Anti-transglutaminase 2 antibodies

Abstract

The invention provides antibodies and antigen-binding fragments thereof that selectively bind to an epitope within the core region of transglutaminase type 2 (TG2). Novel epitopes within the TG2 core are provided. The invention provides human TG2 inhibitory antibodies and uses thereof, particularly in medicine, for example in the treatment and/or diagnosis of conditions including Celiac disease, scarring, fibrosis-related diseases, neurodegenerative/neurological diseases and cancer.

Claims

1. An antibody or an antigen-binding fragment thereof that binds human transglutaminase type 2 (TG2), wherein the antibody or antigen-binding fragment thereof comprises the amino acid sequences set forth in SEQ ID NO: 7 (LCDR1; KASQDINSYLT), SEQ ID NO: 13 (LCDR2; LTNRLMD), SEQ ID NO: 14 (LCDR3; LQYVDFPYT), SEQ ID NO: 10 (HCDR1; SSAMS), SEQ ID NO: 15 (HCDR2; TISSGGRSTYYPDSVKG) and SEQ ID NO: 16 (HCDR3; LISPY).

2. The antibody or antigen-binding fragment thereof of claim 1, wherein the antibody or antigen-binding fragment thereof has a light chain variable region comprising the amino acid sequence set forth in SEQ ID NO: 260 and a heavy chain variable region comprising the amino acid sequence set forth in SEQ ID NO: 258.

3. The antibody of claim 1, wherein the antibody thereof comprises or consists of an intact antibody.

4. The antibody of claim 1 comprising or consisting of an antigen-binding fragment selected from the group consisting of: an Fv fragment an Fab fragment; and an Fab-like fragment.

5. The antibody or antigen-binding fragment thereof according to claim 1, further comprising a moiety selected from the group consisting of a detectable moiety, and a cytotoxic moiety.

6. A pharmaceutical composition comprising an antibody or antigen-binding fragment thereof according to claim 1 and a pharmaceutically acceptable excipient, adjuvant, diluent or carrier.

7. The pharmaceutical composition according to claim 6 further comprising one or more further active ingredients.

8. The pharmaceutical composition according to claim 6, wherein the composition is formulated for intravenous, intramuscular, or subcutaneous delivery to a patient.

9. A kit comprising an antibody or antigen-binding fragment thereof according to claim 1 and a further agent.

10. An in vitro method of reducing or inhibiting TG2 enzyme activity, the method comprising contacting the antibody or antigen-binding fragment thereof according to claim 1 with a sample comprising human TG2.

11. The method according to claim 10, wherein the sample comprising human TG2 is a tissue sample comprising human TG2 or a call sample comprising human TG2.

12. A method of reducing or inhibiting TG2 enzyme activity in an individual in need thereof, comprising contacting the antibody or antigen-binding fragment thereof according to claim 1 with human TG2 in the individual.

Description

(1) The invention is now described in more detail by reference to the following, non-limiting, Figures and Examples.

(2) FIG. 1: Generation of a Human TG2 Recombinant Protein

(3) A: The TG2 catalytic core cDNA was generated by PCR from the pClineo-hTG2 vector and inserted into the pET 21a plasmid. Following amplification in E. Coli this was digested with Nhe I and Hind III to release the TG2 core cDNA and run on a 1% Agarose gel (lane 3). Bands were sized by reference to a 100 bp ladder (lane 1) and λ DNA molecular weight marker (Lane 2).

(4) B: The pET21a TG2 core vector was used to transform E. coli strain BL21-CodonPlus (DE3)-RIPL. Expression was induced using IPTG for 4 hours. TG2 core protein formed insoluble bodies that were recovered from lysates by centrifugation. These were re solubilised, and the 37 kDa His tagged TG2 core purified on a nickel column. 10 ng was separated by SDS-PAGE, western blotted and probed with CUB7402 anti TG2 antibody (lane 2) with reference to a precision plus molecular weight marker (lane 1).

(5) FIG. 2: Immunological Response in Mice to rhTG2 Core Protein

(6) A: Test bleeds were taken from 4 catalytic core immunised mice at day 45 after the first immunisation and 10 days after the second boost. Serum was serially diluted and reactivity checked by ELISA against immobilised TG2 core protein.

(7) B Reactivity was further checked by screened against human rh TG2 and rh TG2 catalytic core domain. 20, 40, 80 ng of protein was fractionated by SDS PAGE and western blotted onto a PVDF membrane. This was immunoprobed with a 1:1000 dilution of serum. Antibody binding was revealed using anti-mouse γ-chain specific HRP. For size reference to a precision plus molecular weight marker was used.

(8) FIG. 3: Hybridoma Reactivity Against TG Family Members

(9) A: ELISA were carried out using plates coated with recombinant TGs (100 ng/well) to determine TG type specificity in 109 hybridoma supernatants that showed good reactivity to TG2. Antibody binding was revealed using anti-mouse γ-chain specific HRP. A random selection of those screened is shown including EF4, CG9 & FD8 that showed cross reactivity.

(10) B: Nine selected hybridomas were double cloned. IgG was purified and tested for reactivity at 0.1μ/ml against recombinant human TG1, TG2, TG3, TG7 and Factor XIIIa using ELISA with plates coated with 100 ng of each TG. Data represents mean OD value from 3 separate ELISA±SEM.

(11) Factor XIIIa is denoted on graphs as TG13.

(12) FIG. 4: Identification of Hybridoma with Inhibitory Activity Against TG2

(13) Conditioned media from 32 hybridoma wells with specificity to TG2 were screened for their effects on 100 ng of rhTG2 activity using the .sup.3H putrescine incorporation assay. The chemical pan TG2 inhibitor 1,3-Dimethyl-2-[(2-oxo-propyl)thio]imidazolium chloride was used as a positive control for inhibition. RPMI (unconditioned medium) was used a negative control. 500 ng of a TG2 inhibitory antibody piloted by Quark biotechnology was included for comparison. Data represents mean CPM incorporated in 30 mins from at least three experiments done in duplicate±SEM. Bars shown in grey show significant TG2 inhibition (p<0.05).

(14) FIG. 5: Mapping of Inhibitory Antibody Epitopes.

(15) Each inhibitory monoclonal antibody was bound to an ELISA plate and panned against a human TG2 phage library. Phage binding to the antibody were rescued, amplified and subjected to 4 further rounds of panning. TG2 library fragments in the phage were then sequenced and overlapping sequences used to determine the epitope for each antibody. Common sequences between antibodies were then used to determine a consensus sequence for a particular inhibitory epitope and antibodies grouped accordingly. 3 inhibitory epitopes were identified.

(16) FIG. 6: Structural location of Inhibitory Epitopes with the TG2 catalytic Core The TG2 amino acid sequence was entered into Pymol and a 3D graphical representation of the structure generated in its open, Ca.sup.2+ activated state with putative calcium binding sites (turquoise) and the catalytic triad (grey) shown for reference (left panel). The consensus inhibitory epitopes were then added in blue (Antibody group 1—AB1 site), red (antibody group 2—DF4 site) and yellow (antibody group 3, DD9 site).

(17) FIG. 7: VL Sequence of Inhibitory Antibodies

(18) RNA from each inhibitory hybridoma was extracted, reverse transcribed and amplified by PCR using a degenerate FR1 primers, MH1 and MH2 primers and 3 constant region primers to amplify VH genes. The resulting VH and VK sequences are shown for AB1.

(19) FIG. 8. Efficacy of AB1 to Inhibit TG2 Activity in a Cell Homogenate

(20) A: Hep2G cells were lysed and 45 ug of protein mixed with 750 ng of IgG from AB1, DH2, DD9, BB7, DC1 and EH6 for 20 minutes. This was subsequently assayed using the .sup.3H Putrescine incorporation TG activity assay with sampling over 1 hour. The rate of reaction was calculated and expressed as a percentage of the same lysate incubated with a random antibody (MAB002). Data represents the mean percentage inhibition±SEM from 2 separate experiments done in duplicate. *p<0.05

(21) B: HepG2 cells were exposed to increasing glucose concentrations for 96 hours to up regulate TG2 expression. Cells were harvested, lysed and 25 ug of lysate fractionated by SDS-PAGE, western blotted and then immunoprobed with a 1 ng/ml solution of AB1 IgG using a chemiluminesant end point.

(22) FIG. 9. (Table 1): Comparative IC50 Values for TG2 Inhibitory Antibodies

(23) To determine an IC50 value for each antibody against human, rat and mouse the .sup.3H Putrescine assay was used. 100 ng of human TG2 or 25 ng of mouse and rat TG2 was used to generate a reaction where approximately 3000 cpm of Putrescine were incorporated per hour in 10 ul of the reaction mixture. Serial dilutions of each antibody were then applied starting from adding 500 ng (5 ug/ml final concentration) to the reaction mixture and incubated with the TG2 for 20 minutes prior to activating the reaction. IC50 values were calculated by determining the concentration at which the enzymatic rate of reaction was reduced by 50% using an appropriate curve fit in graphpad prism. Values are expressed as the amount of IgG in mg/ml in the reaction that would inhibit 1 ng of TG2.

(24) FIG. 10. Extracellular TG ASctivity in HK2 Cells in Response to TG2 Inhibition.

(25) HK2 cells were plated onto fibronectin and incubated for 2 hours in the presence of 0.1M biotin cadaverine with either 4 ng/μl of human anti-TG2 antibody (AB1) (part A), 4 ng/μl of human anti-TG2 antibody (DC1) (part B) or 400 μM of the site-specific pan TG inhibitor 1,3-Dimethyl-2-[(2-oxo-propyl)thio]imidazolium chloride. Extracellular TG activity was measured by the incorporation of biotin cadaverine into fibronectin with incorporation revealed using extravadin-HRP and a TMB substrate. Changes in optical density were measured at 450 nm in a 96 well plate reader. Data represents mean OD at 450 nm corrected to 1 mg of cell protein. n=6 wells per experimental group.

(26) FIG. 11. Comparison of TG2 Inhibition by Antibody AB1 to a Fab Fragment of Quark's TG2 Inhibitory Antibody Using a .sup.3H Putrescine Incorporation Assay

(27) 100 ng of hTG2 was assayed for TG2 activity based on the incorporation of .sup.3H Putrescine into dimethylcasein over a 60 minute period with the addition of either 1 μg of a fab fragment of an antibody described by Quark in WO2006/100679 and synthesised at Sheffield University or 500 ng of AB1. Data represents mean TG activity as incorporation of .sup.3H putrescine (CPM)±SEM from 3 independent experiments done in duplicate.

(28) FIG. 12. Percentage Comparison of TG2 Inhibition by Antibody AB1 with a Fab Fragment of Quark's TG2 Inhibitory Antibody Using a .sup.3H Putrescine Incorporation Assay

(29) Data from FIG. 11 is alternatively expressed as a percentage of TG activity at each time point to display the relative comparative knockdown of TG2 activity by application of AB1 and the Quark antibody fab fragment.

(30) FIG. 13. Comparison of TG2 Inhibition by Antibody AB1 to a Recombinant Rat IgG of Quark's TG2 Inhibitory Antibody Using a .sup.3H Putrescine Incorporation Assay

(31) 100 ng of hTG2 was assayed for TG2 activity based on the incorporation of .sup.3H Putrescine into dimethylcasein over a 60 minute period with the addition of either 500 ng of a recombinant rat version of a TG2 inhibitory antibody described by Quark in WO2006/100679 and synthesised at Medical Research Council Technology or 500 ng of AB1. Data represents mean TG activity as incorporation of .sup.3H putrescine (CPM)±SEM from 3 independent experiments done in duplicate.

(32) FIG. 14. Percentage Comparison of TG2 Inhibition by Antibody AB1 to a Recombinant Rat IgG of Quark's TG2 Inhibitory Antibody Using a .sup.3H Putrescine Incorporation Assay

(33) Data from FIG. 13 is alternatively expressed as a percentage of TG activity at each time point to display the relative comparative knockdown of TG2 activity by application of AB1 and the Quark recombinant rat IgG.

(34) FIG. 15. Effect of AB1 on ECM Levels in HK2 Cells

(35) Mature collagen levels in HK-2 cells were measured by the incorporation of .sup.3H proline into the ECM over a 76 hour period either with or without the addition of TG2 inhibitory antibody AB1. Data represents the incorporation of .sup.3H proline per mg of cellular protein expressed as a percentage of the mean level in untreated cells±SEM. n=2.

(36) FIG. 16. Binding ELISA of Humanised Versions of Antibodies.

(37) Supernatants from HEK293F cells co-transfected with different combinations of humanised light chains and heavy chain vectors were assayed in an anti-human IgG ELISA to determine concentration and in an anti huTG2 ELISA. Each supernatant was assayed in triplicate and IC.sub.50's determined. The most potent combination was selected for further studies and as the candidate humanised antibody.

(38) FIG. 17. Testing MRC Quark CTD190 on Human Tg2 by ELISA.

(39) 96 well plates were plated with hTG2 (1 μg/ml) in carbonate buffer overnight and ELISA detection performed using 100 ng/ml primary antibody. Detection was performed using anti-mouse IgG (SIGMA 3673) for CUB and anti-rat IgG (SIGMA A5795) for the Quark (both 1:5000). The Quark antibody made by MRC T reacts with human TG2.

(40) FIG. 18: RNA from the AB1 hybridoma was extracted, reverse transcribed and amplified by PCR using the degenerate signal sequence primer MHV4 with heavy chain constant region primer MHCG1, or using the degenerate signal sequence primer MKV4 with a kappa light chain constant region primer MKC. The resulting VH and VK sequences are shown.

(41) FIG. 19: RNA from the BB7 hybridoma was extracted, reverse transcribed and amplified by PCR using the degenerate signal sequence primer MHV4 with heavy chain constant region primer MHCG1, or using the degenerate signal sequence primer MKV4 with a kappa light chain constant region primer MKC. The resulting VH and VK sequences are shown.

(42) FIG. 20: RNA from the DC1 hybridoma was extracted, reverse transcribed and amplified by PCR using the degenerate signal sequence primer MHV4 with heavy chain constant region primer MHCG1, or using the degenerate signal sequence primer MKV4 with a kappa light chain constant region primer MKC. The resulting VH and VK sequences are shown.

(43) FIG. 21: RNA from the JE12 hybridoma was extracted, reverse transcribed and amplified by PCR using a 5′ RACE PCR with heavy chain constant region primer MHCG1, or using the signal sequence primer MKV1 with a kappa light chain constant region primer MKC. The resulting VH and VK sequences are shown.

(44) FIG. 22: RNA from the EH6 hybridoma was extracted, reverse transcribed and amplified by PCR using a 5′ RACE PCR with heavy chain constant region primer MHCG2B, or using the signal sequence primer MKV with a kappa light chain constant region primer MKC. The resulting VH and VK sequences are shown.

(45) FIG. 23: RNA from the AG9 hybridoma was extracted, reverse transcribed and amplified by PCR using the degenerate signal sequence primer MHV7 with heavy chain constant region primer MHCG1, or using a mix of degenerate signal sequence primers MKV1-11 with a kappa light chain constant region primer MKC. The resulting VH and VK sequences are shown.

(46) FIG. 24: RNA from the AH3 hybridoma was extracted, reverse transcribed and amplified by PCR using the degenerate signal sequence primer MHV7 with heavy chain constant region primer MHCG2B, or using the signal sequence primer MKV1 with a kappa light chain constant region primer MKC. The resulting VH and VK sequences are shown.

(47) FIG. 25: RNA from the DD9 hybridoma was extracted, reverse transcribed and amplified by PCR using a 5′ RACE PCR with heavy chain constant region primer MHCG2A, or using the degenerate signal sequence primer MKV5 with a kappa light chain constant region primer MKC. The resulting VH and VK sequences are shown.

(48) FIG. 26: RNA from the DH2 hybridoma was extracted, reverse transcribed and amplified by PCR using a 5′ RACE PCR with heavy chain constant region primer MHCG2B, or using the degenerate signal sequence primer MKV45 with a kappa light chain constant region primer MKC. The resulting VH and VK sequences are shown.

(49) FIG. 27: RNA from the DD6 hybridoma was extracted, reverse transcribed and amplified by PCR using a 5′ RACE PCR with heavy chain constant region primer MHCG2B, or using a 5′ RACE PCR with a lambda light chain constant region primer MLC. The resulting VH and VL sequences are shown.

(50) FIG. 28: RNA from the IA12 hybridoma was extracted, reverse transcribed and amplified by PCR using the degenerate signal sequence primer MHV9 with heavy chain constant region primer MHCG1, or using the degenerate signal sequence primer CL14 with a kappa light chain constant region primer MKC. The resulting VH and VK sequences are shown.

(51) FIG. 29. Dose response curves and IC50 values for enzymatic inhibition of recombinant human TG2 by chimeric anti-TG2 antibodies, (a) cAB003, (b) cBB001, (c) cDC001, (d) cDD9001, (e) cDH001 and (f) the commercial TG2 antibody CUB7402. IC50 values are mean of 3 independent experiments.

(52) FIG. 30. Dose response curves and 1050 values for enzymatic inhibition of recombinant cynomogulus monkey TG2 by chimeric anti-TG2 antibodies (a) cDC001 and (b) the commercial TG2 antibody CUB7402.

(53) FIG. 31. Dose response curves and 1050 values for enzymatic inhibition of recombinant human TG2 by humanized anti-TG2 antibodies, (a) hBB001AA, (b) hBB001 BB, (c) hAB005 and (d) hAB004.

(54) FIG. 32. Dose response curves and 1050 values for enzymatic inhibition of recombinant cynomogulus monkey TG2 by humanized anti-TG2 antibodies (a) hBB01AA and (b) hAB004.

(55) FIG. 33. Dose response curves and 1050 values for enzymatic inhibition of recombinant human TG2 by murine monoclonal anti-TG2 antibodies, (a) mAB003, (b) mBB001, (c) mDC001, (d) mDD9001, (e) mDH001 and (f) mDD6001.

(56) FIG. 34—Binding ELISA of Humanised Versions of AB1 Aantibodies.

(57) Supernatants from HEK293F cells co-transfected with different combinations of humanised AB1 light chains and AB1 heavy chain vectors were assayed in an anti-human IgG ELISA to determine concentration and in an anti huTG2 ELISA. Each supernatant was assayed in triplicate and IC.sub.50's determined. The most potent combination was selected for further studies and as the candidate humanised antibody.

(58) FIG. 35. Dose response ELISA binding curves and EC50 data for antibodies binding to human TG2 (a) chimeric antibodies cDD9001, cDH001, cDC001, commercial TG2 antibody CUB7402 and isotype-matched control, (b) chimeric antibody cBB001 and isotype-matched control and (c) chimeric antibody cAB003 and isotope-matched control.

(59) FIG. 36. Dose response ELISA binding curves and EC50 data for antibodies binding to cynomogulus monkey TG2 (a) chimeric antibodies cDD9001, cDH001, cDC001, commercial TG2 antibody CUB7402 and isotype-matched control, (b) chimeric antibody cBB001 and isotype-matched control and (c) chimeric antibody cAB003 and isotope-matched control.

(60) FIG. 37. Dose response ELISA binding curves and EC50 data for antibodies binding to human TG2 (a) humanized antibodies hBB001AA, HBB001BB, commercial TG2 antibody CUB7402 and isotype-matched control and (b) humanized antibody hAB004.

(61) FIG. 38. Dose response ELISA binding curves and EC50 data for antibodies binding to cynomogulus monkey TG2 (a) humanized antibodies hBB001AA, HBB001BB, commercial TG2 antibody CUB7402 and isotype-matched control and (b) humanized antibody hAB004 and isotope-matched control.

(62) FIG. 39: Humanised AB1 binding activity with extracellular TG2 Inhibition of Extracellular TG2 activity produced by HK2 cells was assayed using an ELISA measuring the incorporation of biotin cadaverine into fibronectin. An exemplar curve showing the inhibition of TG2 activity by humanised AB1 (hAB005) and the IC obtained is shown.

(63) FIG. 40: Humanised BB7 Binding Activity with Extracellular TG2.

(64) Inhibition of Extracellular TG2 activity produced by HK2 cells was assayed using an ELISA measuring the incorporation of biotin cadaverine into fibronectin. An exemplar curve showing the inhibition of TG2 activity by versions of humanised BB7 (hBB001AA and hBB001BB) and the ICs obtained is shown.

(65) FIG. 41: Cytochalasin D, R281 and ZDON Control Screatch Assay Results and Commercial Antibody CUB7402 Scratch Assay Results.

(66) Scratch wound assays were performed using WI-38 cell, after plating and overnight growth, cells were washed in media without serum and a scratch wound generated using an Essen Wound Maker. Media was removed and replaces with 95 ul/well serum free media containing controls and test antibodies. The plate was placed in an Essen Incycte and the closure of the wound analysed using Incucyte software. Relative wound density was plotted against time for the controls cytochalasin D, R281 and Z-Don (panel A) and the commercial antibody CUB7402 and cytochalasin (panel B).

(67) FIG. 42: Humanised BB7 Scratch Assay Results.

(68) Scratch wound assays were performed using WI-38 cell, after plating and overnight growth, cells were washed in media without serum and a scratch wound generated using an Essen Wound Maker. Media was removed and replaces with 95 ul/well serum free media containing controls and test antibodies. The plate was placed in an Essen Incycte and the closure of the wound analysed using Incucyte software. Relative wound density was plotted against time for the humanised hBB001 AA and the control cytochalasin D (panel A) and hBB001BB and the control cytochalasin D (panel B).

(69) FIG. 43: Humanised AB1 Scratch Assay Results.

(70) Scratch wound assays were performed using WI-38 cell, after plating and overnight growth, cells were washed in media without serum and a scratch wound generated using an Essen Wound Maker. Media was removed and replaces with 95 ul/well serum free media containing controls and test antibodies. The plate was placed in an Essen Incycte and the closure of the wound analysed using Incucyte software. Relative wound density was plotted against time for the humanised hAB005 and the control cytochalasin D

(71) FIG. 44: Chimeric DC1 Scratch Assay Results.

(72) Scratch wound assays were performed using WI-38 cell, after plating and overnight growth, cells were washed in media without serum and a scratch wound generated using an Essen Wound Maker. Media was removed and replaces with 95 ul/well serum free media containing controls and test antibodies. The plate was placed in an Essen Incycte and the closure of the wound analysed using Incucyte software. Relative wound density was plotted against time for the chimeric antibody cDC001 and the control cytochalasin D

(73) FIG. 45: Human TG2 binding to cAB003 immobilised antibody by Biacore. The association phases of human TG2 injections over the cAB003-coated biosensor at 25, 50, 100 and 200 nM, including in duplicate at 50 nM, are shown on the left. From the same experiment, two long dissociation phases were collected, as shown on the right. Fits are shown as solid black lines and the results are shown in Table 25.

(74) FIG. 46: Cynomolgus monkey TG2 binding to hAB004 immobilised antibody by Biacore. The association phases of cynomolgus monkey TG2 injections over the hAB004-coated biosensor at 25, 50, 100, 200 and 400 nM, including in duplicate at 50 nM, are shown on the left. From the same experiment, two long dissociation phases were collected, as shown on the right. Fits are shown as solid black lines and the results are shown in Table 26.

(75) FIG. 47: Human TG2 binding to cDH001 immobilised antibody by Biacore in the absence of calcium. The association phases of human TG2 injections over the cDH001-coated biosensor at 25, 50, 100 and 200 nM, including in duplicate at 50 nM, are shown on the left. From the same experiment, two long dissociation phases were collected, as shown on the right. Fits are shown as solid black lines and the results are shown in Table 25.

EXAMPLE 1: DEVELOPING A TG2 INHIBITORY ANTIBODY SUITABLE FOR THERAPEUTIC USE IN MAN WITH THE IDENTIFICATION OF 3 SPECIFIC INHIBITORY EPITOPES

(76) Transglutaminase type 2 (TG2) catalyses the formation of an ε-(γ-glutamyl)-lysine isopeptide bond between adjacent peptides or proteins including those of the extracellular matrix (ECM). Elevated extracellular TG2 leads to accelerated ECM deposition and reduced clearance that underlies tissue scarring and fibrosis. It also is linked to celiac disease, neurodegenerative disorders and some cancers. While numerous compounds have been developed that inhibit transglutaminases, none of these are specific to TG2, inhibiting all transglutaminases to some extent. While these have allowed proof of concept studies for TG2's role in these pathologies, the lack of isoform specificity has prevented their application in man. To address this, we set out to develop a high affinity TG2 specific antibody that would inhibit only TG2 activity.

(77) A recombinant protein encompassing amino acids 143 to 473 of the human TG2 core was produced in Escherichia coli, re-folded and 100 μg injected into 4 mice with boosts at 2, 5, 7, and 10 weeks. Spleens were recovered 4 days after the final boost and splenocytes fused to Sp2/0-Ag-14 myeloma cells. Seventy-five hybridoma supernatants showed specificity to TG2. These hybridoma supernatants were screened for their ability to inhibit TG2 activity in a putrescine incorporation assay containing 100 μg of TG2. Ten TG2 specific supernatants were inhibitory. These were subsequently double cloned. Using phage display to screen a TG2 fragment library, each antibody was mapped to a precise epitope in the TG2 core domain and 3 distinct inhibitory epitopes determined. The amount of antibody to reduce the activity from 100 ng of TG2 by 50% was determined.

(78) The 2 most effective antibodies, AB1 and DC1 bound to amino acids 304 to 327 and had an IC.sub.50 of 1.1×10.sup.−5 mg/ml IgG per ng of recombinant TG2. Application of AB1 & DC1 was able to inhibit TG2 successfully in human Hep2G cells and extracellular TG2 in human HK-2 cells when applied to the culture media.

(79) Thus, immunisation of mice with the TG2 core domain surprisingly enabled the generation of monoclonal antibodies that target previously unreported epitopes within the catalytic core. These antibodies are specific, inhibit TG2 activity effectively and are suitable for in vivo application.

(80) Materials and Methods

(81) Transglutaminase 2 Catalytic Core Domain Production The catalytic core domain of human TG2 (residues Cys143-Met 473 of TG2) was expressed, refolded and purified to permit immunisation in mice. The catalytic core domain (PCR sense primer GCG CGC GCT AGC TGC CCA GCG GAT GCT GTG TAC CTG GAC (SEQ ID NO: 97), anti-sense GCG CGC AAG CTT CAT CCC TGT CTC CTC CTT CTC GGC CAG (SEQ ID NO: 98)) was cloned into the expression vector pET21a(+) and expressed as insoluble inclusion bodies in E. coli strain BL21-CodonPlus (DE3)-RIPL (Agilent Technologies). In brief, 50 μl of competent BL21 (DE3) pLysS cells were transformed with 1 μl of the expression plasmid (30 ng/μl) and plated onto LB agar plates containing the selective antibiotics (100 μg/ml ampicillin, 34 μg/ml chloramphenicol) and 1% glucose and incubated overnight at 37° C. A single colony was picked to seed 10 ml of fresh LB medium containing 100 μg/ml ampicillin, 34 μg/ml chloramphenicol and 1% glucose in shaking incubator at 37° C. and at 200 rpm. After overnight growth, cultures were transferred in 100 ml 2×YT media with 1% glucose and grown to an OD.sub.600 nm of 0.8 and then transferred to 1 L 2×YT medium until the OD.sub.600 nm reached 0.8 again. After 4 hours induction under 1 mM IPTG to stimulate expression, pelleted and bacteria were lysed by sonication in buffer A (10 mM Tris; 1 mM EDTA; 10 mM DTT; 1 mM PMSF; 0.5 mg/ml lysozyme protease inhibitor tablets (Roche), pH 8.0). Inclusion bodies were harvested by centrifugation at 40,000×g and washed three times in wash buffer B (50 mM Tris; 1 mM EDTA; 10 mM DTT; 2% sodium deoxycholate, pH 8.0) before a final wash in deionised water.

(82) Inclusion bodies were solubilised in 3.5 mls of resolubilisation buffer (40 mM Tris-HCl, 8 M urea, and 10 mM DTT pH12) and refolded over a period of 16 hours in refolding buffer (40 mM Tris HCl; 150 mM NaCl; 20% glycerol; 5 mM cysteine; 0.5 mM cystine pH 8) at 4° C. in the dark.

(83) The resolubilised inclusion bodies were loaded onto a 1 ml Nickel column. Briefly, the column was pre-equilibrated with binding buffer (40 mM Tris; 300 mM NaCl; 10 mM Imidazole) and the inclusion bodies applied. The column was extensively washed (40 mM Tris; 300 mM NaCl; 30 mM imidazole). The recombinant protein was eluted by high concentration imidazole buffer (40 mM Tris; 300 mM NaCl; 300 mM imidazole). Eluted protein containing fractions were pooled and dialysed overnight against an appropriate buffer (40 mM Tris; 300 mM NaCl pH 8). Protein was assessed using the Bradford protein assay

(84) HepG2 Cell Culture & Lysates

(85) HepG2 cells were kindly supplied by Richard Ross (University of Sheffield). Cells were routinely grown at 37° C. in a 95% humidified atmosphere of 5% CO.sub.2 in DMEM/4.5 g per litre glucose supplemented with 10% foetal calf serum (FCS), 100 IU penicillin and 100 pg/ml streptomycin, 2 mM I-glutamine (all GIBCO). Two million cells were seeded on a 10 cm dishes and grown for 48 hours. Cells were lysed in 250 μl of STE buffer (0.32M sucrose, 5 mM Tris, 1 mM EDTA containing protease inhibitors Phenylmethylsulphonyl fluoride (1 mM), benzamidine (5 mM), and leupeptin (10 pg/ml) and sonicated on ice to produce a cell lysate usable in TG2 activity assay.

(86) Human Kidney 2 (HK2) Cells:

(87) HK-2 cells (kidney proximal tubular epithelium) were purchased from the European cell culture collection at passage 3. Cells were routinely grown at 37° C. in a 95% humidified atmosphere of 5% CO.sub.2 in keratinocyte serum free medium (KSFM, Gibco 17005-042) with L-glutamine supplemented with recombinant EGF (0.1-0.2 ng/ml) and bovine pituitary extract (20-30 ug/ml). For passage, media was removed and washed once with 1×PBS before trypsinising with 1 ml of 0.25% trypsin/EDTA (T75 flask) for 1 minute at 37° C. Cells were resuspended in 10 mls of KSFM and centrifuged at 400 g for 1 minute. Media was removed and cells plated in KSFM (1:3 to 1:5 split is normal). Cells were used experimentally at passages 5-14. Cells typically grew well to 95% confluence.

(88) Coomassie Staining and Western Blotting

(89) The purity of recombinant proteins was checked by running 5 μg of the recovered protein on a 10% (w/v) polyacrylamide denaturing gel and staining with Coomassie Brilliant Blue R staining solution (Sigma).

(90) Confirmation of TG2 core protein synthesis as well as TG2 and TG2 core reactivity levels following immunisation were all measured by western blotting. Recombinant proteins (10 to 80 ng) were loaded on a 10% (w/v) polyacrylamide denaturing or non-denaturing gel as required and transferred onto PVDF membranes (Transblot SD, Biorad, UK) for one hour at 100 V. Membranes were blocked overnight at 4° C. with 3% (w/v) BSA in TBS/0.1% (v/v) Tween 20. The membranes were then washed and probed with monoclonal mouse anti-transglutaminase antibodies in TBS/Tween containing 1% BSA.

(91) For proof of recombinant TG2 core protein and as a positive control for antibody screening the commercial antibody Cub7402 (neomarkers) was used at a 1:1000 dilution. Binding of primary antibody was detected with the anti-mouse gamma-chain-HRP linked secondary antibody (Sigma, Poole, UK). Bands were visualised using ECL chemiluminescent detection system (Amersham, UK).

(92) Mouse Immunisation and Fusion

(93) Each mouse was immunised with a mixture of 50 μg of antigen (made up to a volume of 50 μl with sterile PBS) and 50 μl of complete Freund's adjuvant. Four (8-12 week old) BALB/C mice were injected. Two boost immunisations were carried out (day 14 and day 35) using the same procedure with the exception that incomplete Freund's adjuvant was used for these injections. At day 45, test bleeds were taken from all animals and assessed for reactivity to TG2 by ELISA.

(94) The two best responders were further boosted by injection of 100 μg of core protein (in PBS) again mixed with incomplete Freuds Adjuvant at 10 weeks, and 4 days later the animals were sacrificed for splenocyte recovery and fusion with Sp2/0-Ag-14 myeloma cells. From this fusion, approximately 1000 wells were screened for reactivity to TG2 protein by ELISA.

(95) Screening for TG2 Specificity

(96) Conditioned medium or purified IgG were tested for reactivity to transglutaminase family members. The ability of each to bind to each transglutaminase (TG1, TG2, TG3, TG5, TG7 and Factor XIIIa; all Zedira) was determined using a plate binding assay. Microtiter plates (Costar, Cambridge, UK) were coated with recombinant TG (Zedira, Darmstadt, Germany) in 50 μl of 0.1 M bicarbonate/carbonate buffer (pH 9.6) overnight at 4° C. Plates were blocked for 2 h at 37° C. with 200 μl PBS containing 3% w/v BSA. Plates were washed three times with PBS containing 0.05% Tween 20 (washing buffer) and 100 μl of diluted conditioned medium (dilution 1:5 to 1:20) or purified anti-TG2 catalytic core mAbs was added. Plates were incubated for a further 1 h at room temperature. The washing step was repeated and anti-mouse gamma chain-horseradish peroxidase (1:5000) in PBS containing 0.05% Tween 20 (v:v) and 1% BSA (w:v) (Sigma, Poole UK) was added for 1 h. After eight washes, binding was revealed with 50 μl of 3,3′,5,5′-tetramethylbenzidine substrate. The reaction was stopped by adding 25 μl of 0.1 M H.sub.2SO.sub.4 and the absorbance at 450 nm was determined.

(97) Screening for TG2 Inhibition

(98) TG activity is measured by the Ca.sup.2+ dependent incorporation of .sup.3H-putrescine into N′,N′-dimethylcasein. Recombinant human TG2 (100 ng) was pre-incubated for twenty minutes at room temperature with the test sample (conditioned medium or purified IgG) before starting the reaction. Twenty-five μl of reaction mix (5 μl of 25 mM CaCl.sub.2, 5 μl of 40 mM dithiothreitol, 5 μl .sup.3H-putrescine mix, and 10 μl 25 mg/ml of N,N′ dimethylcasein (replace 25 mM CaCl.sub.2 with 100 mM EDTA for a non-enzymatic control) was added to start the reaction and the samples incubated at 37° C. for up to 1 hour. Aliquots of 10 μl were spotted onto a strip of 3 MM Whatman filter paper and plunged immediately into ice-cold 10% trichloroacetic acid (TCA) in order to precipitate the cross-linked proteins typically at time 0, 10, 30 and 60 minutes into the reaction. After three extensive washes in ice-cold 5% TCA followed by 3 rinses with ice-cold 95% ethanol, the air dried filter was counted in 2 ml of scintillation fluid (Ultima Gold Packard, Perkin Elmer). The rate of reaction was calculated. 1 TG unit is equivalent to the incorporation of 1 nmol of putrescine per hour at 37° C.

(99) The same protocol was used to assess TG inhibition in cell lysates by replacing the 25 μl of recombinant protein with 25 μl of cell lysate.

(100) Hybridoma Cloning & Purification of Antibodies from Conditioned Medium

(101) Monoclonal antibody isolation was undertaken from the cloned inhibitory hybridomas. Initially identified hybridoma wells were doubly cloned by a limiting dilution process (to ensure stability and clonality) according to conventional methods (Loirat M J et al, 1992) with sub-clones tested as described by ELISA and activity screens. Selected antibody producing clones were expanded in 25 and 75 cm.sup.2 flasks and fed with serum free medium (Hyclone, Fisher Scientific, Loughborough, UK). As cells were expanded, conditioned medium was collected for IgG purification using affinity chromatography on protein G column (Amersham Life Sciences). The conditioned medium was diluted in an equal volume of 10 mM sodium phosphate, pH 7.25, and applied to the protein G column at a flow rate of 1.0 to 2.0 ml/min. The column was extensively washed with 10 column volumes of the same buffer. Bound antibody was eluted in glycine solution (0.1M; pH 2.7) and neutralised by 0.15% volumes of 1M Tris/HCl pH 9. Samples were dialysed against 1000 volumes of phosphate buffer saline solution for 24 hours with 2 buffer changes.

(102) Phage Display Mapping of Antibody Epitopes

(103) The full length coding sequence of human TG2 was amplified by polymerase chain reaction using the following primers; TG2-FL-1 5′ ATGGCCGAGGAGCTGGTCTTAGAGA 3′(SEQ ID NO: 99) and TG2-FL-2 5′ GGCGGGGCCAATGATGACATTCCGGA 3′ (SEQ ID NO: 100). The approximately 2 kb amplification product was purified using Quiagen PCR cleanup kit (Qiagen) and digested into random fragments using RQ DNAse I (Promega). The RQ DNAse reaction was treated with Klenow fragment of DNA polymerase I and T4 DNA polymerase to generate blunt-ended fragments. These were purified by gel electrophoresis, and fragments in the range of 50-150 bp extracted using Qiagen gel recovery kit (Qiagen, Crawley UK).

(104) A phage display vector was digested with EcoRV, treated with alkaline phosphatase and purified by gel electrophoresis and the Qiagen gel recovery kit. 100 ng of purified vector was ligated to 15 ng of prepared blunt fragments of human TG2 cDNA. The resultant ligation was electroporated into XL1-Blue electrocompetent cells (Agilent Technologies) and the fragment library rescued with VCSM13 helper phage (Agilent). Phage particles were precipitated with 2% glucose and 4% PEG 6000 and resuspended in PBS 0.1% Tween 20 (v:v) 1% BSA (w:v).

(105) Epitope mapping was carried out using the following procedure. ELISA wells were coated overnight at 4° C. with 30 μg of monoclonal antibody in 100 μl of coating buffer. The coated well was washed with PBS/Tween and blocked with 400 μl of 3% BSA in PBS (w:v) for 1 h at room temperature. Approximately 10.sup.10 phage particles (100 μl) were added to the blocked well and incubated at room temperature for 1 h. The well was washed 8 times with 400 μl PBS/0.5% Tween (v:v) and adherent phage eluted with 0.2 M glycine pH 2.2. Eluted phage were used to infect 1 ml of XL1-Blue host and samples plated onto LB agar (60 μg/ml ampicillin, 15 μg/ml tetracycline), the remaining host was added to 100 ml LB media (60 μg/ml ampicillin, 15 μg/ml tetracycline) and grown overnight at 37° C. in a shaking incubator at 200 rpm to generate the enriched library of selected fragments. This enrichment process was repeated 5 times and random colonies from the final round were selected for sequencing.

(106) Determining the Sequence of the Antibody VL Region

(107) Primers

(108) Heavy chain sense primers—A pair of highly degenerate FR1 primers, MH1 and MH2 (Wang et al 2000), were combined with 3 constant region primers to amplify VH genes.

(109) TABLE-US-00014 MH1 (SEQ ID NO: 101) 5′ CGCGCGCTCGAGSARGTNMAGCTGSAGTC 3′ MH2 (SEQ ID NO: 102) 5′CGCGCGCTCGAGSARGTNMAGCTGSAGSAGTC 3′ Mouse-G1 (SEQ ID NO: 103) 5′ AGGCGCAGTACTACAATCCCTGGGCACAATTTTCTTGTCCACC 3′ Mouse-G2a (SEQ ID NO: 104) 5′ AGGCGCAGTACTACAGGGCTTGATTGTGGGCCCTCTGGG 3′ Mouse-G2b (SEQ ID NO: 105) 5′ AGGCGCAGTACTACAGGGGTTGATTGTTGAAATGGGCCCG 3′

(110) Kappa Primers

(111) TABLE-US-00015 VK1 (SEQ ID NO: 106) 5′ CGCTGCGAGCTCGATATTGTGATGACBCAGDC 3′ VK2 (SEQ ID NO: 107) 5′ CGCTGCGAGCTCGAGRTTKTGATGACCCARAC 3′ VK3 (SEQ ID NO: 108) 5′ CGCTGCGAGCTCGAAAATGTGCTCACCCAGTC 3′ VK4 (SEQ ID NO: 109) 5′ CGCTGCGAGCTCGAYATTGTGATGACACAGTC 3′ VK5 (SEQ ID NO: 110) 5′ CGCTGCGAGCTCGACATCCAGATGACACAGAC 3′ VK6 (SEQ ID NO: 111) 5′ CGCTGCGAGCTCGAYATTGTGCTSACYCARTC 3′ VK7 (SEQ ID NO: 112) 5′ CGCTGCGAGCTCGACATCCAGATGACYCARTC 3′ VK8 (SEQ ID NO: 113) 5′ CGCTGCGAGCTCCAAATTGTTCTCACCCAGTC 3′ K-CONST (SEQ ID NO: 114) 5′ GCGCCGTCTAGAATTAACACTCATTCCTGTTGAA 3′

(112) Total RNA was extracted from monoclonal hybridoma cells (˜10.sup.5 cells) using Trizol (GIBCO) according to the manufacturer's protocol and quantified by A.sub.260 nm. cDNA was synthesised using ImProm II reverse transcriptase (Promega) and random hexamer primers. The reaction mix was as follows; 1 μg total RNA, 0.1 μg oligo (dN).sub.6, 12 μl ImProm II buffer, 1 μl 10 mM dNTPs (Promega), 8 μl 25 mM MgCl.sub.2, 4 μl ImProm II reverse transcriptase (Promega), DEPC-treated H.sub.2O up to total reaction volume of 60 μl. The RNA and random primer mix was heated to 70° C. for 10 min and then placed on ice. The remaining reaction components were added and then incubated at 20° C. for 10 min, then at 40° C. for a further 40 min.

(113) Amplification of VH and VK genes was carried out with GoTaq polymerase (Promega). Each 50 μl reaction contained the following; cDNA 2 μl, 20 μmol sense and antisense primers, 10 μl GoTaq reaction buffer, 1 μl 10 mM dNTPs, 5 μl 25 mM MgCl.sub.2, 2.5 u GoTaq polymerase, H.sub.2O to a final volume of 50 μl. Reactions were cycled 35 times using the following conditions: initial denature 95° C. 2 min; denature 94° C. 1 min, anneal 56° C. 1 min, extension 72° C. 1 min. PCR products were analysed by gel electrophoresis and cloned using the TOPO TA cloning kit (Invitrogen). Random minipreps of heavy and light chain PCR products were selected for sequencing.

(114) Measurement of Extracellular TG Activity

(115) Extracellular TG activity was measured by modified cell ELISA. HK-2 epithelial cells were harvested using 0.1M EDTA or 0.25% trypsin/EDTA and plated at a density of 8×10.sup.4 cells/well in serum free medium onto a 96 well plate that had been coated overnight with 100 μl/well of fibronectin (5 μg/ml in 50 mM Tris-HCl pH 7.4) (Sigma, Poole UK). Cells were allowed to attach for 2.5 h at 37° C. in the presence of the 0.1 mM biotin cadaverine [N-(5 amino pentyl biotinamide) trifluoroacetic acid] (Molecular Probes, Eugene Oreg., USA). Plates were washed twice with 3 mM EDTA/PBS and cells removed with 0.1% (w/v) deoxycholate in 5 mM EDTA/PBS. The supernatant was collected and used for protein determination. Plates were washed with 50 mM Tris-HCl and incorporated biotin cadaverine revealed using 1:5000 extravidin HRP (Sigma, Poole, UK) for 1 h at room temperature followed by a TMB (3,3′,5,5′-tetramethylbenzidine) substrate. The reaction was stopped with 50 μl 2.5 M H2SO4 and the absorbance read at 450 nm.

(116) Measurement of Collagen Levels by Radiolabelling

(117) Cells were seeded at a density of 3.75×10.sup.6/10 cm.sup.2 Petri dish or 1×10.sup.6/well of a 6 well plate. ECM collagen was assessed by labelling with 20 iCi of .sup.3,4H proline (1.0 mCi/ml, ICN). Labelling was performed for 72 h under standard cell culture conditions. Following labelling, the media was removed, cells washed with PBS and removed with 2 ml of 0.25 M ammonium hydroxide in 50 mM Tris pH 7.4 at 37° C. for 10 min. The soluble fraction was collected and protein concentration determined using the bicinchoninic acid (BCA) assay. The dishes were washed extensively with increasing volumes of PBS before the ECM was solubilised with 2 ml of 2.5% (w/v) SDS in 50 mM Tris pH 6.8. The dish was scraped to ensure complete removal of the ECM and 200 μl was measured for radioactivity in a beta scintillation counter. Counts were corrected per mg of solubilised cell protein and expressed as a percentage of the mean control value.

(118) Generation of Recombinant Ratified Quark IgG

(119) For experimental purposes a human-rat chimeric antibody from the sequence of a ‘human’ single-chain Fv of an antibody against human type-II transglutaminase was generated. The antibody is called QPCDTGII (shortened to QCT), and the sequences of the variable regions are available in WO 2006/100679A2.

(120) A rat γ2a subclass was selected for the heavy chain constant regions, removing the glycosylation site to reduce the chance of an ADCC reaction in the rat test animals. The selected rat constant region for the heavy chain was 013593 (Bruggemann, M. Gene 74: 473-482 (1988); Bruggemann, M., Free, J., Diamond, A., Howard, J., Cobbold, S. and Waldmann, H. Proc. Natl. Acad. Sci USA 83: 6075-6079 (1986)) from the Kabat database (Kabat, E. A., Wu, T. T., Perry, H. M., Gottesman, K. S., and Foeller, C. Sequences of Proteins of Immunological Interest. (NIH National Technical Information Service, 1991)). That for the kappa light chain was 013718 (Sheppard, H. W. and Gutman, G. A. Proc. Natl. Acad. Sci. USA 78: 7064-7068 (1981)) from Kabat.

(121) In brief, heavy chain and kappa chain coding sequences were generated by DNA synthesis (codon usage was adapted to a mammalian codon bias).

(122) The heavy chain gene synthesis product was amplified by PCR using the primers QCT_HindIII and QCT_H_rev. The PCR-product was cut with HindIII and NgoMIV and ligated into MRCT expression vector. Clones of competent DH5α bacteria chemically transformed by the ligation product were PCR-screened using the primers HCMVi and rat_gamma1. Three clones generating a PCR product of the predicted size were sequenced.

(123) The kappa chain gene synthesis product was amplified by PCR using the primers QCT_HindIII and QCT_L_rev. The PCR-product was cut with HindIII and PpuMI and ligated into the expression vector pKN100. Clones of competent DH5α bacteria chemically transformed by the ligation product were PCR-screened using the primers HCMVi and rat_kappa. Three clones generating a PCR product of the predicted size were sequenced.

(124) A double insert expression vector coding for both Heavy and kappa chains was generated and transfected into HEK293T cells. Cell culture supernatant from two large scale HEK293T transfections was pooled and affinity purified on a 1 ml Protein L-agarose column using an ÄKTA Explorer chromatography system, in accordance with the manufacturer's protocols. A single OD 280 nm peak eluted with IgG Elution Buffer, and was dialysed against two changes of PBS. This was assayed both by UV absorption at 280 nm, and by rat IgG.sub.2a ELISA. The total yield was approximately 700 μg (by OD.sub.280 nm); 303.5 μg (by ELISA).

(125) Humanisation of AB1 Antibody

(126) Human VH and VK cDNA Databases

(127) The protein sequences of human and mouse immunoglobulins from the International Immunogenetics Database 2009.sup.101 and the Kabat Database Release 5 of Sequences of Proteins of Immunological Interest (last update 17 Nov. 1999).sup.102 were used to compile a database of human immunoglobulin sequences in a Kabat alignment. Our database contains 10,606 VH and 2,910 VK sequences.

(128) Molecular Model of AB1

(129) A homology model of the mouse antibody AB1 variable regions has been calculated using the Modeller program.sup.103 run in automatic mode. The atomic coordinates of 1 MQK.pdb, 3LIZ.pdb and 1MQK.pdb were the highest identity sequence templates for the Interface, VL and VH respectively as determined by Blast analysis of the Accelrys antibody pdb structures database. These templates were used to generate 20 initial models, the best of which was refined by modeling each CDR loop with its 3 best loop templates.

(130) hAB1 Framework Selection

(131) The sequence analysis program, gibsSR, was used to interrogate the human VH and VK databases with the AB1 VHc, VKc and VKc1 protein sequences using various selection criteria. FW residues within 5 Å of a CDR residue (Kabat definition) in the homology model of mouse antibody AB1, were identified, and designated as the “5 Å Proximity” residues.

(132) AF06220 was chosen as the FW on which to base the initial humanised AB1 VHc construct. Table 1 shows the alignment and residue identity of AF06220 to murine Ab1. Table 2 shows the 5 Å proximity envelope of the sequences. AF062260 has only 1 somatic mutation away from its germline VH gene Z12347 (Table 3).

(133) AY247656 was chosen as the FW on which to base the initial humanised AB1 VKc construct. The alignment and residue identity to murine AB1 are shown in Table 4; Table 5 shows the 5 Å proximity envelope of the sequences. The sequence shows 5 somatic mutations from its germline VK gene X93620 (Table 6).

(134) AF193851 was chosen as the FW on which to base the initial AB1 VKc.sub.1 construct. The alignment and residue identity to murine AB1 are shown in Table 7. Table 8 shows the 5 Å proximity envelope of the sequences. The sequence shows no somatic mutations from its germline VK gene J00248 (Table 9).

(135) Binding ELISA

(136) HEK 293F cells were co-transfected with combinations of different humanised light chain vectors in association with different humanised heavy chain vectors. Recombinant human TG2 was used to measure antibody binding by ELISA. The results indicated that the Heavy Chain version RHA (Table 10), in combination with either Light Chain versions RKE and RKJ (Table 11) (representing the different Light Chain versions humanised) showed optimal binding (FIG. 16).

(137) Heavy Chain version RHA is an un-modified graft of the mouse CDR regions of the AB1 antibody onto the Human donor sequence. However, both Light Chain versions RKE and RKJ, have the same single 5 Å proximity reside backmutation, F72 (Kabat numbering—shown in green). This backmutation lies outside the Vernier.sup.104, Canonical.sup.105 or Interface.sup.106 residues (see Table 11). 101. Lefranc, M. P. IMGT, the international ImMunoGeneTics Database®. Nucleic Acids Res. 31, 307-310 (2003). 102. Kabat, E. A., Wu, T. T., Perry, H. M., Gottesman, K. S. & Foeller, C. Sequences of Proteins of Immunological Interest. NIH National Technical Information Service, (1991). 103. Eswar, N. et al. Comparative protein structure modeling using Modeller. Curr. Protoc. Bioinformatics. Chapter 5: Unit 5.6., Unit (2006). 104. Foote, J. & Winter, G. (1992). Antibody framework residues affecting the conformation of the hypervariable loops. J Mol. Biol. 224, 487-499. 105. Morea, V., Lesk, A. M. & Tramontano, A. (2000). Antibody modeling: implications for engineering and design. Methods 20, 267-279. 106. Chothia, C., Novotny, J., Bruccoleri, R. & Karplus, M. (1985). Domain association in immunoglobulin molecules. The packing of variable domains. J Mol. Biol. 186, 651-663.
Tables

(138) TABLE-US-00016 TABLE 1

(139) TABLE-US-00017 TABLE 2 5Å Proximity Residues AB_VHc EVQLCAFTLSWVRWVARFTISRNLYCAKWG SEQ ID NO: 117 AF062260 ........F......S.............. SEQ ID NO: 118

(140) TABLE-US-00018 TABLE 3 ---------+---------+---------+---------+---------+---------+---------          10        20        30        40        50        60         ---------+---------+---------+---------+---------+---------+--------- Z12347.seq EVQLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSAISGSGGSTYYADSVKGRFT AF062260.seq .....................................................................                               <--->              <---------------> +---------+---------+-------- 70        80        90 +---------+---------+-------- Z12347.seq ISRDNSKNTLYLQMNSLRAEDTAVYYCAK SEQ ID NO: 119 AF062260.seq ............................R SEQ ID NO: 120

(141) TABLE-US-00019 TABLE 4

(142) TABLE-US-00020 TABLE 5 5Å Proximity Residues AB_VKc EIVLTQTCWFTLIYGVPFSGSGSGQDFFYCFG SEQ ID NO: 123 AY247656 .........YL.............T..T.... SEQ ID NO: 124

(143) TABLE-US-00021 TABLE 6 ---------+---------+---------+---------+---------+---------+---------+---------+---------+------         10         20        30        40        50        60        70        80        90  ---------+---------+---------+---------+---------+---------+---------+---------+---------+------ X93 DIQMTQSPSSLSASVGDRVTITCQASQDISNYLNWYQQKPGKAPKLLIYDASNLETGVPSRFSGSGSGTQFTFTISSLQPEDIATYYCQQYDNLPP 620. SEQ ID NO: 125 SEQ AY2 E.VL..............................................................................FG.......NTY.L 476                          <--------->              <----->                               <------- 56. SEQ ID NO: 126 SEQ

(144) TABLE-US-00022 TABLE 7

(145) TABLE-US-00023 TABLE 8 5Å Proximity Residues AB_VKc DIQMTQTCWFTLIYGVPFSGSGSGQDFFYCFG SEQ ID NO: 129 AF193851 ..........S.............T..T.... SEQ ID NO: 130

(146) TABLE-US-00024 TABLE 9 ---------+---------+---------+---------+---------+---------+---------+---------+---------+------         10         20        30        40        50        60        70        80        90  ---------+---------+---------+---------+---------+---------+---------+---------+---------+------ J00 DIQMTQSPSSLSASVGDRVTITCRASQGISNYLAWFQQKPGKAPKSLIYAASSLQSGVPSRFSGSGSGTOFTLTISSLQPEDFATYYCQQYNSYPP 248. SEQ ID NO: 131 SEQ AF1 .............................R......................N.....................................H.T..W 938                          <--------->              <----->                               <------- 51. SEQ ID NO: 132 SEQ

(147) TABLE-US-00025 TABLE 10

(148) TABLE-US-00026 TABLE 11        

(149) Results

(150) Generation of rh TG2 Core Protein.

(151) To force the generation of antibodies that would be more likely to target epitopes critical for TG2 activity, rather than favoured sites on the TG2 molecule, we immunised mice with the TG2 catalytic core rather than the full-length TG2 molecule. To generate the recombinant TG2 domain, a PCR construct was generated running from bases 329 to 1419 and inserted into the Pet21+(a) vector (FIG. 1A). Insertion and expression of this vector in BL21-CodonPlus (DE3)-RIPL bacteria resulted in the generation of an insoluble protein spanning amino acids 143 to 473 encompassing the entire catalytic core. This protein was solubilised and refolded in 40 mM Tris HCl; 150 mM NaCl; 20% glycerol; 5 mM cysteine; 0.5 mM cystine pH 8. 10 ng of this was run on a non-reducing polyacrylamide gel, western blotted and immunoprobed with CUB7402. A clear band was visible at 37 kDa which is consistent with the predicted size of the TG2 core (FIG. 1B). Larger bands were also immunoreactive to CUB7402 which most likely represent aggregates of the core protein as these were not present when a reducing gel was run (data not shown).

(152) Immunisation and Fusion.

(153) Four mice were immunised with 50 μg of rhTG2 core. At approximately five and nine weeks post immunisation, a serum sample was taken from each mouse and tested for reactivity against rh TG2 by ELISA using a serial dilution of the serum. All mice showed a strong immune reaction to rhTG2 core, even at the highest dilution used (1:51 000) (FIG. 2A). To confirm the antibodies would also recognise the full-length TG2, rhTG2 and rhTG2 core protein were run on a non-denaturing gel, western blotted and immunoprobed with a 1 in 1000 dilution of mouse serum (FIG. 2B). The mouse with the strongest reactivity (mouse C) against both proteins was boosted and splenocytes recovered for fusion using the University of Sheffield's Hybridoma service, Bioserv.

(154) Selection of Positive Hybridoma and Cloning.

(155) Out of 400 hybridoma wells selected by Bioserv as highest positives, supernatants from 109 showed persistent reactivity to TG2, however only 34 did not react with other key TG family members when tested in ELISA (representative examples shown in FIG. 3A). Those that were specific to TG2 had the supernatant screened by .sup.3H-putrescine incorporation assay for the ability to inhibit TG2 activity resulting from 100 ng of TG2 (FIG. 4). This initial screen indicated that 10 hybridoma supernatants were able to inhibit TG2 activity (AB1; DC1; BB7; EH6; DH2; DD9; JE12; AG9; AH3; DF4). Nine of the ten were successfully cloned by limited dilution. For the clone DF4, although clones were isolated post-cloning, they did not appear to be inhibitory. Post cloning, IgG was purified from each cloned hybridoma and retested for selective reactivity to TG2 (FIG. 3B).

(156) TG2 Inhibitory Potential.

(157) Each cloned hybridoma had its IgG tested for TG2 inhibitory activity against human, rat and mouse TG2 and the IC.sub.50 calculated based on the amount of IgG required to inhibit 1 ng of TG2. There was approximately a 12 fold range in IC.sub.50 values against human TG2 ranging from the most effective AB-1 at 1.1×10.sup.−5 mg/ml of IgG to the least effective JE12 at 12.3 1.1×10.sup.−5 mg/ml of IgG (FIG. 9; Table 1). Interestingly we were only able to determine an IC.sub.50 for 4 antibodies (DH2, DD9, EH6 and BB7) against rat TG2 with the best, DH2 having a IC.sub.50 of 2.23×10.sup.−4 mg/ml of IgG being some 6 fold less active than against human TG2 and comparatively 38 fold less active that the best AB-1 inhibitor against TG2 (FIG. 9; Table 1). None of the inhibitory antibodies were able to inhibit mouse TG2, probably due to immune tolerance.

(158) Mapping the Epitopes of Inhibitory Antibodies

(159) To establish which epitopes in TG2 were immunologically unique to TG2 while inhibitory, as well as establishing if these 10 antibodies were targeting the same or different sites, each antibody was mapped using phage display. A TG2 phage library was constructed and panned against each mAb. The epitope was then determined by consensus sequencing of the binding phages.

(160) AB1, AG1, AH1, BB7, DC1, EH6 and JE12 all appeared to bind in whole or part to a single epitope (FIG. 5) which encompasses amino acids 304 to 326 and appears to sit in front of active site within a substrate binding pocket (FIG. 6). This region we termed the AB-1 site and called antibodies targeting this site Group 1 antibodies. DF4 uniquely targeted a sequence running from amino acid 351 to 365 (FIG. 5) which runs from front to rear of core encompassing Asp 358 in the catalytic triad (FIG. 6). This we termed Group 2.

(161) DH2 and DD9 bound to a sequence spanning amino acids 450 to 467 (FIG. 5). These Group 3 antibodies bind to a region at the rear of core near the junction with β barrel-1, that we termed the DH2 site. The epitope encompasses a putative calcium binding site (FIG. 6).

(162) Antibody Sequencing

(163) In order to establish the variable light chain sequence for each antibody, RNA from each inhibitory hybridoma was extracted, reverse transcribed and amplified by PCR using a pair of highly degenerate FR1 primers, MH1 and MH2 primers being combined with 3 constant region primers to amplify VH genes.

(164) The resulting VH and VK sequences are shown in FIG. 7 for AB1.

(165) The Ability of AB1 to Inhibit TG2 Activity in a Protein Mixture In Vitro.

(166) The most potent inhibitory antibody against recombinant TG2 is AB1. To be of value therapeutically, it must be able to not only inhibit TG2 activity in a pure solution, but also in a complex protein solution and not associate in anyway with other proteins. To test this, a homogenate of the human hepatocyte cell line HepG2 was prepared. Application of 0.5 μg of AB1 was able to inhibit 70% of the TG2 activity (FIG. 8A). However BB7 produced a significantly better inhibition knocking down 90% of the TG2 activity. Immunoprobing 25 μg of this lysate with AB1 showed no off target association with a single immunoreactive band at a size corresponding to TG2 (FIG. 8B).

(167) The Ability of AB1 and DC1 to Inhibit Extracellular TG2 Activity.

(168) To acces if these antibodies could inhibit TG2 activity in a cell system. AB1 (FIG. 10a) and DC1 (FIG. 10b) were applied to human kidney 2 (HK-2) tubular epithelial cells in culture and extracellular TG activity assayed using the biotin cadaverine assay. AB1 was able to achieve a 60% inhibition and DC1 a 55% inhibition of activity when applied at 4 ng/ul in the culture media which was comparable to the chemical pan TG inhibitor 1,3-Dimethyl-2-[(2-oxo-propyl)thio]imidazolium chloride applied at 400 uM.

(169) Comparison of Antibody AB1 with Other Known Inhibitory Antibodies

(170) To test the effectiveness of AB1 in comparison to other known TG2 inhibitory antibodies both fab fragments (FIGS. 11, 12) and full IgG (FIGS. 13, 14) of an antibody as described by Quark biotechnology in patent application number WO2006/100679 were tested to inhibit TG2 activity in the .sup.3H Putrescine incorporation assay. The activity from 100 ng of human TG2 could be inhibited by 60 to 80% by 500 ng of AB1. In comparison neither the fab fragment or the full IgG of the Quark antibody could inhibit TG2 significantly in this assay.

(171) Discussion

(172) There is a clear need to validate TG2 as a therapeutic target in man across a range of diseases where experimental studies have suggested its involvement. These include tissue scarring, celiac disease, neurodegenerative diseases and chemo-resistance in some cancers. Limiting this has been the lack of truly TG2 specific compounds that can selectively inhibit TG2 activity in man.

(173) In this study we have for the first time immunised mice with a fragment of TG2 with the aim of being able to isolate a wider range of anti TG2 antibodies against the enzyme's catalytic core in the search of an inhibitory epitope. This elicited a good immune response with antibodies recognising both the rhTG2 core and native TG2 but no other TG.

(174) 10 of the antibodies isolated showed inhibitory activity. These were subsequently mapped to 3 TG2 specific, yet inhibitory epitopes. These antibodies have been cloned, sequenced and IgG isolated with IC.sub.50 values calculated. Three antibodies (AB1, DC1 and BB7) targeting a substrate pocket proved particularly effective inhibitors. Most importantly these antibodies also worked well both in a cell lysate and in cell culture indicating that these antibodies have the potential to function in a protein rich environment which is critical for in vivo application.

(175) We believe a key element in the successful generation of these inhibitory antibodies has been the decision to immunise with just the core protein. To our knowledge none of the commercial TG2 antibodies have inhibitory potential of any significance. Our own attempts to use full length TG2 resulted in a large number of antibodies, few of which were specific to TG2 and none of which were inhibitory. This would appear to be due to a clear immunogenic preference for protein loops within full length TG2 many of which fall on the rear of the catalytic core in similar positions to the most widely used anti TG2 antibody, CUB7402 (aa447 aa478).

(176) It is surprising that our approach has led to the production of much more effective antibodies. Without being bound by any theory we think that by simply raising antibodies to a smaller protein covering just the central core, we not only eliminate some of the favoured immunological epitopes, but we also force core targeting. This alone increases the variety of antibodies available for selection and thus wider coverage of the core. However, immunising with just the core means that much of the folding of the core is lost and thus some of the epitopes that perhaps may be less available within a whole TG2 molecule may be more attractive epitopes with the core in this format. Given that all 10 of the antibodies recognised linear epitopes (i.e. bound to TG2 on a reducing gel), while 80% of the antibodies we previously isolated using full length TG2 as an immunogen were conformation dependent, does suggest this may be a major factor.

(177) There have previously been other studies that have postulated the idea of a TG2 inhibitory antibody for human application. Esposito and colleagues developed recombinant antibodies from patients with celiac disease where it has been postulated that TG2 antibodies may have an inhibitory role [19]. One of these antibodies was developed for commercial application by Quark Biotechnology and a patent application filed (WO2006/100679). This antibody demonstrated some exciting early data in the prevention of kidney fibrosis in the rat UUO model. However, we produced a recombinant version of this antibody and while it reacted with TG2 in ELISA (FIG. 17) and western blot, we achieved little inhibition up to 500 ng of IgG per ng of TG2 for this antibody at which all of the antibodies developed in this study block essentially all TG2 activity. Furthermore WO2006/100679 describes the generation of a mouse version of this human antibody, and as such, long application in recognised rat models of kidney disease would prove difficult.

(178) Of note in the present study is the mapping of the 3 inhibitory epitopes within the TG2 core. The AB1 epitope is by far the most potent to target, which is perhaps surprising given the position of the epitope. Examination of its position within the predicted TG2 active structure [20] suggests it binds in the entry port to the catalytic triad in what may be a substrate pocket. Given the substrates we used in our screening assay are relatively small (putrescine and dimethyl casein), it is perhaps surprising that this site is so effective. However the position of the epitope must be such that the large IgG (150 kDa) is positioned tightly into the catalytic site. From the epitope data one may have predicted that the DD9 site may be more effective as it is associated with a putative calcium binding site [21]. However examination of the literature suggests 5 or more putative Ca.sup.2+ binding sites [21] and while it clearly has a dramatic effect, is not critical for all TG2 activity.

(179) The DF4 site would be hypothetically the most effective epitope as the antibody binds to 1 of the essential amino acids in the catalytic triad. However it has not been possible to successfully clone out DF4 producing this inhibitory antibody and as such the production of sufficient IgG to adequately perform IC.sub.50 tests has not been possible. It may in fact be very difficult to clone out antibodies that have too high efficacy given the work from Gunzler et al (1982) FEBS Lett. 150(2): 390-6 that suggested that lymphocytes needed TG2 activity to proliferate and thus antibodies with better inhibitory potential may only be possible using recombinant approaches or a continual IgG extraction system.

(180) One of the most frustrating problems in undertaking this work has been the apparent inability of all antibodies developed to efficiently block non-human TG2 activity, which is critical for preclinical testing. All antibodies reacted with rat and mouse TG2 in both western blot and ELISA, in some cases with little difference in intensity. However out of the 9 antibodies we produced IgG for, it was only possible to determine an IC.sub.50 for 4 in rat and none in mouse. The 4 where an IC.sub.50 was calculated against rat TG2 showed a 30 fold or lower IC.sub.50 against rat TG2 than AB1 against human TG2 meaning any in vivo dose would be prohibitively large. Further none would inhibit at all in a rat cell lysate. Given the reactivity in ELISA and western blots, plus there are just 5 mismatches between species for AB1 and 3 for DD9 the significant species specificity for inhibition was surprising and clearly demonstrates the critical importance of affinity for effective inhibition. Thus having identified these inhibitory epitopes for human TG2 it is now critical that analogue antibodies are developed for these sites in rat TG2 if their value is to be established in in vivo pre clinical models of disease.

(181) There are a wide range of TG inhibitors available. Notably the thiomidazole based compounds originally developed by Merke Sharpe Dome [22] the CBZ-glutamyl analogues developed by Griffin and colleagues [23] which we have used very successfully to treat experimental kidney scarring [16] and the dihydroisoxazole type inhibitors developed by Khosla and collegues [24-27] used successfully in various cancer models. There has been hope that continual refinement of these compounds may yield a viable human TG2 inhibitor, but cross TG family reactivity or the potential toxic nature of the compounds seems to have prevented this. More recently Acylideneoxoindoles have been described as a new reversible class of TG2 inhibitors [24], but data regarding their cross reactivity to other TG family members is lacking. At the 2010 Gordon conference on TG2 in human disease Pasternack and collegues from Zedira presented details of a range of compounds that use side chain Michael acceptors as TG2 inhibitors with claims of suitability for in vivo application and TG2 selectivity, however a full publication on these has not materialised to date. At the same meeting early work from Macdonald et al demonstrated some interesting developments in designing a TG2 inhibitor for treatment of Huntington's Chorea, but again a full publication is still awaited. Undoubtedly a small molecule inhibitor of TG2 would be highly desirable should it be achievable. Tissue penetration, the ability to cross the blood brain barrier, production, cost and easy dosing are just some of the benefits. However, an antibody inhibitor as developed here may in some way be preferable.

(182) TG2 clearly is a multifunction enzyme and has been linked to a range of cellular functions including nuclear stabilisation and transport [28, 29], endocytosis [30, 31], GTPase signalling [32-34], Apoptosis [35, 36], cell adhesion [37-39], cytoskeletal integrity [28, 29] and ECM stabilisation [9]. Clearly a small molecule inhibitor may impede on all of these functions as in general they have free access to the extracellular space and cell interior. An antibody cannot enter the cell and as such the intracellular roles of TG2 would not be affected. Importantly most of the pathological roles of TG2 appear to be extracellular such as its role in tissue scarring and fibrosis, celiac disease and cancer. Thus using an antibody would bring an additional degree of selectivity preventing undesired intracellular effects. Therefore an antibody would offer advantages in blocking TG2 in fibrotic and scarring diseases where TG2 crosslinks ECM proteins, in celiac disease where gliadin is deamidated in the extracellular space and in chemo-resistance in cancer where cell adhesion appears to be the protective factor. However, unless a small Fab fragment could be designed that could cross the blood brain barrier a TG2 inhibiting antibody would be little use in treating neurological pathologies.

(183) In conclusion, for the first time we have been able to develop TG2 inhibitory antibodies that selectively target TG2. We have also identified 3 novel inhibitory epitopes within the core domain of TG2. Humanisation of antibody AB1 will open up the possibility for the first time of targeted TG2 therapy in man.

REFERENCES

(184) 1. Tissue transglutaminase in normal and abnormal wound healing: review article. Verderio, E. A., T. Johnson, and M. Griffin, Amino Acids, 2004. 26(4): p. 387-404. 2. Transglutaminase-mediated cross-linking is involved in the stabilization of extracellular matrix in human liver fibrosis. Grenard, P., S. Bresson-Hadni, S. El Alaoui, M. Chevallier, D. A. Vuitton, and S. Ricard-Blum, J Hepatol, 2001. 35(3): p. 367-75. 3. Changes in transglutaminase activity in an experimental model of pulmonary fibrosis induced by paraquat. Griffin, M., L. L. Smith, and J. Wynne, Br J Exp Pathol, 1979. 60(6): p. 653-61. 4. Cardiac specific overexpression of transglutaminase II (G(h)) results in a unique hypertrophy phenotype independent of phospholipase C activation. Small, K., J. F. Feng, J. Lorenz, E. T. Donnelly, A. Yu, M. J. Im, G. W. Dorn, 2nd, and S. B. Liggett, J Biol Chem, 1999. 274(30): p. 21291-6. 5. Tissue transglutaminase and the progression of human renal scarring. Johnson, T. S., A. F. El-Koraie, N. J. Skill, N. M. Baddour, A. M. El Nahas, M. Njloma, A. G. Adam, and M. Griffin, J Am Soc Nephrol, 2003. 14(8): p. 2052-62. 6. Thrombin upregulates tissue transglutaminase in endothelial cells: a potential role for tissue transglutaminase in stability of atherosclerotic plaque. Auld, G. C., H. Ritchie, L. A. Robbie, and N. A. Booth, Arterioscler Thromb Vasc Biol, 2001. 21(10): p. 1689-94. 7. Cross-linking of fibronectin to collagenous proteins. Mosher, D. F., Mol Cell Biochem, 1984. 58(1-2): p. 63-8. 8. Transglutaminases. Lorand, L. and S. M. Conrad, Mol Cell Biochem, 1984. 58(1-2): p. 9-35. 9. Modulation of tissue transglutaminase in tubular epithelial cells alters extracellular matrix levels: a potential mechanism of tissue scarring. Fisher, M., R. A. Jones, L. Huang, J. L. Haylor, M. El Nahas, M. Griffin, and T. S. Johnson, Matrix Biol, 2009. 28(1): p. 20-31. 10. Transglutaminase transcription and antigen translocation in experimental renal scarring. Johnson, T. S., N. J. Skill, A. M. El Nahas, S. D. Oldroyd, G. L. Thomas, J. A. Douthwaite, J. L. Haylor, and M. Griffin, J Am Soc Nephrol, 1999. 10(10): p. 2146-57. 11. Do changes in transglutaminase activity alter latent transforming growth factor beta activation in experimental diabetic nephropathy? Huang, L., J. L. Haylor, M. Fisher, Z. Hau, A. M. El Nahas, M. Griffin, and T. S. Johnson, Nephrol Dial Transplant, 2010. 25(12): p. 3897-910. 12. Expression induced by interleukin-6 of tissue-type transglutaminase in human hepatoblastoma HepG2 cells. Suto, N., K. Ikura, and R. Sasaki, J Biol Chem, 1993. 268(10): p. 7469-73. 13. TNF-alpha modulates expression of the tissue transglutaminase gene in liver cells. Kuncio, G. S., M. Tsyganskaya, J. Zhu, S. L. Liu, L. Nagy, V. Thomazy, P. J. Davies, and M. A. Zern, Am J Physiol, 1998. 274(2 Pt 1): p. G240-5. 14. Inhibition of transglutaminase activity reduces extracellular matrix accumulation induced by high glucose levels in proximal tubular epithelial cells. Skill, N. J., T. S. Johnson, I. G. Coutts, R. E. Saint, M. Fisher, L. Huang, A. M. El Nahas, R. J. Collighan, and M. Griffin, J Biol Chem, 2004. 279(46): p. 47754-62. 15. Transglutaminase inhibition reduces fibrosis and preserves function in experimental chronic kidney disease. Johnson, T. S., M. Fisher, J. L. Haylor, Z. Hau, N. J. Skill, R. Jones, R. Saint, I. Coutts, M. E. Vickers, A. M. El Nahas, and M. Griffin, J Am Soc Nephrol, 2007. 18(12): p. 3078-88. 16. Transglutaminase inhibition ameliorates experimental diabetic nephropathy. Huang, L., J. L. Haylor, Z. Hau, R. A. Jones, M. E. Vickers, B. Wagner, M. Griffin, R. E. Saint, I. G. Coutts, A. M. El Nahas, and T. S. Johnson, Kidney Int, 2009. 76(4): p. 383-94. 17. Tissue transglutaminase contributes to interstitial renal fibrosis by favoring accumulation of fibrillar collagen through TGF-beta activation and cell infiltration. Shweke, N., N. Boulos, C. Jouanneau, S. Vandermeersch, G. Melino, J. C. Dussaule, C. Chatziantoniou, P. Ronco, and J. J. Boffa, Am J Pathol, 2008. 173(3): p. 631-42. 18. GPR56, an atypical G protein-coupled receptor, binds tissue transglutaminase, TG2, and inhibits melanoma tumor growth and metastasis. Xu, L., S. Begum, J. D. Hearn, and R. O. Hynes, Proc Natl Acad Sci USA, 2006. 103(24): p. 9023-8. 19. Anti-tissue transglutaminase antibodies from celiac patients inhibit transglutaminase activity both in vitro and in situ. Esposito, C., F. Paparo, I. Caputo, M. Rossi, M. Maglio, D. Sblattero, T. Not, R. Porta, S. Auricchio, R. Marzari, and R. Troncone, Gut, 2002. 51(2): p. 177-81. 20. Transglutaminase 2 undergoes a large conformational change upon activation. Pinkas, D. M., P. Strop, A. T. Brunger, and C. Khosla, PLoS Biol, 2007. 5(12): p. e327. 21. Functional significance of five noncanonical Ca2+-binding sites of human transglutaminase 2 characterized by site-directed mutagenesis. Kiraly, R., E. Csosz, T. Kurtan, S. Antus, K. Szigeti, Z. Simon-Vecsei, I. R. Korponay-Szabo, Z. Keresztessy, and L. Fesus, Febs J, 2009. 276(23): p. 7083-96. 22. 3,5 substituted 4,5-dihydroisoxazoles as transglutaminase inhibitors. Syntex, U.S. Pat. No. 4,912,120, 1990. March. 23. Griffin M, Coutts I G, and S. R, Novel Compounds and Methods of Using The Same., in International Publication Number WO 2004/113363, 2004: GB patent PCT/G B2004/002569. 24. Acylideneoxoindoles: a new class of reversible inhibitors of human transglutaminase 2. Klock, C., X. Jin, K. Choi, C. Khosla, P. B. Madrid, A. Spencer, B. C. Raimundo, P. Boardman, G. Lanza, and J. H. Griffin, Bioorg Med Chem Lett. 21(9): p. 2692-6. 25. Transglutaminase 2 inhibitors and their therapeutic role in disease states. Siegel, M. and C. Khosla, Pharmacol Ther, 2007. 115(2): p. 232-45. 26. Structure-based design of alpha-amido aldehyde containing gluten peptide analogues as modulators of HLA-DQ2 and transglutaminase 2. Siegel, M., J. Xia, and C. Khosla, Bioorg Med Chem, 2007. 15(18): p. 6253-61. 27. Novel therapies for celiac disease. Sollid, L. M. and C. Khosla, J Intern Med. 269(6): p. 604-13. 28. Transglutaminase 2: an enigmatic enzyme with diverse functions. Fesus, L. and M. Piacentini, Trends Biochem Sci, 2002. 27(10): p. 534-9. 29. Transglutaminases: crosslinking enzymes with pleiotropic functions. Lorand, L. and R. M. Graham, Nat Rev Mol Cell Biol, 2003. 4(2): p. 140-56. 30. Transglutaminase 2 is needed for the formation of an efficient phagocyte portal in macrophages engulfing apoptotic cells. Toth, B., E. Garabuczi, Z. Sarang, G. Vereb, G. Vamosi, D. Aeschlimann, B. Blasko, B. Becsi, F. Erdodi, A. Lacy-Hulbert, A. Zhang, L. Falasca, R. B. Birge, Z. Balajthy, G. Melino, L. Fesus, and Z. Szondy, J Immunol, 2009. 182(4): p. 2084-92. 31. Transglutaminase is essential in receptor-mediated endocytosis of alpha 2-macroglobulin and polypeptide hormones. Davies, P. J., D. R. Davies, A. Levitzki, F. R. Maxfield, P. Milhaud, M. C. Willingham, and I. H. Pastan, Nature, 1980. 283(5743): p. 162-7. 32. GTP binding and signaling by Gh/transglutaminase II involves distinct residues in a unique GTP-binding pocket. Iismaa, S. E., M. J. Wu, N. Nanda, W. B. Church, and R. M. Graham, J Biol Chem, 2000. 275(24): p. 18259-65. 33. The core domain of the tissue transglutaminase Gh hydrolyzes GTP and ATP. Iismaa, S. E., L. Chung, M. J. Wu, D. C. Teller, V. C. Yee, and R. M. Graham, Biochemistry, 1997. 36(39): p. 11655-64. 34. Gh: a GTP-binding protein with transglutaminase activity and receptor signaling function. Nakaoka, H., D. M. Perez, K. J. Baek, T. Das, A. Husain, K. Misono, M. J. Im, and R. M. Graham, Science, 1994. 264(5165): p. 1593-6. 35. Searching for the function of tissue transglutaminase: its possible involvement in the biochemical pathway of programmed cell death. Fesus, L. and V. Thomazy, Adv Exp Med Biol, 1988. 231: p. 119-34. 36. Induction and activation of tissue transglutaminase during programmed cell death. Fesus, L., V. Thomazy, and A. Falus, FEBS Lett, 1987. 224(1): p. 104-8. 37. Fibronectin-tissue transglutaminase matrix rescues RGD-impaired cell adhesion through syndecan-4 and beta1 integrin co-signaling. Telci, D., Z. Wang, X. Li, E. A. Verderio, M. J. Humphries, M. Baccarini, H. Basaga, and M. Griffin, J Biol Chem, 2008. 283(30): p. 20937-47. 38. Regulated expression of tissue transglutaminase in Swiss 3T3 fibroblasts: effects on the processing of fibronectin, cell attachment, and cell death. Verderio, E., B. Nicholas, S. Gross, and M. Griffin, Exp Cell Res, 1998. 239(1): p. 119-38. 39. A novel RGD-independent cel adhesion pathway mediated by fibronectin-bound tissue transglutaminase rescues cells from anoikis. Verderio, E. A., D. Telci, A. Okoye, G. Melino, and M. Griffin, J Biol Chem, 2003. 278(43): p. 42604-14.

EXAMPLE 2: SEQUENCING OF NOVEL TG2 INHIBITORY ANTIBODIES OF THE INVENTION

(185) Antibody Sequencing

(186) In order to establish the sequences of the variable regions of each antibody of the invention, a pellet of the hybridoma cells was processed using the Qiagen RNeasy Mini Kit to extract the RNA following the manufacturer's protocols. The extracted RNA was reverse transcribed to produce a cDNA using a 1.sup.st Strand cDNA Synthesis Kit (GE Healthcare), using a NotI-dT.sub.18 primer, in accordance with the manufacturer's protocols. The cDNA preparation was cleaned up using the Qiagen PCR Purification Kit, in accordance with the manufacturer's protocols.

(187) To determine the heavy chain sequence, the mouse cDNA was amplified by PCR using a set of degenerate primers (MHV1-12) with a constant region primer (MHCG1, MHCG2A, MHCG2B, MHCG3, or a mixture of the four) as shown in Table 12. Similarly, to determine the light chain sequence, the mouse cDNA was amplified using a set of degenerate primers (MVK1-11) with a constant region primer MKC as shown in Table 13.

(188) If no amplification products were seen using the initial set of Heavy Chain PCR a 5′ RACE PCR (Invitrogen) was carried out, using the NotI-dT.sub.18 primer to generate cDNA; and the constant region primers (MHCG1, MHCG2A, MHCG2B, MHCG3, or a mixture of the four) and the 5′ RACE Anchor Primer, GGCCACGCGTCGACTAGTACGGGIIGGGIIGGGIIG (SEQ ID NO: 138) (where I is the base for deoxyinosine) for the PCR.

(189) The resulting amplification bands were ligated into the pCR2.1®-TOPO® vector using the TOPO-TA Cloning® kit (Invitrogen) using the manufacturer's protocol and sent to GATC Biotech AG for sequencing.

(190) TABLE-US-00027 TABLE 12 PCR Primers for Cloning Mouse Heavy Chain Variable Regions Name Sequence MHV1 ATGAAATGCAGCTGGGGCATCTTCTTC SEQ ID NO: 139 MHV2 ATGGGATGGAGCTRTATCATSYTCTT SEQ ID NO: 140 MHV3 ATGAAGWTGTGGTTAAACTGGGTTTTT SEQ ID NO: 141 MHV4 ATGRACTTTGGGYTCAGCTTGRTTT SEQ ID NO: 142 MHV5 ATGGACTCCAGGCTCAATTTAGTT SEQ ID NO: 143 TTCCTT MHV6 ATGGCTGTCYTRGSGCTRCTCTTCTGC SEQ ID NO: 144 MHV7 ATGGRATGGAGCKGGRTCTTTMTCTT SEQ ID NO: 145 MHV8 ATGAGAGTGCTGATTCTTTTGTG SEQ ID NO: 146 MHV9 ATGGMTTGGGTGTGGAMCTT SEQ ID NO: 147 GCTATTCCTG MHV10 ATGGGCAGACTTACATTCTCATTCCTG SEQ ID NO: 148 MHV11 ATGGATTTTGGGCTGATTTTTTTTATTG SEQ ID NO: 149 MHV12 ATGATGGTGTTAAGTCTTCTGTACCTG SEQ ID NO: 150 MHCG1 CAGTGGATAGACAGATGGGGG SEQ ID NO: 151 MHCG2A CAGTGGATAGACCGATGGGGC SEQ ID NO: 152 MHCG2b CAGTGGATAGACTGATGGGGG SEQ ID NO: 153 MHCG3 CAAGGGATAGACAGATGGGGC SEQ ID NO: 154 Ambiguity codes: R = A or G; Y = C or T; M = A or C; K = G or T; S = G or C; W = A or T. MHV indicates primers that hybridize to the leader sequences of mouse heavy chain variable region genes, MHCG indicates primers that hybridize to the mouse constant region genes.

(191) TABLE-US-00028 TABLE 13 PCR Primers for Cloning Mouse Kappa Light Chain Variable Regions Name Size Sequence MKV1 30-mer ATGAAGTTGVVTGTT SEQ ID NO: 155 AGGCTGTTGGTGCTG MKV2 29-mer ATGGAGWCAGACACA SEQ ID NO: 156 CTCCTGYTATGGGTG MKV3 30-mer ATGAGTGTGCTCACT SEQ ID NO: 157 CAGGTCCTGGSGTTG MKV4 33-mer ATGAGGRCCCCTGCT SEQ ID NO: 158 CAGWTTYTTGGMWTC TTG MKV5 30-mer ATGGATTTWAGGTGC SEQ ID NO: 159 AGATTWTCAGCTTC MKV6 27-mer ATGAGGTKCKKTGKT SEQ ID NO: 160 SAGSTSCTGRGG MKV7 31-mer ATGGGCWTCAAGATG SEQ ID NO: 161 GAGTCACAKWYYCWG G MKV8 31-mer ATGTGGGGAYCTKTT SEQ ID NO: 162 TYCMMTTTTTCAATT G MKV9 25-mer ATGGTRTCCWCASCT SEQ ID NO: 163 CAGTTCCTTG MKV10 27-mer ATGTATATATGTTT SEQ ID NO: 164 GTTGTCTATTTCT MKV11 28-mer ATGGAAGCCCCAGC SEQ ID NO: 165 TCAGCTTCTCTTCC CL12A ATGRAGTYWCAGAC SEQ ID NO: 166 CCAGGTCTTYRT CL12B ATGGAGACACATTC SEQ ID NO: 167 TCAGGTCTTTGT CL13 ATGGATTCACAGGC SEQ ID NO: 168 CCAGGTTCTTAT CL14 ATGATGAGTCCTGC SEQ ID NO: 169 CCAGTTCCTGTT CL15 ATGAATTTGCCTGTT SEQ ID NO: 170 CATCTCTTGGTGCT CL16 ATGGATTTTCAATTG SEQ ID NO: 171 GTCCTCATCTCCTT CL17A ATGAGGTGCCTARC SEQ ID NO: 172 TSAGTTCCTGRG CL17B ATGAAGTACTCTGC SEQ ID NO: 173 TCAGTTTCTAGG CL17C ATGAGGCATTCTCT SEQ ID NO: 174 TCAATTCTTGGG MKC 20-mer ACTGGATGGTGGGA SEQ ID NO: 175 AGATGG Ambiguity codes: R = A or G; Y = C or T; M = A or C; K = G or T; S = G or C; W = A or T. MKV indicates primers that hybridise to leader sequences of the mouse kappa light chain variable region genes, MKC indicates the primer that hybridises to the mouse kappa constant region gene.

(192) Sequence Data

(193) Antibody AB1 was sequenced in addition to Antibodies BB7, DC1, JE12, EH6, AG9, AH3, DD9, DH2, DD6 and IA12. The sequences are provided in FIGS. 18 to 28.

EXAMPLE 3: CONSTRUCTION AND CHARACTERISATION OF CHIMERIC AND HUMANISED NOVEL ANTI-TG2 ANTIBODIES OF THE INVENTION

(194) To further characterise the antibodies of the invention and to enable ranking and prioritisation of antibodies for humanisation, a panel of chimeric TG2 antibodies were constructed (murine variable regions and human IgG1 and human kappa). The methodology used to produce the chimeric antibodies is set out below.

(195) Methods

(196) Human VH and VK cDNA Databases

(197) The protein sequences of human and mouse immunoglobulins from the International Immunogenetics Database 2009.sup.1 and the Kabat Database Release 5 of Sequences of Proteins of Immunological Interest (last update 17 Nov. 1999).sup.2 were used to compile a database of human immunoglobulin sequences in a Kabat alignment. Our database contains 10,606 VH and 2,910 VK sequences.

(198) Molecular Model of AB1

(199) As a representative of the Group 1 antibodies (i.e. antibodies that bind the epitope spanning amino acids 304 to 326 of human TG2), a homology model of the mouse antibody AB1 variable regions has been calculated using the Modeller program.sup.3 run in automatic mode. The atomic coordinates of 1 MQK.pdb, 3LIZ.pdb and 1MQK.pdb were the highest identity sequence templates for the Interface, VL and VH respectively as determined by Blast analysis of the Accelrys antibody pdb structures database. These templates were used to generate 20 initial models, the best of which was refined by modeling each CDR loop with its 3 best loop templates.

(200) hAB1 Framework Selection

(201) The sequence analysis program, gibsSR, was used to interrogate the human VH and VK databases with the AB1 VHc, VKc and VKc1, the BB7 VHc and VKc and the DC1 VHc and VKc protein sequences using various selection criteria. FW residues within 5 Å of a CDR residue (Kabat definition) in the homology model of mouse antibody AB1, were identified, and designated as the “5 Å Proximity” residues.

(202) AF06220 was chosen as the FW on which to base the initial humanised heavy chain versions. Table 14 shows the alignment and residue identity of AF06220 to murine antibodies. Table 15 shows the 5 Å proximity envelope of the sequences. AF062260 has only 1 somatic mutation away from its germline VH gene Z12347 (Table 16).

(203) AY247656 was chosen as the FW on which to base the initial AB1 humanised kappa light chain. The alignment and residue identity to murine AB1 antibody kappa light chain are shown in Table 17; Table 18 shows the 5 Å proximity envelope of the sequences. The sequence shows 5 somatic mutations from its germline VK gene X93620 (Table 19).

(204) AF193851 was chosen as the FW on which to base the other humanised kappa light chain constructs. The alignment and residue identity to the murine antibodies are shown in Table 20. Table 21 shows the 5 Å proximity envelope of the sequences. The sequence shows no somatic mutations from its germline VK gene J00248 (Table 22).

(205) Generation of Expression Vectors

(206) Construction of chimeric expression vectors entails adding a suitable leader sequence to VH and VL, preceded by a Hind III restriction site and a Kozak sequence. The Kozak sequence ensures efficient translation of the variable region sequence. It defines the correct AUG codon from which a ribosome can commence translation, and the most critical base is the adenine at position −3, upstream of the AUG start.

(207) For the heavy chain, the construction of the chimeric expression vectors entails introducing a 5′ fragment of the human γ1 constant region, up to a natural ApaI restriction site, contiguous with the 3′ end of the J region of the variable region. The CH is encoded in the expression vector downstream of the inserted VH sequence but lacks the V-C intron.

(208) For the light chain, the natural splice donor site and a BamHI site is added downstream of the V region. The splice donor sequence facilitates splicing out the kappa V:C intron which is necessary for in-frame attachment of the VL to the constant region.

(209) The DNA sequences of the variable regions were optimized and synthesized by Genet®. The leader sequence has been selected as one that gives good expression of antibody in cultured mammalian cells.

(210) Heavy Chain variable region constructs were excised from the cloning vector using HindIII+ApaI digestion, purified and ligated into the similarly-cut and phosphatase-treated MRCT heavy chain expression vector, and were used to transform TOP10 bacteria.

(211) Kappa chain variable region constructs were excised using HindIII+BamHI digestion, purified, ligated into the similarly-cut and phosphatase treated MRCT kappa light chain expression vector, and were used to transform TOP10 bacteria.

(212) Antibody Expression

(213) A double insert expression vector coding for both Heavy and kappa chains was generated and transfected into HEK293T cells. Cell culture supernatant was purified by affinity chromatography on Protein G-agarose in accordance with the manufacturer's protocols.

(214) Binding ELISA

(215) HEK 293F cells were co-transfected with combinations of different humanised light chain vectors in association with different humanised heavy chain vectors. Recombinant human TG2 was used to measure antibody binding by ELISA. The results indicated that the Heavy Chain version RHA (Table 23), in combination with either Light Chain versions RKE and RKJ (Table 24) (representing the different Light Chain versions humanised) showed optimal binding (FIG. 34), and was therefore selected for further characterization. A similar approach was used to identify optimal pairs of humanized BB7 heavy and light chains, and humanised DC1 heavy and light chains.

(216) Heavy Chain version RHA is an un-modified graft of the mouse CDR regions of the AB1 antibody onto the Human donor sequence. However, both Light Chain versions RKE and RKJ, have the same single 5 Å proximity reside backmutation, F72 (Table 24). This back mutation lies outside the Vernier.sup.4, Canonical.sup.5 or Interface.sup.6 residues.

REFERENCES

(217) 1. Lefranc, M. P. IMGT, the international ImMunoGeneTics Database®. Nucleic Acids Res. 31, 307-310 (2003). 2. Kabat, E. A., Wu, T. T., Perry, H. M., Gottesman, K. S. & Foeller, C. Sequences of Proteins of Immunological Interest. NIH National Technical Information Service, (1991). 3. Eswar, N. et al. Comparative protein structure modeling using Modeller. Curr. Protoc. Bioinformatics. Chapter 5: Unit 5.6., Unit (2006). 4. Foote, J. & Winter, G. (1992). Antibody framework residues affecting the conformation of the hypervariable loops. J Mol. Biol. 224, 487-499. 5. Morea, V., Lesk, A. M. & Tramontano, A. (2000). Antibody modeling: implications for engineering and design. Methods 20, 267-279. 6. Chothia, C., Novotny, J., Bruccoleri, R. & Karplus, M. (1985). Domain association in immunoglobulin molecules. The packing of variable domains. J Mol. Biol. 186, 651-663.

(218) The following table A summarises the chimeric and humanised antibodies produced with a cross reference to the identifiers used in the figures.

(219) TABLE-US-00029 Murine Antibody Chimeric Antibody Humanised Antibody AB1 cAB001 hAB004 (hAB001AE) cAB003 hAB005 (hAB001AJ) BB7 cBB001 hBB001AA hBB001BB DC1 cDC001 hDC001AA hDC001BB DD6 cDD6001 DD9 cDD9001 DH2 cDH001

(220) Tables

(221) TABLE-US-00030 TABLE 14 Kabat Numbers.sup.2   1          10          20         30             40          50               60            70        80              90       100   110   -|---------|---------|---------|-----AB----|---------|--ABC-------|---------|---------|--ABC-------|---------|ABCDEFGHIJK---------|- Vernier.sup.4  -.*.........................****..................***....................*.*.*.*....*.................**.................*.......... Canonical.sup.5  -........................1.11.1....1....................2...22...............2.........................1............................ Interface.sup.6  -...................................I...I.I.....I.I.................................................I.I.I................I.......... 5Å Proximity  ****                 * *  ****        ***        ***                  ****** *     *               ****                 **  CDR                                <----->               <---------------->                                 <--------------> AB_VHc  -EVQLVESGGGLVKPGGSLKLSCAASGFILSSSAMS--WVRQTPDRRLEWVATISV--GGGKTYYPDSVKGRFTISRDNAKNTLYLQMNSLRSEDTAMYYCAKLI------------SLYWGQGTTLTVSS (mAB001VH) (SEQ ID NO: 176_ BB7_VHc  -AVQLVESGGGLVKPGGSLKLSCAASGIIFSSSAMS--WVRQTPEKRLEWVATISS--GGRSTYYPDSVKGRFTVSRDSAKNTLYLQMDSLRSEDTAIYYCAKLI------------SPYWGQGTTLTVSS (mBB7001VH) (SEQ ID NO: 177 DC1_VHc  -EVQLVESGGGLVKPGGSLKLSCAASGFILSTHAMS--WVRQTPEKRLEWVATISS--GGRSTYYPDSVKGRFTISRDNVKNTLYLQLSSLRSEDTAVYFCARLI------------STYWGQGTTLTVSS (mDC001VH) (SEQ ID NO: 178 AF062260  C....L.......Q.....R.........F..Y...--....A.GKG....SA..G--S..S...A.............S...........A....V......DG------------GV......LV.... SEQ ID NO: 179 with reference to AB_VHc SEQ ID NO: 180 with reference to BB7_VHc SEQ ID NO: 181 with reference to DC1_VHc  

(222) Table 14 showing the alignment and residue identity of AF062260 to the murine antibodies. Residue identities are shown by a dot (.) character. Residue differences are shown where applicable. Gaps (-) are used to maintain Kabat numbering, and to show residue insertion or deletion where applicable.

(223) TABLE-US-00031 TABLE 15 5Å Proximity Residues AB_VHc EVQLCAFTLSWVRWVARFTISRNLYCAKWG SEQ ID NO: 182 (mAB001VH) BB7_VHc AVQLCAIIFSWVRWVARFTVSRSLYCAKWG SEQ ID NO: 183 (mBB7001VH) DC1_VHc EVQLCAFTLSWVRWVARFTISRNLFCARWG SEQ ID NO: 184 (mDC001VH AF062260 .......F......S............... SEQ ID NO: 185  with reference to  AB_VHc  SEQ ID NO: 186  with reference to  BB7_VHc  SEQ ID NO: 187  with reference to  DC1_VHc

(224) Table 15 showing the antibody heavy chain framework residues that lie within a 5 Å envelope of the CDR's. Residue identities are shown by a dot (.) character. Residue differences are shown where applicable.

(225) TABLE-US-00032 TABLE 16 ---------10--------20--------30--------40--------50--------60-------- ---------+---------+---------+---------+---------+---------+--------- Z12347.seq EVQLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSAISGSGGSTYYADSVKGRFT AF062260.seq .....................................................................                               <--->              <---------------> 70--------80--------90------- +---------+---------+-------- Z12347.seq ISRDNSKNTLYLQMNSLRAEDTAVYYCAK SEQ ID NO: 119 AF062260.seq ............................R SEQ ID NO: 120

(226) Table 16 showing AF062260 has 1 somatic mutation away from the germline VH gene Z12347. Residue identities are shown by a dot (.) character. Residue differences are shown where applicable.

(227) TABLE-US-00033 TABLE 17 Kabat Numbers.sup.2  1       10        20              30        40        50        60        70        80        90             100 -|--------|---------|-------ABCDEF--|---------|---------|---------|---------|---------|---------|-----ABCDEF----|------A- Vernier.sup.4 -.*.*....................................**.........****..............*.*.**.*................................*.......... Canonical.sup.5 -........................1.......1111..1..............2..22..........2......1...................3....3................... Interface.sup.6 -.........................................I.I.....I..........................................I..I............II.......... 5Å Proximity  ******               **                 **         ****       ***  ***********              **               ** CDR                         <--------------->               <----->                                <-------------> AB_VKc (mAB001VK) -EIVLTQSPSSMYASLGERVTITCKASQ------DINSYLTWFQQKPGKSPKTLIYRTNRLFDGVPSRFSGSGSGQDFFLTISSLEYEDMGIYYCLQYDDFP------YTFGGGTKLEI-K (SEQ ID NO: 188) AY247656 (SEQ ID NO: -..........LS..V.D......Q...------..SN..N.Y......A..L...DASN.ET............T..TF.....QP..F.T...Q..NTY.------L..........-. 189)

(228) Table 17 showing the alignment and residue identity of AY247656 to the murine AB1 antibody. Residue identities are shown by a dot (.) character. Residue differences are shown where applicable. Gaps (-) are used to maintain Kabat numbering, and to show residue insertion or deletion where applicable.

(229) TABLE-US-00034 TABLE 18 5Å Proximity Residues AB_VKc (mAB001VK) EIVLTQTCWFTLIYGVPFSGSGSGQDFFYCFG SEQ ID NO: 190 AY247656 .........YL.............T..T.... SEQ ID NO: 191

(230) Table 18 showing the AB1 antibody kappa light chain framework residues that lie within a 5 Å envelope of the CDR's. Residue identities are shown by a dot (.) character. Residue differences are shown where applicable.

(231) TABLE-US-00035 TABLE 19 ---------+---------+---------+--------+----------+---------+---------          10        20        30       40         50        60          ---------+---------+---------+--------+----------+---------+--------- X93620.seq DIQMTQSPSSLSASVGDRVTITCQASQDISNYLNWYQQKPGKAPKLLIYDASNLETGVPSRFSGSGSGT AY247656.seq E.VL.................................................................                        <--------->               <-----> +---------+---------+------ 70        80        90 X93620.seq +---------+---------+------ AY247656.seq DFTFTISSLQPEDIATYYCQQYDNLPP SEQ ID NO: 125 .............FG.......NTY.L SEQ ID NO: 126                    <-------

(232) Table 19 showing AY247656 has 5 somatic mutation away from the germline VK gene X93620. Residue identities are shown by a dot (.) character. Residue differences are shown where applicable.

(233) TABLE-US-00036 TABLE 20 Kabat Numbers.sup.2  1       10        20              30        40        50        60        70        80        90             100 -|--------|---------|-------ABCDEF--|---------|---------|---------|---------|---------|---------|-----ABCDEF----|------A- Vernier.sup.4 -.*.*....................................**.........****..............*.*.**.*................................*.......... Canonical.sup.5 -.1......................1.......1111..1..............2..22...........2......1..................3....3................... Interface.sup.6 -.........................................I.I.....I..........................................I..I............II.......... 5Å Proximity  ******               **                 **         ****       ***  ***********              **               ** CDR                         <--------------->               <----->                                <-------------> AB_VKc1 (mAB002VK) -DIQMIQSPSSMYASLGERVTITCKASQ------DINSYLTWFQQKPGKSPKTLIYRTNRLFDGVPSRFSGSGSGQDFFLTISSLEYEDMGIYYCLQYDDFP------YTFGGGTKLEI-K (SEQ ID NO: 192) AB1_VKc2 (mAB003VK) -DIQKTQSPSSMYASLGERVTITCKASQ------DINSYLTWFQQKPGKSPKTLIYRTNRLFDGVPSRFSGSGSGQDFFLTISSLEYEDMGIYYCLQYDDFP------YTFGGGTKLEI-K (SEQ ID NO: 193 BB7_VKc (mBB001VK)) -AIKMTQSPSSMYASLGERVIITCKASQ------DINSYLTWFQQKPGKSPKTLIYLINRLMDGVPSRFSGSGSGQEFLLTISGLEHEDMGIYYCLQYVDFP------YTFGGGTKLEI-K (SEQ ID NO: 194 DCl_VKc (mDC001VK) -DITMTQSPSSIYASLGERVTITCKASQ------DINSYLTWFQQKPGKSPKILIYLVNBLVDGVPSRFSGSGSGQDYALTISSLEYEDMGIYYCLQYDDFP------YTFGGGTKLEI-K (SEQ ID NO: 195) AF193851 SEQ ID NO: 196 -...M......LS..V.D......R...------G.RN..A........A..S...AASN.QS............T..T......QP..FAT...Q.HNTY.------W...Q...V..-. with reference to AB_VKcl SEQ ID NO: 197 with refer- ence to ABl_VKc2 SEQ ID NO: 198 with refer- ence to BB7_VKc SEQ ID NO: 199 with refer- ence to DCl_VKc

(234) Table 20 showing the alignment and residue identity of AF193851 to the murine antibodies. Residue identities are shown by a dot (.) character. Residue differences are shown where applicable. Gaps (-) are used to maintain Kabat numbering, and to show residue insertion or deletion where applicable.

(235) TABLE-US-00037 TABLE 21 Å Proximity Residues AB_VKc1 (mAB002VK) DIQMTQTCWFTLIYGVPFSGSGSGQDFFYCFG AB1_VKc2 (mAB003VK) DIQKTQTCWFTLIYGVPFSGSGSGQDFFYCFG SEQ ID NO: 200 BB7_VKc (mBB001VK) AIKMTQTCWFTLIYGVPFSGSGSGQEFLYCFG SEQ ID NO: 201 DC1_VKc (mDC001VK) DITMTQTCWFILIYGVPFSGSGSGQDYAYCFG SEQ ID NO: 202 AF193851 ...M......S.............T..T.... SEQ ID NO: 203 with reference to AB_VKc1 SEQ ID NO: 204 with reference to AB1_VB1_c2 SEQ ID NO: 205 with reference to BB7_VKc SEQ ID NO: 206 with reference to DC1_VKc

(236) Table 21 showing the antibody kappa light chain framework residues that lie within a 5 Å envelope of the CDR's. Residue identities are shown by a dot (.) character. Residue differences are shown where applicable.

(237) TABLE-US-00038 TABLE 22 ---------+---------+---------+---------+---------+---------+---------          10        20        30        40        50        60          ---------+---------+---------+---------+---------+---------+--------- J00248.seq DIQMTQSPSSLSASVGDRVTITCRASQGISNYLAWFQQKPGKAPKSLIYAASSLQSGVPSRFSGSGSGT AF193851.seq .............................R......................N................                        <--------->               <-----> +---------+---------+------ 70        80        90 +---------+---------+------ J00248.seq DFTLTISSLQPEDFATYYCQQYNSYPP SEQ ID NO: 131 AF193851.seq .....................H.T..W SEQ ID NO: 132                    <-------

(238) Table 22 showing AF193851 has no somatic mutation away from the germline VK gene J00248. Residue identities are shown by a dot (.) character. Residue differences are shown where applicable.

(239) TABLE-US-00039 TABLE 23 Kabat Numbers  1       10        20        30          40        50           60        70        80           90       100                    110 -|--------|---------|---------|-----AB----|---------|--ABC-------|---------|---------|--ABC-------|---------|ABCDEFGHIJK---------|- Vernser.sup.4 -.*........................****..................***....................*.*.*.*....*.................**.................*.......... Canonscal.sup.5 -.......................1.11.1....1....................2...22...............2.........................1............................ Interface.sup.6 -..................................I...I.I.....I.I...................................................I.I.I..............I.......... 5Å Proximity  ****                 * *  ****       ***        ***                   ****** *    *                 ****               ** CDR                                <----->               <----------------->                                 <---------------> AB_RHA -EVQLLESGGGLVQPGGSLRLSCAASGFTFSSSAMS--WVRQAPGKGLEWVSTISV--GGGKTYYPDSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCAKLI------------SLYWGQGTLVTVSS (hAB001HA) (SEQ ID NO: 134) BB7_RHA -EVQLLESGGGLVQPGGSLRLSCAASGFTFSSSAMS--WVRQAPGKGLEWVSTISS--GGRSTYYPDSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCAKLI------------SPYWGQGTLVTVSS (hBB001HA) (SEQ ID NO: 64) BB7_RHB -EVQLLESGGGLVQPGGSLRLSCAASGIIFSSSAMS--WVRQAPGKGLEWVATISS--GGRSTYYPDSVKGRFTVSRDSSKNTLYLQMNSLRAEDTAVYYCAKLI------------SPYWGQGTLVTVSS (hBB001HB) (SEQ ID NO: 55) DCl_RHA -EVQLLESGGGLVQPGGSLRLSCAASGFTFSTHAMS--WVRQAPGKGLEWVSTISS--GGRSTYYPDSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCAKLI------------STYWGQGTLVTVSS (hDC001HA) (SEQ ID NO: 74) DCl_RHB -EVQLLESGGGLVQPGGSLRLSCAASGFTLSTHAMS--WVRQAPGKGLEWVATISS--GGRSTYYPDSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYFCARLI------------STYWGQGTLVTVSS (hDC001HB) (SEQ ID NO: 75)

(240) Table 23 showing the sequence alignments of the final humanised versions of AB1, BB7 and DC1 heavy chains. Gaps (-) are used to maintain Kabat numbering, and to show residue insertion or deletion where applicable.

(241) TABLE-US-00040 TABLE 24 Kabat Numbers  1       10        20              30        40        50        60        70        80        90             100 -|--------|---------|-------ABCDEF--|---------|---------|---------|---------|---------|---------|-----ABCDEF----|------A- Vernier.sup.4 -.*.*....................................**.........****..............*.*.**.*................................*.......... Canonical.sup.5 -.1......................1.......1111..1..............2..22...........2......1..................3....3................... Interface.sup.6 -.........................................I.I.....I..........................................I..I............II.......... 5+521 Proximity  ******               **                 **         ****       ***  ***********              **               ** CDR                         <--------------->               <----->                                <-------------> AB_RKE (hAB001KE)(SEQ -EIVLIQSPSSLSASVGDRVTITCKASQ------DINSYLTWYQQKPGKAPKLLIYRTNRLFDGVPSRFSGSGSGTDFFFTISSLQPEDFGTYYCLQYDDFP------YTFGGGTKLEI-K ID NO: 136) AB_RKJ (hAB001KJ) (SEQ -DIQMIQSPSSLSASVGDRVTITCKASQ------DINSYLTWFQQKPGKAPKSLIYRTNRLFDGVPSRFSGSGSGTDFFLTISSLQPEDFATYYCLQYDDFP------YTFGQGTKVEI-K ID NO: 137) BB7_RKA (hBB001KA (SEQ -DIQMIQSPSSLSASVGDRVTITCKASQ------DINSYLTWFQQKPGKAPKSLIYLINRLMDGVPSRFSGSGSGTDFFLTISSLQPEDFATYYCLQYVDFP------YTFGQGTKVEI-K ID NO: 62 BB7_RKB (hBB001KB) (SEQ -DIKMTQSPSSLSASVGDRVTITCKASQ------DINSYLTWFQQKPGKAPKTLIYLINRLMDGVPSRFSGSGSGQEFLLTISSLQPEDFATYYCLQYVDFP------YTFGQGTKVEI-K ID NO: 63) DC1_RKA (hDC001KA) (SEQ -DIQMTQSPSSLSASVGDRVTITCKASQ------DINSYLTWFQQKPGKAPKSLIYLVNRLVDGVPSRFSGSGSGTDFFLTISSLQPEDFATYYCLQYDDEP------YTFGQGTKVEI-K ID NO″ 72) DC1_RKB (hDC001KB) (SEQ -DITMTQSPSSLSASVGDRVTITCKASQ------DINSYLTWFQQKPGKAPKILIYLVNRLVDGVPSRFSGSGSGQDYALTISSLQPEDFATYYCLQYDDEP------YTFGQGTKVEI-K ID NO: 73)

(242) Table 24 showing the sequence alignments of the final humanised versions of AB1, BB7 and DC1 kappa light chains. Gaps (-) are used to maintain Kabat numbering, and to show residue insertion or deletion where applicable.

(243) TABLE-US-00041 TABLE 24A Table 24a summarises the sequence information presented in Tables 23 and 24, in particular showing the sequence of the CDRs, and the CDRs with flanking regions, in the heavy- and light-chains of the AB, BB-7 and DC-1 antibodies. Heavy chain - CDR1 AB CAASGFTFSSSAMSWVR (SEQ ID NO: 207) or FTFSSSAMSWVR (SEQ ID NO: 42) or SSAMS (SEQ ID NO: 10). BB7 RHB CAASGFTFSSSAMSWVR (SEQ ID NO: 207) or FTFSSSAMSWVR (SEQ ID NO: 42) or SSAMS (SEQ ID NO: 10). BB7 RHA CAASGIIFSSSAMSWVR (SEQ ID NO: 208) or IIFSSSAMSWVR (SEQ ID NO: 209) or SSAMS (SEQ ID NO: 10). DC1 RHA CAASGFTFSTHAMSWVR (SEQ ID NO: 201) or FTFSTHAMSWVR (SEQ ID NO: 211) or THAMS (SEQ ID NO: 212). DC1 RHB CAASGFTLSTHAMSWVR (SEQ ID NO: 213) or FTLSTHAMSWVR (SEQ ID NO: 214) or THAMS (SEQ ID NO: 212). Heavy chain - CDR2 AB WVSTISVGGGKTYYPDSVKGRFTISRDNSKNTL (SEQ ID NO: 215) or WVSTISVGGGKTYYPDSVKGRFTISRDN (SEQ ID NO: 216) or WVSTISVGGGKTYYPDSVKGRFTISR (SEQ ID NO: 44) or TISVGGGKTYYPDSVKG (SEQ ID NO: 11). BB7 RHB WVSTISSGGRSTYYPDSVKGRFTISRDNSKNTL (SEQ ID NO: 217) or WVSTISSGGRSTYYPDSVKGRFTISRDN (SEQ ID NO: 218) or WVSTISSGGRSTYYPDSVKGRFTISR (SEQ ID NO: 219) or TISSGGRSTYYPDSVKG (SEQ ID NO: 15). BB7 RHA WVATISSGGRSTYYPDSVKGRFTVSRDSSKNTL (SEQ ID NO: 220) or WVATISSGGRSTYYPDSVKGRFTVSRDS (SEQ ID NO: 221) or WVATISSGGRSTYYPDSVKGRFTVSR (SEQ ID NO: 222) or TISSGGRSTYYPDSVKG (SEQ ID NO: 15). DC1 RHA WVSTISSGGRSTYYPDSVKGRFTISRDNSKNTL (SEQ ID NO: 215) or WVSTISSGGRSTYYPDSVKGRFTISRDN (SEQ ID NO: 218) or WVSTISSGGRSTYYPDSVKGRFTISR (SEQ ID NO: 219) or TISSGGRSTYYPDSVKG (SEQ ID NO: 15). DC1 RHB WVATISSGGRSTYYPDSVKGRFTISRDNSKNTL (SEQ ID NO: 223) or WVATISSGGRSTYYPDSVKGRFTISRDN (SEQ ID NO: 224) or WVATISSGGRSTYYPDSVKGRFTISR (SEQ ID NO: 225) or TISSGGRSTYYPDSVKG (SEQ ID NO: 15). Heavy chain - CDR3 AB YCAKLISLYWG (SEQ ID NO: 32) or LISLY (SEQ ID NO: 12). BB7 RHB YCAKLISPYWG (SEQ ID NO: 226) or LISPY (SEQ ID NO: 16). BB7 RHA YCAKLISPYWG (SEQ ID NO: 226) or LISPY (SEQ ID NO: 16). DC1 RHA YCAKLISTYWG (SEQ ID NO: 227) or LISTY (SEQ ID NO: 19). DC1 RHB FCARLISTYWG (SEQ ID NO: 228) or LISTY (SEQ ID NO: 19). Light chain - CDR1 AB RKE TCKASQDINSYLTWY (SEQ ID NO: 37) or KASQDINSYLT (SEQ ID NO: 7). AB RKJ TCKASQDINSYLTWF (SEQ ID NO: 24) or KASQDINSYLT (SEQ ID NO: 7). BB7 RKB TCKASQDINSYLTWF (SEQ ID NO: 24) or KASQDINSYLT (SEQ ID NO: 7). BB7 RKA TCKASQDINSYLTWF (SEQ ID NO: 24) or KASQDINSYLT (SEQ ID NO: 7). DC1 RKA TCKASQDINSYLTWF (SEQ ID NO: 24) or KASQDINSYLT (SEQ ID NO: 7). DC1 RKB TCKASQDINSYLTWF (SEQ ID NO: 24) or KASQDINSYLT (SEQ ID NO: 7). Light chain - CDR2 AB RKE LLIYRTNRLFDGVPSRFSGSGSGTDFF (SEQ ID NO: 229) or LLIYRTNRLFDGVP (SEQ ID NO: 38) or RTNRLFD (SEQ ID NO: 8) AB RKJ SLIYRTNRLFDGVPSRFSGSGSGTDFF (SEQ ID NO: 230) or SLIYRTNRLFDGVP (SEQ ID NO: 39) or RTNRLFD (SEQ ID NO: 8) BB7 RKB SLIYLTNRLMDGVPSRFSGSGSGTDFF (SEQ ID NO: 231) or SLIYLTNRLMDGVP (SEQ ID NO: 232) or LTNRLMD (SEQ ID NO: 13) BB7 RKA TLIYLTNRLMDGVPSRFSGSGSGQEFL (SEQ ID NO: 233) or TLIYLTNRLMDGVP (SEQ ID NO: 234) or LTNRLMD (SEQ ID NO: 13) DC1 RKA SLIYLVNRLVDGVPSRFSGSGSGTDFF (SEQ ID NO: 235) or SLIYLVNRLVDGVP (SEQ ID NO: 236) or LVNRLVD (SEQ ID NO: 17) DC1 RKB ILIYLVNRLVDGVPSRFSGSGSGQDYA (SEQ ID NO: 237) or ILIYLVNRLVDGVP (SEQ ID NO: 238) or LVNRLVD (SEQ ID NO: 17) Light chain - CDR3 AB RKE YCLQYDDFPYTFG (SEQ ID NO: 27) or LQYDDFPYT (SEQ ID NO: 9). AB RKJ YCLQYDDFPYTFG (SEQ ID NO: 27) or LQYDDFPYT (SEQ ID NO: 9). BB7 RKB YCLQYVDFPYTFG (SEQ ID NO: 239) or LQYVDFPYT (SEQ ID NO: 14). BB7 RKA YCLQYVDFPYTFG (SEQ ID NO: 239) or LQYVDFPYT (SEQ ID NO: 14). DC1 RKA YCLQYDDFPYTFG (SEQ ID NO: 27) or LQYDDFPYT (SEQ ID NO: 9). DC1 RKB YCLQYDDFPYTFG (SEQ ID NO: 27) or LQYDDFPYT (SEQ ID NO: 9).

(244) Characterization of Chimeric and Humanised Antibodies

(245) Chimeric and humanised Abs were assayed for binding to human and cynomolgus monkey TG2 and for enzymatic inhibition of these enzymes according to the methodology described below.

(246) Methods

(247) ELISA Assay for TG2 Binding

(248) Antibody binding to human and cynomolgus monkey TG2 was determined in an ELISA assay. Clear polystyrene “Maxisorp” 96-well plates (Nunc) were coated with 50 ng purified human or cynomogulus monkey TG2 in 50 μl 0.05 M carbonate-bicarbonate buffer pH 9.6 at 4° C. overnight. Control wells were coated with 50 μl 100 μg/ml bovine serum albumin (BSA). Plates were washed 3× with 300 μl phosphate-buffered saline pH7.4 (PBS) containing 0.1% Tween 20 (PBST) and blocked with 300 μl 3% w/v Marvel skimmed milk in PBS for 1 hour at room temperature. After 3× wash with PBST, 50 μl protein-A purified chimeric or humanised anti-TG2 antibodies or human IgG1 kappa isotope control antibody or CUB7402 (Abcam) were serially diluted 4-fold from a top concentration of 50 nM in PBS, and added to the plate in duplicate. After 1 hour at room temperature, the plates were washed 3× in PBST and incubated with 50 μl peroxidase-conjugated goat anti-human IgG (Fc) (Serotec) diluted 1/5,000 in 3% w/v Marvel skimmed milk in PBS or for wells containing CUB7402 peroxidase-conjugated 1/5,000 goat anti-mouse IgG (Fc) (Sigma) for 1 hour at room temperature. After 3× washes with PBST, the plates were developed with 50 μl TMB substrate (Sigma) for 5 min at room temperature before stopping the reaction with 25 μl 0.5M H2SO4 and reading absorbance at 450 nM in a microtiter plate reader (BioTek EL808). Dose response curves were analysed and EC50 values and other statistical parameters determined using a 4-parameter logistical fit of the data (Graph Pad Prism).

(249) Fluorescence-Based Transglutaminase Assay of TG2 Inhibition by Antibodies of the Invention.

(250) Transglutaminase activities of purified human (Zedira) or cynomogolus monkey TG2 enzymes (Trenzyme) were measured by incorporation of dansylated lysine K×D (Zedira) into N,N-dimethylated casein (DMC, Sigma). Human or cynomogolus monkey TG2 were diluted in transamidation buffer (25 mM HEPES pH 7.4 containing 250 mM NaCl, 2 mM MgCl.sub.2, 5 mM CaCl.sub.2, 0.2 mM DTT and 0.05% v/v Pluronic F-127) to 1 nM and 10 nM respectively and mixed with various concentrations of protein-A purified murine, chimeric or humanised TG2 antibodies for 180 min at room temperature in 384-well black microtiter plates (Corning). Reactions were initiated by addition of DMC and K×D to a final concentration of 10 uM and 20 uM respectively and a final reaction volume of 30 ul, and allowed to proceed at RT for 180 min and the increase in fluorescence (RFU) (excitation at 280 nm, emission 550 nm) monitored using a Tecan Safire.sup.2 plate reader. Data were normalised to percentage activity where % activity=(RFU test antibody−RFU low controls)/RFU high controls−RFU low controls)×100, where low controls contained all components except enzyme and high controls contained all components except antibody.

(251) Antibody dose response curves were plotted using GraphPad prism software and fitted using a 4-parameter logistical model to return IC50 and other statistical parameters. The results are illustrated in FIGS. 29 to 33.

(252) Results and Discussion of Enzyme Inhibition and ELISA Binding Experiments by Humanized and Murine Anti-TG2 Antibodies

(253) The ability of chimeric and humanized TG2 antibodies to inhibit transamidation by human TG2 was determined by dose-dependent inhibition of TG2-dependent incorporation of dansylated lysine into N,N-dimethylated casein (exemplified in FIGS. 29 and 31). Both chimeric and humanized antibodies from group 1 (e.g. cAB003, cBB001, cDC001, hBB001AA, hAB001BB, hAB005 and hAB004) show potent inhibition of TG2 activity in the low nanomolar range, consistent with ELISA data that shows binding to immobilized human TG2 in the same range (FIGS. 35 and 37). In contrast, the commercial antibody CUB7402 failed to inhibit human or cynomogulus monkey TG2 enzymatic activities (FIGS. 29F and 30B), despite comparable binding to group 1 antibodies in the ELISA assays (FIGS. 35A, 36A, 37A and 38A) consistent with recognition by CUB7402 of an epitope that does not interfere with the transamidation function of the enzyme. Therefore group 1 antibodies can be distinguished by their ability to inhibit enzyme function from other antibodies such as CUB7402, which bind but have no effect on enzymatic activity. Similarly, murine and chimeric antibodies representative of group 3 (e.g. mDD9001, mDH001, cDD9001 and cDH001) consistently inhibited human TG2 transamidation, but with lower potency than group 1 antibodies (FIGS. 29 and 33). Inhibition of human TG2 by exemplified parent murine monoclonal antibodies from group 1 and group 3 (FIG. 33) show comparable potencies to their chimeric and humanized versions, indicating that the functional potency of the murine antibodies has been retained in the humanized versions. Similarly the human TG2 ELISA binding data for the exemplified humanized antibodies hBB001AA, hBB001BB and hAB004 (FIG. 37) show comparable EC50 values with those obtained for the chimeric versions cBB001 and cAB003, indicating that binding affinity has also been preserved in the humanized versions. Chimeric and humanized antibodies also demonstrate potent inhibition of cynomogulus monkey TG2 (FIGS. 30 and 32) and comparable ELISA EC50's (FIGS. 36 and 38) across the species, consistent with the conservation of the cognate epitope in cynomogulus monkey TG2, which has overall 99% sequence identity. In contrast CUB7402 shows comparable binding to TG2 of both species as the group 1 antibodies, but inhibits neither enzyme activity (FIGS. 35-38 and 29-30).

(254) Cell Based Assays

(255) Binding of antibodies of the invention to extracellular TG2 from HK-2 epithelial cells was assayed using the following protocol.

(256) Measurement of Extracellular TG Activity

(257) Extracellular TG activity was measured by modified cell ELISA. HK-2 epithelial cells were harvested using Accutase and plated at a density of 2×10.sup.4 cells/well in serum free medium onto a 96 well plate that had been coated overnight with 100 μl/well of fibronectin (5 μg/ml in 50 mM Tris-HCl pH 7.4) (Sigma, Poole UK). Cells were allowed to attach for κ/N at 37° C. Media was replaced with DMEM (Life Technologies) and compounds, antibodies or controls were added and allowed to bind at 37° C. 0.1 mM biotin cadaverine [N-(5 amino pentyl biotinamide) trifluoroacetic acid] (Zedira) was added to wells and the plate returned to 37° C. for 2 hours. Plates were washed twice with 3 mM EDTA/PBS and cells removed with 0.1% (w/v) deoxycholate in 5 mM EDTA/PBS. The supernatant was collected and used for protein determination. Plates were washed with PBS/Tween and incorporated biotin cadaverine revealed using 1:5000 extravidin HRP (Sigma, Poole, UK) for 1 h at room temperature followed by K Blue substrate (SkyBio). The reaction was stopped with Red Stop (SkyBio) and the absorbance read at 4 650 nm. Each antibody was tested on at least three separate occasions.

(258) Results are provided in FIGS. 39 and 40 and show an exemplar curve and table of IC50 values obtained for the antibodies tested. FIG. 39 displays the results with Humanised AB1 and FIG. 40 displays the results with Humanised BB7.

(259) hAB005 inhibited the extracellular TG2 of the HK2 cells with an IC50 of 71.85 nM and a maximal inhibition of about 30% control activity. hBB001AA inhibited the activity with an IC50 of 19.8 nM and a maximal inhibition of 40% control activity. hBB001 BB had a better IC50 of 4.9 nM but a maximal inhibition of about 55% control.

(260) Scratch Assays

(261) Scratch wound assays were also performed to assess binding activity of humanised and or chimeric anti-TG2 antibodies of the invention.

(262) TG2 has been shown to have an important role in lung fibrosis and TG2 knockout mice show reduced scarring and fibrosis in the bleomycin model. (Keith C. Olsen, Ramil E. Sapinoro, R. M. Kottmann, Ajit A. Kulkarni, Siiri E. Iismaa, Gail V. W. Johnson, Thomas H. Thatcher, Richard P. Phipps, and Patricia J. Sime. (2011) Transglutaminase 2 and Its Role in Pulmonary Fibrosis. Am. J of Respiratory & Critical Care Med. 184 0699-707) Migration of cells from TG2 knockout mice on wounding was reduced compared to wild type. Scratch wound assay were performed to assess the effect of humanised and/or chimeric anti-TG2 antibodies of the invention on the rate of wound closure in a layer of normal lung fibroblasts (WI-38 cells).

(263) Scratch Assay Protocol:

(264) WI-38 cells (normal human lung fibroblasts ATCC cat #CCL-75) were plated in a 96 well Image Lock plate (Essen cat #4379) at 2×10.sup.4/well in αMEM media (Life Technologies cat #32561) with 10% FBS and grown O/N to >97% confluence. Cells were washed 2× with αMEM media without serum and a scratch wound was generated using an Essen

(265) Wound Maker and the manufacturers protocol. The media was removed and replaced with 95 μl/well serum free media. Controls and test antibodies were added to the wells. The plate was placed in an Essen Incucyte and the closure of the wound was analysed using the scratch wound protocol.

(266) Cytochalasin D was used as an assay control at 0.1 μM. R281, a small molecule non-specific transglutaminase inhibitor, was tested at 100 μM. Z DON, a peptide non reversible transglutaminase inhibitor was tested at 10 μM and 100 μM. The commercially available TG2 antibody Cub7402 (ABcam cat #ab2386) was tested at 5 μg/ml. Antibodies of the invention were tested on at least three occasions at various concentrations as indicated. In all experiments controls were Cytochalasin D at 0.1 uM and ZDON at two concentrations to show a dose dependant effect.

(267) Exemplar results of the scratch assays are shown in FIGS. 41 to 44. As can be seen in FIG. 41, Cytochalasin D, R281 and ZDON all inhibited wound closure (ZDON was shown to inhibit in a dose dependant manner) but the antibody Cub7402 did not inhibit wound closure. Humanised BB7, Humanised AB1 and Chimeric DC1 all inhibited wound closure.

(268) Affinities of Chimeric and Humanised anti TG2 Abs

(269) The binding affinities (Kds and off rates) for a panel of chimeric and humanised Abs of the invention against human TG2 and cyano TG2 were assessed using Biacore techniques. The protocols and results are described below and shown in FIGS. 45 to 47.

(270) Biacore Methods

(271) Recombinant human TG2 was obtained from Zedira GmbH (cat. no.: T002). Recombinant cynomolgus monkey TG2 was obtained from Trenzyme. Surface plasmon resonance (SPR) was measured on a Biacore T200 instrument (GE Healthcare). CM5 chips (GE Healthcare cat. no.: BR-1006-68) were coated with monoclonal mouse anti-human IgG1 (Fc) (MAH) antibody (GE Healthcare cat. no.: BR-1008-39) by amine-coupling as described in the manufacturers instructions. HBS-EP+ buffer (0.01 M HEPES, 0.15 M NaCl, 3 mM EDTA, 0.05% v/v Surfactant P20) and HBS-P+ buffer (0.01 M HEPES, 0.15 M NaCl, 0.05% Surfactant P20) were purchased from GE Healthcare as 10× stocks (cat. nos.: BR-1006-69 and BR-1006-71). Calcium Chloride solution was obtained from Sigma Aldrich (cat. no.: 21115).

(272) The method employed to determine the affinity of the anti-TG2 antibodies involved the capture of the chimeric or humanised antibodies on a MAH coated CM5 chip, followed by the injection of a series of TG2 samples in running buffer. The running buffer was 1× HBS-P+ containing 1 mM CaCl.sub.2, or 1×HBS-EP+ for calcium-free experiments. Antibody capture was carried out for a contact time of 120 seconds at a flow rate of 10 μl/min resulting in the capture of approximately 40-80 RU. TG2 was injected over the immobilised antibody at concentrations ranging from 25 nM to 400 nM with a contact time up to 600 seconds at a flow rate of 30 μl/min. Dissociation of TG2 was typically measured for up to 5400 seconds (1.5 hours). Regeneration of the chip was then performed using 3 M MgCl.sub.2, for a contact time of 60 seconds at a flow rate of 30 μl/ml, followed by a 300 s stabilisation period before the next sample. For each of human and cynomolgus monkey TG2, at least 5 injections at a variety of concentrations were performed in at least two separate experiments.

(273) Kinetic data were exported from the Biacore T200 Evaluation Software and analysed using GraphPad Prism, where the association phases and dissociation phases were analysed separately using a one phase association model and one-phase exponential decay model respectively. Association rates (k.sub.on) were calculated for each curve individually, and dissociation rates (k.sub.off) values from the long dissociation phase data collected. Where k.sub.off values were calculated to be <1×10.sup.−5 s.sup.−1, values were set at 1×10.sup.−5 for analysis, as rates slower than this could not be estimated accurately. Values for k.sub.on and k.sub.off are presented in the tables below as the mean of the individual calculated values for each antibody for each TG2 species from multiple concentrations+/−1 standard deviation. K.sub.D values are calculated as mean k.sub.off/mean k.sub.on.

(274) Results of Biacore Experiments

(275) TABLE-US-00042 TABLE 25 Human TG2 k.sub.off k.sub.on K.sub.D Antibody (s.sup.−1) st. dev (M.sup.−1s.sup.−1) st. dev (M) cAB003 +Ca.sup.2+ <10.sup.−5 — 1.7 × 10.sup.5 3.2 × 10.sup.4 <6 × 10.sup.−11 −Ca.sup.2+ <10.sup.−5 — 8.6 × 10.sup.4 2.1 × 10.sup.4 <1 × 10.sup.−10 cBB001 +Ca.sup.2+ <10.sup.−5 — 2.1 × 10.sup.5 6.9 × 10.sup.4 <5 × 10.sup.−11 −Ca.sup.2+ <10.sup.−5 — 1.5 × 10.sup.5 1.8 × 10.sup.4 <7 × 10.sup.−11 hAB004 (hAB001AE) 2.4 × 10.sup.−5 1.0 × 10.sup.−5 2.0 × 10.sup.5 1.4 × 10.sup.5 1.2 × 10.sup.−10  hAB005 (hAB001AJ) <10.sup.−5 — 1.9 × 10.sup.5 6.8 × 10.sup.4 <5 × 10.sup.−11 hBB001AA <10.sup.−5 — 2.7 × 10.sup.5 1.1 × 10.sup.5 <4 × 10.sup.−11 hBB001BB <10.sup.−5 — 2.4 × 10.sup.5 1.2 × 10.sup.5 <4 × 10.sup.−11 cDC001 <10.sup.−5 — 3.2 × 10.sup.5 5.7 × 10.sup.4 <3 × 10.sup.−11 cDH001 +Ca.sup.2+ 1.8 × 10.sup.−5 4.9 × 10.sup.−6 2.8 × 10.sup.4 1.3 × 10.sup.4 6.4 × 10.sup.−10  −Ca.sup.2+ 4.9 × 10.sup.−4 1.6 × 10.sup.−5 2.1 × 10.sup.4 1.2 × 10.sup.4 2.3 × 10.sup.−8   cDD9001 +Ca.sup.2+ 1.3 × 10.sup.−5 1.8 × 10.sup.−6 3.0 × 10.sup.4 2.3 × 10.sup.4 4.3 × 10.sup.−10  −Ca.sup.2+ 7.1 × 10.sup.−5 1.6 × 10.sup.−5 2.3 × 10.sup.4 7.6 × 10.sup.3 3.1 × 10.sup.−9  

(276) Table 25 shows the kinetic data obtained against human TG2. Where k.sub.off rates were calculated to be less than 10.sup.−5 s.sup.−1, values were set to 10.sup.−5 s.sup.−1 for analysis, as rates slower than this could not be accurately determined.

(277) TABLE-US-00043 TABLE 26 Cynomolgus TG2 k.sub.off k.sub.on K.sub.D Antibody (s.sup.−1) st. dev (M.sup.−1s.sup.−1) st. dev (M) cAB003 +Ca.sup.2+ 1.2 × 10.sup.−5 8.1 × 10.sup.−6 2.4 × 10.sup.5 1.1 × 10.sup.5 5.0 × 10.sup.−11 −Ca.sup.2+ 1.3 × 10.sup.−5 1.5 × 10.sup.−6 1.4 × 10.sup.5 1.9 × 10.sup.4 9.3 × 10.sup.−11 cBB001 +Ca.sup.2+ <10.sup.−5 — 2.9 × 10.sup.5 9.0 × 10.sup.4  <3 × 10.sup.−11 −Ca.sup.2+ <10.sup.−5 — 1.9 × 10.sup.5 2.1 × 10.sup.4  <5 × 10.sup.−11 hAB004 (hAB001AE) 1.8 × 10.sup.−5 2.3 × 10.sup.−6 1.6 × 10.sup.5 7.3 × 10.sup.4 1.1 × 10.sup.−11 hAB005 (hAB001AJ) 3.4 × 10.sup.−5 4.6 × 10.sup.−6 1.4 × 10.sup.5 4.1 × 10.sup.4 2.4 × 10.sup.−10 hBB001AA <10.sup.−5 — 3.0 × 10.sup.5 1.4 × 10.sup.5  <3 × 10.sup.−11 hBB001BB 3.7 × 10.sup.−5 3.4 × 10.sup.−6 2.7 × 10.sup.5 1.3 × 10.sup.5 1.4 × 10.sup.−10 cDC001 <10.sup.−5 — 4.2 × 10.sup.5 3.3 × 10.sup.5 2.4 × 10.sup.−11 cDH001 +Ca.sup.2+ <10.sup.−5 — 1.5 × 10.sup.4 9.5 × 10.sup.3  <7 × 10.sup.−10 −Ca.sup.2+ 6.3 × 10.sup.−5 1.2 × 10.sup.−5 2.6 × 10.sup.4 1.0 × 10.sup.4 2.4 × 10.sup.−9  cDD9001 +Ca.sup.2+ 1.6 × 10.sup.−5 1.4 × 10.sup.−5 3.1 × 10.sup.4 1.6 × 10.sup.4 5.2 × 10.sup.−10 −Ca.sup.2+ 4.1 × 10.sup.−5 1.6 × 10.sup.−5 4.6 × 10.sup.4 1.3 × 10.sup.4 8.9 × 10.sup.−10

(278) Table 26 shows the kinetic data obtained against cynomolgus TG2. Where k.sub.off rates were calculated to be less than 10.sup.−5 s.sup.−1, values were set to 10.sup.−5 s.sup.−1 for analysis, as rates slower than this could not be accurately determined.

(279) FIGS. 45 to 47 provide example Biacore data sets. As can be seen, the humanised and chimeric antibodies cAB003, cBB001, hAB004, hAB005, hBB001AA, hBB001BB, cDC001 for TG2 have excellent affinity for human and cynomolgus TG2, with K.sub.D values of 120 pM or better. The chimeric antibodies cDH001 and cDD9001 exhibit slower association rates to human and cynomolgus TG2 and weaker overall affinity.

(280) Examination of a selection of antibodies in the absence of calcium shows that there is little or no effect, except in the case of cDH001 and cDD9001, where binding is weaker due to faster dissociation rates (k.sub.off).