125.6% +/− 121.4% +/− 120.9% +/− 111.1% +/− 102.0% +/− 50 ng/ml 0.8% 1.8% 4.2% 10.0% 4.2% 3.3% 2.3% 4.5%

(73) The results shown in Table 5 demonstrate that Ab7326 can displace already complexed BMP-2 or BMP4/7 from Gremlin 1-BMP complexes. Ab7326 can achieve this displacement at much lower concentrations that the comparison antibody 2484. This provides evidence that Ab7326 is an allosteric inhibitor, consistent with our finding that the binding site for Ab7326 is distal from the known BMP binding regions on gremlin-1. Thus Ab7326 is able to access the allosteric binding site even when BMP is complexed to gremlin-1, resulting in significantly improved inhibition of gremlin activity.

Example 6—Obtaining the crystal structure of Gremlin-1 in complex with the 7326 Fab

(74) The crystal structure of human Gremlin-1 in complex with Ab7326 Fab was solved at a resolution of 2.1 Å. Fab sequences are shown below:

(75) TABLE-US-00007 Heavy chain: SEQ ID NO: 18 QVQLVESGAEVKKPGATVKISCKVSGYTFTDYYMHWVQQAPGKGLEWMGL VDPEDGETIYAEKFQGRVTITADTSTDTAYMELSSLRSEDTAVYYCATDA RGSGSYYPNHFDYWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALG CLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSL GTQTYICNVNHKPSNTKVDKKVEPKSC Light chain: SEQ ID NO: 19 DIVMTQSPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPP KLLIYWASTRESGVPDRFSGSGSGTDFTLTINSLQAEDVAVYFCQQYYDT PTFGQGTRLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAK VQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEKHKVYACE VTHQGLSSPVTKSFNRGEC

(76) The CCP4 software NCONT was then used to identify all contacts at 4 Å between Gremlin-1 and the Fab. The following residues were identified: Ile131, Lys147, Lys148, Phe149, Thr150, Thr151, Arg169, Lys174 and GIn175 (numbering based on the UniProt Sequence of SEQ ID NO: 1 (numbered as Ile110, Lys126, Lys127, Phe128, Thr129, Thr130, Arg148, Lys153 and GIn154 in the structure file which matches the numbering of mouse Gremlin-2).

(77) FIG. 4 shows structural models of the Gremlin-Fab complex, with the Fab epitope residues shown relative to the BMP binding regions.

(78) Ab7326 is an inhibitory antibody which acts allosterically, i.e. it binds away from the BMP binding regions.

Example 7—Affinity Measurements for Binding of Anti-Gremlin-1 Antibody Ab7326 to Gremlin-1.

(79) Method

(80) The affinity of anti-Gremlin mlgG for human Gremlin 1 was determined by biamolecular interaction analysis using surface plasmon resonance (SPR) technology on a Biacore T200 system, GE Healthcare Bio-Sciences AB. Anti-Gremlin mlgG was captured by an immobilised anti-mouse Fc surface and Gremlin 1 was titrated over the captured mlgG. The capture ligand (affinipure F(ab′).sub.2 fragment of goat anti-mouse IgG, Fc fragment specific, 115-006-071, Jackson ImmunoResearch Inc.) was immobilised at 50 μg/ml in 10 mM NaAc, pH5.0 on flow cell 2 of a CM4 Sensor Chip via amine coupling chemistry, using 600 s activation and deactivation injections, to a level of ˜1600 response units (RU). HBS-EP+ buffer (0.01 M HEPES pH 7.4, 0.15 M NaCl, 3 mM EDTA, 0.05% Surfactant P20) was used as the running buffer with a flow rate of 10 μl/min. A reference surface was prepared on flow cell 1 by activating and deactivating the surface as for flow cell 2 but omitting the capture ligand.

(81) The assay buffer was HBS-EP+ plus an extra 150 mM NaCl to give a final NaCl concentration of 300 mM plus 1% CMD40. A 60 s injection of anti-Gremlin mlgG (at 5 μg/ml in running buffer) was passed over flow cells 1 and 2 to give a capture level of approximately 100 RU on the immobilised anti-mouse IgG, Fc surface. Recombinant human Gremlin 1 was titrated in running buffer from 5 nM (using 2-fold dilutions) and injected over flow cells 1 and 2 at a flow rate of 30 μl/min for 3 min followed by a 5 min dissociation phase. A buffer only control was also included. The surface was regenerated at a flow rate of 10 μl/min by a 60 s injection of 50 mM HCl, a 30 s injection of 5 mM NaOH and a 30 s injection of 50 mM HCl.

(82) The kinetic data was determined using Biacore T200 evaluation software.

(83) The affinity measurements were made at 25° C.

(84) Results

(85) Binding affinity, taken as the average K.sub.D value for 5 determinations, was found to be below 100 pM.

Example 8. Inhibition of Gremlin-1 Activity Accelerates Healing and Bridging in an in Vivo Model of Bone Fracture Repair

(86) 8.1. Materials and Methods

(87) Rat Fracture Model and Drug Administration

(88) Long bone segmental defect models have been widely used for the research of bone healing and regeneration (Sato et al; 2014). In the present study, a 3 mm femoral defect was created in 10-week old male rats and stabilised using an 8-hole PEEK plate (RIS. 602.105, RISystem, Switzerland). The plate was fixed to the bone with a forceps in the middle of the diaphysis, before the bone was drilled and fixed with screws. A 3 mm fracture gap was created using a 0.44 mm Gigly saw. The defect size/consistency/fixation was quality controlled by X-Ray imaging using Faxitron (MX-20-DCS, Faxitron Bioptics LLC, USA), this time point was defined as Day 0.

(89) Weekly dosing was commenced on Day 1 for a period of 8 weeks as outlined in Table 1. X-ray images were subsequently acquired during the in-life phase of the study at day 11, 25, 39 and 57 in order to assess the callus formation and the progress of healing. Definiens image analysis was utilized to quantify the area of the defect that was devoid of bone tissue in the captured X-Ray images.

(90) TABLE-US-00008 TABLE 6 Treatment Groups. Animal Dosing Group Number Treatment Dose Regimen Time 1 10 Vehicle Vehicle: Once/week Total of 1 ml, s.c. 57 days 2 10 Anti- 30 mg/kg, Once/week Gremlin-1 1 ml, s.c.

(91) Micro-CT Analysis of Fracture Healing

(92) Femora (fractured side) were scanned at 17.2 μm resolution using micro-CT (SkyScan 1076). A region of approximately 15 mm of the callus with the fracture in the centre was acquired. The scans were reconstructed using the Skyscan NRECON software (1.7.10) and then the reconstructed slices were further segmented to exclude fixator pins at a 3 mm defined region calculated from the mid-point of the femoral fracture defect.

(93) Histomorphometric analysis of fracture callus in 3D was performed by SkyScan software (v. 1.13.1). The mid-point within the 3mm femoral fracture defect was determined and slices 1.5 mm distal and proximal to the mid-point were segmented for each limb measured. Subsequently, the binarization of the reconstructed datasets and segmentation were performed following two defined thresholds, one to delineate the low mineralized callus (thus quantifying newly formed bone) and the other one to define mature bone. Further segmentation of these data was carried out on femora of animals classified as low responders based on satisfying the criteria of incomplete bridging of the femoral defect or high responders exhibiting bridging of the fracture site.

(94) Histomorphometry Analysis of Fracture

(95) Femora were fixed in 10% neutral-buffered formalin for 24 h, dehydrated and embedded in methyl methacrylate (MMA) at low temperature. 50-μm-thick sections were stained with Toluidine Blue to quantify the bone elements of the healing gap defect. Histomorphometric parameters were measured on the trabecular bone of the fracture defect site. Measurements were performed through image analysis.

(96) Statistical Analysis

(97) The results were presented as mean values±SD. Statistical analysis was performed using a two-tailed Mann Whitney U test with GraphPad Prism software unless otherwise stated.

(98) 8.2 Results

(99) Analysis of X-Ray images obtained during the in-life phase of the study indicated that the anti-gremlin-1 antibody accelerated fracture healing with the control and treated groups significantly diverging after 25 days (P<0.05). This effect was apparent for the remainder of the study (FIG. 5).

(100) Micro-CT analysis of terminal samples revealed that treatment with anti-gremlin-1 antibody (30 mg/kg/once weekly) led to an increase in newly formed bone within the fracture callus site (P=0.06).

(101) The incidence of fracture non-union in this model is approximately 60% with no intervention (Sato et al; 2014). To test whether gremlin-1 inhibition reduced the incidence of non-union development, the animals were classified as low responders (LR) and high responders (HR). Gremlin-1 inhibition resulted in a significant increase in the percentage of bone volume/tissue volume (BV/TV %) within LMB (low mineral bone; newly formed bone) (P<0.01) and HMB (high mineral bone; mature bone) (P<0.01) in the low responder group compared to controls, thus indicating progressive repair of the cohort likely to form non-union.

(102) Additionally, there was a trend (non-significant) towards increased LMB and HMB BV/TV % in the high responder group in response to anti-gremlin-1 treatment (FIG. 6A, representative images of LR and HR are shown in FIG. 6B).

(103) Two-dimensional histomorphometric analysis of bone parameters was performed on histological sections of the fracture site (FIG. 7). Treatment with anti-gremlin-1 antibody significantly increased percentage of bone volume/tissue volume (BV/TV %) (P<0.05) compared to control. Anti-gremlin-1 significantly increased trabecular number (Tb.N) (P<0.001) and significantly decreased trabecular separation (Tb.Sp) (P<0.01) indicating increased trabecular bone due to treatment with anti-gremlin-1.

(104) Correlations were performed between two-dimensional histomorphometric analysis and three-dimensional μCT analysis by comparing the LMB and HMB groups segmented in μCt analysis and the two-dimensional histomorphometry analysis of fracture sections (FIG. 8). Comparisons measured by Pearson's correlation revealed a positive and significant correlation between histomorphometry and μCT analysis in the LMB (P<0.0001) and HMB groups (P<0.0001) thus validating the data from each data-set.

(105) 8.3. Conclusion

(106) Inhibition of gremlin-1 activity using a neutralising anti-gremlin-1 antibody resulted in accelerated fracture repair, with significant differences between control and treated groups evident after 25 days (3 doses of antibody). Additionally, terminal analysis of the fracture site indicated the enhanced formation of bone tissue in the low responder animals, which otherwise would likely form non-union. Therefore, inhibition of gremlin-1 activity is a promising therapy for the prevention or treatment of non-union fractures and may be of particular value for the treatment of fractures that are prone to non-union development, for example, tibia, distal radius, femoral neck and scaphoid.

(107) TABLE-US-00009 SEQUENCE LISTING (Human Gremlin-1; Uniprot ID: O60565) SEQ ID NO: 1 MSRTAYTVGALLLLLGTLLPAAEGKKKGSQGAIPPPDKAQHNDSEQTQSPQQPGSRNRGRG QGRGTAMPGEEVLESSQEALHVTERKYLKRDWCKTQPLKQTIHEEGCNSRTIINRFCYGQC NSFYIPRHIRKEEGSFQSCSFCKPKKFTTMMVTLNCPELQPPTKKKRVTRVKQCRCISIDL D (Human truncated Gremlin-1 used in crystallography with N-terminal tag) SEQ ID NO: 2 MGSSHHHHHHSSGENLYFQGSAMPGEEVLESSQEALHVTERKYLKRDWCKTQPLKQTIHEE GCNSRTIINRFCYGQCNSFYIPRHIRKEEGSFQSCSFCKPKKFTTMMVTLNCPELQPPTKK KRVTRVKQCRCISIDLD (Ab7326 HCDR1 combined Kabat & Chothia) SEQ ID NO: 3 GYTFTDYYMH (Ab7326 HCDR1 Kabat) SEQ ID NO: 4 DYYMH (Ab7326 HCDR2 Kabat) SEQ ID NO: 5 LVDPEDGETIYAEKFQG (Ab7326 HCDR3 Kabat) SEQ ID NO: 6 DARGSGSYYPNHFDY (Ab7326 LCDR1 Kabat) SEQ ID NO: 7 KSSQSVLYSSNNKNYLA (Ab7326 LCDR2 Kabat) SEQ ID NO: 8 WASTRES (Ab7326 LCDR3 Kabat) SEQ ID NO: 9 QQYYDTPT (Ab7326 Heavy chain variable region variant 1) SEQ ID NO: 10 QVQLVESGAEVKKPGATVKISCKVSGYTFTDYYMHWVQQAPGKGLEWMGLVDPEDGETIYA EKFQGRVTITADTSTDTAYMELSSLRSEDTAVYYCATDARGSGSYYPNHFDYWGQGTLVTV SS (Ab7326 Light chain variable region variant 1) SEQ ID NO: 11 DIVMTQSPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTRE SGVPDRFSGSGSGTDFTLTINSLQAEDVAVYFCQQYYDTPTFGQGTRLEIK (Ab7326 Heavy chain variable region variant 2) SEQ ID NO: 12 QVQLVQSGAEVKKPGATVKISCKVSGYTFTDYYMHWVQQAPGKGLEWMGLVDPEDGETIYA EKFQGRVTITADTSTDTAYMELSSLRSEDTAVYYCATDARGSGSYYPNHFDYWGQGTLVTV SS (Ab7326 Light chain variable region variant 2) SEQ ID NO: 13 DIVMTQTPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTRE SGVPDRFSGSGSGTDFTLTINSLQAEDVAVYFCQQYYDTPTFGQGTRLEIK (Mouse full length IgG1 heavy chain variant 1) SEQ ID NO: 14 QVQLVESGAEVKKPGATVKISCKVSGYTFTDYYMHWVQQAPGKGLEWMGLVDPEDGETIYA EKFQGRVTITADTSTDTAYMELSSLRSEDTAVYYCATDARGSGSYYPNHFDYWGQGTLVTV SSAKTTPPSVYPLAPGSAAQTNSMVTLGCLVKGYFPEPVTVTWNSGSLSSGVHTFPAVLQS DLYTLSSSVTVPSSTWPSETVTCNVAHPASSTKVDKKIVPRDCGCKPCICTVPEVSSVFIF PPKPKDVLTITLTPKVTCVVVDISKDDPEVQFSWFVDDVEVHTAQTQPREEQFNSTFRSVS ELPIMHQDWLNGKEFKCRVNSAAFPAPIEKTISKTKGRPKAPQVYTIPPPKEQMAKDKVSL TCMITDFFPEDITVEWQWNGQPAENYKNTQPIMDTDGSYFVYSKLNVQKSNWEAGNTFTCS VLHEGLHNHHTEKSLSHSPGK (Mouse full length IgG1 light chain variant 1) SEQ ID NO: 15 DIVMTQSPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTRE SGVPDRFSGSGSGTDFTLTINSLQAEDVAVYFCQQYYDTPTFGQGTRLEIKRTDAAPTVSI FPPSSEQLTSGGASVVCFLNNFYPKDINVKWKIDGSERQNGVLNSWTDQDSKDSTYSMSST LTLTKDEYERHNSYTCEATHKTSTSPIVKSFNRNEC (Human full length IgG1 heavy chain variant 2) SEQ ID NO: 16 QVQLVQSGAEVKKPGATVKISCKVSGYTFTDYYMHWVQQAPGKGLEWMGLVDPEDGETIYA EKFQGRVTITADTSTDTAYMELSSLRSEDTAVYYCATDARGSGSYYPNHFDYWGQGTLVTV SSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQS SGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPELLGG PSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNS TYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELT KNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQG NVFSCSVMHEALHNHYTQKSLSLSPGK (Human full length IgG1 light chain variant 2) SEQ ID NO: 17 DIVMTQTPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTRE SGVPDRFSGSGSGTDFTLTINSLQAEDVAVYFCQQYYDTPTFGQGTRLEIKRTVAAPSVFI FPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSST LTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC (Fab heavy chain variant 1) SEQ ID NO: 18 QVQLVESGAEVKKPGATVKISCKVSGYTFTDYYMHWVQQAPGKGLEWMGLVDPEDGETIYA EKFQGRVTITADTSTDTAYMELSSLRSEDTAVYYCATDARGSGSYYPNHFDYWGQGTLVTV SSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQS SGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSC (Fab light chain variant 1) SEQ ID NO: 19 DIVMTQSPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTRE SGVPDRFSGSGSGTDFTLTINSLQAEDVAVYFCQQYYDTPTFGQGTRLEIKRTVAAPSVFI FPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSST LTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC (Human truncated Gremlin-1 used in crystallography without N-terminal tag) SEQ ID NO: 20 AMPGEEVLESSQEALHVTERKYLKRDWCKTQPLKQTIHEEGCNSRTIINRFCYGQCNSFYI PRHIRKEEGSFQSCSFCKPKKFTTMMVTLNCPELQPPTKKKRVTRVKQCRCISIDLD (Mature Gremlin-1 sequence of SEQ ID NO: 1 lacking the signal peptide of amino acids 1-21) SEQ ID NO: 21 KKKGSQGAIPPPDKAQHNDSEQTQSPQQPGSRNRGRGQGRGTAMPGEEVLESSQEALHVTE RKYLKRDWCKTQPLKQTIHEEGCNSRTIINRFCYGQCNSFYIPRHIRKEEGSFQSCSFCKP KKFTTMMVTLNCPELQPPTKKKRVTRVKQCRCISIDLD (Human IgG4P heavy chain variant 1) SEQ ID NO: 22 QVQLVESGAEVKKPGATVKISCKVSGYTFTDYYMHWVQQAPGKGLEWMGLVDPEDGETIYA EKFQGRVTITADTSTDTAYMELSSLRSEDTAVYYCATDARGSGSYYPNHFDYWGQGTLVTV SSASTKGPSVFPLAPCSRSTSESTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQS SGLYSLSSVVTVPSSSLGTKTYTCNVDHKPSNTKVDKRVESKYGPPCPPCPAPEFLGGPSV FLFPPKPKDTLMISRTPEVTCVVVDVSQEDPEVQFNWYVDGVEVHNAKTKPREEQFNSTYR VVSVLTVLHQDWLNGKEYKCKVSNKGLPSSIEKTISKAKGQPREPQVYTLPPSQEEMTKNQ VSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSRLTVDKSRWQEGNVF SCSVMHEALHNHYTQKSLSLSLGK (Human IgG4P light chain variant 1) SEQ ID NO: 23 DIVMTQSPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTRE SGVPDRFSGSGSGTDFTLTINSLQAEDVAVYFCQQYYDTPTFGQGTRLEIKRTVAAPSVFI FPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSST LTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC (Human IgG1 heavy chain DNA variant 1) SEQ ID NO: 24 caagtgcaactggtggaatccggggccgaagtgaaaaagcccggagccactgtgaagatct cttgcaaagtgtccggctacaccttcaccgactattacatgcactgggtccagcaggcacc tgggaagggccttgagtggatgggtctggtcgatcccgaggacggcgaaactatctacgcc gagaagttccagggtcgcgtcaccatcaccgccgacacttccaccgacaccgcgtacatgg agctgtccagcttgaggtccgaggacacagccgtgtactactgcgccacggatgctcgggg aagcggcagctactacccgaaccacttcgactactggggacagggcactctcgtgactgtc tcgagcgcttctacaaagggcccctccgtgttcccgctcgctccatcatcgaagtctacca gcggaggcactgcggctctcggttgcctcgtgaaggactacttcccggagccggtgaccgt gtcgtggaacagcggagccctgaccagcggggtgcacacctttccggccgtcttgcagtca agcggcctttactccctgtcatcagtggtgactgtcccgtccagctcattgggaacccaaa cctacatctgcaatgtgaatcacaaacctagcaacaccaaggttgacaagaaagtcgagcc caaatcgtgtgacaagactcacacttgtccgccgtgcccggcacccgaactgctgggaggt cccagcgtctttctgttccctccaaagccgaaagacacgctgatgatctcccgcaccccgg aggtcacttgcgtggtcgtggacgtgtcacatgaggacccagaggtgaagttcaattggta cgtggatggcgtcgaagtccacaatgccaaaactaagcccagagaagaacagtacaattcg acctaccgcgtcgtgtccgtgctcacggtgttgcatcaggattggctgaacgggaaggaat acaagtgcaaagtgtccaacaaggcgctgccggcaccgatcgagaaaactatctccaaagc gaagggacagcctagggaacctcaagtctacacgctgccaccatcacgggatgaactgact aagaatcaagtctcactgacttgtctggtgaaggggttttaccctagcgacattgccgtgg agtgggaatccaacggccagccagagaacaactacaagactacccctccagtgctcgactc ggatggatcgttcttcctttactcgaagctcaccgtggataagtcccggtggcagcaggga aacgtgttctcctgctcggtgatgcatgaagccctccataaccactatacccaaaagtcgc tgtccctgtcgccgggaaag (Human IgG1 light chain DNA variant 1) SEQ ID NO: 25 gacattgtgatgacccagtcccccgattcgcttgcggtgtccctgggagaacgggccacca ttaactgcaagagctcacagtccgtcctgtattcatcgaacaacaagaattacctcgcatg gtatcagcagaagcctggacagcctcccaagctgctcatctactgggctagcacccgcgaa tccggggtgccggatagattctccggatcgggttcgggcactgacttcactctgactatca actcactgcaagccgaggatgtcgcggtgtacttctgtcagcagtactacgacaccccgac ctttggacaaggcaccagactggagattaagcgtacggtggccgctccctccgtgttcatc ttcccaccctccgacgagcagctgaagtccggcaccgcctccgtcgtgtgcctgctgaaca acttctacccccgcgaggccaaggtgcagtggaaggtggacaacgccctgcagtccggcaa ctcccaggaatccgtcaccgagcaggactccaaggacagcacctactccctgtcctccacc ctgaccctgtccaaggccgactacgagaagcacaaggtgtacgcctgcgaagtgacccacc agggcctgtccagccccgtgaccaagtccttcaaccggggcgagtgc (Human IgG4P heavy chain DNA variant 1) SEQ ID NO: 26 caagtgcaactggtggaatccggggccgaagtgaaaaagcccggagccactgtgaagatct cttgcaaagtgtccggctacaccttcaccgactattacatgcactgggtccagcaggcacc tgggaagggccttgagtggatgggtctggtcgatcccgaggacggcgaaactatctacgcc gagaagttccagggtcgcgtcaccatcaccgccgacacttccaccgacaccgcgtacatgg agctgtccagcttgaggtccgaggacacagccgtgtactactgcgccacggatgctcgggg aagcggcagctactacccgaaccacttcgactactggggacagggcactctcgtgactgtc tcgagcgcttctacaaagggcccctccgtgttccctctggccccttgctcccggtccacct ccgagtctaccgccgctctgggctgcctggtcaaggactacttccccgagcccgtgacagt gtcctggaactctggcgccctgacctccggcgtgcacaccttccctgccgtgctgcagtcc tccggcctgtactccctgtcctccgtcgtgaccgtgccctcctccagcctgggcaccaaga cctacacctgtaacgtggaccacaagccctccaacaccaaggtggacaagcgggtggaatc taagtacggccctccctgccccccctgccctgcccctgaatttctgggcggaccttccgtg ttcctgttccccccaaagcccaaggacaccctgatgatctcccggacccccgaagtgacct gcgtggtggtggacgtgtcccaggaagatcccgaggtccagttcaattggtacgtggacgg cgtggaagtgcacaatgccaagaccaagcccagagaggaacagttcaactccacctaccgg gtggtgtccgtgctgaccgtgctgcaccaggactggctgaacggcaaagagtacaagtgca aggtgtccaacaagggcctgccctccagcatcgaaaagaccatctccaaggccaagggcca gccccgcgagccccaggtgtacaccctgccccctagccaggaagagatgaccaagaaccag gtgtccctgacctgtctggtcaagggcttctacccctccgacattgccgtggaatgggagt ccaacggccagcccgagaacaactacaagaccaccccccctgtgctggacagcgacggctc cttcttcctgtactctcggctgaccgtggacaagtcccggtggcaggaaggcaacgtcttc tcctgctccgtgatgcacgaggccctgcacaaccactacacccagaagtccctgtccctga gcctgggcaag (Human IgG4P light chain DNA variant 1) SEQ ID NO: 27 gacattgtgatgacccagtcccccgattcgcttgcggtgtccctgggagaacgggccacca ttaactgcaagagctcacagtccgtcctgtattcatcgaacaacaagaattacctcgcatg gtatcagcagaagcctggacagcctcccaagctgctcatctactgggctagcacccgcgaa tccggggtgccggatagattctccggatcgggttcgggcactgacttcactctgactatca actcactgcaagccgaggatgtcgcggtgtacttctgtcagcagtactacgacaccccgac ctttggacaaggcaccagactggagattaagcgtacggtggccgctccctccgtgttcatc ttcccaccctccgacgagcagctgaagtccggcaccgcctccgtcgtgtgcctgctgaaca acttctacccccgcgaggccaaggtgcagtggaaggtggacaacgccctgcagtccggcaa ctcccaggaatccgtcaccgagcaggactccaaggacagcacctactccctgtcctccacc ctgaccctgtccaaggccgactacgagaagcacaaggtgtacgcctgcgaagtgacccacc agggcctgtccagccccgtgaccaagtccttcaaccggggcgagtgc (Mouse full length IgG1 heavy chain variant 2) SEQ ID NO: 28 QVQLVQSGAEVKKPGATVKISCKVSGYTFTDYYMHWVQQAPGKGLEWMGLVDPEDGETIYA EKFQGRVTITADTSTDTAYMELSSLRSEDTAVYYCATDARGSGSYYPNHFDYWGQGTLVTV SSAKTTPPSVYPLAPGSAAQTNSMVTLGCLVKGYFPEPVTVTWNSGSLSSGVHTFPAVLQS DLYTLSSSVTVPSSTWPSETVTCNVAHPASSTKVDKKIVPRDCGCKPCICTVPEVSSVFIF PPKPKDVLTITLTPKVTCVVVDISKDDPEVQFSWFVDDVEVHTAQTQPREEQFNSTFRSVS ELPIMHQDWLNGKEFKCRVNSAAFPAPIEKTISKTKGRPKAPQVYTIPPPKEQMAKDKVSL TCMITDFFPEDITVEWQWNGQPAENYKNTQPIMDTDGSYFVYSKLNVQKSNWEAGNTFTCS VLHEGLHNHHTEKSLSHSPGK (Mouse full length IgG1 light chain variant 2) SEQ ID NO: 29 DIVMTQTPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTRE SGVPDRFSGSGSGTDFTLTINSLQAEDVAVYFCQQYYDTPTFGQGTRLEIKRTDAAPTVSI FPPSSEQLTSGGASVVCFLNNFYPKDINVKWKIDGSERQNGVLNSWTDQDSKDSTYSMSST LTLTKDEYERHNSYTCEATHKTSTSPIVKSFNRNEC (Human full length IgG1 heavy chain variant 1) SEQ ID NO: 30 QVQLVESGAEVKKPGATVKISCKVSGYTFTDYYMHWVQQAPGKGLEWMGLVDPEDGETIYA EKFQGRVTITADTSTDTAYMELSSLRSEDTAVYYCATDARGSGSYYPNHFDYWGQGTLVTV SSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQS SGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPELLGG PSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNS TYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELT KNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQG NVFSCSVMHEALHNHYTQKSLSLSPGK (Human full length IgG1 light chain variant 1) SEQ ID NO: 31 DIVMTQSPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTRE SGVPDRFSGSGSGTDFTLTINSLQAEDVAVYFCQQYYDTPTFGQGTRLEIKRTVAAPSVFI FPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSST LTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC (Fab heavy chain variant 2) SEQ ID NO: 32 QVQLVQSGAEVKKPGATVKISCKVSGYTFTDYYMHWVQQAPGKGLEWMGLVDPEDGETIYA EKFQGRVTITADTSTDTAYMELSSLRSEDTAVYYCATDARGSGSYYPNHFDYWGQGTLVTV SSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQS SGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSC (Fab light chain variant 2) SEQ ID NO: 33 DIVMTQTPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTRE SGVPDRFSGSGSGTDFTLTINSLQAEDVAVYFCQQYYDTPTFGQGTRLEIKRTVAAPSVFI FPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSST LTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC (Human IgG4P heavy chain variant 2) SEQ ID NO: 34 QVQLVQSGAEVKKPGATVKISCKVSGYTFTDYYMHWVQQAPGKGLEWMGLVDPEDGETIYA EKFQGRVTITADTSTDTAYMELSSLRSEDTAVYYCATDARGSGSYYPNHFDYWGQGTLVTV SSASTKGPSVFPLAPCSRSTSESTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQS SGLYSLSSVVTVPSSSLGTKTYTCNVDHKPSNTKVDKRVESKYGPPCPPCPAPEFLGGPSV FLFPPKPKDTLMISRTPEVTCVVVDVSQEDPEVQFNWYVDGVEVHNAKTKPREEQFNSTYR VVSVLTVLHQDWLNGKEYKCKVSNKGLPSSIEKTISKAKGQPREPQVYTLPPSQEEMTKNQ VSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSRLTVDKSRWQEGNVF SCSVMHEALHNHYTQKSLSLSLGK (Human IgG4P light chain variant 2) SEQ ID NO: 35 DIVMTQTPDSLAVSLGERATINCKSSQSVLYSSNNKNYLAWYQQKPGQPPKLLIYWASTRE SGVPDRFSGSGSGTDFTLTINSLQAEDVAVYFCQQYYDTPTFGQGTRLEIKRTVAAPSVFI FPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSST LTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

REFERENCES

(108) Bostrom, M. P. & Seigerman, D. A. (2005), HSS journal: The Musculoskeletal Journal of Hospital for Special Surgery 1, 9-18. The clinical use of allografts, demineralized bone matrices, synthetic bone graft substitutes and osteoinductive growth factors: a survey study.

(109) Buza, J. A., 3rd & Einhorn, T. (2016), Clinical cases in mineral and bone metabolism: The Official Journal of the Italian Society of Osteoporosis, Mineral Metabolism, and Skeletal Diseases 13, 101-105. Bone healing in 2016.

(110) Canalis, E., Parker, K. & Zanotti, S. (2012), J. Cell Physiol 227, 269-277. Gremlin1 is required for skeletal development and postnatal skeletal homeostasis.

(111) Cho, T. J., Gerstenfeld, L. C. & Einhorn, T. A. (2002), Journal of Bone and Mineral Research: The Official Journal of the American Society for Bone and Mineral Research 17, 513-520. Differential temporal expression of members of the transforming growth factor beta superfamily during murine fracture healing.

(112) Einhorn, T. A. & Gerstenfeld, L. C. (2015), Nat. Rev. Rheumatol. 11, 45-54. Fracture healing: mechanisms and interventions.

(113) Ferguson, C., Alpern, E., Miclau, T. & Helms, J. A. (1999), Mechanisms of development 87, 57-66. Does adult fracture repair recapitulate embryonic skeletal formation?

(114) Gazzerro, E. et al. (2005), Endocrinology 146, 655-665. Skeletal overexpression of gremlin impairs bone formation and causes osteopenia.

(115) Gazzerro, E. et al. (2007), J. Biol.Chem. 282, 31549-31557. Conditional deletion of gremlin causes a transient increase in bone formation and bone mass.

(116) Goulet, J. A., Senunas, L. E., DeSilva, G. L. & Greenfield, M. L. (1997), Clinical Orthopaedics and Related Research, 76-81. Autogenous iliac crest bone graft. Complications and functional assessment.

(117) Hsu, D. R., Economides, A. N., Wang, X., Eimon, P. M. & Harland, R. M. (1998), Mol. Cell 1, 673-683. The Xenopus dorsalizing factor Gremlin identifies a novel family of secreted proteins that antagonize BMP activities.

(118) Sato, K., Watanabe, Y., Harada, N., Abe, S., Matsushita, T., Yamanaka, K., Kaneko, T., & Sakai, Y. (2014) Tissue Eng Part C. Methods 20, 1037-1041.

(119) Schmid, G. J., Kobayashi, C., Sandell, L. J. & Ornitz, D. M. (2009), Developmental Dynamics: An Official Publication of the American Association of Anatomists 238, 766-774. Fibroblast growth factor expression during skeletal fracture healing in mice.

(120) Yu, Y. Y. et al. (2010), Bone 46, 841-851. Immunolocalization of BMPs, BMP antagonists, receptors, and effectors during fracture repair.