NZ786225A - Anti-tnfalpha-antibodies and functional fragments thereof - Google Patents

Abstract

The present invention relates to antibody molecules and functional fragments thereof, capable of binding to tumor necrosis factor alpha (TNFa), to processes for their production, and to their therapeutic uses.

Description

Anti-TNFα-antibodies and functional fragments thereof The present application is a divisional ation from New Zealand patent application number 745999, the entire sure of which is incorporated herein by reference.

FIELD OF THE INVENTION The present invention relates to antibody molecules and functional fragments thereof, capable of binding to tumor necrosis factor alpha , to processes for their production, and to their therapeutic uses.

BACKGROUND TNFα is ahomo-trimeric pro-inflammatory cytokine that is released by and interacts with cells of the immune system. TNFα has also been shown to be up-regulated in a number of human es, ing chronic diseases such as rheumatoid arthritis, Crohn's disease, ulcerative colitis and multiple sclerosis.

Antibodies to TNFα have been proposed for the prophylaxis and ent of endotoxic shock (Beutler et al., Science, 234, 470-474, 1985). Bodmer et al., (Critical Care Medicine, 21, S441-S446, 1993) and Wherry et al., (Critical Care Medicine, 21, S436-S440, 1993) discuss the therapeutic ial of anti-TNFα antibodies in the treatment of septic shock.

The use of anti-TNFα antibodies in the treatment of septic shock is also discussed by Kirschenbaum et al., (Critical Care Medicine, 26, 1625-1626, 1998). Collagen-induced arthritis can be treated effectively using an anti-TNFα monoclonal antibody (Williams et al.

(PNAS-USA, 89, 9784-9788, 1992)).

The use of anti-TNFα antibodies in the treatment of toid arthritis and Crohn's disease is discussed in Feldman et al. (Transplantation Proceedings, 30, 4126-4127, 1998), Adorini et al. (Trends in Immunology Today, 18, 209-211, 1997) and in Feldman et al.

(Advances in Immunology, 64, 0, 1997). The dies to TNFα previously used in such treatments are generally chimeric antibodies, such as those described in U.S. Pat. No. ,919,452.

Monoclonal antibodies t TNFα have been described in the prior art. Meager et al.

(Hybridoma, 6, 305-311, 1987) describe murine monoclonal antibodies against inant TNFα. Fendly et al. (Hybridoma, 6, 0, 1987) describe the use of murine monoclonal antibodies against recombinant TNFα in defining neutralising epitopes on TNFα.

Furthermore, in International Patent Application WO 92/11383, recombinant antibodies, including CDR-grafted antibodies, specific for TNFo are disclosed. Rankin et al. (British J.

Rheumatology, 34, 334-342, 1995) describe the use of such CDR-grafted antibodies in the treatment of rheumatoid arthritis. U.S. Pat No. 5,919,452 discloses anti-TNFo chimeric antibodies and their use in treating pathologies associated with the presence of TNFo.

Further anti-TNch antibodies are disclosed in Stephens et al. (Immunology, 85, 668-674, 1995), GB-A-2 246 570, GB-A—2 297 145, US 310, US 193400, EP 2 390 267 B1, US 8,293,235, US 8,697,074, WO 55723 A2 and The prior art inant anti-TNch antibody molecules generally have a reduced affinity for TNFo compared to the antibodies from which the hypervariable regions or CDRs are derived. All currently marketed inhibitors of TNFo are administered enously or subcutaneously in weekly or longer intervals as bolus injections, resulting in high ng concentrations that are steadily decreasing until the next injection.

Currently approved anti-TNFo rapeutics include (i) infliximab, a chimeric lgG anti- human monoclonal antibody (Remicade®; ski M et al: "lnfliximab (Remicade)", Handbook of Therapeutic Antibodies, WILEY-VCH; Weinheim, 200701, p.885-904); (ii) etanercept, a TNFR2 dimeric fusion protein, with an lgG1 Fc (Enbrel®); (iii) adalimumab, a fully human monoclonal antibody (mAb) (Humira®; Kupper H et al: "Adalimumab (Humira)", Handbook of Therapeutic Antibodies, WILEY-VCH; Weinheim, 200701, p.697-732); (iv) izumab, a PEGylated Fab fragment (Cimzia®; Melmed G Y et al: "Certolizumab pegol", Nature Reviews. Drug ery, Nature Publishing Group, GB, Vol. 7, No. 8, 2008- 08-01, p.641-642); (v) Golimumab, a human lgGlK monoclonal antibody (Simponi®; ar S et al: umab", mAbs, Landes ence, US, Vol. 1, No. 5, 200901, p.422-431). However, various biosimilars are in development, and a mimic of infliximab known as Remsima has already been approved in Europe. lnfliximab has a relatively low affinity to TNFo (KD > 0.2 nM; Weir et al., 2006, Therapy 3: 535) and a limited neutralization potency in an L929 assay. In addition, infliximab shows substantially no reactivity with TNFo from Cynomo/gus or Rhesus monkeys. For anti- TNFo dies, however, cross-reactivity with TNFo from monkeys is highly desirable, as this allows for animal tests with primates, reflecting the situation in humans in many aspects.

Etanercept, gh a bivalent molecule, binds TNFo at a ratio of one trimer per one etanercept molecule, ding the formation of large antigen-biotherapeutics complexes (Wallis, 2008, Lancet Infect Dis, 8: 601). It does not inhibit LPS—induced cytokine secretion in monocytes (Kirchner et al., 2004, Cytokine, 28: 67).

The potency of adalimumab is similar to that of infliximab. Another disadvantage of adalimumab is its poor stability, e.g. as determined in a thermal unfolding test. The melting ature (Tm) of adalimumab in such a test was determined to be 67.5 °C. The lower the Tm value of an antibody, however, the lower is its l stability. A lower Tm makes antibodies less suitable for pharmaceutical use, e.g. for oral administration.

The potency of certolizumab is slightly greater than that of infliximab, but still not satisfying.

Certolizumab does not inhibit T-cell proliferation in a MLR (Vos et al., 2011, Gastroenterology, 140: 221 ). 515 A1 discloses humanized anti-TNFoc antibodies and antigen-binding fragments (Fab) thereof. As becomes clear from the disclosed examples, the potency of the resulting humanized Fab nts is comparable to that of infliximab in a L929 neutralization assay (see Table 2 and 5). The sole anti-TNch lgG dy tested for cross-reactivity binds only weakly to Rhesus TNF-cx (see [0069]; Fig. 3). Cross-reactivity with Cynomolgus TNch was not tested. er, there is weak binding to human TNFB (see Fig. 3). Therefore, EP2623515 A1 does not disclose anti-TNFo antibodies or onal fragments thereof, which have a y to inhibit TNFo-induced apoptosis in L929 cells greater than that of imab and which are cross-reactive with Rhesus TNFoc and Cynomolgus TNch. in the form of fusion proteins comprising one or more single antigen binding domains that bind to one or more targets (e.g. TNFoc), a linker and one or more polymer molecules. The only specific example given is termed SDAB-01 and includes two antigen binding domains, which bind to TNFoc, connected with a flexible linker, and a C-terminal Cysteine supporting the site ic PEGylation (see Fig. 3). of SDAB-01 to known TNFoc antibodies like infliximab in a L929 cell-based neutralization assay, or to assess other SDAB-Ol-specific parameters like the effectiveness to block TNFoc -TNFR|/ll interaction and the selectivity for binding TNFoc over TNFB. In an assay where the treatment with SDAB-01 and imab are compared in a enic mouse model for polyarthritis that overexpresses human TNFoc (see page 54, Example 8), the two seem to be similarly effective in preventing r development of arthritis (e.g. Fig. 17&18).

However, the dosage given in this example is ding as the molecular weight of SDAB- 01 is less than half of that of infliximab. Thus, WO 07880 A2 does not disclose anti- TNch antibodies having a y to inhibit TNFor-induced sis in L929 cells r than that of infliximab.

This scFv showed a TNch neutralization capacity in a PK-15 cell assay that was comparable to that of infliximab (see [0236]). In addition, the scFv seems to have some cross-reactivity to TNF-cx from rhesus macaque and cynomolgus monkey (see Ex. 8). No affinity data are reported in DLX105 (also known as ESBA105), however, which is known to have only moderate affinity 7 pM; see Urech et al. 2010 Ann Rheum Dis 69: 443), shows a better binding to TNF-o than scFv1 (see Fig. 1 of does not disclose anti-TNF-cx antibodies having high affinity for human TNch (KD < 125 pM). antibodies binding to both antigens. Certain anti-TNch antibodies showed some cross- reactivity with TNFo from Cynomolgus (Fig. 17). The anti-TNFo antibodies, however, exhibited a significantly lower potency than infliximab in an L929 neutralization assay ([0152]; Fig. 5). Therefore, having a y to inhibit nduced apoptosis in L929 cells greater than that of infliximab.

Drugs in R&D, Vol. 4 No. 3, 2003, pages 174-178 desribes the zed antibody "Humicade" (CDP 571; BAY 103356), a monoclonal anti-TNFoc antibody with high affinity.

The potency of Humicade to inhibit TNFor-induced apoptosis in L929 cells, however, appears to be limited (see, e.g., US 2003/0199679 A1 at [0189]). The reference therefore does not disclose anti-TNF-d antibodies having a potency to inhibit TNFd-induced apoptosis in L929 cells greater than that of infliximab.

Saldanha J W et al: "Molecular Engineering I: Humanization", Handbook of Therapeutic Antibodies, Chapter 6, 200701, WILEY-VCH, Weinheim, p.119-144 discloses ent strategies for humanization of monoclonal antibodies including CDR Grafting, ResurfacingNeneering, SDR transfer and Delmmunization Technology.

There is a need for improved antibody molecules to treat c inflammatory diseases such as inflammatory bowel disorders. The dy molecules should at least have (i) high affinity for human TNFq (Le. a KD < 125 pM), (ii) a high potency to inhibit TNFq-induced apoptosis in L929 cells, (iii) a high potency to inhibit LPS-induced cytokine secretion, (iv) substantial affinity to TNFq from Cynomolgus and Rhesus (e.g. a KD < 1 nM), and (v) a high melting temperature of the le domain as determined in a thermal unfolding experiment (e.g. a Tm > 70°C).

SUMMARY OF THE INVENTION The inventors of the present application found that certain anti-TNFq antibodies and functional fragments thereof exhibit a combination of favorable properties, including high ty for human TNFq (KD < 125 pM), a potency to inhibit nduced apoptosis in L929 cells greater than that of infliximab, a potency to t duced cytokine secretion greater than that of adalimumab, and substantial affinity (KD< 1 nM) to TNFq from animals such as Cynomolgus monkey (Macaca fascicularis) and/or Rhesus macaques (Macaca mulatta). In addition, the antibodies and functional fragments thereof were specific for TNFq in that they did not icantly bind to TNFB, and exhibit a icant stability, as determined in a thermal unfolding assay of the variable .

The invention provides antibody molecules and onal fragments thereof.

The present invention therefore relates to the subject matter defined in the following items (1) to (48): (1) An antibody or a functional fragment thereof capable of binding to human tumor necrosis factor alpha (TNFq), wherein said antibody or functional fragment thereof comprises (i) a VL domain comprising a CDR1 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:1, a CDR2 region having an amino acid sequence in ance with the amino acid sequence as shown in SEQ ID N02, and a CDR3 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:3, and (ii) a VH domain comprising a CDR1 region having an amino acid sequence in ance with the amino acid sequence as shown in SEQ ID NO:4, a CDR2 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:5, and a CDR3 region having an amino acid sequence in accordance with the amino acid ce as shown in SEQ ID NO:6. (2) The antibody or onal fragment of item (1), n said antibody or functional fragment comprises (i) a VL domain comprising a CDR1 region having the amino acid sequence as shown in SEQ ID NO:7, a CDR2 region having the amino acid sequence as shown in SEQ ID NO:8, and a CDR3 region having the amino acid sequence as shown in SEQ ID NO:9, and (ii) a VH domain comprising a CDR1 region having the amino acid sequence as shown in SEQ ID NO:10, a CDR2 region having the amino acid sequence as shown in SEQ ID NO:11, and a CDR3 region having the amino acid sequence as shown in SEQ ID NO:12. (3) The antibody or functional fragment of any one of the preceding items, wherein said antibody or functional fragment comprises a VH domain having the amino acid ce as shown in SEQ ID NO:13. (4) The antibody or functional fragment of any one of the preceding items, wherein said antibody or functional fragment ses a VL domain having an amino acid sequence ed from SEQ ID NO:14 and SEQ ID NO:54, preferably having the amino acid sequence as shown in SEQ ID NO:14. (5) The antibody or functional fragment of any one of the preceding items, wherein said antibody or functional fragment thereof specifically binds to human TNFq. (6) The antibody or functional fragment of any one of any one of the preceding items, wherein said dy or functional fragment thereof does not significantly bind to TNFB. (7) The antibody or functional fragment of any one of the preceding items, wherein said antibody or functional fragment (i) binds to human TNFq with a dissociation constant (KB) of less than 125 pM; (ii) is cross-reactive with Macaca mulatta TNFq and with Macaca fascicu/aris TNFq; (iii) has a greater potency than mab, as determined by an L929 assay; and/or (iv) is capable of g to human TNFGTrimer in a stoichiometry ody : TN er) of at least 2. (8) The antibody or functional fragment of any one of the preceding items, which binds to human TNFo with a KB of less than 100 pM, preferably of less than 50 pM. (9) The antibody or onal fragment of any one of the preceding items, which binds to TNFo from Macaca mulatta with a KB of less than 1 nM. (10) The dy or functional fragment of any one of the preceding items, which binds to TN For from Macaca fascicu/aris with a KB of less than 1 nM. (11) The antibody or functional fragment of any one of the preceding items, wherein the potency of the antibody or functional fragment to inhibit nduced apoptosis relative to that of infliximab (relative potency), determined in an L929 assay, is greater than 5, and wherein said relative potency is the ratio of the |C50 value in ng/mL of infliximab in the L929 assay to the |C50 value in ng/mL of the antibody in scFv format in the L929 assay. (12) The antibody or functional fragment of any one of the preceding items, wherein the melting temperature of the variable domain of the antibody in scFv format, determined by differential scanning fluorimetry, is at least 65°C. (13) The antibody or functional fragment of any one of the preceding items, wherein the melting temperature of the variable domain of the antibody in scFv format, determined by differential ng fluorimetry, is at least 68°C. (14) The antibody or functional fragment of any one of the preceding items, wherein the melting temperature, determined by differential scanning fluorimetry, is at least 70°C. (15) The antibody or functional fragment of any one of the preceding items, wherein the loss in monomer content, after five consecutive freeze-thaw cycles, is less than 0.2%. (16) The antibody or functional fragment of any one of the preceding items, wherein the loss in monomer content, after e for four weeks at 4°C, is less than 1%. (17) The antibody or onal fragment of any one of the ing items, wherein the potency of the antibody or functional fragment to block the interaction between human TNFo and TNF receptor l (TNFRI), ve to that of infliximab (relative potency), as determined in an inhibition ELISA, is at least 2, wherein said relative potency is the ratio of the |C50 value in ng/mL of infliximab to the |C50 value in ng/mL of the dy in scFv format. (18) The antibody or functional fragment of any one of the ing items, wherein the y of the antibody or functional fragment to block the interaction between human TNFo and TNF receptor ll (TNFRII), relative to that of imab (relative y), as determined in an tion ELISA, is at least 2, wherein said relative potency is the ratio of the |C50 value in ng/mL of infliximab to the |C50 value in ng/mL of the antibody in scFv format. (19) The antibody or functional fragment of any one of the preceding items, which is capable of inhibiting cell eration of peripheral blood mononuclear cells in a mixed lymphocyte reaction. (20) The antibody or functional fragment of any one of the preceding items, which is capable of inhibiting LPS—induced secretion of interleukin-1B from CD14+ monocytes. (21) The antibody or functional fragment of item (20), wherein the |C50 value for inhibiting LPS—induced secretion of interleukin-1B is less than 1 nM. (22) The antibody or functional fragment of item (21), wherein said |C50 value for inhibiting LPS—induced secretion of interleukin-1B, on a molar basis, is lower than that of adalimumab. (23) The antibody or functional fragment of any one of the preceding items, which is capable of inhibiting LPS—induced secretion of TNFo from CD14+ monocytes. (24) The antibody or functional fragment of item (23), wherein the |C50 value for inhibiting LPS—induced secretion of TNch is less than 1 nM. (25) The antibody or functional nt of item (24), wherein said |C50 value for inhibiting LPS—induced secretion of TNFo, on a molar basis, is lower than that of adalimumab. (26) The antibody of any one of the preceding items, which is an immunoglobulin G (lgG)- (27) The functional fragment of any one of items (1) to (25), which is a single-chain variable fragment (scFv). (28) The onal fragment of item (27), wherein said scFv comprises or consists of an amino acid sequence selected from SEQ ID NO:15 and SEQ ID NO:55, preferably the amino acid sequence as shown in SEQ ID NO:15. (29) The functional fragment of any one of items (1 ) to (25), which is a diabody. (30) The functional fragment of item (29), wherein said diabody comprises or consists of the amino acid sequence as shown in SEQ ID NO:51. (31) An antibody or functional fragment thereof g to essentially the same epitope on human TNFq as an antibody comprising a VH domain having the amino acid sequence as shown in SEQ ID NO:13 and a VL domain having the amino acid sequence as shown in SEQ ID NO:14, in ular wherein said antibody or functional fragment exhibits one or more of the features referred to in items (1) to (30) herein above. (32) The antibody or functional fragment of any one of the ing items, n the sum of (i) the number of amino acids in framework regions I to III of the variable light domain of said antibody or functional fragment that are different from the respective human VK1 consensus sequences with SEQ ID NOs: 56 to 58 (see Table 15), and (ii) the number of amino acids in framework region IV of the variable light domain of said antibody or functional fragment that are different from the most similar human A germIine-based sequence ed from SEQ ID NOs: 59 to 62 (see Table 16), is less than 7, preferably less than 4. (33) The antibody or functional fragment of any one of the preceding items, wherein the framework s I to III of the variable light domain of said antibody or onal nt consist of human VK1 consensus sequences with SEQ ID NOs:56 to 58, tively, and framework region IV consists of a A germIine-based sequence selected from SEQ ID NOs:59 to 62. (34) A nucleic acid encoding the antibody or functional fragment of any one of the preceding items. (35) A vector or plasmid comprising the nucleic acid of item (34). (36) A cell comprising the nucleic acid of item (35) or the vector or plasmid of item (34). (37) A method of ing the antibody or functional nt of any one of items (1) to (33), comprising culturing the cell of item (36) in a medium under conditions that allow expression of the nucleic acid encoding the antibody or functional nt, and recovering the antibody or functional fragment from the cells or from the medium. (38) A pharmaceutical composition comprising the antibody or onal fragment of any one of items (1) to (33), and optionally a pharmaceutically acceptable carrier and/or excipient. (39) The dy or functional fragment as defined in any one of items (1) to (33) for use in a method of treating an inflammatory disorder or a TNFd-related disorder. (40) The antibody or functional fragment for use according to item (39), wherein said inflammatory disorder is selected from the list of diseases and disorders listed in Section ders to be treated" below. (41) The antibody or functional fragment for use according to item (39), wherein said inflammatory disorder is an inflammatory disorder of the gastrointestinal tract. (42) The antibody or functional fragment for use according to item (41), wherein said inflammatory disorder of the intestinal tract is inflammatory bowel e. (43) The antibody or functional fragment for use according to item (41) or (42), wherein said inflammatory disorder of the gastrointestinal tract is Crohn’s disease. (44) The antibody or functional fragment for use according to item (43), wherein said Crohn’s disease is ed from the group consisting of ileal, colonic, ileocolonic, and/or isolated upper s disease (gastric, duodenal and/or jejunal) and including non-stricturing/non-penetrating, stricturing, penetrating and perianal disease behavior, allowing any combination of zation and disease behavior of any of the above mentioned. (45) The antibody or functional fragment for use according to item (41) or (42), wherein said inflammatory disorder of the gastrointestinal tract is ulcerative colitis. (46) The antibody or functional fragment for use ing to item (45), wherein said ulcerative colitis is selected from the group consisting of ulcerative proctitis, proctosigmoiditis, left-sided colitis, pan-ulcerative colitis, and pouchitis. (47) The antibody or functional fragment for use ing to item (41) or (42), wherein said inflammatory disorder of the gastrointestinal tract is microscopic colitis. (48) The antibody or functional fragment for use according to any one of items (39) to (47), wherein said method comprises orally administering the antibody or functional fragment to a subject.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1: Schematic representation of the humanization process.

Figure 2: SE-HPLC chromatograms of purified humanized scFv preparations of an scFv.

The scFv monomer elutes at retention times between 8.5 and 9.5 minutes, while buffer components elute at >10 min. All peaks from the dead volume of the column up to the respective scFv monomer were integrated as aggregates/oligomers and used for the calculation of the relative peak area.

Figure 3: Thermal unfolding curves from DSF measurements of two scFv constructs. For each construct duplicate measurements are shown. The resulting Tm values have been determined by fitting the data to a Boltzmann equation to obtain the midpoint of transition.

Figure 4: ourse of the monomer content of the two scFv constructs during storage.

The r content as determined by C has been plotted for the storage temperatures 4, -20 and <-65°C for the duration of 4 weeks.

Figure 5: Overlay of SE-HPLC chromatograms for two scFv molecules. For each scFv the sample (10 mg/ml) at d0 and after storage for 4 weeks at 4°C is shown. In addition, the togram of the sample after 5 cycles of freezing and g is shown. The ed panel shows an . 15-fold zoom of the y-axis for each molecule to visualize also minuscule changes in er content.

Figure 6: Time-course of the monomer t of the humanized scFvs during storage. The monomer content as determined by SE-HPLC has been plotted for the 10 mg/mL samples at a storage temperature of 37°C for the duration of 4 weeks.

Figure 7: Potency to neutralize human TNch in the L929 assay of two scFvs. Dose- response curves for the scFvs and the reference antibody infliximab are shown for each experiment. The highest scFv and imab concentrations as well as negative controls were set to 100% and 0% of growth.

Figure 8: Potency of two scFvs to neutralize non-human primate and human TNFo in the L929 assay. Dose-response curves for neutralization of human, Cynomolgus monkey and Rhesus monkey TNFo are shown. The highest scFv concentration and negative controls were set to 100% and 0% of growth.

Figure 9: Potency of two scFvs to block the TNch-TNFRI interaction. Dose-response curves are shown. The highest scFv concentration and negative controls were set to 0% and 100% of binding of TNFo to TNFRI.

Figure 10: Potency of two scFvs to block the TNFo-TNFRII interaction. Dose-response curves are shown. The highest scFv concentration and negative controls were set to 0% and 100% of binding of TNFo to TNFRII.

Figure 11: Target specificity of an scFv. The potential to inhibit the interaction of biotinylated TNFoc with the scFv by TNFoc and TNFB was ed by ition ELISA. Dose- dependent s of TNFoc and TNFB are shown.

Figure 12 depicts the formation of 16H5-scFv:TNFo complexes determined by SE- HPLC (Example 4).

Figure 13 depicts the simultaneous g of two TNch molecules to 16H5-scDb determined by SPR (Example 5).

Figure 14A depicts the formation of 16H5-lgG:TNFd complexes (Example 5).

Figure 148 depicts the formation of 16H5-scDb:TNFd complexes le 5).

Figure 15 depicts the tion of cell proliferation in a MLR ing anti-TNFo treatment. *: p < 0.05; **: p < 0.01 compared to lgG l (Example 6).

Figure 16 shows the ability of the different antibody formats of 16H5 and adalimumab to inhibit the LPS—induced secretion of lL-1B (Figure 16A) and TNFo e 168) in monocytes in a dose-dependent manner (Example 7).

DETAILED DESCRIPTION The present invention pertains to an dy or a functional fragment thereof capable of binding to human TNFd.

In the context of the present ation, the term "antibody" is used as a synonym for "immunoglobulin" (lg), which is defined as a protein belonging to the class lgG, lgM, lgE, lgA, or lgD (or any subclass thereof), and es all conventionally known dies and functional fragments thereof. In the context of the present invention, a "functional fragment" of an antibody/immunoglobulin is defined as antigen-binding fragment or other derivative of a parental antibody that essentially maintains one or more of the properties of such parental antibody referred to in items (1) to (30) herein above. An "antigen-binding fragment" of an antibody/immunoglobulin is defined as fragment (e.g., a variable region of an lgG) that retains the n-binding region. An "antigen-binding region" of an antibody typically is found in one or more hypervariable region(s) of an dy, i.e., the CDR-1, -2, and/or -3 regions. "Antigen-binding fragments" of the invention include the domain of a F(ab')2 fragment and a Fab fragment. "Functional fragments" of the invention include, scFv, dst, diabodies, triabodies, tetrabodies and Fc fusion proteins. The F(ab')2 or Fab may be engineered to minimize or completely remove the intermolecular disulphide interactions that occur between the CH1 and CL domains. The antibodies or functional fragments of the present invention may be part of bi- or multifunctional constructs.

Preferred functional fragments in the present invention are scFv and diabodies.

An scFv is a single chain Fv fragment in which the le light ("VL") and le heavy ("VH") domains are linked by a e bridge.

A diabody is a dimer consisting of two fragments, each having variable regions joined together via a linker or the like (hereinafter referred to as diabody-forming fragments), and typically contain two VLs and two VHs. Diabody-forming fragments include those consisting of VL and VH, VL and VL, VH and VH, etc., preferably VH and VL. ln diabody-forming fragments, the linker joining le s is not specifically limited, but preferably enough short to avoid noncovalent bonds n variable regions in the same fragment.

The length of such a linker can be determined as appropriate by those skilled in the art, but typically 2—14 amino acids, preferably 3-9 amino acids, especially 4-6 amino acids. In this case, the VL and VH encoded on the same fragment are joined via a linker short enough to avoid noncovalent bonds between the VL and VH on the same chain and to avoid the formation of single-chain le region fragments so that dimers with another fragment can be formed. The dimers can be formed via either covalent or noncovalent bonds or both between y-forming fragments.

Moreover, y-forming fragments can be joined via a linker or the like to form - chain diabodies (sc(Fv)2). By joining diabody-forming fragments using a long linker of about 15-20 amino acids, noncovalent bonds can be formed between diabody-forming nts ng on the same chain to form dimers. Based on the same principle as for preparing diabodies, polymerized antibodies such as s or tetramers can also be prepared by joining three or more diabody-forming fragments.

Preferably, the antibody or functional fragment of the invention specifically binds to TNFo.

As used herein, an antibody or functional fragment thereof "specifically recognizes", or "specifically binds to" human TNFo, when the antibody or onal fragment is able to discriminate n human TNFo and one or more reference molecule(s). Preferably, the IC50 value for binding to each of the reference molecules is at least 1,000 times greater than the IC50 value for binding to TNFo, ularly as described in Example 2, section 2.1.4. In its most general form (and when no defined reference is mentioned), "specific binding" is referring to the ability of the antibody or functional fragment to minate between human TNch and an unrelated biomolecule, as determined, for example, in accordance with a specificity assay methods known in the art. Such methods comprise, but are not limited to, Western blots and ELISA tests. For example, a standard ELISA assay can be carried out.

Typically, determination of binding specificity is performed by using not a single reference biomolecule, but a set of about three to five unrelated biomolecules, such as milk , BSA, transferrin or the like. In one embodiment, specific binding refers to the ability of the antibody or fragment to discriminate between human TNFo and human TN F8.

The antibody of the ion or the functional fragment of the invention comprises a VL domain and a VH domain. The VL domain comprises a CDR1 region (CDRL1), a CDR2 region ), a CDR3 region (CDRL3) and Framework regions. The VH domain comprises a CDR1 region (CDRH1), a CDR2 region (CDRH2), a CDR3 region (CDRH3) and Framework regions.

The term "CDR" refers to one of the six hypervariable regions within the variable domains of an antibody that mainly contribute to antigen binding. One of the most commonly used definitions for the six CDRs was provided by Kabat E. A. et aI., (1991) ces of proteins of immunological interest. NIH Publication 91-3242). As used herein, Kabat’s tion of CDRs only apply for CDR1, CDR2 and CDR3 of the light chain variable domain (CDR L1, CDR L2, CDR L3, or L1, L2, L3), as well as for CDR2 and CDR3 of the heavy chain variable domain (CDR H2, CDR H3, or H2, H3). CDR1 of the heavy chain variable domain (CDR H1 or H1), however, as used herein is d by the following residues (Kabat numbering): It starts with position 26 and ends prior to position 36.

The CDR1 region of the VL domain consists of an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:1. Preferably, the CDR1 region of the VL domain consists of an amino acid sequence selected from the group consisting of SEQ ID NO:7, SEQ ID NO:16, SEQ ID NO:17, SEQ ID NO:18, SEQ ID NO:19, SEQ ID NO:20, SEQ ID N021 and SEQ ID NO:22. Most preferably, the CDR1 region of the VL domain consists of the amino acid sequence as shown in SEQ ID NO:7.

The CDR2 region of the VL domain consists of an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:2. Preferably, the CDR2 region of the VL domain consists of an amino acid ce selected from the group ting of SEQ ID NO:8, SEQ ID NO:23, SEQ ID NO:24, SEQ ID N025, and SEQ ID NO:26. Most ably, the CDR2 region of the VL domain ts of the amino acid sequence as shown in SEQ ID NO:8.

The CDR3 region of the VL domain consists of an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:3. Preferably, the CDR3 region of the VL domain consists of an amino acid sequence selected from the group consisting of SEQ ID NO:9, SEQ ID NO:27, SEQ ID NO:28, SEQ ID NO:29, SEQ ID NO:30, SEQ ID NO:31 and SEQ ID NO:32. Most preferably, the CDR3 region of the VL domain consists of the amino acid sequence as shown in SEQ ID NO:9.

The CDR1 region of the VH domain consists of an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:4. Preferably, the CDR1 region of the VH domain consists of an amino acid sequence selected from the group consisting of SEQ ID NO:10, SEQ ID NO:33, SEQ ID NO:34, SEQ ID NO:35 and SEQ ID NO:36. Most preferably, the CDR1 region of the VH domain consists of the amino acid sequence as shown in SEQ ID NO:10.

The CDR2 region of the VH domain consists of an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:5. Preferably, the CDR2 region of the VH domain consists of an amino acid sequence ed from the group consisting of SEQ ID NO:11, SEQ ID NO:37, SEQ ID NO:38, SEQ ID NO:39, SEQ ID NO:40 and SEQ ID NO:41. Most preferably, the CDR2 region of the VH domain consists of the amino acid sequence as shown in SEQ ID NO:11.

The CDR3 region of the VH domain consists of an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:6. Preferably, the CDR3 region of the VH domain consists of an amino acid sequence selected from the group consisting of SEQ ID NO:12, SEQ ID NO:42, SEQ ID NO:43, SEQ ID NO:44, SEQ ID NO:45, SEQ ID NO:46, SEQ ID NO:47 and SEQ ID NO:48. Most preferably, the CDR3 region of the VH domain consists of the amino acid ce as shown in SEQ ID NO:12.

In a particular embodiment, the antibody of the invention or the functional fragment of the invention comprises (i) a VL domain comprising a CDR1 region having an amino acid sequence in ance with the amino acid sequence as shown in SEQ ID NO:1 , a CDR2 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID N02, and a CDR3 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:3, and (ii) a VH domain comprising a CDR1 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:4, a CDR2 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:5, and a CDR3 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID In a particular embodiment, the antibody of the invention or the functional fragment of the invention comprises (i) a VL domain sing a CDR1 region having the amino acid sequence as shown in SEQ ID NO:7, a CDR2 region having the amino acid sequence as shown in SEQ ID NO:8, and a CDR3 region having the amino acid sequence as shown in SEQ ID NO:9, and (ii) a VH domain comprising a CDR1 region having the amino acid ce as shown in SEQ ID NO:10, a CDR2 region having the amino acid sequence as shown in SEQ ID NO:11, and a CDR3 region having the amino acid sequence as shown in SEQ ID NO:12.

In a more preferred ment, the antibody of the invention or the functional fragment of the ion comprises a VH domain having the amino acid sequence as shown in SEQ ID NO:13. In another more red embodiment the antibody or functional fragment comprises a VL domain having the amino acid sequence as shown in SEQ ID NO:14 or SEQ ID NO:54. Most preferably, the antibody of the invention or the functional fragment of the invention comprises (i) a VH domain having the amino acid sequence as shown in SEQ ID NO:13, and (ii) a VL domain having the amino acid sequence as shown in SEQ ID NO:14.

In a particularly preferred embodiment, the functional fragment is a single chain antibody (scFv) comprising a VH domain having the amino acid sequence as shown in SEQ ID NO:13 and a VL domain having the amino acid sequence as shown in SEQ ID NO:14 or SEQ ID NO:54. The VH domain and the VL domain are preferably linked by a peptide linker. The peptide linker (hereinafter referred to as "linkerA") lly has a length of about 10 to about amino acids, more preferably of about 15 to about 25 amino acids. The linkerA typically comprises Gly and Ser residues, but other amino acids are also possible. In preferred embodiments the linker comprises multiple repeats of the sequence GGGGS (SEQ ID NO:50), e.g. 2 to 6, or 3 to 5, or 4 consecutive repeats of the amino acid ce as shown in SEQ ID NO:50. Most preferably, the linkerA consists of the amino acid sequence as shown in SEQ ID NO:49. The scFv may have the following ure (with the N-terminus being left and the C-terminus being right): kerA-VH; or VH-LinkerA-VL.

More preferably, the functional fragment is a single chain antibody (scFv) consisting of the amino acid sequence as shown in SEQ ID NO:15 or SEQ ID NO:55. Most preferably, the functional fragment is a single chain antibody (scFv) consisting of the amino acid sequence as shown in SEQ ID NO:15.

In another particularly preferred embodiment, the functional nt is a diabody comprising a VH domain having the amino acid sequence as shown in SEQ ID NO:13 and a VL domain having the amino acid sequence as shown in SEQ ID NO:14 or SEQ ID NO:54.

The VH domain and the VL domain are linked by a peptide linker. The peptide linker (hereinafter referred to as "linkerB") preferably has a length of about 2 to about 10 amino acids, more preferably of about 5 amino acids. The linkerB typically comprises Gly and Ser residues, but other amino acids are also le. Most preferably, the B ts of the amino acid sequence as shown in SEQ ID NO:50.

The diabody preferably is a monospecific diabody, i.e. it is directed to one epitope only. The diabody is preferably a homodimer. The diabody may be a dimer of two polypeptide chains that are non-covalently bound to each other. Each monomer may be a polypeptide chain having the structure: kerB-VH; or VH-LinkerB-VL.

Moreover, diabody-forming fragments can be joined via a linkerA or the like to form single- chain ies (sc(Fv)2). By joining diabody-forming fragments using a long linker of about -20 amino acids, noncovalent bonds can be formed between diabody-forming nts existing on the same chain to form dimers. Examples of the arrangements of -chain diabodies include the following.

VH — linkerB — VL — A - VH linkerB - VL VL - linkerB - VH - linkerA - VL - linkerB - VH Preferably the diabody of the invention has the following structure: VL - B - VH - linkerA - VL - linkerB - VH Most preferably the diabody consists of the amino acid sequence as shown in SEQ ID NO:51.

Based on the same principle as for preparing diabodies, polymerized antibodies such as trimers or tetramers can also be prepared by joining three or more diabody-forming fragments.

In another particular embodiment the dy of the invention is an immunoglobulin, preferably an immunoglobulin G (lgG). The subclass of the lgG of the invention is not limited and includes lgG1, lgG2, lgG3, and lgG4. Preferably, the lgG of the invention is of subclass 1, Le. it is an lgG1 molecule. In one embodiment, each light chain of the IgG molecule of the invention has the amino acid sequence as shown in SEQ ID NO:52, and/or each heavy chain of the IgG molecule of the invention has the amino acid ce as shown in SEQ ID NO:53. A specific lgG of the invention ts of two light chains and two heavy chains, wherein each of the two light chains has the amino acid ce as shown in SEQ ID NO:52, and each of the two heavy chains has the amino acid sequence as shown in SEQ ID NO:53.

Affinity The antibody or functional fragment of the invention has a high affinity to human TNFq. The term "KD," refers to the dissociation equilibrium constant of a particular antibody-antigen interaction. Typically, the antibody or functional fragment of the invention binds to human TNFq with a dissociation equilibrium constant (KB) of less than approximately 2x10'10 M, preferably less than 1.5x10'10 M, preferably less than 0'10 M, more preferably less than 1x10'10 M, most preferably less than 7.5x10'11 M or even less than 5x10'11 M, as determined using surface n resonance (SPR) technology in a BIACORE instrument.

In particular, the determination of the KB is carried out as described in Example 2, section 2.1.1. reactivity to TNFa from Cynomolgus s or from Rhesus macaques In particular embodiments, the antibody or functional fragment of the invention has substantial affinity to TNFq from animals such as Cynomolgus monkeys (Macaca fascicularis) and/or Rhesus macaques (Macaca mulatta). This is advantageous, as preclinical tests of anti-human TNFq antibodies such as ty s are preferably performed with such animals. Accordingly, the antibody or functional fragment of the invention is preferably cross-reactive with TNFq from animals such as Cynomolgus monkeys and/or Rhesus macaques. Affinity measurements are d out as described in Example 2, section 2.1 .1.

In one embodiment, the antibody or functional fragment of the invention is cross-reactive with TNFq from Macaca fascicularis. The dy or functional fragment of the invention preferably has an affinity to Macaca fascicularis TNFq that is less than 20-fold, particularly less than 10-fold, even more particularly less than 5-fold different to that of human TNFq.

Typically, the antibody or onal fragment of the invention binds to TNFq from Macaca fascicularis with a dissociation equilibrium constant (KD), wherein the ratio RMfascicularis of (i) the KB for g to TNFq from Macaca fascicularis to (ii) the KB for binding to human TNFq is less than 20.

KD (M. fascicularis) RMfascicularis. =W RM_fascicu.aris is preferably less than 20, particularly less than 10, even more particularly less than 5.

In another embodiment, the antibody or functional fragment of the invention is cross- reactive with TNFo from Macaca mulatta. The antibody or functional fragment of the ion preferably has an affinity to Macaca mulatta TNFo that is less than 20-fold, more particularly less than 10-fold ent to that of human TNFd. Typically, the antibody or functional fragment of the invention binds to TNFo from Macaca mulatta with a dissociation equilibrium constant (KD), wherein the ratio RMmulafia of (i) the KB for binding to TNFo from Macaca mulatta to (ii) the KB for binding to human TN Fox is less than 20.

KD (M. mulatta) RMmulatta =W RMmulafia is preferably less than 20, particularly less than 10.

In yet another embodiment, the antibody or functional fragment of the invention is cross- reactive with TNFo from Macaca fascicu/aris and with TNFo from Macaca a. The antibody or functional fragment of the invention preferably has an affinity to Macaca fascicu/aris TNFo that is less than 20-fold, ularly less than d, even more particularly less than 5-fold different to that of human TNFd, and it preferably has an affinity to Macaca mulatta TNch that is less than 20-fold, more particularly less than 10-fold different to that of human TNFo. The ratio RMfascicularis of the antibody or functional fragment is preferably less than 20, ularly less than 10, even more ularly less than 5, and the ratio RMmulafia of the antibody or functional fragment is preferably less than 20, particularly less than 10.

Potency to inhibit TNFa-induced apoptosis of L929 cells The antibody or onal fragment of the invention has a high potency to inhibit TNFd- induced sis of L929 cells. In a particular embodiment, the antibody or functional fragment of the invention has a potency to t TNFd-induced sis of L929 cells greater than that of the known antibody infliximab.

Potency ve to infliximab can be determined in an L929 assay as described in e 2, section 2.1.2 of this application. The relative potency of the antibody or functional fragment of the invention is greater than 1.5, preferably greater than 2, more preferably greater than 3, more preferably greater than 5, more preferably greater than 7.5, or even greater than 10, wherein the relative potency is the ratio of (i) the |C50 value of infliximab in an L929 assay over (ii) the leo value of the antibody or functional fragment of the invention in the L929 assay, and wherein the |C50 indicates the concentration in ng/mL of the respective molecule necessary to achieve 50% of maximal inhibition of TNFor-induced apoptosis of L929 cells.

In another embodiment, the relative potency of the antibody or functional fragment of the invention is r than 1.5, preferably greater than 2, more ably greater than 3, more preferably greater than 5, more preferably greater than 7.5, or even greater than 10, wherein the relative potency is the ratio of (i) the |C90 value of infliximab in an L929 assay over (ii) the |C90 value of the antibody or functional fragment of the invention in the L929 assay, and wherein the ngo value indicates the concentration in ng/mL of the respective molecule necessary to achieve 90% of l inhibition of TNFd-induced apoptosis of L929 cells .

Inhibition of LPS—induced cytokine secretion Typically, the antibody or functional fragment of the invention is capable of inhibiting LPS- induced ne ion from monocytes. LPS—induced cytokine secretion from monocytes can be ined as described in Example 7.

In one embodiment, the antibody or functional fragment of the invention is capable of ting LPS—induced ion of interleukin-1B from CD14+ monocytes. The |C50 value for inhibiting LPS—induced secretion of eukin-1B is preferably less than 1 nM and/or less than 100 pg/mL. The |C50 value for inhibiting duced secretion of eukin-1B, on a molar basis and/or on a weight-per-volume basis, is preferably lower than that of adalimumab.

In another embodiment, the antibody or functional fragment of the invention is capable of inhibiting LPS—induced secretion of TNFo from CD14+ tes. The |C50 value for inhibiting LPS—induced secretion of TNch is preferably less than 1 nM and/or less than 150 pg/mL. The IC50 value for inhibiting LPS—induced secretion of TNch, on a molar basis and/or on a weight-per-volume basis, is ably lower than that of adalimumab.

Inhibition of cell proliferation The antibody or functional fragment of the invention is typically capable of inhibiting cell proliferation of peripheral blood mononuclear cells in a mixed lymphocyte reaction. The inhibition of cell proliferation can be determined as described in Example 6. The stimulation index of the dy or functional fragment, e.g. of the scFv or diabody of the invention, determined according to e 6, is preferably less than 5, more preferably less than 4.5.

In particular embodiments, the stimulation index of the antibody, e.g. of the lgG of the invention, is less than 4 or even less than 3.

Inhibition of interaction between TNFa and TNF receptor Typically, the antibody or functional fragment of the invention is capable of inhibiting the interaction between human TNFo and TNF receptor l (TNFRI). The inhibition of the interaction between human TNFo and TNFRI can be determined in an tion ELISA as bed below in e 2, section 2.1.3.

The potency of the antibody or functional fragment of the invention to inhibit the interaction between human TNFo and TNFRI, relative to that of infliximab (relative potency), as determined in an inhibition ELISA, is preferably at least 2, wherein said relative potency is the ratio of the |C50 value in ng/mL of infliximab to the |C50 value in ng/mL of the antibody or functional fragment thereof.

Typically, the antibody or functional fragment of the invention is capable of inhibiting the interaction between human TNFo and TNF receptor ll (TNFRII). The inhibition of the interaction between human TNFo and TNFRII can be determined in an inhibition ELISA as described below in Example 2, section 2.1.3.

The y of the antibody or functional fragment of the invention to inhibit the ction between human TNFo and TNFRII, relative to that of infliximab (relative potency), as determined in an inhibition ELISA, is preferably at least 2, more preferably at least 3, wherein said relative potency is the ratio of the |C50 value in ng/mL of infliximab to the |C50 value in ng/mL of the antibody or functional fragment thereof.

Stoichiometry and inking The antibody or functional fragment of the invention is typically capable of binding to human TNdemer in a iometry ody:TNForTrimer) of at least 2. The stoichiometry (antibody: TNFaTrimer) is preferably greater than 2, or at least 2.5, or at least 3. In one embodiment, the stoichiometry (antibody:TNFdTrimer) is about 3. The stoichiometry (antibody : TNFGTrimer) can be determined as described in Example 4 below.

In another embodiment, the antibody or functional fragment of the invention is e of forming a complex with human TNch, wherein said complex comprises at least two molecules of TNch and at least three molecules of antibody or functional fragment. The functional fragment in accordance with this ment comprises at least two separate binding sites for TNch such as, eg. diabodies. Complex formation can be determined as described in Example 5 below.

In one embodiment, the antibody is an lgG, and is capable of forming a complex of at least 600 kDa with TNch. In another embodiment, the functional fragment is a diabody, and is capable of forming a x of at least 300 kDa with TNch.

Target selectivity In certain embodiments, the antibody or the functional fragment of the invention has a high target selectivity, i.e. it can discriminate between TNFo and TNFB. ably, the IC50 value of TNFB is at least 1,000 times greater than the IC50 value of TNFd, as determined in a competition ELISA as described in e 2, section 2.1.4. More preferably, the IC50 value of TNFB is at least 5,000 times, most preferably at least 10,000 greater than the IC50 value of TNFd, as determined in a competition ELISA as described in Example 2, section 2.1.4.

Expression yield and ing yield In other embodiments, the antibody or functional fragment of the invention, preferably the scFv or diabody, can be recombinantly expressed in high yield in microorganisms such as bacteria or in other cells. Preferably, the expression yield in E. coli, determined as described in Example 2, is at least 0.25 g/L. This particularly applies to functional fragments such as scFvs.

The refolding yield, ined as described in e 2, is at least 5 mg/L, more preferably at least 10 mg/L, more preferably at least 15 mg/L, and most preferably at least 20 mg/L. This ularly applies to functional fragments such as scFvs.

Stability Typically, the antibody or functional fragment of the invention, preferably the scFv or diabody, has a high ity. Stability can be assessed by different methodologies. The ng temperature" Tm of the variable domain of the antibody or onal fragment of the invention, determined by differential scanning fluorimetry (DSF) as described in Example 2, section 2.2.4, is preferably at least 65°C, more preferably at least 68°C, most ably at least 70°C. The "melting temperature of the variable domain", as used herein, refers to the melting ature of an scFv consisting of the sequence VL — LinkerA — VH, n the amino acid sequence of LinkerA consists of the amino acid sequence as shown in SEQ ID NO:49. For example, the melting temperature of the variable domain of an lgG is defined as the melting temperature of its corresponding scFv as defined above.

The loss in monomer content (at a concentration of 10 g/L; initial monomer content > 95%) after storage for four weeks at 4°C, determined by analytical size-exclusion chromatography as described in Example 2, section 2.2.5, is preferably less than 5%, more preferably less than 3%, more preferably less than 1%, most preferably less than 0.5%. The loss in monomer content (at a concentration of 10 g/L; initial monomer content > 95%) after storage for four weeks at -20°C, determined by analytical size-exclusion chromatography as described in Example 2, section 2.2.5, is preferably less than 5%, more preferably less than 3%, more preferably less than 1%, most preferably less than 0.5%. The loss in monomer content (at a concentration of 10 g/L; initial monomer content > 95%) after storage for four weeks at -65°C, determined by analytical size-exclusion chromatography as described in e 2, n 2.2.5, is preferably less than 5%, more preferably less than 3%, more preferably less than 1%, most preferably less than 0.5%.

The monomer loss after five consecutive freeze-thaw cycles, determined as described in Example 2, is less than 5%, more preferably less than 1%, more preferably less than 0.5%, most preferably less than 0.2%, e.g. 0.1% or 0.0%.

Antibodies and functional fragments Particular ments of the invention relate to functional fragments of the antibodies described herein. Functional fragments e, but are not limited to, 2 fragment, a Fab fragment, scFv, diabodies, dies and tetrabodies. Preferably, the functional fragment is a single chain antibody (scFv) or a diabody. More preferably, the non-CDR sequences of the scFv or of the diabody are human ces.

Preferably in order to minimize ial for immunogenicity in humans the chosen acceptor ld is composed of framework regions derived from human consensus sequences or human germline sequences. In particular framework s I to III of the variable light domain consist of human VK1 consensus sequences according to SEQ ID NOs: 56 to 58 and a framework region IV of a A germline-based sequence selected from SEQ ID NOs:59 to 62. As residues that are not human consensus or human germline residues may cause immune reactions the number of such es in each variable domain (VH or VL) should be as low as possible, ably lower than 7, more preferably lower than 4, most preferably 0.

Preferably the antibody is a monoclonal antibody. The term "monoclonal antibody" as used herein is not limited to dies produced through hybridoma technology. The term lonal antibody" refers to an antibody that is derived from a single clone, including any eukaryotic, prokaryotic, or phage clone, and not the method by which it is produced. onal antibodies can be prepared using a wide variety of techniques known in the art including the use of hybridoma, recombinant, and phage display technologies, or a combination thereof. (Harlow and Lane, odies, A Laboratory Manual" CSH Press 1988, Cold Spring Harbor N.Y.).

In other embodiments, including embodiments relating to the in vivo use of the anti-TNFq antibodies in humans, chimeric, ized, humanized, or human antibodies can be used.

In a preferred embodiment, the antibody is a human antibody or a humanized antibody, more preferably a monoclonal human antibody or a onal humanized antibody.

The term "chimeric" antibody as used herein refers to an antibody having variable sequences derived from a non-human immunoglobulin, such as a rat or mouse dy, and human immunoglobulins constant regions, typically chosen from a human immunoglobulin template. Methods for producing chimeric antibodies are known in the art.

See, e.g., Morrison, 1985, Science 229(4719): 1202-7; Oi et al, 1986, BioTechniques 4:214- 221; Gillies et al., 1985, J. lmmunol. Methods 125: 191-202; US. Pat. Nos. 5,807,715; 4,816,567; and 4,816397, which are incorporated herein by reference in their entireties.

Different inant methodologies are available to one of ordinary skill in the art to render a non-human (e.g., murine) antibody more human-like by generating immunoglobulins, immunoglobulin chains or fragments thereof (such as Fv, Fab, Fab', F(ab')2 or other - g uences of antibodies), which contain minimal sequences derived from such non-human immunoglobulin. In general, the resulting recombinant antibody will comprise substantially all of at least one, and typically two, le domains, in which all or substantially all of the CDR regions correspond to those of a non-human immunoglobulin and all or substantially all of the FR regions are those of a human immunoglobulin sequence, particularly a human immunoglobulin consensus sequence. CDR-grafted antibodies are antibody molecules having one or more complementarity determining regions (CDRs) from an antibody originally generated in a non-human species that bind the desired antigen and framework (FR) s from a human immunoglobulin molecule 400; PCT publication WO 91/09967; U.S. Patent Nos. 5,225,539; 5,530,101 and 5,585,089). Often, in a process called "humanization", framework residues in the human framework regions will additionally be substituted with the ponding residue from the CDR donor antibody to alter, preferably improve, antigen binding. These framework substitutions are identified by methods well known in the art, e.g., by ng of the interactions of the CDR and framework residues to identify ork residues important for n binding and sequence comparison to identify unusual framework es at particular ons. See, e.g., Riechmann et al., 1988, Nature 332:323-7 and Queen et al, U.S. Patent Nos: 5,530,101; 5,585,089; 5,693,761; 5,693,762; and 6,180,370 (each of which is incorporated by reference in its entirety). Antibodies can be rendered more human using a variety of onal techniques known in the art including, for example, veneering or acing (EP592106; EP519596; Padlan, 1991, Mol. lmmunol, 28:489-498; Studnicka et al, 1994, Prot. Eng. 7:805-814; Roguska et al, 1994, Proc. Natl.

Acad. Sci. 91:969-973, and chain shuffling (U.S. Patent No. 5,565,332), all of which are hereby incorporated by reference in their entireties. A CDR-grafted or humanized antibody can also comprise at least a portion of an immunoglobulin constant region (Fc), typically that of a human immunoglobulin te chosen.

In some embodiments, humanized antibodies are prepared as described in Queen et al, U.S. Patent Nos: 5,530,101; 5,585,089; 5,693,761; 5,693,762; and 6,180,370 (each of which is incorporated by reference in its entirety).

In some embodiments, the anti-TNFq antibodies are human antibodies. Completely "human" NFo antibodies can be desirable for eutic treatment of human patients.

As used herein, "human antibodies" include antibodies having the amino acid ce of a human immunoglobulin and include antibodies isolated from human globulin libraries or from animals transgenic for one or more human immunoglobulin and that do not express nous immunoglobulins. Human antibodies can be made by a variety of methods known in the art including phage display methods described above using antibody libraries d from human immunoglobulin sequences. See U.S. Patent Nos. 4,444,887 and 4,716,111; and PCT publications WO 98/46645; WO 98/50433; WO 98/24893; WO 98/16654; WO 96/34096; WO 96/33735; and WO 91/10741, each of which is incorporated herein by reference in its entirety. Human antibodies can also be produced using transgenic mice which are ble of expressing functional endogenous immunoglobulins, but which can s human immunoglobulin genes. See, e.g., PCT publications WO 98/24893; WO 92/01047; WO 96/34096; WO 96/33735; U.S. Patent Nos. 5,413,923; 5,625,126; 5,633,425; ,569,825; 5,661,016; 806; 5,814,318; 5,885,793; 771; and 598, which are incorporated by reference herein in their ties. Completely human antibodies that recognize a selected epitope can be generated using a technique referred to as "guided selection.' In this ch a selected non-human monoclonal antibody, e.g., a mouse dy, is used to guide the selection of a completely human antibody recognizing the same epitope (Jespers et al, 1988, Biotechnology 12:899-903).

In some embodiments, the anti-TNFo antibodies are primatized antibodies. The term "primatized antibody" refers to an antibody comprising monkey variable regions and human constant regions. Methods for producing primatized antibodies are known in the art. See e.g., U.S. Patent Nos. 5,658,570; 5,681,722; and 5,693,780, which are orated herein by reference in their entireties.

In some embodiments, the anti-TNFo antibodies are derivatized antibodies. For example, but not by way of limitation, the derivatized antibodies that have been modified, e.g., by glycosylation, acetylation, pegylation, phosphorylation, amidation, tization by known protecting/blocking groups, proteolytic cleavage, linkage to a cellular ligand or other protein (see below for a discussion of antibody conjugates), etc. Any of numerous chemical cations may be carried out by known techniques, including, but not limited to, specific al cleavage, acetylation, formylation, metabolic synthesis of tunicamycin, etc.

Additionally, the derivative may contain one or more non-classical amino acids.

In yet other aspects, an anti-TNFo antibody has one or more amino acids inserted into one or more of its hypervariable region, for example as bed in US 2007/0280931.

Antibody Conjugates In some embodiments, the anti-TNFd antibodies are antibody conjugates that are modified, e.g., by the covalent attachment of any type of molecule to the antibody, such that covalent attachment does not interfere with binding to TNFo.Techniques for conjugating effector moieties to antibodies are well known in the art (See, e.g., Hellstrom et al., Controlled Drag Delivery, 2nd Ed., at pp. 623-53 (Robinson et al., eds., 1987)); Thorpe et al., 1982, lmmunol. Rev. 62: 119-58 and Dubowchik et al., 1999, Pharmacology and eutics 83:67-123).

WO 58092 In one example, the antibody or nt thereof is fused via a covalent bond (e.g., a e bond), at optionally the N-terminus or the C-terminus, to an amino acid sequence of another protein (or portion thereof; preferably at least a 10, 20 or 50 amino acid portion of the protein). Preferably the antibody, or nt thereof, is linked to the other protein at the N- terminus of the constant domain of the antibody. Recombinant DNA procedures can be used to create such fusions, for example as described in WO 86/01533 and EP0392745. In another example the effector molecule can increase half-life in vivo. Examples of suitable or molecules of this type include rs, albumin, albumin binding ns or albumin binding compounds such as those described in In some embodiments, anti-TNFd antibodies can be attached to poly(ethyleneglycol) (PEG) moieties. For example, if the antibody is an antibody nt, the PEG moieties can be ed through any ble amino acid side-chain or terminal amino acid functional group located in the antibody fragment, for example any free amino, imino, thiol, hydroxyl or carboxyl group. Such amino acids can occur naturally in the antibody fragment or can be engineered into the fragment using recombinant DNA methods. See for example US.

Patent No. 5,219,996. Multiple sites can be used to attach two or more PEG molecules.

Preferably PEG moieties are covalently linked through a thiol group of at least one cysteine residue located in the antibody fragment. Where a thiol group is used as the point of attachment, appropriately activated effector moieties, for example thiol selective derivatives such as maleimides and cysteine derivatives, can be used.

In another example, an anti-TNFo antibody conjugate is a modified Fab' fragment which is PEGylated, i.e., has PEG (poly(ethyleneglycol)) ntly attached thereto, e.g., according to the method disclosed in EP0948544. See also thyleneglycol) Chemistry, Biotechnical and Biomedical Applications, (J. Milton Harris (ed.), Plenum Press, New York, 1992); thyleneglycol) Chemistry and Biological Applications, (J. Milton Harris and S.

Zalipsky, eds., American Chemical Society, gton D. C, 1997); and Bioconjugation Protein Coupling Techniques for the Biomedical Sciences, (M. Aslam and A. Dent, eds., Grove Publishers, New York, 1998); and Chapman, 2002, Advanced Drug Delivery Reviews 54:531- 545.

Pharmaceutical Compositions and ent ent of a disease encompasses the ent of patients already diagnosed as having any form of the disease at any clinical stage or manifestation; the delay of the onset or evolution or aggravation or deterioration of the symptoms or signs of the disease; and/or preventing and/or reducing the severity of the disease.

A "subject" or "patient" to whom an anti-TNch antibody or functional fragment thereof is administered can be a mammal, such as a non-primate (e.g., cow, pig, horse, cat, dog, rat, etc.) or a primate (e.g., monkey or human). In certain aspects, the human is a pediatric t. In other aspects, the human is an adult patient.

Compositions comprising an anti-TNch antibody and, ally one or more additional therapeutic agents, such as the second therapeutic agents described below, are bed herein. The compositions typically are supplied as part of a sterile, pharmaceutical composition that includes a pharmaceutically acceptable carrier. This ition can be in any suitable form ding upon the desired method of administering it to a patient).

The anti-TNch antibodies and functional fragments can be administered to a patient by a variety of routes such as orally, transdermally, subcutaneously, intranasally, intravenously, intramuscularly, intrathecally, topically or locally. The most suitable route for stration in any given case will depend on the particular antibody, the subject, and the nature and severity of the disease and the physical condition of the subject. Typically, an anti-TNch antibody or functional fragment f will be stered intravenously.

In a particularly preferred embodiment, the antibody or functional fragment of the invention is administered orally. If the administration is via the oral route the functional fragment is preferably a single chain dy (scFv), diabody or lgG.

In typical ments, an anti-TNch antibody or functional fragment is present in a pharmaceutical composition at a concentration sufficient to permit intravenous administration at 0.5 mg/kg body weight to 20 mg/kg body weight. In some embodiments, the concentration of antibody or fragment suitable for use in the compositions and s described herein includes, but is not limited to, 0.5 mg/kg, 0.75 mg/kg, 1 mg/kg, 2 mg/kg, 2.5 mg/kg, 3 mg/kg, 4 mg/kg, 5 mg/kg, 6 mg/kg, 7 mg/kg, 8 mg/kg, 9 mg/kg, 10 mg/kg, 11 mg/kg, 12 mg/kg, 13 mg/kg, 14 mg/kg, 15 mg/kg, 16 mg/kg, 17 mg/kg, 18 mg/kg, 19 mg/kg, mg/kg, or a concentration ranging n any of the foregoing values, e.g., 1 mg/kg to mg/kg, 5 mg/kg to 15 mg/kg, or 10 mg/kg to 18 mg/kg.

The ive dose of an anti-TNFo antibody or functional fragment can range from about 0.001 to about 750 mg/kg per single (e.g., bolus) administration, le administrations or continuous administration, or to achieve a serum concentration of 001-5000 ug/ml serum concentration per single (e.g., bolus) administration, multiple administrations or continuous administration, or any effective range or value therein depending on the condition being treated, the route of administration and the age, weight and ion of the subject. In certain embodiments, each dose can range from about 0.5 mg to about 50 mg per kilogram of body weight or from about 3 mg to about 30 mg per kilogram body weight. The antibody can be formulated as an aqueous solution.

Pharmaceutical compositions can be conveniently presented in unit dose forms containing a predetermined amount of an anti-TNch antibody or functional fragment per dose. Such a unit can contain 0.5 mg to 5 g, for example, but without limitation, 1 mg, 10 mg, 20 mg, 30 mg, 40 mg, 50 mg, 100 mg, 200 mg, 300 mg, 400 mg, 500 mg, 750 mg, 1000 mg, or any range between any two of the ing , for example 10 mg to 1000 mg, 20 mg to 50 mg, or 30 mg to 300 mg. Pharmaceutically acceptable carriers can take a wide variety of forms depending, e.g., on the condition to be treated or route of administration.

Determination of the effective dosage, total number of doses, and length of treatment an NFo antibody or functional nt f is well within the capabilities of those skilled in the art, and can be ined using a standard dose escalation study. eutic formulations of the anti-TNFo antibodies and functional fragments suitable in the methods described herein can be prepared for storage as |ized formulations or s solutions by mixing the antibody having the desired degree of purity with optional pharmaceuticaIIy-acceptable carriers, excipients or stabilizers typically employed in the art (all of which are referred to herein as "carriers"), i.e., buffering agents, stabilizing agents, vatives, isotonifiers, non-ionic detergents, antioxidants, and other miscellaneous additives. See, Remington's Pharmaceutical Sciences, 16th edition (Osol, ed. 1980). Such additives must be nontoxic to the recipients at the dosages and concentrations employed.

Buffering agents help to maintain the pH in the range which approximates physiological conditions. They can present at concentration ranging from about 2 mM to about 50 mM.

Suitable buffering agents include both organic and inorganic acids and salts thereof such as citrate buffers (e.g., dium citrate-disodium citrate e, citric acid-trisodium citrate mixture, citric acid-monosodium citrate mixture, etc.), succinate buffers (e.g., succinic acid- monosodium succinate mixture, succinic acid-sodium hydroxide mixture, succinic acid- disodium succinate mixture, etc.), tartrate buffers (e.g., tartaric acid-sodium tartrate mixture, tartaric acid-potassium tartrate mixture, tartaric acid-sodium hydroxide mixture, etc.), fumarate buffers (e.g., fumaric acid-monosodium fumarate mixture, fumaric acid-disodium fumarate mixture, dium fumarate-disodium fumarate mixture, etc.), gluconate s (e.g., gluconic acid-sodium glyconate mixture, gluconic odium hydroxide mixture, gluconic acid-potassium glyuconate mixture, etc.), oxalate buffer (e.g., oxalic acid- sodium oxalate mixture, oxalic acid-sodium hydroxide mixture, oxalic acid-potassium oxalate mixture, etc), e buffers (e.g., lactic acid-sodium lactate mixture, lactic acid- sodium hydroxide mixture, lactic acid-potassium lactate mixture, etc.) and e s (e.g., acetic acid-sodium e mixture, acetic acid-sodium hydroxide mixture, etc.). onally, phosphate buffers, histidine buffers and hylamine salts such as Tris can be used.

Preservatives can be added to retard ial growth, and can be added in amounts ranging from 0.2%- 1% (w/v). Suitable preservatives include phenol, benzyl alcohol, meta- cresol, methyl paraben, propyl n, octadecyldimethylbenzyl ammonium chloride, benzalconium halides (e.g., chloride, bromide, and iodide), hexamethonium chloride, and alkyl parabens such as methyl or propyl paraben, catechol, resorcinol, cyclohexanol, and 3- pentanol. lsotonicifiers sometimes known as "stabilizers" can be added to ensure isotonicity of liquid compositions and include polyhydric sugar alcohols, preferably trihydric or higher sugar alcohols, such as glycerin, erythritol, ol, xylitol, sorbitol and mannitol. Stabilizers refer to a broad category of excipients which can range in function from a bulking agent to an additive which solubilizes the therapeutic agent or helps to prevent denaturation or adherence to the container wall. Typical stabilizers can be polyhydric sugar alcohols (enumerated above); amino acids such as ne, lysine, glycine, glutamine, asparagine, ine, alanine, ornithine, L-leucine, 2-phenylalanine, glutamic acid, ine, etc., organic sugars or sugar alcohols, such as lactose, trehalose, stachyose, mannitol, sorbitol, xylitol, ribitol, myoinisitol, galactitol, glycerol and the like, including cyclitols such as inositol; polyethylene glycol; amino acid polymers; sulfur containing reducing agents, such as urea, glutathione, thioctic acid, sodium thioglycolate, ycerol, d-monothioglycerol and sodium thio sulfate; low lar weight polypeptides (e.g., peptides of 10 residues or fewer); proteins such as human serum albumin, bovine serum albumin, gelatin or immunoglobulins; hydrophylic polymers, such as polyvinylpyrrolidone ccharides, such as , mannose, fructose, glucose; disaccharides such as lactose, maltose, sucrose and trisaccacharides such as raffinose; and polysaccharides such as dextran. Stabilizers can be present in the range from 0.1 to 10,000 s per part of weight active protein.

Non-ionic surfactants or ents (also known as "wetting agents") can be added to help solubilize the therapeutic agent as well as to protect the therapeutic protein t agitation-induced aggregation, which also permits the ation to be exposed to shear surface stressed t causing denaturation of the protein. Suitable non-ionic surfactants include polysorbates (20, 80, etc.), polyoxamers (184, 188 etc.), Pluronic polyols, polyoxyethylene sorbitan monoethers (TWEEN®-20, TWEEN®-80, etc.). Non-ionic surfactants can be present in a range of about 0.05 mg/ml to about 1.0 mg/ml, or in a range of about 0.07 mg/ml to about 0.2 mg/ml.

Additional miscellaneous excipients include bulking agents (e.g., starch), chelating agents (e.g., EDTA), idants (e.g., ascorbic acid, methionine, vitamin E), protease inhibitors and co-solvents.

The formulation herein can also contain a second therapeutic agent in addition to an anti- TNFo dy or functional fragment thereof. Examples of suitable second therapeutic agents are provided below.

The dosing schedule can vary from once a month to daily depending on a number of clinical factors, including the type of disease, severity of disease, and the patient's sensitivity to the anti-TNFo antibody or functional fragment. ln specific embodiments, an anti-TNFo antibody or functional fragment thereof is administered daily, twice weekly, three times a week, every other day, every 5 days, every 10 days, every two weeks, every three weeks, every four weeks or once a month, or in any range between any two of the foregoing values, for e from every four days to every month, from every 10 days to every two weeks, or from two to three times a week, etc.

The dosage of an anti-TNFo antibody or functional nt to be administered will vary according to the particular antibody, the subject, and the nature and severity of the disease, the physical ion of the subject, the therapeutic regimen (e.g., whether a second eutic agent is used), and the selected route of administration; the appropriate dosage can be readily determined by a person skilled in the art.

WO 58092 It will be recognized by one of skill in the art that the optimal quantity and g of individual dosages of an anti-TNFo antibody or functional fragment thereof will be determined by the nature and extent of the ion being treated, the form, route and site of administration, and the age and condition of the particular subject being treated, and that a physician will ultimately determine appropriate dosages to be used. This dosage can be repeated as often as appropriate. If side effects develop the amount and/or frequency of the dosage can be altered or reduced, in accordance with normal clinical practice.

Disorders to be treated The invention relates to a method of treating or preventing a human TNch-related disease in a subject, comprising administering to the subject the antibody or onal fragment as defined herein. The term "TNFd-related disorder" or "TNFd-related disease" refers to any disorder, the onset, progression or the persistence of the symptoms or e states of which requires the participation of TNch. Exemplary TNFd-related disorders include, but are not limited to, chronic and/or autoimmune states of inflammation in general, immune mediated inflammatory disorders in l, inflammatory CNS disease, inflammatory diseases affecting the eye, joint, skin, mucous membranes, central s system, gastrointestinal tract, urinary tract or lung, states of uveitis in general, retinitis, HLA—BZ7+ uveitis, 's disease, dry eye syndrome, glaucoma, n syndrome, diabetes mellitus (incl. diabetic neuropathy), insulin ance, states of arthritis in general, toid arthritis, osteoarthritis, reactive arthritis and Reiter's syndrome, juvenile arthritis, ankylosing spondylitis, multiple sclerosis, Guillain-Barré syndrome, myasthenia gravis, amyotrophic l sclerosis, sarcoidosis, glomerulonephritis, chronic kidney disease, is, psoriasis (incl. psoriatic arthritis), hidradenitis suppurativa, panniculitis, pyoderma gangrenosum, SAPHO syndrome itis, acne, pustulosis, stosis and osteitis), acne, Sweet's sydrome, pemphigus, s disease (incl. extraintestinal manifestastations), ulcerative colitis, asthma bronchiale, hypersensitivity pneumonitis, general allergies, allergic rhinitis, allergic sinusitis, chronic obstructive pulmonary disease (COPD), lung fibrosis, Wegener's granulomatosis, Kawasaki syndrome, Giant cell arteritis, Churg-Strauss vasculitis, polyarteritis nodosa, burns, graft versus host e, host versus graft reactions, rejection episodes following organ or bone marrow lantation, systemic and local states of vasculitis in general, systemic and cutaneous lupus erythematodes, polymyositis and dermatomyositis, dermia, pre-eclampsia, acute and chronic pancreatitis, viral hepatitis, alcoholic hepatitis, postsurgical inflammation such as after eye surgery (e.g. cataract (eye lens replacement) or glaucoma surgery), joint surgery (incl. arthroscopic surgery), surgery at joint-related structures (e.g. nts), oral and/or dental surgery, minimally ve cardiovascular procedures (e.g. PTCA, atherectomy, stent placement), laparoscopic and/or endoscopic abdominal and gynecological procedures, endoscopic urological procedures (e.g. prostate surgery, oscopy, cystoscopy, interstitial cystitis), or perioperative inflammation (prevention) in general, bullous dermatitis, neutrophilic dermatitis, toxic epidermal necrolysis, pustular dermatitis, al malaria, hemolytic uremic syndrome, allograft rejection, otitis media, snakebite, erythema nodosum, myelodysplastic syndromes, primary sclerosing cholangitis, seronegative spondylartheropathy, autoimmune hematolytic anemia, orofacial amatosis, matitis vegetans, aphthous stomatitis, phic tongue, migratory stoimatitis, Alzheimer disease, Parkinson's e, Huntington's disease, Bell's palsy, Creutzfeld- Jakob disease and neuro-degenerative conditions in general.

Cancer-related osteolysis, cancer-related inflammation, cancer-related pain, cancer-related cachexia, bone metastases, acute and c forms of pain, ective whether these are caused by central or peripheral effects of TNFo and whether they are fied as inflammatory, nociceptive or neuropathic forms of pain, sciatica, low back pain, carpal tunnel syndrome, complex regional pain syndrome (CRPS), gout, postherpetic neuralgia, fibromyalgia, local pain states, c pain syndroms due to metastatic tumor, dismenorrhea.

Particular disorders to be treated include states of arthritis in general, rheumatoid arthritis, osteoarthritis, reactive arthritis, juvenile arthritis; psoriasis incl. tic arthritis; inflammatory bowel e, including Crohn's disease, ulcerative colitis incl. proctitis, sigmoiditis, left-sided colitis, ive colitis and pancolitis, undetermined colitis, microscopic colitis incl. collagenous and lymphocytic colitis, colitis in connective tissue disease, diversion colitis, s in diverticular disease, eosinophilic colitis and pouchitis.

Most preferably, the antibody or functional fragment of the invention is used to treat an inflammatory bowel disease, in particular s disease, ulcerative colitis or microscopic colitis. The Crohn’s disease may be ileal, colonic, ileocolonic or isolated upper s disease (gastric, duodenal and/or jejunal) including non-stricturing/non-penetrating, stricturing, penetrating and perianal e behavior, allowing any combination of localization and disease behavior of any of the above mentioned. The ulcerative colitis may be ulcerative proctitis, proctosigmoiditis, left-sided colitis, pan-ulcerative colitis and pouchitis. ation Therapy and other aspects Preferably, the patient being treated with an anti-TNFor antibody or functional fragment thereof is also treated with another conventional medicament. For e, a patient suffering from inflammatory bowel disease, especially if having moderate to severe disease is typically also being treated with mesalazine or derivatives or prodrugs thereof, corticosteroids, e.g. budesonide or prednisolone (oral or i.v.), suppressants, e.g. oprine/6-mercaptopurine (6-MP) or methotrexate, cyclosporine or tacrolimus. Other medicaments which can be co-administered to the patient e biologics such as infliximab, adalimumab, etanercept, certolizumab pegol or others. Further medicaments which can be inistered to the patient include immunosupressants (e.g. azathioprine/6-MP or methotrexate or oral cyclosporine) in order to maintain stable and longer remission. Yet another aspect of the invention is the use of an anti-TNFor antibody or functional fragment as defined hereinabove for reducing inflammation.

Yet another aspect of the invention is an anti-TNFor antibody or functional fragment as defined hereinabove for use in reducing inflammation in a patient ing from an inflammatory condition.

A further aspect of this invention is a method of treating an inflammatory condition, comprising administering to a patient in need thereof an effective amount of an anti-TNFor antibody or functional fragment as defined hereinabove. The inflammatory condition is preferably one of the ions described above.

A further aspect of this invention is a method of preventing an inflammatory condition, sing administering to a patient in need thereof an effective amount of an NFor antibody or functional fragment as defined above. The inflammatory condition is preferably one of the conditions described above.

Table 1. Summary of the amino acid sequences SEQ ID NO: Description CDO'I-hCJONA c CDR L1 c CDR L2 Generic CDR L3 Generic CDR H1 Generic CDR H2 Generic CDR H3 SEQ ID NO: Description 7 CDR L1 of clone 16H05 8 CDR L2 of clone 16H05 and clone 16H10 9 CDR L3 of clone 16H05 and clone 16H10 CDR H1 of clone 16H05, clone 16B08 and clone 16H10 11 CDR H2 of clone 16H05 12 CDR H3 of clone 16H05 13 VHocif humanized scFv of clone 16H05-scO2 and clone H05-so 14 VL of humanized scFv of clone 16H05-scO2 Humanized scFv of clone 16H05-scO2 16 CDR L1 of clone 16B11 17 CDR L1 of clone CO8 18 CDR L1 of clone 16E05 19 CDR L1 of clone 16B08 CDR L1 of clone 16H10 21 CDR L1 of clone 17G07 22 CDR L1 of clone E06 23 CDR L2 of clone 16B11, clone 16B08 and clone 17G07 24 CDR L2 of clone 16CO8 CDR L2 of clone 16E05 26 CDR L2 of clone 17E06 27 CDR L3 of clone 16B11 28 CDR L3 of clone 16CO8 29 CDR L3 of clone 16E05 CDR L3 of clone 16B08 31 CDR L3 of clone 17G07 32 CDR L3 of clone 17E06 33 CDR H1 of clone 16B11 and clone 16E05 34 CDR H1 of clone 16CO8 CDR H1 of clone 17G07 36 CDR H1 of clone 17E06 37 CDR H2 of clone 16B11 and clone 17G07 38 CDR H2 of clone 16CO8 39 CDR H2 of clone E05 40 CDR H2 of clone B08 and clone 17E06 41 CDR H2 of clone 16H10 42 CDR H3 of clone 16B11 43 CDR H3 of clone 16CO8 44 CDR H3 of clone 16E05 45 CDR H3 of clone 16B08 46 CDR H3 of clone H10 47 CDR H3 of clone 17G07 48 CDR H3 of clone 17E06 49 Linker sequence in scFv 50 Linker sequence in diabody 51 Humanized diabody of clone 16H05 52 Light chain of humanized IgG of clone 16H05 53 Heavy chain of humanized IgG of clone H05 54 VL of humanized scFv of clone 16H05-scO4 SEQ ID NO: Description 55 zed scFv of clone 16H05-sc04 56 VK1 consensus sequence of framework l (Kabat positions 1-23) 57 VK1 consensus sequence of framework ll (Kabat positions 35-49) 58 VK1 consensus ce of ork lll (Kabat positions 57-88) 59 VA germline-based sequence of ork IV (see Table 16) 60 VA germline-based sequence of framework IV (see Table 16) 61 VA germline-based ce of framework lV (see Table 16) 62 VA germline-based sequence of framework lV (see Table 16) Examples Example 1: Generation of rabbit antibodies directed against human TNFd 1 . Results 1.1 Immunization Rabbits have been immunized with purified recombinant human TNFor (Peprotech, Cat.

No. 300-01A). During the course of the immunization, the strength of the humoral immune response t the antigen was qualitatively assessed by determining the maximal dilution (titer) for the serum of each rabbit that still ed detectable binding of the polyclonal serum antibodies to the antigen. Serum antibody titers against immobilized recombinant human TNFor were assessed using an -linked immunosorbent assay (ELISA, see 2.2.1). All three rabbits showed very high titers with a 10 x 106-fold dilution of the serum still resulting in a positive signal (at least 3-fold higher than the signal ed with serum from a nai've unrelated animal which was used as background control) in the ELISA. In addition, the ability of different rabbit sera to inhibit the biological activity of TNFor was assessed using a mouse L929 ased assay (see 2.2.3). All three sera inhibited TNFq-induced apoptosis of mouse L929 fibroblasts. Rabbit #3 showed the strongest neutralizing activity with 50% inhibition (IC50) reached at a serum dilution of 1:155’000.

Compared to rabbit #3, rabbit #2 and rabbit #1 showed approximately 3 and 21-fold lower activity, reaching 50% inhibition at a serum dilution of 1:55’500 and 1:7’210, respectively.

Lymphocytes isolated from spleens of all three animals were chosen for the subsequent hit identification procedures. The animals were prioritized based on the potency to inhibit the biological activity of TNFo in the L929 assay. Therefore, the highest number of hits that were cultivated originated from rabbit #3, and the lowest number of hits was derived from rabbit #1. 1.2 Hit Identification 1.2.1 Hit g Prior to the hit identification ure, a flow-cytometry-based g procedure was developed that specifically detects and allows for the isolation of high-affinity TNFo binding B-cells (see 2.1 ).

A total of 33 x 106 lymphocytes (corresponding to 1.5% of total cytes ed) derived from all three rabbits were characterized in two independent sorting campaigns. Out of the 33 x 106 cells analyzed in total 3452 B-cells expressing TNFd-specific antibodies (lgG) were isolated. The numbers of lymphocytes cloned were different for the three rabbits, as more cells were isolated from those rabbits whose sera showed strong inhibition of TNFo in the L929 assay. Of the isolated B-cells, 792 clones were derived from rabbit #1, 1144 clones from rabbit #2 and 1408 clones from rabbit #3. For 108 clones the respective rabbit origin is not known, because they are derived from a mixture of al lymphocytes from all 3 rabbits to allow optimal recovery of small amount of lymphocytes from the vials. 1.2.2 Hit Screening The results ed during the screening phase are based on assays performed with non- purified antibodies from culture supernatants of antibody secreting cells (ASC), as the scale of the high-throughput e does not allow for purification of the individual rabbit antibodies. Such supernatants were used to rank large s of antibodies ve to each other, however not to provide absolute values (e.g. for inhibition of biological activity of TNFo), except for binding affinity. ASC supernatants were screened in a high-throughput ELISA for binding to recombinant human TNFo. TNFo-binding supernatants were further terized for binding to Cynomolgus monkey TNFo by ELISA, binding kinetics and for their ial to neutralize the biological activity of human TNFo in the L929 assay. With the exception of binding kinetics, the reporting values of the high-throughput screenings should be interpreted as "yes" or "no" answers, which are based on single-point measurements (no dose-response). Affinity to Cynomolgus monkey and mouse TNFo was analyzed for all the 102 clones that were selected for amplification and sequencing of the antibody heavy and light chain le domains. 1.2.2.1 Binding to human TNFOi The aim of the primary screening is to identify ASC clones that produce antibodies specific for human TNFq. For this purpose, cell culture supernatants of 3452 ASC clones were analysed for the presence of antibodies to human TNFq by ELISA (see 2.2.1). The ELISA method used assesses the "quantity" of antibodies of the IgG subtype bound to recombinant human TNFOi, gives however no information about the affinity or the concentration of the dies. In this assay, supernatants from 894 ASC clones produced a signal that was clearly above background. The hit rate in screening was similar for rabbit #1 and rabbit #2 with 153 hits out of 792 (19.3%) identified from rabbit #1 and 225 hits out of 1144 fied from rabbit #2 ). Rabbit #3 showed a significantly higher hit rate of 34.4% resulting in the identification of 484 hits out of 1408. All 894 hits identified in this primary screening, were subjected to the measurement of binding kinetics by SPR (secondary screening). 1.2.2.2 TNFOi Binding Kinetics The aim of the secondary screening is to obtain quantitative information on the quality of target g for each hit from the primary screening by surface pIasmon resonance (SPR, see 2.2.2). In contrast to the ELISA used during the primary screening, this method assesses the kinetics of target binding as a function of time. This allows determination of the rate constants for association (ka) and dissociation (kd) of the dy from its target. The ratio kd/ka provides the equilibrium dissociation constant (KD), which reflects the affinity of an antibody to its target. Of the 894 hits identified in the primary screening, binding ties to human TNFq could be determined for 839 onal rabbit antibodies. For the ing 55 antibodies affinity could not be ed because the antibody concentration in the ASC supernatant was below the detection limit of the SPR instrument in the respective setup. The 839 anti-TNFOi antibodies that could be measured showed dissociation constants (KD) ranging from below 1.36 x 10'13 M to 1.14 x 10'8 M. 69% of all antibodies analyzed had a KD below 0.5 nM.

The median KDs of 2.21 x 10'10 M and 2.09 x 10'10 M for screening hits identified from rabbits #2 and #3 were similar while rabbit #1 showed about 2-fold higher values with a median KB of 4.65 x 10'10 M. When considering only lizing screening hits, the affinity distributions were r for all three animals with lower values for the median KD (median KDs between 1.4 x 10'10 M and 1.27 x 10'10 M). Affinities below 0.041 nM, 0.029 nM and 0.026 nM were measured for 5% of screening hits for rabbits #1, #2 and #3, respectively. For 2% of atants, affinities were even in the low picomolar range (below 6.2 pM, 7.9 pM and 11 pM). The excellent yield of high-affinity antibodies resulting from the secondary screening provides a broad basis for the selection of the most appropriate antibodies for humanization and reformatting. 3 Potency For the assessment of potency, a cell-based assay (L929 assay) has been developed (see 2.2.3). 506 out of the 894 selected antibodies (56.6%), inhibited TNFq-induced apoptosis in the L929 assay by more than 50%. In line with results obtained during titer analysis, the highest percentage of neutralizing hits was derived from rabbit #3 with a hit rate of 62.8%, followed by rabbit #2 with a hit rate of 56.4% and rabbit #1 with the lowest hit rate of 39.9%.

Affinities of these neutralizing antibodies ranged between 1.36 x 10'13 to 1.19 x 10'9 M. 1.2.2.4 Species reactivity (Cynomolgus monkey) All 894 hits identified in the primary screening, were analyzed for species cross-reactivity to Cynomolgus monkey TNFq by ELISA (see . The aim of this additional ing was to allow selection of ASC clones that are known to cross-react with Cynomolgus monkey TNFq. The ELISA method used assesses the "quantity" of antibodies of the IgG subtype bound to recombinant Cynomolgus monkey TNFq, gives however no information about the affinity or the concentration of the antibodies. Supernatants from 414 (46%) ASC clones produced a clear signal (optical density (OD) 2 1). The percentage of hits reactive to Cynomolgus monkey TNFq was r for rabbit #1 and rabbit #3 with 81 hits out of 153 (52.9%) identified from rabbit #1 and 236 hits out of 484 identified from rabbit #3 (48.8%).

With 37.8%, rabbit #2 showed a slightly lower percentage of cross-reactive hits resulting in the identification of 82 hits out of 225. 1.2.2.5 Selection of Clones for RT-PCR As a prerequisite for hit confirmation, gene sequence analysis and subsequent zation of the rabbit antibodies, the genetic information ng the rabbit antibody variable domain needs to be retrieved. This was done by reverse ription (RT) of the respective messenger RNA into the complementary DNA (cDNA), followed by amplification of the double stranded DNA by the polymerase chain reaction (PCR). The selection of ASC clones subjected to RT-PCR was primarily based on affinity and neutralizing activity. As additional criterion cross-reactivity to Cynomolgus monkey TNFo was considered. In total 102 ASC clones were selected for gene cloning by RT-PCR. First, the 93 best ranking ASC (in terms of affinity) with a KD below 80 pM, that inhibited the biological ty of TNFo in the L929 assay by more than 50% and that showed significant binding to Cynomolgus monkey TNFo were selected. Additionally, all the 9 best ranking ASC clones with KD below 20 pM that neutralized TNFo activity by more than 50% but did not bind to lgus monkey TNFo nevertheless were chosen as well. In total, 12, 13 and 66 ASC clones were successfully amplified and sequenced from rabbits #1, #2 and #3, respectively. 1.2.2.6 Identification of Related Clones with Desired Properties In order to characterize the genetic diversity of the panel of isolated ASC clones the sequences of the complementary ining regions (CDRs) were extracted and ted to a multiple sequence ent thus allowing sequence clustering in a phylogenetic tree.

While this analysis on one hand allows the selection of a diverse set of clonal ces to be carried forward into humanization and re-formatting experiments it also identifies homologous clusters of clonal sequences that appeared to share a common al B-cell clone in the rabbit. The rk of these sequence clusters are high sequence homology in the CDRs and a consistent pattern of codynamic properties. Both of these features are summarized for a cluster of eight clones in Tables 2 and 3. Despite the functional conservation of this sequence cluster the consensus ce in Table 3 reveals that a certain variability of the CDRs is tolerated, while still resulting in the desired pharmacodynamic profile.

Table 2: Pharmacodynamic properties of onal antibodies in ASC supernatants. .

ASC SN Affinity to human TNFa Affinity to Cynomolgus TNFa all—$9525: Clone ID ka(M'1s'1) Kd (s'l) KD (M) ka(M'1s'1) Kd (s'l) KD (M) % inh. 16B11 3.22E+06 1.87E-04 11 2.15E+06 1.37E-03 6.34E-10 71 16C08 1.77E+06 05 5.33E-11 1.80E+06 1.77E-04 9.82E-11 93 16E05 2.78E+06 8.27E-05 2.97E-11 2.53E+06 2.99E-04 1.18E-10 94 16B08 2.27E+06 4.53E-05 1.99E-11 2.35E+06 1.52E-04 6.45E-11 73 16H 10 2.03E+06 1.59E-04 7.85E-11 06 4.61E-04 2.10E-10 70 H05 1.90E+06 7.26E-05 3.82E-11 2.17E+06 6.97E-05 3.21E-11 67 GO7 05 1.37E-06 1.73E-12 7.80E+05 7.34E-05 9.41E-11 109 17E06 1.19E+06 3.59E-06 3.03E-12 1.45E+06 3.27E-05 2.26E-11 101 Table 3: The following sequence data regarding the CDRs were obtained for the above clones: CDR clone Sequence* SEQ ID NO: CDR L1 16—22—{05 QASQSIFSGJA 7 16—12—311 SNYJA 16 2.6—13—C08 QASQSISTAJA 17 16—13—305 QASQSIGRNJA 18 16—16—308 QASQSISNSJA 19 16—16—{10 QASQSIYSGJA 20 L7—13—G07 QASQSIGSNJA 21 17—20—306 QASQSISSSJA 22 QASQSIXXXJA 1 CDR L2 16—22—{05 GASKJAS 8 16—12—311 RASTJAS 23 2.6—13—C08 RASE‘ ES 24 16—13—305 QASKJAS 25 16—16—308 RASTJAS 23 16—16—{10 S 8 17—13—G07 RASTJAS 23 17—20—306 S 26 XASXJXS 2 CDR L3 {05 QSYYYSSSSSDGSYA 9 16—12—311 QSYYYSSSSSDGFFA 27 L6C08 QSYYYSSSSSDGSFA 28 16—13—305 QSYYYSSSNSDGSLA 29 16—16—308 QSYYYSSISSDGSYA 30 16—16—{10 SSSSDGSYA 9 17—13—G07 QSYYYSSSSSDGSVA 31 17306 QSYYYTSSTSDGSYA 32 QSYYYXSXXSDGXXA 3 CDR H1 16105 GIDFNNYGIG 10 16—12—311 GIDFSNYGIC 33 L6C08 GIDFSNYGIS 34 16—13—305 GIDFSNYGIC 33 16308 YGIG 10 16{10 GIDFNNYGIG 10 CDR clone Sequence* SEQ ID NO: L7—13—G07 GIDFSTYGIS 35 L7—20—306 GIDFSNYGIG 36 GIDFXXYGIX 4 CDR H2 L6{05 YIYPGFAITNFARSVKG ll 311 YIYPGFGITNYARSVKG 37 L6-l3-C08 YIYPGFGIRNYAPSVKG 38 L6-l3-EO5 YIYPGFGIRNYARSLKG 39 L6-l6-308 YIYPGFAIRNYARSVKG 40 L6-l6-{lO YIYPGFGITNFARSVKG 41 L7-l3-G07 YIYPGFGITNYARSVKG 37 L7EO6 YIYPGFAIRNYARSVKG 40 YIYPGFXIXNXAXSXKG 5 CDR H3 L6{05 DPVYATSSGYFDJ l2 L6-l2-311 SSGY.D. 42 L6C08 DPVYSSDWGYFNJ 43 L6-l3-EO5 DPVYASSSGY.D. 44 L6-l6-308 DPJYATSSGYFDJ 45 L6-l6-{lO DPVYASSSGYFDJ 46 L7-l3-G07 DPVYASSSAYYNJ 47 L7EO6 DPJYSTSSGYFNJ 48 XXXYXXJ 6 *Amino acids designated "X" have the g as defined in the accompanying sequence listing. 1.2.2.7 Cross-reactivity to Cynomolgus monkey TNFd (by SPR) e of the high number of high affinity hits that potently neutralized TNFd, species cross-reactivity was assessed for all monoclonal rabbit antibodies that were ted to RT-PCR in order to facilitate the selection of ASC clones for Hit confirmation. Affinities to Cynomolgus monkey TNFd were determined by SPR measurements similarly as described above (see also 2.2.2). The ties of the 93 tested antibodies for Cynomolgus monkey TNFd ranged from 9.6 x 10'12 to 2.1 x 10'9 M. 38 of the 93 cross-reactive antibodies bound human and Cynomolgus TNFd with similar affinity (less than two-fold difference in KD). Moreover, the difference in affinity between human and Cynomolgus was less than -fold for 79 of the 93 cross-reactive antibodies and less than 10-fold for 62 of them, which makes them acceptable for the preclinical development in the Cynomolgus monkey. 2. s 2.1 Sorting Assay ytometry based sorting procedure for the isolation of antigen-specific B-cells from rabbit lymphatic tissue was med as outlined by Lalor et al (Eur J lmmunol.1992;22.3001-2011) 2.2 Screening Assays 2.2.1 TNFa Binding ELISA (human and Cynomolgus monkey TNFa) Recombinant human TNFor (Peprotech, Cat. No. 300-01) was coated on a 96 well microtiter ELISA plate. Binding of rabbit antibodies in the ASC culture supernatants to the immobilized TNch was detected by a secondary HRP-labelled anti-rabbit lgG (Jacksonlmmuno Research, Cat. No. 5-046). TMB substrate (3,3',5,5'- tetramethylbenzidine, KPL, Cat.

No. 5300) was added and the colour reaction was d by the addition of H2804.

Plates were read using a microtiter plate reader ity reader M200 Pro, Tecan) at a wavelength of 450 nm.

Assay performance during the screening campaigns was monitored by a commercially available positive control Nch rabbit polyclonal antibody (AbD Serotec, Cat. No. 9295- 0174). For this purpose the positive control antibody was tested at 100 and 250 ng/ml in ate on each ing plate. Robustness and precision of the response of the positive l was monitored for each plate. At the final assay conditions, the signal-to-background ratio was between 30 to 40 for the positive control at 250 ng/ml and coefficient of variation (CV) of the positive control were below 10%. A signal with an optical density of 2100% relative to the 250 ng/ml positive control was considered as a primary screening hit.

For serum titer determinations, the same ELISA setup was used as described above. A serum dilution was considered positive when the binding signal of the immune serum was at least 3-fold higher compared to the signal of a naive unrelated animal.

Species cross-reactivity to Cynomolgus monkey was determined using a similar ELISA as described above. inant Cynomolgus monkey TNFor (Sino Biological, Cat. No. 90018-CNAE) was coated on 96 well microtiter ELISA plates. Binding of rabbit antibodies in the ASC culture supernatants to the immobilized Cynomolgus monkey TNch was detected by the HRP-labelled ary antibody as specified above. Immune serum from rabbit #2 was used as positive control at a dilution of 00 and 1:320’000. Robustness and precision of the response of the positive control was monitored for each plate. At the final assay conditions, the -to-background ratio was between 20 to 30 for the positive control at a on of 1:80’000 and CVs of the positive control were below 10%. 2.2.2 Binding Kinetics to TNFa by SPR (human and Cynomolgus monkey) g affinities of antibodies towards human TNch were measured by surface plasmon resonance (SPR) using a MASS-1 SPR ment (Sierra Sensors). Performance of the instrument was ied by means of standard reference solutions as well as by analysis of a reference dy-antigen interaction such as infliximab-TNch interaction.

For affinity screening, an antibody specific for the Fc region of rabbit lgGs (Bethyl Laboratories, Cat. No. A120-111A) was immobilized on a sensor chip (SPR-2 Affinity Sensor, High Capacity Amine, Sierra Sensors) using a standard amine-coupling ure.

Rabbit monoclonal antibodies in ASC supernatants were captured by the immobilized anti- rabbit lgG antibody. After capturing of the monoclonal antibodies, human TNFo (Peprotech, Cat. No. 300-01) was injected into the flow cells for 3 min at a concentration of 90 nM, and dissociation of the protein from the lgG captured on the sensor chip was allowed to proceed for 5 min. After each injection cycle, surfaces were regenerated with two injections of 10 mM Glycine-HCI. The apparent dissociation (kd) and association (ka) rate nts and the apparent dissociation equilibrium constant (KD) were calculated with the MASS-1 analysis software (Analyzer, Sierra Sensors) using one-to-one Langmuir binding model and quality of the fits was monitored based on relative Chi2 (Chi2 normalized to the extrapolated maximal binding level of the analyte), which is a measure for the quality of the curve fitting.

For most of the Hits the relative Chi2 value was below 15%. Results were deemed valid if the response units (RU) for ligand binding were at least 2% of the RUs for antibody capturing. s with RUs for ligand binding with less than 2% of the RUs for antibody capturing were considered to show no specific binding of TNch to the ed antibody.

Species cross-reactivity to Cynomolgus monkey TNch (Sino Biological, Cat. No. 90018- CNAE) was measured using the same assay setup and TNch concentrations and ng the same quality measures. The ve Chi2 was below 15% for most of the ASC atants ed. 2.2.3 TNFa-induced Apoptosis in L929 Fibroblasts The ability of rabbit lgGs from ASC culture supernatants to neutralize the ical activity of recombinant human TNch was assessed using mouse L929 fibroblasts (ATCC/LGC Standards, Cat. No. CCL-1). L929 cells were sensitized to TNch-induced apoptosis by addition of 1 ug/ml actinomycin D. Cells were cultured in l flat-bottom microtiter plates in the presence of 50% ASC culture supernatant and 100 pM (5.2 ng/ml) human TNFq (Peprotech, Cat. No. 300-01) for 24 h. ed to purified antibodies, higher concentrations of TNFq have to be used in the ce of ASC supernatants for hit screening. al of the cells was determined by a colorimetric assay using the WST-8 (2- (2-methoxynitrophenyl)(4-nitrophenyl)(2,4-disulfophenyl)-2H-tetrazolium, mono- sodium salt) cell proliferation reagent (Sigma Aldrich, Cat. No. 96992).WST-8 is reduced by cellular dehydrogenases to an orange an product. The amount of formazan produced is directly proportional to the number of living cells. Data were analyzed using a four- parameter logistic curve fit using the Softmax Data Analysis Software (Molecular s), and the concentration of infliximab required to neutralize TNFq-induced sis by 50% (le0) was calculated at a tration of 36.2 ng/ml. Therefore, the estimated lower limit of detection for this assay is between 30 to 40 ng/ml. This value is only a rough estimate for the detection limit, since the potential to block TNFq is not only ent on the concentration of the monoclonal dy but also on affinity of the antibody to the target.

However, the sensitivity of the assay is sufficient for screening of ASC supernatants since lgG concentrations in most ASC supernatants are above a concentration of 40 ng/ml.

Supernatants resulting in 50% neutralization of TNFq-induced apoptosis were considered positive.

To assure robust assay performance during the screening campaigns the positive control antibody infliximab was tested at 115 ng/ml (0.8 nM) and at 58 ng/ml (0.4 nM) in duplicates on each screening plate. Percent tion and precision of the response for the positive control was monitored for each screening plate. The acceptance criteria for each plate were set as follows: at least 60% inhibition with the positive control antibody at a concentration of 115 ng/ml with a coefficient of variation (CV) below 20%.

Example 2: Humanization and Generation of scFv 1 . Results 1.1 Hit Confirmation & Selection of Hits for Humanization 73 unique sets of parental rabbit light and heavy chain variable s were retrieved during hit screening and analyzed by sequence alignment. Based on the screening assay results and the sequence homology of the individual rabbit lgG clones, 30 candidates were selected for hit confirmation. 29 monoclonal dies were manufactured and the best performing clones in terms of affinity and potency were selected for the zation and lead candidate generation. The criteria for the selection of clones were i) lization of human TNFq in L929 assay, ii) high affinity to human TNFq, iii) cross-reactivity to Cynomolgus and Rhesus monkey TNFq, and iv) sequence diversity. One clone (16-22—H05) has been selected for humanization - one of the best ranking lgGs in terms of potency to neutralize human TNFq in the L929 assay. With respect to binding strength, the best possible ty is desired since a n loss of affinity needs to be pated as a result of humanization and reformatting into the scFv format.

The data for lgG clone No. 16-22—H05 are summarized in Table 4.

Table 4: In vitro binding and activity properties of purified monoclonal antibody (16-22—H05) Neutratizatien of flocking of Thin Blocking ofmnmmz Affinity to human INF Affinity to was TN? fiftinity to rhesus TM? TM? in 1929 assay INFR: interaction interaction "‘i his"? gnaw} mm?) ms‘i; mm) Mme; «dish warm) maid; rei.:<:,,,’ mug; 4.04E+{)5 1.31984 3.25%11 3.509506 4.892436 1.40312 2.315% 5.31E4M 2.65E~EG 335 1.62 0.88 2 £59, Mamas/mm Test 1.2 tion and Selection of Humanized scFv fragments The sequences encoding the complementarity determining regions (CDRs) were transferred in silico by CDR-loop ng onto a human variable domain scaffold sequence as described in clone, which transferred additional amino acids from the donor sequence at positions with structural relevance for the immunoglobulin domains and CDR positioning. An cial gene (with an optimized codon usage for bacterial expression) encoding the respective humanized single-chain antibody Fv (scFv) was synthesized (from the corresponding variable light and heavy chains). The polypeptide was then ed and subsequently characterized using similar assays as bed during the hit confirmation. 1.2.1 Humanization and cture of Humanized scFv (APls) The humanization of the selected clone comprised the transfer of the rabbit CDRs onto a scFv acceptor framework of the VK1NH3 type as described in process, which is schematically shown in Figure 1, the amino acid sequence of the six CDR regions was identified on the donor sequence (rabbit mAb) and grafted into the acceptor ld sequence, resulting in the constructs termed "CDR graft".

In addition, a second graft was designed, which included additional amino acids modifications from the rabbit donor on the positions L15, L22, L48, L57, L74, L87, L88, L90, L92, L95, L97, L99 and H24, H25, H56, H82, H84, H89, H108 (AHo numbering), which have been described to potentially nce CDR positioning and thus antigen binding (Borras et al. JBC. 2010; 285:9054-9066). These humanized construct is termed "structural (STR) graft". In case the comparison of the characterization data for these two initial constructs revealed a significant advantage of the STR uct onal variants were designed that ed the CDR grafted VL with STR grafted VH. This combination has been proven to be often sufficient to retain the activity of the STR graft (Borras et al., 2010, JBC, 285:9054-9066) and would generally be preferred as fewer man alterations in the human acceptor ld reduce the risk for ed stability and also the potential for immunogenicity.

Once the in-silico construct design described in the us section was completed the corresponding genes were synthesized and bacterial expression vectors were constructed.

The sequence of the expression constructs was confirmed on the level of the DNA and the constructs were manufactured ing to generic expression and cation protocols.

The logous expression of the proteins was performed in E.coli as insoluble inclusion bodies. The expression culture was inoculated with an exponentially growing starting culture. The cultivation was performed in shake flasks in an l shaker using commercially available rich media. The cells were grown to a defined ODBOO of 2 and induced by overnight expression with 1 mM lsopropyl B-Dthiogalactopyranoside (IPTG).

At the end of fermentation the cells were harvested by centrifugation and homogenized by sonication. At this point the expression level of the different constructs was determined by SDS—PAGE analysis of the cell lysate. The inclusion bodies were isolated from the homogenized cell pellet by a centrifugation protocol that included several washing steps to remove cell debris and other host cell impurities. The purified ion bodies were solubilized in a denaturing buffer (100 mM Tris/HCI pH 8.0, 6 M Gdn-HCI, 2 mM EDTA) and the scFvs were refolded by a scalable refolding protocol that generated milligram amounts of natively folded, ric scFv. A standardized protocol was employed to purify the scFvs, which included the following steps. The product after refolding was captured by an affinity chromatography employing Capto L agarose (GE Healthcare) to yield the purified scFvs. Lead candidates that met the affinity and y criteria in l testing were further purified by a polishing size-exclusion chromatography using a HiLoad Superdex75 column (GE Healthcare). Subsequent to the purification protocol the proteins were formulated in a buffered saline solution and terized by the various biophysical, protein interaction and biological methods, as described in the following. The ibility of the ent constructs was compared by determining the final yield of purified protein for the batch and normalizing this value to 1 L of refolding volume. 1.2.2 sical Characterization of zed scFv The biophysical characterization of the scFv with respect to stability and ibility were compiled in Table 5. The producibility and stability of the scFv construct was characterized by the different reporting points as discussed in the subsequent sections.

The scFv was investigated as to certain criteria, as explained in the following.

The producibility criterion shall ensure that the selected scFv entity can be expressed, refolded and purified in sufficient amounts to support later development of the lead molecule. The defined criteria were the expression yield of scFv per liter of fermentation broth, as assessed by GE, and the purification yield achieved in the generic lab- scale process, as assessed by measurement of the amount of purified protein by UV spectrometry, calculated back to 1 liter of refolding solution.

The criteria for stability were ed to assess the aggregation propensity during the manufacturing process of the molecules and their structural integrity during storage and further handling. The monomer content determined by SE-HPLC allows assessing the colloidal stability of molecules during the purification process (2.2.3). In a subsequent stability study the monomer content was tested over a duration of 4 weeks at 1 and 10 mg/mL and storage at 4, -20 and < -65°C. In addition, the colloidal stability of the proteins was tested after 5 freezing and thawing cycles. As an additional ity indicating ter, the midpoint of thermal unfolding was determined by differential scanning fluorimetry (DSF) (2.2.4) to e a read-out for the conformational stability of the lead candidates.

Table 5: Summary of the biophysical characterization data for the humanized scFvs.

Clone ll) Construct ity Producibility Tm Steerag Freeze/Thaw Purity EXpression Purfiieclagmn I "CI lAVa} {0/a} {g/L1 [A963 [mg/L} 455°C ~25"C 4°C 16»22~H05r5501 ISZZ-HBS-scflz 05 1622«H05-sc04 1.2.2.1 Producibility Assessment The lead candidate scFv molecules were expressed by shake flask fermentation in batch mode and purified by a generic lab-scale process to yield the n s for further characterization. During this process some key performance parameters were monitored to compare the candidate molecules and to identify potentially difficult to develop constructs.

The expression titer was determined on the level of the crude Eco/i lysate after the harvest of the cells by centrifugation. During the harvest a small loss of cells is anticipated, however, this factor was chosen to be neglected for the ation of the expression yield in favor of a more conservative estimation of the productivity. For the quantification of the scFv t in the lysate coomassie stained reducing SDS—PAGE (2.2.1) was chosen due to the high specificity of the method that allows discriminating the product from the host cell proteins in the sample.

A second criterion to assess the producibility is the cation yield of scFv ated per liter of refolding solution. This parameter addresses the potential neck in the anticipated manufacturing process that includes a protein refolding step. Since the efficiency of the refolding procedure has proven to be limiting in comparable manufacturing processes it has been decided to e the performance of the different constructs with respect to the producibility normalized to a defined refolding volume. For the calculation of the yield the final protein sample from each batch was quantified by UV absorbance (2.2.2) and divided by the actual refolding volume of the respective purification (Table 6).

Table 6: Summary of producibility data for two humanized scFvs. The expression titer was determined by quantitative SDS—PAGE on lysates of end-of-production cells. The batch yield was determined by UV absorbance measurement of the final purification pool. The purification yield is calculated as the purified scFv per liter of refolding volume.

Producibility Re 9f Ed'mg Pur: Ea Km'f‘ t‘ Expression Titer Batch Yield Construct lD Volume Yield [g/L] {mg} [Li {mg/L} 16~229H05-sc02 (3.2? 14.3 0.48 29.8 16—22’HOS-sc0fi 0.26 13.9 0.49 24.1 1.2.2.2 Stability Assessment The assessment of the conformational ity, monodispersity and structural integrity of the scFv ucts is an al component for the ranking of the different molecules with respect to the developability. A prerequisite for the meaningful comparison of the ent constructs is the preparation of purified molecules of similar quality. The criterion er purity" determined by SE-HPLC is intended to ensure compatible quality of the different test substances. In addition to the C analysis, SDS—PAGE for the determination of protein purity and identity was performed to confirm able quality of the tested preparations.

The SE-HPLC results of the two scFvs reveal that all preparations could be purified to a monomer t of 2 99% (Figure 2).

The thermal unfolding behavior of the lead candidates was tested by differential scanning fluorimetry (DSF) to allow ranking of the molecules with respect to their expected conformational stability. A normalized plot of the fluorescence raw data is shown in Figure 3, which depicts duplicate measurements of each sample. A cooperative unfolding behavior was observed. The two molecules 16H05-sc02 and 16H05-sc04 showed a Tm of 71.3 and 726°C, respectively.

In a second arm of the stability assessment the monodispersity of the molecules was red over the duration of 4 weeks at different temperatures. The results for the ity study and the resulting monomer contents are shown in Figure 4. Both molecules (16 H05-sc02 and H05-sc04) start at a monomer content exceeding the minimum of 95% monomer and lose less than 5% of monomer with respect to the respective starting value at a concentration of 10 mg/ml. In the frozen state at -20°C and <-65°C the samples only showed minimal differences over time. At the most stringent condition (4°C) the le 16H05-sc02 lost as little as 0.2% of r during the 4 weeks. In on a stress stability study was conducted at a temperature of 37°C and a scFv concentration of 10 mg/ml for up to 4 weeks. At this condition a more stringent discrimination of the propensity for aggregation of the different constructs is expected. The resulting data summarized in Figure 6 revealed a monomer loss of 15% after 28 days. Both scFv demonstrated good monomer stability at stress conditions. Chromatograms of the stability study at 4°C are provided in Figure 5, where the sample at day 0 and after 28 days at 4°C is shown. In this chromatogram overlay also the results of the freeze / thaw stability is shown. For this part of the study the samples were repeatedly frozen and thawed for a total of 5 cycles. The resulting quantification of the monomer content by analytical C did not reveal any changes in the two samples (Table 5).

A SDS—PAGE analysis was performed for the two scFvs to generate supportive data for the quantification by UV absorbance, confirming the purity of the sample preparation and thereby conferring specificity for the content fication. In another aspect of this analysis the SDS—PAGE results revealed the e of protein degradation during the stability study (28 days at 4°C and a concentration of 10 mg/ml compared to sample from day 0 stored at <-65°C), which is an important characteristic from a developability perspective.

It is important to note that the different s med within the scope of this assessment address distinct mechanistic aspects of protein stability. The determination of the thermal unfolding temperature of a protein will give complementary results to the ement of the monodispersity by SE-HPLC upon storage at elevated temperature.

While both methods are designed to give an estimation of the ial product shelf live and stability the mechanisms addressed are profoundly different. The midpoint of transition (Tm) assessed by thermal unfolding is a qualitative measure for protein domain stability (does not allow for thermodynamic ination of AG). Highly stable protein domains (high Tm) are less likely to spontaneously unfold at ambient temperature and thus less prone to irreversible aggregation/precipitation driven by unfolded domain interactions. High domain stability indicates dense packaging of amino acid residues, which also ates with resistance towards protease cleavage. The SE-HPLC assessment on the other hand tatively determines the content of the monomeric fraction as well as of soluble ers/aggregates. Such soluble oligomers are imes reversible and relatively loose associations driven by electrostatic or hydrophobic interactions between correctly folded proteins. There is some correlation between Tm as assessed by thermal unfolding and the propensity for oligomer/aggregate formation as assessed by SE-HPLC ularly for proteins with "border line" stability. Beyond a certain threshold Tm of approximately 60°C dy variable domains are generally sufficiently stable to be resistant toward aggregation/precipitation and lytic degradation due to partial domain unfolding at ambient temperature. Oligomerization driven by hydrophobic and/or electrostatic interactions of surface residues may, however, still occur. lmportantly, in an accelerated (stress) ity study at elevated temperature (e.g. 37°C) the various mechanisms of oligomer formation and precipitation may occur simultaneously. 1.2.3 Characterization of in vitro binding and activity of zed scFvs In the following the humanized scFvs were characterized in vitro for their target binding properties and potencies. Binding kinetics (ka, kd and KD) to human TNFo and potency to neutralize TNFd-induced apoptosis of L929 fibroblasts was analyzed. Additionally, the y to inhibit Cynomolgus monkey (Macaca fascicularis) and Rhesus monkey (Macaca mulatta) TNFo d sis as well as the y to inhibit the interaction n human TNFo and TNFRI/TNFRII by ELISA and target selectivity for binding to TNFoc over TNFB was determined.

For the understanding of the results below it is important to note that both, the transfer of the rabbit CDRs onto a human variable domain scaffold as well as the change in the format from the full-size lgG to the scFv fragment may impact on pharmacological properties. For example, a certain loss of affinity is usually associated with humanization. Further, due to the smaller size of the scFv compared to the lgG the ability of a scFv to interfere with interaction rs through steric hindrance is largely reduced. Last but not least it shall be noted that due to its bivalent mode of binding to the homo-trimeric TNFo, the ty of the parent lgG may have been ed too high (SPR artifact). Consequently, when comparing affinities n the parental bivalent rabbit lgG and the humanized monovalent scFv, the reported "loss in affinity" may be overestimated. 1.2.3.1 Affinity Affinity of humanized scFvs to human TNFo was determined by SPR measurements (see also 2.1.1). Affinity was determined using 2-fold serial dilutions of the respective scFvs. The scFvs were derived from a rabbit monoclonal dy. Two scFv variants were generated , named "CDR" (CDR) and "structural graft" (STR). To assess the relative contribution of framework substitutions in the light and the heavy chain and to possibly reduce the number of rabbit amino acid residues introduced in the human framework, domain shuffling experiments were performed. Therefore, scFv constructs containing a CDR grafted light chain and a structural grafted heavy chain (CDR/STR) were generated for clone 16H05.

The top ranking scFvs H05-sc02 (STR) and 16H05-sc04 (CDR/STR) bound with affinities of 4.5 x 10'11 and 1.1 x 10'10 M, respectively. H05-sc04 showed only a slight reduction of affinity when compared to its "structural graft" variant (16H04-sc02) (see Table 7). These results suggest that affinity of the humanized scFvs mainly depends on the few rabbit amino acids introduced into the human heavy chain framework. 1.2.3.2 Potency The ability of the humanized scFvs to neutralize human TN For was analyzed using the L929 assay (see also 2.1.2). The potency (le0 and ngo) to neutralize TNFd induced apoptosis was ed for 16H5 derived scFvs and compared to the y of the reference antibody infliximab to allow for direct comparison of |C50 and |C90 values from different assay plates. Relative |050 and ngo values were calculated in mass units (ng/ml) of infliximab and the scFvs. y analysis was performed several times on different days with different lots of antibody fragments. Figure 7 shows representative dose-response curves from one experiment for each of the two scFvs. Mean values of replicate measurements are shown in Table 7 ard deviations are ized in the legend of the table).

The humanized scFvs inhibited TNch induced apoptosis with lower |050 and ngo values than imab (see Table 7). Consistent with SPR results, the domain shuffled variant 16- 22-H05-sc04 (CDR/STR) exhibited equal potency when compared to the structural graft 16- 22-H05-sc02 (STR). The scFvs 16H05-sc04 and 16H05-sc02 showed ent TNFo-neutralizing activities, with |C50 values of 14.6- and 13.1-fold better than infliximab, respectively. ngo values of 16H05-sc04 and 16H05-sc02 were 13.1- and 12.6-fold better than for infliximab, respectively. As observed for the parental rabbit monoclonal antibodies there was no clear ation n affinity and potency of antibodies (correlation not shown). Nevertheless, scFv derived from 16H05 showing the highest affinities (16H05-sc02 (STR) and 16H05-sc04 (CDR/STR)) also showed highest potency. Additionally, results from the neutralization assays suggest that a certain threshold ty needs to be achieved for efficient inhibition of TNFo signaling. For example, scFvs 16D08—sc01 (CDR), 16CO9-sc01 (CDR), 16H07-sc01 (CDR) and 17G01- sc01 (CDR), all binding to TNFo with affinities above 1 nM, show poor potential to neutralize TNFo (not . 1.2.3.3 Species cross-reactivity (Cynomolgus and Rhesus monkey TNch) Species cross-reactivity for the top ranking scFvs was determined by two methods: 1) potency to neutralize Cynomolgus monkey and Rhesus monkey TNch in the L929 assay and 2) affinity to Cynomolgus monkey and Rhesus monkey TNFo by SPR. The y to neutralize TNch from the different species was determined by the L929 assay similarly as described above for human TNch using Cynomolgus monkey and Rhesus monkey TNch, respectively (see also 2.1.2). TNFo from both species showed very similar potency to induce L929 apoptosis (data not shown). Therefore, same trations of human and monkey TNch were used for species cross-reactivity testing. Additionally, binding kinetics (by SPR) to Cynomolgus monkey and Rhesus monkey TNch were determined using a r assay as for human TNch (see also 2.1.1).

All scFvs d from the clone 16H05 showed cross-reactivity to Cynomolgus monkey and Rhesus monkey TNch (see Table 7). The affinities were similar, namely 2.0 x 10'10 and 2.3 x 10'10 M for Cynomolgus monkey and Rhesus monkey, respectively. The ence in affinity between human and monkey TNch was about 5-fold. Potencies to neutralize Cynomolgus monkey, Rhesus monkey and human TNch correlated well with the affinities to the respective TNchs. uently the two clones d from 16H05 showed between 5- to 7-fold lower potencies towards monkey TNch as compared to human TNch (see Table 7 and Figure 8). To summarize, the two scFvs showed species cross-reactivity to Cynomolgus and Rhesus TNch. 1.2.3.4 Blocking of the human TNFd-TNFRI/ll interaction In on to the L929 assay, the potency of each humanized scFv to inhibit the interaction between human TNFo and TNFRI/ll was assessed by ELISA (see . Similarly to the L929 assay, individual |050 values on each plate were calibrated against the IC50 of the reference molecule infliximab that was taken along on each plate and relative leo and ngo values were calculated in mass units (ng/ml) of lnfliximab and the scFvs.

Neutralization assays can distinguish potencies of target ng antibodies only if they bind their target with an equilibrium binding constant (KD) that is higher than the target concentration used in the y assay (KD > target concentration). For the L929 assay a TNFo concentration of 5 pM was used while in the TNFRI/II inhibition ELISAs a TNFo concentration of 960 pM was used. ore, theoretically, the L929 assay can differentiate potencies between scFvs with KD > 5 pM, while the inhibition ELISA can only differentiate potencies between scFvs with KD > 960 pM. Since all of the scFvs analyzed showed KDs below 960 pM, potencies between scFvs with different affinities (but similar mechanism of ) can be differentiated only in the L929 assay. 16H5-sc02 and 1605-sc04 showed potencies for blocking of the TNFor-TNFRI interaction between 2.8 and 3.5-fold higher compared to infliximab while the potency compared to infliximab in the L929 assay was significantly higher (13.1 and 14.6-fold).

When comparing the relative IC50 values for the parental rabbit IgG (see Table 2) with the relative |C50 values for the humanized scFvs (Table 7) potencies of the scFvs are in general slightly higher ed to the parental lgG although affinities in general are in the same range for the parental rabbit lgG. Since potencies of antibodies and scFvs were compared in mass units, the number of valencies (TNFor g sites) at each concentration is about 2.9-fold higher for the monovalent scFvs compared to the more than five-fold heavier but bivalent lgG. With very high-affinity binding scFvs, this results in more potent blocking of the TNFo and TNFRI/II interaction e the lack of avidity is no longer al for activity. In contrast, with low-affinity monovalent domains the opposite has been published eters et al. Arthritis & Rheumatism, 2006; 54:1856-1866). For the reasons mentioned above, results from the inhibition ELISA were not used for g of potencies n the different dies but primarily for comparison of the potential of antibodies to block the interaction with TNFRI versus TNFRII. The investigated scFvs blocked the interaction between both TNFor receptors with comparable potencies (Table 9, Figure 9 and Figure 10). 1.2.3.5 Target specificity (Selectivity for binding to TNFoc versus TNFB) Specificity of the two scFvs (16H05-sc02 and 16H05-sc04) for TNFoc over TNFB was confirmed by assessment of the relative potential of TNFB as compared to TNFor to aximally inhibit TN For binding to each scFv and was measured in a competition ELISA (see also 2.1.4). The quality of recombinant human TNFB has been ed 1) for purity by SDS-page and HPLC analysis, and 2) for biological activity in the mouse L929 cytotoxicity assay, by the manufacturer of the protein. As shown in Figure 11, the interaction between each of the scFvs with biotinylated TNFoc was blocked by unlabeled TNFoc with IC50 values g from 60 to 260 ng/ml, while TNFB did not show any significant effect even at the highest concentration of TNFB tested (1250 ug/ml). Hence, all of the scFvs analyzed bind specifically to TNFoc but not to its closest homologue, TNFB. TNFB did not show any significant inhibition of TNFoc binding to scFvs at the concentrations tested. Therefore, the TNFB tration required to half-maximally inhibit TNFoc g has to be significantly higher than the highest concentration of TNFB used in the assay (1250 . When comparing concentrations of TNFoc and TNFB required to half-maximally inhibit TNFoc binding to the scFvs, the selectivity for binding to TNFoc over TNFB is significantly higher than approximately 5000 to 20000 fold for all of the fragments tested (see also Table 7).

Therefore, off-target binding of any of the scFvs appears highly unlikely.

The results of the experiments described above are summarized in tables 7 to 9.

W0 58092 V6 w> ooo_o_\ L 89 ooo_o_\ 58 I— 509% EZH Un_Z._. ._9 AA AA mm; QN QM m_m>_mcm Noom-moI-NN-©_‘ b_o_:omaw-w®_omqm 3519; QN Wm >890; md md "mco:m_>mc 8% .955 .vN ".39.? QNF _\.m_\ $2 H >oc90n_ 32:33: 9228 30 0.0 o $5525.. Emccflm nomo_ ¥omo_._9 _\.m_\ Q: $8952 m, 956:8 ._®._ cmz._. Evox o_\-m_m.N 8% "Eb a Ucm do SN. 8.0 3.0 wswmcm $2 .955 9: ©.o be; vo-m_m._\ 30 v: £59m 8 6,3 3 5.9 .mg H a £_>> 2 2, 0% No N50: >=cE< to mEmEmmt ©.v_‘nom0_ $12me mo+m5m 58952 v "moo ._®._ mEEouE: .m>u_om 359:5 HANHS Evox o_\-m_o.N cmz._. UL Q: 9% UmNEmEsc US voom-moI-NN-©_‘ be; mommd v: 2 .co_Em>c_ EELmE. 9.2 >=Et< EthE $12me mo+mm.v "6 "comma mmEona "Wm 9: £_>> @529me 22> Evox _\_\-m_m.v o_\-m__\._\ ho 5 395 mam EN m>mc cmz._. 33:8 ©.N_‘nom0_ m>u_om cmEsc EthE be; mommé mo-m__\N ._®._ Ucm Ucm >=Et< $12me mo+mm.m mo+mw.m UmNEmEsc 8598qu 8698qu 22> 32 owe owe mcficfi we 3::ch 3::ch US 92:: H 0:3 _Sm>mm c980 mkmioo Eomdmg E0399 _‘.m_‘nom0_ >45:ch >45:ch mcozmoEoQO N ._®._ .w 05m... 82:0th "8ch >n_ow Noow-mOI-NN-m_\ voow-mOI-NN-m_\ 35ngde "3&ngde 05m... moom-moz-mm-$ voom-moz-mm-$ 8:09.35 ac mém NNH‘ __mu_z._.-_ou_z._. :39E ac o.om odm 36 339 920on 0m me 339 Em 0m .cozomBEH He NmmZHlomZH. :39E m9» 8:09.35 ac mdm HmquH.-_ou_z._. :39E mZH. mc QmH‘ NC 36 339 920on 2 9: Hum 9. 339 33.3.0 3335 0m QN m>u_om ._9 3330_ 3330_ 5851 .3335 .3335 0.9m... E03 Noom-moI-NN-©_‘ voom-moI-NN-o_‘ 33330_ 33339 .3 .3 2. Methods 2.1 Lead terization Assays 2.1.1 Binding cs and Species Cross-reactivity by SPR Binding affinities of scFvs towards human TNFq were measured by surface plasmon resonance (SPR) using a MASS-1 SPR instrument (Sierra Sensors). Performance of the SPR assay was qualified by analysis of a reference antibody antigen interaction such as certolizumab-TNFq interaction. The pegylated Fab-fragment certolizumab was selected as reference due to its monovalent binding mode r to that of scFvs. Using the same assay setup as for affinity measurements of the scFvs, a value of 9.94 x 10'11 M was determined for the affinity of certolizumab to TNFd. This value is in good agreement with published KD values of 9.02 i 1.43 x 10'11 M (BLA certolizumab; BLA number: 125160; submission date: April 30, 2007).

For affinity measurements of scFvs human TNFq (Peprotech, Cat. No. 300-01) was immobilized on a sensor chip (SPR-2 Affinity Sensor, Amine, Sierra Sensors) by amine- coupling to reach an lization level of 50 to 100 RUs (immobilization levels achieved during SPR analysis were between 40 to 120 RUs). In a first step, affinity screening of scFvs was med using only one scFv concentration (90 nM). In a second step, for the best performing scFvs, Single ion Cycle Kinetics (SiCK) were measured from a single injection cycle by simultaneously injecting six analyte samples at different concentrations into each of the eight parallel channels in the MASS-1 system. For affinity ings, humanized scFvs were injected into the flow cells at a concentration of 90 nM for three minutes and dissociation was monitored for 12 minutes. For the subsequent more precise affinity determinations, two-fold serial dilutions of scFv ranging from 45 to 1.4 nM were ed into the flow cells for three minutes and dissociation of the n from TNFd lized on the sensor chip was d to proceed for 12 minutes. The apparent dissociation (kd) and association (ka) rate constants and the apparent dissociation equilibrium constant (KD) were calculated with the MASS-1 analysis software (Analyzer, Sierra Sensors) using one-to-one Langmuir binding model and quality of the fits was monitored based on Chi2, which is a measure for the quality of the curve fitting. The smaller the value for the Chi2 the more accurate is the fitting to the one-to-one Langmuir binding model. For affinity screenings, s were deemed valid if the Chi2 was below 10 for the concentration analyzed. In cases where several scFv concentrations were analyzed, results were deemed valid if the average Chi2 over all the concentrations tested was below 10. ance criteria were met for all scFvs tested.

Species cross-reactivity to lgus monkey (Sino Biological, Cat. No. CNAE) and Rhesus monkey (R&D Systems, Cat. No. 1070-RM-025/CF) TNch tech, Cat. No. 315- 01A) was measured using the same assay setup and applying the same quality measures as described above for human TNch. For Cynomolgus and Rhesus monkey TNch immobilization levels ranging from 50 to 180 RUs and from 90 to 250 RUs, respectively, were achieved. The scFvs were analyzed using two-fold serial dilutions with concentrations ranging from 45 to 1.4 nM. The e Chi2 values were below 10 for all of the scFvs . 2.1.2 TNFa—induced sis in L929 Fibroblasts (neutralization of human, non-human primate and TNFa by scFvs) The ability of scFvs to neutralize the biological activity of recombinant human TNFo was assessed using mouse L929 lasts (ATCC/LGC Standards, Cat. No. CCL-1). L929 cells were sensitized to TNFd-induced apoptosis by addition of 1 ug/ml actinomycin D. Three-fold serial dilutions of anti-TNch reference antibody or scFvs (3000-0.05 ng/ml) and 5 pM recombinant human TNch (Peprotech, Cat. No. 300-01) were pre-incubated at room temperature for 1 hour. The used TNch concentration (5 pM) induces submaximal L929 apoptosis (ECgo). After addition of the agonist/inhibitor mixtures the cells were ted for 24 hours. Survival of the cells was determined by a colorimetric assay using the WST-8 (2- (2-methoxynitrophenyl)(4-nitrophenyl)(2,4-disulfophenyl)—2H-tetrazolium, mono- sodium salt) cell proliferation reagent (Sigma Aldrich, Cat. No. 96992). WST-8 is reduced by cellular dehydrogenases to an orange formazan product. The amount of formazan ed is directly proportional to the number of living cells. Data were analyzed using a four- parameter logistic curve fit using the Softmax Data is Software (Molecular Devices), and the concentration of reference antibody or scFvs required to neutralize TNFd-induced apoptosis by 50% and 90% (le0 and ngo) was calculated (see also Figure 7). In order to render IC50 and ngo values directly comparable between experiments that were performed on different days or on ent assay plates, IC50 and ngo values were calibrated against the reference antibody infliximab. To control precision of the response, the dose-response curves were analyzed in duplicate. Standard deviations and CVs were calculated for each measurement point (CV < 20%).

Species cross-reactivity to Cynomolgus monkey (Sino Biological, Cat. No. 90018—CNAE) and Rhesus monkey (R&D s, Cat. No. 1070-RM-025/CF) TNch was measured using the same assay setup and applying the same quality measures as described above for human TNFq. Similarly to the human counterpart, TNFq concentrations that induce submaximal L929 sis (ECgo) were used for species cross-reactivity g. TNFq from both species showed very similar potency to human TNFq to induce L929 mouse fibroblast sis. uently the same concentration of TNFq (5 pM) was used for oth species tested. During species cross-reactivity testing CVs of most of the duplicate measurement points were below 10%. 2.1.3 TNFa tion ELISA The inhibitory effect of scFvs on ligand binding was assessed using an ELISA, a biochemical method solely reproducing the interaction between TNFq and TNFRI and TNFRII.

For the first inhibition ELISA, the extracellular domain of TNFRI fused to the Fc region of human lgG (R&D Systems, Cat. No. 372-Rl) was coated on a 96-well Maxisorp ELISA at a concentration of 0.5 ug/ml. For the second inhibition ELISA, the extracellular domain of TNFRII fused to the Fc region of human lgG (R&D Systems, Cat. No. 726-R2) was coated at a concentration of 2 ug/ml. All subsequent steps were cal for both . In order to detect binding of TNFq to TNFRI and TNFRII, TNFq was biotinylated prior to its use.

Biotinylated human TNFq (960 pM, 50 ng/ml) was first incubated with 3-fold serially d humanized anti-TNFq scFvs and infliximab (10’000 ng/ml-0.2 ng/ml) for 1 hour at room ature. The TNFq/antibody fragment mixtures were transferred to the TNF receptor immobilized plates and binding of unblocked TNFq to the immobilized TNFq receptor was detected after incubation at room temperature for 20 minutes with the biotin-binding streptavidin-HRP (SDT Reagents, Cat. No. SP4OC). Addition of 3’,5,5’-tetramethylbenzidine (TMB) substrate ed in a colorimetric read-out that was proportional to the binding of TNFq to TNFRI and TNFRII. Before use in the competition ELISA, the biological activity of the biotinylated TNFq was confirmed in the L929 assay. The E050 of biotinylated TNFq was similar to the E050 of unlabeled TNFq (data not shown). Similar to the L929 assay described above, data were ed using a four-parameter logistic curve fit using the Softmax Data Analysis Software (Molecular Devices), and the concentration of scFvs required to inhibit interaction of TNFq and TNFR by 50% and 90% (le0 and ngo) was calculated. In order to render |050 and ngo values directly comparable between experiments that were performed on different days or on different assay plates, |050 and ngo values were calibrated against the reference antibody infliximab.

To control precision of the se, the esponse curves were analyzed in duplicate.

Standard deviations and CVs were calculated for each measurement point (CV < 25%). 2.1.4 Target specificity To m specificity of the anti-TNFq scFvs, binding to the most homologous family member TNFB was assessed. The potential to inhibit the ction of biotinylated TNFq with scFvs by unlabeled TNFB (Peprotech, Cat. No. 300-018) and TNFq (Peprotech, Cat.

No. 300-01) was analyzed by competition ELISA. For this purpose, the scFvs were coated on a 96-well Maxisorp ELISA plate at a concentration of 1 ug/ml. Binding of biotinylated TNFq (75 ng/ml) to the coated scFvs in ce of 5-fold serially diluted unlabeled TNFq (50 ug/ml — 0.00013 ug/ml) or TNFB (1250 ug/ml — 0.00013 ug/ml) was detected using the biotinbinding streptavidin-HRP (SDT Reagents, Cat. No. SP40C) as described above. For the dose-response curve with TNFq data were ed using a four-parameter logistic curve fit using the Softmax Data Analysis Software (Molecular Devices), and the concentration of unlabeled TNFq ed to block the interaction of biotinylated TNFq with the coated scFv by 50% (IC50) was calculated. TNFB did not show any significant inhibition of the interaction between biotinylated TNFq and scFvs (see also Figure 11). To quantify the relative potential of TNFB as ed to TNFq to inhibit TNFq binding to each scFv the |C50 to inhibit the ction by TNFB relative to TNFq was calculated. Since no significant inhibition was observed when using TNFB at an approximately 5000 to 20’000-fold higher concentration than the |C50 of TNFq, the selectivity for binding to TNFq over TNFB was determined to be significantly higher than 5000 to 20’000-fold. To control precision of the response, the dose- response curves were analyzed in duplicate. Standard deviations and CVs were calculated for each measurement point (CV < 25% for all but one of the TN Foe/[3 concentrations tested).

All scFv fulfilled this criterion. 2.1 CMC Analytics 2.2.1 Reducing SDS-PAGE Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE) is an is technique used for qualitative characterization and to control purity of proteins. According to the United States Pharmacopeia (USP) (USP Chapter 1056) analytical gel ophoresis is an appropriate and routine method to identify and to assess the homogeneity of proteins in drug substances.

WO 58092 The method is used to quantify the amount of scFv product from Eco/i lysates to derive the expression yield after fermentation. Another ation of the method is to verify the identity of test substances based on their lar weight with respect to the theoretical values. For supportive purposes this method is used to quantify the purity of test samples with respect to process-related impurities (host cell proteins) and product related impurities (degradation products or adducts).

The SDS-PAGE analyses were performed with commercially available precast gel system "Mini Protean" obtained from Bio-Rad Laboratories Inc. Humanized scFvs were analyzed on "Any kD" ing gels (#456-9036). In both cases the Tris/Glycine buffer system recommended by the cturer was used. For the ion of protein bands either coomassie staining with SimplyBlueTM ng solution (Life Technologies Corp., 0) or silver staining with the Pierce Silver Stain Kit (Thermo Fisher ific Inc, #24612) was employed. For the staining procedures the protocols of the respective supplier were followed.

The documentation and analysis of the stained protein gels was performed with the documentation system ChemiDoc XRS System (Bio-Rad Laboratories Inc, #170-8265) and re Image Lab, Version 4.0.1 (Bio-Rad Laboratories Inc, # 170-9690).

Titer determination of lysate samples SDS-PAGE allows for specific detection of the protein of interest in the mixture of host cell proteins. A reference standard on series in the linear range of the method (which was determined in advance) was included on every gel. A linear regression of band intensities (measured by densitometry) versus nominal concentrations of the reference standard were used to calculate a standard curve, which in turn, was used to extrapolate scFv content in the sample.

The lysate samples of unknown product concentrations were loaded in different dilutions (at least 1:10 in dilution buffer) to have at least one scFv concentration in the linear range of the method. The product amount was calculated based on the ed band intensities of the scFv and the concentration was determined using the dilution factors of the sample preparation. The values were averaged for all samples that were within the linear range of the standard curve.

As an additional test of the suitability of the method for the fication of lysate sample an inhibition/enhancement test was performed by spiking a lysate sample with a known amount of reference standard. Calculation of the spike recovery at a sample on of 1:10 in dilution buffer resulted in a value of 95.4% which is at the same level of precision as observed with the reference rd in dilution buffer. Thus, no significant matrix interference in cell lysates was observed and the method was deemed suitable for the quantification of scFv content in cell lysates. n Purity and Content To show the ility of the method to determine the content and thereby also the purity of test samples, the lower limit of detection (LOD) for a reference scFv was determined visually (by identifying the protein band) at a nominal load of 0.02 ug, the tion of the intensity histogram of the respective lane shows a signal-to-noise ratio at this load of approximately 2.

In addition, the linear range for the quantification was determined by analyzing the main bands densitometrically.

The fit of the data with a linear regression, results in a coefficient of determination (R2) of 0.9998, thus indicating a good quality of the fit. In addition to the overall quality of the fit the ve error of each individual data point was determined to document the suitability of the method in the chosen range. The relative errors are below 10% for all data points indicating good accuracy of this method. 2.2.2 UVAbsorbance at 280 nm The method UV ance at 280 nm is a total protein assay as outlined in USP Chapter 1057. Protein solutions absorb UV light at a wavelength of 280 nm due to the presence of aromatic amino acids. The UV absorbance is a function of the content of tyrosine and tryptophan es in the protein and is proportional to the protein concentration. The absorbance of an unknown protein on can be determined according to USP Chapter 851 on spectroscopy by applying Beer’s law: A: s*l*c, where the absorbance (A) is equal to the product of the molar absorptivity (a), the absorption path length and the concentration of the nce. The molar absorptivity for the scFv was calculated with the re Vector NT|® (Life Technologies Corporation).

The measurement of the UV absorbance was performed with the ty reader M200 Pro equipped with Nanoquant plate (Tecan Group Ltd.). The absorbance of the protein samples were measured at 280 nm and 310 nm, where the latter wavelength was serving as a reference signal that was subtracted from the 280 nm signal. To account for potential interference of the sample matrix a blank subtraction was performed for each measurement.

The final absorbance signal of a protein sample obtained was used to calculate the protein concentration using Lambert-Beer’s law.

All measurements were performed within the range given by the instruments specifications in the measurement range of 0-4 OD, where a reproducibility of < 1% and a uniformity of < 3% is specified by the manufacturer. 2.2.3 SE-HPLC (Size Exclusion High-pressure Liquid tography) SE-HPLC is a separation technique based on a solid stationary phase and a liquid mobile phase as outlined by the USP r 621. This method separates molecules based on their size and shape ing a hydrophobic stationary phase and aqueous mobile phase. The separation of molecules is occurring between the void volume (V0) and the total permeation volume (VT) of a specific column. Measurements by SE-HPLC were performed on a Chromaster HPLC system (Hitachi High-Technologies Corporation) equipped with automated sample injection and a UV detector set to the detection wavelength of 280nm. The equipment is controlled by the software EZChrom Elite (Agilent Technologies, Version 3.3.2 SP2) which also supports analysis of resulting chromatograms. Protein samples were cleared by centrifugation and kept at a temperature of 6°C in the autosampler prior to injection. For the analysis of scFv samples the column Shodex KW402.5-4F (Showa Denko lnc., 201) was employed with a rdized buffered saline mobile phase (50 mM Sodium acetate pH 6.0, 250 mM sodium chloride) at the recommended flow rate of 0.35 mL/min. The target sample load per ion was 5 ug. Samples were detected by an UV detector at a ngth of 280 nm and the data recorded by a le software suite.

The resulting chromatograms were analyzed in the range of V0 to VT thereby excluding matrix associated peaks with >10 min n time.

To ensure intermediate precision of the method, a reference standard was routinely measured at the ing and end of each HPLC sequence. The nce rd used for this system suitability test was a scFv that had been produced as a batch and was aliquoted to be used for each measurement timepoint. 2.2.4 DSF (Differential Scanning Fluorimetry) The method DSF is a non-compendial method to measure temperature-dependent protein unfolding. The measurement of the thermal unfolding temperature by DSF were performed with a MX3005P qPCR machine (Agilent Technologies) controlled with the MX Pro software package (Agilent Technologies) and equipped with an excitation/emission filter set at 0 nm. The reactions were set-up in Thermo fast 96 white PCR plates (Abgene; #AB- ). For the detection of protein unfolding a commercially available stock solution of the dye SYPRO orange (Molecular Probes; # 86650) was used at a final dilution of 1:1’000. The protein s were diluted for the ing measurements to a final concentration of 50 ug/mL in a standardized buffered saline solution. The thermal unfolding was performed by a temperature program starting at 25°C ramping up to 96°C in 1°C steps with a duration of seconds. During the ature program the scence emission of each sample was recorded. The recorded raw data was processed and evaluated with a package of Microsoft Excel templates (Niesen, Nature Protocols 2007, Vol. 2 No.9) and the fluorescence data was fitted with a Boltzmann equation using the program GraphPad Prism (GraphPad Software, Inc.) to obtain the midpoint of transition (Tm).

In order to produce reliable and robust measurements of the midpoint of unfolding at least duplicate measurements were med. With respect to the data quality only measurements with a goodness of fit (R2) >0.9900 and a 95% confidence interval of the Tm of smaller than 0.5% were considered.

For an assessment of the ediate precision a reference standard (known characterized scFv) was ed with every measurement to allow for comparison of assay performance on different days. 2.2.5 Stability Study In order to assess the stability of ent scFv constructs as a read-out for the developability of these molecules a short-term stability study protocol was designed. The protein constructs were concentrated in a simple buffered saline formulation (see above) to the target concentrations of 1 and 10 mg/mL. The monomer content was determined by SE-HPLC to confirm that the purity is exceeding the success criteria of > 95%. Subsequently the n samples were stored at <-65, -20, 4 and 37°C for the duration of 4 weeks and aliquots were analyzed at various time points. The primary read-out is the analysis by SE-HPLC, which allows the quantification of soluble higher molecular weight oligomers and aggregates. As supportive measurements the protein content is determined by UV absorbance at 280 nm, which gives an indication r during the storage period substantial amounts of protein were lost by itation. For the storage screw cap tubes were used (Sarstedt, Cat. No. 72692005) with filling amounts of 30-1500 pg per aliquot. Additionally purity is determined by SDS—PAGE that indicates the stability of the construct with t to degradation or nt multimerization.

WO 58092 Example 3: Generation of humanized Diabody and lgG The single-chain diabody construct was designed by arranging the variable domains in a VLA-L1-VHB-L2-VLB-L3-VHA configuration. In these constructs the VLA and VHA and VLB and VHB domains jointly form the binding site for TNFq. The e linkers L1-L3 connecting the variable domains were constructed of glycine/serine repeats. The two short s L1 and L3 are composed of a single G48 repeat, whereas the long linker L2 is composed of the sequence (G4S)4. The tide sequences encoding the zed variable domains (Example 2; 1.2.1.) were de novo synthesized and cloned into an adapted vector for E.coli expression that is based on a pET26b(+) backbone (Novagen). The expression and purification was performed as described for the scFvs in Example 2; 1.2.1.

The humanized lgG was constructed by cloning the variable s a suitable mammalian expression vector for transient heterologous expression containing a leader sequence and the respective constant domains e.g. the pFUSE-rlgG vectors (lnvivogen). The transient expression of the functional lgG was performed by co-transfection of s encoding the heavy and light chains with the FreeStyleTM MAX system in CH0 8 cells. After cultivation for l days the supernatant of the antibody secreting cells was recovered for purification.

Subsequently the secreted lgGs were affinity purified by Protein A sepharose (GE Healthcare). The elution fractions were analyzed by SDS—PAGE, UV absorbance at 280 nm and SE-HPLC.

The affinities of the antibody les were determined using a e instrument as described in Example 2 under 2.1 .1).

Table 10 ka(M-1S-1) kd(s'1) KD (M) IgG 1.90x106 7.92x10‘5 4.17x10'11 scDb 9.40x105 0'5 2.15x10'11 scFv 8.93 x 105 3.38 x 10'5 3.79 x 10'11 The potencies of the antibody molecules were determined in an L929 assay (the method is bed in Example 2 under 2.1 .2).

Table 11 Potency |C50 (nM) lgG 0.02 scDb 0.01 scFv 0.03 Example 4: Determination of Stoichiometry of TNFa binding The binding stoichiometry of 16H5 to TNFo was determined using SE-HPLC. 16H5- scFv and TNFo were incubated at two different molar ratios, namely at a 1:1 and 4.5:1 molar ratio. Since TNFo exists as a trimer in solution the indicated molar ratios refer to the TNFommer. Thus, in the 4.5:1 ratio the 16H5-scFv is in excess and should occupy all TNdemer binding ons resulting in complexes of 1 TNFormmer with 3 scFv. r, under equimolar conditions there is not enough scFv present to saturate all 3 tical TNFo binding sites. Therefore, also complex variants with less than 3 scFv bound are expected. TNFo and 16H5-scFv were incubated for 2 hours at RT to allow for complex formation. Samples were then centrifuged at 4°C for 10 min. 10 uL of each sample were analysed on SE-HPLC. The C analysis was performed with 50 mM phosphate buffer pH 6.5, 300 mM NaCl as eluent at a flow rate of 0.35 . Eluted protein peaks were detected at a wavelength of 280 nm. The column was calibrated using the Gel filtration Calibration Kit from GE Healthcare (LMW, HMW) in advance for the determination of apparent molecular weights.

The bottom panel of Figure 12 shows the elution profile with equimolar amounts of scFv and TNFo which is overlayed with the profiles of TNFormmer alone and scFv alone. Due to the ization of the TNFo in on there are theoretically up to three equivalent binding sites for the scFv present on each trimer and hence the scFv molecules are limiting. Under these conditions all three complex species (3:1, 2:1, 1:1) were identified. The top panel of figure 12 shows the elution profile of the complex with excess amounts of scFv. The surplus of unbound scFv eluted at the expected retention time. The TNFo peak was quantitatively consumed for complex formation and disappeared completely. The peak of this complex shifted s lower retention times, and correlated well with the retention time of the peak with the largest molecular weight of the equimolar setup. For this reason it was concluded that all available binding sites on the TNFq were occupied by scFv and thus, the binding stoichiometry is 3:1 (scFv:TNFo) if the scFv is available in . r to these qualitative observations, the nt binding stoichiometry was also ated based on the apparent MW of the 16H5-scFv:TNFq complex as determined by SE-HPLC. Based on retention time, the apparent MW was calculated to be 139.7 kDa.

According to equation (1) below the apparent binding stoichiometry was calculated to be 3.3.

This correlates well with the theoretical number of three equivalent g sites available for scFv on the TNqurner and the observations above where a 3:1 binding stoichiometry was determined.

MW(complex app)— MW(TNFOC theo) Equation (1 ): binding stochiometry (scFv: TNFa) = v theo) MW (complex app): 139.7 kDa MW (TNFo theo): 52.2 kDa MW (scFv theo): 26.5 kDa Example 5: ion of ntibody complexes (cross-linking of TNFa) The ability of the 16H5-scDb to bind simultaneously to two TNFq molecules was tested on a Biacore T200 instrument in HEPES buffer containing 10 mM HEPES, 150 mM NaCl and 0.05% Tween. Biotinylated TNFq (Acro Biosystems) was captured via a biotinylated ssDNA oligo using the Biotin CAPture kit (GE Healthcare) ing to the manufacturer’s instructions. 0.25 ug/mL of biotinylated TNFo were injected with a flow rate of 10 uL/min for 3 min to reach a capture level of approximately 200 to 300 RUs (resonance units). The antibodies 16H5-scDb and 16H5-scFv, as a control, were injected over the TNFq immobilized surface for 2 min at a flow rate of 30 uL/min at a concentration of 90 nM. ing association of the antibody fragments, TNFo (Peprotech) was injected for 5 min with a flow rate of 30 uL/min at 90 nM. The antibody and TNFq concentrations were ed close to saturation of the binding. The measurement was performed at 25°C. Figure 13 illustrates that the bivalent 16H5-scDb is able to bind two TNFq molecules simultaneously while, as expected, the monovalent 16H5-scFv binds only to one TNFq molecule.

Further, the formation of TN For-antibody complexes was assessed at different ratios of TN For and 16H5 antibody formats using SE-HPLC. H5-lgG (150 kDa) and 16H5- scDb (52 kDa) were incubated with TNFo (52 kDa) at different molar ratios (1:3, 1:1, 3:1) in respect to binding sites. Thus, lgG and scDb have 2 and TNFo has 3 binding sites. The antibody-TNch mixtures were incubated for at least 30 min at 37°C, cooled down for 10 min at RT and stored overnight at 2 — 8°C. Five to 10 uL of protein mixture at a tration of approx. 1 mg/mL were ed onto a TOSHO TSngl UP-SW3000 column. The analysis was performed with 150 mM phosphate buffer pH 6.8, 100 mM NaCl as eluent at a flow rate of 0.3 mL/min. Eluted protein peaks were detected at a wavelength of 214 nm. The column was calibrated using the BEH450 SEC protein standard mix (Waters) in advance for the determination of the approximate molecular s of the complexes. Figure 14A shows the formation of 16H5-lgG:TNch complexes. Complexes that are 2 600 kDa indicate the formation of complexes consisting of 2 2 TNFo and 2 3 lgG molecules. Figure 148 shows the formation of 16H5-scDb:TNFd complexes. Complexes that are 2 300 kDa indicate the formation of complexes consisting of 2 2 TNFo and 2 3 scDb molecules.

Example 6: tion of cell proliferation The capacity of ent antibody formats of 16H5 and adalimumab to inhibit the proliferation of peripheral blood mononuclear cells (PBMC) was tested in a mixed lymphocyte reaction (MLR). PBMC from 2 healthy donors were cultured (RPMI1640) in a 1:1 ratio in 96- well plates for 48 h at 37°C/5 % COZ. After activation, cells were treated with anti-TNFo antibodies or lgG control antibody (all at a final concentration of 10 ug/mL) in sextuplicates for another 5 d at 37°C/5 % 002. 24 h before the end of incubation BrdU (20 l) was added to each well and proliferation was determined by measuring BrdU uptake using a commercially ble cell proliferation ELISA (Roche Diagnostics). The stimulation index was determined by calculating the ratio of BrdU uptake n the antibody treated cells and mitomycin C (25 ng/mL) treated cells. Table 12 and Figure 15 illustrate that all tested antibody s of 16H5 significantly inhibited T-cell proliferation comparable to adalimumab.

Table 12 concentration Stimulation Index ) mean SD IgG control Adalimumab 16H5-IgG 16H5-scDb 16H5-scFv *p < 0.05; **p < 0.01 e 7: Inhibition of LPS-induced Cytokine Secretion CD14+ monocytes in RPM|1640 were seeded in 96-well plates and incubated for 16 h at 37°C/5 % C02 in a humidified incubator. Then cells were treated with anti-TNch antibodies or lgG control antibody in duplicates for 1 h using final antibody concentrations ranging from 2 to 2000 ng/mL. The monocytes were washed 3times with cell e medium and subsequently incubated with LPS (100 ng/mL) for 4 h at 37°C/5% C02. |L-1B and TNFd concentrations in the cell culture supernatants were determined using commercially available ELISA kits (R&D Systems). The Results are shown in Tables 13 and 14 and Figures 16A and B. |C50 was determined using a four-parameter logistic curve fit. Regarding secretion of |L-1B the IC50 values for 16H5-IgG, 16-22—H5-scDb, 16H5-scFv and adalimumab is ized in Table 13 below.

Table 13. Secretion of |L-1B IC50 ) Adalimumab 121.3 16H5-IgG 81.55 16H5-scDb 30.51 16H5-scFv 16.98 As regards TNFd secretion, the ined |C50 values for 16H5-lgG, 16-22—H5-scDb, 16H5-scFv are summarized in Table 14.

Table 14. Secretion of TNFq ICso (pg/mL) Adalimumab 174.0 16H5-IgG 120.5 16H5-scDb 17.18 16H5-scFv 13.48 Table 15. VK1 consensus ces (rearranged) Positions according SEQ ID NO: Sequence to Kabat: Framework I 1 to 23 56 SPSSLSASVGDRVTITC Framework || 35 to 49 57 WYQQKPGKAPKLLIY Framework III 57 to 88 58 GVPSRFSGSGSGTDFTLTISSLQPEDFATYYC Table 16: Vk germline-based framework IV sequences SEQ ID NO: Sequence 59 FGTGTKVTVL 50 FGGGTKLTVL 51 FGGGTQLIIL 62 FGSGTKVTVL Claims An antibody or a functional fragment thereof capable of binding to human tumor necrosis factor alpha (TNFq), wherein said antibody or functional fragment comprises (i) a VL domain comprising a CDR1 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:1, a CDR2 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID N02, and a CDR3 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:3, and (ii) a VH domain comprising a CDR1 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:4, a CDR2 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:5, and a CDR3 region having an amino acid sequence in accordance with the amino acid sequence as shown in SEQ ID NO:6.

The antibody or functional fragment of claim 1, n said antibody or onal fragment comprises (i) a VL domain comprising a CDR1 region having the amino acid sequence as shown in SEQ ID NO:7, a CDR2 region having the amino acid ce as shown in SEQ ID NO:8, and a CDR3 region having the amino acid sequence as shown in SEQ ID NO:9, and (ii) a VH domain comprising a CDR1 region having the amino acid sequence as shown in SEQ ID NO:10, a CDR2 region having the amino acid sequence as shown in SEQ ID NO:11, and a CDR3 region having the amino acid ce as shown in SEQ ID NO:12.

The antibody or functional fragment of claim 1 or 2, wherein said antibody or functional nt (i) binds to human TNFq with a dissociation constant (KB) of less than 125 pM; (ii) is reactive with Macaca mulatta (Rhesus) TNFq and with Macaca fascicu/aris (Cynomolgus) TN For; (iii) has a greater y to inhibit TNFq-induced apoptosis than infliximab, as determined by an L929 assay; (iv) comprises a variable domain having a melting temperature, determined by differential scanning fluorimetry, of at least 70°C and/or (v) is capable of binding to human TNFqTrimer in a stoichiometry (antibody : TN FqTrimer) of at least 2.

The antibody or functional fragment of any one of the preceding claims, which binds to human TNFq with a KB of less than 50 pM.

The antibody or functional nt of any one of the preceding claims, wherein said antibody or functional fragment comprises a VH domain having the amino acid sequence as shown in SEQ ID NO:13.

The antibody or functional fragment of any one of the preceding claims, wherein said antibody or functional fragment comprises a VL domain having an amino acid ce selected from SEQ ID NO:14 and SEQ ID NO:54.

The functional fragment of any one of the preceding claims, which is a single-chain variable nt (scFv).

The functional fragment of claim 7, wherein said scFv has the amino acid sequence as shown in SEQ ID NO:15 or SEQ ID NO:55.

The antibody of any one of claims 1 to 6, which is an immunoglobulin G (IgG).

. An antibody or functional fragment thereof binding to essentially the same epitope as the functional fragment of claim 8. 11. The antibody or functional fragment of any one of the preceding claims, wherein the sum of (i) the number of amino acids in framework s I to III of the le light domain of said antibody or functional fragment that are different from the tive human VK1 consensus sequences with SEQ ID NOs: 56 to 58, and (ii) the number of amino acids in framework region IV of the variable light domain of said antibody or functional fragment that are different from the most similar human A germline-based sequence ed from SEQ ID NOs: 59 to 62, is less than 7, preferably less than 4. 12. The antibody or functional fragment of any one of the preceding claims, wherein the ork regions I to III of the variable light domain of said antibody or functional nt consist of human VK1 consensus sequences with SEQ ID NOs:56 to 58, tively, and framework region IV consists of a A ne-based sequence selected from SEQ ID NOs:59 to 62. 13. A nucleic acid encoding the antibody or functional fragment of any one of the preceding claims.

A vector or plasmid comprising the c acid of claim 13.

. A cell comprising the nucleic acid of claim 13 or the vector or plasmid of claim 14. 16. A method of preparing the antibody or functional fragment of any one of claims 1 to 12, comprising culturing the cell of claim 15 in a medium under conditions that allow expression of the nucleic acid encoding the antibody or functional fragment, and recovering the antibody or functional fragment from the cells or from the medium. 17. A pharmaceutical composition comprising the antibody or functional fragment of any one of claims 1 to 12, and optionally a pharmaceutically acceptable carrier and/or 18. The antibody or functional fragment as defined in any one of claims 1 to 12 for use in a method of treating an inflammatory disorder or a TNch-related disorder. 19. The dy or functional fragment for use ing to claim 18, wherein said matory disorder is an inflammatory disorder of the gastrointestinal tract.

. The antibody or functional fragment for use according to claim 19, wherein said inflammatory disorder of the gastrointestinal tract is inflammatory bowel disease. 21. The antibody or functional fragment for use according to claim 19 or 20, wherein said inflammatory disorder of the gastrointestinal tract is Crohn’s disease or tive colitis. 1/1 2 Figure 1 a ffamawmk rabbi? W rabbit mm whiff: mm; 2/1 2 Figure 2 — 021mm O 1 2 3 4 5 6 7 8 9 1O 11 ‘32 13 14 15 Minutes (EV-"Ll Regain Pk ?¢ Name Retemmn I‘m}? mm 33010031" Area 1 1936 (3,6! 63% 2 £13000:sz 5?sz @0338 €39.31}? :3 NEW T0003 100,00 5 3.483022% 1 6-22—H05—8002 — wwwm mAU 4o 0‘ '1' '2'3 '4 "5' "a 7 ‘8‘ 9 "10 11 12 13 14 15 {313111 Rfialt‘s F1; a Name Rah—3mm: Time A103 Pfii‘fiféflt 33000 1 gamma my 0023 0.03 4303 3 maximums: 30R" 8&6? 9993 513490 Tataia 20000 5310900 3/1 2 lntensity .0co.0 \4 Fluorescence .0m0 U1 wiG-ZZ—HOS—SCOZ P4; VVVVVVVVVVVVVVVVVVVVVVVVVVVVV ~e—16H05-sc02 Normalized 53w.0N 85 90 95 .0 to Intensity .0oo.0 \l scence .0m0 U1 le-ZZ—HOS—scflll .04:. +16H05-SCO4 Normalized .0w.0N Temperature [‘C] Figure 4 caz i 1‘30 m a A a X g 333$} :93 , 1 Ire]. cement manomer Bl?’ a £25 «20%; «86‘1" 16~22~H35~sc94 209 a a; g C: [is Q: %] mas; {:3 mama: ma: monamer mm: ii} (If: {3:403 3:13 "28°C W35:30°(: /1 2 Figure 5 1 62245662 90; 30 f 25 sex 29 - 15 ?0 ’fi , 0.5 I 0x) 80? 415 50f 45 f .29 40-; ~25 mAU 3m 3 2 34 58 78 8161f‘i2131415 .4; Q gwmwwwwWMWMAWW.‘WWWWWWMMWWWV,WWWWWMWW. Mum—WWWMMWWWWWWWWWW. t=28£§;4"C ‘25 f ’- ............................................. .. .. _ _ ., ........................................ ffeeze/thaw{5 CYCEES) 3622449558134 992' 3.0 E 2.5 80: 20 i 7,5 7‘6 m } 05 68- 0.0"" , «)5 53f 433w" 401 .23; mAU ~25 , "3'80 1 2 34 5678 9102112131415 ,mwmwmwMMMWWMWWWWMMMMMWWWMM__W, t=28fi;4"C freeze/thaw {5 cycies} Figure 6 aways: at 37%: :2 2:; M a» w: 90 g 85 E 83 E 735 a. 16~2Z«H{35~$502 a} H9§£€0$ 9 5 10 15 2:3 iime {a} Figure 7 16H05-sc02 16H05-sc04 1 00 1 00 g § 0 0 ’5, 5a} 3: 5° ‘9‘ infliximab '3' infliximab + 16~22-H05'SCG2 ‘F' 18.22»H05-5c04 9.01 0.1 1 10 100 1000 10900 0-01 0-1 1 10 190 100%) 10000 cone. [lag/ml} conc. [ngimn Figure 8 16H05-3602 16H05~SCO4 100 196 z z E -E' humanTNF E -e- human TNF E 4- cynomokgus TNF O * cynomoigusTNF + rhesusTNF 3 _-_ rhesus TNF :3 g {3 0 9.01 0.1 1 18 100 14300 10000 0.01 0,1 1 10 1GB 1000 143000 nglml] conc.{ngim3} Flgure9 16H85-5602 16H05-sc04 TNFRI 190 ..9 - 10%} §\ -e~ imab TNFRI §\‘ -9 - infiixim ab $6-22—H95—sc02 to to ofTNF 0' @ TNF Q" G Binding Binding % "/0 cone. [ngimi] conc.[ng1m!] WO 58092 8/1 2 Figure 10 16H05-sc02 1SH05-sc04 TNFRH 108 100 .6.5‘ -9. mfliximab TNFRH '9‘ {nfiixémab ! + 16—224-{05-5002 + 16~22~H05—SCO4 to ‘ to ofTNF l 5o | ofTNF 50 Binding | 0 lllllllllllllllllllllllllllllllllllllllllllll ............................................. Binding a "/0 % x x a Q} <3 § a Q N f\ k 6 Q ‘5 (I) a a" N v " u" ,3" ’39" @253 Q," Q "Q 3% KS") "n"? " " conc‘ [nglmf] conc. {nglml} Figure 11 ’i S-ZZ-H 35-8602 0450~£50 conc. [pg/ml] 9/1 2 Figure 12 4.5:1 ratio scFv to TNFOLWmer 1:1 ratio scFv to rim mAU 10.0 2.5 ; 0.0 3 6.0 6.5 ?'.0 7.5 8.0 8.5 9.0 9.5 10.0 12.0 Minutes /12 Figure 13 pmtein iniectlan TNFa injectinn - 22-H5-SCDb $13!] WmmmmwmmWMJ-Wh capturéfiknwl} ca3 nume- 16- 22-H5-SCFV Jim , _ 1——"—"——1’*‘———l mu WEI 15m 1290 ME"?! TEDEI 13m: 2051] TIME- E Figure 14A 476 m x—u—n 450- "A" W*_—-—i ————— lgG TNFX. lgGv complexes. A 2 IgG 400‘ lgG:TNF = 1:3 WWW-m4 . __ . ____________ > 600 kDa lgG.TNF _ 1_1 350 ‘ ‘ lgG:TNF = 3:1 300 ‘ 250 ‘ 200 ‘ .24 7" 215 3.0 410 50 6.0 0 8,0 O 1210 1310 #40 15,0 16.0 1710 17,4 [2? (:31? ‘17 £27 «9 a: £33 <8" ‘3? g?» (a "’3 1 1/12 Figure 14B 3181 mAu300 4—! TNF wwwwww SCDb 250} TNFX: scDby complexes r scDthNF = 1:3 ‘ i > 300 kDa scDszNF = 1:1 ., , scDb:TNF = 3:1 200 1 180 A SCDb 160 3; 140 1 1 : g 120 , 1‘ 1 2 1 mo . g t: 80 1' i" {a 60 5 a. £2 40 . l E 3 1 l 1 I , 2o 5‘ \fi 1 3 1 «17 .. . _ L, . . "‘"T 235 VVV 3.0 ""476 5.0 Fe w7.6 Eer— 9.0 10.0 We ‘ "32:0 13.0 We 15b 16.0 17.51714 flfifl? lb (g? ‘3‘? 58m £5) £9 £9 (an .22? {33‘ gr: Figure 15 ** ** * ** .E 4 .5 m 66° 9 (9 ‘0 4 (5° X?" <9 s8 0 ‘0 f9 , o°° 6" a? $’ 90’ \Qo v55 (05" 50’ N bit" Nb '\, *: p < 0.05 **: p < 0.01 1 2/1 2 Figure 16A +|gG control -E-Ada|imumab (pg/mL) +16—22-H5-IgG IL-1B +16H5-scDb +16H5-SCFV 1 10 100 1000 10000 antibody ) Figure 16B 1000 +IgG control -E-Ada|imumab E 800 - En +16H5-IgG ‘6’ 600 +16H5-scDb +16H5-SCFV 1 10 100 1000 10000 antibody (ng/mL) SEQUENCE LISTING <110> ts Pharma AG <120> Anti-TNF-alpha antibodies and functional fragments thereof <130> PWO00278TIL <150> EP16160907.8 <151> 201617 <160> 62 <170> PatentIn version 3.5 <210> 1 <211> 11 <212> PRT <213> Artificial Sequence <220> <223> CDR <220> <221> MISC_FEATURE <222> (7)..(7) <223> Xaa is Phe, Ser, Gly or Tyr <220> <221> MISC_FEATURE <222> 8) <223> Xaa is Ser, Asn, Thr or Arg <220> <221> MISC_FEATURE <222> (9)..(9) <223> Xaa is Gly, Tyr, Ala, Asn or Ser <400> 1 Gln Ala Ser Gln Ser Ile Xaa Xaa Xaa Leu Ala 1 5 10 <210> 2 <211> 7 <212> PRT <213> Artificial Sequence <220> <223> CDR Page 1 <220> <221> MISC_FEATURE <222> (1)..(1) <223> Xaa is Gly, Arg or Gln <220> <221> MISC_FEATURE <222> (4)..(4) <223> Xaa is Lys or Thr <220> <221> MISC_FEATURE <222> (6)..(6) <223> Xaa is Ala or Glu <400> 2 Xaa Ala Ser Xaa Leu Xaa Ser 1 5 <210> 3 <211> 15 <212> PRT <213> Artificial Sequence <220> <223> CDR <220> <221> EATURE <222> (6)..(6) <223> Xaa is Ser or Thr <220> <221> MISC_FEATURE <222> (8)..(8) <223> Xaa is Ser or Ile <220> <221> MISC_FEATURE <222> (9)..(9) <223> Xaa is Ser, Thr or Asn <220> <221> MISC_FEATURE <222> (13)..(13) <223> Xaa is Ser or Phe <220> <221> MISC_FEATURE Page 2 <222> (14) <223> Xaa is Tyr, Phe, Leu or Val <400> 3 Gln Ser Tyr Tyr Tyr Xaa Ser Xaa Xaa Ser Asp Gly Xaa Xaa Ala 1 5 10 15 <210> 4 <211> 10 <212> PRT <213> Artificial Sequence <220> <223> CDR <220> <221> MISC_FEATURE <222> (5)..(5) <223> Xaa is Asn or Ser <220> <221> MISC_FEATURE <222> (6)..(6) <223> Xaa is Asn or Thr <220> <221> MISC_FEATURE <222> (10)..(10) <223> Xaa is Gly, Cys or Ser <400> 4 Gly Ile Asp Phe Xaa Xaa Tyr Gly Ile Xaa 1 5 10 <210> 5 <211> 17 <212> PRT <213> Artificial Sequence <220> <223> CDR <220> <221> MISC_FEATURE <222> (7)..(7) <223> Xaa is Ala or Gly Page 3 <220> <221> MISC_FEATURE <222> (9)..(9) <223> Xaa is Thr or Arg <220> <221> MISC_FEATURE <222> (11)..(11) <223> Xaa is Phe or Tyr <220> <221> MISC_FEATURE <222> (13)..(13) <223> Xaa is Asn or His <220> <221> MISC_FEATURE <222> (15)..(15) <223> Xaa is Val or Leu <400> 5 Tyr Ile Tyr Pro Gly Phe Xaa Ile Xaa Asn Xaa Ala Xaa Ser Xaa Lys 1 5 10 15 <210> 6 <211> 13 <212> PRT <213> Artificial ce <220> <223> CDR <220> <221> MISC_FEATURE <222> (3)..(3) <223> Xaa is Val, Ile or Leu <220> <221> MISC_FEATURE <222> (5)..(5) <223> Xaa is Ala or Ser <220> <221> MISC_FEATURE <222> (6)..(6) <223> Xaa is Thr or Ser Page 4 <220> <221> MISC_FEATURE <222> (7)..(7) <223> Xaa is Asp or Ser <220> <221> EATURE <222> (8)..(8) <223> Xaa is Trp or Ser <220> <221> MISC_FEATURE <222> (9)..(9) <223> Xaa is Gly or Ala <220> <221> MISC_FEATURE <222> (11)..(11) <223> Xaa is Phe, Leu or Tyr <220> <221> MISC_FEATURE <222> (12)..(12) <223> Xaa is Asp or Asn <400> 6 Asp Pro Xaa Tyr Xaa Xaa Xaa Xaa Xaa Tyr Xaa Xaa Leu 1 5 10 <210> 7 <211> 11 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 7 Gln Ala Ser Gln Ser Ile Phe Ser Gly Leu Ala 1 5 10 <210> 8 <211> 7 <212> PRT <213> Artificial Sequence <220> <223> CDR Page 5 <400> 8 Gly Ala Ser Lys Leu Ala Ser 1 5 <210> 9 <211> 15 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 9 Gln Ser Tyr Tyr Tyr Ser Ser Ser Ser Ser Asp Gly Ser Tyr Ala 1 5 10 15 <210> 10 <211> 10 <212> PRT <213> cial Sequence <220> <223> CDR <400> 10 Gly Ile Asp Phe Asn Asn Tyr Gly Ile Gly 1 5 10 <210> 11 <211> 17 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 11 Tyr Ile Tyr Pro Gly Phe Ala Ile Thr Asn Phe Ala Asn Ser Val Lys 1 5 10 15 Page 6 <210> 12 <211> 13 <212> PRT <213> cial Sequence <220> <223> CDR <400> 12 Asp Pro Val Tyr Ala Thr Ser Ser Gly Tyr Phe Asp Leu 1 5 10 <210> 13 <211> 122 <212> PRT <213> Artificial Sequence <220> <223> humanized VH domain of clone 16H05-sc02 and clone 16H05-sc04 <400> 13 Glu Val Gln Leu Val Glu Ser Gly Gly Gly Leu Val Gln Pro Gly Gly 1 5 10 15 Ser Leu Arg Leu Ser Cys Lys Ala Ser Gly Ile Asp Phe Asn Asn Tyr 25 30 Gly Ile Gly Trp Val Arg Gln Ala Pro Gly Lys Gly Leu Glu Trp Ile 40 45 Thr Tyr Ile Tyr Pro Gly Phe Ala Ile Thr Asn Phe Ala Asn Ser Val 50 55 60 Lys Gly Arg Phe Thr Ile Ser Ser Asp Asn Ser Lys Asn Thr Val Tyr 65 70 75 80 Leu Gln Met Asn Ser Leu Arg Ala Glu Asp Thr Ala Val Tyr Tyr Cys 85 90 95 Ala Arg Asp Pro Val Tyr Ala Thr Ser Ser Gly Tyr Phe Asp Leu Trp 100 105 110 Page 7 Gly Gln Gly Thr Leu Val Thr Val Ser Ser 115 120 <210> 14 <211> 114 <212> PRT <213> cial Sequence <220> <223> humanized VL domain of clone 16H05-sc02 <400> 14 Asp Ile Gln Met Thr Gln Ser Pro Ser Ser Leu Ser Ala Ser Val Gly 1 5 10 15 Asp Arg Val Thr Ile Lys Cys Gln Ala Ser Gln Ser Ile Phe Ser Gly 25 30 Leu Ala Trp Tyr Gln Gln Lys Pro Gly Lys Ala Pro Lys Leu Leu Ile 40 45 Tyr Gly Ala Ser Lys Leu Ala Ser Gly Val Pro Ser Arg Phe Ser Gly 50 55 60 Ser Gly Ser Gly Thr Glu Phe Thr Leu Thr Ile Ser Gly Leu Gln Pro 65 70 75 80 Ala Asp Phe Ala Thr Tyr Tyr Cys Gln Ser Tyr Tyr Tyr Ser Ser Ser 85 90 95 Ser Ser Asp Gly Ser Tyr Ala Phe Gly Gly Gly Thr Lys Leu Thr Val 100 105 110 Leu Gly <210> 15 <211> 257 <212> PRT <213> Artificial Sequence <220> <223> humanized scFv of clone 16H05-sc02 Page 8 <400> 15 Met Asp Ile Gln Met Thr Gln Ser Pro Ser Ser Leu Ser Ala Ser Val 1 5 10 15 Gly Asp Arg Val Thr Ile Lys Cys Gln Ala Ser Gln Ser Ile Phe Ser 25 30 Gly Leu Ala Trp Tyr Gln Gln Lys Pro Gly Lys Ala Pro Lys Leu Leu 40 45 Ile Tyr Gly Ala Ser Lys Leu Ala Ser Gly Val Pro Ser Arg Phe Ser 50 55 60 Gly Ser Gly Ser Gly Thr Glu Phe Thr Leu Thr Ile Ser Gly Leu Gln 65 70 75 80 Pro Ala Asp Phe Ala Thr Tyr Tyr Cys Gln Ser Tyr Tyr Tyr Ser Ser 85 90 95 Ser Ser Ser Asp Gly Ser Tyr Ala Phe Gly Gly Gly Thr Lys Leu Thr 100 105 110 Val Leu Gly Gly Gly Gly Gly Ser Gly Gly Gly Gly Ser Gly Gly Gly 115 120 125 Gly Ser Gly Gly Gly Gly Ser Glu Val Gln Leu Val Glu Ser Gly Gly 130 135 140 Gly Leu Val Gln Pro Gly Gly Ser Leu Arg Leu Ser Cys Lys Ala Ser 145 150 155 160 Gly Ile Asp Phe Asn Asn Tyr Gly Ile Gly Trp Val Arg Gln Ala Pro 165 170 175 Gly Lys Gly Leu Glu Trp Ile Thr Tyr Ile Tyr Pro Gly Phe Ala Ile 180 185 190 Thr Asn Phe Ala Asn Ser Val Lys Gly Arg Phe Thr Ile Ser Ser Asp 195 200 205 Page 9 Asn Ser Lys Asn Thr Val Tyr Leu Gln Met Asn Ser Leu Arg Ala Glu 210 215 220 Asp Thr Ala Val Tyr Tyr Cys Ala Arg Asp Pro Val Tyr Ala Thr Ser 225 230 235 240 Ser Gly Tyr Phe Asp Leu Trp Gly Gln Gly Thr Leu Val Thr Val Ser 245 250 255 <210> 16 <211> 11 <212> PRT <213> cial Sequence <220> <223> CDR <400> 16 Gln Ala Ser Gln Ser Ile Ser Asn Tyr Leu Ala 1 5 10 <210> 17 <211> 11 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 17 Gln Ala Ser Gln Ser Ile Ser Thr Ala Leu Ala 1 5 10 <210> 18 <211> 11 <212> PRT <213> Artificial Sequence <220> <223> CDR Page 10 <400> 18 Gln Ala Ser Gln Ser Ile Gly Arg Asn Leu Ala 1 5 10 <210> 19 <211> 11 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 19 Gln Ala Ser Gln Ser Ile Ser Asn Ser Leu Ala 1 5 10 <210> 20 <211> 11 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 20 Gln Ala Ser Gln Ser Ile Tyr Ser Gly Leu Ala 1 5 10 <210> 21 <211> 11 <212> PRT <213> Artificial ce <220> <223> CDR <400> 21 Gln Ala Ser Gln Ser Ile Gly Ser Asn Leu Ala 1 5 10 <210> 22 <211> 11 <212> PRT <213> Artificial Sequence Page 11 <220> <223> CDR <400> 22 Gln Ala Ser Gln Ser Ile Ser Ser Ser Leu Ala 1 5 10 <210> 23 <211> 7 <212> PRT <213> Artificial ce <220> <223> CDR <400> 23 Arg Ala Ser Thr Leu Ala Ser 1 5 <210> 24 <211> 7 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 24 Arg Ala Ser Thr Leu Glu Ser 1 5 <210> 25 <211> 7 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 25 Gln Ala Ser Lys Leu Ala Ser 1 5 <210> 26 Page 12 <211> 7 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 26 Arg Ala Ser Lys Leu Ala Ser 1 5 <210> 27 <211> 15 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 27 Gln Ser Tyr Tyr Tyr Ser Ser Ser Ser Ser Asp Gly Phe Phe Ala 1 5 10 15 <210> 28 <211> 15 <212> PRT <213> cial Sequence <220> <223> CDR <400> 28 Gln Ser Tyr Tyr Tyr Ser Ser Ser Ser Ser Asp Gly Ser Phe Ala 1 5 10 15 <210> 29 <211> 15 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 29 Gln Ser Tyr Tyr Tyr Ser Ser Ser Asn Ser Asp Gly Ser Leu Ala 1 5 10 15 Page 13 <210> 30 <211> 15 <212> PRT <213> Artificial ce <220> <223> CDR <400> 30 Gln Ser Tyr Tyr Tyr Ser Ser Ile Ser Ser Asp Gly Ser Tyr Ala 1 5 10 15 <210> 31 <211> 15 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 31 Gln Ser Tyr Tyr Tyr Ser Ser Ser Ser Ser Asp Gly Ser Val Ala 1 5 10 15 <210> 32 <211> 15 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 32 Gln Ser Tyr Tyr Tyr Thr Ser Ser Thr Ser Asp Gly Ser Tyr Ala 1 5 10 15 <210> 33 <211> 10 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 33 Page 14 Gly Ile Asp Phe Ser Asn Tyr Gly Ile Cys 1 5 10 <210> 34 <211> 10 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 34 Gly Ile Asp Phe Ser Asn Tyr Gly Ile Ser 1 5 10 <210> 35 <211> 10 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 35 Gly Ile Asp Phe Ser Thr Tyr Gly Ile Ser 1 5 10 <210> 36 <211> 10 <212> PRT <213> Artificial ce <220> <223> CDR <400> 36 Gly Ile Asp Phe Ser Asn Tyr Gly Ile Gly 1 5 10 <210> 37 <211> 17 <212> PRT <213> Artificial Sequence <220> Page 15 <223> CDR <400> 37 Tyr Ile Tyr Pro Gly Phe Gly Ile Thr Asn Tyr Ala Asn Ser Val Lys 1 5 10 15 <210> 38 <211> 17 <212> PRT <213> Artificial ce <220> <223> CDR <400> 38 Tyr Ile Tyr Pro Gly Phe Gly Ile Arg Asn Tyr Ala His Ser Val Lys 1 5 10 15 <210> 39 <211> 17 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 39 Tyr Ile Tyr Pro Gly Phe Gly Ile Arg Asn Tyr Ala Asn Ser Leu Lys 1 5 10 15 <210> 40 <211> 17 <212> PRT <213> Artificial Sequence Page 16 <220> <223> CDR <400> 40 Tyr Ile Tyr Pro Gly Phe Ala Ile Arg Asn Tyr Ala Asn Ser Val Lys 1 5 10 15 <210> 41 <211> 17 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 41 Tyr Ile Tyr Pro Gly Phe Gly Ile Thr Asn Phe Ala Asn Ser Val Lys 1 5 10 15 <210> 42 <211> 13 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 42 Asp Pro Ile Tyr Ala Ser Ser Ser Gly Tyr Leu Asp Leu 1 5 10 <210> 43 <211> 13 <212> PRT <213> cial Sequence <220> <223> CDR Page 17 <400> 43 Asp Pro Val Tyr Ser Ser Asp Trp Gly Tyr Phe Asn Leu 1 5 10 <210> 44 <211> 13 <212> PRT <213> Artificial ce <220> <223> CDR <400> 44 Asp Pro Val Tyr Ala Ser Ser Ser Gly Tyr Leu Asp Leu 1 5 10 <210> 45 <211> 13 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 45 Asp Pro Leu Tyr Ala Thr Ser Ser Gly Tyr Phe Asp Leu 1 5 10 <210> 46 <211> 13 <212> PRT <213> Artificial Sequence <220> <223> CDR <400> 46 Asp Pro Val Tyr Ala Ser Ser Ser Gly Tyr Phe Asp Leu 1 5 10 <210> 47 <211> 13 <212> PRT <213> Artificial Sequence Page 18 <220> <223> CDR <400> 47 Asp Pro Val Tyr Ala Ser Ser Ser Ala Tyr Tyr Asn Leu 1 5 10 <210> 48 <211> 13 <212> PRT <213> Artificial ce <220> <223> CDR <400> 48 Asp Pro Leu Tyr Ser Thr Ser Ser Gly Tyr Phe Asn Leu 1 5 10 <210> 49 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> linker sequence in scFv <400> 49 Gly Gly Gly Gly Ser Gly Gly Gly Gly Ser Gly Gly Gly Gly Ser Gly 1 5 10 15 Gly Gly Gly Ser <210> 50 <211> 5 <212> PRT <213> Artificial Sequence <220> <223> linker sequence in diabody <400> 50 Gly Gly Gly Gly Ser 1 5 Page 19 <210> 51 <211> 503 <212> PRT <213> Artificial Sequence <220> <223> zed diabody of clone 16H05 <400> 51 Met Asp Ile Gln Met Thr Gln Ser Pro Ser Ser Leu Ser Ala Ser Val 1 5 10 15 Gly Asp Arg Val Thr Ile Lys Cys Gln Ala Ser Gln Ser Ile Phe Ser 25 30 Gly Leu Ala Trp Tyr Gln Gln Lys Pro Gly Lys Ala Pro Lys Leu Leu 40 45 Ile Tyr Gly Ala Ser Lys Leu Ala Ser Gly Val Pro Ser Arg Phe Ser 50 55 60 Gly Ser Gly Ser Gly Thr Glu Phe Thr Leu Thr Ile Ser Gly Leu Gln 65 70 75 80 Pro Ala Asp Phe Ala Thr Tyr Tyr Cys Gln Ser Tyr Tyr Tyr Ser Ser 85 90 95 Ser Ser Ser Asp Gly Ser Tyr Ala Phe Gly Gly Gly Thr Lys Leu Thr 100 105 110 Val Leu Gly Gly Gly Gly Gly Ser Glu Val Gln Leu Val Glu Ser Gly 115 120 125 Gly Gly Leu Val Gln Pro Gly Gly Ser Leu Arg Leu Ser Cys Lys Ala 130 135 140 Ser Gly Ile Asp Phe Asn Asn Tyr Gly Ile Gly Trp Val Arg Gln Ala 145 150 155 160 Pro Gly Lys Gly Leu Glu Trp Ile Thr Tyr Ile Tyr Pro Gly Phe Ala Page 20 165 170 175 Ile Thr Asn Phe Ala Asn Ser Val Lys Gly Arg Phe Thr Ile Ser Ser 180 185 190 Asp Asn Ser Lys Asn Thr Val Tyr Leu Gln Met Asn Ser Leu Arg Ala 195 200 205 Glu Asp Thr Ala Val Tyr Tyr Cys Ala Arg Asp Pro Val Tyr Ala Thr 210 215 220 Ser Ser Gly Tyr Phe Asp Leu Trp Gly Gln Gly Thr Leu Val Thr Val 225 230 235 240 Ser Ser Gly Gly Gly Gly Ser Gly Gly Gly Gly Ser Gly Gly Gly Gly 245 250 255 Ser Gly Gly Gly Gly Ser Asp Ile Gln Met Thr Gln Ser Pro Ser Ser 260 265 270 Leu Ser Ala Ser Val Gly Asp Arg Val Thr Ile Lys Cys Gln Ala Ser 275 280 285 Gln Ser Ile Phe Ser Gly Leu Ala Trp Tyr Gln Gln Lys Pro Gly Lys 290 295 300 Ala Pro Lys Leu Leu Ile Tyr Gly Ala Ser Lys Leu Ala Ser Gly Val 305 310 315 320 Pro Ser Arg Phe Ser Gly Ser Gly Ser Gly Thr Glu Phe Thr Leu Thr 325 330 335 Ile Ser Gly Leu Gln Pro Ala Asp Phe Ala Thr Tyr Tyr Cys Gln Ser 340 345 350 Tyr Tyr Tyr Ser Ser Ser Ser Ser Asp Gly Ser Tyr Ala Phe Gly Gly 355 360 365 Gly Thr Lys Leu Thr Val Leu Gly Gly Gly Gly Gly Ser Glu Val Gln 370 375 380 Page 21 Leu Val Glu Ser Gly Gly Gly Leu Val Gln Pro Gly Gly Ser Leu Arg 385 390 395 400 Leu Ser Cys Lys Ala Ser Gly Ile Asp Phe Asn Asn Tyr Gly Ile Gly 405 410 415 Trp Val Arg Gln Ala Pro Gly Lys Gly Leu Glu Trp Ile Thr Tyr Ile 420 425 430 Tyr Pro Gly Phe Ala Ile Thr Asn Phe Ala Asn Ser Val Lys Gly Arg 435 440 445 Phe Thr Ile Ser Ser Asp Asn Ser Lys Asn Thr Val Tyr Leu Gln Met 450 455 460 Asn Ser Leu Arg Ala Glu Asp Thr Ala Val Tyr Tyr Cys Ala Arg Asp 465 470 475 480 Pro Val Tyr Ala Thr Ser Ser Gly Tyr Phe Asp Leu Trp Gly Gln Gly 485 490 495 Thr Leu Val Thr Val Ser Ser <210> 52 <211> 221 <212> PRT <213> Artificial Sequence <220> <223> light chain of zed IgG of clone 16H05 <400> 52 Asp Ile Gln Met Thr Gln Ser Pro Ser Ser Leu Ser Ala Ser Val Gly 1 5 10 15 Asp Arg Val Thr Ile Lys Cys Gln Ala Ser Gln Ser Ile Phe Ser Gly 25 30 Leu Ala Trp Tyr Gln Gln Lys Pro Gly Lys Ala Pro Lys Leu Leu Ile Page 22 40 45 Tyr Gly Ala Ser Lys Leu Ala Ser Gly Val Pro Ser Arg Phe Ser Gly 50 55 60 Ser Gly Ser Gly Thr Glu Phe Thr Leu Thr Ile Ser Gly Leu Gln Pro 65 70 75 80 Ala Asp Phe Ala Thr Tyr Tyr Cys Gln Ser Tyr Tyr Tyr Ser Ser Ser 85 90 95 Ser Ser Asp Gly Ser Tyr Ala Phe Gly Gly Gly Thr Lys Leu Thr Val 100 105 110 Leu Gly Arg Thr Val Ala Ala Pro Ser Val Phe Ile Phe Pro Pro Ser 115 120 125 Asp Glu Gln Leu Lys Ser Gly Thr Ala Ser Val Val Cys Leu Leu Asn 130 135 140 Asn Phe Tyr Pro Arg Glu Ala Lys Val Gln Trp Lys Val Asp Asn Ala 145 150 155 160 Leu Gln Ser Gly Asn Ser Gln Glu Ser Val Thr Glu Gln Asp Ser Lys 165 170 175 Asp Ser Thr Tyr Ser Leu Ser Ser Thr Leu Thr Leu Ser Lys Ala Asp 180 185 190 Tyr Glu Lys His Lys Val Tyr Ala Cys Glu Val Thr His Gln Gly Leu 195 200 205 Ser Ser Pro Val Thr Lys Ser Phe Asn Arg Gly Glu Cys 210 215 220 <210> 53 <211> 452 <212> PRT <213> cial Sequence <220> Page 23 <223> heavy chain of zed IgG of clone 16H05 <400> 53 Glu Val Gln Leu Val Glu Ser Gly Gly Gly Leu Val Gln Pro Gly Gly 1 5 10 15 Ser Leu Arg Leu Ser Cys Lys Ala Ser Gly Ile Asp Phe Asn Asn Tyr 25 30 Gly Ile Gly Trp Val Arg Gln Ala Pro Gly Lys Gly Leu Glu Trp Ile 40 45 Thr Tyr Ile Tyr Pro Gly Phe Ala Ile Thr Asn Phe Ala Asn Ser Val 50 55 60 Lys Gly Arg Phe Thr Ile Ser Ser Asp Asn Ser Lys Asn Thr Val Tyr 65 70 75 80 Leu Gln Met Asn Ser Leu Arg Ala Glu Asp Thr Ala Val Tyr Tyr Cys 85 90 95 Ala Arg Asp Pro Val Tyr Ala Thr Ser Ser Gly Tyr Phe Asp Leu Trp 100 105 110 Gly Gln Gly Thr Leu Val Thr Val Ser Ser Ala Ser Thr Lys Gly Pro 115 120 125 Ser Val Phe Pro Leu Ala Pro Ser Ser Lys Ser Thr Ser Gly Gly Thr 130 135 140 Ala Ala Leu Gly Cys Leu Val Lys Asp Tyr Phe Pro Glu Pro Val Thr 145 150 155 160 Val Ser Trp Asn Ser Gly Ala Leu Thr Ser Gly Val His Thr Phe Pro 165 170 175 Ala Val Leu Gln Ser Ser Gly Leu Tyr Ser Leu Ser Ser Val Val Thr 180 185 190 Val Pro Ser Ser Ser Leu Gly Thr Gln Thr Tyr Ile Cys Asn Val Asn Page 24 195 200 205 His Lys Pro Ser Asn Thr Lys Val Asp Lys Lys Val Glu Pro Lys Ser 210 215 220 Cys Asp Lys Thr His Thr Cys Pro Pro Cys Pro Ala Pro Glu Leu Leu 225 230 235 240 Gly Gly Pro Ser Val Phe Leu Phe Pro Pro Lys Pro Lys Asp Gln Leu 245 250 255 Met Ile Ser Arg Thr Pro Glu Val Thr Cys Val Val Val Asp Val Ser 260 265 270 His Glu Asp Pro Glu Val Lys Phe Asn Trp Tyr Val Asp Gly Val Glu 275 280 285 Val His Asn Ala Lys Thr Lys Pro Arg Glu Glu Gln Tyr Asn Ser Thr 290 295 300 Tyr Arg Val Val Ser Val Leu Thr Val Leu His Gln Asp Trp Leu Asn 305 310 315 320 Gly Lys Glu Tyr Lys Cys Lys Val Ser Asn Lys Ala Leu Pro Ala Pro 325 330 335 Ile Glu Lys Thr Ile Ser Lys Ala Lys Gly Gln Pro Arg Glu Pro Gln 340 345 350 Val Tyr Thr Leu Pro Pro Ser Arg Glu Glu Met Thr Lys Asn Gln Val 355 360 365 Ser Leu Thr Cys Leu Val Lys Gly Phe Tyr Pro Ser Asp Ile Ala Val 370 375 380 Glu Trp Glu Ser Asn Gly Gln Pro Glu Asn Asn Tyr Lys Thr Thr Pro 385 390 395 400 Pro Val Leu Asp Ser Asp Gly Ser Phe Phe Leu Tyr Ser Lys Leu Thr 405 410 415 Page 25 Val Asp Lys Ser Arg Trp Gln Gln Gly Asn Val Phe Ser Cys Ser Val 420 425 430 Leu His Glu Ala Leu His Asn His Tyr Thr Gln Lys Ser Leu Ser Leu 435 440 445 Ser Pro Gly Lys <210> 54 <211> 114 <212> PRT <213> Artificial ce <220> <223> humanized VL of clone 16H05-sc04 <400> 54 Asp Ile Gln Met Thr Gln Ser Pro Ser Ser Leu Ser Ala Ser Val Gly 1 5 10 15 Asp Arg Val Thr Ile Thr Cys Gln Ala Ser Gln Ser Ile Phe Ser Gly 25 30 Leu Ala Trp Tyr Gln Gln Lys Pro Gly Lys Ala Pro Lys Leu Leu Ile 40 45 Tyr Gly Ala Ser Lys Leu Ala Ser Gly Val Pro Ser Arg Phe Ser Gly 50 55 60 Ser Gly Ser Gly Thr Glu Phe Thr Leu Thr Ile Ser Gly Leu Gln Pro 65 70 75 80 Ala Asp Phe Ala Thr Tyr Tyr Cys Gln Ser Tyr Tyr Tyr Ser Ser Ser 85 90 95 Ser Ser Asp Gly Ser Tyr Ala Phe Gly Gly Gly Thr Lys Leu Thr Val 100 105 110 Leu Gly Page 26 <210> 55 <211> 257 <212> PRT <213> Artificial Sequence <220> <223> zed scFv of clone 16H05-sc04 <400> 55 Met Asp Ile Gln Met Thr Gln Ser Pro Ser Ser Leu Ser Ala Ser Val 1 5 10 15 Gly Asp Arg Val Thr Ile Thr Cys Gln Ala Ser Gln Ser Ile Phe Ser 25 30 Gly Leu Ala Trp Tyr Gln Gln Lys Pro Gly Lys Ala Pro Lys Leu Leu 40 45 Ile Tyr Gly Ala Ser Lys Leu Ala Ser Gly Val Pro Ser Arg Phe Ser 50 55 60 Gly Ser Gly Ser Gly Thr Glu Phe Thr Leu Thr Ile Ser Gly Leu Gln 65 70 75 80 Pro Ala Asp Phe Ala Thr Tyr Tyr Cys Gln Ser Tyr Tyr Tyr Ser Ser 85 90 95 Ser Ser Ser Asp Gly Ser Tyr Ala Phe Gly Gly Gly Thr Lys Leu Thr 100 105 110 Val Leu Gly Gly Gly Gly Gly Ser Gly Gly Gly Gly Ser Gly Gly Gly 115 120 125 Gly Ser Gly Gly Gly Gly Ser Glu Val Gln Leu Val Glu Ser Gly Gly 130 135 140 Gly Leu Val Gln Pro Gly Gly Ser Leu Arg Leu Ser Cys Lys Ala Ser 145 150 155 160 Page 27 Gly Ile Asp Phe Asn Asn Tyr Gly Ile Gly Trp Val Arg Gln Ala Pro 165 170 175 Gly Lys Gly Leu Glu Trp Ile Thr Tyr Ile Tyr Pro Gly Phe Ala Ile 180 185 190 Thr Asn Phe Ala Asn Ser Val Lys Gly Arg Phe Thr Ile Ser Ser Asp 195 200 205 Asn Ser Lys Asn Thr Val Tyr Leu Gln Met Asn Ser Leu Arg Ala Glu 210 215 220 Asp Thr Ala Val Tyr Tyr Cys Ala Arg Asp Pro Val Tyr Ala Thr Ser 225 230 235 240 Ser Gly Tyr Phe Asp Leu Trp Gly Gln Gly Thr Leu Val Thr Val Ser 245 250 255 <210> 56 <211> 23 <212> PRT <213> Artificial Sequence <220> <223> Vk1 consensus sequence of framework I (Kabat positions 1-23) <400> 56 Asp Ile Gln Met Thr Gln Ser Pro Ser Ser Leu Ser Ala Ser Val Gly 1 5 10 15 Asp Arg Val Thr Ile Thr Cys <210> 57 <211> 15 <212> PRT <213> Artificial ce <220> <223> Vk1 consensus sequence of framework II (Kabat positions 35-49) Page 28 <400> 57 Trp Tyr Gln Gln Lys Pro Gly Lys Ala Pro Lys Leu Leu Ile Tyr 1 5 10 15 <210> 58 <211> 32 <212> PRT <213> Artificial ce <220> <223> Vk1 consensus sequence of framework III (Kabat positions 57-88) <400> 58 Gly Val Pro Ser Arg Phe Ser Gly Ser Gly Ser Gly Thr Asp Phe Thr 1 5 10 15 Leu Thr Ile Ser Ser Leu Gln Pro Glu Asp Phe Ala Thr Tyr Tyr Cys 25 30 <210> 59 <211> 10 <212> PRT <213> Artificial Sequence <220> <223> V lamda germline-based sequence of framework IV <400> 59 Phe Gly Thr Gly Thr Lys Val Thr Val Leu 1 5 10 <210> 60 <211> 10 <212> PRT <213> Artificial Sequence <220> <223> V lamda germline-based sequence of framework IV <400> 60 Phe Gly Gly Gly Thr Lys Leu Thr Val Leu 1 5 10 Page 29 <210> 61 <211> 10 <212> PRT <213> Artificial Sequence <220> <223> V lamda germline-based sequence of framework IV <400> 61 Phe Gly Gly Gly Thr Gln Leu Ile Ile Leu 1 5 10 <210> 62 <211> 10 <212> PRT <213> Artificial Sequence <220> <223> V lamda ne-based sequence of framework IV <400> 62 Phe Gly Ser Gly Thr Lys Val Thr Val Leu 1 5 10 Page 30