Patent Grant
9567398
This application is the U.S. National Phase Application of PCT/EP2011/070375, filed Nov. 17, 2011, which claims priority to EP 10191664.1, filed Nov. 18, 2010, the contents of each are hereby incorporated by reference in their entireties.
The present invention relates to antibodies with specificity for granulocyte-macrophage colony stimulating factor (GM-CSF). More particularly, the invention relates to humanized monoclonal antibodies that bind specifically to human GM-CSF with high affinity. The invention also relates to nucleic acids encoding the antibodies, vectors for expression of these nucleic acids, and host cells for producing said antibodies. Further, the invention relates to the use of said antibodies in the diagnosis or treatment of autoimmune or inflammatory diseases.
Granulocyte-macrophage colony stimulating factor (GM-CSF) is a hematopoietic cytokine that stimulates the differentiation of progenitor cells into monocytes and granulocytes (eosinophils, neutrophils, and basophils). GM-CSF is mainly produced from activated T cells, macrophages and endothelial cells, but secretion of this cytokine has been observed from a plethora of cells and cell lines.
GM-CSF is also a potent. inflammatory cytokine, whose activity or overexpression can have significant detrimental effects. It is implicated in a variety of autoimmune and inflammatory diseases including rheumatoid arthritis, asthma, multiple sclerosis and idiopathic thrombocytopenic purpura. For example, it has been reported that GM-CSF−/− mice develop no disease in a collagen-induced arthritis model. Further, in asthma and rheumatoid arthritis, elevated levels of GM-CSF have been detected and correlated with the inflammatory process.
Neutralization or blockage of GM-CSF activity is therefore a promising approach in the treatment and/or prevention of autoimmune or inflammatory conditions. In particular, GM-CSF neutralization or blockage may be achieved by antibodies directed to GM-CSF. In order to neutralize GM-CSF, or be neutralizing, binding of an antibody partially or completely inhibits one or more biological activities of GM-CSF. For example, the antibody reduces inflammatory responses mediated by GM-CSF.
WO 2008/141391 discloses antibodies that bind to GM-CSF or fragments thereof and antagonize or neutralize GM-CSF activity. The content of this document is herein incorporated by reference. In particular, this document refers to a humanized form of murine monoclonal antibody 4K21 deposited on 17 May 2007 at the European Collection of Cell Cultures (ECACC), Centre for Applied Microbiology and Research, Porton Down, Salisbury, United Kingdom, under Accession No. 07051601.
The genomic sequence encoding the epitope recognized by this antibody contains a single nucleotide polymorphism (SNP), leading to two diverse codons coding for isoleucine (I) and threonine (T), respectively, at position 117 of human GM-CSF protein (see
Experimental data have shown that the anti-GM-CSF 4K21 recombinant chimeric and humanized molecules, e.g. clone hGM4/34 of WO 2008/141391, have a preference for the human GM-CSF I117 target protein, with a ≧9 fold lower affinity for the T117 target protein in solution as determined by Kd measurements. Likewise, these antibodies show lower potency (IC50 of GM-CSF dependent TF-1 cell proliferation and IL-8 secretion in U937 cells) for the T117 containing GM-CSF protein.
GM-CSF protein is highly conserved among different species of the order Primates, e.g. man, macaca and marmoset (
Accordingly, there was a need for anti-GM-CSF antibodies with improved properties. It was an object of the present invention to provide antibodies with specificity to human GM-CSF to overcome the problems of state-of-the-art-antibodies such as 4K21. In particular, the newly developed antibodies should have improved potency to both GM-CSF protein isoforms as well as improved cross-reactivity to the macaca ortholog, better affinity maturation and biochemical properties, such as Fab stability. Preferably, the newly developed antibodies are based on 4K21.
Surprisingly, it could be shown that mutation of a single amino acid, histidine at position 93 in the variable light chain region of the humanized 4K21 antibody, significantly improved the properties of the resulting antibody.
A first aspect of the invention therefore relates to an antibody that binds to human GM-CSF, which comprises
(i) a heavy chain variable region comprising complementarity determining regions CDR1, CDR2, and CDR3 as shown in SEQ ID NOs: 1, 2, and 3, or variants thereof wherein optionally 1, 2, or 3 amino acids in each CDR may be substituted by different amino acids; and
(ii) a light chain variable region comprising complementarity determining regions CDR1, CDR2; and CDR3 as shown in SEQ ID NOs: 4, 5 and 6, or variants thereof wherein optionally 1, 2, or 3 amino acids in each CDR may be substituted by different amino acids, with the proviso that the amino acid tyrosine at position 93 in the light chain CDR3 as set forth in SEQ ID NO:6 is not substituted,
or an antigen-binding fragment thereof.
Examples for “human GM-CSF” are the naturally occurring I117-(SEQ ID NO: 26) and T117-(SEQ ID NO: 27) isoforms, as well as further naturally occurring isoforms and artificially produced mutants. Such mutants may carry one, two, three, four, five, ten or more amino acid substitutions, deletions, and/or insertions, as long as the biological activity of the protein is retained. One example for a GM-CSF mutant carrying a single amino acid substitution is set forth in SEQ ID NO: 28 (A117).
The term “antibody” as used in the present invention comprises all the various forms of antibodies including, but not limited to whole antibodies and antigen-binding fragments thereof, e.g. Fv, Fab, Fab′ and F(ab′)2, diabodies, minibodies, domain antibodies (dAb), camelid antibodies (IgG) and nanobodies (VHH), monovalent antibodies, divalent or multivalent antibodies comprising a fragment of more than one antibody, single chain variable fragments (scFv), and antibody derivatives such as immunoglobulin-derived polypeptides, disulfide stabilized Fv fragments (dsFv), CDR-comprising peptides etc. Antibody fragments such as Fabs (approx. 60 kDa) or scFvs (about 20-30 kDa) and still smaller fragments of antibodies are of great interest and advantageous for pharmaceutical applications or as imaging agents.
Preferably, the antibody comprises at least one heavy immunoglobulin chain and at least one light immunoglobulin chain, for example two heavy chains and two light chains, which may be covalently linked via disulfide bonds. A functional structure analogous to the isotype G of immunoglobulins (IgG) is preferred. Among these, antibodies of the IgG1-, IgG2-, IgG3-, or IgG4-type are preferred. However, the antibody or antigen-binding fragment thereof according to the invention may also be of the IgA-, IgD-, IgE, or IgM-type, preferably of the IgM1- and IgM2-type.
An immunoglobulin chain comprises a variable domain and optionally a constant domain. A variable domain (also referred to as variable region) comprises complementarity determining regions (CDRs), e.g. a CDR1, CDR2 and/or CDR3 region, and framework regions (FRs). The term “complementarity determining region” (CDR) is well-defined in the art (see, for example, Harlow and Lane (eds.), Antibodies: A Laboratory Manual, CSHL Press, Cold Spring Harbor, N.Y., 1988; incorporated herein by reference in its entirety) and refers to the stretches of amino acids within the variable domain of an antibody that primarily make contact with the antigen and determine antibody specificity. The antibodies of the invention have at least one antigen binding site, e.g. one or two antigen binding sites. In certain embodiments, e.g. for CDR-comprising peptides, a variable domain may contain only CDRs linked via short linker peptides instead of complete framework regions.
The term “bind” or “binding” of an antibody means an at least temporary interaction or association with or to a target antigen, e.g. GM-CSF, comprising fragments thereof containing an epitope.
The antibody according to the invention binds to GM-CSF or a fragment thereof. Typically, antibodies show species-specificity for GM-CSF. In the present invention, high specificity for human GM-CSF is preferred. However, depending on the degree of sequence identity between GM-CSF homologs of different species, a given antibody or antigen-binding fragment may show cross-reactivity with GM-CSF from at least one other species, e.g. macaca, marmoset, dog, rabbit, mouse and/or rat. For antibodies directed towards human GM-CSF, some level of cross-reactivity with other mammalian forms of GM-CSF may be desirable in certain circumstances, for example when testing antibodies in animal models of a particular disease or for conducting toxicology, safety and dosage studies.
The antibody according to the invention comprises a variable heavy chain region (VH) and/or a variable light chain region (VL), whose CDRs show a high degree of sequence identity to murine and/or humanized 4K21 complementarity determining region sequences (VH: SEQ ID NOs: 1, 2, 3; VL: SEQ ID NOs: 4, 5, 6). In the VL CDR3 sequence, the amino acid tyrosine is substituted for histidine. According to the numbering of Kabat et al. (Sequences of Proteins of Immunological Interest, US Dept. Health and Human Services, NIH Publication No. 91-3242 (1991); incorporated herein by reference), which is used throughout this invention, this histidine is assigned position number 93; the substitution is therefore referred to as VL H93Y. The VL CDR3 is shown in SEQ ID NO:6 with the H9 3Y substitution, and in SEQ ID NO:7 without the substitution.
Besides the H93Y substitution in the variable light chain region and murine back mutations introduced during humanization (see below), which are also encompassed by the present invention, antibodies or fragments thereof according to the invention may contain one or more, e.g. 2, 3, 4, 5, 6, 7, 10, 15 and up to 18 additional mutations. These polypeptide chains are encompassed by the term “variants”.
The term “mutation” or “mutated” as used in the context of the present invention includes the deletion, insertion and substitution of amino acids and/or nucleotides. Substitutions are preferred, but deletions and/or insertions are also possible, as long as binding of the resulting antibody to GM-CSF is not abolished.
It is contemplated that the framework regions flanking the CDRs may be entirely or partially substituted. However, it is preferred that the degree of sequence identity of the VL and/or VH regions according to the invention to those of the murine and/or humanized 4K21 is at least 70%, preferably at least 80%, more preferably at least 90%, and most preferably at least 95%. In preferred embodiments, the antibody comprises a variable light chain region sequence selected from the group consisting of SEQ ID NOs: 8, 9 and 10.
The percentage “sequence identity” of two sequences may be determined using the BLASTP algorithm (Altschul et al., Gapped BLAST and PSI-BLAST: a new generation of protein database search programs, Nucleic Acids Res. 25:3389-3402, 1997; incorporated herein by reference) using default parameters.
Additional mutations are preferably located in the CDRs of both the heavy and the light chain. In particular, 1, 2, or 3 amino acids in each heavy and/or light chain CDR may be mutated. In preferred embodiments, at least one of the amino acids serine at position 55 (VH S55), serine at position 56 (VH S56), threonine at position 62 (VH T62) in the heavy chain CDR2, and/or asparagine at position 28 (VL N28), histidine at position 34 (VL H34) in the light chain CDR1, and/or asparagine at position 53 (VL N53) in the light chain CDR2 can be substituted by another amino acid. Optionally also the serine at position 27e (VL S27e) of the light chain CDR1 may be substituted by another amino acid.
Still more preferably, one or more specific amino acids may be substituted for one or more of the above-indicated amino acids. Accordingly, in especially preferred embodiments of the invention, VH S55 is substituted by lysine or arginine (VH S55K or VH S55R), and/or VH S56 is substituted by valine (VH S56V), and/or VH T62 is substituted by serine (VH T62S), and/or VL H34 is substituted by serine (VL H34S), and/or VL N28 is substituted by isoleucine (VL N28I), and/or VL N53 is substituted by arginine (VL N53R). In a further embodiment, the amino acid tyrosine is substituted for the serine at position 27 of the light chain CDR1 (VL S27eY).
Still more preferably, the antibody according to the invention comprises a variable heavy chain region comprising complementarity determining regions CDR1 as shown in SEQ ID NO:1, CDR2 selected from SEQ ID NOs: 11, 12, and 13, and CDR3 as shown in SEQ ID NO:3, and a variable light chain region comprising complementarity determining regions CDR1 selected from SEQ ID NOs: 14 and 15, CDR2 as shown in SEQ ID NO: 16, and CDR3 as shown in SEQ ID NO:6.
In a preferred embodiment, the affinity of the antibody of the invention to the human GM-CSF T117 and I117 alleles differs by a factor of less than 5, preferably less than 2. This means that the strength of binding of the antibody to human GM-CSF does not depend on the amino acid at position 117. The affinity of an antibody is defined as the monovalent binding strength of one molecule to another, e.g. one epitope to one paratope. The value of the antibody affinity can be expressed as an equilibrium constant K, e.g. the equilibrium dissociation constant, Kd which is defined as koff/kon, i.e. the ratio of the reverse rate constant (koff) to the forward rate constant (kon). The Kd values are given in mol/l or M. The affinity can be determined e.g. via measurement of binding constants and calculation of Kd using a kinetic exclusion assay (KINEXA® system, Sapidyne Instruments, Inc.).
In a further preferred embodiment, the potency of the antibody according to the invention to the human GM-CSF T117 and I117 alleles differs by a factor of less than 5, preferably less than 2. This means that the antibody can be used to neutralize both isoforms of human GM-CSF to a similar extent. The potency may be expressed as the concentration of antibody/antigen-binding fragment which produces the half-maximal effect at a given antigen concentration. For example, the “effect” of an antibody may be inhibition or neutralization of an antigen. In this case, the antibody concentration producing the half-maximal inhibition may be referred to as IC50, which is given in mol/l or M. Potency is usually influenced by affinity until, at a given antigen concentration, an affinity is reached beyond which further improvements in affinity will not enhance binding of the antigen anymore (so-called potency ceiling). Potency may for example be determined by measuring the IC50 values of GM-CSF dependent TF-1 cell proliferation and/or of GM-CSF dependent IL-8 secretion in U937 cells.
In yet a further preferred embodiment, the antibody is a monoclonal antibody, a chimeric antibody, a recombinant antibody and/or a human or humanized antibody. A monoclonal antibody (also referred to as mAB) is a single molecular species of antibody and is usually produced by creating hybrid antibody-forming cells from a fusion of nonsecreting myeloma cells with immune spleen cells. Polyclonal antibodies, by contrast, are produced by injecting an animal (such as a rodent, rabbit or goat) with an antigen, and extracting serum from the animal. A chimeric antibody is an antibody in which the variable domain of e.g. a murine antibody is combined with the constant region of a human antibody. Recombinant antibodies are obtained via genetic engineering without having to inject animals. Human antibodies according to the invention may be prepared using transgenic mice or by phage display; these methods are well known in the art.
A humanized antibody is a genetically engineered (monoclonal) antibody in which the CDRs from a donor antibody, e.g. an antibody from mouse, rat, hamster etc. are grafted onto an acceptor antibody, i.e. a human antibody. Typically, a humanized antibody comprises a light chain comprising three CDRs from a murine antibody, a variable domain framework from a human antibody, and a human constant region, and a heavy chain comprising three CDRs from a murine antibody, a variable domain framework from a human antibody, and a human constant region. Humanization of antibodies may be performed to avoid an immune response, resulting in generation of human anti-mouse antibodies (HAMA), the so-called HAMA response. Methods for the humanization of antibodies, e.g. CDR grafting, are well known in the art. In the context of the present invention, humanized monoclonal antibodies are particularly preferred.
Preferred examples of antibodies according to the invention are based on humanized 4K21 variable region heavy chain (4K21 hVH; SEQ ID NO: 18) and the corresponding light chain 4K21-hVL carrying the H93Y mutation (SEQ ID NO: 8). Exemplary heavy chains based on humanized 4K21 with preferred mutations are 4K21 hVH-S55K-T62S (SEQ ID NO: 21), 4K21 hVH-S55R-T62S (SEQ ID NO: 22), and 4K21 hVH-S56V-T62S (SEQ ID NO: 23). Exemplary light chains based on humanized 4K21 with preferred mutations are 4K21 hVL-S27eY-N28I-H34S-N53R-H93Y (SEQ ID NO: 9) and 4K21 hVL-N28I-H34S-N53R-H93Y (SEQ ID NO: 10).
The antibody according to the invention may be glycosylated. Typically, monosaccharaides such as N-acetylglucosamine, mannose, glucose, galactose, fucose, sialic acid etc. are assembled to oligosaccharides at individual glycosylation sites. The glycosylation pattern may be the same as or different from the pattern found in a host species, e.g. mouse, macaca, marmoset, or man.
It is also contemplated that the antibody according to the invention may be conjugated with a further molecule. The further molecule may be conjugated to the antibody directly or via a spacer of suitable length. For therapeutic purposes, conjugation with a therapeutic effector group, such as a radioactive group, i.e. a group consisting of or comprising a radioisotope or radionuclide (e.g. 3H, 14C, 15N, 33p, 35S, 90Y, 99Tc, 111In, 123I, 125I, 131I, 201TI, 213Bi), a toxin, or a cytotoxic group, e.g. a cell growth inhibitor may be suitable. In another aspect, the antibody or antigen-binding fragment thereof can be coupled to a labeling group (labeled antibody), which may then be used e.g. for diagnostic purposes. Suitable labeling groups may be selected from radioisotopes (e.g. those mentioned supra) or groups containing a radioisotope, radionuclides, fluorescent groups (e.g. fluorescent proteins such as GFP, RFP etc., Alexa-Fluor® dyes, rhodamines, fluorescein and its derivatives such as FITC, cyanine dyes such as Cy3® and Cy5®, enzymatic groups (e.g. horseradish peroxidase, alkaline phosphatase, β-galactosidase), chemiluminescent groups, biotinyl groups, metal particles, (e.g. gold particles), magnetic particles (e.g. with a core containing magnetite (Fe3O4) and/or maghemite (Fe2O3)), predetermined polypeptide groups, etc.
In another aspect, the invention relates to a nucleic acid, preferably an isolated nucleic acid, encoding the antibody of the invention. The nucleic acid of the invention encoding the above-described antibody or antigen-binding fragment may be, e.g., a natural DNA or RNA or a recombinantly or synthetically produced DNA, RNA or LNA or a recombinantly produced chimeric nucleic acid molecule comprising any of these nucleic acids either alone or in combination. The nucleic acid may also be genomic DNA corresponding to an entire gene or a substantial portion thereof or to fragments and derivatives thereof. The nucleotide sequence may correspond to the naturally occurring nucleotide sequence or may contain single or multiple nucleotide substitutions, deletions or additions. The nucleic acid may also be fused to another nucleic acid. In a particularly preferred embodiment of the present invention, the nucleic acid is a CDNA molecule.
The nucleic acid according to the invention is preferably selected from
(i) one or more nucleic acid sequences set forth in SEQ ID NOs. 31-40, with SEQ ID NOs. 31-35 and 38-40 being particularly preferred,
(ii) nucleic acid sequences comprising sections encoding CDRs as set forth in SEQ id NOs: 1-6 and 11-16,
(iii) nucleic acid sequences encoding a polypeptide selected from the group consisting of SEQ ID NOs: 8-10, 17-18, and 21-23.
(iv) nucleic acid sequences complementary to any of the sequences in (i), (ii) or (iii), and
(v) nucleic acid sequences capable of hybridizing to (i), (ii), (iii) or (iv) under stringent conditions and encoding a polypeptide, wherein an antibody or functional fragment thereof comprising said polypeptide binds to GM-CSF.
The term “stringent conditions” is readily understood by the skilled person and may refer, e.g., to hybridization in 6.0×SSC at about 45° C. followed by a washing step with 2.0×SSC at 50° C. (see e.g. Sambrook et al., Molecular Cloning: A Laboratory Manual, 3rd ed., CSHL Press, Cold Spring Harbor, N.Y., 2001; incorporated herein by reference in its entirety).
The invention also relates to a vector or vector system comprising a nucleic acid of the invention. A vector according to the invention may be, for example, a plasmid, a phage, a phagemid, a viral or a retroviral vector. Retroviral vectors may be replication competent or replication defective. In the latter case, viral propagation generally will occur only in complementing host cells. A vector system according to the invention comprises a vector as defined supra and optionally further vectors which may be required for the respective application.
In a preferred embodiment, the vector or vector system comprises an expression vector. An expression vector is generally used to introduce a specific gene into a target cell in order to express it, usually to produce high quantities of nucleic acids or proteins encoded by said gene, and typically contains an origin of replication, expression control sequences as well as specific genes capable of providing phenotypic selection of the transformed cells.
In an expression vector, the nucleic acid of the invention is operably linked to one or more expression control sequences allowing the transcription and optionally translation in prokaryotic and/or eukaryotic host cells. The term “operably linked” means that a nucleic acid sequence is placed into a functional relationship with another nucleic acid sequence. The expression control sequences define, inter alia, the location at which the RNA polymerase binds and at which the ribosomes bind and interact with the mRNA (the product of transcription) to initiate translation of the mRNA into protein. The expression control sequence may be constitutively or regulatably active. Suitable expression control sequences include prokaryotic promoter sequences, e.g. lac, trp, tac, T7, araB or phoA promoters, eukaryotic promoter sequences, e.g. ADH1, AOX, CYC1, GAL1, or PGK1 promoters in yeast or the CMV-, SV40-, RSV and other promoters in higher eukaryotic cells, operator sequences such as lacO or tetO, enhancer sequences, e.g. CMV- or SV40-enhancer, and terminator sequences.
Besides elements which are responsible for the initiation of transcription, expression control sequences may also comprise transcription termination signals, such as the SV40-poly-A site or the tk-poly-A site, located downstream of the nucleic acid sequence of interest. Additional expression control sequences may include translational enhancers. In general, vectors containing expression control sequences which facilitate the efficient transcription and/or translation of the inserted nucleic acid molecule are chosen in accordance with the respective host.
Suitable expression vectors advantageously contain elements enabling selection in the host cell. For example, the vector comprising a nucleic acid of the invention may contain at least one selectable marker, e.g. an antibiotic resistance marker or an auxotrophic marker.
Methods to construct the vectors according to the invention and to introduce them into suitable hosts or host cells, e.g. by (co)transformation, (co)transfection, or (co)infection, are well known to those skilled in the art (see e.g. Sambrook et al., 2001, supra).
The invention further relates to a non-human host organism or a host cell for expressing the nucleic acids according to the invention. The nucleic acid or vector of the invention, which is present in the host, may either be integrated into the genome of the host or host cell or it may be maintained extrachromosomally.
Suitable non-human hosts include all bacterial species which can be transformed or transfected with a nucleic acid sequence for its expression, for example Escherichia coli, Salmonella typhimurium, Serratia marcescens and Bacillus subtilis. Preferred fungal hosts are, inter alia, of the family Saccharomycetaceae, e.g. Saccharomyces cerevisiae, Hansenula polymorphs or Pichia pastoris. Among mammalian hosts, goats, cows, horses, pigs, rats, mice, rabbits, hamsters and chicken are preferred. There, antibodies can be produced in and recovered from tissue or body fluids, such as milk, blood or urine.
The host cell can be any prokaryotic or eukaryotic cell, such as a bacterial, fungal, plant, or animal cell. Suitable prokaryotic cells include cells of bacterial species as described above. Likewise, preferred fungal cells are those of Saccharomycetaceae species as described above. Preferred animal cells include but are not limited to insect, amphibian, mammalian or, more preferably, human cells. Suitable insect cells are e.g. cells of the SF9 cell lines. Preferred mammalian cells include Chinese hamster ovary cells (CHO, ATCC CCL-61), baby hamster kidney (BHK) cells, monkey kidney cells (COS, ATCC CRL-1650), 3T3 cells or NIH3T3 cells derived from a mouse fibroblast (ATCC No. CRL-1658), NSO cells and a number of other cell lines. Suitable human cells include, but are not limited to human embryonic kidney cells (HEK293, 293T), Hela cells, human hepatocellular carcinoma cells (e.g., Hep G2), and A549 cells.
According to another embodiment of the invention, antibodies or antigen-binding fragments thereof are produced by cell lines. These cell lines may be hybridomas, i.e. hybrid cell lines obtained by fusing a specific antibody-producing B lymphocyte with a myeloma cell and appropriate selection, which is well known in the art. Also contemplated are cell lines, e.g. 293 or CHO cell lines, transformed with a vector or vector system as described supra.
Methods for preparing fused and/or operably linked genes and expressing them in bacteria or animal cells are well-known in the art (Sambrook et al., 2001, supra). The genetic constructs and methods described therein can be utilized for expression of antibodies according to the invention in, e.g., prokaryotic hosts. The nucleic acid sequences encoding individual polypeptide chains of an antibody may be expressed from a single vector or co-expressed from two or more distinct vectors. For example, one vector may encode an IgG heavy chain, and the other may encode a light chain. The transformed prokaryotic hosts or transfected higher eukaryotic host cells may be cultured according to techniques known in the art to achieve optimal cell growth. The antibodies or antigen-binding fragments thereof can then be isolated from the growth medium, cellular lysates, or cellular membrane fractions. The isolation and purification of the microbially or otherwise expressed antibodies or antigen-binding fragments thereof can be performed with conventional means such as preparative chromatographic separations, immunological separations, etc.
The present invention further relates to a pharmaceutical composition comprising an antibody or antigen-binding fragment thereof. It is also possible that the pharmaceutical composition comprises a nucleic acid encoding said antibody or antigen-binding fragment thereof, a vector or vector system containing said nucleic acid and/or a preferably human cell producing said antibody or antigen-binding fragment thereof. Optionally, the pharmaceutical composition comprises pharmaceutically acceptable excipients, adjuvants and/or carriers.
As exemplary excipients, disintegrators, binders, fillers, and lubricants may be mentioned. Examples of disintegrators include agar-agar, algins, calcium carbonate, cellulose, colloid silicon dioxide, gums, magnesium aluminium silicate, methylcellulose, and starch. Examples of binders include microcrystalline cellulose, hydroxymethyl cellulose, hydroxypropylcellulose, and polyvinylpyrrolidone. Examples of fillers include calcium carbonate, calcium phosphate, tribasic calcium sulfate, calcium carboxymethylcellulose, cellulose, dextrin, dextrose, fructose, lactitol, lactose, magnesium carbonate, magnesium oxide, maltitol, maltodextrins, maltose, sorbitol, starch, sucrose, sugar, and xylitol. Examples of lubricants include agar, ethyl oleate, ethyl laureate, glycerin, glyceryl palmitostearate, hydrogenated vegetable oil, magnesium oxide, stearates, mannitol, poloxamer, glycols, sodium benzoate, sodium lauryl sulfate, sodium stearyl, sorbitol, and talc.
Usual stabilizers, preservatives, wetting and emulsifying agents, consistency-improving agents, flavour-improving agents, salts for varying the osmotic pressure, buffer substances, solubilizers, diluents, emollients, colorants and masking agents and antioxidants come into consideration as pharmaceutical adjuvants.
Suitable carriers include but are not limited to magnesium carbonate, magnesium stearate, talc, sugar, lactose, pectin, dextrin, starch, gelatine, tragacanth, methylcellulose, sodium carboxymethylcellulose, a low melting-point wax, cocoa butter, water, alcohols, polyols, glycerol, vegetable oils and the like.
The pharmaceutical composition may also comprise at least one further active agent, e.g. one or more further antibodies or antigen-binding fragments thereof, peptides, proteins, nucleic acids, organic and inorganic molecules.
In a preferred embodiment of the invention, the pharmaceutical compositions comprising an antibody or antigen-binding fragment thereof are for use in medicine or diagnostics. Preferably, the pharmaceutical compositions are for use in human medicine, but they may also be used for veterinary purposes.
In particular, the antibodies or antigen-binding fragments thereof, the nucleic acids, the vector or vector system, the host or host cell of the invention, or a pharmaceutical composition comprising an antibody or antigen-binding fragment thereof are for use in the diagnosis, prevention or treatment of disorders associated with, caused by or accompanied by elevated levels and/or activity of GM-CSF, and other diseases or conditions which may be beneficially diagnosed, prevented, or treated by inhibiting and/or neutralizing GM-CSF activity via the administration of antibodies or antigen-binding fragments thereof as described supra.
In a further embodiment, the present invention relates to methods for the diagnosis, prevention or treatment of disorders associated with, caused by or accompanied by elevated levels and/or activity of GM-CSF, and other diseases or conditions which may be beneficially diagnosed, prevented, or treated by inhibiting and/or neutralizing GM-CSF activity.
Preferred medical indications are autoimmune or inflammatory diseases or conditions associated with elevated levels and/or activity of GM-CSF. The disease or condition may be selected from, for example, asthma, rheumatoid arthritis, chronic obstructive pulmonary disease, idiopathic thrombocytopenic purpura, acute respiratory distress syndrome, multiple sclerosis, Alzheimer's disease, Crohn's disease, irritable bowel syndrome, colitis, psoriasis, macular degeneration, uveitis, Wallerian degeneration, antiphospholipid syndrome, restinosis, atherosclerosis, idiopathic pulmonary fibrosis, relapsing polychondritis, hepatitis, glomerulonephritis, nephritis, lupus, atopic dermatitis, sarcoidosis, osteoarthritis, osteoporosis and osteolytic tumors, the treatment of bone metastasis in myeloma, prostate or breast cancer, or, alternatively, prevention of the formation of breast cancer metastasis, leukemia and other metabolic diseases. Particularly preferred indications are rheumatoid arthritis, multiple sclerosis, asthma and chronic obstructive pulmonary disease.
An antibody or antigen-binding fragment thereof or a pharmaceutical composition according to the invention can be administered to a subject in need thereof in an amount effective to obtain the desired therapeutic or prophylactic effect. For example, one desired effect to be achieved by said administration may be to block, inhibit and/or neutralize one or more biological function(s) of GM-CSF. In this context, administration may comprise contacting the antibody of the invention with cells or a tissue suspected of expressing GM-CSF, preferably at high and/or aberrant levels, under conditions, wherein the antibody is capable of blocking, inhibiting and/or neutralizing GM-CSF function. The contacting may be in vitro or in vivo.
Administration of suitable compositions may be effected in different ways, e.g., by intravenous, intraperitoneal, subcutaneous, intramuscular, topical, oral, intradermal, intranasal or intrabronchial administration. Administration may also be conducted directly at the target site.
The dosage regimen will be determined by the attending physician and clinical factors. As is well known in the medical arts, dosage for any one patient depends upon many factors, including the patient's size, weight, body surface area, age, the particular compound to be administered, the activity of the employed antibody or antibodies, time and route of administration, general health, and combination with other therapies or treatments. Proteinaceous pharmaceutically active matter may be present in amounts between 1 μg and 100 mg/kg body weight per dose; however, doses below or above this exemplary range are also envisioned. If the regimen is a continuous infusion, it may be in the range of 1 pg to 100 mg per kilogram of body weight per minute.
A neutralizing antibody or antigen-binding fragment thereof according to the invention may be employed at a concentration of, e.g., 0.01, 0.1, 0.5, 1, 2, 5, 10, 20 or 50 μg/ml in order to inhibit and/or neutralize a biological function of GM-CSF by at least about 50%, preferably 75%, more preferably 90%, 95% or up to 99%, and most preferably approximately 100% (essentially completely) as assayed by methods well known in the art.
According to further aspects of the invention, the antibodies and antigen-binding fragments thereof may be used additional applications in vivo and in vitro. For example, antibodies or antigen-binding fragments of the invention may be employed for diagnostic purposes, e.g. in assays designed to detect and/or quantify the presence of GM-CSF and/or to purify GM-CSF. Antibodies may also be tested in animal models of particular diseases and for conducting toxicology, safety and dosage studies.
Finally, the invention relates to a kit comprising at least one antibody or antigen-binding fragment thereof according to the invention, at least one nucleic acid sequence encoding said components, the vector or vector system of the invention, and/or a host cell according to the invention. It is contemplated that the kit may be offered in different forms, e.g. as a diagnostic kit.
The present invention shall be further illustrated by the following Figures and Examples.
b) Alignment of human (SEQ ID NO:26), macaca (SEQ ID NO:29), and the predicted marmoset (SEQ ID NO: 30) GM-CSF protein sequences. Bottom paragraph gives scores of pairwise alignment.
A recombinant chimeric form of the murine hybridoma-derived 4K21 antibody was obtained by fusing the murine 4K21 variable heavy and light chain DNA sequences (4K21-mVH and -mVL; SEQ ID NOs: 17 and 19) into expression plasmids containing human constant IgG1 heavy chain and kappa light chain domains, respectively. Co-expression of these plasmids resulted in anti-GM-CSF 4K21 hIgG1K antibody (MSB0010259).
Humanization of 4K21 was done by grafting 4K21's murine CDR regions into the frameworks of human VH3 (V-base VH3-07, i.e. IGHV3-7; SEQ ID NO: 24) and VK2 (V-base VK2-A3, i.e. IGKV2-28; SEQ ID NO: 25) germ line genes. Only four back mutations (A28, R44, L47 and T93) to the corresponding murine sequence in the framework were reintroduced in the humanized VH, and none in the VL. The 4K21-VH3-07.1AL heavy chain (4K21-hVH; SEQ ID NO: 18) and 4K21-VK2-A3.2 light chain (4K21-hVL; SEQ ID NO: 20) thus constitute the humanized parental anti-GM-CSF 4K21 VH3-07.1AL+VK2-A3.2 hIgG1K antibody (MSB0010439) (
To determine potency and affinity of the antibody, Kd measurements and a proliferation bioassay of TF-1 cells were performed. TF-1 is an erythroleukemic cell line from blood, which proliferates in response to e.g. GM-CSF. Kd determination was carried out with the Kinetic Exclusion Assay (KINEXA®; Sapidyne Instruments, Inc., Boise, Id., USA).
In the TF-1 proliferation assay and the KINEXA®, the humanized parental 4K21 antibody (MSB0010439) had a similar potency and affinity to the chimeric (MSB0010259) and murine 4K21 antibody, including the >10× stronger preference for GM-CSF I117 versus T117, and was <10 fold less efficient on neutralizing the orthologs of cynomolgus (Macaca fascicularis) and rhesus (Macaca mulatta) monkeys.
During the screening procedure using calibrated IgG supernatants of transiently transfected cells or small-scale purified antibody samples, 9 individual amino acid substitutions at 8 different sites were identified that contribute to the improvement of 4K21: 5 individual VH- & VL-mutations (VH-S55K, S55R or S56V, and VL-N53R and H93Y) partly reduced the human SNP and macaca binding differences and improved overall potency as compared to the parental antibody; VL-H34S improved overall potency for all target species and expression; VL-N28I removed the NG deamidation site without affecting potency; VL-S27eY totally cancelled the macaca ortholog difference and further destroyed the potential NS deamidation site adjacent to NG; and T62S increased homology to human germline gene.
Small-scale purified humanized parental antibody with the H93Y mutant alone (MS100305-A) improved the IC50 of the humanized parental 4K21 antibody (MSB0010439) by 2.3-, 6-, and 5-fold in the presence of hGM-CSF-I117, T117, and the macaca ortholog, respectively. Consequently, the affinity for hGM-CSF T117 or macaca GM-CSF were still 3.7- or 2.6-fold lower than for hGM-CSF I117, respectively (
Strikingly, the combination of the above-identified mutations at the following positions: VH-(S55 or S56) and T62 plus VL-S27e, N28, H34, N53 and H93, and pairing of the heavy and light chains in medium scale purified antibody batches reproducibly resulted in total issue resolution and an overall improvement in several respects of the new antibody compared to parental 4K21 (
In summary, a maximum of seven mutations including H93Y is sufficient to resolve all issues of 4K21, and lead to an overall improvement of the antibody. The sequences and alignments of the optimized 4K21 hVHs and hVLs are presented in
The humanized and optimized Fab molecules were showing a melting temperature superior to 84° C. compared to the chimeric 4K21 molecule (Tm=71° C.) as measured by Differential Scanning calorimetry (DSC) (
The 4K21 and reference antibodies were incubated at concentration below the expected Kd with serial dilutions of human GM-CSF I117 or T117, or macaca GM-CSF up to equilibrium. The remaining free antibody was measured in a Kinetic exclusion assay with the KINEXA® instrument. Kd was calculated by fitting the curve of the signal obtained versus the GM-CSF concentrations. The average of two experiments using different IgG concentrations was calculated. Compared to chimeric (MSB0010259) and humanized parental 4K21 (MSB0010439), the affinity of the optimized 4K21 antibody increased 3- to 5-fold with respect to hGM-CSF I117, 25- to 35-fold with respect to hGM-CSF T117, and 5- to 9-fold with respect to macaca GM-CSF. The affinity differences of the optimized 4K21 antibody between the 2 hGM-CSF isoforms and the macaca ortholog is reduced to less than 3-fold from up to 10-fold for the chimeric 4K21 (Table 3)). The kinetics data is in agreement with IC50 data in the TF-1 proliferation assay, i.e. the affinity of the optimized 4K21 antibodies 4K21 hVH S55K-T62S hVL N28I-H34S-N53R plus H93Y is almost equal for all GM-CSF variants (Kd=5-12 pM).
The potency of the optimized antibodies and the humanized parental antibody on the three GM-CSF targets was compared in this assay. Strikingly, the differences in potency observed with the humanized parental 4K21 antibody for the macaca and the human GM-CSF T117 compared to the human GM-CSF I117 target are absent for the optimized antibodies (
Epitope mapping based on a competition enzyme-linked immunosorbent assay (ELISA) protocol has shown that 4K21 interacts with hGM-CSF through one key residue at position 117. Substitution of this residue with alanine dramatically reduced binding of the parental chimeric or humanized 4K21 antibodies. Mapping the epitope of the lead optimized 4K21 mutants revealed that in all cases they bind equally well to the hGM-CSF A117 mutant as to the naturally occurring hGM-CSF isoforms (
Neutrophils from two donors were purified from buffy coats and induced with GM-CSF I117 or T117 in the presence of different anti-GM-CSF antibodies in dose response. After 2 days of incubation, CD11b cell-surface expression was measured by Fluorescence Activated Cell Sorting (FAGS). IC50 was single digit nM for optimized antibody 4K21 hVH S55K-T62S hVL N28I-H34S-N53R-H93Y, but not for humanized parental 4K21 on human GM-CSF T117, which was 10-fold worse than on GM-CSF I117 and than other antibodies (Table 5). Interestingly, only the optimized 4K21 antibody showed >80% maximum activity, whereas the humanized parental antibody was only partially inhibiting C011b upregulation.
The rhesus macaque monkey (Macaca mulatta) GM-CSF sequence was obtained from Swissprot (Q9GL44_MACMU; SEQ ID NO:29) and the Crab eating cynomolgus macaque monkey (Macaca fascicularis) GM-CSF was cloned from cynomolgus monkey lung RNA using RT-PCR primers that are conserved between the 5′ and 3′ untranslated regions of the human and rhesus GM-CSF mRNA sequences. The amino acid sequences of both rhesus and cynomolgus macaque GM-CSF orthologs are identical.
The human GM-CSF sequence was obtained from Swissprot (CSF2_HUMAN; SEQ ID NO:26) and blasted against the common marmoset (Callithrix jacchus) genomic sequences available in GenBank and translated in the 6 frames. The best score was obtained with the contig AC151012 (216132 bp). Using the GeneWise program (see e.g. Birney E, Clamp M, Durbin R. GeneWise and Genomewise. Genome Res. 2004; 14:988-995.), marmoset eDNA and protein sequences were predicted. The human and marmoset protein sequences were aligned and show an identity of 84%. The predicted marmoset protein sequence is now available in Uniprot (BOKWQ4_CALJA; SEQ ID NO:30).
In order to consolidate the gene prediction, the marmoset predicted sequence was aligned with human and rhesus sequences, these 3 sequences are very similar (see
Recombinant marmoset GM-CSF was shown to be active in the TF-1 proliferation assay with an EC50 of 90 ng/ml. The potency of the optimized 4K21 antibodies for the neutralization of the marmoset monkey GM-CSF was compared to that of the chimeric and humanized parental 4K21 antibodies. Strikingly, only the optimized 4K21 molecules (i.e. S55K, S55R and S56V hVH T62S variants+hVL N28I-H34S-N53R-H93Y with or without S27eY) were able to neutralize marmoset GM-CSF whereas the parental 4K21 was not (
For the anti-GM-CSF antibody humanized variant 4K21 hVH-S55K-T62S hVL-N28I-H34S-N53R-H93Y, greatly improved potency to GM-CSF I117 (5.3×), T117 (68×) and macaca (15× without S27eY, 33× with S27eY as compared to the parental antibody 4K21 has been determined. Further, the potency ratio T/I=0.9×, I117IC50=23 pM, T117IC50=21 pM, compared to the parental 4K21 antibody, which had a 11.6× difference for both SNPs.
Furthermore, a better potency on macaca GM-CSF of IC50=53 PM with a reduced difference to human I117 (2.3×), compared to the humanized parental 4K21 (6.6×), could be achieved. With the S27eY mutation, the IC50 was improved to 24 pM.
Further improved properties include the lack of the potential deamidation site in the light chain CDR1, a high thermal Fab stability (84.8° C.), and a single digit pM affinity for both human GM-CSF SNPs.
The antibodies of the invention are thus not binding anymore to human GM-CSF in a manner dependent to position 117, as exemplified by the 4K21 hVH-S55K-T62S hVL-N28I-H34SN53R-H93Y variant Moreover, these antibodies are thus fully potent on both I117 and T117 GM-CSF SNPs for the inhibition of GM-CSF induced CD11b upregulation in neutrophills, as well as fully cross-reactive to and potently neutralizing marmoset GM-CSF.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10191664 | Nov 2010 | EP | regional |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/EP2011/070375 | 11/17/2011 | WO | 00 | 5/20/2013 |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO2012/066089 | 5/24/2012 | WO | A |
| Number | Date | Country |
|---|---|---|
| WO 2008141391 | Nov 2008 | AU |
| 2008141391 | Nov 2008 | WO |
| 2009038760 | Mar 2009 | WO |
| 2009062238 | May 2009 | WO |
| Entry |
|---|
| Rudikoff et al., Proc Natl Acad Sci USA 79: 1979-1983, 1982. |
| Colman, Research in Immunology 145: 33-36, 1994. |
| Kussie et al., J. Immunol. 152: 146-152, 1994. |
| Chen et al., EMBO J., 14: 2784-2794, 1995. |
| Altschul, Stephen F. et al, “Gapped BLAST and PSI-BLAST: a new generation of protein database search programs,” Nucleic Acids Research (1997), vol. 25, No. 17, pp. 3389-3402. |
| Birney, Ewan et al., “GeneWise and Genomewise,” Genome Research (2004), vol. 14, pp. 988-995. |
| Greenfield, Edward A. (Editor), Antibodies: A Laboratory Manual (2nd Ed., 1988), Cold Spring Harbor Laboratory Press. |
| Kabat, Elvin A. et al., “Sequences of Proteins of Immunological Interest,” (5th Ed., 1991), U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health, NIH Publication No. 91-3242. |
| Sambrook, Joseph et al., “Molecular Cloning,” (3rd Ed., 2001), Cold Spring Harbor Laboratory Press. |
| International Search Report and Written Opinion of PCT/EP2011/070375 filed Nov. 17, 2011. |
| Number | Date | Country | |
|---|---|---|---|
| 20140065163 A1 | Mar 2014 | US |
