WO2011133756A1 - Monoclonal antibodies which specifically bind and neutralize human pdgf-dd - Google Patents

Abstract

The invention relates to monoclonal antibodies, which bind specifically to full length human PDGF-DD, and inhibit its processing to the active form of the molecule. Also described are the hybridomas, which produce these monoclonal antibodies, methods for identifying them, and methods for using them to inhibit the processing event.

Description

MONOCLONAL ANTIBODIES WHICH SPECIFICALLY BIND AND

NEUTRALIZE HUMAN PDGF-DD

RELATED APPLICATIONS

[01] This application claims priority from U.S. Provisional Application No. 61/342,977, filed April 22, 2010, incorporated by reference in its entirety.

FIELD OF THE INVENTION

[02] This invention relates to antibodies, monoclonal antibodies in particular, which bind to and neutralize full length human PDGF-DD, have no cross reactivity with other species of PDGF-DD, and do not bind with truncated forms of the molecule.

BACKGROUND OF THE INVENTION

[03] Platelet derived growth factor DD or "PDGF-DD," is the most recently discovered member of the PDGF/VEGF family of growth factors. See, e.g., Bergsten, et al,, Nature Cell Biol., 3:512-516 (2001); U.S. Patent Nos. 6,706,687; 7,148,037; and 7,476,654, all of which are incorporated by reference. As with all members of the PDGF family, the molecule is a dimer, and is secreted as an extracellular ligand, It acts via two, closely related tyrosine kinase receptors: PDGFRa and PDGFRp. PDGF-DD shares a feature with PDGF-CC not shared with the other members of the family, which is the need for proteolytic activation by a plasminogen activator. See, e.g., Ustach, et al., Mol. Cell Biol.. 25(14):6279-88 (2005); Fredriksson, et al., EMBO J., 23(19):3793-3802 (2004). Fredriksson, et al., J. Biol. Chem., 280(29):26856-26862 (2005), have characterized the structural requirements for interaction between PDGF-CC and t-PA, and also those required for interaction of PDGF-DD and uPA. See, Ehnmann, et al., Oncogene, 28(4):534-44 (2009), incorporated by reference in its entirety.

[04] An additional unique feature of these molecules is their multi domain structure. Full length PDGFs, including PDGF-DD, contains a VEGF/PDGF homology domain, referred to as "VHD" with a "CUB" domain positioned in front of the VHD. Urokinase plasminogen activator (uPA) activates full length PDGF-DD. [05] The amino acid sequence or primary structure of human PDGF-DD is well known as per, e.g., SEQ ID NO: 2 of U.S. Patent No. 6,630,142 to Hart, et al., the disclosure of which is incorporated by reference. Hart, et al. also delineate the various domains of the molecule. Specifically, with reference to SEQ ID NO: 2 of the patent, residues 1-18 constitute a secretory peptide. The CUB domain consists of amino acids 52-179. Hart refers to a "propeptide like sequence," region herein, as starting at amino acid 180 and extending to anywhere from amino acid 247 to amino acid 257. This is also called the "hinge" region. The growth factor domain or "GFD" also referred to as the "VHD" region, consists of amino acids 258-370, which is the C terminus of the molecule.

[06] Various mAbs to PDGF-DD are known. Hart, et al., e.g., describes production of mAbs, which were produced using, as immunogens, peptides derived from the CUB, hinge, GFD, and VHD domains. See, e.g., Example 6. Such mAbs would be expected to bind to full length PDGF-DD, as well as to any trunctated form of the molecule, which contains the region from which the immunogen derives. Hart et al. do not speak of the ability of the mAbs to neutralize PDGF-DD in any way.

[07] Several studies have implicated PDGF-DD in tumorigenesis, progressive renal disease, and fibrosis. Hence, it would be useful to have a tool available which inhibits or neutralizes the activation of PDGF-DD, to serve as a useful agent in alleviating these conditions. Such is the subject of the invention, as set forth in the disclosure which follows.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS EXAMPLE 1

[08] These experiments describe the production of PDGF-DD specific monoclonal antibodies.

[09] Recombinant, full length ("fl" hereafter) human PDGF-DD was used as an immunogen, and was produced following Bergsten, et al., Nature Cell Biol, 3:512-516 (2001), incorporated by reference. A His₆ tag was added to the C-terminal end of the fl molecule, in order to facilitate its purification, also as described by Bergsten, et al., supra.

[10] Once the fl protein was purified, subject animals (mice) received injections of 30 μg protein emulsified in FCA per mouse. The mice received booster injections every 3 weeks, using the same amount of emulsified fl PDGF-DD. Immunizations continued until serum titers of antibody were deemed sufficiently high.

[11] Animals with sufficiently high serum titers were selected, and splenocytes were removed, and fused with SP2/0 myeloma cells, using polyethylene glycol (PEG). The resulting hybridomas were selected in standard fashion, using HAT containing HM20 medium which had been supplemented with 20% fetal calf serum, 50 Mg/ml gentamycin, 200 mM L-glutamate, and 10% of hybridoma cloning factor, at 37°C in a 5% CO2 atmosphere.

[12] Following selection, the hybridomas were screened for antibody production, To elaborate, ELISA plates were coated with 1 μg/ml fl human PDGF-DD in 50 μΐ of 100 mM NaHCC>3, at 4°C, overnight, The plates were then washed, three times, with 1 X PBS / 0,1% Tween 20 followed by blocking with 3% BSA/PBST, at room temperature, for one hour. Supernatant from the hybridomas was then applied to the plates, and incubated for 1 hour, at room temperature.

[13] After the incubation and washing, 50 μΐ of goat anti-mouse Ig alkaline phosphatase conjugate which had been diluted 1 :2500 in 3% bovine serum albumin, was applied to the plates for 1 hour, also at room temperature. Any bound alkaline phosphatase was determined using the substrate pNPP, using an Elisa plate reader. The reading was at OD405_nm.

[14] Selected hybridomas were subjected to Western Blot analysis, in accordance with Li, et al., Nature Cell Biol., 2:302-309 (2000), incorporated by reference. Any hybridomas which were positive in both the ELISA and Western blotting assays were selected, grown in the medium described, supra, sub-cloned, and expanded. [15] The mAbs were purified via expanding the cultures and reducing serum content to a minimum. Other steps were also taken to avoid unspeciflc IgG contamination,

[16] Purification was carried out using HiTrap-Protein G polymers. The samples of medium were passed through columns;, and unspecific binding molecules were washed away using PBS buffer, Antibodies were eluted using 0.1 M glycine in HC1 buffer at pH3.0. The recovered mAbs were dialyzed against PBS, and used in the examples which follow. Prior to use, the hybridomas were frozen, and then thawed in Dulbecco's modified Eagle Medium, 4.5 g/ml glucose, supplemented with 20% fetal bovine serum, 2mM L-glutamine, 1 mg/ml gentamicin, and 10% hybridoma cloning factor. Following subcloning and the gradual reduction of both the fetal bovine serum and the hybridoma cloning factor, the clones of interest were maintained in 8% fetal bovine serum, without the cloning factor.

EXAMPLE 2

[17] These experiments were carried out to determine if the mAbs described supra were cross reactive with other materials. The other materials were the very closely related molecule human PDGF-CC and the His₆ tag referred to supra.

[18] Immunoblotting was carried out, under reducing conditions using 50 ng of recombinantly fi human PDGF-DD per lane. The control molecules (PDGF-CC and His₆) were used under the same conditions. The immunoblotting methodology employed was standard and known to the art, employing a one step Western blotting assay at a 1:5 dilution.

[19] Two clones, referred to hereafter as clone 1 and clone 6, showed the strongest signals, while two others, i.e., clones 10 and 12, also appeared to have promise. None of the fourteen hybridomas tested produced mAbs, which were cross reactive with human PDGF-CC or His₆. [20] In subsequent experiments, the clones were tested for their ability to detect murine fl PDGF-DD, as well as truncated, growth factor domain ("GFD") PDGF for mice and humans. The assays were carried out under reduced and non-reduced conditions.

[21] None of the mAbs recognized either form of the murine proteins, nor did they recognize the GFD form of human PDGF-DD, Clone 6 showed the strongest reactivity with non-reduced protein, but the other three species mentioned, i.e., clones 1, 10, and 12 were also reactive with the non-reduced protein. Hence, the mAbs do not bind to an epitope involving amino acids 258-370.

EXAMPLE 3

[22] The hybridoma supernatants were then tested in an ELISA. Recombinant fl human PDGF-DD and PDGF-DD (GFD) were diluted in 100 mM HC0₃ coating buffer, were placed in EIA/RIA plates, and incubated overnight at 4°C.

[23] Hybridoma supernatants were then diluted, 1:2, in PBS/1% bovine serum albumin ("BSA") and added to the plates, followed by incubation for 1.5 - 2 hours, at room temperature, followed by washing with PBS. Controls were established, using a rabbit IgG purified polyclonal antibody.

[24] Any binding was determined by using either an anti-mouse, or anti-rabbit IgG, coupled to alkaline phosphatase, in the same manner discussed supra.

[25] All of the supernatants tested recognized fl human PDGF-DD, and none recognized the human PDGF-DD (GFD) form, confirming the results of the prior experiments.

EXAMPLE 4

[26] The interaction between PDGF-DD and urokinase plasminogen activator (uPA), has been referred to, supra. Ehmann et al., supra show that the interaction between full length human PDGF-DD and uPA occurs at around amino acids 247-249. These experiments were designed to begin study of the interaction of PDGF-DD and high molecular weight uPA ("HMWuPA") in the presence of the described antibodies.

[27] A 20 μl sample of each hybridoma supernatant was diluted, in 130 μΐ of serum free DMEM, followed by incubation with a Protein A sepharose slurry, for 2 hours, at 4°C. This resulted in antibodies coupled to beads, which were then washed with 20 mM Tris-HCl, at pH 7.5. These samples were then added to 1 ml of conditioned medium which had been collected from COS-1 cells that had been transfected with a vector encoding fl human PDGF-DD, After 1 hour of incubation at 4°C, 10 μg of HMWuPA were added, and the mixture was incubated, for 2.25 hours at room temperature. The beads were then collected via mild centrifugation, washed with 20 mM Tris-HCl, pH 7,5, and then medium was TCA precipitated, This resulted in a first sample, i.e., the Protein-A sepharose beads, and the medium.

[28] These two samples were analyzed via immunoblotting, as described supra.

[29] The results indicated that the aforementioned clones 1 and 6 produced high affinity mAbs against fl human PDGF-DD. This was evident from the strength of the signal produced in the Sepharose- A bead samples following immunoblotting. Processed PDGF-DD which had been processed by the HMWuPA was evident in the precipitated samples.

EXAMPLE 5

[30] Follow up experiments were carried out to ascertain the ability of the mAbs to bind to fixed proteins.

[31] COS-1 cells were transfected with known vectors which express one of fl PDGF-DD, fl PDGF-CC, PDGF-D (GFD), (all human) or with an empty, or "mock" vector, and were cultured in medium. After one day, the medium was exchanged for serum free medium to which 0,5 Brifeldin was added, in order to inhibit the secretion of proteins. [32] Thirty hours after transfection, the cells were washed in cold PBS, detached using 1 mM PBS-EDTA, and fixed in single cell suspension, using 4% paraformaldehyde, for at lest one hour.

[33] The cells were subjected to mild cenrrifugaiion, washed, twice, for 15 minutes each time, in cold PBS, and then dehydrated in 70% ethanol, 95% ethanol, and absolute alcohol Two, fifteen minute treatments at each concentration were used, Finally, cells were incubated in xylene, and then fixed in 3 ml of histowax at 60°C, overnight. A parallel set of experiments changed the xylene to isopropanol,

[34] Sectioning was carried out following cooling to room temperature, and staining was carried out. To elaborate, cells were re-hydrated, washed in PBS, and treated to quench endogenous PBS with 3% H₂O₂ in PBS -Tween, for 10 minutes. TNT buffer (0.1 M Tris-HCl, pH 7.5; 0.15 M NaCl, 0.05% Tween-20), was used for subsequent washes and TNB buffer (0.1M Tris-HCl, pH 7.5, 0.15 M NaCL 0.5% blocking agent) for blocking following the manufacturer's instructions.

[35] The fixed cells were incubated with hybridoma supernatants, at a 1:20 dilution, and a biotinylated anti-mouse IgG was used for detection, using standard avidin/biotin methods, with single developed using 3,3'-diaminobenzidine.

[36] Both of clones 1 and 6 proved to produce mAbs which stained fl PDGF- DD in fixed materials, as did clones 2 and 7. None of these clones recognized other tested materials.

EXAMPLE 6

[37] In further experiments, the ability of the mAbs to stain proteins via immunofluorescence was tested.

[38] Transfected COS-1 cells were used to produce recombinant fl human and murine PDGF-D and PDGF-D (GFD), and also human PDGF-D (CUB), as described by Ehnmann, et al., Oncogene, 28(4):534-44 (2009), incoiporated by reference. Cells were recovered and, after substitution of serum free DMEM, cells were fixed 40 hours after transfection, in 3% PFA. A 10 mM glycine solution was used for quenching, and permeabilization was carried out in 0.1% BSA/PBS. Incubation was carried out using hybridoma supernatants that had been diluted at 1:6, for 30 minutes at room temperature. An Alexa 488 conjugated anti-mouse IgG was used, in accordance with standard methods. As with the preceding experiments, a negative control was used as well.

[39] Clones 1 and 6 exhibited strong specific staining for full length human PDGF-DD, while only one other clone (clone 13), showed minimum binding with one species (PDGF-D (CUB)). This establishes a lack of binding with amino acids 52-179, i.e., the CUB region.

EXAMPLE 7

[40] In order to determine which antibodies should be selected for purification, 25 μΐ of the supernatant for each hybridoma culture was analyzed for IgG expression via immunoblotting, using a commercially available, one-step Western blotting kit, under reducing conditions.

[41] To cany out these assays, 350 ml of supernatant was diluted in 150 ml binding buffer (20 mM sodium phosphate, pH 7.0), and was then passed through a standard Western column. The same binding buffer was used for equilibration and washing.

[42] Bound IgG was eluted with 0.1 M glycine-HCl, pH 2.7. In order to preserve the activity of acid labile IgG, 150 μΐ of 1 M Tris-HCl, pH 9.0, was added to all collection vials beforehand, so as to ensure neutralization. Absorbance was monitored at 280 nm and fractions with a value > 0.5 were saved, pooled, and diluted, in order to allow for accurate protein concentration measurements.

[43] Buffer exchange to PBS was made overnight at 4°C, via dialysis, with final purity being determined by separation of 3 μg of sample on an SDS gel, followed by Coomassie staining. EXAMPLE 8

[44] It is well known that the chromosomes of hybridomas are unstable. For this reason, the most promising hybridomas were thawed and subcloned as soon as possible.

[45] Single cell colonies required at least 10 days before collection of cell culture supernatants and screening for immunoreactivity. When secretion of appropriate antibodies was confirmed, three subclones from each hybridoma were saved for further expansion, in cell culture flasks. Concentrations of both HCF and FBS were reduced gradually and eventually the conditioned media collected for IgG purification contained no HCF, and just 8% FBS with ultra low IgG (< 5μ^ιη1).

EXAMPLE 9

[46] These experiments detail investigations into the ability of the mAbs to neutralize PDGF-DD.

[47] Samples of 10 μg of fl human PDGF-DD were diluted in 1 ml of serum free conditioned medium that had been collected from COS-1 cells, Then, either 5 μg or 20 μg of IgG from clones 1 and 6 were added, and incubated with the target protein, for 1 hour at 4°C. Following this, 10 pg of HMWuPA was added, and left to compete for binding to PDGF-DD for 1.5 hours, at room temperature, (As a control, a parallel experiment used a rabbit polyclonal antibody against fl human PDGF-DD).

[48] Following this period, the proteins were precipitated with 20% TCA on ice, for 1 hour, followed by a wash with 80% ethanol. The PDGF-DD was then analyzed via immunoblotting under reducing conditions, using a PDGF-DD goat polyclonal antibody, and a commercially available one step Western blot kit.

[49] Effective inhibition of activation of the fl human PDGF-DD was seen at both concentrations. Given this result, all of the mAbs were tested, using only 5 pg/ml IgG, and incubation of 1.25 hours with each of fl PDGF-DD and HMWuPA. Four clones exhibited inhibition of HMWuPA mediated processing.

EXAMPLE 10

[50] The mAbs identified as inhibitory in the foregoing experiment were analyzed in a tyrosine phosphorylation assay. In this assay, porcine aortic endothelial cells ("PAE" hereafter) which expressed human PDGFRp were used to detect active PDGF-DD species.

[51] The protocol of example 9, supra, was used but the medium employed was serum free, F12/0.1T BSA. Presence of PDGF-DD was analyzed via induction of cellular receptor activation, after 1 hour of stimulation, on ice, using a non-specific, IgG purified, and rabbit polyclonal antibody as the control. Cells were washed in cold PBS, mildly centrifuged, and lysed in lysis buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM EDTA, 0.5% deoxycholic acid, 200 μΜ orthovandate, and complete protease inhibitor). Any released proteins were analyzed using a mouse monoclonal antibody which recognized phosphotyrosine PY99, and a goat polyclonal antibody which detected calnexin, as loading control.

[52] Neutralization of PDGF-DD activity was confirmed, for all of the four clones referred to supra, via the inhibition of formation of phosphotyrosine PYPP.

[53] The foregoing examples describe the invention, which includes, inter alia, monoclonal antibodies ("mAbs"), which specifically bind to and inhibit full length, human PDGF-DD. To elaborate, in addition to binding full length human PDGF-DD, the mAbs do not bind to other species of PDGF-DD, including the muiine counterpart, nor do they bind to any truncated forms of human PDGF-DD, including the CUB domain, the activated molecule, or the GFD region. To elaborate further, the mAbs of the invention bind to an epitope defined by or involving amino acids 180-247/249 of human PDGF- DD. The sequence referred to is that of Hart, et al., supra. [54] "Inhibits" PDGF-DD, as used herein, refers to the ability of the mAbs to inhibit the processing of full length, human PDGF-DD, to its active form, i.e., PDGF-DD consisting of the VHD and GFD regions of the full length molecule. It is well known that this form of the molecule interacts with receptors; hence, by inhibiting processing (by uP A), the mAbs inhibit binding of PDGF-DD to receptors, and consequently, the cascade of events which result from that binding.

[55] Also a part of the invention are the hybridomas which produce the mAbs described herein, as well as a method for screening a population of antibodies, be these polyclonal or monoclonal, to identify antibodies which possess the properties of the claimed invention.

[56] Also a feature of the invention is a method for inhibiting full length human PDGF-DD, by contacting the molecule with the mAb of the invention so as to inhibit processing thereof by urokinase plasminogen activator.

[57] Other features of the invention will be clear to the skilled artisan and need not be reiterated here.

[58] The terms and expression which have been employed are used as terms of description and not of limitation, and there is no intention in the use of such terms and expression of excluding any equivalents of the features shown and described or portions thereof, it being recognized that various modifications are possible within the scope of the invention.

Claims

We Claims:

1. A hybridoma cell line which produces a monoclonal antibody that:

(i) binds specifically to full length human platelet derived growth factor-DD ("PDGF-DD");

(ii) does not cross react with PDGF-DD from non-human species;

(iii) does not bind to isolated CUB domains, or activated forms of human PDGF-DD, and

(iv) inhibits processing of full length human PDGF-DD by urokinase plasminogen activator.

2. A monoclonal antibody produced by the hybridoma of claim 1.

3. The monoclonal antibody of claim 2, which binds to an epitope which involves a portion of amino acids 180-247/249 of full length human PDGF-DD.

4. A method for identifying the monoclonal antibody of claim 2, comprising:

(a) Contacting a sample of monoclonal antibodies with each of full length human PDGF-DD: a PDGF-DD from a non-human species, CUB domain of human PDGF-DD, and activated human PDGF-DD, and selecting any monoclonal antibodies which bind only to full length human PDGF-DD.

5. A method for inhibiting processing of full length human PDGF-DD to its active form, comprising contacting full length human PDGF-DD to the monoclonal

, antibody of claim 2, under conditions favoring inhibition of processing of said full length, human PDGF-DD by urokinase plasminogen activator.