AU2013202859A1 - Crystalline anti-hTNFalpha antibodies - Google Patents

Abstract

CRYSTALLINE ANTI-HTNFALPHA ANTIBODIES Abstract The present invention relates to a batch crystallization method for crystallizing an anti-bTNFalpba antibody which allows the production of said antibody on an industrial scale; antibody crystals as obtained according to said method; compositions containing said crystals as well as methods of use of said crystals and compositions. (1l60686_2)KZA (1l60686_3)KZA

Description

Crystalline anti-hTNFalpha antibodies The present invention relates to a batch crystallization method for crystallizing an anti hTNFalpha antibody which allows the production of said antibody on an industrial 5 scale; antibody crystals as obtained according to said method; compositions containing said crystals as well as methods of use of said crystals and compositions. Technical Background 10 a) Antibody crystals With over 100 monoclonal antibodies currently being evaluated in clinical study phases 2 or 3, the mAb market can be considered one of the most promising biopharmaceuti cal markets. As these drugs have to be delivered in single doses often exceeding 100 15 mg, there is an urgent need to find suitable formulation strategies satisfying stability, safety and patient compliance. However, highly concentrated liquid mAb formulations show increased viscosity, hin dering syringe ability through patient friendly thin needles. Furthermore, the aggrega 20 tion tendency of mAb molecules at such high concentrations exponentially increases when compared to moderately concentrated solutions. This is unacceptable, in all means, regarding safety and stability requirements. Thus, the delivery of high mAb doses is restrained to large volumes, which generally 25 have to be delivered via infusion. This way of dosing is cost intensive and significantly reduces the patient's compliance. Therefore, pharmaceutically applicable low volume mAb crystal suspensions for subcu taneous injection would be highly desirable. Theoretically, degradation pathways influ 30 encing the mAb integrity should be significantly decelerated due to the rigidity of a crys tal lattice, where motions in the protein structure are hindered. Moreover, it can be ex pected that the increase in viscosity would be significantly reduced when comparing highly concentrated crystal suspensions with liquid formulations. With respect to sus tained release, it might be possible to generate or alter protein crystals in that way that 35 they dissolve slowly when brought into the patient's body. This would be a very elegant 1 way to deliver a sustained release formulation, as the extensive use of excipients and processes harming the mAb structure would be prevented. Despite the great potential in using protein crystals as drug substance, few attempts 5 have been made to systematically evaluate this strategy. A well-known exemption is insulin, which was successfully crystallized decades ago. Today, the use of crystal suspensions of insulin is well described, offering stable and long acting formulations being well established on the market. The discrepancy be 10 tween the development of insulin crystals and crystallization of all other proteins might be related to the fact that ordered insulin aggregates are natively formed in the pan creas. Thus, insulin crystals are easily obtained when insulin is brought in contact with an excess of zinc ions. Most other proteins tend to form unordered precipitates rather than crystals, and therefore, finding crystallization conditions for a protein is a time 15 consuming, non-trivial task. Despite a great interest in harvesting protein crystals for x-ray diffraction analysis, the quest of finding suitable crystallization conditions still is an empirical science, as in principle any protein behaves differently. To date, no general rule has been found 20 which might reliably predict by reason alone a successful crystallization condition for a protein of choice. Thus, obtaining crystals of a given protein always is referred to be the "bottle neck" of whatever intended application is planned later on. To make things even more challenging, antibodies are described to be especially hard 25 to crystallize, due to the flexibility of the molecule. Nevertheless, examples of immunoglobulin crystals have been known for a long time. The first example of immunoglobulin crystals were described 150 years ago by an Eng lish physician, Henry Bence Jones; he isolated crystals of an abnormal Ig light chain 30 dimer from the urine of a myeloma patient (Jones 1848). Such abnormal Igs have been known ever since as Bence Jones proteins. In 1938, the spontaneous crystallization of a distinct abnormal Ig from the serum of a myeloma patient was described (von Bons dorf, Groth et al. 1938), apparently an Ig heavy chain oligomer (MW 200 kDa). 2 Crystalline human immunoglobulins of normal structure (two heavy chains linked to two light chains) were described over the next thirty years, again mostly isolated from mye loma patients (Putnam 1955). Davies and co-workers were the first to characterize the structure of an intact human myeloma antibody, named "Dob", using x-ray crystallogra 5 phy (Terry, Matthews et al. 1968), and they determined its three-dimensional structure in 1971 (Sarma, Silverton et al. 1971). Their pioneering work was followed by that of others, yielding the crystal structures of the IgG "Kol" (Huber, Deisenhofer et al. 1976), the igG "Mcg"(Rajan, Ely et al. 1983), and a canine lymphoma IgG2a (Harris, Larson et al. 1992). 10 Crystals of immunoglobulins retain their distinctive immunological activities upon re dissolution. Nisonoff et al. reported in 1968 on a rabbit anti-p-azobenzoate antibody, "X4", that was easily crystallized. Antibody X4 was extensively characterized before crystallization as well as after re-dissolution of the crystals. [ 12 5 ]-p-iodobenzoate was 15 found to bind specifically and potently to re-dissolved X4; the re-dissolved crystals also exhibited multiple specific Ouchterlony immunodiffusion reactions typical of the unpuri fied rabbit serum (Nisonoff, Zappacosta et al. 1968). Connell and co-workers described a human myeloma gamma-immunoglobulin-1 kappa (IgG-K), called "Tem", that crystal lized spontaneously from serum at cold temperatures (Connell, Freedman et al. 1973). 20 Tem crystals were found to be well-formed and possessed rhombohedral symmetry. Tem-containing serum was extensively characterized by agarose immunodiffusion techniques. Electrophoresis and immunodiffusion of a re-dissolved solution of the Tern crystals showed them to be identical with the material obtained from the serum by cryoprecipitation, and with the isolated myeloma protein (Connell, Freedman et al. 25 1973). Mills and co-workers reported in 1983 an unusual crystallocryoglobulinemia resulting from human monoclonal antibodies to albumin (Mills, Brettman et al. 1983). Here, very similar cuboidal crystals were isolated from two patients. Redissolution of the crystals 30 followed by electrophoresis and immunoelectrophoresis indicated that the crystals were composed of two protein components, a monoclonal IgG-lambda and human serum albumin in a 1:2 ratio (Jentoft, Dearborn et al. 1982). The components were separated on preparative scale by dissolution of the original crystals followed by column chroma tography. Although neither separated component crystallized on its own, upon recom 35 bination the original bipartite complex reformed and then crystallized. Further study of 3 the distinctive sedimentation characteristics and immunological reactivity of the redis solved, separated IgG and its Fab fragment with human serum albumin indicated that reassociation of the two redissolved, separated components was immunologic in na ture, i.e. that the crystalline antibody once redissolved still possessed its native, highly 5 specific (for human serum albumin) binding characteristics (Mills, Brettman et al. 1983). Recently, Margolin and co-workers reported on the potential therapeutic uses of crys talline antibodies (Yang, Shenoy et al. 2003). They found that the therapeutic mono clonal antibody trastuzumab (Herceptin@) could be crystallized (Shenoy, Govardhan et 10 al. 2002). Crystalline trastuzumab suspensions were therapeutically efficacious in a mouse tumor model, thus demonstrating retention of biological activity by crystalline trastuzumab (Yang, Shenoy et al. 2003). b) Crystallization techniques 15 Unlike some other scientific or engineering endeavors, the crystallization of diverse proteins cannot be carried out successfully using defined methods or algorithms. Cer tainly, there have been great technical advances in the last 20-30 years, as noted by the world-renowned expert in protein crystallization, A. McPherson. McPherson pro 20 vides extensive details on tactics, strategies, reagents, and devices for the crystalliza tion of macromolecules. He does not, however, provide a method to ensure that any given macromolecule can indeed be crystallized by a skilled person with reasonable expectation of success. McPherson states for example : 'Whatever the procedure, no effort must be spared in refining and optimizing the parameters of the system, both 25 solvent and solute, to encourage and promote specific bonding interactions between molecules and to stabilize them once they have formed. This latter aspect of the prob lem generally depends on the specific chemical and physical properties of the particu lar protein or nucleic acid being crystallized." (McPherson 1999, p. 159) 30 It is widely accepted by those skilled in the art of protein crystallization that no algo rithm exists to take a new protein of interest, apply definite process steps, and thereby obtain the desired crystals. Several screening systems a commercially available (for example Hampton 1 and 2, 35 Wizzard I and 1l) which allow, on a microliter scale, to screen for potentially suitable 4 crystallization conditions for a specific protein. However, positive results obtained in such a screening system do not necessarily allow successful crystallization in a larger, industrially applicable batch scale. Conversion of microliter-size crystallization trials into industrial dimensions is described to be a challenging task (see Jen et al., 2001). 5 Baldock et al (1996) reported on a comparison of microbatch and vapor diffusion for initial screening of crystallization conditions. Six commercially available proteins were screened using a set of crystallization solutions. The screens were performed using the most common vapor diffusion method and three variants of a microbatch crystalli 10 zation method, including a novel evaporation technique. Out of 58 crystallization condi tions identified, 43 (74%) were identified by microbatch, while 41 (71 %) were identified by vapor diffusion. Twenty-six conditions were found by both methods, and 17 (29%) would have been missed if microbatch had not been used at all. This shows that the vapor diffusion technique, which is most commonly used in initial crystallization screens 15 does not guarantee positive results. c) hTNFalpha antibody crystals Human TNFalpha (hTNFalpha) is considered as a causative agent of numerous dis 20 eases. There is, therefore, a great need for suitable methods of treating such hTNFal pha related disorders. One promising therapeutic approach comprises the administra tion of pharmaceutically effective doses of anti-human TNFalpha antibodies. Recently one such antibody, designated D2E7, or generically adalimumab, has been put on the market and is commercialised under the trade name HUMIRA@. 25 WO-A-02/072636 disclosed the crystallization of the whole, intact antibodies Rituximab, Infliximab and Trastuzumab. Most of the crystallization experiments were performed with chemicals with unclear toxicity, like imidazole, 2-cyclohexyl-ethanesulfonate (CHES), methylpentanediol, copper sulphate, and 2-morpholino-ethanesulfonate 30 (MES). Most of the examples used seed crystals to initiate crystallization. WO-A-2004/009776 disclosed crystallization experiments in the microliter scale using the sitting drop vapor diffusion technique by mixing equal volumes (1 pl) of different crystallization buffers and D2E7 F(ab)' 2 or Fab fragments. While several experimental 5 conditions were reported for each of said fragments, no successful crystallization of the whole, intact D2E7 antibody was reported. Methods for preparing crystals of any given anti-human TNFalpha whole antibodies, in 5 particular of D2E7, therefore are not available. The problem to be solved according to the present invention is, therefore, to develop suitable batch crystallization conditions for anti-hTNFalpha antibodies, in particular for the human anti-hTNFalpha antibody D2E7, and to establish crystallization process 10 conditions applicable to volumes relevant for industrial antibody crystal production. At the same time a crystallization process should be established that does not make use of toxic agents, which might negatively affect the pharmaceutical applicability of such antibodies. 15 Summary of the Invention: The above-mentioned problem was, surprisingly, solved by the finding that it is possible to obtain crystals of a whole anti-hTNFalpha antibody in batch crystallization volumes above the microliter scale by applying physiologically acceptable inorganic phosphate 20 salts as the crystallization-inducing agent. Preferred Embodiments In a first aspect the invention relates to a batch crystallization method for crystallizing 25 an anti-hTNFalpha antibody, comprising the steps of: a) providing an aqueous solution of said antibody in admixture with an inorganic phosphate salt as crystallization agent, as for example by mixing an aqueous so lution of said antibody, wherein the antibody preferably is present in dissolved form, with an aqueous crystallization solution comprising an inorganic phosphate 30 salt as crystallization agent in dissolved form, or alternatively by adding said crys tallization agent in solid form; and b) incubating said aqueous crystallization mixture until crystals of said antibody are formed. 6 The crystallization method of the invention generally is performed at a pH of said aque ous crystallization mixture in the range of about pH 3 to about 5, in particular about 3.5 to about 4.5, or about 3.7 to about 4.2. 5 Moreover, said aqueous crystallization mixture may contain at least one buffer. Said buffer may, in particular, comprise an acetate component as main component, espe cially an alkali metal salt, in particular, sodium acetate. Said salt is adjusted by addition of an acid, in particular acetic acid, to the required pH. In a preferred embodiment of the crystallization method, the buffer concentration (total acetate) in said aqueous crys 10 tallization mixture is 0 to about 0.5 M, or about 0.02 to about 0.5 M, as for example about 0.05 to about 0.3 M, or about 0.15 to about 0.2 M. In a further particular embodiment of the crystallization method according to the inven tion, the phosphate salt used as the precipitating agent is selected from hydrogen 15 phosphate salts, such as mono- or dihydrogenphosphate salts, in particular an ammo nium salt or an alkali metal salt, for example a salt containing Na* or K' ions, or a mix ture thereof comprising of at least two different salts. Suitable examples are: NaH 2

PO

4 , Na 2

HPO

4 , KH 2

PO

4 , K 2

HPO

4 , NH 4

H

2

PO

4 , (NH 4

)

2 HP0 4 , and mixtures thereof. 20 In particular, the phosphate salt concentration in the crystallization mixture is in the range of about 1 to about 6 M, for example a range of about 1.0 to about 4.0 M, or about 1.0 to about 3.0 M, or about 1.5 to about 2.8 M, or about 2.0 to about 2.6M. In a preferred embodiment of the invention, protein solution and crystallization solution 25 are combined in a ratio of about 1:1. Thus, molarities of the buffering agents / crystalli zation agents in the original crystallization solution are about double as high as in the crystallization mixture. Typically, the crystallization method is performed in a batch volume in the range of 30 about 1 ml to about 20000 1, or 1 ml to about 15000 1, or 1 ml to about 12000 1, or about 1 ml to about 10000 1, or 1 ml to about 6000 1, or 1 ml to about 3000 1, or 1 ml to about 1000 1, or 1 ml to about 100 1, as for example about 50 ml to about 8000 ml, or about 100 ml to about 5000 ml, or about 1000 ml to about 3000 ml; or about 1 1 to about 1000 1; or about 10 I to about 500 1. 35 7 In addition, the crystallization method of the invention may be performed so that at least one of the following additional crystallization conditions is achieved: a) incubation is performed for between about 1 hour to about 60 days, for example about 1 to about 30 days, or about 2 to 10 days; 5 b) incubation is performed at a temperature between about 0 0C and about 50 OC, for example about 4 0C and about 37 OC; c) the antibody concentration (i.e. protein concentration) in the crystallization mix ture is in the range of about 1 to 200 mg/ml or 1 to 100 mg/ml, for example 1.5 to 20 mg/ml, in particular in the range of about 2 to 15 mg/ml, or 5 to 10 mg/ml. The 10 protein concentration may be determined according to standard procedures for protein determination. According to a particularly preferred method, crystallization is performed under the fol lowing conditions of the crystallization mixture: 15 Phosphate salt: NaH 2

PO

4 , 1.5 to 2.5 M buffer: total acetate, 0 to 0.3 M pH: 3.6 to 4.2 anti-hTNFalpha concentration: 3 to 10 mg/ml 20 Temperature: 18 to 24 0C Batch volume: 1 to 100 1 Agitation: None Duration: 4 to 15 days 25 The crystallization mixtures as outlined above are usually obtained by adding a crystal lization agent in solution or as solid to the protein solution. Both solutions may be, but do not have to be buffered. Crystallization agent molarity and buffer molarity in the original crystallization solution is usually higher than in the crystallization mixture as it is 30 "diluted" with the protein solution. In a further embodiment, the crystallization method of the invention may further com prise the step of drying the obtained crystals. Suitable drying methods comprise evapo rative drying, spray drying, lyophilization, vacuum drying, fluid bed drying, spray freeze 35 drying, near critical drying, supercritical drying, and nitrogen gas drying. 8 In a further embodiment, the crystallization method of the invention may further com prise the step of exchanging the crystallization mother liquor with a different buffer, e.g. a buffer containing polyethylene glycol (PEG) with a molar mass in the range of about 5 300 to 8000 Daltons or mixtures of PEGs, by centrifugation, diafiltration, ultrafiltration or other commonly used buffer exchange techniques. The present invention also relates to a crystal of an anti-hTNFalpha antibody, obtain able by a crystallization method as defined above and in general to crystals of an anti 10 hTNFalpha antibody The crystals of the invention are typically characterized by a needle-like morphology with a maximum length I of about 2 - 500 pm or about 100 - 300 pm and an l/d ratio of about 3 to 30, but may also have other geometrical appearances 15 Said crystal may be obtained from a polyclonal antibody or, preferably, a monoclonal antibody. In particular, said antibody is selected from the group consisting of: non-chimeric or 20 chimeric antibodies, humanized antibodies, non-glycosylated antibodies, human anti bodies and mouse antibodies. In particular the antibody to be crystallized is a non chimeric, human antibody optionally further processed for improving the antigen binding. 25 Preferably, said crystals are obtained from an igG antibody such as, for example, an IgG1, IgG2, IgG3 or IgG4 antibody. In particular, said antibody is a whole anti-human TNFalpha antibody of the group IgG1. In a preferred embodiment, the crystals are prepared from an isolated human antibody, 30 that dissociates from hTNFalpha with a Kd of 1 x10 8 M or less, more preferably 1x 10-9 M or less, and even more preferably 5 x 10-10 M or less, and a Ka rate constant of 1 x 10' s' or less, both determined by surface plasmon resonance, and neutralizes hTNFalpha cytotoxicity in a standard in vitro L929 assay with an IC50 of 1 x10- 7 M or less. 35 9 In particular, said crystals may be prepared from an isolated human antibody with the following characteristics: a) dissociates from human TNFalpha with a koff rate constant of 1 x 10- s-1 or less, as determined by surface plasmon resonance; b) has a light chain CDR3 domain comprising the amino acid sequence of SEQ ID NO: 3, or modified 5 from SEQ ID NO: 3 by a single alanine substitution at position 1, 4, 5, 7 or 8, or by one to five conservative amino acid substitutions at positions 1, 3, 4, 6, 7, 8 and/or 9; c) has a heavy chain CDR3 domain comprising the amino acid sequence of SEQ ID NO: 4, or modified from SEQ ID NO: 4 by a single alanine substitution at position 2, 3, 4, 5, 6, 8, 9, 10 or 11, or by one to five conservative amino acid substitutions at positions 2, 3, 4, 10 5, 6, 8, 9, 10, 11 and/or 12. More preferably, the antibody, or antigen-binding portion thereof, dissociates from hu man TNFalpha with a koff of 5 x 10-4 s-1 or less. Even more preferably, the antibody, or antigen-binding portion thereof, dissociates from human TNFalpha with a koff of 1 x 15 10-4 s-1 or less. In a particularly preferred embodiment, said crystals are prepared from an isolated hu man antibody with a light chain variable region (LCVR) comprising the amino acid se quence of SEQ ID NO: 1 and a heavy chain variable region (HCVR) comprising the 20 amino acid sequence of SEQ ID NO: 2. Most preferred are crystals prepared from the antibody D2E7, as disclosed in WO-A 97/29131 or a functional equivalent thereof. Said antibody is recombinantly produced in Chinese hamster ovary cells and comprises a heavy chain sequence according to SEQ 25 ID NO:6 and a light chain sequence according to SEQ ID NO: 5. In a further embodiment, the invention relates to a solid, liquid or semi-solid pharma ceutical composition comprising: (a) crystals of an anti-hTNFalpha antibody as defined in any one of claims 15 to 26, and (b) at least one pharmaceutically acceptable excipi 30 ent stably maintaining the antibody crystals. Another aspect of this invention relates to a solid, liquid or semi-solid pharmaceutical composition comprising: (a) crystals of an anti-hTNFalpha antibody as defined herein, and (b) at least one pharmaceutically acceptable excipient encapsulating or embedding 35 said antibody crystals. The composition may further comprise (c) at least one pharma 10 ceutically acceptable excipient stably maintaining the antibody crystals. Moreover, en capsulation and embedding may be implemented in conjunction. In particular, said compositions may have an antibody crystal concentration higher than 5 about 1 mg/ml, in particular about 200 mg/ml or more, for example about 200 to about 600 mg/ml, or about 300 to about 500 mg/ml. Said excipients may comprise at least one polymeric, optionally biodegradable carrier or at least one oil or lipid carrier. 10 Said polymeric carrier may be a polymer selected from one or more of the group con sisting of: poly (acrylic acid), poly (cyanoacrylates), poly (amino acids), poly (anhy drides), poly (depsipeptide), poly (esters), poly (lactic acid), poly (lactic-co-glycolic acid) or PLGA, poly (B-hydroxybutryate), poly (caprolactone), poly (dioxanone); poly (ethyl 15 ene glycol), poly (hydroxypropyl) methacrylamide, poly (organo) phosphazene, poly (ortho esters), poly (vinyl alcohol), poly (vinylpyrrolidone), maleic anhydride alkyl vinyl ether copolymers, pluronic polyols, albumin, alginate, cellulose and cellulose deriva tives, collagen, fibrin, gelatin, hyaluronic acid, oligosaccharides, glycaminoglycans, sulfated polysaccharides, blends and copolymers thereof. 20 Said oil (or oily liquid) may be an oil (or oily liquid) selected from one or more of the group consisting of: oleaginous almond oil, corn oil, cottonseed oil, ethyl oleate, isopro pyl myristate, isopropyl palmitate, mineral oil, light mineral oil, octyldodecanol, olive oil, peanut oil, persic oil, sesame oil, soybean oil, squalane, liquid triglycerides, liquid 25 waxes, higher alcohols. Said lipid carrier may be a lipid selected from one or more of the group consisting of: fatty acids and salts of fatty acids, fatty alcohols, fatty amines, mono-, di-, and triglyc erides of fatty acids, phospholipids, glycolipids, sterols and waxes and related similar 30 substances. Waxes are further classified in natural and synthetic products. Natural ma terials include waxes obtained from vegetable, animal or minerals sources such as beeswax, carnauba or montanwax. Chlorinated naphthalenes and ethylenic polymers are examples for synthetic wax products. 11 In a preferred embodiment, said composition is an injectable composition comprising anti-hTNFalpha antibody crystals as defined above and having an antibody crystal concentration in the range of about 10 to about 400 or about 50 to about 300 mg/mL. 5 In a further aspect the invention relates to a crystal slurry comprising anti-hTNFalpha antibody crystals as defined above having an antibody crystal concentration higher than about 100 mg/ml, for example about 150 to about 600 mg/ml, or about 200 to about 400 mg/ml. 10 The present invention also relates to a method for treating a mammal comprising the step of administering to the mammal an effective amount of whole anti-hTNFalpha an tibody crystals as defined above or an effective amount of the composition as defined above. Preferably, said composition is administered by parenteral route, oral route, or by injection. 15 Furthermore, the present invention relates to a method of treating a hTNFalpha-related disorder in a subject that comprises administering a therapeutically effective amount of antibody crystals as defined above. 20 In particular, said hTNFalpha-related disorder is selected from: an autoimmune disease, in particular rheumatoid arthritis, rheumatoid spondylitis, os teoarthritis and gouty arthritis, an allergy, multiple sclerosis, autoimmune diabetes, autoimmune uveitis and nephrotic syndrome; an infectious disease, transplant rejection or graft-versus-host disease, malignancy, pulmonary disorder, intestinal disorder, car 25 diac disorder, inflammatory bone disorders, bone resorption disease, alcoholic hepati tis, viral hepatitis, fulminant hepatitis, coagulation disturbances, burns, reperfusion in jury, keloid formation, scar tissue formation, pyrexia, periodontal disease, obesity and radiation toxicity; a spondyloarthropathy, a pulmonary disorder, a coronary disorder, a metabolic disorder, anemia, pain, a hepatic disorder, a skin disorder, a nail disorder, or 30 vasculitis, Behcet's disease, ankylosing spondylitis, asthma, chronic obstructive pulmo nary disease (COPD), idiopathic pulmonary fibrosis (IPF), restenosis, diabetes, ane mia, pain, a Crohn's disease-related disorder, juvenile rheumatoid arthritis (JRA), a hepatitis C virus infection, psoriasis, psoriatic arthritis, chronic plaque psoriasis, age related cachexia, Alzheimer's disease, brain edema, inflammatory brain injury, chronic 35 fatigue syndrome, dermatomyositis, drug reactions, edema in and/or around the spinal 12 cord, familial periodic fevers, Felty's syndrome, fibrosis, glomerulonephritides (e.g. post-streptococcal glomerulonephritis or IgA nephropathy), loosening of prostheses, microscopic polyangiitis, mixed connective tissue disorder, multiple myeloma, cancer and cachexia, multiple organ disorder, myelo dysplastic syndrome, orchitism osteoly 5 sis, pancreatitis, including acute, chronic, and pancreatic abscess, periodontal disease polymyositis, progressive renal failure, pseudogout, pyoderma gangrenosum, relapsing polychondritis, rheumatic heart disease, sarcoidosis, sclerosing cholangitis, stroke, thoracoabdominal aortic aneurysm repair(TAAA), TNF receptor associated periodic syndrome (TRAPS), symptoms related to Yellow Fever vaccination, inflammatory dis 10 eases associated with the ear, chronic ear inflammation, or pediatric ear inflammation, uveitis, sciatica, prostatitis, endometriosis, choroidal neovascularization, lupus, Sjogren's syndrome, and wet macular degeneration. Moreover, the present invention relates to the use of whole anti-hTNFalpha antibody 15 crystals as defined above for preparing a pharmaceutical composition for treating a hTNFalpha-related disease as defined above. Finally, the present invention provides anti-hTNFalpha antibody crystals as defined above for use in medicine. 20 Description of Figures Fig. 1: D2E7 crystals from Example 37 after 6 days. 25 Fig. 2: D2E7 crystals manufactured in 1 mL batch volume, ambient temperature. Fig. 3: D2E7 crystals manufactured in 50 mL batch volume, ambient temperature. 30 Fig. 4: D2E7 crystals manufactured in 10 L batch volume, ambient temperature. Fig. 5: D2E7 crystals manufactured according to the invention and birefringence thereof. Fig. 6: D2E7 crystal suspensions at different concentrations injected via different 35 gauge needles. 13 Fig. 7: FT-IR analysis of D2E7 crystal suspension. Detailed Description of the Invention 5 A. Definitions A "batch method of crystallization" comprises the step of adding the crystallization solu tion comprising the crystallization agent, preferably in dissolved form, to the solution of 10 the antibody to be crystallized. A "micro scale crystallization method", which may for example be based upon vapor diffusion, comprises the steps of admixing a small volume of antibody solution in the microliter range with a reservoir buffer containing a crystallization agent; placing a 15 droplet of said mixture in a sealed container adjacent to an aliquot of said reservoir buffer; allowing exchange of solvent between the droplet and the reservoir by vapor diffusion, during which the solvent content in said droplet changes and crystallization may be observed if suitable crystallization conditions are reached. 20 A "crystallization agent", in the present case a phosphate salt, favors crystal formation of the antibody to be crystallized. A "crystallization solution" contains said crystallization agent in dissolved form. Pref erably said solution is an aqueous system, i.e. the liquid constituents thereof predomi 25 nantly, consist of water. As for example, 80 to 100 wt.-% or 95 to 100 wt.-% or 98 to 100 wt.-% may be water. Antibody "crystals" are one form of the solid state of matter of said protein, which is distinct from a second solid form, i.e. the amorphous state, which exists essentially as 30 an unorganized, heterogeneous solid. Crystals have a regular three-dimensional struc ture, typically referred to as a lattice. An antibody crystal comprises a regular three dimensional array of antibody molecules. See Giege, R. and Ducruix, A. Barrett, Crys tallization of Nucleic Acids and Proteins, a Practical Approach, 2nd ed., pp. 1-16, Ox ford University Press, New York (1999). 35 14 A "whole" or "intact" anti-hTNFalpha antibody as crystallized according to this invention, is a functional antibody that is able to recognize and bind to its antigen human TNFal pha in vitro and/or in vivo. The antibody may initiate subsequent immune system reac tions of a patient associated with antibody-binding to its antigen, in particular Direct 5 Cytotoxicity, Complement-Dependent Cytotoxicity (CDC), and Antibody-Dependent Cytotoxicity (ADCC). The antibody molecule has a structure composed of two identical heavy chains (MW each about 50 kDa) covalently bound to each other, and two identi cal light chains (MW each about 25 kDa), each covalently bound to one of the heavy chains. The four chains are arranged in a classic "Y" motif. Each heavy chain is com 10 prised of a heavy chain variable region (abbreviated herein as HCVR or VH) and a heavy chain constant region. The heavy chain constant region is comprised of three domains, CH1, CH2 and CH3. Each light chain is comprised of a light chain variable region (abbreviated herein as LCVR or VL) and a light chain constant region. The light chain constant region is comprised of one domain, CL. The VH and VL regions can be 15 further subdivided into regions of hypervariability, termed complementarity determining regions (CDR), interspersed with regions that are more conserved, termed framework regions (FR). Each VH and VL is composed of three CDRs and four FRs, arranged from amino-terminus to carboxy-terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. The complete antibody molecule has two antigen binding 20 sites, i.e. is "bivalent". The two antigen binding sites are specific for one hTNFalpha antigen, i.e. the antibody is "mono-specific". "Monoclonal antibodies" are antibodies that are derived from a single clone of B lym phocytes (B cells), and recognize the same antigenic determinant. Whole monoclonal 25 antibodies are those that have the above-mentioned classic molecular structure that includes two complete heavy chains and two complete light chains. Monoclonal anti bodies are routinely produced by fusing the antibody-producing B cell with an immortal myeloma cell to generate B cell hybridomas, which continually produce monoclonal antibodies in cell culture. Other production methods are available, as for example ex 30 pression of monoclonal antibodies in bacterial, yeast, insect, or mammalian cell culture using phage-display technology; in vivo production in genetically modified animals, such as cows, goats, pigs, rabbits, chickens, or in transgenic mice which have been modified to contain and express the entire human B cell genome; or production in ge netically modified plants, such as tobacco and corn. Anti-hTNFalpha antibodies from all 35 such sources may be crystallized according to this invention. 15 The monoclonal antibodies to be crystallized according to the invention include "chi meric" anti-hTNFalpha antibodies in which a portion of the heavy and/or light chain is identical with or homologous to corresponding sequences in antibodies derived from a 5 particular species or belonging to a particular antibody class or subclass, while the re mainder of the chain(s) is identical with or homologous to corresponding sequences in antibodies derived from another species or belonging to another antibody class or sub class. As an example there may be mentioned a mouse/human chimera containing variable antigen-binding portions of a murine antibody and constant portions derived 10 from a human antibody. "Humanized" forms of non-human (e.g. murine) anti-hTNFalpha antibodies are also encompassed. These are chimeric antibodies that contain minimal sequence derived from a non-human immunoglobulin. For the most part, humanized antibodies are hu 15 man immunoglobulins in which residues from a complementarity determining region (CDR) or hypervariable loop (HVL) of the human immunoglobulin are replaced by resi dues from a CDR or HVL of a non-human species, such as mouse, rat, rabbit or non human primate, having the desired functionality. Framework region (FR) residues of the human immunoglobulin may replaced by corresponding non-human residues to 20 improve antigen binding affinity. Furthermore, humanized antibodies may comprise residues that are found neither in the corresponding human or non-human antibody portions. These modifications may be necessary to further improve antibody efficacy. A "human antibody" or "fully human antibody" is one, which has an amino acid se 25 quence which corresponds to that of an antibody produced by a human or which is recombinantly produced. The term "human antibody", as used herein, is intended to include antibodies having variable and constant regions derived from human germline immunoglobulin sequences. The human antibodies of the invention may include amino acid residues not encoded by human germline immunoglobulin sequences (e.g. muta 30 tions introduced by random or site-specific mutagenesis in vitro or by somatic mutation in vivo), for example in the CDRs and in particular CDR3. However, the term "human antibody", as used herein, is not intended to include antibodies in which CDR se quences derived from the germline of another mammalian species, such as a mouse, have been grafted onto human framework sequences. 35 16 The term "recombinant human antibody", as used herein, is intended to include all hu man antibodies that are prepared, expressed, created or isolated by recombinant means, such as antibodies expressed using a recombinant expression vector trans fected into a host cell, antibodies isolated from a recombinant, combinatorial human 5 antibody library, antibodies isolated from an animal (e.g. a mouse) that is transgenic for human immunoglobulin genes (see e.g. Taylor, L.D., et al. (1992) Nucl. Acids Res. 20:6287-6295) or antibodies prepared, expressed, created or isolated by any other means that involves splicing of human immunoglobulin gene sequences to other DNA sequences. Such recombinant human antibodies have variable and constant regions 10 derived from human germline immunoglobulin sequences. In certain embodiments, however, such recombinant human antibodies are subjected to in vitro mutagenesis (or, when an animal transgenic for human Ig sequences is used, in vivo somatic mutagenesis) and thus the amino acid sequences of the VH and VL regions of the re combinant antibodies are sequences that, while derived from and related to human 15 germline VH and VL sequences, may not naturally exist within the human antibody germline repertoire in vivo. A "neutralizing antibody", as used herein (or an "antibody that neutralized hTNFalpha activity"), is intended to refer to an antibody whose binding to hTNFalpha results in in 20 hibition of the biological activity of hTNFalpha. This inhibition of the biological activity of hTNFalpha can be assessed by measuring one or more indicators of hTNFalpha bio logical activity, such as hTNFalpha-induced cytotoxicity (either in vitro or in vivo), hTNFalpha-induced cellular activation and hTNFalpha binding to hTNFalpha receptors. These indicators of hTNFalpha biological activity can be assessed by one or more of 25 several standard in vitro or in vivo assays known in the art. Preferably, the ability of an antibody to neutralize hTNFalpha activity is assessed by inhibition of hTNFalpha induced cytotoxicity of L929 cells. As an additional or alternative parameter of hTNFaI pha activity, the ability of an antibody to inhibit hTNFalpha-induced expression of ELAM-1 on HUVEC, as a measure of hTNFalpha-induced cellular activation, can be 30 assessed. An "affinity matured" anti-hTNFalpha antibody is one with one or more alterations in one or more hypervariable regions, which result in an improvement in the affinity of the antibody for antigen, compared to a parent antibody. Affinity matured antibodies will 35 have nanomolar or even picomolar affinities values for the target antigen. Affinity ma 17 tured antibodies are produced by procedures known in the art. Marks et al., Bio/Technology 10:779-783 (1992) describes affinity maturation by VH and VL domain shuffling. Random mutagenesis of CDR and/or framework residues is described by: Barbas et al., Proc. Nat. Acad. Sci. USA 91:3809-3813 (1994); Scier et al., Gene 5 169:147-155 (1995); Yelton et al., J. Immunol. 155:1994-2004 (1995); Jackson et al., J. Immunol. 154(7):3310-9 (1995); and Hawkins et al., J. Mol Biol. 226:889-896 (1992). An "isolated antibody", as used herein, is intended to refer to an antibody that is sub stantially free of other antibodies having different antigenic specificities (e.g. an isolated 10 antibody that specifically binds hTNFalpha is substantially free of antibodies that spe cificallybind antigens other than hTNFalpha). An isolated antibody that specifically binds hTNFalpha may, however, have cross-reactivity to other antigens, such as hTNFalpha molecules from other species. Moreover, an isolated antibody may be sub stantially free of other cellular material and/or chemicals. 15 The term "human TNFalpha" (abbreviated herein as hTNFalpha, or simply hTNF), as used herein, is intended to refer to a human cytokine that exists as a 17 kDa secreted form and a 26 kDa membrane-associated form, the biologically active form of which is composed of a trimer of noncovalently bound molecules. The structure of hTNFalpha 20 is described further in, for example, Pennica, D., et al. (1984) Nature 312:724-729; Davis, J.M., et al. (1987) Biochemistry 26:1322-1326; and Jones, E.Y., et al. (1989) Nature 338:225-228. The term human TNFalpha is intended to include recombinant human TNFalpha (rhTNFalpha), which can be prepared by standard recombinant ex pression methods or purchased commercially (R & D Systems, Catalog No. 210-TA, 25 Minneapolis, MN). The term "koff", as used herein, is intended to refer to the off rate constant for dissocia tion of an antibody from the antibody/antigen complex. 30 The term "Kd", as used herein, is intended to refer to the dissociation constant of a par ticular antibody-antigen interaction. A "functional equivalent" of a specific parentf' anti-hTNFalpha antibody as crystallized according to the invention is one which shows the same antigen-specificity, differs 35 however with respect to the molecular composition of the "parent" antibody on the 18 amino acid level or glycosylation level. Said differences, however, may be merely such that the crystallization conditions do not deviate from the parameter ranges as dis closed herein. 5 "Encapsulation" of antibody crystals refers to a formulation where the incorporated crystals are individually coated by at least one layer of a coating material. In a preferred embodiment, such coated crystals may have a sustained dissolution rate. "Embedding" of antibody crystals refers to a formulation where the crystals, which 10 might be encapsulated or not, are incorporated into a solid, liquid or semi-solid carrier in a disperse manner. Such embedded crystallized antibody molecules may be re leased or dissolved in a controlled, sustained manner from the carrier. B. Method of crystallization 15 The crystallization method of the invention is in principle applicable to any anti hTNFalpha antibody. Said antibody may be a polyclonal antibody or, preferably, a monoclonal antibody. Said antibody may be chimeric antibodies, humanized antibod ies, human antibodies or non-human, as for example mouse antibodies, each in glyco 20 sylated or non-glycosylated form. In particular the method is applicable to D2E7 and functional equivalents thereof. Preferably said anti-hTNFalpha antibody is an IgG antibody, in particular an anti human TNFalpha antibody of the group IgG1. 25 Unless otherwise stated the crystallization method of the invention makes use of tech nical equipment, chemicals and methodologies well known in the art. However, as ex plained above, the present invention is based on the surprising finding that the selec tion of specific crystallization conditions, in particular, the selection of specific crystalli 30 zation agents, optionally further combined with specific pH conditions and/or concentra tion ranges of the corresponding agents (buffer, antibody, crystallization agent), allows for the first time to prepare reproducibly and in a large scale stable crystals of antibod ies, in particular non-chimeric, human antibodies, directed against hTNF alpha, which can be further processed to form an active ingredient of a superior, highly advanta 35 geous pharmaceutical composition. 19 The starting material for performing the crystallization method normally comprises a concentrated solution of the antibody to be crystallized. The protein concentration may, for example, be in the range of about 5 to 75 mg/ml. Said solution may contain addi 5 tives stabilizing said dissolved antibody, and it may be advisable to remove said addi tives in advance. This can be achieved by performing a buffer exchange step. Preferably said starting material for performing the crystallization contains the antibody in an aqueous solution, having a pH adjusted in the range of about 3.2 to 8.2, or about 10 4.0 to 8.0, in particular about 4.5 to 6.5, preferably around 5.0 to 5.5. The pH may be adjusted by means of a suitable buffer applied in a final concentration of about 1 to 50 mM, in particular about 1 to 10 mM. The solution may contain additives, as for example in a proportion of about 0.01 to 15, or 0.1 to 5, or 0.1 to 2 wt.-% based on the total weight of the solution, like salts, sugars, sugar alcohols and surfactants, in order to 15 further stabilize the solution. The excipients should preferably be selected from physio logically acceptable compounds, routinely applied in pharmaceutical preparations. As non-limiting examples there may be mentioned salts, like NaCl;; surfactants, like poly sorbate 80 (Tween 80), polysorbate 20 (Tween 20); sugars, like sucrose, trehalose; sugar alcohols, like mannitol, sorbitol; and buffer agents, like phosphate-based buffer 20 systems, as sodium and potassium hydrogen phosphate buffers as defined above, acetate buffer, phosphate buffer, citrate buffer, TRIS buffer, maleate buffer or succinate buffer, histidine buffer; amino acids, like histidine, arginine and glycine. The buffer exchange may be performed by means of routine methods, for example 25 dialysis or ultrafiltration. The initial protein concentration of the aqueous solution used as starting material should be in the range of about 0.5 to about 200 or about 1 to about 50 mg/ml. 30 Depending on the intended final batch size (which may be in the range of 1 ml to 20000 (twenty thousand) litres) an initial volume of said aqueous antibody solution is placed in an appropriate container (as for example a vessel, bottle or tank) made of inert mate rial, as for example glass, polymer or metal. The initial volume of said aqueous solution may correspond to about 30 to 80%, normally about 50% of the final batch size. 35 20 If necessary the solution after having been filled into said container will be brought to standardized conditions. In particular, the temperature will be adjusted in the range of about 4 0 C and about 37 0C. 5 Then the crystallization solution, containing the crystallization agent in an appropriate concentration, optionally pre-conditioned in the same way as the antibody solution, is added to the antibody solution. The addition of the crystallization solution is performed continuously or discontinuously 10 optionally under gentle agitation in order to facilitate mixing of the two liquids. Prefera bly the addition is performed under conditions where the protein solution is provided under agitation and the crystallization solution (or agents in its solid from) is / are added in a controlled manner. 15 The formation of the antibody crystals is initiated by applying a phosphate salt, in par ticular a hydrogen phosphate salt, and preferably an alkali metal salt, or a mixture of at least two different alkali metal salts as defined above as the crystallization agent. The crystallization solution contains the agent in a concentration which is sufficient to afford a final concentration of the phosphate salt in said crystallization mixture in the range of 20 about 1 to 6 M. Preferably, the crystallization solution additionally contains an acidic buffer, i.e. different from that of the antibody solution, in a concentration suitable to allow the adjustment of the pH of the crystallization mixture in the range of about 3 to 5. 25 After having finished the addition of said crystallization solution, the thus obtained mix ture may be further incubated for about 1 hour to about 60 days in order to obtain a maximum yield of antibody crystals. If appropriate, the mixture may, for example, be agitated, gently stirred, rolled or moved in a manner known per se. 30 Finally, the crystals obtained may be separated by known methods, for example filtra tion or centrifugation, as for example by centrifugation at about 200 - 20000 rpm, pref erably 500 - 2000 rpm, at room temperature or 40C. The remaining mother liquor may be discarded or further processed. 35 21 If necessary, the thus isolated crystals may be washed and subsequently dried, or the mother liquor can be exchanged by a different solvent system suitable for storage and /or final use of the antibodies suspended therein. 5 Antibody crystals formed according to the present invention may vary in their shape. Shapes typically may include needles, cone-like, spherical and sea urchin like shapes. The size of the crystals can be on the order of higher nm to mm size (as for example length). In some embodiments, the crystals are at least about 10 pm in size, and may be visible to the naked eye. For therapeutic administration, the size of the crystals will 10 vary depending on the route of administration, for example, for subcutaneous admini stration the size of the crystals may be larger than for intravenous administration. The shape of the crystals may be altered by adding specific additional additives to the crystallization mixture, as has been previously described for both protein crystals and 15 crystals of low molecular weight organic and inorganic molecules. If necessary, it may be verified that the crystals are in fact crystals of said antibody. Crystals of an antibody can be analyzed microscopically for birefringence. In general, crystals, unless of cubic internal symmetry, will rotate the plane of polarization of polar 20 ized light. In yet another method, crystals can be isolated, washed, resolubilized and analyzed by SDS-PAGE and, optionally, stained with an anti-Fc receptor antibody. Op tionally, the resolubilized antibody can also be tested for binding to its hTNFalpha utiliz ing standard assays. 25 Crystals as obtained according to the invention may also be crosslinked to one an other. Such crosslinking may enhance stability of the crystals. Methods for crosslinking crystals described, for example, in U.S. Patent No. 5,849,296. Crystals can be crosslinked using a bifunctional reagent such as glutaraldehyde. Once crosslinked, crystals can be lyophilized and stored for use, for example, in diagnostic or therapeutic 30 applications. In some cases, it may be desirable to dry the crystal. Crystals may be dried by means of inert gases, like nitrogen gas, vacuum oven drying, lyophilization, evaporation, tray drying, fluid bed drying, spray drying, vacuum drying or roller drying. Suitable methods 35 are well known. 22 Crystals formed according to the invention can be maintained in the original crystalliza tion solution, or they can be washed and combined with other substances, like inert carriers or ingredients to form compositions or formulations comprising crystals of the 5 invention. Such compositions or formulations can be used, for example, in therapeutic and diagnostic applications. A preferred embodiment is to combine a suitable carrier or ingredient with crystals of the invention in that way that crystals of the formulation are embedded or encapsulated 10 by an excipient. Suitable carriers may be taken from the non limiting group of: poly (acrylic acid), poly (cyanoacrylates), poly (amino acids), poly (anhydrides), poly (dep sipeptide), poly (esters), poly (lactic acid), poly (lactic-co-glycolic acid) or PLGA, poly (B-hydroxybutryate), poly (caprolactone), poly (dioxanone); poly (ethylene glycol), poly (hydroxypropyl) methacrylamide, poly (organo) phosphazene, poly (ortho esters), poly 15 (vinyl alcohol), poly (vinylpyrrolidone), maleic anhydride alkyl vinyl ether copolymers, pluronic polyols, albumin, alginate, cellulose and cellulose derivatives, collagen, fibrin, gelatin, hyaluronic acid, oligosaccharides, glycaminoglycans, sulfated polysaccharides, blends and copolymers thereof, SAIB, fatty acids and salts of fatty acids, fatty alcohols, fatty amines, mono-, di-, and triglycerides of fatty acids, phospholipids, glycolipids, 20 sterols and waxes and related similar substances. Waxes are further classified in natu ral and synthetic products. Natural materials include waxes obtained from vegetable, animal or minerals sources such as beeswax, carnauba or montanwax. Chlorinated naphthalenes and ethylenic polymers are examples for synthetic wax products. 25 C. Compositions Another aspect of the invention relates to compositions/formulations comprising anti hTNFalpha antibody crystals in combination with at least one carrier/excipient. 30 The formulations may be solid, semisolid or liquid. Formulations of the invention are prepared, in a form suitable for storage and/or for use, by mixing the antibody having the necessary degree of purity with a physiologi cally acceptable additive, like carrier, excipient and/or stabilizer (see for example Rem 35 ington's Pharmaceutical Sciences, 16th Edn., Osol, A. Ed. (1980)), in the form of sus 23 pensions, lyophilized or dried in another way. Optionally further active ingredients, as for example different antibodies, biomolecules, chemically or enzymatically synthesized low-molecular weight molecules may be incorporated as well. 5 Acceptable additives are non-toxic to recipients at the dosages and concentrations employed. Nonlimiting examples thereof include: - Acidifying agents, like acetic acid, citric acid, fumaric acid, hydrochloric acid, malic acid, nitric acid, phosphoric acid, diluted phosphoric acid, sulfuric acid, tartaric acid. 10 - Aerosol propellants, like butane, dichlorodifluoromethane, dichlorotetrafluoroethane, isobutane, propane, trichloromonofluoromethane. - Air displacements, like carbon dioxide, nitrogen; 15 - Alcohol denaturants, like methyl isobutyl ketone, sucrose octacetate; - Alkalizing agents, like ammonia solution, ammonium carbonate, diethanolamine, diisopropanolamine, potassium hydroxide, sodium bicarbonate, sodium borate, sodium 20 carbonate, sodium hydroxide, trolamine; - Antifoaming agents, like dimethicone, simethicone. - Antimicrobial preservatives, like benzalkonium chloride, benzalkonium chloride solu 25 tion, benzelthonium chloride, benzoic acid, benzyl alcohol, butylparaben, cetylpyridin ium chloride, chlorobutanol, chlorocresol, cresol, dehydroacetic acid, ethylparaben, methylparaben, methylparaben sodium, phenol, phenylethyl alcohol, phenylmercuric acetate, phenylmercuric nitrate, potassium benzoate, potassium sorbate, propylpara ben, propylparaben sodium, sodium benzoate, sodium dehydroacetate, sodium propi 30 onate, sorbic acid, thimerosal, thymol. - Antioxidants, like ascorbic acid, acorbyl palmitate, butylated hydroxyanisole, butylated hydroxytoluene, hypophosphorous acid, monothioglycerol, propyl gallate, sodium for maldehyde sulfoxylate, sodium metabisulfite, sodium thiosulfate, sulfur dioxide, toco 35 pherol, tocopherols excipient; 24 - Buffering agents, like acetic acid, ammonium carbonate, ammonium phosphate, boric acid, citric acid, lactic acid, phosphoric acid, potassium citrate, potassium metaphos phate, potassium phosphate monobasic, sodium acetate, sodium citrate, sodium lac 5 tate solution, dibasic sodium phosphate, monobasic sodium phosphate, histidine. - Chelating agents, like edetate disodium, ethylenediaminetetraacetic acid and salts, edetic acid; 10 - Coating agents, like sodium carboxymethylcellulose, cellulose acetate, cellulose ace tate phthalate, ethylcellulose, gelatin, pharmaceutical glaze, hydroxypropyl cellulose, hydroxypropyl methylcellulose, hydroxypropyl methylcellulose phthalate, methacrylic acid copolymer, methylcellulose, polyethylene glycol, polyvinyl acetate phthalate, shel lac, sucrose, titanium dioxide, carnauba wax, microcystalline wax, zein, poly amino 15 acids, other polymers like PLGA etc., and SAIB. - Coloring agent, like ferric oxide. - Complexing agents, like ethylenediaminetetraacetic acid and salts (EDTA), edetic 20 acid, gentisic acid ethanolamide, oxyquinoline sulfate. - Desiccants, like calcium chloride, calcium sulfate, silicon dioxide. - Emulsifying and/or solubilizing agents, like acacia, cholesterol, diethanolamine 25 (adjunct), glyceryl monostearate, lanolin alcohols, lecithin, mono-and di-glycerides, monoethanolamine (adjunct), oleic acid (adjunct), oleyl alcohol (stabilizer), poloxamer, polyoxyethylene 50 stearate, polyoxyl 35 caster oil, polyoxyl 40 hydrogenated castor oil, polyoxyl 10 oleyl ether, polyoxyl 20 cetostearyl ether, polyoxyl 40 stearate, polysorbate 20, polysorbate 40, polysorbate 60, polysorbate 80, propylene glycol 30 diacetate, propylene glycol monostearate, sodium lauryl sulfate, sodium stearate, sorbitan monolaurate, soritan monooleate, sorbitan monopalmitate, sorbitan monostearate, stearic acid, trolamine, emulsifying wax. - Filtering aids, like powdered cellulose, purified siliceous earth. 35 25 - Flavors and perfumes, like anethole, benzaldehyde, ethyl vanillin, menthol, methyl salicylate, monosodium glutamate, orange flower oil, peppermint, peppermint oil, pep permint spirit, rose oil, stronger rose water, thymol, tolu balsam tincture, vanilla, vanilla tincture, vanillin. 5 - Glidant and/or anticaking agents, like calcium silicate, magnesium silicate, colloidal silicon dioxide, talc. - Humectants, like glycerin, hexylene glycol, propylene glycol, sorbitol; 10 - Ointment bases, like lanolin, anhydrous lanolin, hydrophilic ointment, white ointment, yellow ointment, polyethylene glycol ointment, petrolatum, hydrophilic petrolatum, white petrolatum, rose water ointment, squalane. 15 - Plasticizers, like castor oil, lanolin, mineral oil, petrolatum, benzyl benyl formate, chlorobutanol, diethyl pthalate, sorbitol, diacetylated monoglycerides, diethyl phthalate, glycerin, glycerol, mono-and di-acetylated monoglycerides, polyethylene glycol, propyl ene glycol, triacetin, triethyl citrate, ethanol. 20 - Polypeptides, like low molecular weight (less than about 10 residues); Proteins, such as serum albumin, gelatin, or immunoglobulins; - Polymer membranes, like cellulose acetate membranes. 25 - Solvents, like acetone, alcohol, diluted alcohol, amylene hydrate, benzyl benzoate, butyl alcohol, carbon tetrachloride, chloroform, corn oil, cottonseed oil, ethyl acetate, glycerin, hexylene glycol, isopropyl alcohol, methyl alcohol, methylene chloride, methyl isobutyl ketone, mineral oil, peanut oil, polyethylene glycol, propylene carbonate, pro pylene glycol, sesame oil, water for injection, sterile water for injection, sterile water for 30 irrigation, purified water, liquid triglycerides, liquid waxes, higher alcohols. - Sorbents, like powdered cellulose, charcoal, purified siliceous earth, Carbon dioxide sorbents, barium hydroxide lime, soda lime. 26 - Stiffening agents, like hydrogenated castor oil, cetostearyl alcohol, cetyl alcohol, cetyl esters wax, hard fat, paraffin, polyethylene excipient, stearyl alcohol, emulsifying wax, white wax, yellow wax. 5 - Suppository bases, like cocoa butter, hard fat, polyethylene glycol; - Suspending and/or viscosity-increasing agents, like acacia, agar, alginic acid, alumi num monostearate, bentonite, purified bentonite, magma bentonite, carbomer 934p, carboxymethylcellulose calcium, carboxymethylcellulose sodium, carboxymethycellu 10 lose sodium 12, carrageenan, microcrystalline and carboxymethylcellulose sodium cel lulose, dextrin, gelatin, guar gum, hydroxyethyl cellulose, hydroxypropyl cellulose, hy droxypropyl methylcellulose, magnesium aluminum silicate, methylcellulose, pectin, polyethylene oxide, polyvinyl alcohol, povidone, propylene glycol alginate, silicon diox ide, colloidal silicon dioxide, sodium alginate, tragacanth, xanthan gum; 15 - Sweetening agents, like aspartame, dextrates, dextrose, excipient dextrose, fructose, mannitol, saccharin, calcium saccharin, sodium saccharin, sorbitol, solution sorbitol, sucrose, compressible sugar, confectioner's sugar, syrup; 20 - Tablet binders, like acacia, alginic acid, sodium carboxymethylcellulose, microcrystal line cellulose, dextrin, ethylcellulose, gelatin, liquid glucose, guar gum, hydroxypropyl methylcellulose, methycellulose, polyethylene oxide, povidone, pregelatinized starch, syrup. 25 - Tablet and/or capsule diluents, like calcium carbonate, dibasic calcium phosphate, tribasic calcium phosphate, calcium sulfate, microcrystalline cellulose, powdered cellu lose, dextrates, dextrin, dextrose excipient, fructose, kaolin, lactose, mannitol, sorbitol, starch, pregelatinized starch, sucrose, compressible sugar, confectioner's sugar; 30 - Tablet disintegrants, like alginic acid, microcrystalline cellulose, croscarmellose so dium, corspovidone, polacrilin potassium, sodium starch glycolate, starch, pregelati nized starch. - Tablet and/or capsule lubricants, like calcium stearate, glyceryl behenate, magnesium 35 stearate, light mineral oil, polyethylene glycol, sodium stearyl fumarate, stearic acid, 27 purified stearic acid, talc, hydrogenated vegetable oil, zinc stearate; - Tonicity agent, like dextrose, glycerin, mannitol, potassium chloride, sodium chloride Vehicle: flavored and/or sweetened aromatic elixir, compound benzaldehyde elixir, iso 5 alcoholic elixir, peppermint water, sorbitol solution, syrup, tolu balsam syrup. - Vehicles, like oleaginous almond oil, corn oil, cottonseed oil, ethyl oleate, isopropyl myristate, isopropyl palmitate, mineral oil, light mineral oil, myristyl alcohol, octyldode canol, olive oil, peanut oil, persic oil, sesame oil, soybean oil, squalane; solid carrier 10 sugar spheres; sterile bacteriostatic water for injection, bacteriostatic sodium chloride injection, liquid triglycerides, liquid waxes, higher alcohols - Water repelling agents, like cyclomethicone, dimethicone, simethicone; 15 - Wetting and/or solubilizing agents, like benzalkonium chloride, benzethonium chlo ride, cetylpyridinium chloride, docusate sodium, nonoxynol 9, nonoxynol 10, octoxynol 9, poloxamer, polyoxyl 35 castor oil, polyoxyl 40, hydrogenated castor oil, polyoxyl 50 stearate, polyoxyl 10 oleyl ether, polyoxyl 20, cetostearyl ether, polyoxyl 40 stearate, polysorbate 20, polysorbate 40, polysorbate 60, polysorbate 80, sodium lauryl sulfate, 20 sorbitan monolaureate, sorbitan monooleate, sorbitan monopalmitate, sorbitan monostearate, tyloxapol; The crystals may be combined with a polymeric carrier to provide for stability and/or sustained release. Such polymers include biocompatible and biodegradable polymers. 25 A polymeric carrier may be a single polymer type or it may be composed of a mixture of polymer types. Nonlimiting examples of polymeric carriers have already been stated above. Examples of preferred ingredients or excipients include: 30 - salts of amino acids such as glycine, arginine, aspartic acid, glutamic acid, lysine, asparagine, glutamine, proline, histidine; - monosaccharides, such as glucose, fructose, galactose, mannose, arabinose, xylose, ribose; 35 - disaccharides, such as lactose, trehalose, maltose, sucrose; 28 - polysaccharides, such as maltodextrins, dextrans, starch, glycogen; - alditols, such as mannitol, xylitol, lactitol, sorbitol; - glucuronic acid, galacturonic acid; - cyclodextrins, such as methyl cyclodextrin, hydroxypropyl- (3-cyclodextrin) 5 - inorganic salts, such as sodium chloride, potassium chloride, magnesium chloride, phosphates of sodium and potassium, boric acid ammonium carbonate and ammonium phosphate; - organic salts, such as acetates, citrate, ascorbate, lactate; - emulsifying or solubilizing agents like acacia, diethanolamine, glyceryl monostearate, 10 lecithin, monoethanolamine, oleic acid, oleyl alcohol, poloxamer, polysorbates, sodium lauryl sulfate, stearic acid, sorbitan monolaurate, sorbitan monostearate, and other sorbitan derivatives, polyoxyl derivatives, wax, polyoxyethylene derivatives, sorbitan derivatives; and - viscosity increasing reagents like, agar, alginic acid and its salts, guar gum, pectin, 15 polyvinyl alcohol, polyethylene oxide, cellulose and its derivatives propylene carbonate, polyethylene glycol, hexylene glycol and tyloxapol. Formulations described herein also comprise an effective amount of crystalline anti body. In particular, the formulations of the invention may include a "therapeutically ef 20 fective amount" or a "prophylactically effective amount" of antibody crystals of the in vention. A "therapeutically effective amount" refers to an amount effective, at dosages and for periods of time necessary, to achieve the desired therapeutic result. A "thera peutically effective amount" of the antibody crystals may vary according to factors such as the disease state, age, sex, and weight of the individual, and the ability of the anti 25 body to elicit a desired response in the individual. A therapeutically effective amount is also one in which any toxic or detrimental effects of the antibody are outweighed by the therapeutically beneficial effects. A "prophylactically effective amount" refers to an amount effective, at dosages and for periods of time necessary, to achieve the desired prophylactic result. Typically, since a prophylactic dose is used in subjects prior to or 30 at an earlier stage of disease, the prophylactically effective amount will be less than the therapeutically effective amount. Suitable dosages can readily be determined using standard methodology. The antibody is suitably administered to the patient at one time or over a series of treatments. De 35 pending on the above mentioned factors, about 1 pg/kg to about 50 mg/kg , as for ex 29 ample 0.1-20 mg/kg of antibody is an initial candidate dosage for administration to the patient, whether, for example, by one or more separate administrations, or by continu ous infusion. A typical daily or weekly dosage might range from about 1 pg/kg to about 20 mg/kg or more, depending on the condition, the treatment is repeated until a desired 5 suppression of disease symptoms occurs. However, other dosage regimens may be useful. In some cases, formulations comprise a concentration of antibody of at least about 1 g/L or greater when resolubilized. In other embodiments, the antibody concen tration is at least about 1 g/L to about 100 g/L when resolubilized. 10 Crystals of an antibody, or formulations comprising such crystals, may be administered alone or as part of a pharmaceutical preparation. They may be administered by par enteral, oral or topical routes. For example, they may be administered by oral, pulmo nary, nasal, aural, anal, dermal, ocular, intravenous, intramuscular, intraarterial, intrap eritoneal, mucosal, sublingual, subcutaneous, transdermal, topical or intracranial 15 routes, or into the buccal cavity. Specific examples of administration techniques com prise pulmonary inhalation, intralesional application, needle injection, dry powder inha lation, skin electroporation, aerosol delivery, and needle-free injection technologies, including needle-free subcutaneous administration. 20 The present invention will now be explained in more detail by means of the following, non-limiting, illustrative examples. Guided by the general part of the description and on the basis of his general knowledge a skilled reader will be enabled to provide further embodiments to the invention without undue experimentation. 25 Experimental Part A. Materials 30 a) Protein Frozen monoclonal antibody (mAb) D2E7 was obtained from Abbott Laboratories. All experiments were performed from a drug product lot where the original mAb concentra tion was 50 mg/ml. 35 b) Fine chemicals 30 Sodium acetate was obtained from GrOssing GmbH, Filsum. Polyethyleneglycols of different polymerization grades were obtained from Clariant GmbH, Sulzbach. Fur thermore, commercial crystallization screens and reagents (Hampton Research, Nextal Biotechnologies) were used for certain microscale experiments. All other chemicals 5 were from Sigma-Aldrich, Steinheim, or Merck, Darmstadt. B. General methods a) Thawing of D2E7 drug substance 10 D2E7 was thawed at 25 0C in agitated water baths. b) Buffer exchange - method A An aliquot of D2E7 solution was pipetted into a 30 KDa MWCO Vivaspin 20 concentra tor (Vivascience). The protein sample was diluted with the new buffer in a ratio of 1:10, 15 and by centrifugation at 5,000 x g at 4 *C (Sigma 4 K 15 lab centrifuge) the sample volume was brought back to the original sample volume. The dilution / centrifugation steps were repeated once, resulting in a dilution of 1:100 of the original sample buffer. After adjustment of protein concentration, the solution was sterile filtered through a 0.2 pm syringe driven filter unit. 20 b) Buffer exchange - method B An aliquot of D2E7 solution was placed into a SLIDE-A-LYZER dialysis cassette (Pierce Biotechnology Inc.). The dialysis cassette was placed into a beaker containing the buffer of choice, and the buffer exchange was performed at 4 0C overnight with 25 stirring. After adjustment of protein concentration, the solution was sterile filtered through a 0.2 pm syringe driven filter unit. c) OD280 - protein concentration measurements A ThermoSpectronics UV1 device was used to assess protein concentration at a wave 30 length of 280 nm, applying an extinction coefficient of 1.39 cm 2 mg 1 . For this purpose, aliquots of crystallization slurries were centrifuged at 14,000 rpm, and residual protein concentration was determined in the supernatant. d) pH measurements 31 pH measurements were conducted by using a Mettler Toledo MP220 pH meter. Inlab 413 electrodes and Inlab 423 microelectrodes were utilized. e) Crystallization methods 5 el) Microscale crystallization - Sitting drop vapor diffusion Hydra 11 Initial crystallization screens were performed using a Hydra || crystallization robot and Greiner 96 well plates (three drop wells, Hampton Research). After setting up the plates, the wells were sealed with Clearseal film (Hampton Research). 10 e2) Microscale crystallization - Hanging drop vapor diffusion Hanging drop vapor diffusion experiments were conducted using VDX plates (with sealant, Hampton Research) and OptiClear plastic cover slides (squares, Hampton Research) or siliconized glass cover slides (circle, Hampton Research), respectively. 15 After preparation of reservoir solutions, one drop of reservoir solution was admixed with one drop of the protein solution on a cover slide, and the well was sealed with the in verted cover slide in such a way that the drop was hanging above the reservoir. e3) Batch crystallization - method A (24 well plate) 20 Batch crystallization was performed by admixing the protein solution with an equal amount of crystallization buffer (500 pl) in a well. The well was subsequently sealed with adhesive tape to prevent water evaporation. e4) Batch crystallization - method B (Eppendorff reaction tube) 25 Batch crystallization was performed by admixing the protein solution with an equal amount of crystallization buffer in a 1.5 mL or a 2 mL Eppendorff reaction tube. e5) Batch crystallization - method C (Falcon tubes, agitation) Batch crystallization was performed by admixing the protein solution with an equal 30 amount of crystallization buffer in a 50 mL Falcon tube. Right after closing, the tube was put on a laboratory shaker (GFL 3013 or GFL 3015) or was alternatively agitated by tumbling. By application of these methods, introduction of stirrers into the sample was avoided. 35 e6) Batch crystallization - method D (1 liter polypropylene container) 32 Batch crystallization was performed by admixing the protein solution with an equal amount of crystallization buffer in a sterilized 1 liter polypropylene bottle. Right after closing, the container was stored at ambient temperature without agitation. By applica tion of this method, introduction of stirrers into the sample was avoided. 5 f) SDS-PAGE Samples were prepared by adjusting protein concentration to 8 pg / 20 pL. The sam ples were diluted with an SDS / Tris / Glycerine buffer containing bromphenolblue. 10 Qualitative SDS PAGE analysis was performed using Invitrogen NuPage 10% Bis-Tris Gels, NuPage MES SDS Running Buffer and Mark12 Wide Range Protein Standards. 20 pL of sample was pipetted into a gel pocket. After running the gel and fixation with acetic acid / methanol reagent, staining was performed using the Novex Colloidal Blue Stain Kit. Gels were dried using Invitogen Gel-Dry drying solution. 15 g) Light Microscopy Crystals were observed using a Zeiss Axiovert 25 or a Nikon Labophot microscope. The latter was equipped with a polarization filter set and a JVC TK C1 380 color video camera. 20 h) SE-HPLC Aggregation levels of D2E7 samples were assessed by SE-HPLC. A Dionex P680 pump, ASI-100 autosampler and UVD170U detector device were used. Aggregated species were separated from the monomer by an Amersham Bioscience Superose 6 25 10/300 GL gel filtration column, applying a validated Abbott standard protocol (CL16 PS-02, Adalimumab purity). C. Vapor diffusion crystallization experiments 30 Concentration values given in the following examples are initial values referring to the antibody solution and the reservoir solution before mixing of the two solutions. All pH values, if not described otherwise, refer to the pH of an acetate buffer stock be 35 fore it was combined with other substances, like the crystallization agent. 33 All buffer molarities, if not described otherwise, refer to sodium acetate concentrations in a stock solution before pH adjustment, typically performed using acetic acid glacial. 5 Example 1 - PEG 4,000 / sodium acetate grid screen in hanging drop vapor diffu sion mode D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of 10 around 5.2. The protein concentration was adjusted to 10 mg/mL. A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, 50% w/v PEG 4,000 solution and Milli Q water (fully desalted and optionally pre-destilled) in each 15 well. In this example, the acetate buffer molarity was kept constant at around 0.1 M, and PEG 4,000 was varied from around 6% w/v to around 28% w/v in 2% steps. The pH was around 5.2 throughout. Each condition was assessed in duplicate. Around 1 pL of protein solution was admixed with around 1 pL of a particular reservoir solution on a square OptiClear plastic cover slide, and the well was sealed with the inverted slide, 20 generating a hanging drop experiment. The plates were stored at ambient temperature. Microscopy of the drops was performed multiple times during the following thirty days. The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops containing mixtures of precipitated species and crystals. 25 RESULTS: From the 24 wells assessed, no crystals were observed. Example 2 - PEG 4,000 / sodium acetate grid screen in hanging drop vapor diffu 30 sion mode, different protein concentration D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of around 5.2. The protein concentration was adjusted to 50 mg/mL. Except for the pro tein concentration the process conditions were identical with those of Example 1. 35 34 RESULTS: From the 24 wells assessed, no crystals were observed. Example 3 - PEG 400 / sodium acetate grid screen in hanging drop vapor diffu 5 sion mode D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of around 5.2. The protein concentration was adjusted to 10 mg/mL. 10 A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, 50% w/v PEG 400 solution and Milli Q water in each well. In this example, the acetate buffer molarity was kept constant at around 0.1 M, and PEG 400 was varied from around 30% w/v to around 40% w/v in 2% steps. The pH was around 5.2 throughout. Each condition was 15 assessed in duplicate. Around 1 pL of protein solution was admixed with around 1 pL of a particular reservoir solution on a square OptiClear plastic cover slide, and the well was sealed with the inverted slide, generating a hanging drop experiment. The plates were stored at ambient temperature. Microscopy of the drops was performed multiple times during the following thirty days. The conditions were classified into clear drops, 20 drops containing random precipitation, drops containing crystals and drops containing mixtures of precipitated species and crystals. RESULTS: From the 12 wells assessed, no crystals were observed. 25 Example 4 - PEG 400 / sodium acetate grid screen in hanging drop vapor diffu sion mode, different protein concentration D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of 30 around 5.2. The protein concentration was adjusted to 50 mg/mL. Except for the pro tein concentration the process conditions were identical with those of Example 3. RESULTS: From the 12 wells assessed, no crystals were observed. 35 35 Example 5 - PEG 10,000 / sodium acetate grid screen in hanging drop vapor dif fusion mode D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of 5 around 5.2. The protein concentration was adjusted to 10 mg/mL. A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, 50% w/v PEG 400 solution and Milli Q water in each well. In this example, the acetate buffer molarity 10 was kept constant at around 0.1 M, and PEG 10,000 was varied from around 4% w/v to around 14% w/v in 2% steps. The pH was around 5.2 throughout. Each condition was assessed in duplicate. Around 1 pL of protein solution was admixed with around 1 pL of a particular reservoir solution on a square OptiClear plastic cover slide, and the well was sealed with the inverted slide, generating a hanging drop experiment. The plates 15 were stored at ambient temperature. Microscopy of the drops was performed multiple times during the following thirty days. The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops containing mixtures of precipitated species and crystals. 20 RESULTS: From the 12 wells assessed, no crystals were observed. Example 6 - PEG 10,000 / sodium acetate grid screen in hanging drop vapor dif fusion mode, different protein concentration 25 D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of around 5.2. The protein concentration was adjusted to 45 to 55 mg/mL, preferably 50 mg/mL. Except for the protein concentration the process conditions were identical with those of Example 5. 30 RESULTS: From the 12 wells assessed, no crystals were observed. Example 7 - PEG 400 / sodium acetate grid screen in hanging drop vapor diffu 35 sion mode, different set up 36 D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of around 5.2. The protein concentration was adjusted to 10 mg/mL. 5 A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, 50% w/v PEG 10,000 solution and Milli Q water in each well. In this example, the acetate buffer mo larity was kept constant at around 0.1 M, and PEG 400 was around 32% w/v and around 34% w/v. The pH was around 4.2, 4.7, 5.2, 5.7, 6.2 or 6.7. Each condition was 10 assessed in duplicate. Around 1 pL of protein solution was admixed with around 1 pL of a particular reservoir solution on a square OptiClear plastic cover slide, and the well was sealed with the inverted slide, generating a hanging drop experiment. The plates were stored at ambient temperature. Microscopy of the drops was performed multiple times during the following thirty days. The conditions were classified into clear drops, 15 drops containing random precipitation, drops containing crystals and drops containing mixtures of precipitated species and crystals. RESULTS: From the 24 wells assessed, no crystals were observed. 20 Example 8 - PEG 400 / sodium acetate grid screen in hanging drop vapor diffu sion mode, different protein concentration and set up D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of 25 around 5.2. The protein concentration was adjusted to 50 mg/mL. Except for the pro tein concentration the process conditions were identical with those of Example 7. RESULTS: From the 24 wells assessed, no crystals were observed. 30 Example 9 - PEG 400 / sodium acetate grid screen in hanging drop vapor diffu sion mode, different set up D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of 35 around 5.2. The protein concentration was adjusted to 10 mg/mL. 37 A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, 50% w/v PEG 400 solution and Milli Q water in each well. In this example, the acetate buffer molarity 5 was used at around 0.025 M, 0.05 M, 0.075 M, 0.15 M, 0.2 M or 0.25 M. PEG 400 was varied from around 32% w/v to around 34% w/v. The pH was around 5.7 or 4.2. Each condition was assessed in duplicate. Around 1 pL of protein solution was admixed with around 1 pL of a particular reservoir solution on a square OptiClear plastic cover slide, and the well was sealed with the inverted slide, generating a hanging drop experiment. 10 The plates were stored at ambient temperature. Microscopy of the drops was per formed multiple times during the following thirty days. The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops containing mixtures of precipitated species and crystals. 15 RESULTS: From the 48 wells assessed, no crystals were observed. Example 10 - PEG 400 / sodium acetate grid screen in hanging drop vapor diffu sion mode, different set up 20 D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of around 5.2. The protein concentration was adjusted to 10 mg/mL. A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a 25 particular reservoir solution was prepared by admixing acetate buffer, 50% w/v PEG 400 solution and Milli Q water in each well. In this example, the acetate buffer molarity was used at around 0.025 M, 0.05 M or 0.1 M. PEG 400 was around 28% w/v or around 30% w/v. The pH was around 5.2, 5.7, 6.2 or 6.7. Each condition was assessed in duplicate. Around 1 pL of protein solution was admixed with around 1 pL of a particu 30 lar reservoir solution on a square OptiClear plastic cover slide, and the well was sealed with the inverted slide, generating a hanging drop experiment. The plates were stored at ambient temperature. Microscopy of the drops was performed multiple times during the following thirty days. The conditions were classified into clear drops, drops contain ing random precipitation, drops containing crystals and drops containing mixtures of 35 precipitated species and crystals. 38 RESULTS: From the 48 wells assessed, no crystals were observed. 5 Example 11 - PEG 400 combined with PEG 4,000 / sodium acetate grid screen in hanging drop vapor diffusion mode D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of around 5.2. The protein concentration was adjusted to 10 mg/mL. 10 A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, 50% w/v PEG 4,000 solution and Milli Q water in each well. In this example, the acetate buffer molar ity was kept constant at around 0.1 M, and PEG 4,000 was varied from around 4% w/v 15 to around 8% w/v in 2% steps. Simultaneously, PEG 400 was added to the PEG 4,000 / acetate solutions at concentrations of around 24% w/v, 26% w/v, 28% w/v or 30% w/v. The pH was around 5.2 throughout. Each condition was assessed in duplicate. Around 1 pL of protein solution was admixed with around 1 pL of a particular reservoir solution on a square OptiClear plastic cover slide, and the well was sealed with the 20 inverted slide, generating a hanging drop experiment. The plates were stored at ambi ent temperature. Microscopy of the drops was performed multiple times during the fol lowing thirty days. The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops containing mixtures of pre cipitated species and crystals. 25 RESULTS: From the 24 wells assessed, no crystals were observed. Example 12 - PEG 400 combined with PEG 4,000 / sodium acetate grid screen in 30 hanging drop vapor diffusion mode, different set up D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of around 5.2. The protein concentration was adjusted to 10 mg/mL. 39 A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, 50% w/v PEG 4,000 solution and Milli Q water in each well. In this example, the acetate buffer molar ity was kept constant at around 0.1 M, and PEG 4,000 was varied from around 4% w/v 5 to around 8% w/v in 2% steps. Simultaneously, PEG 400 was added to the PEG 4,000 / acetate solutions at concentrations of around 30% w/v, 32% w/v, 34% w/v or 36% w/v. The pH was around 4.2 throughout. Each condition was assessed in duplicate. Around 1 pL of protein solution was admixed with around 1 pL of a particular reservoir solution on a square OptiClear plastic cover slide, and the well was sealed with the 10 inverted slide, generating a hanging drop experiment. The plates were stored at ambi ent temperature. Microscopy of the drops was performed multiple times during the fol lowing thirty days. The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops containing mixtures of pre cipitated species and crystals. 15 RESULTS: From the 24 wells assessed, no crystals were observed. Example 13 - PEG 4,000 / sodium acetate grid screen in hanging drop vapor dif 20 fusion mode, different set up D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of around 6.5, 6.0, 5.5, 5.0, 4.5 or 4.0. The protein concentration was adjusted to 10 mg/mL. 25 A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, 50% w/v PEG 4,000 solution and Milli Q water in each well. In this example, the acetate buffer molar ity was kept constant at around 0.1 M, and PEG 4,000 was varied from around 4% w/v 30 to around 26% w/v in 2% steps. The pH of the acetate buffer used was the same as the corresponding protein buffer. Each condition was assessed in duplicate. Around 1 pL of protein solution was admixed with around 1 pL of a particular reservoir solution on a square OptiClear plastic cover slide, and the well was sealed with the inverted slide, generating a hanging drop experiment. The plates were stored at ambient temperature. 35 Microscopy of the drops was performed multiple times during the following thirty days. 40 The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops containing mixtures of precipitated species and crystals. 5 RESULTS: From the 144 wells assessed, no crystals were observed. Example 14 - PEG 4,000 / sodium acetate grid screen in hanging drop vapor dif fusion mode, different set up 10 The experimental conditions were identical to Example 13, except for the acetate buffer molarity, which was kept constant at around 0.2 M (molarity of precipitation buffer) RESULTS: From the 144 wells assessed, no crystals were observed. 15 Example 15 - PEG 4,000 / sodium acetate grid screen in hanging drop vapor dif fusion mode, different set up The experimental conditions were identical to Example 13, except for the acetate buffer 20 molarity which was kept constant at 0.1 M (molarity of precipitation buffer). RESULTS: From the 144 wells assessed, no crystals were observed. 25 Example 16 - PEG 4,000 / sodium acetate grid screen in hanging drop vapor dif fusion mode, different set up The experimental conditions were identical to Example 13, except for the acetate buffer molarity which was kept constant at around 0.4 M (molarity of precipitation buffer). 30 RESULTS: From the 144 wells assessed, no crystals were observed. Example 17 - PEG 4,000 / sodium acetate bulk experiments 35 41 D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of around 5.5. The protein concentration was adjusted to 10 mg/mL. Four aliquots of 500 pL each were pipetted into four Eppendorff reaction tubes. A 24% 5 PEG 4,000 in 0.1 M sodium acetate buffer at pH 5.5 solution was titrated to the protein solutions until the solution became slightly opaque. Subsequently, water was pipetted to the solutions just until the solutions became clear again. This method is referred to as bulk crystallization. Titration was performed at ambient temperature for two samples and at 4 0C for the two other samples. Subsequently, one of each pair of samples was 10 stored at ambient temperature or at 4 0C, respectively. Microscopy of 1 pL aliquots of the samples was performed multiple times during the following week RESULTS: From the four samples, none rendered crystals. 15 Example 18 - PEG 4,000 / sodium acetate grid screen in hanging drop vapor dif fusion mode, different temperature The experimental conditions were identical to Example 13. However, the tubes were 20 set up and stored at 4 *C. RESULTS: From the 144 wells assessed, no crystals were observed. 25 Example 19 - PEG 4,000 / sodium acetate grid screen in hanging drop vapor dif fusion mode, different protein concentration Except for the protein concentration, which was adjusted to 5 mg/mL the experimental conditions were identical to Example 13. 30 RESULTS: From the 144 wells assessed, no crystals were observed. Example 20 - Ammonium sulfate / sodium acetate grid screen in hanging drop 35 vapor diffusion mode 42 D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of around 5.5. The protein concentration was adjusted to 10 mg/mL. 5 A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, ammonium sul fate stock solution and Milli Q water in each well. In this example, the acetate buffer molarity was kept constant at around 0.1 M, and the ammonium sulfate concentration was varied from 0.5 M to 2.5 M in steps of 0.25 M. The pH of the acetate buffer was 10 around 5.5 throughout. Each condition was assessed in duplicate. Around 1 pL of pro tein solution was admixed with around 1 pL of a particular reservoir solution on a square OptiClear plastic cover slide, and the well was sealed with the inverted slide, generating a hanging drop experiment. The plates were set up and stored at ambient temperature. Microscopy of the drops was performed multiple times during the follow 15 ing two weeks. The conditions were classified into clear drops, drops containing ran dom precipitation, drops containing crystals and drops containing mixtures of precipi tated species and crystals. RESULTS: From the 18 wells assessed, no crystals were observed. 20 Example 21 - Sodium chloride / sodium acetate grid screen in hanging drop va por diffusion mode 25 D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of around 5.5. The protein concentration was adjusted to 10 mg/mL. A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, sodium chloride 30 stock solution and Milli Q water in each well. In this example, the acetate buffer molar ity was kept constant at around 0.1 M, and the sodium chloride concentration was var ied from 1.5 M to 2.5 M, varied in steps of 0.5 M. The pH of the acetate buffer was around 5.5 throughout. Each condition was assessed in duplicate. Around 1 pL of pro tein solution was admixed with around 1 pL of a particular reservoir solution on a 35 square OptiClear plastic cover slide, and the well was sealed with the inverted slide, 43 generating a hanging drop experiment. The plates were set up and stored at ambient temperature. Microscopy of the drops was performed multiple times during the follow ing two weeks. The conditions were classified into clear drops, drops containing ran dom precipitation, drops containing crystals and drops containing mixtures of precipi 5 tated species and crystals. RESULTS: From the 6 wells assessed, no crystals were observed. 10 Example 22 - PEG 4,000 / sodium acetate grid screen in hanging drop vapor dif fusion mode, influence of detergents D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of around 5.5. The protein concentration was adjusted to 5 mg/mL. 15 A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, 50% w/v PEG 4,000 solution and Milli Q water in each well. In this example, the acetate buffer molar ity was kept constant at around 0.1 M, and PEG 4,000 was varied from around 10% 20 w/v to around 20% w/v in 2% steps. The pH of the acetate buffer was around 5.5 throughout. Furthermore, polysorbate 20, polysorbate 80 and Pluronic F 68 were added to any resulting buffer as described above at concentrations of 0%, 0.02% and 0.1%, respectively. Around 1 pL of protein solution was admixed with around 1 pL of a particular reservoir solution on a square OptiClear plastic cover slide, and the well was 25 sealed with the inverted slide, generating a hanging drop experiment. The plates were set up and stored at ambient temperature. Microscopy of the drops was performed mul tiple times during the following two weeks. The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops con taining mixtures of precipitated species and crystals. 30 RESULTS: From the 84 wells assessed, no crystals were observed. No influence of the assessed detergents on the behaviour of the crystallization system could be observed. 44 Example 23 - Zinc acetate / sodium acetate grid screen in hanging drop vapor diffusion mode D2E7 was buffered into a buffer containing around 0.1 M sodium acetate at a pH of 5 around 5.5. The protein concentration was adjusted to 10 mg/mL. A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, zinc acetate stock solution and Milli 0 water in each well. In this example, the acetate buffer molar 10 ity was kept constant at around 0.1 M, and the zinc acetate concentration was varied from 0.1 M to 0.9 M in steps of around 0.2 M. The pH of the acetate buffer was around 5.5 throughout. Each condition was assessed in duplicate. Around 1 pL of protein solu tion was admixed with around 1 pL of a particular reservoir solution on a square Opti Clear plastic cover slide, and the well was sealed with the inverted slide, generating a 15 hanging drop experiment. The plates were set up and stored at ambient temperature. Microscopy of the drops was performed multiple times during the following two weeks. The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops containing mixtures of precipitated species and crystals. 20 RESULTS: From the 12 wells assessed, no crystals were observed. Example 24 - Broad screening of conditions in vapor diffusion mode 25 D2E7 was buffered into a 20 mM HEPES / 150 mM sodium chloride buffer at pH 7.4. The protein concentration was adjusted to 5 mg/mL, 10 mg/mL, or 20 mg/mL. Using the Hydra 11 crystallization robot, 96 well Greiner plates were set up at ambient 30 temperature, using several commercially available crystallization screens. The protein solution and the crystallization agent were admixed in a ratio of around 1:1, preferably 1:1. The following screens were used: 35 Hampton Crystal Screen 1 & 2 (Hampton Research), 45 Hampton Index Screen (Hampton Research), Hampton SaltRX Screen (Hampton Research), Nextal The Classics, The Classics Lite, The PEGs, The Anions, The pH clear and The Ammonium sulfate (all from Nextal Biotechnologies). 5 After addition of protein to the crystallization agent (three drops per condition, contain ing the three different protein concentrations as described above), the plates were sealed with Clearseal film. Each plate was set up in quadruplicate and then stored at ambient temperature, 4 0C, 27 *C and 37 0C, respectively. Microscopy of the drops was 10 performed after five days and twelve days, respectively. The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops containing mixtures of precipitated species and crystals. RESULTS: From the 864 commercial conditions evaluated, 2 rendered crystals at pro 15 tein concentrations and temperatures as defined below, at least after two weeks. - 0.1 M sodium acetate anhydrous pH 4.6, 0.9 M sodium dihydrogen phosphate, 0.9 M potassium dihydrogen phosphate (= Nextal The Anions, E3), 10 or 20 mg/mL, and 27 *C, or 20 mg/mL and 37 *C. 20 - 0.1 M Bis-Tris Propane pH 7.0, 1.5 M ammonium sulfate (= Hampton SaltRX, F2), 5, 10, or 20 mg/mL. and 27 *C. The crystals showed needle cluster-like morphologies. 25 The following conditions from commercially available screens did not render crystals. For detailed solution compositions, please refer to www.hamptonresearch.com and www.nextalbiotech.com: 30 Hampton Crystal Screen 1 - all conditions (48) Hampton Crytsal Screen 2 - all conditions (48) Hampton Index Screen - all conditions (96) Hampton SaltRX Screen - all conditions despite "F2" (95) Nextal - The Classics - all conditions (96) 35 Nextal - The Classics Lite - all conditions (96) 46 Nextal - The PEGs - all conditions (96) Nextal - The Anions - all conditions despite "E3" (95) Nextal - The pH Clear - all conditions (96) Nextal - The AmmoniumSulfate - all conditions (96) 5 Example 25 - PEG 4,000 / sodium acetate grid screen in hanging drop vapor dif fusion mode, different set up 10 D2E7 was buffered into a 20 mM HEPES / 150 mM sodium chloride buffer at pH 7.4. The protein concentration was adjusted to 5 mg/mL, 10 mg/mL, or 20 mg/mL. A greased VDX plate and circle siliconized glass cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, 50% w/v PEG 15 4,000 solution and Milli Q water in each well. In this example, the acetate buffer molar ity was kept constant at around 0.1 M, and PEG 4,000 concentration was varied from 4% to 26% in 2% steps. The pH was around 5.5 throughout. Each condition was set up with the three protein concentrations as described above. Around 1 pL of protein solu tion was admixed with around 1 pL of a particular reservoir solution on a circle sili 20 conized glass cover slide, and the well was sealed with the inverted slide, generating a hanging drop experiment. The plates were stored at ambient temperature. Microscopy of the drops was performed after six days. The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops con taining mixtures of precipitated species and crystals. 25 RESULTS: From the 72 wells assessed, no crystals were observed. Example 26 - Hanging drop vapor diffusion experiments applying the Hampton 30 Detergent Screen D2E7 was buffered into a 20 mM HEPES / 150 mM sodium chloride buffer at pH 7.4. The protein concentration was adjusted to 5 mg/mL. 47 A greased VDX plate and circle siliconized glass cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, 50% w/v PEG 4,000 solution and Milli Q water in each well. In this example, the acetate buffer molar ity was kept constant at around 0.1 M, and the PEG 4,000 concentration around 12% 5. w/v or 14% w/v. The pH was around 5.5 throughout. Around 4 pL of protein solution was admixed with around 1 pL of a particular detergent solution of the Hampton screen on a circle siliconized glass cover slide. The drop was subsequently admixed with 5 pL of a particular reservoir solution, and the well was sealed with the inverted slide, gen erating a hanging drop experiment. The plates were stored at ambient temperature. 10 Microscopy of the drops was performed after six days. The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops containing mixtures of precipitated species and crystals. RESULTS: From the 144 wells assessed, no crystals were observed. 15 Example 27 - Hanging drop vapor diffusion using Hampton PEG / Ion screen D2E7 was buffered into a 20 mM HEPES / 150 mM sodium chloride buffer at pH 7.4. 20 The protein concentration was adjusted to 5 mg/mL or 10 mg/mL. Greased VDX plates and circle siliconized glass cover slides were used. 500 pL of each of the 48 buffer formulations was pipetted into a well and admixed wit 250 pL of Milli Q water, respectively. Around 1 pL of protein sample was pipetted onto a cover 25 slide and subsequently admixed with around 1 pL of the reservoir solution of a particu lar well. The well was sealed with the inverted cover slide, generating a hanging drop experiment. The plates were stored at ambient temperature. Microscopy of the drops was performed multiple times during the following seven days. The conditions were classified into clear drops, drops containing random precipitation, drops containing 30 crystals and drops containing mixtures of precipitated species and crystals. RESULTS: From the 96 conditions tested, no crystals were observed. 48 Example 28 - Hanging drop vapor diffusion using Hampton PEG / Ion screen, dif ferent set up 5 D2E7 was buffered into a 20 mM HEPES / 150 mM sodium chloride buffer at pH 7.4. The protein concentration was adjusted to 5 mg/mL. The experimental conditions were identical with those of Example 27 with the exception that 500 pL of each of the 48 buffer formulations was pipetted into a well and admixed 10 with 500 pL of Milli Q water, respectively. RESULTS: From the 48 conditions tested, no crystals were observed. 15 Example 29 - Hanging drop vapor diffusion using Hampton Low Ionic Strength Screen D2E7 was buffered into a 20 mM HEPES / 150 mM sodium chloride buffer at pH 7.4. The protein concentration was adjusted to 5 mg/mL. 20 Greased VDX plates and circle siliconized glass cover slides were used. 1 mL of 24% w/v PEG 3,350 dehydrant solution was pipetted into 108 wells, respectively. Around 2 pL of protein sample were pipetted onto a cover slide and subsequently admixed with around 1 pL of one of the 18 particular buffer reagents. Thereafter, around 2.5 pL of 25 PEG 3,350 precipitant of one of six different concentrations was added to the drop. The wells were sealed with the inverted cover slides, generating 108 different hanging drop experiments. The plates were stored at ambient temperature. Microscopy of the drops was per 30 formed multiple times during the following seven days. The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops containing mixtures of precipitated species and crystals. RESULTS: From the 108 conditions tested, none rendered crystals. 35 49 Example 30 - Ammonium sulfate / Bis-Tris Propane grid screen in hanging drop vapor diffusion mode 5 D2E7 was buffered into a 20 mM HEPES / 150 mM sodium chloride buffer at pH 7.4. The protein concentration was adjusted to 5 mg/mL, 10 mg/mL, or 20 mg/mL. A greased VDX plate and circle siliconized glass cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing ammonium sulfate stock solu 10 tion, Bis-Tris propane stock solution and Milli Q water in each well. In this example, ammonium sulfate molarity was around 0.5 M, 1 M, 1.5 M or 2 M. The Bis-Tris Propane molarity was 0.1 M throughout, and the Bis-Tris Propane buffer pH was around 5.5, 6.0, 6.5, 7.0, 7.5 or 8.0. The resulting 24 conditions were assessed with all of the three protein concentrations as described above, and with storage at ambient temperature or 15 storage at around 27 *C, respectively. Around 1 pL of protein solution was admixed with around 1 pL of a particular reservoir solution on a circle siliconized glass cover slide, and the well was sealed with the inverted slide, generating a hanging drop ex periment. The plates were stored at ambient temperature. Microscopy of the drops was performed after three days. The conditions were classified into clear drops, drops con 20 taining random precipitation, drops containing crystals and drops containing mixtures of precipitated species and crystals. RESULTS: From the 144 conditions tested, none rendered crystals after three days. 25 Example 31 - Sodium potassium dihydrogen phosphate / sodium acetate grid screen in hanging drop vapor diffusion mode D2E7 was buffered into a 20 mM HEPES / 150 mM sodium chloride buffer at pH 7.4. 30 The protein concentration was adjusted to 5 mg/mL, 10 mg/mL, or 20 mg/mL. A greased VDX plate and square OptiClear plastic cover slides were used. 500 pL of a particular reservoir solution was prepared by admixing acetate buffer, sodium dihydro gen phosphate stock solution, potassium dihydrogen phosphate stock solution and Milli 50 o water in each well. In this example, the acetate buffer molarity was kept constant at around 0.1 M, and the acetate buffer pH was around 4.1, 4.6, 5.1 or 5.6. The following combinations of sodium dihydrogen phosphate and potassium dihydro 5 gen phosphate were applied: - around 0.3 M sodium dihydrogen phosphate and around 0.3 M potassium di hydrogen phosphate; - around 0.6 M sodium dihydrogen phosphate and around 0.6 M potassium di hydrogen phosphate; 10 - around 0.9 M sodium dihydrogen phosphate and around 0.9 M potassium di hydrogen phosphate; - around 1.8 M sodium dihydrogen phosphate, - around 2.1 M sodium dihydrogen phosphate, - around 2.4 M sodium dihydrogen phosphate. 15 Each condition was set up with the three protein concentrations as described above. Around 1 pL of protein solution was admixed with around 1 pL of a particular reservoir solution on a square OptiClear plastic cover slide, and the well was sealed with the inverted slide, generating a hanging drop experiment. The plates were stored at ambi 20 ent temperature. Microscopy of the drops was performed multiple times during the fol lowing month. The conditions were classified into clear drops, drops containing random precipitation, drops containing crystals and drops containing mixtures of precipitated species and crystals. 25 RESULTS: From the 72 wells assessed, the following crystallization buffers generated crystals in the shape of needle clusters: - around 0.9 M sodium dihydrogen phosphate and around 0.9 M potassium di hydrogen phosphate, at pH around 4.1; 30 - around 1.8 M sodium dihydrogen phosphate without the potassium salt, at pH around 4.6. Crystals were obtained with these conditions at all three protein concentrations. 35 51 Example 32 - Sodium potassium dihydrogen phosphate / sodium acetate grid screen in hanging drop vapor diffusion mode, different temperature The experimental conditions were identical with those of Example 31, except that the 5 storage temperature was increased to 30 *C. RESULTS: From the 72 wells assessed, following crystallization buffers generated crystals in the shape of needle clusters: 10 Protein concentration of around 5 mg/mL: - around 0.9 M sodium dihydrogen phosphate and around 0.9 M potassium di hydrogen phosphate at pH around 4.1; - around 1.8 M sodium dihydrogen phosphate without the potassium salt, at pH around 4.1. 15 - around 1.8 M sodium dihydrogen phosphate without the potassium salt, at pH around 4.6. - Around 1.8 M sodium dihydrogen phosphate without the potassium salt, at pH around 5.1. 20 Protein concentration of around 10 mg/mL: - around 0.9 M sodium dihydrogen phosphate and around 0.9 M potassium di hydrogen phosphate, at pH around 4.1. - around 1.8 M sodium dihydrogen phosphate without the potassium salt, at pH around 4.6. 25 - around 1.8 M sodium dihydrogen phosphate without the potassium salt, at pH around 5.1. Protein concentration of around 20 mg/mL: - around 0.9 M sodium dihydrogen phosphate and around 0.9 M potassium di 30 hydrogen phosphate at pH around 4.1 and - around 1.8 M sodium dihydrogen phosphate without the potassium salt, at pH around 4.1. Discussion of results of vapor diffusion crystallization experiments: 35 52 Crystallization experiments were initially performed using a well-described micro scale methodology. Since a PEG 4,000 / sodium acetate buffer condition was described as a promising crystallization condition by other inventors who were working with different antibodies of different antigen specificity or origin, it was decided to start with these 5 agents. It was found after extensive experimentation that PEG 4,000 in an acetate buffer did not provide crystals, at least at investigated combinations of factors influenc ing crystallization (protein concentration, precipitating agent concentration, buffer ionic strength and pH, temperature), and thus it was decided to continue with broad crystalli zation screens, thereby introducing a wide variety of chemicals into the screening 10 process. Finally, it was surprisingly found that sodium dihydrogen phosphate in acetate buffer is a powerful crystallization agent for D2E7, which does not introduce any toxic reagent unacceptable from a pharmaceutical point of view. 15 D. Batch crystallization experiments Concentration values given in the following examples are initial values referring to the antibody solution and the crystallization solution before mixing of the two solutions. 20 All pH values, if not described otherwise, refer to the pH of an acetate buffer stock be fore it was combined with other substances, like the crystallization agent. All buffer molarities, if not described otherwise, refer to sodium acetate concentrations in a stock solution before pH adjustment, typically performed using acetic acid glacial. 25 Example 33 - Sodium potassium dihydrogen phosphate / sodium acetate batch crystallization at 800 pL batch volume 30 D2E7 was buffered into a 20 mM HEPES / 150 mM sodium chloride buffer at pH 7.4. The protein concentration was adjusted to 5 mg/mL, 10 mg/mL, or 20 mg/mL. Batch crystallization was performed by admixing around 400 pL of each protein solu tion with an equal amount of crystallization solution in a 1.5 mL Eppendorff reaction 35 tube. 400 pL of a particular crystallization solution was prepared by admixing acetate 53 buffer, sodium dihydrogen phosphate stock solution, potassium dihydrogen phosphate stock solution and Milli Q water. In this example, the acetate buffer molarity was 0.1 M, and the acetate buffer pH was around 4.1. The following combination of sodium dihy drogen phosphate and potassium dihydrogen phosphate was used: around 0.9 M so 5 dium dihydrogen phosphate and around 0.9 M potassium dihydrogen phosphate. The reaction tubes were stored at ambient temperature. Microscopy of 1 pL aliquots was performed after 11 days. RESULTS: No crystals were observed after 11 days. 10 Example 34 - Sodium dihydrogen phosphate / sodium acetate batch crystalliza tion at 600 pL batch volume 15 D2E7 was buffered into a 20 mM HEPES / 150 mM sodium chloride buffer at pH 7.4. The protein concentration was adjusted to 10 mg/mL. Batch crystallization was performed by admixing around 300 pL of the protein solution with an equal amount of crystallization solution in a 1.5 mL Eppendorff reaction tube. 20 300 pL of a particular crystallization solution was prepared by admixing acetate buffer, sodium dihydrogen phosphate stock solution and Milli 0 water. In this example, the acetate buffer molarity was 0.1 M, and the acetate buffer pH was around 4.1. Sodium dihydrogen phosphate molarity was around 1.5 M, 1.8 M, 2.1 M and 2.4 M, respec tively. The reaction tubes were stored at ambient temperature. Microscopy of 1 pL ali 25 quots was performed after 11 days. RESULTS: No crystals were observed after 11 days. 30 Example 35 - Sodium potassium dihydrogen phosphate / sodium acetate grid screen batch crystallization at 1 mL batch volume D2E7 was used without exchanging the buffer. Thus, the initial composition was D2E7 50 mg/mL, mannitol 12 mg/mL, polysorbate 80 1 mg/mL, citric acid monohydrate 1.305 35 mg/mL, sodium citrate 0.305 mg/mL, disodium hydrogen phosphate dihydrate 1.53 54 mg/mL, sodium dihydrogen phosphate dehydrate 0.86 mg/mL, and sodium chloride 6.16 mg/mL, pH 5.2. D2E7 was brought to a concentration of around 10 mg/mL by dilution with Milli 0 water. 5 Batch crystallization was performed by admixing around 500 pL of the protein solution with an equal amount of crystallization solution in well of a 24 well plate. 500 pL of a particular crystallization solution was prepared by admixing acetate buffer, sodium di hydrogen phosphate stock solution, potassium dihydrogen phosphate stock solution 10 and Milli Q water in a well. In this example, the acetate buffer molarity was 0.1 M, and the acetate buffer pH was around 4.1, 4.6, 5.1 or 5.6. The following combinations of sodium dihydrogen phosphate and potassium dihydrogen phosphate were used: - around 0.7 M sodium dihydrogen phosphate and around 0.7 M potassium di 15 hydrogen phosphate, - around 0.9 M sodium dihydrogen phosphate and around 0.9 M potassium di hydrogen phosphate, - around 1.8 M sodium dihydrogen phosphate without the potassium salt, - around 2.1 M sodium dihydrogen phosphate without the potassium salt, 20 - around 2.4 M sodium dihydrogen phosphate without the potassium salt. The wells were subsequently sealed after preparation of the crystallization mixture to prevent water evaporation. Microscopy of the plate was performed after 4 days. 25 RESULTS: No crystals were observed after 4 days. Example 36 - Sodium potassium dihydrogen phosphate / sodium acetate grid screen batch crystallization at 1 mL batch volume 30 D2E7 was buffered into a 20 mM HEPES / 150 mM sodium chloride buffer at pH 7.4. The protein concentration was adjusted to 10 mg/mL. Batch crystallization was performed by admixing around 500 pL of the protein solution 35 with an equal amount of crystallization solution in well of a 24 well plate. 500 pL of a 55 particular crystallization solution was prepared by admixing acetate buffer, sodium di hydrogen phosphate stock solution, potassium dihydrogen phosphate stock solution and Milli Q water in a well. In this example, the acetate buffer molarity was 0.1 M, and the acetate buffer pH was around 4.1 or 4.6. The following combinations of sodium 5 dihydrogen phosphate and potassium dihydrogen phosphate were applied: - around 1.8 M sodium dihydrogen phosphate and around 0.8 M potassium di hydrogen phosphate, - around 2.2 M sodium dihydrogen phosphate and around 0.6 M potassium di 10 hydrogen phosphate, - from around 2.6 M sodium dihydrogen phosphate to around 4.4 M sodium di hydrogen phosphate in 0.2 M steps without the potassium salt, respectively. The wells were subsequently sealed after preparation of the crystallization mixture to 15 prevent water evaporation. Microscopy of the plate was performed multiple times dur ing the following week. Furthermore, the crystal yield of three batches was determined by OD280. An aliquot of the suspension was centrifuged at 14,000 rpm, and the protein concentration in the supernatant was assessed. 20 RESULTS: Needle cluster like crystals were found in the following eight batches: - acetate buffer pH 4.1 and sodium dihydrogen phosphate molarity of around 3.6 M to around 4.4 M (in 0.2 M steps), 25 - acetate buffer pH 4.6 and sodium dihydrogen phosphate molarity of around 4.0 M to around 4.4 M (in 0.2 M steps). Crystal yield was assessed for the batches at acetate buffer pH 4.1 and sodium dihy drogen phosphate molarity of around 4.0 M to around 4.4 M. The crystal yield as de 30 termined by OD280 from residual protein concentration in the supernatant was above 95% after five days. Precipitated species were obviously present in these batches immediately after com bining the protein solution and the crystallization solution (milky suspension, typical 35 light microscopic picture). As no precipitated species were observed after five days, it 56 was concluded that formerly precipitated species rearranged into crystalline species. The protein is highly supersaturated in the crystallization mixture, and protein precipi tates immediately. Some protein may still be dissolved, now either only slightly super saturated or perhaps even below saturation. Crystals form, thereby further lowering the 5 concentration of dissolved protein. Furthermore, the precipitated species clearly redis solve over time and are incorporated into the growing crystals. Example 37 - sodium dihydrogen phosphate / sodium acetate grid screen batch 10 crystallization at 1 mL batch volume, different protein concentration D2E7 was used without exchanging the buffer (see Example 35). D2E7 was brought to a concentration of around 10 mg/mL by diluting the liquid with 15 Milli Q water. Batch crystallization was performed by admixing around 500 pL of the protein solution with an equal amount of crystallization solution in well of a 24 well plate. 500 pL of a particular crystallization solution was prepared by admixing acetate buffer, sodium di 20 hydrogen phosphate stock solution, potassium dihydrogen phosphate stock solution and Milli Q water in a well. In this example, the acetate buffer molarity was 0.1 M, and the acetate buffer pH was around 4.1 or 4.6. Sodium dihydrogen phosphate molarity was varied from around 2.6 M sodium dihydrogen phosphate to around 4.4 M sodium dihydrogen phosphate in 0.2 M steps. The wells were subsequently sealed after prepa 25 ration of the crystallization mixture to prevent water evaporation. Microscopy of the plate was performed multiple times during the following week. Furthermore, the crystal yield of one particular batch was determined by OD280. An aliquot of the suspension was centrifuged at 14,000 rpm, and the protein concentration in the supernatant was assessed. 30 RESULTS: Needle cluster-like crystals were found in the following six batches: - acetate buffer pH 4.1 and sodium dihydrogen phosphate molarity of around 35 3.4 M to around 4.4 M (in 0.2 M steps). 57 Crystal yield was assessed for the batch at acetate buffer pH 4.1 and sodium dihydro gen phosphate molarity of around 4.2 M. The crystal yield as determined by OD280 from residual protein concentration in the supernatant was above 95% after eight days. 5 Precipitated species were obviously present in these batches immediately after com bining the protein solution and the crystallization solution (milky suspension, typical light microscopic picture). As no precipitated species were observed after six days, it was concluded that a phase transition occured where formerly precipitated species 10 rearranged into crystalline species. Example 38 - sodium dihydrogen phosphate / sodium acetate batch crystalliza tion at 2 mL batch volume 15 D2E7 was buffered into a 20 mM HEPES / 150 mM sodium chloride buffer at pH 7.4. The protein concentration was adjusted to 10 mg/mL. Batch crystallization was performed by admixing around 1 mL of the protein solution with an equal amount of crystallization solution in a 2 mL Eppendorff reaction tube. 1 20 mL of a particular crystallization solution was prepared by admixing acetate buffer, so dium dihydrogen phosphate stock solution and Milli Q water. In this example, the ace tate buffer molarity was 0.1 M, and the acetate buffer pH was around 4.1. Sodium di hydrogen phosphate molarity was around 4.0 M, 4.2 M or 4.4 M. The reaction tubes were stored at ambient temperature. Microscopy of 1 pL aliquots was performed multi 25 ple times during the following week. RESULTS: Needle cluster-like crystals were found in all batches after six days. Precipitated species were obviously present in these batches immediately after com 30 bining the protein solution and the crystallization solution (milky suspension, typical light microscopic picture). Formerly precipitated species rearranged into crystalline species as described in Example 36. 58 Example 39 - sodium dihydrogen phosphate / sodium acetate grid screen batch crystallization at 1 mL batch volume, different protein concentration D2E7 was used without exchanging the buffer (see Example 35). 5 Batch crystallization was performed by admixing around 500 pL of the protein solution with an equal amount of crystallization solution in well of a 24 well plate. 500 pL of a particular crystallization solution was prepared by admixing acetate buffer, sodium di hydrogen phosphate stock solution, potassium dihydrogen phosphate stock solution 10 and Milli Q water in a well. In this example, the acetate buffer molarity was 0.1 M, and the acetate buffer pH was around 4.1. Sodium dihydrogen phosphate molarity was var ied from around 0.2 M to around 4.4 M in 0.2 M steps. The wells were subsequently sealed after preparation of the crystallization mixture to prevent water evaporation. Mi croscopy of the plate was performed multiple times during the following week. Further 15 more, the crystal yield of the batch was determined by OD280. An aliquot of the sus pension was centrifuged at 14,000 rpm, and the protein concentration in the super natant was assessed. RESULTS: Needle cluster-like crystals were found in the following two batches: 20 - sodium dihydrogen phosphate molarity of around 3.4 M and around 3.6 M. Precipitated species and oily precipitation phases were also present in these crystal containing batches 25 Example 40 - sodium dihydrogen phosphate / sodium acetate batch crystalliza tion at 20 mL batch volume, agitation 30 D2E7 was used without exchanging the buffer (see Example 35). D2E7 was brought to a concentration of around 10 mg/mL by dilution with Milli Q water. Batch crystallization was performed by admixing around 10 mL of protein solution with 35 an equal amount of crystallization solution in a 50 mL Falcon tube. 10 mL of the crystal 59 lization solution was prepared by admixing acetate buffer, sodium dihydrogen phos phate stock solution and Milli Q water in the tube. In this example, the acetate buffer molarity was 0.1 M, and the acetate buffer pH was around 4.1. Sodium dihydrogen phosphate molarity was 4.2 M. The tube was stored at ambient temperature, agitating 5 the batch on a laboratory shaker. Microscopy of a 1 pL aliquot of the solution was per formed multiple times during the following month. RESULTS: Precipitated matter was observed in this batch. 10 Example 41a - Sodium dihydrogen phosphate / sodium acetate batch crystalliza tion at 100 mL batch volume, no agitation D2E7 was used without exchanging the buffer (see Example 35). 15 D2E7 was brought to a concentration of around 10 mg/mL by dilution with Milli Q water. Batch crystallization was performed by admixing around 50 mL of protein solution with an equal amount of crystallization solution in a clean 1 L polypropylene bottle. 50 mL of 20 the crystallization solution was prepared by admixing acetate buffer, sodium dihydro gen phosphate stock solution and Milli Q water in the tube. In this example, the acetate buffer molarity was 0.1 M, and the acetate buffer pH was around 4.1. Sodium dihydro gen phosphate molarity was 4.2 M. The container was stored at ambient temperature. Microscopy of a 1 pL aliquot of the solution was performed multiple times during the 25 following month. RESULTS: Needle cluster like crystals were observed after seven days. The crystal yield as determined by OD280 from residual protein concentration in the supernatant was above 95% after seven days. 30 Precipitated species were present in this batch immediately after combining the protein solution and the crystallization solution (milky suspension, typical light microscopic pic ture). Since no precipitated species were observed after seven days, it was concluded that a phase transition occurred where formerly precipitated species rearranged into 35 crystalline species. 60 Example 41b - Sodium dihydrogen phosphate / sodium acetate batch crystalliza tion at 1 mL, 50 mL and 10 L batch volume, No agitation 5 Large-scale crystallization of D2E7 was also performed by combining 1 L of 50 mg/mL D2E7 in Adalimumab commercial buffer formulation pH 5.2 (see Example 35) and 4 L water for injection (WFI) in a 10 L polypropylene vessel (Nalgene@). The solution was homogenized by gentle shaking. This 5 L D2E7 solution (10 mg/mL) was then mixed 10 with 5 L of precipitating agent solution (5 M sodium dihydrogen phosphate, 4,400 mL, 1 M sodium acetate buffer, pH 4.1, 500 mL, WFI (Ampuwa), 100 mL) The precipitating agent solution was added in 500 mL portions. After addition of each portion the solution was homogenized by gently rotating/inverting the bottle. After addition of around 2,500 to -3,000 mL of the precipitating agent solution, a white precipitate appeared. The re 15 maining precipitating agent was added all at once. Then the crystal preparation was homogenized (gently rotating/inverting) again. Immediately after batch manufacture (i.e. after admixing of 5 L D2E7 solution and 5 L precipitating agent), 1 mL (filled into low volume Eppendorf sample vials) and 50 mL 20 aliquots (filled into 50 mL Falcon sample tubes) were pulled and stored adjacent to the 10 L vessel for control and for evaluation of the impact of batch size on D2E7 crystalli zation. As outlined by Figs. 2 to 4, the batch volume (i.e. 1 mL, 50 mL and 10 L, re spectively) had no impact on D2E7 crystal needle/needle cluster size. 25 Discussion of results of batch crystallization experiments: As the applied micro scale technique (see Section D. supra) is not feasible for large scale production of protein crystals, the crystallization conditions discovered by these 30 micro scale methods were transferred into a scaleable batch mode. D2E7 was successfully crystallized at 100 mL batch volume with ultimately high yield (> 95%) and reproducibility, indicating that this crystallization system is applicable for in dustrial processing. By SDS-PAGE analysis, the protein character of the crystals was 35 proven. SE-HPLC analysis of redissolved crystals showed only a slight increase in ag 61 gregated species. Washing of the crystals was possible by using an acetate buffer con taining sodium dihydrogen phosphate around 4.2M sodium dihydrogen phosphate in around 0.1M sodium acetate at a pH around 4.1. No measurable solubility of D2E7 crystals in such a washing buffer occurs, as analyzed by OD280, recovering more than 5 90% of the crystals. The experimental conditions of the above batch experiments are summarized in the following Table 1: 62 Table 1 - Batch Experiments KO 0 M0 CO S~ 0E 0 'Z EE2. X . :( cc 33 800 pi 0.1 M NaAc, NaH2PO4 0.9 M, yes none 4.1 2.5-10 amb 11 d KH2PO4 0.9 M 34 600 p1 0.1M NaAc, NaH2PO41,5-24M yes none 4.1 5 amb -. d 35 1 ml 0.1 M NaAc, NaH2PO4 0.7 M, none 4.1-5.6 5 amb 4 d KH2PO4 0.7 M 0.1 M NaAc, NaH2PO4 0.9 M, 4.1-5.6 KH2PO4 0.9 M 0.1lM NaAc, NaH2PO4 1.8 M, 4.1-5.6 0.1 M NaAc, NaH2PO4 2.1 M, 4.1-5.6 0.1 M NaAc, NaH2PO4 2.4 M, 4.1-5.6 36 1 ml 0.1M NaAc, NaH2PO4 3.6-4.4 M yes >95% 4.1 3.9- 5 amb 5 d 3.7 0.1 M NaAc, NaH2PO4 4.0 - 4.4 M 4.6 4.0- amb 3.9 37 1 ml 0.1 M NaAc, NaH2PO4 3.4 - 4.4 M >95% 4.1 3.9- 5 amb 6d 3.7 38 2 ml .1M NaAc, NaH2PO4 4.0-4.4 M yes n.d. 4.1 3.9- 5 amb 6d 3.7 39 1 ml 0.1M NaAc, NaH2PO4 3.4 - 3.6 M n.d + 4.1 25 amb 6d precipi tate 40 20 ml 0.1M NaAc, NaH2PO4 4.2M precipi 4.1 3.8 5 amb 4d with tate agitatio n 41a 100 ml 0.1M NaAc, NaH2PO4 4.2 M >95% 4.1 3.8 5 amb 7d no agitatio n 41b 1ml, 0.1M NaAc, NaH2PO4 4.4 M n.d. 4.1 n.d. 5 amb 50ml or 101 5 63 E. Methods for crystal processing and analysis Example 42 - Washing of crystals 5 After formation of the crystals, a washing step without redissolving the crystals is fa vourable. Therefore, after the crystallization process was finished, the crystal slurry was transferred into a centrifugation tube and centrifuged at 500 to 1000 x g for twenty minutes. The centrifugation was performed at 4 0C, but might also be performed at 10 other feasible temperatures, e.g. room temperature. After centrifugation, the super natant was discarded, and the crystal pellet was resuspended in a buffer containing around 4.2 M sodium dihydrogen phosphate in around 0.1 M sodium acetate at a pH around 4.1. No measurable solubility of D2E7 crystals in the washing buffer occured, as analyzed by OD280. The centrifugation / resuspension steps were subsequently 15 repeated for one to three times, and after this washing procedure, the pellet was re suspended and stored in a buffer containing around 4.2 M sodium dihydrogen phos phate in around 0.1 M sodium acetate at a pH around 4.1. 20 Example 43 - Analysis of crystals by SDS-PAGE To prove the protein character of the crystals, the crystals were washed with a washing buffer as described in example 42. After assuring by OD280 that no more dissolved protein was present in the supernatant after centrifugation, the supernatant was dis 25 carded, and the crystals were subsequently dissolved in distilled water. OD280 meas urement of this solution revealed that the crystals essentially consisted of protein, as the absorbance of the sample was now significantly higher as in the residual washing buffer. SDS-PAGE analysis of this solution of redissolved crystals, when compared to an original D2E7 sample, showed the same pattern. 30 F. Miscellaneous Examples Concentration values given in the following examples are initial values referring to the antibody solution and the crystallization solution before mixing of the two solutions. 35 64 All pH values, if not described otherwise, refer to the pH of an acetate buffer stock be fore it was combined with other substances, like the crystallization agent. All buffer molarities, if not described otherwise, refer to sodium acetate concentrations 5 in a stock solution before pH adjustment, typically performed using acetic acid glacial. Example 44 - Solid crystallization agent D2E7 was used without exchanging the buffer (see Example 35). 10 D2E7 was brought to a concentration of around 10 mg/mL by diluting the liquid with Milli Q water. Batch crystallization was performed by admixing around 500 pL of the protein solution 15 with an equal amount of acetate buffer (0.1 M, pH 4.1 or 4.6, respectively) in a well of a 24 well plate. Subsequently, solid sodium dihydrogen phosphate dihydrate was added at six different ratios to each pH setting: around 0.23g, 0.27g, 0.30g, 0.33g, 0.36g and 0.39g. Thus, after complete dissolution of the crystallization agent, the concentration was around 1.5M to 2.5M in 0.2M steps. The wells were subsequently sealed and the 20 plate was agitated on a laboratory shaker until complete dissolution of the crystalliza tion agent. Thereafter, the 24 well plate was stored at ambient temperature without agitation. Microscopy of the plate was performed after five days. RESULTS: Needle cluster-like crystals were found in the following seven batches: 25 - acetate buffer pH 4.1 and sodium dihydrogen phosphate molarity of around 2.1 M, 2.3M and 2.5M, respectively. - acetate buffer pH 4.6 and sodium dihydrogen phosphate molarity of around 1.9M, 2.1 M, 2.3M and 2.5M, respectively. 30 Example 45 - Different buffer preparation protocol and preparation of crystals In this example, the acetate buffers were prepared as described in the following: 3 g of 35 acetic acid glacial were diluted with around 42 mL of purified water. The pH was ad 65 justed with sodium hydroxide solution and the volume adjusted to 50 mL. In this case, the total acetate amount is fixed at 1 M (100 mM in the crystallization solution, or 50 mM in the crystallization mixture) and not expanded by pH adjustment. 5 D2E7 was used without exchanging the buffer (see Example 35). D2E7 was brought to a concentration of around 10 mg/mL by diluting the liquid with Milli Q water. 10 Batch crystallization was performed by admixing around 500 pL of the protein solution with an equal amount of crystallization solution in well of a 24 well plate. 500 pL of a particular crystallization solution was prepared by admixing acetate buffer, sodium di hydrogen phosphate stock solution, potassium dihydrogen phosphate stock solution and Milli Q water in a well. In this example, the acetate buffer molarity was 0.1 M, and 15 the acetate buffer pH was around 4.1 and 4.6, respectively. Sodium dihydrogen phos phate molarity was varied from around 3.4 M to around 4.4 M in 0.2 M steps. The wells were subsequently sealed after preparation of the crystallization mixture to prevent water evaporation. Microscopy of the plate was performed after five days. 20 RESULTS: Needle cluster-like crystals were found in the following nine batches: - acetate buffer pH 4.1 and sodium dihydrogen phosphate molarity of around 3.6 to 4.4, in 0.2 steps. - acetate buffer pH 4.6 and sodium dihydrogen phosphate molarity of around 25 3.6 to 4.2, in 0.2 steps. Example 46 - Preparation of encapsulated crystals 30 Crystals as obtained in example 41 are positively charged as determined via zeta po tential measurement using a Malvern Instruments Zetasizer nano. The crystals are washed and suspended in a buffer containing excipients which con serve crystallinity, and which has a pH that keeps the crystals charged. Subsequently, 35 an appropriate encapsulating agent is added to the crystal suspension. In this context, 66 an appropriate encapsulating agent is a (polymeric) substance with low toxicity, biode gradability and counter ionic character. Due to this counter ionic character, the sub stance is attracted to the crystals and allows coating. By this technique, the dissolution of crystals in media which do not contain any other excipient maintaining crystallinity is 5 preferably sustained. Example 47 - Preparation of encapsulated / embedded crystals Crystals are obtained as described in example 41. 10 The crystals are washed and suspended in a buffer containing excipients which con serve crystallinity. The crystals can then be - embedded by drying the crystals and combining these dried crystals with a car 15 rier, e.g. by compression, melt dispersion, etc. - encapsulated / embedded by combining a crystal suspension with a carrier so lution which is not miscible with water. The carrier precipitates after removal of the solvent of the carrier. Subsequently, the material is dried. - encapsulated / embedded by combining a crystal suspension with a water mis 20 cible carrier solution. The carrier precipitates as its solubility limit is exceeded in the mixture. - embedded by combining dried crystals or a crystal suspension with a water miscible carrier solution. - embedded by combining dried crystals with a carrier solution which is not water 25 miscible. G. Crystal characterization 30 In the following section experiments that were performed to determine whether crystal line monoclonal antibody D2E7 retains the bioactivity characteristic of never crystallized D2E7 upon redissolution of the crystalline material are summarized. G1. Bioactivity test with murine L-929 cells 35 67 a) General method The neutralizing effect of D2E7 solution against the cytotoxic effect of recombinant hu man TNF (rHuTNF) was determined. This involved incubating mouse L-929 cells as 5 indicator in a 96-well microtiter plate in the presence of various D2E7 concentrations for 48 hours with a defined amount of rHuTNF at 37 0C. The surviving cells were stained with crystal violet. The intensity of color was measured by spectrophotometry in the individual wells of the microtiter plate and evaluated. The ICo was measured, i.e. the concentration of D2E7 which reduced the cytotoxic effect of rHuTNF on L-929 cells 10 such that 50% of the cells survived. In a separate dilution box, starting from the 1 pg protein/mL dilutions, the 9 titer curve measuring points (curve dilutions) were prepared individually in the dilution tubes for sample and reference standard. 15 The L-929 cell suspension to be used was diluted with medium to provide a concentra tion of 60,000 cells/mL. Subsequently 100 pL per well of the respective cell concentra tion were pipetted into columns 1 -11 of the test plate. The wells in column 12 con tained only 100 pL of medium each. Incubation was applied at 37 0C and 5% (v/v) C02 20 for 24 hours in the test plate. After 24 hours' incubation, 50 pL of each of the 9 titer curve dilutions were transferred from the dilution box to the test plate for the reference standard or sample, i.e. for the reference standard to wells in rows A - D in columns 1 - 9 and for the sample to the 25 wells in rows E - H in columns 1 to 9. 50 pL of medium were pipetted into column 10; and 100 pL each were pipetted into columns 11 and 12. 30 50 pL of TNF reference standard (12.5 ng protein/mL medium) were pipetted into the wells in column 1 to 10, row A to H, whereby column 10 corresponded to the 100% lysis value (TNF control). Column 11 was a 100% growth control, and column 12 contained no cell material and thus acted as a blank. The final volume per well was 200 pL. 35 68 Incubation of the test plates was performed for 48 hours at 37 0C and with 5% Co 2 . Following incubation for 2 days, the liquids from the test plate wells were discarded by turning quickly and giving a single, vigorous downward shake. Then 50 pL of crystal violet solution (0.75% crystal violet, 0.35% sodium chloride, 32.4% ethanol and 8.6% 5 formaldehyde) were pipetted into each well. The solution was left in the wells for 15 minutes and then discarded as described above. Then the plates were washed and dried at room temperature for about 30 minutes. Subsequently, 100 pL of reagent solu tion (50% ethanol and 0.1% acetic acid) were pipetted into each well. Agitation of the plates (at about 300 rpm for 15 min) produced an evenly colored solution in each of the 10 wells. The absorbance of the dye in the test plate wells was measured in a plate photometer at 620 nm. Individual values were plotted on a graph, with the absorbance (y axis) be ing plotted against the respective dilution or concentration ng/mL (x axis) of antibody. 15 From the 4-parameter plot, the concentration was read off at which half the cells sur vive and half die (IC5o value). This concentration was calculated by parameter 3 of the 4-parameter function of the curve data. The mean values of the reference standard concentrations were calculated. The relative biological activity of the sample was calcu lated by dividing the mean IC5o value of the reference standard by the individual ICso 20 values of the sample and multiplication by 100%. The relative activities were then av eraged. b) Relative activity for D2E7 Crystals 25 The test was performed as a comparison of the biological activity of the sample to that of a reference standard. The absorption values, plotted versus the concentration of D2E7 and assessed by a 4-parameter nonlinear regression, revealed the IC50 values for the inhibition of the TNF effect by the antibody. Since both samples were run in four repeats on one microplate this results in four IC50 values for D2E7 reference standard 30 and sample respectively. Subsequently, the mean of the IC50 values of the reference standard was calculated and the relative activity of each repeat of the sample was as sessed by dividing the mean IC50 value of the reference standard by the relevant IC0o value of the sample and multiplication by 100%. 69 The test of the sample (crystal suspension 2.7 mg/mL, prepared as described in Ex ample 36) revealed a relative biological activity of 111 %. Thus, the sample can be considered as fully biologically active. 5 G2. Microscopic characterization In the following, data on microscopic characterization of crystals of D2E7 will be pre sented. a) Optical analysis of mAb crystal batch samples 10 After homogenization, aliquots of 1 to 10 pL sample volume were pipetted onto an ob ject holder plate and were covered with a glass cover slide. The crystal preparations were assessed using a Zeiss Axiovert 25 inverted light microscope equipped with E-PI 1 Ox oculars and 1 Ox, 20x and 40x objectives, respectively. Pictures were taken using a 15 digital camera (Sony Cybershot DSC S75). b) Optical analysis of vapor diffusion experiments, assessment of approximate crystal sizes and detection of birefringence 20 For this purpose, a Nikon Labophot microscope was used, equipped with CFW 10x oculars and 4x, 1 Ox, 20x and 40x objectives, respectively. For assessment of vapor diffusion experiments, the sample drops in the 24 well plates were screened. 25 Crystal sizes were assessed by transferring the microscopic picture onto a computer screen by means of a JVC TK C1380 color video camera, and by measuring the length or diameter of representative needle-like or needle cluster-like crystals, applying the JVC Digital Screen Measurement Comet software version 3.52a. Furthermore, the mi 30 croscope was equipped with a filter set (polarizer and analyzer) to assess the birefrin gent behaviour of samples. If the polarization directions of the polarizer and analyzer filters are set at a 90* angle relative to each other ("crossed polarizers"), no light will pass through to the micro 70 scope eyepiece; the image will appear dark or black. If now a sample, which is placed into the light beam between the crossed polarizers, is capable of rotating the polariza tion plane of the light, a distinct glimmering of the sample against a dark background will be observed. This behaviour, termed "birefringence", distinguishes ordered crystal 5 line (anisotropic) from unordered amorphous (isotropic) matter. As birefringence is characteristic for anisotropic matter, this glimmering appearance proves the existence of crystalline matter. However, the absence of birefringence does not exclude the exis tence of crystalline matter, as the crystals might also exhibit cubic symmetry and there fore be isotropic, like amorphous matter. 10 c) Results In the attached Figures 1 to 4 representative pictures of D2E7 crystals are presented. 15 Figure 1 shows D2E7 crystals obtained by small-scale batch crystallization according to Example 37 after 6 days at room temperature (5 mg/ml final protein concentration; Crystallization buffer: 4.2 M sodium dihydrogen phosphate in 0.1 M sodium acetate, pH 4.1). The crystals exhibited birefringence. 20 Figures 2 to 4 show D2E7 crystals obtained by large-scale batch crystallization accord ing to Example 41 b. Syringeability: A D2E7 crystal suspension 200 mg/mL protein incorporated in crystals and formulated in a buffer containing 20% (m/v) PEG 4,000 is syringeable through a 27 25 G needle. G3. Birefringence In order to demonstrate that Adalimumab crystallization in fact provides crystalline ma 30 terial its birefringence was analyzed. Protein: Diluted 70 mg/ml Adalimumab in formulation buffer with double-distilled water to 10mg/ml. 35 Precipitant:: 4 M NaH 2

PO

4 (dissolved powder in double-distilled water) 71 Method: Micro batch crystallization in a hanging-drop tray with 2 ml compartment well, Mixed 500 pl protein solution with 500 pl protein; no NaOAc in the solution. 5 Temperature: 24 *C. Technical Equipment for birefringence measurement: Nikon SMZ1500 stereo dis secting microscope equipped with a Nikon CoolPix CCD camera. Crystal birefringence was photographed under crossed polarizers. Magnification is approximately 200x. 10 Corresponding micrographs are shown in Figure 5 A. A marked birefringence of the clusters of Adalimumab needle-like crystals is observed. The colour of the crystals changes from blue to red, then back to blue, as the orientation of the crystal needle axis rotates relative to the light polarization direction. 15 A further set of micrographs is depicted in Figures 5 B, C and D. All images were taken with a Nikon Eclipse E600 POL microscope and a Nikon DXM 1200 digital camera. Magnification is approximately 40x. 20 The image of Figure 5B (grey) is taken with plane polars and shows the particle mor phology. The white crystals on the black background (Figure 5D) show birefringence and were taken with crossed polars. The blue and orange crystals on the purple back ground (Figure 5C) show birefringence and were taken with crossed polar and a red 25 compensator or quarter wave plate. H. Crystal Syringeability In the following section, experiments were performed to determine the syringeability of 30 crystalline suspensions (in PEG) of monoclonal antibody D2E7 (10-200 mg/ml) using different gauge needles. PEG buffer: 35 20% PEG 4,000 m/v 12 mg/mL mannitol 72 0.1 mg/mL polysorbate 80 1.305 mg/mL citric acid monohydrate 0.305 mg/mL sodium citrate 1.53 mg/mL di sodium hydrogen phosphate dehydrate 5 0.86 mg/mL sodium di hydrogen phosphate dehydrate pH was adjusted to 5.2 with sodium hydroxide Syringe depletion (1 mL filling volume) was performed as it would be manually by a patient in the course of administration. 20 - 27.5G needle sizes were evaluated. 10 Syringes: 20 / 23 / 26G: Henke Sass Wolf GmbH 1 mL Norm-Ject syringes, equipped with - Henke Sass Wolf GmbH Fine-Ject* 20G needles 15 - Terumo* 23G needles - Neopoint* 26G needles 27.5G: BD HyPak SCFM 1 mL long syringes, equipped with 27.5G RNS needles 38800 Le Pont du Claix 20 The results (Figure 6) suggest that higher gauge needles provide a slower delivery of the crystals at at high concentrations. I. Stability data (SE HPLC, FT-IR) 25 In the following section, experiments were performed to determine the stability of crys talline suspensions of monoclonal antibody D2E7 (50 and 200 mg/ml) over 12 month storage at 2-8*C. Crystals suspended in medium as used in syringeability studies: 30 20% PEG 4,000 m/v 12 mg/mL mannitol 0.1 mg/mL polysorbate 80 1.305 mg/mL citric acid monohydrate 0.305 mg/mL sodium citrate 35 1.53 mg/mL di sodium hydrogen phosphate dehydrate 0.86 mg/mL sodium di hydrogen phosphate dehydrate pH was adjusted to 5.2 with sodium hydroxide 73 SE-HPLC 50 mg/mL Adalimumab crystal suspension, stable at 2-8*C over time Time point Aggregates (%) Monomer (%) Fragments (%) TO 1.3 98.5 0.2 1 month 1.4 98.3 0.3 3 month 2.2 97.5 0.3 6 month 3.2 96.3 0.5 9 month 4.0 95.5 0.5 12 month 4.2 95.1 0.7 5 200 mg/mL Adalimumab Time point Aggregates (%) Monomer (%) Fragments (%) TO 1.3 98.5 0.2 1 month 1.3 98.4 0.3 3 month 1.9 97.8 0.3 6 month 2.4 97.2 0.4 9 month 2.5 97.0 0.5 12 month 2.6 96.8 0.6 A Dionex HPLC system (P680 pump, ASI 100 autosampler, UVD170U) was used for stability anaysis by SEC. D2E7 samples were separated on a GE Superose* 6 col 10 umn, applying a flow rate of 0.5 mL/min. UV quantitation (detection) was performed at a wavelength of 214 nm. The running buffer consisted of 0.15M sodium chloride in 0.02M sodium phosphate buffer pH 7.5. IR spectra were recorded with a Confocheck system on a Bruker Optics Tensor 27. Liquid samples were analyzed using a MicroBio lytics AquaSpec cell. Each sample was assessed performing at least two measure 15 ments of 120 to 500 scans at 25 0 C. Blank buffer spectra were subtracted from the pro tein spectra, respectively. Protein second derivative spectra were generated by Fourier transformation and vector normalized from 1580-1720 cm- 1 for relative comparison. Redissolution of crystals was performed as follows: Crystal suspensions were diluted with Humira commercial buffer to 10 mg/mL protein concentration. By decreasing PEG 20 concentration crystals redissolved. Blue - standard, Humira* after freeze / thaw Red - redissolved crystals after 6 month storage at 25 0 C, 50 mg/mL 74 Green - redissolved crystals after 6 month storage at 250C, 200 mg/mL Results: Figure 7 illustrates that there were no conformational differences over 6 months of storage at 250C 5 J. Morphology After 12 months of storage at 2-80C no significant morphological changes were observed by light microscopy analysis of the crystals. 10 Aliquots of 1 to 10 pL sample volume were pipetted onto an object holder plate, diluted with formulation buffer (20% PEG) and covered with a glass cover slide. The preparations were assessed using a Zeiss Axiovert 25 inverted light microscope equipped with E-PI 1Ox oculars and 1Ox, 20x and 40x objectives, respectively. 15 References: Baldock Peter; Mills, Vaughan; Stewart, Patrick Shaw, Journal of Crystal Growth (1996), 168(1-4, Crystallization of Biological Macromolecules), 170-174. 20 Connell, G. E., M. H. Freedman, et al. (1973). "Human IgG myeloma protein crystalliz ing with rhombohedral symmetry." Canadian Journal of Biochemistry 51(8): 1137-41. Harris, L. J., S. B. Larson, et al. (1992). "The three-dimensional structure of an intact monoclonal antibody for canine lymphoma." Nature (London, United Kingdom) 25 360(6402): 369-72. Huber, R., J. Deisenhofer, et al. (1976). "Crystallographic structure studies of an IgG molecule and an Fc fragment." Nature 264(5585): 415-20. Jen , A., Merkle, H. P. (2001), Diamonds in the rough: Protein Crystals from a from a formulation perspective, Pharm. Res. (2001),18, 11, 1483 30 Jentoft, J. E., D. G. Dearborn, et al. (1982). "Characterization of a human cryoglobulin complex: a crystalline adduct of a monoclonal IgG and albumin." Biochemistry 21(2): 289-294. Jones, H. B. (1848). "On a new substance occurring in the urine of a patient with mo/l ties ossium." Philosophical Transactions of the Royal Society, London 138: 55 35 62. McPherson, A. (1999). Crystallization of Biological Macromolecules. Cold Spring Har bor, New York, Cold Spring Harbor Laboratory Press. 75 Mills, L. E., L. R. Brettman, et al. (1983). "Crystallocryoglobulinemia resulting from hu man monoclonal antibodies to albumin." Annals of internal medicine 99(5): 601 4. Nisonoff, A., S. Zappacosta, et al. (1968). "Properties of crystallized rabbit anti-p 5 azobenzoate antibody." Cold Sprinq Harbor Symposia on Quantitative Biology 32: 89-93. Putnam, F. W. (1955). "Abnormal human serum globulins." Science (Washington, DC, United States) 122: 275-7. Rajan, S. S., K. R. Ely, et al. (1983). "Three-dimensional structure of the Mcg IgG1 im 10 munoglobulin." Molecular Immunology 20(7): 787-99. Sarma, V. R., E. W. Silverton, et al. (1971). "Three-dimensional structure at 6 Ang. Resolution of a human gG1 immunoglobulin molecule." Journal of Biological Chemistry 246(11): 3753-9. Shenoy, B., C. P. Govardhan, et al. (2002). Pharmaceutical compositions comprising 15 crystals of polymeric carrier-stabilized antibodies and fragments for therapeutic uses. PCT Int. Apple. WO, (Altus Biologics Inc., USA). 173 pp. Terry, W. D., B. W. Matthews, et al. (1968). "Crystallographic studies of a human im munoglobulin." Nature 220(164): 239-41. von Bonsdorf, B., H. Groth, et al. (1938). "On the Presence of a High-molecular Crys 20 tallizable Protein in Blood Serum in Myeloma." Folia Haematologia 59:184-208. Yang, M.X., B. Shenoy, et al. (2003). "Crystalline monoclonal antibodies for subcuta neous delivery." Proceedings of the National Academy of Sciences of the United States of America 100(12): 6934-6939. 25 76

Claims (26)

1. A batch crystallization method for crystallizing an anti-hTNFalpha antibody, 5 comprising the steps of: (a) providing an aqueous solution of said antibody in admixture with an inorganic phosphate salt as crystallization agent; and (b) incubating said aqueous crystallization mixture until crystals of said antibody are formed. 10

2. The crystallization method according to one of the preceding claims, wherein the pH of said aqueous crystallization mixture is in the range of about pH 3 to 5.

3. The crystallization method according to one of the preceding claims, wherein 15 said aqueous crystallization mixture contains a buffer.

4. The crystallization method according to claim 3, wherein said buffer com prises an acetate buffer. 20 5. The crystallization method according to claim 4, wherein the buffer comprises sodium acetate.

6. The crystallization method according to one of the claim 3 to 5, wherein the buffer concentration in said aqueous crystallization mixture is 0 M to 0.5 M. 25

7. The crystallization method according to one of the preceding claims, wherein the phosphate salt is a hydrogenphosphate salt.

8. The crystallization method according to claim 7, wherein the phosphate salt is 30 an alkali metal salt, or a mixture of at least two different alkali metal salts.

9. The crystallization method according to one of the preceding claims, wherein the phosphate salt concentration in the crystallization mixture is in the range of about 1 to 6 M. 35

10. The crystallization method according to claim 9, wherein the salt concentra tion in the crystallization mixture is in the range of about 1.0 to 3.0 M.

11. The crystallization method of one of the preceding claims, wherein at least 40 one of the following additional crystallization conditions are met: 77 a) incubation is performed for between about 1 hour to about 60 days; b) incubation is performed at a temperature between about 4 oC and about 37 0C; 5 c) the antibody concentration is in the range of about 0.5 to about 100 mg/ml.

12. The crystallization method according to one of the preceding claims, further comprising the step of drying said crystals. 10

13. The crystallization method according to one of the preceding claims, further comprising the step of exchanging the crystallization mother liquor with a differ ent buffer. 15 14. The crystallization method according to one of the preceding claims, wherein the batch volume is in the range of about 1 ml to 20.000 1.

15. A crystal of an anti-hTNFalpha antibody, obtainable by a crystallization method as defined in anyone of the preceding claims. 20

16. A crystal of an anti-hTNFalpha antibody, with the proviso that said antibody is not INFLIXIMAB.

17. The crystal of anyone of the claims 15 and 16 having a needle-like morphol 25 ogy with a maximum length I of about 2 - 500 pm and an l/d ratio of about 3 to

30. 18. The crystal according to anyone of claims 15 to 17, wherein said antibody is a polyclonal antibody or a monoclonal antibody. 30 19. The crystal according to claim 18, wherein said antibody is selected from the group consisting of: chimeric antibodies, humanized antibodies, non-glycosylated antibodies, human antibodies and mouse antibodies. 35 20. The crystal according to anyone of claims 17 or 18, wherein said antibody is an IgG antibody. 21. The crystal according to claim 19, wherein said antibody is selected from the group consisting of: IgG1, IgG2, IgG3 and IgG4, antibodies. 40 78 22. The crystal according to claim 21, wherein said antibody is an anti humanTNFalpha antibody of the group IgG1. 23. The crystal according to claim 22, prepared from an isolated human antibody, 5 that dissociates from hTNFalpha with a Kd of 1 x10 8 M or less and a Ka rate constant of 1 x 10,3 s- or less, both determined by surface plasmon resonance, and neutralizes hTNFalpha cytotoxicity in a standard in vitro L929 assay with an IC50 of 1 x10- 7 M or less. 10 24. The crystal according to claim 22, prepared from an isolated human antibody with the following characteristics: a) dissociates from human TNFalpha with a Kff rate constant of 1 x 10- s- or less, as determined by surface plasmon reso nance; b) has a light chain CDR3 domain comprising the amino acid sequence of SEQ ID NO: 3, or modified from SEQ ID NO: 3 by a single alanine substitution at 15 position 1, 4, 5, 7 or 8 or by one to five conservative amino acid substitutions at positions 1, 3, 4, 6, 7, 8 and/or 9; c) has a heavy chain CDR3 domain comprising the amino acid sequence of SEQ ID NO: 4, or modified from SEQ ID NO: 4 by a single alanine substitution at position 2, 3, 4, 5, 6, 8, 9, 10 or 11 or by one to five conservative amino acid substitutions at positions 2, 3, 4, 5, 6, 8, 9, 10, 11 and/or 20 12. 25. The crystal according to claim 22, prepared from an isolated human antibody with a light chain variable region (LCVR) comprising the amino acid sequence of SEQ ID NO: 1 and a heavy chain variable region (HCVR) comprising the amino 25 acid sequence of SEQ ID NO: 2. 26. The crystal according to claim 25, prepared from the antibody D2E7. 27. A pharmaceutical composition comprising: (a) crystals of an anti-hTNFalpha 30 antibody as defined in anyone of claims 15 to 26, and (b) at least one pharma ceutical excipient; which composition is provided in as solid, semisolid or liquid formulation, each formulation containing said antibody in crystalline form. 28. A pharmaceutical composition comprising: (a) crystals of an anti-hTNFalpha 35 antibody as defined in anyone of claims 15 to 26, and (b) at least one pharma ceutical excipient, which embeds or encapsulates crystals of said antibody. 29. The composition of claim 26, wherein said composition has an antibody con centration greater than about 1 mg/ml. 40 79 30. The composition of claim 27, wherein said composition has an antibody con centration greater than about 200 mg/ml.

31. The composition according to anyone of the claims 26 to 28, wherein said 5 excipient comprises at least one polymeric optionally biodegradable carrier or at least one oil or lipid carrier.

32. The composition according to claim 29, wherein said polymeric carrier is a polymer selected from one or more of the group consisting of: poly (acrylic acid), 10 poly (cyanoacrylates), poly (amino acids), poly (anhydrides), poly (depsipeptide), poly (esters), poly (lactic acid), poly (lactic-co-glycolic acid) or PLGA, poly (13 hydroxybutryate), poly (caprolactone), poly (dioxanone); poly (ethylene glycol), poly (hydroxypropyl) methacrylamide, poly[ (organo) phosphazene, poly (ortho esters), poly (vinyl alcohol), poly (vinylpyrrolidone), maleic anhydride alkyl vinyl 15 ether copolymers, pluronic polyols, albumin, alginate, cellulose and cellulose de rivatives, collagen, fibrin, gelatin, hyaluronic acid, oligosaccharides, glycami noglycans, sulfated polysaccharides, blends and copolymers thereof.

33. An injectable liquid composition comprising anti-hTNFalpha antibody crystals 20 as defined in anyone of claims 15 to 26 and having an antibody concentration in the range of about 10 to 400 mg/ml.

34. A crystal slurry comprising anti-hTNFalpha antibody crystals as defined in anyone of claims 15 to 26, having a antibody concentration greater than about 25 100 mg/ml.

35. A method for treating a mammal comprising the step of administering to the mammal an effective amount of antibody anti-hTNFalpha crystals as defined in anyone of claims 15 to 26. 30

36. A method for treating a mammal comprising the step of administering to the mammal an effective amount of the composition according to anyone of the claims 27 to 32. 35 37. The method according to anyone of the claims 34 and 35, wherein the com position is administered by parenteral route, oral route, or by injection.

38. A method of treating a TNFalpha-related disorder in a subject, which method comprises administering a therapeutically effective amount of antibody crystals 40 as defined in anyone of claims 15 to 26. 80

39. The method of claim 37, wherein said TNFalpha- related disorder is selected from the group consisting of 5 an autoimmune disease, in particular rheumatoid arthritis, rheumatoid spondylitis, osteoarthritis and gouty arthritis, an allergy, multiple sclerosis, autoimmune diabetes, autoimmune uveitis and nephrotic syndrome; an infectious disease, transplant rejection or graft versus-host disease, malignancy, pulmonary disorder, intestinal disorder, cardiac disor der, inflammatory bone disorders, bone resorption disease, alcoholic hepatitis, viral 10 hepatitis, fulminant hepatitis, coagulation disturbances, burns, reperfusion injury, keloid formation, scar tissue formation, pyrexia, periodontal disease, obesity and radiation tox icity; a spondyloarthropathy, a pulmonary disorder, a coronary disorder, a metabolic disorder, anemia, pain, a hepatic disorder, a skin disorder, a nail disorder, or vasculitis, Behcet's disease, ankylosing spondylitis, asthma, chronic obstructive pulmonary dis 15 ease (COPD), idiopathic pulmonary fibrosis(IPF), restenosis, diabetes, anemia, pain, a Crohn's disease-related disorder, juvenile rheumatoid arthritis (JRA), a hepatitis C virus infection, psoriasis, psoriatic arthritis, and chronic plaque psoriasis, age related cachexia, Alzheimer's disease, brain edema, inflammatory brain injury, chronic fatigue syndrome, dermatomyositis, drug reactions, edema in and/or 20 around the spinal cord, familial periodic fevers, Felty's syndrome, fibro sis,glomerulonephritides (e.g. post-streptococcal glomerulonephritis or IgA nephropathy), loosening of prostheses, microscopic polyangiitis, mixed connec tive tissue disorder, multiple myeloma, cancer and cachexia, multiple organ dis order, myelo dysplastic syndrome, orchitism osteolysis, pancreatitis, including 25 acute, chronic, and pancreatic abscess, periodontal disease polymyositis, pro gressive renal failure, pseudogout, pyoderma gangrenosum, relapsing poly chondritis, rheumatic heart disease, sarcoidosis, sclerosing cholangitis, stroke, thoracoabdominal aortic aneurysm repair(TAAA), TNF receptor associated peri odic syndrome (TRAPS), symptoms related to Yellow Fever vaccination, in 30 flammatory diseases associated with the ear, chronic ear inflammation, or pedi atric ear inflammation, uveitis, sciatica, prostatitis, endometriosis, choroidal ne ovascularization, lupus, Sjogren's syndrome, and wet macular degeneration.

40. The use of anti-hTNFalpha antibody crystals as defined in anyone of claims 35 15 to 26 for preparing a pharmaceutical composition for treating a TNFalpha related disease as defined in claim 38. 81

41. Anti-hTNFalpha antibody crystals as defined in anyone of claims 15 to 26 for use in medicine. 82