Polynucleotides encoding IL33 antibodies and methods of using the same

Abstract

The present invention provides isolated IL-33 proteins, active fragments thereof and antibodies, antigen binding fragments thereof, against IL-33 proteins. Also provided are methods of modulating cytokine activity, e.g., for the purpose of treating immune and inflammatory disorders.

Claims

1. A polynucleotide encoding an antibody or antigen binding fragment thereof which specifically binds to IL-33, wherein the antibody or antigen binding fragment thereof comprises a variable heavy domain (VH) comprising a VHCDR1 having the sequence of SEQ ID NO: 543, a VHCDR2 having the sequence of SEQ ID NO: 544, and a VHCDR3 having the sequence of SEQ ID NO: 545, and a variable light domain (VL) comprising a VLCDR1 having the sequence of SEQ ID NO: 548, a VLCDR2 having the sequence of SEQ ID NO: 549, and a VLCDR3 having the sequence of SEQ ID NO: 550.

2. The polynucleotide according to claim 1, wherein the VH and VL of said antibody or antigen-binding fragment thereof comprise amino acid sequences at least 95%, 90%, or 85% identical to SEQ ID NO: 542 and SEQ ID NO: 547, respectively.

3. The polynucleotide according to claim 2, wherein said antibody or antigen-binding fragment thereof comprises a VH having the sequence of SEQ ID NO: 542 and a VL having the sequence of SEQ ID NO: 547.

4. The polynucleotide according to claim 1, wherein the VH and VL of said antibody or antigen-binding fragment thereof comprise amino acid sequences at least 95%, 90%, or 85% identical to SEQ ID NO: 616 and SEQ ID NO: 618, respectively.

5. The polynucleotide according to claim 4, wherein said antibody or antigen-binding fragment thereof comprises a VH having the sequence of SEQ ID NO: 616 and a VL having the sequence of SEQ ID NO: 618.

6. The polynucleotide of any one of claim 1, 2, 3, 4, or 5, wherein said antibody or antigen-binding fragment thereof is selected from the group consisting of a human antibody, a chimeric antibody, and a humanized antibody.

7. The polynucleotide of any one of claim 1, 2, 3, 4, or 5, wherein said antibody or antigen-binding fragment thereof is selected from the group consisting of a naturally-occurring antibody, an scFv fragment, an Fab fragment, an F(ab′)2 fragment, a minibody, a diabody, a triabody, a tetrabody, and a single chain antibody.

8. The polynucleotide of any one of claim 1, 2, 3, 4, or 5, wherein said antibody or antigen-binding fragment thereof is a monoclonal antibody.

9. A vector comprising the polynucleotide of any one of claim 1, 2, 3, 4, or 5.

10. A composition comprising the polynucleotide of any one of claim 1, 2, 3, 4, or 5.

11. A host cell comprising the polynucleotide of any one of claim 1, 2, 3, 4, or 5.

12. A host cell comprising at least a first and a second vector, wherein said first and said second vectors are non-identical, wherein said first vector comprises a polynucleotide encoding SEQ ID NO: 542, which encodes an immunoglobulin heavy chain variable region, and wherein said second vector comprises a polynucleotide encoding SEQ ID NO: 547, which encodes an immunoglobulin light chain variable region.

13. A method of producing an anti-IL33 antibody or antigen-binding fragment thereof, comprising culturing the host cell of claim 12, and recovering said antibody or antigen-binding fragment thereof.

14. A host cell comprising at least a first and a second vector, wherein said first and said second vectors are non-identical, wherein said first vector comprises a polynucleotide encoding SEQ ID NO: 616, which encodes an immunoglobulin heavy chain variable region, and wherein said second vector comprises a polynucleotide encoding SEQ ID NO: 618, which encodes an immunoglobulin light chain variable region.

15. A method of producing an anti-IL33 antibody or antigen-binding fragment thereof, comprising culturing the host cell of claim 14, and recovering said antibody or antigen-binding fragment thereof.

Description

DESCRIPTION OF THE FIGURES

(1) FIG. 1 Shows a HTRF assay for human IL33-ST2 binding in the presence of unpurified scFv periplasmic preparations. A well containing antibody IL330004 is highlighted.

(2) FIG. 2 Shows neutralization of human (A) and cynomolgus (B) IL33 by purified scFv preparations in an IL33-ST2 HTRF assay.

(3) FIG. 3 Shows neutralization of human (A) and cynomolgus (B) IL33 by purified IgG preparations in an IL33-ST2 HTRF assay.

(4) FIG. 4 Shows neutralization of human IL33 by purified IgG preparations in a luciferase NFκB reporter assay (A) and a HUVEC NFκB translocation assay (B).

(5) FIG. 5 Shows detection of endogenous IL-33 in bronchial smooth muscle cells by immunofluorescence staining by IL-33 antibody IL330004 (right panel) compared to CAT-002 negative control (left panel).

(6) FIG. 6 Shows binding data from a single plate screened against human IL-33, cynomolgus IL-33 and insulin. One specific human/cynomolgus cross-reactive IL-33 binder is shown in well C4, and wells A12 and B12 contain control IL-33 binding clone.

(7) FIG. 7 Shows neutralizing activity of antibodies in a TF-1 proliferation assay (A) and in a HUVEC IL-6 production assay (B).

(8) FIG. 8 Shows neutralizing activity of antibodies in a mast cell IL-6 (A), GM-CSF (B), IL-10 (C), IL-8 (D), IL-13 (E) and IL-5 (F) cytokine production assay.

(9) FIG. 9 Shows binding of IL-33 antibodies (IL330065, IL330101, IL330107, and IL330149) to full length human IL-33 by Western blot.

(10) FIG. 10 Shows neutralizing activity of anti-IL-33 antibodies IL330065 and IL330101 on mast cell IL-6 (A) and IL-13 (B) production stimulated by full length IL-33 cell lysates.

(11) FIG. 11 Shows a HTRF® receptor-ligand competition assay in the presence of antibodies IL330065, IL330099, IL330101, IL330107, IL33149, and IL330180 (A) and Huvec NFkB (p65/RelA) translocation assay in the presence of antibodies IL330065, IL330099, IL330101, IL330107, and IL330149 (B).

(12) FIG. 12 Shows competitive binding of purified scFv preparations with mAb IL330101 for binding to biotinylated human IL-33 (A) and competitive binding of purified scFv preparations with mAb IL330180 for binding to biotinylated human IL-33 (B).

(13) FIG. 13 Shows competitive binding IL330259 scFv with mAb IL330101 for binding to biotinylated human IL-33.

(14) FIG. 14 Shows competitive binding IL330259 IgG with mAb IL330101 for binding to biotinylated human IL-33 (A) and competitive binding antibodies with mAb H338L293 for binding to biotinylated human IL-33 (B).

(15) FIG. 15 Shows neutralizing activity of antibodies in HUVEC (A) or mast cell (B) IL-6 production assays.

(16) FIG. 16 Shows a Schild analysis of IL330388 and H338L293 in a mast cell IL-6 production assay.

(17) FIG. 17 Shows the activity of human IL-33 (A), cysteine-biotinylated IL-33 (B) or cell culture media-pretreated IL-33 (C), measured in HUVEC signaling assays (30 minutes) and IL-6 production assay (18-24 hours).

(18) FIG. 18 Shows SDS-PAGE of: human IL-33 (BK349) before or after treatment with Iscoves Modified Dulbeccos Medium (IMDM) (A), of human IL-33 Flag® His, non-biotinylated versus biotinylated under non-reducing conditions (B), of the mouse IL-33 Flag® His after treatment with IMDM (C).

(19) FIG. 19 Shows the purification of cell culture media-treated human IL-33 by SEC.

(20) FIG. 20 Shows the intact mass of PBS versus media-treated IL-33 determined by LC-MS. IMDM-treated IL-33 displayed a 4 Da loss compared with PBS-treated compatible with the formation of two disulphide bonds.

(21) FIG. 21 Shows the disulphide mapping of media-treated human IL-33. Data were consistent with the formation of two disulphide bridges. A shows combined, deconvoluted mass spectra from non-reduced and reduced Lys-C peptide mapping analysis of DSB IL-33. B shows isolated spectra for cysteine containing peptides. C shows sequences of disulphide bonded peptides identified by non-reduced and reduced Lys-C peptide mapping analysis of disulphide bonded IL-33. Disulphide linkages are represented by two hyphens (--). Lys-C miscleavages are represented by square brackets.

(22) FIG. 22 Shows SDS-PAGE analysis of high concentration redIL-33 and DSB IL-33 for NMR analysis (A), NMR heteronuclear multiple quantum coherence (HMQC) analysis with overlay of the .sup.1H-.sup.15N HMQC spectra for .sup.15N-labeled human IL-33 for redIL-33 and DSB IL-33 (B), near-UV circular dichroism (CD) Spectrum (C), key features of IL-33 (Trp193, cysteines, and ST2 binding site) indicated within the solved IL-33 structure (D) and Far-UV circular dichroism (CD) Spectrum (E).

(23) FIG. 23 Shows hydrogen-deuterium exchange analysis of redIL-33 and DSB IL-33. Differences in deuterium incorporation are mapped onto the published IL-33 structure. (A) shows the comparison of fractional hydrogen exchange (for deuterium) in reduced IL-33 (left panel) and DSB IL-33 (right panel). (B) shows a structural model of differential HX-MS data overlaid with the ST2 binding site (red and magenta).

(24) FIG. 24 Shows redIL-33 (A) or DSB IL-33 (B) binding to ST2.

(25) FIG. 25 Shows analysis of three commercial IL-33 ELISA assays for detection of redIL-33 and DSB IL-33 forms. A and B show that two commercial human IL-33 assays predominantly detect the disulphide-bonded form of IL-33 (IL33-DSB). C shows that the mouse IL-33 assay detects both reduced and oxidized forms of mouse IL-33.

(26) FIG. 26 Shows ELISA assays that are specific for detection of redIL-33. (A) shows signal intensity as a function of DSB IL-33 or red-IL-33 concentration using IL330004 as a capture probe and IL330425 as a detection probe. (B) shows signal intensity as a function of DSB IL-33 or red-IL-33 concentration using IL330425 as a capture probe and biotinylated sST2.Fc as a detection probe.

(27) FIG. 27 Shows a time course of human IL-33 incubated in cell culture media (IMDM) or human serum. redIL-33 or DSB IL-33 forms are measured by ELISA (A) or Western blot (B).

(28) FIG. 28 Shows analysis of BALF from humanized IL-33 mice collected at varying timepoints following Alternaria intranasal challenge, using a combination of multiple ELISA assays. (A) Millipore, (B) R&D systems and (C) IL330425/sST2-biotin assays were used to measure IL-33 in the presence or absence of sST2 (left hand graphs). Signals in the presence of sST2 (signal from the reduced IL-33 fraction eliminated) were compared with a disulphide bonded IL-33 standard to quantify the levels of disulphide bonded IL-33. The reduced IL-33 signal was calculated as the difference in signal between IL-33 measurements in the presence and absence of ST2, quantified against a reduced IL-33 standard. Estimations for reduced IL-33 are shown on the right hand graphs.

(29) FIG. 29 Shows analysis of BALF from wild type BALB/c mice collected at varying timepoints following Alternaria intranasal challenge. Mouse IL-33 ELISA (R&D systems) was used to measure IL-33 in the presence or absence of sST2 (media-treated mouse IL-33 used as standard curve) (A). Signals in the presence of sST2 (signal from the reduced IL-33 fraction eliminated) were compared with a media-treated mouse IL-33 standard to quantify the levels of oxidised IL-33. The reduced IL-33 signal was calculated as the difference in signal between IL-33 measurements in the presence and absence of ST2, quantified against a reduced mouse IL-33 standard (B).

(30) FIG. 30 Shows relative fluorescence units at 100 minutes following incubation of 5 uM antibody with 20 uM IL33 at 25° C. in the presence of 8×SYPRO orange dye. In the presence of IL-33 H338L293, but not IL330004 or the control mAb, the increased fluorescent signal is indicative of protein unfolding (A). (B) shows relative fluorescence units over time following incubation of varying concentrations of H338L293 with 20 uM IL33 at 25° C. in the presence of 8×SYPRO orange dye. Fluorescent signal increased with increasing antibody concentration. (C) Shows SDS-PAGE analysis of IL-33. Preincubation of IL-33 with H338L293, but not control mAb or no mAb, increased the presence of the faster migrating, disulfide bonded form of IL-33 under non-reducing conditions.

(31) FIG. 31 Shows a timecourse of neutralization of IL-33 stimulated NFkB signaling in HUVECs with mAb H338L293. (A) shows NFkB signaling 30 minutes following stimulation. (B) shows NFkB signaling 6 hours following stimulation.

(32) FIG. 32 Shows the inhibition of the FRET signal, produced by human IL-33 binding to human ST2 with increasing concentrations of H338L293 under directly competitive conditions (B) or following preincubation with IL-33 (C). A shows a schematic representation of the assay.

(33) FIG. 33 Shows epitope mapping of H338L293. A shows SEC analysis of IL33:IgG complexes with H338L293 pre and post digestion with trypsin. B shows the truncate peptide that was determined to bind strongly to H338L293 coloured black within the IL-33 structure described by Lingel et al 2009.

(34) FIG. 34 Shows NFkB signaling activity of wild type IL-33 (IL33-01—A) and IL-33 cysteine to serine mutants (IL33-02—B, IL33-03—C, IL33-04—D, IL33-05—E, IL33-06—F, IL33-07—G, IL33-08—H, IL33-09—I, IL33-10—J, IL33-11—K, IL33-12—L, IL33-13—M, IL33-14—N, IL33-15—O, IL33-16—P) before and after treatment for 18 hours with IMDM cell culture media.

(35) FIG. 35 Shows that IL33-11 has greater potency than IL33-01 (WT) in vitro (A) and in vivo (B).

(36) FIG. 36 Shows overlay of the .sup.1H-.sup.15N HMQC spectra for 0.1 mM .sup.15N-labeled IL33-01 and IL33-11 plotted in black and red, respectively. Assignment for relevant residues are indicated. Data show peak shifts around C208 and C259 as expected. However, there are more more peak shifts than expected from T185 to A196 which might indicate a conformation change.

(37) FIG. 37 Shows a timecourse of IL-33 neutralizing activity of 33v20064 scFv in an IL33-ST2 HTRF binding assay. (A) shows the inhibition of the FRET signal after 1 hour incubation produced by human IL-33-01 binding to human ST2 with increasing concentrations of IL-33 scFv antibody 33v20064. (B) shows the inhibition of the FRET signal after 1 hour incubation produced by human IL-33-11 binding to human ST2 with increasing concentrations of IL-33 scFv antibody 33v20064. (C) shows the inhibition of the FRET signal after overnight incubation produced by human IL-33-01 binding to human ST2 with increasing concentrations of IL-33 scFv antibody 33v20064. (D) shows the inhibition of the FRET signal after overnight incubation produced by human IL-33-11 binding to human ST2 with increasing concentrations of IL-33 scFv antibody 33v20064.

(38) FIG. 38 Shows a timecourse of IL-33 neutralizing activity of 33v20064 IgG in an IL33-ST2 HTRF binding assay. (A) shows the inhibition of the FRET signal after 1 hour incubation, produced by human IL-33 binding to human ST2 with increasing concentrations of 33v20064 IgG1 antibody. (B) shows the inhibition of the FRET signal after overnight incubation, produced by human IL-33 binding to human ST2 with increasing concentrations of 33v20064 IgG1 antibody.

(39) FIG. 39 Shows antibody neutralization of IL-6 production in HUVECs stimulated by wild type (A) or mutant IL-33 (B).

(40) FIG. 40 Shows the inhibition of the FRET signal, produced by biotinylated human IL-33-01 binding to DyLight labelled 33v20064, with increasing concentrations of mouse IL-33 FH and cyno IL-33 FH (A), human IL-33 FH and B7H3 avi his (A and B), and human IL-1 beta and human IL-1 alpha (B). Inhibition of the signal corresponds with relative binding affinity of 33v20064 to the test protein.

(41) FIG. 41 Shows a timecourse of IL-33 neutralizing activity in an IL33-ST2 HTRF binding assay of germline variant 33_640001 compared with parent 33v20064 scFv. (A) shows the inhibition of the FRET signal after 1 hour incubation. (B) shows the inhibition of the FRET signal after overnight incubation.

(42) FIG. 42 Shows antibody neutralization of IL-8 production in HUVECs stimulated by truncated (112-270) (A) or full length (1-270) IL-33 (B).

(43) FIG. 43 Shows the effect of IL-33 binding proteins on conversion from redIL-33 to DSB IL-33 in IMDM+1% BSA (A) or PBS+1% BSA (B).

(44) FIG. 44 Shows the inhibition of the FRET signal, produced by biotinylated human IL-33 after 1 hour incubation (A) or after overnight incubation (B), or cynomolgus IL-33 binding to 33_640117 mAb (C), with increasing concentrations of test proteins. Inhibition of the signal corresponds with relative binding affinity of 33v20064 to the test protein.

(45) FIG. 45 Shows antibody neutralization of IL-8 production in HUVECs stimulated by truncated (112-270) (A) or full length (1-270) IL-33 (B).

(46) FIG. 46 Shows that H338L293 dose-dependently inhibits Alternaria (ALT)-induced BAL IL-5 and eosinophilia in wild type BALB/c mice. Test substances were dosed intranasally (10, 30 or 100 mg/kg as indicated in brackets) at −2 hours prior to challenge with 25 ug of ALT. BALF was harvested at 24 hours post ALT challenge and analysed for presence of IL-5 (A) and eosinophils (B). Significant effect of test substances was determined using one-way ANOVA with Bonferroni's multiple comparisons test. ***p<0.001, ˜˜p<0.01 compared to control mAb (n=4-8). Mouse IL-33 Trap was used as a positive control.

(47) FIG. 47 Shows that H338L293 (30 mg/kg) and mouse IL-33 Trap (10 mg/kg), but not IL330004 (30 mg/kg), inhibit ALT-induced BAL IL-5 in humanized IL-33 mice. Test substances were dosed intranasally at −2 hours prior to challenge with 25 ug of ALT. BALF was harvested at 24 hours post ALT challenge and analysed for presence of IL-5. Significant effect of test substances was determined using one-way ANOVA with Bonferroni's multiple comparisons test. ***p<0.001, **p<0.01 (n=4).

(48) FIG. 48 Shows that 33_640050 dose dependently inhibits Alternaria-induced BAL IL-5 in humanized IL-33 mice. Test substances were dosed intraperitoneally (0.3, 3 or 30 mg/kg as indicated in brackets) at −24 hours prior to challenge with 25 ug of Alternaria. BALF was harvested at 24 hours post ALT challenge and analysed for presence of IL-5. Significant effect of test substances was determined using one-way ANOVA with Bonferroni's multiple comparisons test. ***p<0.001, **p<0.01 (n=4-5).

(49) FIG. 49 Shows effect of antibodies in an IL33-ST2 FRET binding assay (A) and on IL-33 stimulated IL-8 release from Huvecs (B).

(50) FIG. 50 Shows mAb specificity for IL-33 from various species or other IL-1 family members using a FRET assay based on (A) IL33/33_640087-7B or (B) 33_640237-2B.

(51) FIG. 51 Shows effect of antibodies on IL-8 release from Huvecs stimulated by human lung lysate.

(52) FIG. 52 Shows that 33_640087-7B dose dependently inhibits Alternaria-induced BAL IL-5 in humanized IL-33 mice.

(53) FIG. 53 Experimental design of a pilot in vivo study to investigate the potential for activity IL-33 independent of ST2 (A). Analysis of human IL-33 exposure in BAL fluid following repeated administration of human IL-33 to BALB/c mice (B). Analysis of IL-33 exposure in plasma following a single intraperitoneal administration of human IL-33 (10 ug) (C). Analysis of IL-33 exposure in plasma following repeated administration of human IL-33 to BALB/c mice (D).

(54) FIG. 54 Shows representative H&E stained paraffin sections of lung tissue from mice administered (A) PBS or (B) IL-33 intranasally for 6 weeks.

(55) FIG. 55 Shows (A) p-p38 MAPK or (B) p-STAT5 nuclear translocation activity in Huvecs in response to reduced IL-33 or DSB IL-33 and Western blot analysis of p-p38 MAPK, p-JAK2 and p-STAT5 in Huvecs stimulated for 15 minutes with reduced IL-33 or DSB IL-33 (C).

(56) FIG. 56 (A) Shows binding of RAGE-Fc to reduced or DSB IL-33 by ELISA. (B) Shows inhibition of the Huvec pSTAT5 response to DSB IL-33 with RAGE-Fc or anti-RAGE mAb. (C) Shows inhibition of the Huvec pSTAT5 response to DSB IL-33 with anti-RAGE mAb.

(57) FIG. 57 Shows the effect of anti-IL-33 versus anti-ST2 on the pSTAT5 response in Huvecs.

(58) FIG. 58 Shows the effect of anti-IL-33, anti-ST2 or anti-RAGE mAbs on IL-33 induced inhibition of A549 cell migration versus (A) reduced IL-33 (B) DSB IL-33.

SUMMARY OF THE SEQUENCES

(59) TABLE-US-00005 SEQ ID NO: Reference/Antibody Name Description 1 IL330002 VH DNA 2 IL330002 VH PRT 3 IL330002 VH CDR1 PRT 4 IL330002 VH CDR2 PRT 5 IL330002 VH CDR3 PRT 6 IL330002 VL DNA 7 IL330002 VL PRT 8 IL330002 VL CDR1 PRT 9 IL330002 VL CDR2 PRT 10 IL330002 VL CDR3 PRT 11 IL330004 VH DNA 12 IL330004 VH PRT 13 IL330004 VH CDR1 PRT 14 IL330004 VH CDR2 PRT 15 IL330004 VH CDR3 PRT 16 IL330004 VL DNA 17 IL330004 VL PRT 18 IL330004 VL CDR1 PRT 19 IL330004 VL CDR2 PRT 20 IL330004 VL CDR3 PRT 21 IL330020 VH DNA 22 IL330020 VH PRT 23 IL330020 VH CDR1 PRT 24 IL330020 VH CDR2 PRT 25 IL330020 VH CDR3 PRT 26 IL330020 VL DNA 27 IL330020 VL PRT 28 IL330020 VL CDR1 PRT 29 IL330020 VL CDR2 PRT 30 IL330020 VL CDR3 PRT 31 IL330071 VH DNA 32 IL330071 VH PRT 33 IL330071 VH CDR1 PRT 34 IL330071 VH CDR2 PRT 35 IL330071 VH CDR3 PRT 36 IL330071 VL DNA 37 IL330071 VL PRT 38 IL330071 VL CDR1 PRT 39 IL330071 VL CDR2 PRT 40 IL330071 VL CDR3 PRT 41 IL330125 VH DNA 42 IL330125 VH PRT 43 IL330125 VH CDR1 PRT 44 IL330125 VH CDR2 PRT 45 IL330125 VH CDR3 PRT 46 IL330125 VL DNA 47 IL330125 VL PRT 48 IL330125 VL CDR1 PRT 49 IL330125 VL CDR2 PRT 50 IL330125 VL CDR3 PRT 51 IL330126 VH DNA 52 IL330126 VH PRT 53 IL330126 VH CDR1 PRT 54 IL330126 VH CDR2 PRT 55 IL330126 VH CDR3 PRT 56 IL330126 VL DNA 57 IL330126 VL PRT 58 IL330126 VL CDR1 PRT 59 IL330126 VL CDR2 PRT 60 IL330126 VL CDR3 PRT 61 IL330425 VH DNA 62 IL330425 VH PRT 63 IL330425 VH CDR1 PRT 64 IL330425 VH CDR2 PRT 65 IL330425 VH CDR3 PRT 66 IL330425 VL DNA 67 IL330425 VL PRT 68 IL330425 VL CDR1 PRT 69 IL330425 VL CDR2 PRT 70 IL330425 VL CDR3 PRT 71 IL330428 VH DNA 72 IL330428 VH PRT 73 IL330428 VH CDR1 PRT 74 IL330428 VH CDR2 PRT 75 IL330428 VH CDR3 PRT 76 IL330428 VL DNA 77 IL330428 VL PRT 78 IL330428 VL CDR1 PRT 79 IL330428 VL CDR2 PRT 80 IL330428 VL CDR3 PRT 81 IL330065 VH DNA 82 IL330065 VH PRT 83 IL330065 VH CDR1 PRT 84 IL330065 VH CDR2 PRT 85 IL330065 VH CDR3 PRT 86 IL330065 VL DNA 87 IL330065 VL PRT 88 IL330065 VL CDR1 PRT 89 IL330065 VL CDR2 PRT 90 IL330065 VL CDR3 PRT 91 IL330099 VH DNA 92 IL330099 VH PRT 93 IL330099 VH CDR1 PRT 94 IL330099 VH CDR2 PRT 95 IL330099 VH CDR3 PRT 96 IL330099 VL DNA 97 IL330099 VL PRT 98 IL330099 VL CDR1 PRT 99 IL330099 VL CDR2 PRT 100 IL330099 VL CDR3 PRT 101 IL330101 VH DNA 102 IL330101 VH PRT 103 IL330101 VH CDR1 PRT 104 IL330101 VH CDR2 PRT 105 IL330101 VH CDR3 PRT 106 IL330101 VL DNA 107 IL330101 VL PRT 108 IL330101 VL CDR1 PRT 109 IL330101 VL CDR2 PRT 110 IL330101 VL CDR3 PRT 111 IL330101_fgl VH DNA 112 IL330101_fgl VH PRT 113 IL330101_fgl VH CDR1 PRT 114 IL330101_fgl VH CDR2 PRT 115 IL330101_fgl VH CDR3 PRT 116 IL330101_fgl VL DNA 117 IL330101_fgl VL PRT 118 IL330101_fgl VL CDR1 PRT 119 IL330101_fgl VL CDR2 PRT 120 IL330101_fgl VL CDR3 PRT 121 IL330107 VH DNA 122 IL330107 VH PRT 123 IL330107 VH CDR1 PRT 124 IL330107 VH CDR2 PRT 125 IL330107 VH CDR3 PRT 126 IL330107 VL DNA 127 IL330107 VL PRT 128 IL330107 VL CDR1 PRT 129 IL330107 VL CDR2 PRT 130 IL330107 VL CDR3 PRT 131 IL330149 VH DNA 132 IL330149 VH PRT 133 IL330149 VH CDR1 PRT 134 IL330149 VH CDR2 PRT 135 IL330149 VH CDR3 PRT 136 IL330149 VL DNA 137 IL330149 VL PRT 138 IL330149 VL CDR1 PRT 139 IL330149 VL CDR2 PRT 140 IL330149 VL CDR3 PRT 141 IL330180 VH DNA 142 IL330180 VH PRT 143 IL330180 VH CDR1 PRT 144 IL330180 VH CDR2 PRT 145 IL330180 VH CDR3 PRT 146 IL330180 VL DNA 147 IL330180 VL PRT 148 IL330180 VL CDR1 PRT 149 IL330180 VL CDR2 PRT 150 IL330180 VL CDR3 PRT 151 IL330259 VH DNA 152 IL330259 VH PRT 153 IL330259 VH CDR1 PRT 154 IL330259 VH CDR2 PRT 155 IL330259 VH CDR3 PRT 156 IL330259 VL DNA 157 IL330259 VL PRT 158 IL330259 VL CDR1 PRT 159 IL330259 VL CDR2 PRT 160 IL330259 VL CDR3 PRT 161 IL330259_fgl VH DNA 162 IL330259_fgl VH PRT 163 IL330259_fgl VH CDR1 PRT 164 IL330259_fgl VH CDR2 PRT 165 IL330259_fgl VH CDR3 PRT 166 IL330259_fgl VL DNA 167 IL330259_fgl VL PRT 168 IL330259_fgl VL CDR1 PRT 169 IL330259_fgl VL CDR2 PRT 170 IL330259_fgl VL CDR3 PRT 171 H338L293 VH DNA 172 H338L293 VH PRT 173 H338L293 VH CDR1 PRT 174 H338L293 VH CDR2 PRT 175 H338L293 VH CDR3 PRT 176 H338L293 VL DNA 177 H338L293 VL PRT 178 H338L293 VL CDR1 PRT 179 H338L293 VL CDR2 PRT 180 H338L293 VL CDR3 PRT 181 H338L293_fgl VH DNA 182 H338L293_fgl VH PRT 183 H338L293_fgl VH CDR1 PRT 184 H338L293_fgl VH CDR2 PRT 185 H338L293_fgl VH CDR3 PRT 186 H338L293_fgl VL DNA 187 H338L293_fgl VL PRT 188 H338L293_fgl VL CDR1 PRT 189 H338L293_fgl VL CDR2 PRT 190 H338L293_fgl VL CDR3 PRT 191 IL330377 VH DNA 192 IL330377 VH PRT 193 IL330377 VH CDR1 PRT 194 IL330377 VH CDR2 PRT 195 IL330377 VH CDR3 PRT 196 IL330377 VL DNA 197 IL330377 VL PRT 198 IL330377 VL CDR1 PRT 199 IL330377 VL CDR2 PRT 200 IL330377 VL CDR3 PRT 201 IL330377_fgl VH DNA 202 IL330377_fgl VH PRT 203 IL330377_fgl VH CDR1 PRT 204 IL330377_fgl VH CDR2 PRT 205 IL330377_fgl VH CDR3 PRT 206 IL330377_fgl VL DNA 207 IL330377_fgl VL PRT 208 IL330377_fgl VL CDR1 PRT 209 IL330377_fgl VL CDR2 PRT 210 IL330377_fgl VL CDR3 PRT 211 IL330388 VH DNA 212 IL330388 VH PRT 213 IL330388 VH CDR1 PRT 214 IL330388 VH CDR2 PRT 215 IL330388 VH CDR3 PRT 216 IL330388 VL DNA 217 IL330388 VL PRT 218 IL330388 VL CDR1 PRT 219 IL330388 VL CDR2 PRT 220 IL330388 VL CDR3 PRT 221 IL330388_fgl VH DNA 222 IL330388_fgl VH PRT 223 IL330388_fgl VH CDR1 PRT 224 IL330388_fgl VH CDR2 PRT 225 IL330388_fgl VH CDR3 PRT 226 IL330388_fgl VL DNA 227 IL330388_fgl VL PRT 228 IL330388_fgl VL CDR1 PRT 229 IL330388_fgl VL CDR2 PRT 230 IL330388_fgl VL CDR3 PRT 231 IL330396 VH DNA 232 IL330396 VH PRT 233 IL330396 VH CDR1 PRT 234 IL330396 VH CDR2 PRT 235 IL330396 VH CDR3 PRT 236 IL330396 VL DNA 237 IL330396 VL PRT 238 IL330396 VL CDR1 PRT 239 IL330396 VL CDR2 PRT 240 IL330396 VL CDR3 PRT 241 IL330396_fgl VH DNA 242 IL330396_fgl VH PRT 243 IL330396_fgl VH CDR1 PRT 244 IL330396_fgl VH CDR2 PRT 245 IL330396_fgl VH CDR3 PRT 246 IL330396_fgl VL DNA 247 IL330396_fgl VL PRT 248 IL330396_fgl VL CDR1 PRT 249 IL330396_fgl VL CDR2 PRT 250 IL330396_fgl VL CDR3 PRT 251 IL330398 VH DNA 252 IL330398 VH PRT 253 IL330398 VH CDR1 PRT 254 IL330398 VH CDR2 PRT 255 IL330398 VH CDR3 PRT 256 IL330398 VL DNA 257 IL330398 VL PRT 258 IL330398 VL CDR1 PRT 259 IL330398 VL CDR2 PRT 260 IL330398 VL CDR3 PRT 261 IL330398_fgl VH DNA 262 IL330398_fgl VH PRT 263 IL330398_fgl VH CDR1 PRT 264 IL330398_fgl VH CDR2 PRT 265 IL330398_fgl VH CDR3 PRT 266 IL330398_fgl VL DNA 267 IL330398_fgl VL PRT 268 IL330398_fgl VL CDR1 PRT 269 IL330398_fgl VL CDR2 PRT 270 IL330398_fgl VL CDR3 PRT 271 ZZ1EBX-E05 (33v20064) VH DNA 272 ZZ1EBX-E05 (33v20064) VH PRT 273 ZZ1EBX-E05 (33v20064) VH CDR1 PRT 274 ZZ1EBX-E05 (33v20064) VH CDR2 PRT 275 ZZ1EBX-E05 (33v20064) VH CDR3 PRT 276 ZZ1EBX-E05 (33v20064) VL DNA 277 ZZ1EBX-E05 (33v20064) VL PRT 278 ZZ1EBX-E05 (33v20064) VL CDR1 PRT 279 ZZ1EBX-E05 (33v20064) VL CDR2 PRT 280 ZZ1EBX-E05 (33v20064) VL CDR3 PRT 281 ZZ1I6V-H02 (33_640001) VH DNA 282 ZZ1I6V-H02 (33_640001) VH PRT 283 ZZ1I6V-H02 (33_640001) VH CDR1 PRT 284 ZZ1I6V-H02 (33_640001) VH CDR2 PRT 285 ZZ1I6V-H02 (33_640001) VH CDR3 PRT 286 ZZ1I6V-H02 (33_640001) VL DNA 287 ZZ1I6V-H02 (33_640001) VL PRT 288 ZZ1I6V-H02 (33_640001) VL CDR1 PRT 289 ZZ1I6V-H02 (33_640001) VL CDR2 PRT 290 ZZ1I6V-H02 (33_640001) VL CDR3 PRT 291 ZZ1JRB-A03 (33_640027) VH DNA 292 ZZ1JRB-A03 (33_640027) VH PRT 293 ZZ1JRB-A03 (33_640027) VH CDR1 PRT 294 ZZ1JRB-A03 (33_640027) VH CDR2 PRT 295 ZZ1JRB-A03 (33_640027) VH CDR3 PRT 296 ZZ1JRB-A03 (33_640027) VL DNA 297 ZZ1JRB-A03 (33_640027) VL PRT 298 ZZ1JRB-A03 (33_640027) VL CDR1 PRT 299 ZZ1JRB-A03 (33_640027) VL CDR2 PRT 300 ZZ1JRB-A03 (33_640027) VL CDR3 PRT 301 ZZ1F7Q-D10 (33_640050) VH DNA 302 ZZ1F7Q-D10 (33_640050) VH PRT 303 ZZ1F7Q-D10 (33_640050) VH CDR1 PRT 304 ZZ1F7Q-D10 (33_640050) VH CDR2 PRT 305 ZZ1F7Q-D10 (33_640050) VH CDR3 PRT 306 ZZ1F7Q-D10 (33_640050) VL DNA 307 ZZ1F7Q-D10 (33_640050) VL PRT 308 ZZ1F7Q-D10 (33_640050) VL CDR1 PRT 309 ZZ1F7Q-D10 (33_640050) VL CDR2 PRT 310 ZZ1F7Q-D10 (33_640050) VL CDR3 PRT 311 ZZ1F7P-E01 (33_640047) VH DNA 312 ZZ1F7P-E01 (33_640047) VH PRT 313 ZZ1F7P-E01 (33_640047) VH CDR1 PRT 314 ZZ1F7P-E01 (33_640047) VH CDR2 PRT 315 ZZ1F7P-E01 (33_640047) VH CDR3 PRT 316 ZZ1F7P-E01 (33_640047) VL DNA 317 ZZ1F7P-E01 (33_640047) VL PRT 318 ZZ1F7P-E01 (33_640047) VL CDR1 PRT 319 ZZ1F7P-E01 (33_640047) VL CDR2 PRT 320 ZZ1F7P-E01 (33_640047) VL CDR3 PRT 321 ZZ1IV4-H06 (33_640166) VH DNA 322 ZZ1IV4-H06 (33_640166) VH PRT 323 ZZ1IV4-H06 (33_640166) VH CDR1 PRT 324 ZZ1IV4-H06 (33_640166) VH CDR2 PRT 325 ZZ1IV4-H06 (33_640166) VH CDR3 PRT 326 ZZ1IV4-H06 (33_640166) VL DNA 327 ZZ1IV4-H06 (33_640166) VL PRT 328 ZZ1IV4-H06 (33_640166) VL CDR1 PRT 329 ZZ1IV4-H06 (33_640166) VL CDR2 PRT 330 ZZ1IV4-H06 (33_640166) VL CDR3 PRT 331 ZZ1IV4-G09 (33_640169) VH DNA 332 ZZ1IV4-G09 (33_640169) VH PRT 333 ZZ1IV4-G09 (33_640169) VH CDR1 PRT 334 ZZ1IV4-G09 (33_640169) VH CDR2 PRT 335 ZZ1IV4-G09 (33_640169) VH CDR3 PRT 336 ZZ1IV4-G09 (33_640169) VL DNA 337 ZZ1IV4-G09 (33_640169) VL PRT 338 ZZ1IV4-G09 (33_640169) VL CDR1 PRT 339 ZZ1IV4-G09 (33_640169) VL CDR2 PRT 340 ZZ1IV4-G09 (33_640169) VL CDR3 PRT 341 ZZ1K7Q-B11 (33_640170) VH DNA 342 ZZ1K7Q-B11 (33_640170) VH PRT 343 ZZ1K7Q-B11 (33_640170) VH CDR1 PRT 344 ZZ1K7Q-B11 (33_640170) VH CDR2 PRT 345 ZZ1K7Q-B11 (33_640170) VH CDR3 PRT 346 ZZ1K7Q-B11 (33_640170) VL DNA 347 ZZ1K7Q-B11 (33_640170) VL PRT 348 ZZ1K7Q-B11 (33_640170) VL CDR1 PRT 349 ZZ1K7Q-B11 (33_640170) VL CDR2 PRT 350 ZZ1K7Q-B11 (33_640170) VL CDR3 PRT 351 ZZ1KAD-C04 (33_640036) VH DNA 352 ZZ1KAD-C04 (33_640036) VH PRT 353 ZZ1KAD-C04 (33_640036) VH CDR1 PRT 354 ZZ1KAD-C04 (33_640036) VH CDR2 PRT 355 ZZ1KAD-C04 (33_640036) VH CDR3 PRT 356 ZZ1KAD-C04 (33_640036) VL DNA 357 ZZ1KAD-C04 (33_640036) VL PRT 358 ZZ1KAD-C04 (33_640036) VL CDR1 PRT 359 ZZ1KAD-C04 (33_640036) VL CDR2 PRT 360 ZZ1KAD-C04 (33_640036) VL CDR3 PRT 361 33_640117 VH DNA 362 33_640117 VH PRT 363 33_640117 VH CDR1 PRT 364 33_640117 VH CDR2 PRT 365 33_640117 VH CDR3 PRT 366 33_640117 VL DNA 367 33_640117 VL PRT 368 33_640117 VL CDR1 PRT 369 33_640117 VL CDR2 PRT 370 33_640117 VL CDR3 PRT 371 ZZ1JLT-F06 (33_640076) VH DNA 372 ZZ1JLT-F06 (33_640076) VH PRT 373 ZZ1JLT-F06 (33_640076) VH CDR1 PRT 374 ZZ1JLT-F06 (33_640076) VH CDR2 PRT 375 ZZ1JLT-F06 (33_640076) VH CDR3 PRT 376 ZZ1JLT-F06 (33_640076) VL DNA 377 ZZ1JLT-F06 (33_640076) VL PRT 378 ZZ1JLT-F06 (33_640076) VL CDR1 PRT 379 ZZ1JLT-F06 (33_640076) VL CDR2 PRT 380 ZZ1JLT-F06 (33_640076) VL CDR3 PRT 381 ZZ1JMB-H05 (33_640081) VH DNA 382 ZZ1JMB-H05 (33_640081) VH PRT 383 ZZ1JMB-H05 (33_640081) VH CDR1 PRT 384 ZZ1JMB-H05 (33_640081) VH CDR2 PRT 385 ZZ1JMB-H05 (33_640081) VH CDR3 PRT 386 ZZ1JMB-H05 (33_640081) VL DNA 387 ZZ1JMB-H05 (33_640081) VL PRT 388 ZZ1JMB-H05 (33_640081) VL CDR1 PRT 389 ZZ1JMB-H05 (33_640081) VL CDR2 PRT 390 ZZ1JMB-H05 (33_640081) VL CDR3 PRT 391 ZZ1JMA-B04 (33_640082) VH DNA 392 ZZ1JMA-B04 (33_640082) VH PRT 393 ZZ1JMA-B04 (33_640082) VH CDR1 PRT 394 ZZ1JMA-B04 (33_640082) VH CDR2 PRT 395 ZZ1JMA-B04 (33_640082) VH CDR3 PRT 396 ZZ1JMA-B04 (33_640082) VL DNA 397 ZZ1JMA-B04 (33_640082) VL PRT 398 ZZ1JMA-B04 (33_640082) VL CDR1 PRT 399 ZZ1JMA-B04 (33_640082) VL CDR2 PRT 400 ZZ1JMA-B04 (33_640082) VL CDR3 PRT 401 ZZ1JLR-D06 (33_640084) VH DNA 402 ZZ1JLR-D06 (33_640084) VH PRT 403 ZZ1JLR-D06 (33_640084) VH CDR1 PRT 404 ZZ1JLR-D06 (33_640084) VH CDR2 PRT 405 ZZ1JLR-D06 (33_640084) VH CDR3 PRT 406 ZZ1JLR-D06 (33_640084) VL DNA 407 ZZ1JLR-D06 (33_640084) VL PRT 408 ZZ1JLR-D06 (33_640084) VL CDR1 PRT 409 ZZ1JLR-D06 (33_640084) VL CDR2 PRT 410 ZZ1JLR-D06 (33_640084) VL CDR3 PRT 411 ZZ1JMC-H09 (33_640086) VH DNA 412 ZZ1JMC-H09 (33_640086) VH PRT 413 ZZ1JMC-H09 (33_640086) VH CDR1 PRT 414 ZZ1JMC-H09 (33_640086) VH CDR2 PRT 415 ZZ1JMC-H09 (33_640086) VH CDR3 PRT 416 ZZ1JMC-H09 (33_640086) VL DNA 417 ZZ1JMC-H09 (33_640086) VL PRT 418 ZZ1JMC-H09 (33_640086) VL CDR1 PRT 419 ZZ1JMC-H09 (33_640086) VL CDR2 PRT 420 ZZ1JMC-H09 (33_640086) VL CDR3 PRT 421 ZZ1JMF-G02 (33_640087) VH DNA 422 ZZ1JMF-G02 (33_640087) VH PRT 423 ZZ1JMF-G02 (33_640087) VH CDR1 PRT 424 ZZ1JMF-G02 (33_640087) VH CDR2 PRT 425 ZZ1JMF-G02 (33_640087) VH CDR3 PRT 426 ZZ1JMF-G02 (33_640087) VL DNA 427 ZZ1JMF-G02 (33_640087) VL PRT 428 ZZ1JMF-G02 (33_640087) VL CDR1 PRT 429 ZZ1JMF-G02 (33_640087) VL CDR2 PRT 430 ZZ1JMF-G02 (33_640087) VL CDR3 PRT 431 33_640076_1 VH DNA 432 33_640076_1 VH PRT 433 33_640076_1 VH CDR1 PRT 434 33_640076_1 VH CDR2 PRT 435 33_640076_1 VH CDR3 PRT 436 33_640076_1 VL DNA 437 33_640076_1 VL PRT 438 33_640076_1 VL CDR1 PRT 439 33_640076_1 VL CDR2 PRT 440 33_640076_1 VL CDR3 PRT 441 33_640081_A VH DNA 442 33_640081_A VH PRT 443 33_640081_A VH CDR1 PRT 444 33_640081_A VH CDR2 PRT 445 33_640081_A VH CDR3 PRT 446 33_640081_A VL DNA 447 33_640081_A VL PRT 448 33_640081_A VL CDR1 PRT 449 33_640081_A VL CDR2 PRT 450 33_640081_A VL CDR3 PRT 451 33_640082_2 VH DNA 452 33_640082_2 VH PRT 453 33_640082_2 VH CDR1 PRT 454 33_640082_2 VH CDR2 PRT 455 33_640082_2 VH CDR3 PRT 456 33_640082_2 VL DNA 457 33_640082_2 VL PRT 458 33_640082_2 VL CDR1 PRT 459 33_640082_2 VL CDR2 PRT 460 33_640082_2 VL CDR3 PRT 461 33_640084_2 VH DNA 462 33_640084_2 VH PRT 463 33_640084_2 VH CDR1 PRT 464 33_640084_2 VH CDR2 PRT 465 33_640084_2 VH CDR3 PRT 466 33_640084_2 VL DNA 467 33_640084_2 VL PRT 468 33_640084_2 VL CDR1 PRT 469 33_640084_2 VL CDR2 PRT 470 33_640084_2 VL CDR3 PRT 471 33_640086_2 VH DNA 472 33_640086_2 VH PRT 473 33_640086_2 VH CDR1 PRT 474 33_640086_2 VH CDR2 PRT 475 33_640086_2 VH CDR3 PRT 476 33_640086_2 VL DNA 477 33_640086_2 VL PRT 478 33_640086_2 VL CDR1 PRT 479 33_640086_2 VL CDR2 PRT 480 33_640086_2 VL CDR3 PRT 481 33_640087_2 VH DNA 482 33_640087_2 VH PRT 483 33_640087_2 VH CDR1 PRT 484 33_640087_2 VH CDR2 PRT 485 33_640087_2 VH CDR3 PRT 486 33_640087_2 VL DNA 487 33_640087_2 VL PRT 488 33_640087_2 VL CDR1 PRT 489 33_640087_2 VL CDR2 PRT 490 33_640087_2 VL CDR3 PRT 491 33_640076_4 VH DNA 492 33_640076_4 VH PRT 493 33_640076_4 VH CDR1 PRT 494 33_640076_4 VH CDR2 PRT 495 33_640076_4 VH CDR3 PRT 496 33_640076_4 VL DNA 497 33_640076_4 VL PRT 498 33_640076_4 VL CDR1 PRT 499 33_640076_4 VL CDR2 PRT 500 33_640076_4 VL CDR3 PRT 501 33_640082_4 VH DNA 502 33_640082_4 VH PRT 503 33_640082_4 VH CDR1 PRT 504 33_640082_4 VH CDR2 PRT 505 33_640082_4 VH CDR3 PRT 506 33_640082_4 VL DNA 507 33_640082_4 VL PRT 508 33_640082_4 VL CDR1 PRT 509 33_640082_4 VL CDR2 PRT 510 33_640082_4 VL CDR3 PRT 511 33_640082_6 VH DNA 512 33_640082_6 VH PRT 513 33_640082_6 VH CDR1 PRT 514 33_640082_6 VH CDR2 PRT 515 33_640082_6 VH CDR3 PRT 516 33_640082_6 VL DNA 517 33_640082_6 VL PRT 518 33_640082_6 VL CDR1 PRT 519 33_640082_6 VL CDR2 PRT 520 33_640082_6 VL CDR3 PRT 521 33_640082_7 VH DNA 522 33_640082_7 VH PRT 523 33_640082_7 VH CDR1 PRT 524 33_640082_7 VH CDR2 PRT 525 33_640082_7 VH CDR3 PRT 526 33_640082_7 VL DNA 527 33_640082_7 VL PRT 528 33_640082_7 VL CDR1 PRT 529 33_640082_7 VL CDR2 PRT 530 33_640082_7 VL CDR3 PRT 531 33_640086_6 VH DNA 532 33_640086_6 VH PRT 533 33_640086_6 VH CDR1 PRT 534 33_640086_6 VH CDR2 PRT 535 33_640086_6 VH CDR3 PRT 536 33_640086_6 VL DNA 537 33_640086_6 VL PRT 538 33_640086_6 VL CDR1 PRT 539 33_640086_6 VL CDR2 PRT 540 33_640086_6 VL CDR3 PRT 541 33_640087_7 VH DNA 542 33_640087_7 VH PRT 543 33_640087_7 VH CDR1 PRT 544 33_640087_7 VH CDR2 PRT 545 33_640087_7 VH CDR3 PRT 546 33_640087_7 VL DNA 547 33_640087_7 VL PRT 548 33_640087_7 VL CDR1 PRT 549 33_640087_7 VL CDR2 PRT 550 33_640087_7 VL CDR3 PRT 551 ZZ1JMY-H09 (33_640201) VH DNA 552 ZZ1JMY-H09 (33_640201) VH PRT 553 ZZ1JMY-H09 (33_640201) VH CDR1 PRT 554 ZZ1JMY-H09 (33_640201) VH CDR2 PRT 555 ZZ1JMY-H09 (33_640201) VH CDR3 PRT 556 ZZ1JMY-H09 (33_640201) VL DNA 557 ZZ1JMY-H09 (33_640201) VL PRT 558 ZZ1JMY-H09 (33_640201) VL CDR1 PRT 559 ZZ1JMY-H09 (33_640201) VL CDR2 PRT 560 ZZ1JMY-H09 (33_640201) VL CDR3 PRT 561 ZZ1M37-E06 (33_640237) VH DNA 562 ZZ1M37-E06 (33_640237) VH PRT 563 ZZ1M37-E06 (33_640237) VH CDR1 PRT 564 ZZ1M37-E06 (33_640237) VH CDR2 PRT 565 ZZ1M37-E06 (33_640237) VH CDR3 PRT 566 ZZ1M37-E06 (33_640237) VL DNA 567 ZZ1M37-E06 (33_640237) VL PRT 568 ZZ1M37-E06 (33_640237) VL CDR1 PRT 569 ZZ1M37-E06 (33_640237) VL CDR2 PRT 570 ZZ1M37-E06 (33_640237) VL CDR3 PRT 571 33_640201_2 VH DNA 572 33_640201_2 VH PRT 573 33_640201_2 VH CDR1 PRT 574 33_640201_2 VH CDR2 PRT 575 33_640201_2 VH CDR3 PRT 576 33_640201_2 VL DNA 577 33_640201_2 VL PRT 578 33_640201_2 VL CDR1 PRT 579 33_640201_2 VL CDR2 PRT 580 33_640201_2 VL CDR3 PRT 581 33_640237_2 VH DNA 582 33_640237_2 VH PRT 583 33_640237_2 VH CDR1 PRT 584 33_640237_2 VH CDR2 PRT 585 33_640237_2 VH CDR3 PRT 586 33_640237_2 VL DNA 587 33_640237_2 VL PRT 588 33_640237_2 VL CDR1 PRT 589 33_640237_2 VL CDR2 PRT 590 33_640237_2 VL CDR3 PRT 591 33_640076_4B VH DNA 592 33_640076_4B VH PRT 593 33_640076_4B VL DNA 594 33_640076_4B VL PRT 595 33_640081_AB VH DNA 596 33_640081_AB VH PRT 597 33_640081_AB VL DNA 598 33_640081_AB VL PRT 599 33_640082_6B VH DNA 600 33_640082_6B VH PRT 601 33_640082_6B VL DNA 602 33_640082_6B VL PRT 603 33_640082_7B VH DNA 604 33_640082_7B VH PRT 605 33_640082_7B VL DNA 606 33_640082_7B VL PRT 607 33_640084_2B VH DNA 608 33_640084_2B VH PRT 609 33_640084_2B VL DNA 610 33_640084_2B VL PRT 611 33_640086_6B VH DNA 612 33_640086_6B VH PRT 613 33_640086_6B VL DNA 614 33_640086_6B VL PRT 615 33_640087_7B VH DNA 616 33_640087_7B VH PRT 617 33_640087_7B VL DNA 618 33_640087_7B VL PRT 619 33_640201_2B VH DNA 620 33_640201_2B VH PRT 621 33_640201_2B VL DNA 622 33_640201_2B VL PRT 623 33_640237_2B VH DNA 624 33_640237_2B VH PRT 625 33_640237_2B VL DNA 626 33_640237_2B VL PRT 627 Mature Human IL-33_FH a.a. 112-270 PRT 628 Mature Mouse IL-33_FH a.a. 109-266 PRT 629 Mature Cynomolugus IL-33 a.a. 112-270 PRT 630 Human ST2 ECD-Fc/his6 a.a. 1-328 PRT 631 Mouse ST2 ECD-Fc/His6 a.a. 1-332 PRT 632 IL33-01 a.a. 112-270 PRT 633 Human 6His TEV mature IL-33 WT a.a. 112-270 PRT 634 IL33-02 a.a. 112-270 PRT 635 IL33-03 a.a. 112-270 PRT 636 IL33-04 a.a. 112-270 PRT 637 IL33-05 a.a. 112-270 PRT 638 IL33-06 a.a. 112-270 PRT 639 IL33-07 a.a. 112-270 PRT 640 IL33-08 a.a. 112-270 PRT 641 IL33-09 a.a. 112-270 PRT 642 IL33-10 a.a. 112-270 PRT 643 IL33-11 a.a. 112-270 PRT 644 IL33-12 a.a. 112-270 PRT 645 IL33-13 a.a. 112-270 PRT 646 IL33-14 a.a. 112-270 PRT 647 IL33-15 a.a. 112-270 PRT 648 IL33-16 a.a. 112-270 PRT 649 Cynomolgus 10His Avitag IL-33 a.a. 112-270 PRT 650 Human ST2 ECD-Flag-his10 a.a. 1-328 PRT

EXAMPLES

(60) Example 1 Isolation of Antibodies to IL-33

(61) Cloning, Expression and Purification of Mature IL-33 from Human, Mouse and Cynomolgus Monkey

(62) Protein sequences for IL-1RAcP and ST2 were obtained from Swiss Prot. Isolation and identification of anti-IL-33 scFv antibodies cDNA molecules encoding the mature component of IL-33 were synthesized by primer extension PCR and cloned into pJexpress404 (DNA 2.0). Accession numbers corresponding to database sequence information for human and mouse IL-33 are shown in Table 2. No Cynomologus monkey sequences were available so based on the high homology between Cynomolgus monkey and Rhesus monkey, the sequence of Rhesus monkey (Accession No. ENSMMUT00000030043) was used to design primers capable of amplifying the coding sequence of the IL-33 gene in Cynomolgus monkey. The Rhesus gene sequence was aligned to the Humans IL-33 cDNA sequence (Accession No. NM_033439), this demonstrated that the Rhesus sequence was mis-assembled and was missing exon 1. A BLAST search was performed against the Rhesus genomic sequence using the human exon 1, and the Rhesus sequence matching exon 1 was identified. Additional primers were designed to amplify exon 1.

(63) The mature IL-33 coding sequence was modified to contain a FLAG® 10×his epitope tag (DYKDDDDKAAHHHHHHHHHH; SEQ ID NO. 627) at the C-terminus of the protein. SEQ ID NOs corresponding to mature Flag® His-tagged human, cynomolgus and mouse IL-33 are shown in Table 2.

(64) TABLE-US-00006 TABLE 2 Sequences for human, mouse and cynomolgus monkey mature IL-33 Accession No. Flag ®His-tagged Species Amino acids (Swiss-Prot) IL-33 Sequences Human 112-270 O95760 SEQ ID NO. 627 Mouse 109-266 Q8BVZ5 SEQ ID NO. 628 Cynomolgus 112-270 Not Available SEQ ID NO. 629

(65) Vectors were transformed into BL21(DE3) competent cells (Merck Biosciences, 69450) and expression induced with 1 mM IPTG. Harvested cells were lysed with Bugbuster (Merck Biosciences, 70584) and expressed protein was purified using Ni-NTA affinity chromatography (Histrap HP column: GE Healthcare, 17-5248-02) followed by Size Exclusion chromatography (Superdex 75 column: GE Healthcare, 17-1068-01).

(66) Protein Modifications

(67) IgGs and modified receptor proteins used herein were biotinylated via free amines using EZ link Sulfo-NHS-LC-Biotin (Thermo/Pierce, 21335) The biotin reagent was dissolved in anhydrous dimethylformamide and PBS based protein solutions were adjusted to pH ˜8 with 1 M NaHCO.sub.3 in D-PBS (Dulbecco's phosphate buffered saline). IL-33 proteins used herein were biotinylated via free cysteines using EZ link Biotin-BMCC (Perbio/Pierce, product no. 21900). The biotin reagent was dissolved in anhydrous dimethylformamide and mixed PBS protein solutions. Label incorporations were assessed by MALDI-TOF mass spectrometry in all cases and unreacted reagents were cleared by buffer exchange using PBS equilibrated disposable Sephadex G25 columns. For biotinylation, the final protein concentrations were determined by 280 nm absorbance using extinction coefficients calculated from amino acid sequences.

(68) Selections

(69) A large single chain Fv (scFv) human antibody library, based upon variable (V) genes isolated from human B-cells from adult naïve donors and cloned into a phagemid vector based on filamentous phage M13 was used for selections (Hutchings, C., “Generation of Naïve Human Antibody Libraries” in Antibody Engineering, Dubel. Berlin, Springer Laboratory Manuals: p. 93 (2001); Lloyd et al., Protein Eng. Des. Sel. 22(3):159-68 (2009)). IL-33-specific scFv antibodies were isolated from the phage display library in a series of repeated selection cycles on recombinant human and/or mouse IL-33 essentially as described in Vaughan et al. (Nat. Biotechnol. 14(3):309-14 (1996)). A list of IL-33 reagents used herein is shown in Table 3.

(70) TABLE-US-00007 TABLE 3 ELISA Binding Assay Reagents Catalogue Number/ ELISA Reagent Supplier Designation assay Human IL-33 Axxora/Adipogen AG-40B-0038 Phage/IgG Human IL-33 Axxora/Alexis ALX-522-098 Phage Human IL-33 Flag ®His In house PS-295 Phage/IgG Human IL-33 Peprotech 200-33 Phage Mouse IL-33 Peprotech 210-33 Phage Mouse IL-33 Axxora/Alexis ALX-522-101 Phage Mouse IL-33 Flag ®His In house PS-296 Phage/IgG Cynomologus IL-33 In house PS-368 Phage/IgG Flag ®His IL-4Rα Flag ®His In house 020629080 Phage/IgG Bovine insulin - biotin Sigma I2258 Phage/IgG

(71) In brief, the scFv-phage particles were incubated with biotinylated recombinant IL-33 in solution (biotinylated via free cysteines using EZ link Biotin-BMCC (Perbio/Pierce, product no. 21900)). Particles were incubated with 100 nM biotinylated recombinant IL-33 for 2 hours. ScFv bound to antigen were then captured on streptavidin-coated paramagnetic beads (Dynabeads®, M-280) following manufacturer's recommendations. Unbound phage was washed away in a series of wash cycles using PBS-Tween. The phage particles retained on the antigen were eluted, infected into bacteria and rescued for the next round of selection. Typically two or three rounds of selection were performed in this way.

(72) Identification of IL-33 Specific Binders by Phage ELISA

(73) A representative number of individual clones from the selection outputs after two or three rounds of selection described above were grown up in 96-well plates. Single-chain Fv fragments were displayed on phage particles and tested in a binding assay to determine cross-reactivity and specificity to a panel of recombinant human, mouse and cynomolgus IL-33 antigens. Phage-displayed scFv supernatant samples were generated in 96-well deep well plates as follows. 5 μl of culture from each well of a 96-well master plate was transferred into a Greiner deep well culture plate containing 500 μl of 2TYAG (2TY+100 mg/ml ampicillin+2% glucose) media and incubated for 5 hours at 37° C., 280 rpm. K07 M13 helper phage (diluted to 1.5×10.sup.11 pfu/ml in 2TYAG) was then added at 100 μl/well and the plate incubated at 37° C., 150 rpm to allow infection. The plate was spun down at 3200 rpm for 10 minutes and the supernatant removed. Bacterial pellets were resuspended in 500 μl/well 2TYAK (2TY+100 mg/ml ampicillin+50 mg/ml kanamycin) and the plate incubated overnight at 25° C., 280 rpm. In the morning, 500 μl of 6% (w/v) skimmed milk powder in 2×PBS was added to each well and the plate incubated for 1 hour at room temperature. The plate was then centrifuged at 3200 rpm for 10 minutes and the blocked phage-displayed scFv supernatants were used directly in ELISA experiments.

(74) For EC50 determinations, typically purified IgGs were diluted 3-fold in 3% (w/v) dried-milk powder in PBS (PBS-M), to give 11 concentration points. 96-well Greiner polypropylene plates (Greiner, 650201) were used for dilution preparation. Generally, each dilution was prepared in duplicate. IgG dilutions were allowed to block in PBS-M for 1 hour at room temperature before being used directly in ELISA experiments.

(75) The IL-33 binding assays were plate-based ELISAs performed essentially as follows. Table 3 above shows the antigens used for these experiments. Not all antigens were used in every experiment, but in all cases a human, a mouse, and a cynomolgus IL-33 antigen was tested. Relevant control antigens (bovine insulin plus IL-4Rα FLAG® His, if appropriate) were also used to test for non-specific binding. With the exception of bovine insulin, all antigens were biotinylated (see subsection 1.1. above) and all were generated using bacterial expression. The method for generation of IL-4Rα FLAG® His, which was used as a control antigen, is described in WO/2010/070346.

(76) Streptavidin plates (Thermo Scientific, AB-1226) were coated with biotinylated antigen at 0.5 μg/ml in PBS and incubated overnight at 4° C. Plates were washed 3× with PBS and blocked with 300 μl/well blocking buffer (PBS-M) for 1 hour. Plates were washed 1× with PBS and blocked samples added, 50 μl/well for 1 hour at room temperature. Plates were washed 3× with PBS-T (PBS+1% (v/v) Tween-20) and detection reagents [anti-human IgG HRP (Sigma, A0170) or anti-M13-HRP antibody (Amersham, 27-9421-01) for detection of IgG or phage-displayed scFv, respectively] at 1:5000 dilutions were added at 50 μl/well in PBS-M for 1 hour at room temperature. Plates were washed 3× with PBS-T and developed with TMB, 50 μl/well (Sigma, T0440). The reaction was quenched with 50 μl/well 0.1M H2504 before reading on an EnVision™ plate reader, or similar equipment, at 450 nm.

(77) Dose response curves were plotted for IgG titrations using Prism (Graphpad) curve-fitting software. Phage-displayed scFv were considered to bind the IL-33 antigen if the absorbance 450 nm was >0.5, and <0.2 for the same sample on controls (insulin and IL-4Rα Flag® His).

(78) Cloning, Expression and Purification of ST2 ECD from Human and Mouse

(79) cDNA molecules encoding the extracellular domains (ECDs) of ST2 from human and mouse were synthesized by primer extension PCR cloning and cloned into pDONR221 (Invitrogen, 12536-017). Database sequences for human and mouse ST2 were used (see Table 4). ST2 ECD cDNA clones in pDONR221 were transferred to mammalian expression vector pDEST12.2 using LR Gateway Clonase II enzyme according to the manufacturer's instructions. The pDEST12.2 vector had been modified to contain the human IgG1 Fc coding region, polyhistidine (His6) tag in-frame with the inserted gene of interest, and also by insertion of the oriP origin of replication from the pCEP4 vector allowing episomal plasmid replication upon transfection into cell lines expressing the EBNA-1 gene product (such as HEK293-EBNA cells).

(80) TABLE-US-00008 TABLE 4 Amino acids and accession numbers for human and mouse ST2 extracellular domain Accession number EDC-Fc-His6 Species Amino acids (Swiss-Prot) Sequences Human 1-328 Q01638 SEQ ID NO: 630 Mouse 1-332 P14719 SEQ ID NO: 631

(81) Expressed ST2.Fc proteins in HEK293-EBNA supernatants were purified using Protein A affinity chromatography (HiTrap Protein A column (GE Healthcare, 17-0402-01)) followed by Size Exclusion chromatography (Superdex 200 column (GE Healthcare, 17-1069-01)).

(82) Inhibition of IL-33 Binding to ST2 by Unpurified scFv

(83) A representative number of individual clones from the selection outputs after two or three rounds of selection described above were grown up in 96-well plates. ScFv were expressed in the bacterial periplasm (Kipriyanov, et al. J Immunol Methods 200(1-2): 69-77 (1997)) and screened for their inhibitory activity in a homogeneous FRET (fluorescence resonance energy transfer) HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) based IL-33:ST2-binding assay. In this assay, samples competed with human or mouse ST2.Fc for binding to FLAG® His-tagged human, cynomolgus or mouse IL-33.

(84) An HTRF® assay is a homogeneous assay technology that utilises fluorescence resonance energy transfer between a donor and acceptor fluorophore that are in close proximity (Mathis, et al. Clin Chem 41(9):1391-7 (1995)). This assay was used to measure macromolecular interactions by directly or indirectly coupling one of the molecules of interest to a donor fluorophore, europium (Eu3+) cryptate, and coupling the other molecule of interest to an acceptor fluorophore XL665, (a stable cross linked allophycocyanin). Excitation of the cryptate molecule (at 337 nm) resulted in fluorescence emission at 620 nm. The energy from this emission was transferred to XL665 in close proximity to the cryptate, resulting in the emission of a specific long-lived fluorescence (at 665 nm) from the XL665. The specific signals of both the donor (at 620 nm) and the acceptor (at 665 nm) were measured, allowing the calculation of a 665/620 nm ratio that compensates for the presence of colored compounds in the assay.

(85) Unpurified anti-IL-33 scFv samples were tested for inhibition of FLAW-His tagged IL-33 binding ST2-Fc by adding 10 microlitres of each dilution of antibody test sample to a 384 well low volume assay plate (Costar, 3676). Next, a solution containing 2 nM human or mouse ST2-Fc and 3 nM anti-human Fc cryptate detection (Cisbio International, 61HFCKLB) was prepared and 5 microlitres of the mix added to the assay plate. This was followed by the addition of 5 microlitres of a solution containing 1.2 nM FLAW-His tagged human, cynomolgus or mouse IL-33 combined with 20 nM anti-FLAG® XL665 detection (Cisbio International, 61FG2XLB). All dilutions were performed in assay buffer containing 0.8 M potassium fluoride (BDH 103444T) and 0.1% bovine serum albumin (BSA, Sigma A9576) in Dulbeccos PBS (Invitrogen, 14190185). Assay plates were incubated for 1 hour at room temperature followed by 16 hour at 4° C. before reading time resolved fluorescence at 620 nm and 665 nm emission wavelengths using an EnVision plate reader (Perkin Elmer).

(86) Data were analysed by calculating the 665/620 nm ratio followed by the % Delta F values for each sample. The 665/620 nm ratio was used to correct for sample interference using Equation 1:

(87) $665/620\mathrm{nm}\mathrm{ratio}=\left(\frac{665\mathrm{nm}\mathrm{signal}}{620\mathrm{nm}\mathrm{signal}}\right)\times 10\text{,}000$

(88) The % Delta F for each sample was then calculated using Equation 2:

(89) $\mathrm{Delta}F\left(\%\right)=\left(\frac{\begin{array}{c}\mathrm{sample}665/620\mathrm{nm}\mathrm{ratio}-\\ \mathrm{negative}\mathrm{control}665/620\mathrm{nm}\mathrm{ratio}\end{array}}{\mathrm{negative}\mathrm{control}665/620\mathrm{nm}\mathrm{ratio}}\right)\times 100$

(90) The negative control (non-specific binding) was defined by replacing test sample with 150 nM non-tagged human or mouse IL-33 (Axxora, human ALX522-098, mouse ALX-522-101) prepared in a dilution buffer comprised of Dulbeccos PBS (Invitrogen, 14190185) containing 0.1% bovine serum albumin (BSA, Sigma A9576).

(91) The % Delta F values were subsequently used to calculate % specific binding as described in Equation 3:

(92) $\%\mathrm{specific}\mathrm{binding}=\left(\frac{\left(\mathrm{Sample}\mathrm{Delta}F\%-\mathrm{NSB}\mathrm{Delta}F\%\right)}{\left(\mathrm{Total}\mathrm{binding}\mathrm{Delta}F\%-\mathrm{NSB}\mathrm{Delta}F\%\right)}\right)\times 100$

(93) IC.sub.50 values were determined using GraphPad Prism software by curve fitting using a four-parameter logistic equation (Equation 4).
Y=Bottom+(Top−Bottom)/(1+10{circumflex over ( )}((Log EC50−X)*HillSlope))  Equation 4:
X is the logarithm of concentration.
Y is specific binding
Y starts at Bottom and goes to Top with a sigmoid shape.

(94) FIG. 1 shows the inhibition of the FRET signal, produced by human IL-33 binding to human ST2, by unpurified scFv periplasmic extracts in a single point screen. The final concentration of the periplasmic extract was 50% v/v. Well B04 (unpurified IL330004 scFv) shows an example ‘hit’ and column 12 contains control wells as indicated.

(95) Inhibition of IL-33 Binding to ST2 by Purified scFv

(96) Single chain Fv clones which showed an inhibitory effect on IL-33:ST2 interaction as unpurified periplasmic extracts or demonstrated a desirable species cross-reactivity and specificity profile by phage binding experiments above, were subjected to DNA sequencing (Osbourn, et al. Immunotechnology 2(3):181-96 (1996); Vaughan, et al. Nat Biotechnol 14(3):309-14 (1996).). Unique scFv were expressed again in bacteria and purified by affinity chromatography (as described in WO01/66754). The potencies of these samples were determined by competing a dilution series of the purified preparation against human or mouse ST2.Fc for binding to FLAG® His-tagged human, cynomolgus or mouse IL-33 as described above. Purified scFv preparations that were capable of inhibiting the IL-33:ST2 interaction to a greater extent than the negative control were selected for conversion to IgG format (e.g., scFv antibodies IL330002, IL330004, IL330020 and IL330071

(97) FIG. 2A: shows the inhibition of the FRET signal, produced by human IL-33 binding to human ST2 with increasing concentrations of IL-33 scFv antibodies IL330002, IL330004, IL330020 and IL330071, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(98) FIG. 2B: shows the inhibition of the FRET signal, produced by cynomolgus monkey IL-33 binding to human ST2 with increasing concentrations of IL-33 scFv antibodies IL330002, IL330004, IL330020 and IL330071, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(99) Identification of Partner Antibodies for IL330004

(100) scFv periplasmic extracts of those clones that demonstrated positive binding to human IL-33 by Phage ELISA were screened by Octet assay (Octet RED 384 system) to identify an antibody that bound IL-33 simultaneously with IL330004. Neat periprep samples were captured on a Nickel NTA biosensor and sequential binding of IL-33 (200 nM) followed by IL330004 (200 nM) was performed. Biosensors were regenerated to minimise use. Sensors are regenerated in glycine (10 mM, pH 1.7), neutralised in buffer (PBS+1 mg/ml (0.1%) BSA+0.02% Tween20) and reloaded with NiS04 (10 mM) to replenish the nickel on the biosensor surface. IL330425 and IL330428 were identified and converted to whole immunoglobulin G1 (IgG1) antibody format.

(101) Reformatting of scFv to IgG1

(102) Single chain Fv clones with desirable properties from the IL-33:ST2 binding assays, plus a panel of phage-displayed scFv with desirable specificities by binding experiments were converted to whole immunoglobulin G1 (IgG1) antibody format essentially as described by Persic et al. (Gene 187(1):9-18 (1997)) with the following modifications. An OriP fragment was included in the expression vectors to facilitate use with CHO-transient cells and to allow episomal replication. The variable heavy (VH) domain was cloned into a vector (pEU1.3) containing the human heavy chain constant domains and regulatory elements to express whole IgG1 heavy chain in mammalian cells. Similarly, the variable light (VL) domain was cloned into a vector (pEU4.4) for the expression of the human light chain (lambda) constant domains and regulatory elements to express whole IgG light chain in mammalian cells. To obtain IgGs, the heavy and light chain IgG expressing vectors were transfected into CHO-transient mammalian cells (Daramola et al. Biotechnol Prog 30(1):132-41 (2014)). IgGs were expressed and secreted into the medium. Harvests were filtered prior to purification, then IgG was purified using Protein A chromatography. Culture supernatants were loaded on a column of appropriate size of Ceramic Protein A (BioSepra) and washed with 50 mM Tris-HCl pH 8.0, 250 mM NaCl. Bound IgG was eluted from the column using 0.1 M Sodium Citrate (pH 3.0) and neutralized by the addition of Tris-HCl (pH 9.0). The eluted material was buffer exchanged into PBS using Nap10 columns (Amersham, #17-0854-02) and the concentration of IgG was determined spectrophotometrically using an extinction coefficient based on the amino acid sequence of the IgG (Mach et al., Anal. Biochem. 200(1):74-80 (1992)). The purified IgG were analyzed for aggregation and degradation purity using SEC-HPLC and by SDS-PAGE. SEQ ID NOs corresponding to the various regions of antibodies IL330002, IL330004, IL330020, IL330071, IL330125, and IL330126 are shown in Table 5.

(103) TABLE-US-00009 TABLE 5 Anti-IL-33 Antibody Sequences VH VL VH CDRs VL CDRs IgG1 Sequence Sequence 1, 2, 3 1, 2, 3 IL330002 SEQ ID SEQ ID SEQ ID SEQ ID NO: 2 NO: 7 NO: 3 NO: 8 SEQ ID SEQ ID NO: 4 NO: 9 SEQ ID SEQ ID NO: 5 NO: 10 IL330004 SEQ ID SEQ ID SEQ ID SEQ ID NO: 12 NO: 17 NO: 13 NO: 18 SEQ ID SEQ ID NO: 14 NO: 19 SEQ ID SEQ ID NO: 15 NO: 20 IL330020 SEQ ID SEQ ID SEQ ID SEQ ID NO: 22 NO: 27 NO: 23 NO: 28 SEQ ID SEQ ID NO: 24 NO: 29 SEQ ID SEQ ID NO: 25 NO: 30 IL330071 SEQ ID SEQ ID SEQ ID SEQ ID NO: 32 NO: 37 NO: 33 NO: 38 SEQ ID SEQ ID NO: 34 NO: 39 SEQ ID SEQ ID NO: 35 NO: 40 IL330125 SEQ ID SEQ ID SEQ ID SEQ ID NO: 42 NO: 47 NO: 43 NO: 48 SEQ ID SEQ ID NO: 44 NO: 49 SEQ ID SEQ ID NO: 45 NO: 50 IL330126 SEQ ID SEQ ID SEQ ID SEQ ID NO: 52 NO: 57 NO: 53 NO: 58 SEQ ID SEQ ID NO: 54 NO: 59 SEQ ID SEQ ID NO: 55 NO: 60 IL330425 SEQ ID SEQ ID SEQ ID SEQ ID NO: 62 NO: 67 NO: 63 NO: 68 SEQ ID SEQ ID NO: 64 NO: 69 SEQ ID SEQ ID NO: 65 NO: 70 IL330428 SEQ ID SEQ ID SEQ ID SEQ ID NO: 72 NO: 77 NO: 73 NO: 78 SEQ ID SEQ ID NO: 74 NO: 79 SEQ ID SEQ ID NO: 75 NO: 80
Inhibition of IL-33 Binding to ST2 by Purified IgG

(104) The ability of anti-IL-33 antibodies to inhibit the binding of FLAG®-His tagged IL-33 to the ST2 receptor was assessed in a biochemical HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) competition assay, the principles of which are described above.

(105) Activity of purified IgG preparations were determined by competing a dilution series of the purified IgG against biotinylated human or mouse ST2.Fc for binding to FLAG® His-tagged human, cynomolgus or mouse IL-33.

(106) Purified or unpurified anti-IL-33 antibody samples were tested for inhibition of FLAG®-His tagged IL-33 binding ST2-Fc by adding 10 microlitres of each dilution of antibody test sample to a 384 well low volume assay plate (Costar 3676). Next, a solution containing 4 nM biotinylated human or mouse ST2-Fc and 20 nM streptavidin XL665 detection (Cisbio International, 611SAXLB) was prepared and 5 microlitres of the mix added to the assay plate. This was followed by the addition of 5 microlitres of a solution containing 1.2 nM FLAG®-His tagged human, cynomolgus or mouse IL-33 combined with 1.72 nM anti-FLAG® cryptate detection (Cisbio International, 61FG2KLB). All dilutions were performed in assay buffer comprised of Dulbeccos PBS (Invitrogen, 14190185) containing 0.8 M potassium fluoride (BDH 103444T) and 0.1% BSA (Sigma A9576). Assay plates were incubated for 2 hour at room temperature followed by 16 hour at 4° C. before reading time resolved fluorescence at 620 nm and 665 nm emission wavelengths using an EnVision plate reader (Perkin Elmer).

(107) Data were analyzed as described above using Equations 1 to 3.

(108) The negative control (non-specific binding) was defined by replacing test sample with 100 nM non-biotinylated ST2 prepared in a dilution buffer comprised of Dulbecco's PBS (Invitrogen, 14190185) containing 0.1% bovine serum albumin (BSA, Sigma A9576).

(109) The representative potencies (IC.sub.50) for purified IgG antibodies IL330002, IL330004, IL330020, IL330071, IL330125, and IL330126 are shown in Table 6.

(110) TABLE-US-00010 TABLE 6 IC.sub.50 results in the IL-33 FLAG ®-His/ST2-Fc competition assay Human IL-33 FLAG ® + Cynomolgus IL-33 FLAG ® + IgG Human ST2-Fc Human ST2-Fc IL330002 27 nM  42 nM IL330004 9 nM 170 nM IL330020 40 nM No inhibition IL330071 59 nM 375 nM IL330125 210 nM No inhibition IL330126 226 nM No inhibition

(111) All of the purified IgG preparations (i.e., IL330002, IL330004, IL330020, IL330071, IL330125, and IL330126) were shown to inhibit the human IL-33: human ST2 interaction. FIG. 3A: shows the inhibition of the FRET signal, produced by human IL-33 binding to human ST2 with increasing concentrations of IL-33 IgG1 antibodies IL330002, IL330004, IL330020, IL330071, IL330125 and IL330126, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(112) The IL330002, IL330004, and IL330071 IgG preparations were also shown to inhibit the cynomolgus IL-33: human ST2 interaction. FIG. 3B: shows the inhibition of the FRET signal, produced by cynomolgus monkey IL-33 binding to human ST2 with increasing concentrations of IL-33 IgG1 antibodies IL330002, IL330004, IL330020 and IL330071, IL330125 and IL330126, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding. No inhibition was detected in the mouse IL-33 FLAG®-His+mouse ST2-Fc competition assay by any of the above tested antibodies.

(113) Inhibition of NFκB Signaling in Hela-ST2 Reporter Cells by IgG

(114) A reporter assay was used to assess the inhibition of IL-33 induced NFκB signaling by anti-IL-33 antibodies IL330002, IL330004, IL330020, IL330071, IL330125, and IL330126, using Hela cells co-transfected with ST2 and an NFκB responsive luciferase reporter construct. Cells were exposed to IL-33 in the presence or absence of test antibody, and NFκB signaling was detected by measuring the activity of the luciferase subsequently produced. Hela cells containing luciferase reporter construct were sourced from Panomics. Human ST2 sequence was cloned into a lentiviral vector from System Biosciences. Lentiviral particles were generated in Ad293 cells (Stratagene) and used to transduce the Hela-luciferase reporter cells.

(115) In the reporter assay, stimulation of the ST2 receptor with IL-33 resulted in activation of the NFκB signaling pathway and through the NFκB promoter, initiated the expression of the enzyme luciferase. Following lysis of the cells, a luciferase substrate was added, which underwent a chemical reaction in the presence of luciferase to produce a luminescent product. The amount of light detected from the cell lysate was quantified using an Envision plate reader (PerkinElmer) and used as a direct measure of IL-33 mediated NFκB signaling.

(116) Hela transfected cells were maintained in media containing hygromycin B to maintain stable receptor expression. Cells were exposed to IL-33 in the presence or absence of test antibody, and NFκB signalling detected by measuring the activity of the luciferase subsequently produced.

(117) Transfected Hela cells were seeded at 1×10.sup.4 cells/well (50 microlitres per well) in DMEM culture medium (Invitrogen, 41966) containing 10% v/v Fetal Bovine Serum (heat inactivated) and 100 microgram per mL Hygromycin B (Invitrogen 10687-010) into 384 well black-walled, Poly-D-Lysine coated plates (Greiner, 781946). Plates were incubated at 37 degrees Celsius, 5% CO.sub.2 for 18-24 hours, and then cell medium was gently aspirated from the wells prior to addition of test samples.

(118) Serial dilutions of samples were prepared by dilution in DMEM culture medium (Invitrogen 41966) containing 10% v/v FBS (heat inactivated) and 100 microgram per mL Hygromycin B (Invitrogen, 10687-010). Fifteen microlitres of test sample was added to cells in duplicate. IL-33 FLAG®-His was diluted to 0.6 nM in DMEM culture medium (Invitrogen, 41966) containing 10% v/v FBS (heat inactivated) and 100 microgram per mL Hygromycin B (Invitrogen 10687-010) and 15 microlitres added to the cells and test samples. This concentration represented the EC50 value of the reporter cell response to IL-33 FLAG®-His (Geomean 0.32 nM, 95% confidence intervals 0.25-0.40 nM, n=5). Background response was defined by the addition of 30 microlitres DMEM culture medium (Invitrogen, 41966) containing 10% v/v FBS (heat inactivated) and 100 microgram per mL Hygromycin B (Invitrogen, 10687-010). Plates were incubated at 37 degrees Celsius, 5% CO.sub.2 for 4 hours and at room temperature for 1 hour.

(119) To measure production of luciferase in response to NFκB signaling, 30 microlitres of Bright Glo® lysis buffer combined with luciferase substrate (Promega, E2620) was added to the plate and incubated at room temperature for 5 minutes. Luminescence produced as a result of oxidation of the substrate by luciferase is read using an EnVision plate reader (PerkinElmer).

(120) The relative light unit (RLU) values were subsequently used to calculate % specific response as described in equation 5:

(121) $\%\mathrm{specific}\mathrm{response}=\left(\frac{\left(\mathrm{Sample}\mathrm{RLU}-\mathrm{Background}\mathrm{RLU}\right)}{\left(\mathrm{Total}\mathrm{RLU}-\mathrm{Background}\mathrm{RLU}\right)}\right)\times 100$

(122) IC.sub.50 values were determined using GraphPad Prism software by curve fitting using a four-parameter logistic equation (Equation 4):

(123) Purified IgG preparations of antibodies IL330002, IL330004, IL330020, IL330071, IL330125, and IL330126 were shown to inhibit the NFκB-driven luciferase activity with representative potencies (IC50) shown in Table 7.

(124) TABLE-US-00011 TABLE 7 IC.sub.50 results in the human ST2 transfected HeLa NFkB reporter assay IgG Human IL-33 FLAG ®-His IL330002 226 nM IL330004 23 nM IL330020 147 nM IL330125 332 nM IL330126 245 nM IL330071 226 nM

(125) FIG. 4A shows the inhibition of NFκB activity in luciferase NFκB reporter assay by IL-33 antibodies IL330002, IL330004 IL330020, IL330071, IL330125, and IL330126 compared to a negative control IgG.

(126) Inhibition of NFκB Signaling in Huvec by IgG

(127) NFκB signaling in Human umbilical vein endothelial cells (Huvecs) in response to IL-33 was assessed by nuclear translocation of the p65/RelA NFkB subunit detected by immunofluorecence staining. Imaging and quantification of the nuclear staining intensity was performed on ArrayScan VTi HCS Reader (Cellomics).

(128) Huvecs were obtained from Cambrex and maintained in complete EBM-2 media (Lonza) according to recommended protocol. Huvecs were harvested from flasks with accutase (PAA, #L11-007) and seeded at 1×10.sup.4/1000 well in culture media [EBM-2 (Lonza, #CC-3156) with EGM-2 SingleQuot Kit Suppl. & Growth Factors (Lonza, #CC-4176)] into 96-well black walled, clear flat-bottomed Collagen I coated plates (Greiner) and incubated at 37° C., 5% CO.sub.2 for 18-24 hours. After this time, media was aspirated, leaving cell monolayer intact, and replaced with assays test samples as prepared below.

(129) Test samples of purified IgG (in duplicate) were diluted to the desired concentration in complete culture media in 96 well U-bottom polypropylene plates (Greiner, 650201). IL-33 (Adipogen) was prepared in complete culture media mixed with appropriate test antibody to give a final IL-33 concentration of 1 ng/mL in a total volume of 120 μl/well. All samples were incubated for 30 mins at 37° C., prior to transfer of 100 μl of IL-33/antibody mixture to the assay plate. Following 30 minute incubation at 37° C., media were aspirated, leaving the cell monolayer intact and cells were fixed for 15 minutes with 3.7% formaldehyde solution that had been pre-warmed to 37° C. Fixative was aspirated and cells were washed twice with 100 μL/well of PBS.

(130) Cells were stained for NFκB using a Cellomics NFκB assay kit (Thermo Scientific, #8400492) according to manufacturer's instructions. Briefly, cells were permeabilised for 15 minutes at room temperature, blocked for 15 minutes and stained for 1 hour with primary antibody solution in a volume of 50 μL. Plates were washed×2 in blocking buffer and stained for 1 hour at room temperature with secondary antibody solution, which included Hoechst nuclear stain as well as secondary antibody. Plates were washed×2 in PBS. Cells were stored in a final volume of 150 μL/well PBS and covered with a black, light-blocking seal (Perkin Elmer, #6005189) before reading on ArrayScan VTi HCS Reader. The intensity of nuclear staining was calculated using a suitable algorithm. Data were analysed using Graphpad Prism software. IC.sub.50 values were determined by curve fitting using a four-parameter logistic equation (Equation 4).

(131) FIG. 4B shows the inhibition of NFκB activity in Huvec NFκB translocation assay by IL-33 antibody IL330004 compared to Anti-NIP IgG1 negative control antibody, NIP228. Nuclear translocation of p65/RelA NFκB was inhibited by antibody IL330004. When tested as a purified IgG, the IC.sub.50 for antibody IL330004 was calculated as being 12 nM.

(132) Binding Affinity Calculation for IL-33 Antibodies Using BIAcore

(133) The binding affinity of purified IgG samples of exemplary binding members to human and cynomolgus IL-33 was determined by surface plasmon resonance using BIAcore 2000 biosensor (BIAcore AB) essentially as described by Karlsson et al., J Immunol Methods 145(1-2):229-40 (1991). In brief, Protein G′ (Sigma Aldrich, P4689) was covalently coupled to the surface of a CM5 sensor chip using standard amine coupling reagents according to manufacturer's instructions (BIAcore). The protein G′ surface was used to capture purified anti-IL-33 antibodies via the Fc domain to provide a surface density of approximately 290 RU per cycle. Human or cynomolgus IL-33 prepared in HBS-EP buffer (BIAcore AB), at a range of concentrations, between 600 nM and 18.75 nM, were passed over the sensor chip surface. The surface was regenerated using two 10 mM Glycine washes of pH 1.7 and pH 1.5 between each injection of antibody. The resulting sensorgrams were evaluated using BIA evaluation 3.1 software and fitted to a 1:1 Langmuir binding model, to provide relative binding data. The binding results (KD, Ka, and Kd) for antibodies IL330002 and IL330004 binding to human or cynomolgus IL-33 are shown in Table 8.

(134) TABLE-US-00012 TABLE 8 BIAcore binding affinity of exemplary binding members KD Ka Kd Antibody Antigen (nM) (1/Ms) (1/s) IL330002 Human IL-33 35 6.16E+04 2.18E−03 FLAG ®-His IL330002 Cynomolgus IL-33 189 2.34E+04 4.42E−03 FLAG ®-His IL330004 Human IL-33 6 1.96E+05 1.10E−03 FLAG ®-His IL330004 Cynomolgus IL-33 365 1.25E+04 4.58E−03 FLAG ®-His
Binding of IL-33 Antibodies to Intracellular IL-33

(135) Selection and activity studies described above used recombinant or commercial sources of mature IL-33 (amino acids 112-270). Studies suggest full length IL-33 may also be active (Cayrol et al., Proc Natl Acad Sci USA 106(22):9021-6 (2009); Hayakawa et al., Biochem Biophys Res Commun 387(1):218-22 (2009); Talabot-Ayer et al., J Biol Chem. 284(29):19420-6 (2009)). The binding of antibodies to full length (“native”) IL-33 was determined by immunofluorescence staining of primary bronchial smooth muscle cells (BSMC).

(136) BSMC were obtained from Cambrex and maintained in complete smooth muscle growth media (SmBM®, Lonza) according to manufacturer's instructions. Cells were harvested from flasks with accutase (PAA #L11-007) and seeded at 2×10.sup.4/100 μl/well in culture media [SMBM (Lonza #CC-3181) with SMGM SingleQuot Kit Suppl. & Growth Factors (Lonza #CC-4149)] into 96-well black walled, clear flat-bottomed Collagen I coated plates (Greiner) and incubated at 37° C., 5% CO.sub.2 for 18-24 hours. After this time, media was aspirated, leaving cell monolayer intact, and cells were fixed for 15 minutes with 3.7% formaldehyde solution that had been pre-warmed to 37° C. Fixative was aspirated and cells were washed twice with 100 μL/well of PBS. Cells were permeabilised for 15 minutes at room temperature using permeabilisation buffer (Thermo Scientific, #8400492), washed×2 in PBS and blocked with 100 μL/well of PBS/1% BSA (Sigma, #A9576) for 30-60 minutes at room temperature. Blocking buffer was flicked out and replaced with a titration of anti-IL-33 or suitable isotype control antibodies, that had been diluted in blocking buffer, for 1 hour at room temperature.

(137) Plates were washed×2 in PBS and stained for 1 hour at room temperature with secondary antibody solution, which included Hoechst dye (10 mg/mL; Thermo Scientific) diluted 1:10000 as well as secondary antibody (Anti-Human IgG (H+L), Alexa Fluor® 488 conjugate 2 mg/mL; Invitrogen, #A11013) diluted 1:1000. Plates were washed three times in PBS. Cells were stored in a final volume of 150 μL/well PBS and covered with a black, light-blocking seal (Perkin Elmer, #6005189) before imaging on ArrayScan VTi HCS Reader.

(138) IL-33 expression by cultured BSMC was confirmed with a commercial polyclonal antibody (R&D Systems, #AF3625), detected with anti-Goat IgG (H+L), Alexa Fluor® 488 conjugate 2 mg/mL; Invitrogen, #A11055). FIG. 5 shows detection of endogenous IL-33 in bronchial smooth muscle cells by immunofluorescence staining by IL-33 antibody IL330004 (right panel) compared to CAT-002 negative control (left panel). Antibody IL330004 showed a clear nuclear staining of BSMC, corresponding with the expected localization of full length IL-33, and that detected with the commercial pAb.

Example 2 Isolation and Identification of Anti-IL-33 scFv Antibodies

(139) Identification of IL-33 Specific Binders by Phage ELISA

(140) Reagents and selections were as described in Example 1. Single-chain Fv fragments were displayed on phage particles and tested as unpurified preparations in a single point ELISA screen. Phage-displayed scFv were considered to bind the IL-33 antigen if the absorbance 450 nm was >0.5, and <0.1-0.2 for the same sample on controls (insulin and IL-4Rα Flag® His).

(141) FIG. 6 shows data from a single plate screened against human IL-33, cynomolgus IL-33 and insulin. One specific human/cynomolgus cross-reactive IL-33 binder is shown in well C4, and wells A12 and B12 contain control IL-33 binding clone.

(142) Identification of IL-33 Binders by Axxora IL33305B Competition

(143) The Axxora IL33305B Competition Assay is a homogeneous assay that utilizes Fluorescence Microvolume Assay Technology (FMAT). The assay assessed the inhibition of Axxora IL33305B mAb (Axxora/Adipogen, #AG-20A-0041-0050) binding to recombinant biotinylated human IL-33 Flag® His in the presence of crude scFv supernatant samples or purified scFv and IgG in a 384-well format.

(144) ScFv were expressed in the bacterial periplasm and screened for their inhibitory activity in an FMAT epitope competition assay against a known biologically active IL33305B mAb. Biotinylated IL-33 was immobilized on streptavidin coated beads (Spherotec, #SVP-60-5) and the interaction with Axxora IL33305B Ab was detected using a goat anti-mouse Alexafluor®-647 labeled antibody (Molecular Probes A21236). The FMAT system is a macroconfocalimager, which measures the red fluorescence associated with the beads.

(145) Plates were read on the Applied Biosystems Cellular Detection system 8200 reader. The Helium neon excitation laser focuses within 100 μm depth of the bottom of the well scanning an area 1 mm.sup.2. The beads settle at the bottom of the well and upon laser excitation at 633 nm those beads with fluorophore bound (where the local concentration of fluorophore is relatively high compared to unbound fluorophore) emit a signal at 650-685 nm that is measured using PMT1. Unbound fluorophore in solution is outside the excitation depth or at a relatively low local concentration and thus does not emit a significant signal. ScFv or IgG samples that effectively block IL33305B binding to IL-33 will therefore cause a reduction in the amount of bead:IL-33:IL33305B:anti-mouse Alexafluor®-647 labeled antibody complexes at the bottom of the well which results in a reduction in measured fluorescence.

(146) For the assay setup, the following were prepared:

(147) (1) IL33305B and anti-mouse AF647 mix, IL33305B was diluted to 2.25 nM in assay buffer [PBS (Gibco, 14190-094) containing 0.1% BSA (Sigma, #A9576) and 0.1% Tween-20 (Sigma, P2287)] and mixed with anti-mouse AF647 diluted to 2 μg/ml (final 400 ng/ml) in assay buffer.
(2) IL-33 and bead mix, 2.5 nM biotinylated human IL-33 FLAG® His was added to 0.0095% w/v streptavidin beads in assay buffer and incubated with rotation at room temperature for 1 hour—before use these particles were spun down at 2000 rpm for 15 minutes and resuspended in original volume assay buffer.
(3) Sample Preparation, crude scFv supernatant samples were generated in 96 deep well plates. 5 μl culture from each well of a 96-well master plate was transferred into a Greiner deep well culture plate containing 900 μl of 2TY+100 μg/ml ampicillin+0.1% glucose media and incubated for 5 hours at 37° C., 280 rpm. 10 mM IPTG in TY was then added at 100 ul/well and the plate incubated overnight at 30° C., 280 rpm. The next morning, the plate was spun down at 3200 rpm for 15 minutes. For high-throughput screening, scFv supernatants from the deep well plate were transferred directly to the assay plate to achieve a final concentration of 20%.

(148) For IC50 determinations, typically purified scFv or IgGs were diluted 3-fold in assay buffer, in duplicate, to give 11 concentration points. 96 well Greiner polypropylene (Greiner, 650201) plates are used for dilution preparation.

(149) Into columns 1-22 of a 384-well clear bottomed non-binding surface black plate (Costar, #3655), the following were added: 10 μl sample, 20 μl IL33305B/anti-mouse AF647 mix, and 20 μl IL-33/bead mix. In all cases, total well volume was 40 μl. Controls typically used in these experiments included: IL-33/bead mix plus anti-mouse AF647 was added (non-specific binding); IL330305B/anti-mouse AF647 mix plus IL-33/bead mix (total binding). The plates were sealed and incubated for four hours at room temperature in the dark and then read on the Applied Biosystems Cellular Detection system 8200 reader. Data was analyzed with the Velocity algorithm, with gating set as color ratio <0.4, size <15 and minute count 20. Hits from the crude scFv supernatant samples were defined as showing 50% or greater inhibition of signal compared to the total binding control wells. Dose response curves were plotted for purified scFv and IgG titrations using Prism (Graphpad) curve-fitting software.

(150) Reformatting of scFv to IgG1

(151) scFv that displayed a desirable species cross-reactivity and specificity profile as determined by phage-displayed scFv binding experiments or showed an inhibitory effect in the epitope competition assay against Axxora Il33305B (as described above), were subjected to DNA sequencing (Osbourn et al., Immunotechnology 2(3):181-96 (1996); Vaughan et al., Nat. Biotechnol. 14(3):309-14 (1996)). First, scFv with desirable properties were converted to whole immunoglobulin G1 (IgG1), or an effector-null isotype IgG1 TM (IgG1 Fc sequence incorporating mutations L234F, L235E and P331S), antibody format as described in Example 1. SEQ ID NOs corresponding to the various regions of antibodies IL330065, IL330099, IL330101, IL330107, IL33149, and IL330180 are shown in Table 9.

(152) TABLE-US-00013 TABLE 9 Anti-IL-33 Antibody Sequences VH VL VH CDRs VL CDRs IgG1 Sequence Sequence 1, 2, 3 1, 2, 3 IL330065 SEQ ID SEQ ID SEQ ID SEQ ID NO: 82 NO: 87 NO: 83 NO: 88 SEQ ID SEQ ID NO: 84 NO: 89 SEQ ID SEQ ID NO: 85 NO: 90 IL330099 SEQ ID SEQ ID SEQ ID SEQ ID NO: 92 NO: 97 NO: 93 NO: 98 SEQ ID SEQ ID NO: 94 NO: 99 SEQ ID SEQ ID NO: 95 NO: 100 IL330101 SEQ ID SEQ ID SEQ ID SEQ ID NO: 102 NO: 107 NO: 103 NO: 108 SEQ ID SEQ ID NO: 104 NO: 109 SEQ ID SEQ ID NO: 105 NO: 110 IL330107 SEQ ID SEQ ID SEQ ID SEQ ID NO: 122 NO: 127 NO: 123 NO: 128 SEQ ID SEQ ID NO: 124 NO: 129 SEQ ID SEQ ID NO: 125 NO: 130 IL330149 SEQ ID SEQ ID SEQ ID SEQ ID NO: 132 NO: 137 NO: 133 NO: 138 SEQ ID SEQ ID NO: 134 NO: 139 SEQ ID SEQ ID NO: 135 NO: 140 IL330180 SEQ ID SEQ ID SEQ ID SEQ ID NO: 142 NO: 147 NO: 143 NO: 148 SEQ ID SEQ ID NO: 144 NO: 149 SEQ ID SEQ ID NO: 145 NO: 150
Binding Assay for IgGs

(153) Species cross-reactivity of anti-IL-33 antibodies was determined using a plate-based ELISA. Streptavidin plates (Thermo Scientific, AB-1226) were coated with biotinylated antigen at 0.5 μg/ml in PBS. Binding of purified IgG preparations was detected with anti-human IgG HRP (Sigma, A0170). EC50 data for binding curves are shown in Table 10.

(154) TABLE-US-00014 TABLE 10 Binding of IL-33 antibodies to Flag ®His- tagged human, cynomolgus or mouse IL-33 EC50 (nM) Human IL-33 Cynomolgus IL-33 Mouse IL-33 Antibody Flag ®His Flag ®His Flag ®His IL330065 0.65 0.66 No binding IL330099 0.40 0.35 0.38 IL330101 1.19 1.20 0.85 IL330107 0.86 1.09 0.83 IL330149 0.23 0.35 0.17 IL330180 Not determined* Not determined* Not determined* IL330180 was determined to bind to human IL-33, but not Cynomolgus or mouse IL-33 in phage-displayed scFv format.
Inhibition of IL-33 Functional Responses by Anti-IL-33 Antibodies
Inhibition of TF-1 Cell Proliferation by IgG

(155) A cell viability assay was used to assess the inhibition of IL-33 induced proliferation/survival from TF-1 cells by anti-IL-33 antibodies. The CellTiter-Glo® Luminescent Cell Viability Assay (Promega) is a homogeneous method to determine the number of viable cells in culture based on quantitation of the ATP present, which signals the presence of metabolically active cells. Cells were exposed to IL-33 in the presence or absence of test antibody. Cell viability was measured by CellTiter-Glo following 72 hour stimulation with IL-33.

(156) In particular, the proliferation assay was used to assess the inhibition of IL-33 induced proliferation from TF-1 cells by anti-IL-33 antibodies. TF-1 cells were a gift from R&D Systems and maintained according to manufacturer's instructions. The assay media comprised RPMI-1640 with GLUTAMAX I (Invitrogen, 61870) containing 5% fetal bovine serum (heat inactivated, gamma-irradiated), 1% sodium pyruvate (Sigma, S8636), 1-2% Penicillin/streptomycin (Invitrogen, 15140-122). Prior to each assay, TF-1 cells were pelleted by centrifugation at 300×g for 5 minutes, the media were removed by aspiration, and the cells were then re-suspended in assay media. This process was repeated twice with cells re-suspended at a final concentration of 2×10.sup.5 cells/ml in assay media. Test solutions of IgG (in duplicate) were titrated to the desired concentration range in assay media in 96 well U-bottom polypropylene plates (Greiner, 650201) and 50 μL transferred to 96-well flat-bottomed tissue culture-treated plates (Costar, #3598). Recombinant human IL-33 (Alexis, ALX-522-098-3010) was added to the appropriate test antibody titrations to give a total volume of 100 μl/well. 100 μl of cell suspension was then added to 100 μl of IL-33 or IL-33 and antibody mixture to give a total assay volume of 200 μl/well and total cell number of 20,000 per well. A final assay concentration of 100 ng/mL IL-33 was used in the assay, which was selected as the dose that gave approximately 80% of maximal proliferative response. Plates were incubated for 72 hours at 37° C. and 5% CO.sub.2. 100 μL of supernatant was carefully removed from the assay plates. 100 μL of CellTiter-Glo (Promega, G7571), reconstituted according to manufacturers instructions, was added per well. Plates were shaken on a plate shaker at 500 rpm for 5 minutes and luminescence read on EnVision plate reader (PerkinElmer). Data were analyzed using Graphpad Prism software. IC50 values were determined by curve fitting using a three or four-parameter logistic equation.

(157) For those antibodies that achieved full inhibition curves, IC50 values were calculated and are summarized in Table 11 below. Purified IgG preparations were capable of inhibiting TF-1 proliferation in response to IL-33. FIG. 7A shows percent inhibition for IL330065 and IL330101 (compared to control mAb and hST2/Fc) for the TF-1 proliferation assay, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is a percentage of the maximum response.

(158) Inhibition of Huvec IL-6 Release by IL-33 Antibodies

(159) A cytokine release assay was used to assess the inhibition of IL-33 induced IL-6 production from human umbilical vein endothelial cells (Huvec) by anti-IL-33 antibodies. Cells were exposed to IL-33 in the presence or absence of test antibody.

(160) Huvecs were obtained from Cambrex and maintained in complete EBM-2 media (Lonza) according to manufacturer's protocol. Cells were harvested from flasks with accutase (PAA, #L11-007) and seeded at 1×10.sup.4/1000 well in culture media (EBM-2 (Lonza, #CC-3156) with EGM-2 SingleQuot Kit Suppl. & Growth Factors (Lonza, #CC-4176)) into 96-well flat-bottomed tissue-culture treated plates (Costar, #3598) and incubated at 37° C., 5% CO.sub.2 for 18-24 hours. After this time, media was aspirated, leaving cell monolayer intact, and replaced with assay test samples as discussed below.

(161) Test solutions of purified IgG (in duplicate) were diluted to the desired concentration in complete culture media in 96 well U-bottom polypropylene plates (Greiner, 650201). IL-33 (Adipogen) was prepared in complete culture media mixed with appropriate test antibody to give a final IL-33 concentration of 30 ng/mL. All samples were incubated for 30 minutes at room temperature, prior to transfer of 120 μl of IL-33/antibody mixture to the assay plate. Following 18-24 hour incubation, IL-6 was measured in cell supernatants by ELISA (R&D Systems, DY206) adapted for europium readout. Black Fluro-Nunc Maxisorp plates (VWR, #437111) were coated with 50 μL capture antibody, washed three times with PBS-Tween (0.01%) using an automated plate washer (Biotek) and blocked with 250 μL/well of PBS/1% BSA (Sigma, #A9576) for 1-2 hours at room temperature. Plates were washed as above and incubated with 50 uL mast cell assay supernatants for 1-2 hours at room temperature. Following 3× wash with PBS-Tween, plates were incubated with detection antibody (50 uL/well) according to manufacturer's instructions. ELISA plates were washed three times in PBS-Tween, and Streptavidin-Europium (PerkinElmer, 1244-360) was diluted 1:1000 in DELFIA® assay buffer (PerkinElmer, 4002-0010) and added at 50 μl/well for 45-60 minutes at room temperature. Plates were then washed 7 times in DELFIA wash buffer before the addition of 50 μl/well of enhancement solution (PerkinElmer, 4001-0010) and analyzed using time resolved fluorometry (Excitation 340 nM, Emission 615 nM). Data were analyzed using Graphpad Prism software. IC50 values were determined by curve fitting using a three or four-parameter logistic equation. For those antibodies that achieved full inhibition curves, IC50 values were calculated and are summarized in Table 11 below. Purified IgG preparations (IgG1 or IgG1-TM) of antibodies IL330065, IL330099, IL330101, IL330107, IL330149, and IL330180 inhibited IL-6 production in comparison with control antibodies. Potency of exemplary binding members was essentially unaffected by the presence of IgG1-TM Fc sequence mutations (L234F, L235E and P331S), and the data as shown combines information for both formats. FIG. 7B shows percent maximal IL-6 release for IL330065 and IL330101 (compared to human ST2-Fc, anti-IL33 pAb AF3625 (R & D Systems), and control mAb), wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is a percentage of the maximum response.

(162) Inhibition of Human Mast Cell Cytokine Release

(163) A cytokine release assay was used to assess the inhibition of IL-33 induced IL-6 production by anti-IL-33 antibodies from human mast cells. In addition to IL-6, other cytokines (IL-5, IL-6, IL-8, IL-10, IL-13, GM-CSF and TNFα) in cell supernatants were measured using alternative duoset ELISAs or Mesoscale Discovery multiplex analysis.

(164) Human mast cells were produced by in vitro differentiation of cord blood CD133+ progenitor cells (Lonza, #2C-108) essentially as described in Andersen et al. (J Immunol Methods 336: 166-174 (2008)). Progenitor cells were thawed according to manufacturer's instructions and cultured in vitro in Serum Free Expansion Media (StemSpan, #09650) supplemented with 1% Penicillin/streptomycin (Invitrogen, 15140-122) and growth factors: 100 ng/mL Stem Cell Factor (Peprotech, #AF-300-07) and 50 ng/mL IL-6 (Peprotech, #AF-200-6) for 8 weeks. In addition, 1 ng/mL IL-3 (R&D Systems, #203-IL) was included in the culture media during the first three weeks. Cells were maintained throughout at <5×10.sup.5/mL.

(165) Mast cells were cultured overnight in assay media (StemSpan, #09650; 1% Penicillin/streptomycin (Invitrogen, 15140-122) and 100 ng/mL Stem Cell Factor (Peprotech, #AF-300-07)), prior to exposure to IL-33 in the presence or absence of test antibody.

(166) For cytokine release assays, cells were removed, pelleted (150 g for 10 minutes) and resuspended in assay media (StemSpan, #09650, 1% Penicillin/streptomycin (Invitrogen, 15140-122) and 100 ng/mL Stem Cell Factor (Peprotech, #AF-300-07)). Cells were returned to a flask and cultured for 18-24 hours prior to assay setup. For sample assessment, test solutions of IgG (in duplicate) were titrated to the desired concentration range in assay media in 96 well U-bottom polypropylene plates (Greiner, 650201) and 50 μL of test solutions transferred to 96-well flat-bottomed tissue-culture treated plates (Costar, #3598). 50 μL of recombinant human IL-33 (Adipogen, #522-098-3010), diluted in assay media to 90 ng/mL was added to the appropriate test antibody titrations to give a total volume of 100p1/well. 50p1 of cell suspension (1.5×10.sup.5) was then added to 100p1 of IL-33 or IL-33 and antibody mixture to give a total assay volume of 150p1/well and total cell number of 5×10.sup.4 per well. A final assay concentration of 30 ng/mL IL-33 was used in the assay, selected as the dose that gave approximately 50-80% of maximal cytokine response. Plates were incubated for 18-24 hours at 37° C. and 5% CO.sub.2.

(167) IL-6 was measured in cell supernatants by ELISA (R&D Systems, DY206) adapted for europium readout. Black Fluro-Nunc Maxisorp plates (VWR, #437111) were coated with 50 μL capture antibody, washed three times with PBS-Tween (0.01%) using an automated plate washer (Biotek) and blocked with 250 μL/well of PBS/1% BSA (Sigma, #A9576) for 1-2 hours at room temperature. Plates were washed as above and incubated with 50 μL mast cell assay supernatants for 1-2 hours at room temperature. Following 3× wash with PBS-Tween, plates were incubated with detection antibody (50 μL/well) according to manufacturer's instructions. ELISA plates were washed three times in PBS-Tween, and Streptavidin-Europium (PerkinElmer, 1244-360) was diluted 1:1000 in DELFIA assay buffer (PerkinElmer, 4002-0010) and added at 50p1/well for 45-60 minutes at room temperature. Plates were then washed 7 times in DELFIA wash buffer before the addition of 50p1/well of enhancement solution (PerkinElmer, 4001-0010) and analyzed using time resolved fluorometry (Excitation 340 nM, Emission 615 nM). Data were analyzed using Graphpad Prism software. IC50 values were determined by curve fitting using a three or four-parameter logistic equation.

(168) FIG. 8A shows the reduction of IL-6 production by increasing concentrations of antibodies IL330065, IL330099, IL330101, IL330107, IL33149, and IL330180, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is % maximum response. Purified IgG preparations of test antibodies were capable of inhibiting the IL-6 activity in comparison with control antibodies. Potency of exemplary binding members was essentially unaffected by the presence of IgG1-TM Fc sequence mutations (L234F, L235E and P331S) and the data combines information for both formats. IC50 results for TF-1 proliferation assay, HUVEC IL-6 production, and mast cell IL-6 production are shown in Table 11.

(169) TABLE-US-00015 TABLE 11 Example potencies of clones identified from naïve human scFv phage display libraries IgG1 Geomean (95% CI) IC50 (nM) TF-1 Huvec IL-6 Mast cell IL-6 Antibody proliferation production production IL330065 81 (19-345) 87 (60-129) 12 (4-40) IL330099 Incomplete curve Incomplete curve 1221 (n = 1) IL330101 149 (n = 1) 497 (167-1474) 49 (31-75) IL330107 94 (n = 1) Incomplete curve 283 (154-522) IL330149 Incomplete curve Incomplete curve 298 (n = 2) IL330180 Not Determined Not Determined 22 (n = 1)

(170) Additional cytokines (IL-5, IL-6, IL-8, IL-10, IL-13, and GM-CSF) in cell supernatants were detected using Meso-Scale Diagnostics Demonstration 10-plex human cytokine assay (#K15002B-1) according to manufacturer's instructions. Cytokines were measured in cell supernatants by ELISA adapted for europium readout using a similar protocol to the IL-6 ELISA described above.

(171) Mast cells were shown to produce a range of cytokines after stimulation with IL-33 (Meso-Scale Diagnostics Demonstration 10-plex human cytokine assay #K15002B-1; R&D Systems, #DY213). FIGS. 8B-F show the inhibition of IL-33-driven production of GM-CSF, IL10, IL-8, IL-13 and IL-5, respectively. These results show that antibodies IL330065, IL330101, IL330107, and IL330149 were able to inhibit IL-33-driven production of all cytokines measured.

(172) Binding and Neutralization of Native Full Length IL-33

(173) Selection and activity studies described above used recombinant in house or commercial sources of mature IL-33 (amino acids 112-270). Full length IL-33 may also be active (Cayrol et al., Proc Natl Acad Sci USA 106(22):9021-6 (2009); Hayakawa et al., Biochem Biophys Res Commun 387(1):218-22 (2009); Talabot-Ayer et al., J Biol Chem. 284(29):19420-6 (2009)). To assess binding of antibodies to full length IL-33, full length IL-33 was cloned and expressed in HEK293-EBNA cells. As described below, selected antibodies were shown to bind to full length IL-33 as determined by Western blot.

(174) Cloning and Expression of Full-Length Human IL-33

(175) The cDNA molecule encoding full-length (FL) IL-33 from human (Swiss Prot accession number O95760 amino acids 1-270) was synthesized by primer extension PCR cloning and cloned into pDONR221 (Invitrogen, 12536-017) and was transferred to the mammalian expression vector pDEST12.2 (Invitrogen) using LR Gateway Clonase II enzyme according to the manufacturer's instructions (Invitrogen, 12538-120). The pDEST12.2 vector had been modified to contain the oriP origin of replication from the pCEP4 vector (Invitrogen) allowing episomal plasmid replication upon transfection into cell lines expressing the EBNA-1 gene product (such as HEK293-EBNA cells). HEK293-EBNA cells were transfected with Lipofectamine 2000 (Invitrogen, 11668-019). Cells expressing FL HuIL-33 (and mock-transfected controls) were lysed using sonication in the presence of protease inhibitors (Roche, 05892791001).

(176) Western Blot Analysis of Cell Lysates Expressing Full-Length Human IL-33

(177) Proteins from cell lysates were denatured and reduced with SDS sample buffer and DTT prior to separation by SDS-PAGE electophoresis and transfer to Nitrocellulose membranes. Membranes were blocked with 5% non-fat dried milk in PBS-T for 1 h, incubated with primary antibody (0.5 m per ml) for 1 h, washed three times in PBS-T, then incubated for 1 h with HRP-conjugated secondary antibody (1 in 10,000 dilution of goat anti-human IgG (Sigma, A0170)) and washed three times in PBS-T. HRP was detected with Amersham ECL plus detection reagent (GE healthcare, RPN2132). Sizes were estimated by comparing migration to that of Magic Mark XP (Invitrogen, LC5602). FIG. 9 shows binding of IL-33 antibodies (IL330065, IL330101, IL330107, and IL330149) to full length human IL-33 by Western blot.

(178) Neutralization of Mast Cell Cytokine Responses to Full Length IL-33 Cell Lysate

(179) HEK293-EBNA cells expressing full length (FL) HuIL-33 (and mock-transfected controls) were harvested 24 hours following transfection with accutase (PAA, #L11-007). Cells were diluted to 5×10.sup.7/mL with PBS and homogenized for 30 seconds using a tissue homogenizer. Cell debris was removed by centrifugation. Mast cells were stimulated with cell lysates at varying concentrations. Stimulation of cytokine production was only observed with full length IL-33-transfected cell lysate and not with mock transfected cell lysate. A concentration of lysate that stimulated a sub-maximal cytokine release (approx EC50) was selected for antibody neutralization studies.

(180) For cytokine release assays, mast cells were cultured overnight in assay media (StemSpan, #09650; 1% Penicillin/streptomycin (Invitrogen, 15140-122) and 100 ng/mL Stem Cell Factor (Peprotech, #AF-300-07)), prior to exposure to FL HuIL-33 in the presence or absence of test antibody. IL-6 and IL-13 production was detected by ELISA of assay supernatants after 18-24 hours. A detailed description of the protocol was described above (Example 2-0007).

(181) FIG. 10 shows the effect of anti-IL-33 antibodies IL330065 and IL330101 on mast cell IL-6 and IL-13 production stimulated by cell lysates of full length IL-33-transfected cells. Purified IgG preparations were capable of inhibiting IL-6 (FIG. 10A) and IL-13 (FIG. 10B) production induced by full length IL-33 cell lysates.

(182) Non-Competitive Mode of Action of IL-33 Antibodies

(183) Inhibition of IL-33 Binding to ST2 by Purified IgG

(184) The ability of anti-IL-33 antibodies to inhibit the binding of Flag®-His tagged IL-33 to the ST2 receptor was assessed in a biochemical HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) competition assay the full methods of which are described in Example 1.

(185) FIG. 11A shows specific binding results for HTRF® receptor-ligand competition assay with increasing concentrations of antibodies IL330065, IL330099, IL330101, IL330107, IL33149, and IL330180. These results show that antibodies IL330065, IL330099, IL330101, IL330107, IL33149, and IL330180 are not competitive inhibitors of the IL-33:ST2 interaction.

(186) Inhibition of NFkB Signalling in Huvec by IgG

(187) NFkB signaling in Huvecs in response to IL-33 was assessed by nuclear translocation of the p65/RelA NFkB subunit, detected by immunofluorecence staining as described in Example 1.

(188) FIG. 11B shows Huvec NFkB translocation with increasing concentrations of antibodies IL330065, IL330099, IL330101, IL330107, and IL330149. These results show that IL330065, IL330099, IL330101, IL330107, and IL330149 did not inhibit nuclear translocation of p65/RelA NFkB in IL-33-stimulated Huvecs 30 minutes following stimulation. The results are consistent with failure of antibodies IL330065, IL330099, IL330101, IL330107, IL33149, and IL330180 to inhibit IL-33 binding to ST2.

(189) Epitope Binning of IL-33 Antibodies in HTRF® Epitope Competition Assays

(190) The ability of antibodies to compete with mAb IL330101 or mAb IL330180 for binding to biotinylated human IL-33 was assessed in a biochemical HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) epitope competition assay.

(191) The HTRF® epitope competition assays, described below, were used to measure the binding of a lead antibody in IgG format to biotinylated IL-33. Test scFv samples which recognize a similar epitope to the lead antibody, will compete with the lead antibody for binding to IL-33, leading to a reduction in assay signal.

(192) Purified anti-IL-33 scFv antibody samples were tested for inhibition of biotinylated human IL-33 binding the lead antibody by adding 5 microliters of each dilution of antibody test sample to a 384 well low volume assay plate (Costar, 3673). Next, a solution containing 8 nM IL330180 IgG1 or 12 nM IL330101 IgG1 and 40 nM anti human Fc detection (Cisbio International, 61HFXLB) was prepared and 2.5 microliters added to the assay plate. This was followed by the addition of 2.5 microliters of a solution containing 4 nM (for IL330180 epitope competition assay) or 18 nM (for IL330101 epitope competition assay) biotinylated human IL-33 (Axxora, AG-40B-0038; biotinylated) and 4.65 nM streptavidin cryptate detection (Cisbio International, 610SAKLB). All dilutions were performed in assay buffer comprised of Dulbeccos PBS (Invitrogen, 14190185) containing 0.8 M potassium fluoride (BDH, 103444T) and 0.1% bovine serum albumin (BSA, Sigma A9576). Assay plates were incubated for 2 hour at room temperature followed by 16 hour at 4° C. before reading time resolved fluorescence at 620 nm and 665 nm emission wavelengths using an EnVision plate reader (Perkin Elmer). Data were analyzed using Equations 1 to 3 as described previously. The negative control (non-specific binding) is defined by replacing biotinylated IL-33/streptavidin cryptate combination with streptavidin cryptate detection only.

(193) Results for IL330101 and IL330180 epitope competition assay are shown in FIGS. 8A and 8B, respectively. FIG. 12A shows the competitive binding of IL330101 scFv, IL330107 scFv, IL330149 scFv, IL330065 scFv, Irrelevant scFv, and IL330180 scFv with mAb IL330101 for binding to biotinylated human IL-33. These results show that IL330101 scFv, IL330107 scFv, IL330149 scFv competitively inhibited IL330101 binding to biotinylated human IL-33.

(194) FIG. 12B shows the competitive binding of IL330180, IL330101, IL330149, IL330065, and Irrelevant scFv with mAb IL330180 for binding to biotinylated human IL-33 assessed in a biochemical HTRF®. These results show that IL330101, IL330149, and IL330065 do not competitively inhibit IL330180 binding to biotinylated human IL-33.

(195) Epitope binning using this method shows three panels of antibodies comprising IL330101, IL330107 and IL330149 in panel 1, IL330065 in panel 2 and IL330180 in panel 3, as detailed in Table 12.

(196) TABLE-US-00016 TABLE 12 Epitope Binning Panels Inhibition in IL330101 Inhibition in IL330180 Epitope Competition Assay Epitope Competition Assay IL330101 ✓ X IL330107 ✓ X IL330149 ✓ X IL330065 X X IL330180 X ✓

Example 3 Optimization of Anti-IL-33 Ab IL330101

(197) Affinity Maturation

(198) IL330101 was optimised using a targeted mutagenesis approach and affinity-based phage display selections. Large scFv-phage libraries derived from the lead clone were created by oligonucleotide-directed mutagenesis of the variable heavy (VH) complementarity determining regions 2 and 3 (CDR2 and CDR3) and light (VL) chain CDR3 using standard molecular biology techniques as described (Clackson, T. and Lowman, H. B. Phage Display—A Practical Approach, 2004. Oxford University Press). The libraries were subjected to affinity-based phage display selections in order to select variants with higher affinity for human and mouse IL-33. The selections were performed essentially as described previously (Thompson, J et al. J Mol Biol, 1996. 256: p. 77-88) using reagents as described in Examples 1 and 2. In brief, the scFv-phage particles were incubated with recombinant biotinylated human IL-33 in solution (Adipogen; biotinylated as described in Protein Modifications within Example 1). ScFv-phage bound to antigen were then captured on streptavidin-coated paramagnetic beads (Dynabeads® M-280) following the manufacturer's recommendations. The selected scFv-phage particles were then rescued as described previously (Osbourn, J. K., et al. Immunotechnology, 1996. 2(3): p. 181-96), and the selection process was repeated in the presence of alternating and decreasing concentrations of human or mouse biotinylated IL-33—typically from 500 nM to 500 pM over four rounds of selection.

(199) Inhibition of IL-33 Binding to mAb by Unpurified scFv

(200) A representative number of individual clones from the selection outputs were grown up in 96-well plates. ScFv were expressed in the bacterial periplasm (Kipriyanov, et al. J Immunol Methods 200(1-2): 69-77 (1997)) and screened for their inhibitory activity in a homogeneous FRET (fluorescence resonance energy transfer) HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) based IL-33:mAb-binding assay. In this assay, samples competed with IL330101 IgG for binding to biotinylated human IL-33 or mouse IL-33 FLAG® His. Such epitope competition assays are based on the principle that a test antibody sample, which recognizes a similar epitope to the anti-IL-33 IgG, will compete with the IgG for binding to biotinylated IL-33 resulting in a reduction in assay signal.

(201) Unpurified anti-IL-33 scFv samples were tested for inhibition of biotinylated human or mouse IL-33 FLAG® His binding to IL330101 by adding 5 microlitres of sample to a 384 well low volume assay plate (Costar, 3673). Next, a solution containing 12 nM IL330101 combined with 40 nM anti human Fc XL665 detection (Cisbio International, 61HFCXLB) was prepared for the human IL-33 assay and 2 nM IL330101 combined with 40 nM anti human Fc XL665 detection (Cisbio International, 61HFCXLB) was prepared for the mouse assay. 2.5 microlitres was added to the assay plates. This was followed by the addition of 2.5 microlitres of a solution containing 18 nM biotinylated human IL-33 (Adipogen, AG-40B-0038) combined with 4.6 nM streptavidin cryptate detection (Cisbio International, 610SAKLB) for the human assay or a solution containing 2 nM biotinylated mouse IL-33 FLAG® His combined with 4.6 nM streptavidin cryptate detection (Cisbio International, 610SAKLB) for the cynomolgus assay. All dilutions were performed in assay buffer containing 0.8 M potassium fluoride (VWR, 26820.236) and 0.1% bovine serum albumin (BSA, PAA, K05-013) in Dulbeccos PBS (Invitrogen, 14190185). Assay plates were incubated for 4 hour at room temperature followed by 16 hour at 4 degrees Celsius and time resolved fluorescence was read at 620 nm and 665 nm emission wavelengths using an EnVision plate reader (Perkin Elmer). Data were analysed by calculating the 665/620 nm ratio followed by the % Delta F values for each sample. The 665/620 nm ratio was used to correct for sample interference using Equation 1. The % Delta F for each sample was then calculated using Equation 2. The negative control (non-specific binding) was defined by replacing biotinylated IL-33 combined with streptavidin cryptate detection with streptavidin cryptate detection only. The % Delta F values were subsequently used to calculate % specific binding as described in Equation 3.

(202) As the epitope competition assay reached its limit of sensitivity, an assay using an intermediate optimised mAb IL330259 was used for testing unpurified scFv samples. Unpurified anti-IL-33 antibody samples were tested for inhibition of biotinylated human IL-33 or biotinylated mouse IL-33 FLAG® His binding DyLight labelled IL330259 by adding 5 microlitres of each sample to a 384 well low volume assay plate (Costar, 3673). Next, a solution containing 20 nM DyLight labelled IL330259 was prepared for the human IL-33 assay and 4 nM DyLight labelled IL330259 was prepared for the mouse assay and 2.5 microlitres added to the assay plates (IgG labelled using kit (Thermo Scientific, 53051) as per manufacturer's instructions). This was followed by the addition of 2.5 microlitres of a solution containing 20 nM biotinylated human IL-33 (Adipogen, AG-40B-0038) or 1.6 nM biotinylated mouse IL-33 FLAG® His combined with 6 nM streptavidin cryptate detection (Cisbio International, 610SAKLB). All dilutions were performed in assay buffer containing 0.8 M potassium fluoride (VWR, 26820.236) and 0.1% bovine serum albumin (BSA, PAA, K05-013) in Dulbeccos PBS (Invitrogen, 14190185). Assay plates were incubated for 4 hour at room temperature followed by 16 hour at 4 degrees Celsius and time resolved fluorescence was read at 620 nm and 665 nm emission wavelengths using an EnVision plate reader (Perkin Elmer). Data were analysed by calculating the 665/620 nm ratio followed by the % Delta F values for each sample. The 665/620 nm ratio was used to correct for sample interference using Equation 1. The % Delta F for each sample was then calculated using Equation 2. The negative control (non-specific binding) was defined by replacing biotinylated IL-33 combined with streptavidin cryptate detection with streptavidin cryptate detection only. The % Delta F values were subsequently used to calculate % specific binding as described in Equation 3.

(203) Inhibition of IL-33 Binding to mAb by Purified scFv

(204) Single chain Fv clones which showed a greater inhibitory effect on IL-33:mAb interaction as unpurified periplasmic extracts compared to IL330101 were subjected to DNA sequencing (Osbourn, et al. Immunotechnology 2(3):181-96 (1996); Vaughan, et al. Nat Biotechnol 14(3):309-14 (1996)). Unique scFv were expressed again in bacteria and purified by affinity chromatography (as described in WO01/66754). The potencies of these samples were determined by competing a dilution series of the purified preparation against IL330101 IgG for binding to biotinylated human IL-33, biotinylated mouse IL-33 FLAG® His or biotinylated cynomolgus IL-33 FLAG® His as described above but with the addition of the biotinylated cynomolgus IL-33 FLAG® His assay (biotinylated cynomolgus IL-33 FLAG® His was added at 12 nM concentration).

(205) FIG. 13: shows inhibition of the FRET signal, produced by biotinylated human IL-33 binding to IL330101 with increasing concentrations of IL-33 scFv antibodies IL330101, IL330259, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(206) Purified scFv preparations that were capable of inhibiting the IL-33:mAb interaction to a greater extent than IL330101 were selected for conversion to IgG format. The method of IgG expression and purification is described in Example 1.

(207) As the epitope competition assay reached its limit of sensitivity, an assay using an intermediate optimised mAb (IL330259) was used for testing purified scFv samples. Purified anti-IL-33 antibody samples were tested for inhibition of biotinylated human IL-33, biotinylated mouse IL-33 FLAG® His or biotinylated cynomolgus FLAG® His IL-33 binding IL330259 as described above with the addition of the biotinylated cynomolgus IL-33 FLAG® His assay (biotinylated cynomolgus IL-33 FLAG® His was added at 12 nM concentration).

(208) Based on sequence and epitope competition data, selected VH and VL outputs were recombined by standard molecular biology techniques to form libraries in which clones contained randomly paired VH and VL sequences (for example, VH CDR2/VL CDR3 and VH CDR3/VL CDR3 libraries). Alternatively, VH CDR3 and VL CDR3 sequences were randomly paired and recombined with specific VH CDR2 sequences selected from a pool of improved variants, to generate libraries in which all three CDRs were non-parental. Typically five rounds of affinity selection using decreasing and alternating concentrations of human and mouse biotinylated IL-33, from 10 nM to 10 pM, were performed on all recombination libraries to identify scFv sequences with improved kinetics. Alternatively, recombination libraries were selected using a fixed concentration of biotinylated IL-33 (for example, 1 nM, 100 pM or 300 pM) in the presence of 1000× unbiotinylated IL-33 for increasing amounts of time (for example 30 minutes, 1 hour, 2 hours or 4 hours) over four rounds of selection—a process known in the art as ‘off rate’ or ‘competition’ selection—to identify scFv's with improved kinetics.

(209) Samples were again screened in an HTRF® epitope competition assay for the ability to inhibit the binding of labelled-hulL-33 to IL330101 parent antibody or VH CDR2 optimized antibody IL330259 as described previously. ScFv's which showed a significantly improved inhibitory effect when compared to IL330101, were subjected to DNA sequencing, and unique variants were produced as purified scFv for further characterisation. Inhibitory scFv's were converted to whole immunoglobulin G1 (IgG1) antibody format as described in Example 1.

(210) Alternatively, individual unique VH CDR2, VH CDR3 and VL CDR3 sequences were specifically and rationally recombined and produced as IgG directly. In this example, IgGs were tested for improved kinetics without any additional affinity selection.

(211) Antibodies with improved kinetics were identified from all strategies and are exemplified by IL330259, IL330377, IL330388, IL330396, IL330398 and H338L293.

(212) The amino acid sequences of the V.sub.H and V.sub.L domains of IL330101 parent and the optimised anti-IL-33 antibodies were aligned to the known human germline sequences in the IMGT database (Lefranc, M. P. et al. Nucl. Acids Res. 2009. 37(Database issue): D1006-D1012), and the closest germline was identified by sequence similarity. For the V.sub.H domain of the IL330101 antibody lineage this was IGHV3-21/IGHJ2. For the V.sub.L domain it was IGLV3-25/IGLJ3. Without considering the Vernier residues (Foote, J., et al. J Mol Biol, 1992. 224: p. 487), which were left unchanged, there were no changes required in the frameworks of the V.sub.H domains and 4 changes in the V.sub.L frameworks (V3E, TSM, A45V and V104L; Kabat numbering). These positions were changed as indicated using standard site directed mutagenesis techniques with appropriate mutagenic primers. Antibodies that were germlined in this way appear in the sequence listings with the ‘fgl’ suffix. SEQ ID NOs corresponding to the various regions of antibodies IL330259, H338L293, IL330377, IL330388, IL330396, and IL330398 are shown in Table 13.

(213) TABLE-US-00017 TABLE 13 Anti-IL-33 Antibody Sequences VH SEQ VL SEQ VH CDRs VL CDRs IgG1 ID NO: ID NO: 1, 2, 3 1, 2, 3 IL330101_fgl 112 117 SEQ ID SEQ ID NO: 113 NO: 118 SEQ ID SEQ ID NO: 114 NO: 119 SEQ ID SEQ ID NO: 115 NO: 120 IL330259 152 157 SEQ ID SEQ ID NO: 153 NO: 158 SEQ ID SEQ ID NO: 154 NO: 159 SEQ ID SEQ ID NO: 155 NO: 160 IL330259_fgl 162 167 SEQ ID SEQ ID NO: 163 NO: 168 SEQ ID SEQ ID NO: 164 NO: 169 SEQ ID SEQ ID NO: 165 NO: 170 H338L293 172 177 SEQ ID SEQ ID NO: 173 NO: 178 SEQ ID SEQ ID NO: 174 NO: 179 SEQ ID SEQ ID NO: 175 NO: 180 H338L293_fgl 182 187 SEQ ID SEQ ID NO: 183 NO: 188 SEQ ID SEQ ID NO: 184 NO: 189 SEQ ID SEQ ID NO: 185 NO: 190 IL330377 192 197 SEQ ID SEQ ID NO: 193 NO: 198 SEQ ID SEQ ID NO: 194 NO: 199 SEQ ID SEQ ID NO: 195 NO: 200 IL330377_fgl 202 207 SEQ ID SEQ ID NO: 203 NO: 208 SEQ ID SEQ ID NO: 204 NO: 209 SEQ ID SEQ ID NO: 205 NO: 210 IL330388 212 217 SEQ ID SEQ ID NO: 213 NO: 218 SEQ ID SEQ ID NO: 214 NO: 219 SEQ ID SEQ ID NO: 215 NO: 220 IL330388_fgl 222 227 SEQ ID SEQ ID NO: 223 NO: 228 SEQ ID SEQ ID NO: 224 NO: 229 SEQ ID SEQ ID NO: 225 NO: 230 IL330396 232 237 SEQ ID SEQ ID NO: 233 NO: 238 SEQ ID SEQ ID NO: 234 NO: 239 SEQ ID SEQ ID NO: 235 NO: 240 IL330396_fgl 242 247 SEQ ID SEQ ID NO: 243 NO: 248 SEQ ID SEQ ID NO: 244 NO: 249 SEQ ID SEQ ID NO: 245 NO: 250 IL330398 252 257 SEQ ID SEQ ID NO: 253 NO: 258 SEQ ID SEQ ID NO: 254 NO: 259 SEQ ID SEQ ID NO: 255 NO: 260 IL330398_fgl 262 267 SEQ ID SEQ ID NO: 263 NO: 268 SEQ ID SEQ ID NO: 264 NO: 269 SEQ ID SEQ ID NO: 265 NO: 270
Inhibition of IL-33 Binding to mAb by Purified IgG

(214) The ability of anti-IL-33 antibodies to inhibit the binding of biotinylated human IL-33, biotinylated mouse IL-33 FLAG® His or cynomolgus IL-33 FLAG® His to the DyLight labelled IL330101 IgG was assessed in a biochemical HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) competition assay.

(215) Purified anti-IL-33 antibody samples were tested for inhibition of biotinylated human IL-33, biotinylated mouse IL-33 FLAG® His or biotinylated cynomolgus FLAG® His IL-33 binding DyLight labelled IL330101 by adding 5 microlitres of each concentration of sample to a 384 well low volume assay plate (Costar, 3673). Next, a solution containing 40 nM DyLight labelled IL330101 was prepared and 2.5 microlitres added to the assay plates (labelled using kit (Thermo Scientific, 53051) as per manufacturer's instructions). This was followed by the addition of 2.5 microlitres of a solution containing 40 nM biotinylated human IL-33 (Adipogen, AG-40B-0038), 2.5 nM biotinylated mouse IL-33 FLAG® His or 12 nM biotinylated cynomolgus IL-33 FLAG® His combined with 4.6 nM streptavidin cryptate detection (Cisbio International, 610SAKLB). All dilutions were performed in assay buffer containing 0.8 M potassium fluoride (VWR, 26820.236) and 0.1% bovine serum albumin (BSA, PAA, K05-013) in Dulbeccos PBS (Invitrogen, 14190185). Assay plates were incubated for 4 hour at room temperature followed by 16 hour at 4 degrees Celsius and time resolved fluorescence was read at 620 nm and 665 nm emission wavelengths using an EnVision plate reader (Perkin Elmer). Data were analysed by calculating the 665/620 nm ratio followed by the % Delta F values for each sample. The 665/620 nm ratio was used to correct for sample interference using Equation 1. The % Delta F for each sample was then calculated using Equation 2. The negative control (non-specific binding) was defined by replacing biotinylated IL-33 combined with streptavidin cryptate detection with streptavidin cryptate detection only. The % Delta F values were subsequently used to calculate % specific binding as described in Equation 3.

(216) FIG. 14A: shows inhibition of the FRET signal, produced by biotinylated human IL-33 binding to IL330101 with increasing concentrations of IL-33 IgG1 antibodies IL330101, IL330259, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(217) As the epitope competition assay reached its limit of sensitivity an assay using an intermediate optimised mAb IL330259 was used for testing purified IgG samples. This is as described for testing purified scFv.

(218) As the IL330259 epitope competition assay reached its limit of sensitivity, a third assay using an optimised mAb (H338L293) was used for testing purified IgG samples. Purified anti-IL-33 antibody samples were tested for inhibition of biotinylated human IL-33, biotinylated cynomolgus IL-33 FLAG® His or biotinylated mouse IL-33 FLAG® His binding DyLight labelled H338L293 by adding 5 microlitres of each sample to a 384 well low volume assay plate (Costar, 3673). Next, a solution containing 20 nM DyLight labelled H338L293 was prepared and 2.5 microlitres added to the assay plates (labelled using kit (Thermo Scientific, 53051) as per manufacturer's instructions). This was followed by the addition of 2.5 microlitres of a solution containing 4 nM biotinylated human IL-33 (Adipogen, AG-40B-0038), 0.8 nM biotinylated mouse IL-33 FLAG® His or 1.6 nM biotinylated cynomolgus IL-33 FLAG® His combined with 4.6 nM streptavidin cryptate detection (Cisbio International, 610SAKLB). All dilutions were performed in assay buffer containing 0.8 M potassium fluoride (VWR, 26820.236) and 0.1% bovine serum albumin (BSA, PAA, K05-013) in Dulbeccos PBS (Invitrogen, 14190185). Assay plates were incubated for 4 hour at room temperature followed by 16 hour at 4° C. and time resolved fluorescence was read at 620 nm and 665 nm emission wavelengths using an EnVision plate reader (Perkin Elmer). Data were analysed by calculating the 665/620 nm ratio followed by the % Delta F values for each sample. The 665/620 nm ratio was used to correct for sample interference using Equation 1. The % Delta F for each sample was then calculated using Equation 2. The negative control (non-specific binding) was defined by replacing biotinylated IL-33 combined with streptavidin cryptate detection with streptavidin cryptate detection only. The % Delta F values were subsequently used to calculate % specific binding as described in Equation 3.

(219) FIG. 14B: shows inhibition of the FRET signal, produced by biotinylated human IL-33 binding to H338L293 with increasing concentrations of IgG1 antibodies H338L293, IL330396 and IL330388 wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(220) Inhibition of Huvec IL-6 Production by IL-33 Antibodies

(221) Antibodies were assessed for ability to inhibit IL-33-stimulated IL-6 production from Huvecs as described in Example 2

(222) FIG. 15A shows the reduction of IL-6 production by increasing concentrations of antibodies (IL330101, IL330377, H338L293, IL330388, IL330396, IL330398) wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is % maximum response. Purified IgG preparations of antibodies inhibited the IL-6 activity by a maximum of ˜70%, compared to a commercial polyclonal antibody which inhibited 100%.

(223) Inhibition of Mast Cell IL-6 Production by IL-33 Antibodies

(224) Antibodies were assessed for ability to inhibit IL-33-stimulated IL-6 production from human cord blood derived mast cells as described in Example 2

(225) FIG. 15B shows the reduction of IL-6 production by increasing concentrations of antibodies (IL330101, H338L293, IL330388, IL330396) wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is % maximum response. Purified IgG preparations of antibodies inhibited the IL-6 activity by 100%, compared to the negative control antibody.

(226) IC50 results for HUVEC IL-6 production and mast cell IL-6 production are shown in Table 14.

(227) TABLE-US-00018 TABLE 14 Example potencies of optimised antibodies IgG1 Geomean (95% CI) IC50 (nM) Clone Huvec IL-6 production Mast cell IL-6 production IL330101 193 93 IL330377_FGL 6.6 1.3 H338L293 1.9 1.7 IL330388_FGL 2.1 1.2 IL330396_FGL 2.8 1.5 IL330398_FGL 1.5 0.9
Antibody Pharmacology

(228) IL-6 production from cord blood-derived mast cells was induced by increasing concentrations of IL-33 (Adipogen) using the method described in Example 2. This dose-response was carried out in the presence of increasing concentrations of H338L293 or IL330388 to produce a rightward shift of the IL-33 dose-response curve. EC50 values for IL-33 in the absence and presence of antibody were calculated using GraphPad PRISM software (La Jolla, Calif., USA), and the dose ratio (DR) was calculated. Data were plotted as log [Antibody] M (x-axis) versus log [DR-1] (y-axis). This clearly shows a non-competitive profile (curved plot), characteristic of an allosteric modulator. To investigate this, data from each experiment were normalised and combined into a single data set. An allosteric model was fitted using Prism Graphpad Software and the Kb and value of alpha determined. Values of alpha were similar for both leads examined, indicating a value for alpha of ˜0.02 (i.e. the maximum reduction in IL-33 affinity/potency when antibody is bound is ˜50-fold). Functional affinity (Kb) can be estimated for H338L293 (˜4.2 nM) and IL330388-1.7 nM).

(229) FIG. 16 shows a Schild analysis of IL330388 and H338L293 in a mast cell IL-6 production assay. Both antibodies display the profile of an allosteric modulator.

(230) Binding Affinity Calculation for IL-33 Antibodies Using BIAcore

(231) The binding affinity of purified IgG samples of exemplary binding members to human, cynomolgus or mouse IL-33 was determined by solution affinity using a BIAcore 2000 (GE healthcare). A biotinylated-IL33 surface was immobilised on a streptavin coated sensor chip (GE healthcare cat. No. BR-1000-32). Anti-IL33 antibodies were incubated with various concentrations of unlabelled IL33 and equilibrated at 25° C. for 48 hours. The amount of free antibody was determined by flowing the samples over the IL33 chip and measuring the response compared to a standard curve of antibody. Affinities were determined using the solution affinity fit in BIAevaluation software. The affinities for antibodies IL330101, H338L293, IL330388, IL330396 binding to human, cynomolgus or mouse IL-33 are shown in Table 15.

(232) TABLE-US-00019 TABLE 15 BIAcore solution affinity of exemplary binding members Antibody Antigen KD (nM) IL3300101 Human IL-33 FLAG ®-His 3.20E−06 H338L293 Human IL-33 FLAG ®-His 5.90E−10 H338L293 Cynomolgus IL-33 FLAG ®-His 5.45E−10 H338L293 Mouse IL-33 FLAG ®-His 4.52E−10 IL3300388 Human IL-33 FLAG ®-His 3.00E−10 IL3300388 Cynomolgus IL-33 FLAG ®-His 2.92E−10 IL3300388 Mouse IL-33 FLAG ®-His 8.82E−11 IL3300396 Human IL-33 FLAG ®-His 5.93E−10

Example 4 Redox Regulation of IL-33

(233) Reagents

(234) cDNA encoding the mature component of Human IL-33 (amino acids 112-270); accession number (Swiss-Prot) O95760 was synthesized by primer extension PCR and cloned into pJexpress404 (DNA 2.0). The coding sequence was modified to contain a 10×his, Avitag, and Factor-Xa protease cleavage site (MHHHHHHHHHHAAGLNDIFEAQKIEWHEAAIEGR) at the N-terminus of the protein. N-terminal tagged His10/Avitag IL33-01 (WT, SEQ ID 632) was generated by transforming E. coli BL21(DE3) cells. Transformed cells were cultured in autoinduction media (Overnight Express™ Autoinduction System 1, Merck Millipore, 71300-4) at 37° C. for 18 hours before cells were harvested by centrifugation and stored at −20° C. Cells were resuspended in BugBuster (Merck Millipore, 70921-5), containing complete protease inhibitor cocktail tablets (Roche, 11697498001), 2.5 u/ml Benzonase nuclease (merck Millipore, 70746-3) and 1 mg/ml recombinant lysozyme. Cell lysate was clarified by centrifugation at 75,000×g for 2 hours at 4° C. IL-33 proteins were purified from the supernatant by Nickel affinity chromatography in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM Imidazole, eluting in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 250 mM Imidazole. IL-33 was further purified by size exclusion chromatography using a Superdex 75 10/300 GL column in Phosphate Buffered Saline pH 7.4. Peak fractions were analysed by SDS PAGE. Fractions containing pure IL-33 were pooled and the concentration measured by Nanodrop A280 measurement. Final samples were analysed by SDS PAGE.

(235) To generate detagged IL-33 (BK349), N-terminal tagged His10/Avitag IL33-01 was incubated with 10 units of Factor Xa (GE healthcare 27-0849-01) per mg of protein in 2×DPBS buffer at RT for 1 hour. Untagged IL33 was purified using SEC chromatography in 2×DPBS on a S75 column (GE healthcare 28-9893-33) with a flow rate of 1 ml/min.

(236) Other reagents outlined in Table 16 were generated as described in Example 1.

(237) TABLE-US-00020 TABLE 16 IL-33 reagents Catalogue Number/ Reagent Supplier Designation SEQ ID Human IL-33 Flag ®His In house SEQ ID NO. 627 Mouse IL-33 Flag ®His In house SEQ ID NO. 628 Human His10/Avitag In house SEQ ID IL33-01 NO. 632 Human IL33-01 detagged In house BK349 Human IL-33 Axxora/Alexis ALX-522-098 Mouse IL-33 Peprotech 210-33 Mouse IL-33 R&D systems 3626/ML
Protein Modifications

(238) IgGs and modified receptor proteins used herein were biotinylated via free amines using EZ link Sulfo-NHS-LC-Biotin (Thermo/Pierce, 21335) as described in Example 1. IL-33 proteins used herein were biotinylated via free cysteines using EZ link Biotin-BMCC (Perbio/Pierce, product no. 21900).

(239) IL-33 Loses Activity Rapidly In Vitro

(240) IL-33 activity was measured in HUVEC signaling assays (30 minutes) and IL-6 production assay (18-24 hours), the methods of which are described in Examples 1 and 2 respectively.

(241) FIG. 17 shows the activity of human IL-33, cysteine-biotinylated IL-33 or cell culture media-pretreated IL-33, measured in HUVEC signaling assays (30 minutes) and IL-6 production assay (18-24 hours). FIG. 17A shows the comparison of human IL-33 activity (Adipogen) as measured by the NFkB or IL-6 assays wherein the x-axis is the concentration of human IL-33 in molar concentration and the y-axis is percent maximal response. IL-33 was significantly less potent in overnight compared with short (30 minute) assays. Human IL-33 Flag® His biotinylated via cysteine residues did not lose activity between short (30 minute) and longer (overnight) assays (FIG. 17B). To investigate this phenomenon, IL-33 (BK349) was pretreated for 18 hours in cell culture media (EBM-2 (Lonza, #CC-3156) with EGM-2 SingleQuot Kit Suppl. & Growth Factors (Lonza, #CC-4176)) and then compared with untreated IL-33 for ability to induce NFkB signaling. IL-33 that has been pre-treated with culture media displayed a significant loss of activity (FIG. 17C).

(242) SDS-PAGE Analysis of IL-33

(243) To investigate potential changes to the IL-33 protein, PBS/0.1% bovine serum albumin (BSA) or Iscoves Modified Dulbeccos Medium (IMDM)-treated human IL-33 (BK349 or human IL-33 Flag® His) and mouse IL-33 Flag® His were analysed by SDS PAGE electrophoresis under reducing or non-reducing conditions. Samples were made up in 1× NuPAGE gel loading buffer (Invitrogen) and denatured at 90° C. for 3 min. Reduced samples contained 2% beta-mercaptoethanol. Samples were run on NuPAGE Novex 12% Bis-Tris mini gels (Invitrogen) with MOPS running buffer (Invitrogen) according to manufacturer's instructions. Reduced and non-reduced samples were run on separate gels. 500 ng of IL-33 was loaded per lane. Gels were washed 3×5 min on shaking platform in ddH20 and then stained for 1 hour using EzBlue (Coomassie brilliant blue G-250 based gel staining reagent, Sigma G1041). Gels were destained in dH2O and scanned using an Epsom Scanner.

(244) FIG. 18 shows SDS-PAGE of human or mouse IL-33 under reducing or non-reducing conditions, before or after treatment with Iscoves Modified Dulbeccos Medium (IMDM). Differences in apparent molecular weight of IL-33 after treatment with IMDM were observed only under non-reducing conditions implying the presence of redox-related modifications for both human and mouse IL-33. FIG. 18A shows difference in apparent molecular weight of the human IL-33 (BK349) after treatment with IMDM that was observed only under non-reducing conditions. FIG. 18B shows human IL-33 Flag® His, non-biotinylated versus biotinylated under non-reducing conditions. Difference in apparent molecular weight was observed for IL-33 Flag® His, but not cysteine biotinylated IL-33 Flag® His after IMDM treatment. FIG. 18C shows difference in apparent molecular weight of the mouse IL-33 Flag® His after treatment with IMDM that was observed only under non-reducing conditions.

(245) Mass Spectrometry Analysis and Disulphide Mapping

(246) The media-treated form of human IL-33 was purified for further analysis. Human IL-33 (BK349) was was incubated with 60% IMDM media or in PBS at a final protein concentration of 300 ug/ml for 18 hours at 37° C. After 18 hours, media-treated IL33 was purified from media components using Size Exclusion Chromatography (SEC) on an S75 16:600 Superdex column (GE healthcare 28-9893-33) in 2×DPBS using an AKTAxpress FPLC system (GE healthcare). Peak fractions were analysed by SDS PAGE, and the non-aggregated, pure fractions were pooled and analysed by LC-MS.

(247) FIG. 19 shows the purification of IMDM-treated human IL-33 by SEC. The monomer fraction was collected for further analysis.

(248) LC-MS

(249) Reverse phase (RP) LC-MS analysis was performed using an Acquity UPLC coupled to a Synapt G1 quadrupole time of flight (QToF) mass spectrometer (Waters, Milford, US). 1 μg of purified protein diluted in 10 mM Tris HCl pH 8 at 1 mg/ml was injected onto a 50 mm×2.1 mm, 1.7 μm particle size BEH300 C4 analytical column held at 65° C. (Waters, Milford, US). Protein was eluted at a constant flow rate of 0.15 mL/min using a 5 minute binary gradient; solvent B was initially increased from 5 to 95% over 1 minute, reduced to 20% over 2 minutes and returned to 5% over a further 2 minutes. The column was cleaned prior to the subsequent injection by oscillating between high (95%) and low (5%) solvent B for 5 minutes. Solvent A (water) and B (acetonitrile) were supplemented with 0.01% (v/v) trifluoroacetic acid and 0.1% (v/v) formic acid. Spectra were acquired between 500-4500 m/z. Key instrument parameters included +ve ionisation mode, source voltage: 3.4 kV, sample cone voltage: 50 V, source temperature: 140° C., desolvation temperature: 400° C. BioPharmaLynx (Waters, Milford, US) was used to deconvolute the charge envelopes.

(250) FIG. 20 shows the intact mass of PBS versus IMDM treated IL-33 determined by LC-MS. IMDM-treated IL-33 displayed a 4 Da loss compared with PBS-treated IL-33 compatible with the formation of two disulphide bonds.

(251) Disulphide Bond Mapping

(252) For each sample, 50 μg of protein was prepared at 3 mg/ml in 100 mM sodium phosphate, 1 mM N-ethylmaleimide, pH 7.0 buffer and incubated for 20 minutes at room temperature. Dried samples were resuspended in 7 M Guanidine HCl, 100 mM NaCl, 10 mM sodium phosphate and incubated at 37° C. for 30 minutes. Denatured protein was diluted to 0.3 mg/ml and digested with Glu-C at an E:S ratio of 1:50 in 2 M Guanadine, 100 mM sodium phosphate, 0.1 mM EDTA, pH7.0 at 37° C. After 2 hours a second, equal aliquot of Lys-C was added. After a further 2 hours the digest was split; for the reduced analysis the digest was incubated with 50 mM Dithiothreitol for 15 min at room temperature. Reduced and non-reduced samples were analysed by RP LC-MS using an Acquity UPLC coupled to a Synapt G2 QToF mass spectrometer (Waters, Milford, US). For each sample, 5 ug of Lys-C digest was injected onto a 150 mm×2.1 mm, 1.7 μm particle size BEH300 C18 analytical column held at 55° C. (Waters, Milford, US). Peptides were eluted at a constant flow rate of 0.2 mL/min using a 75 minute binary gradient; solvent B was increased from 0% to 35%. The column was cleaned prior to the subsequent injection by oscillating between high (95%) and low (5%) solvent B for 5 minutes. Solvent A (water) and B (acetonitrile) were supplemented with 0.02% (v/v) trifluoroacetic acid. Spectra were acquired between 50-2000 m/z using a data independent mode of acquisition. Low and high energy spectra were processed using BioPharmaLynx (Waters, Milford, US).

(253) FIG. 21 shows the disulphide mapping of IMDM-treated human IL-33. FIG. 21A shows combined, deconvoluted mass spectra from non-reduced and reduced Lys-C peptide mapping analysis of DSB IL-33. FIG. 21B shows isolated spectra for cysteine containing peptides. Peptides unique to reduced and non-reduced samples are highlighted in green and blue, respectively. Data were consistent with the formation of two disulphide bridges. One species identified had bridges between cysteines C208-C249 and C227-C232, respectively. However, the predominant peak was not resolved and other species may exist. FIG. 21C shows sequences of disulphide bonded peptides identified by non-reduced and reduced Lys-C peptide mapping analysis of disulphide bonded IL-33. Disulphide linkages are represented by two hyphens (--). Lys-C miscleavages are represented by square brackets.

(254) NMR Analysis of the Disulphide Bonded IL-33

(255) Based on the reported structure of IL-33 (Lingel, A. et al. Structure 17, 1398-1410 (2009); Liu, X. et al. Proc. Natl. Acad. Sci. U.S.A 110, 14918-14923 (2013)), cysteine residues are not in sufficiently close proximity for disulphide bonding to occur without significant conformational change. To investigate this NMR heteronuclear multiple quantum coherence (HMQC) analysis was performed.

(256) Production of .sup.15N-IL-33 Proteins

(257) DNA encoding wild type IL-33 with an N-terminal 6His tag and TEV protease cleavage site (SEQ ID. 633) was used to transform E. coli BL21 Gold cells. Transformed cells were cultured at 37° C. in M9 minimal media supplemented with 5 g/L of .sup.15N-IsoGro™ powder until they reached an OD600 nm of 0.6 to 0.8, when protein expression was induced by addition of 100 mM IPTG. Cultures were continued at 18° C. for a further 20 hours before cells were harvested by centrifugation and stored at −80° C. Cells were resuspended in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM Imidazole, 5 mM BetaMercaptoethanol containing Complete protease inhibitor tablets (Roche, 11697498001), 2.5 U/ml Benzonase nuclease (merck Millipore, 70746-3) and 1 mg/ml recombinant lysozyme. Resuspended cells were lysed using a Constant Systems cell disruptor at 25 kpsi and clarified by centrifugation at 75,000×g for 2 hours at 4° C. IL-33 was purified from the supernatant by Nickel affinity chromatography in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM Imidazole, 5 mM BetaMercaptoethanol, eluting in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 250 mM Imidazole, 5 mM BetaMercaptoethanol. Eluted protein was incubated with TEV protease and dialysed into 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM Imidazole, 5 mM BetaMercaptoethanol at 4° C. De-tagged protein was separated from uncleaved IL-33 by Nickel affinity chromatography in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM Imidazole, 5 mM BetaMercaptoethanol. IL-33 was further purified by size exclusion chromatography using a HiLoad 16/60 Superdex 75 column (GE healthcare) in 20 mM Sodium phosphate pH 6.5, 100 mM NaCl, 5 mM BetaMercaptoethanol, using an AKTAxpress FPLC system (GE healthcare). Peak fractions were analysed by SDS PAGE.

(258) Fractions containing pure IL-33 were pooled and the concentration measured by Nanodrop A280 measurement. Protein was concentrated using an Amicon 10,000 molecular weight cut-off spin concentrator to a final concentration of 9.5 mg/ml for NMR analysis.

(259) Purified .sup.15N labelled protein in PBS pH7.4 was incubated with 60% IMDM media at a final protein concentration of 0.28 mg/ml for 18 hours at 37° C. After 18 hours, the protein was concentrated using an Amicon 10,000 molecular weight cut-off spin concentrator to a concentration of 0.8 mg/ml. The protein was then purified by size exclusion chromatography using a HiLoad 16/60 Superdex 75 in PBS pH 7.4. Peak fractions were analysed by SDS PAGE, and the non-aggregated, pure fractions were pooled. Finally protein was concentrated using an Amicon 10,000 molecular weight cut-off spin concentrator to a concentration of 1.8 mg/ml (100 μM) for NMR analysis.

(260) NMR Analysis

(261) NMR spectra were recorded at 298 K on a Bruker Avance 600 MHz spectrometer running Topspin 2.3 equipped with a 5 mm TCI Cryoprobe with Z-axis gradients. The .sup.15N-labelled IL33 WT sample was prepared as described with the addition of 5% deuterium oxide to allow sample locking. The exemplified .sup.1H-.sup.15N correlation spectra were acquired employing the sofast HMQC pulse sequence (Schanda, P; Brutscher, B; Very fast two-dimensional NMR spectroscopy for real-time investigation of dynamic events in proteins on the time scale of seconds, J. Am. Chem. Soc. (2005) 127, 8014-5) with (F2×F1) 1024×64 complex points (in states-TPPI mode), 9615×1460 Hz sweep width, 53.4 ms×43.8 ms acquisition times.

(262) FIG. 22A shows SDS PAGE analysis of IMDM treated WT IL33. SDS PAGE showing reduced and non-reduced IMDM treated WT IL33 before and after concentration for NMR.

(263) FIG. 22B shows NMR analysis of WT IL33. Overlay of the .sup.1H-.sup.15N HMQC spectra for 0.1 mM .sup.15N-labeled IL33 WT before and after IMDM media treatment plotted in black and red, respectively. Comparison of the two spectra indicate an entirely different, and less ordered, structure after IMDM treatment.

(264) Circular Dichroism (CD) Spectroscopy.

(265) To confirm conformational change and investigate further, Circular Dichroism (CD) Spectroscopy analysis was performed.

(266) Far-UV and near-UV CD analysis were performed on a Jasco-815 instrument (Easton, Md.). For Far-UV CD the spectra were recorded over wavelength range 180-260 nm in a 1 mm pathlength cuvette at sample concentration of 0.14 mg/mL and 0.12 mg/mL for respectively redIL-33 and DSB IL-33 at 20° C. in buffer solution 10 mM Phosphate pH=6.9. For near-UV CD the spectra were recorded over wavelength range 260-350 nm in a 10 mm pathlength cuvette at sample concentration of 1.38 mg/mL and 0.89 mg/mL for respectively redIL-33 and DSB IL-33 at 20° C. in buffer solution DPBS. CD spectra of the buffer solution were recorded and subtracted from all sample spectra to correct for instrument, cuvette and baseline effects. CD Pro software was employed for the deconvolution of spectra into secondary structure elements.

(267) FIG. 22C: Near-UV circular dichroism (CD) Spectroscopy. Spectra were recorded over wavelength range 260-350 nm. The final spectra were the average of 4 scans. Aromatic amino-acids and disulphide absorption bands are adapted from Kelly (Kelly S. M. et al. How to study proteins by Circular Dichroism. Biochimica et Biophysica Acta, 1751, 119-139 (2005)) The difference observed in ellipticity around Trp absorption is consistent with change in environment of the sole Tryptophan (W193) demonstrating changes to tertiary structure in this region between reduced and DSB IL-33. The difference in intensity around 260 nm is consistent with the introduction of additional chromophores from disulphide bond formation.

(268) FIG. 22D: Key features of IL-33. Trp193, cysteines, and ST2 binding site (Liu, X. et al. Structural insights into the interaction of IL-33 with its receptors. Proc. Natl. Acad. Sci. U.S.A 110, 14918-14923 (2013)) are indicated within the solved IL-33 structure (Lingel 2009)

(269) FIG. 22E: Far-UV circular dichroism (CD) Spectroscopy. Spectra were recorded over wavelength range 190-260 nm. The final spectra were the average of 8 scans. Far-UV spectra are consistent with predominantly β-sheet secondary structures as seen previously with this family of proteins (Chang B. S. et al, Formation of an active dimer during storage of interleukin-1 receptor antagonist in aqueous solution. Biophysical Journal. 71, 3399-3406 (1996); Craig S. et al. Conformation, Stability and folding of Interleukin1β. Biochemistry. 26, 3570-3576 (1987); Hailey K. L. et al. Pro-interleukin (IL)-1β shares a core region of stability as compared with mature IL-1β while maintaining a distinctly different configurational landscape. J. Biol. Chem. 284. 26137-26148 (2009); Hazudat D. et al. Purification and characterisation of Human Recombinant Precursor Interleukin 1β. J. Biol. Chem. 264, 1689-1693 (1989); Meyers C. A. et al, Purification and characterization of Human recombinant interleukin-1β. J. Biol. Chem. 262, 11176-11181 (1987)). Spectra are significantly different demonstrating a change in secondary structure in DSB IL-33, relative to reduced IL-33.

(270) CD spectra indicated a significant conformatonal change between the IL-33 forms. redIL-33 spectra were consistent with published data. DSB-IL-33 spectra were consistent with a structured protein that was different to the reduced form. To map the areas that may be most altered between reduced and DSB-IL-33, we performed hydrogen/deuterium-exchange mass spectrometry.

(271) Hydrogen/Deuterium-Exchange Mass Spectrometry (HDX-MS).

(272) Proteins were diluted to 3.5 uM in phosphate buffered saline, pH 7.4. This stock was used to initiate labeling experiments by diluting 10-fold with deuterated (10 mM sodium phosphate, pD 6.6) aqueous solvent. Initial mapping experiments were done to assign species from the mass spectra to peptic peptide sequences from IL33. This was done largely as described.sup.21. Briefly, protonated diluted protein was mixed 1:1 with a quench solution (100 mM potassium phosphate, pH 2.55, 0.1 M TCEP, 1° C.), such that the final mixture pH was 2.55. The quenched protein was injected into a Waters HDX Manager with an immobilized pepsin column (2.0×30 mm; Poroszyme, Life Technologies), C18 trapping column (VanGuard ACQUITY BEH 2.1×5 mm; Waters) and analytical C18 column (1.0×100 mm ACQUITY BEH; Waters). Mobile phases were 0.1% formic acid in H2O (A) and 0.1% formic acid in ACN (B), such that their pH was 2.55. Protein was applied to the pepsin and trapping columns in 100 μL/min buffer A and eluted from the analytical column in a linear gradient of 3-40% B at 40 μL/min. Peptide sequences were assigned from MSE fragment data with Protein Lynx Global Server (Waters) 3.0.2 and DynamX 3.0 (Waters). Labeling data was acquired as for sequencing, except the mass spectrometer was set to MS scans only. Peptide-level data were analyzed in DynamX and MatLab (Mathworks).

(273) FIG. 23 shows hydrogen-exchange mass spectrometry (HX-MS) analysis of reduced and DSB IL-33. FIG. 23A Comparison of fractional hydrogen exchange (for deuterium) in reduced IL-33 (left panel) and DSB IL-33 (right panel). Data are mapped onto the published IL-33 structure.sup.(lingel 2009) in both cases for comparison purposes. Gaps in sequence coverage where no data HX-MS data could be obtained are highlighted in slate blue. Side chains of cysteine residues are displayed as sticks. FIG. 23B shows a structural model of differential HX-MS data overlaid with the ST2 binding site (red and magenta)..sup.(Liu 2013). Dark blue indicates regions of increased hydrogen exchange in DSB IL-33 relative to reduced IL-33. The ST2 binding site 1 is in the area of greatest difference in H/D exchange and has likely been altered in structure.

(274) Binding of Reduced Vs DSB IL-33 to ST2 (BIAcore)

(275) The disulphide bonded IL-33 is likely a very different structure to the reduced, ST2-binding IL-33 form (FIG. 22, 23), and conversion to the disulphide bonded form was associated with loss of functional activity (FIG. 17C). To investigate this, the disulphide bonded form of IL-33 was tested for ability to bind ST2 by BIAcore analysis. Direct binding of IL-33 to the extracellular domain of ST2 was determined by Surface Plasmon Resonance using a BIAcore 2000 (GE healthcare). ST2 was immobilised via the Fc-tag using an anti-human Fc capture (GE healthcare BR-1003-39) on a CM5 sensor chip (GE healthcare BR-1003-99) to give a stable surface of approximately 150 RU. IL-33 was flowed over the surface at 30 ul/min for three minutes to determine association rates. Dissociation was measured by flowing buffer at 30 ul/min for 15 minutes. Sensorgrams were interpreted using BIAevaluation software and kinetics were determined using double reference subtracted sensorgrams using a 1:1 (Langmuir) binding model.

(276) FIG. 24A shows redIL-33 binding to ST2. Sensorgrams from 7.8 nM to 0.24 nM are shown, giving a KD of 0.2 nM.

(277) FIG. 24B shows disulphide bonded IL-33 (IL33-DSB) binding to ST2. Sensorgrams from 500 nM to 0.24 nM are shown with no obvious binding observed.

(278) Loss of ST2 binding and activity led us to hypothesise that oxidation could be a mechanism to terminate IL-33 activity and limit duration of ST2-dependent immunological responses in vivo.

(279) Detection of IL-33 Forms

(280) To ascertain that the disulphide-bonded form of IL-33 indeed exists in vivo, we used three different commercial IL-33 detection assays (2 human IL-33 and one mouse IL-33). Human and mouse IL-33 Duoset ELISAs (RnD Systems) were converted to MSD format (Meso Scale Discovery, Rockville, Md.). Coating concentrations of capture antibodies were as follows: anti-mouse IL-33 pAb 37.5 ug/ml; anti-human IL-33 pAb 18 ug/mL. Capture antibody was diluted in PBS with 0.03% Triton X-100 and 5 μl was spot coated onto standard bind plates (Meso Scale Discovery, Rockville, Md.) into the centre of each well and left to dry overnight at room temperature. Plates were washed ×3 in PBS-Tween and blocked with 25 μl Assay Diluent by sealing plates and incubating for 30 minutes at room temperature with shaking (450 rpm). 25 μl of samples or calibrator diluted in Assay Diluent were transferred to the blocked assay plates, which were incubated for 2 hours at room temperature with shaking (450 rpm). Plates were washed ×3 in PBS-Tween and 25 μl of detection reagent (detection antibody plus streptavidin SulfoTag both diluted to 1 μg/ml in antibody diluent). Plates were sealed and incubated for 1 hour at RT with shaking. Plates were washed ×3 in PBS-Tween. 150 μl of Read Buffer T diluted to 2× in distilled water was added. Plates were read within 15 minutes (Meso Scale Discovery, Rockville, Md.).

(281) The Millipore human IL-33 assay (Cat #HTH17MAG-14K lot 2159117) was performed according to manufacturer's instructions. Briefly, IL-33 standards and samples were diluted in assay buffer and incubated with beads for 1 hour at room temperature with shaking (500 rpm), protected from light. Well contents were removed and washed ×2 with 200 uL of wash buffer. 25 uL of detection antibody was added per well and plates incubated for 1 hour at room temperature. 25 uL streptavidin-PE was added (without washing) and plates incubated for a further 30 minutes, shaking (850 rpm) and protected from light. Well contents were removed and washed ×2 with 200 uL of wash buffer. Samples were resuspended in 125 uL assay buffer, covered and shaken at 850 rpm for 30 seconds. Samples were analysed on Bio-Plex 200 (BioRad). Plate was read at low RP1, counting 50 beads per region (region 44) with doublet detection gates set to 5000 (low) and 25000 (high).

(282) FIG. 25 shows analysis of three commercial IL-33 ELISA assays for detection of reduced and disulphide bonded IL-33 forms. The effect of ST2 on interference with the assay signal for both reduced and disulphide bonded forms is also shown. FIGS. 25A and B shows that the two commercial human IL-33 assays predominantly detect the disulphide-bonded form of IL-33 (IL33-DSB), suggesting that this is the main species that others have measured to date in human ex vivo samples. The ‘reduced’ but not oxidized/disulphide bonded IL-33 assay signal can be eliminated by addition of sST2. FIG. 25C shows the mouse IL-33 assay which detects both reduced and oxidized forms of mouse IL-33. The ‘reduced’ but not oxidized IL-33 assay signal can be eliminated by addition of sST2.

(283) As we were unable to identify a commercial assay specific for the reduced, ST2-active form of human IL-33, we developed our own novel assays. IL330425 mAb (SEQ ID NOs. 62 and 67) or IL330004 mAb (SEQ ID NOs. 12 and 17) were used as capture antibodies. Captured IL-33 was detected with biotinylated sST2.Fc (R&D systems) or biotinylated IL330425 (SEQ ID NOs. 62 and 67) respectively. Capture antibody was diluted to 150 ug/mL in PBS with 0.03% Triton X-100 and 5 μl was spot coated onto standard bind plates (Meso Scale Discovery, Rockville, Md.) into the centre of each well and left to dry overnight at room temperature. Plates were washed ×3 in PBS-Tween and blocked with 25 μl Assay Diluent by sealing plates and incubating for 30 minutes at room temperature with shaking (450 rpm). 25 μl of samples or calibrator diluted in Assay Diluent were transferred to the blocked assay plates, which were incubated for 2 hours at room temperature with shaking (450 rpm). Plates were washed 3× in PBS-Tween and 25 μl of detection reagent (detection antibody plus streptavidin SulfoTag both diluted to 1 μg/ml in antibody diluent). Plates were sealed and incubated for 1 hour at RT with shaking. Plates were washed ×3 in PBS-Tween. 150 μl of Read Buffer T diluted to 2× in distilled water was added. Plates were read within 15 minutes (Meso Scale Discovery, Rockville, Md.).

(284) FIG. 26A, B shows ELISA assays that are specific for detection of reduced IL-33. No detection of the disulphide bonded form is observed.

(285) Timecourse of Conversion to Disulphide Bonded Form of IL-33

(286) Assays detecting different IL-33 forms as described above were used to monitor a time course of conversion from redIL-33 to its disulphide bonded form. 10 ug/mL of detagged redIL-33 was incubated in 100% human serum, PBS/1% BSA or IMDM/1% BSA at 37° C. At timepoints t=0, 15 minutes, 30 minutes, 1 hour, 2 hours, 4 hours, 6 hours, 8 hours and 24 hours a 10 ul aliquot was removed and added to 90 ul PBS/1% BSA (1 in 10 dilution to 1 ug/ml), this was divided into 3×30 ul aliquots and snap frozen on dry ice before storage at −80° C. A t=0 sample was also prepared fresh immediately prior to the ELISA analysis as a control for the freeze/thaw cycle. Samples were analysed using the human MSD (R&D Systems) and the IL33004/1L330425-biotin assays described above to measure disulphide bonded and reduced IL-33 respectively. Together, these assays allowed monitoring of the conversion from the reduced to the disulphide-bonded form of IL-33.

(287) To confirm the result from the ELISAs, IL-33 in the timecourse samples was analysed by Western blot. Samples were subjected to SDS-PAGE under reducing or non-reducing conditions. Samples were made up in 1× NuPAGE gel loading buffer (Invitrogen) and denatured at 90° C. for 3 minutes. Reduced samples contained 2% beta-mercaptoethanol. Samples were run on NuPAGE Novex 12% Bis-Tris mini gels (Invitrogen) with MOPS running buffer (Invitrogen) according to manufacturer's instructions. Reduced and non-reduced samples were run on separate gels. 100 pg of IL-33 was loaded per lane. Proteins were transferred to Nitrocellulose membranes (Invitrogen cat. no. IB3010-02) and detected by Western blotting with anti-IL-33 pAb (R&D systems).

(288) FIG. 27 shows a time course of human IL-33 incubated in IMDM or human serum. FIG. 27A: IL-33 ELISAs (IL33004/1L330425-biotin and human R&D systems MSD assays) were used to detect reduced and disulphide bonded IL-33 respectively. FIG. 27B: Western blot analysis was used to detect reduced and disulphide bonded IL-33 forms. The conversion to the disulphide bonded form of IL-33 occurred rapidly with a 50% conversion in 1-2 hours. The disappearance of reduced IL-33 correlated well with the appearance of oxidised IL-33 in both ELISA and Western blot analysis.

(289) Generation of a Humanized IL-33 Transgenic Mouse

(290) To study the behaviour and lifecycle of endogenous IL-33, we used a transgenic mouse with the gene for mouse IL-33 replaced with the human IL-33 gene. A humanized IL-33 transgenic mouse were generated as described below. Briefly, mouse genomic fragments (obtained from the C57BL/6J RPCIB-731 BAC library), a human genomic fragment (obtained from the human RPCIB-753 BAC library) and selected features (such as recombination sites and selection markers) were combined to make the targeting vector (data not shown).

(291) The targeting vector was linearized with B stBI and electroporated into TaconicArtemis Balb/cJ ES cell line (Balb/c.2) and ES cell clones were selected with Puromycin (positive selection) and Gancyclovir (negative selection). The resultant puromycin-resistant ES cell clones were then screened by a combination of PCR and Southern analyses to identify correctly targeted ES clones. These were expanded and frozen in liquid nitrogen.

(292) After administration of hormones, superovulated C57BL/6 females were mated with C57BL/6 males. Blastocysts were isolated from the uterus at dpc 3.5. For microinjection, blastocysts were placed in a drop of DMEM with 15% FCS under mineral oil. A flat tip, piezo actuated microinjection-pipette with an internal diameter of 12-15 micrometer was used to inject 10-15 targeted BALB/c ES cells into each blastocyst. After recovery, 8 injected blastocysts were transferred to each uterine horn of 2.5 days post coitum, pseudopregnant NMRI females. Chimerism was measured in chimeras (GO) by coat colour contribution of ES cells to the C57BL/6 host (white/black). Highly chimeric mice were bred to strain BALB/cJBomTac females mutant for the presence of a Flp recombinase gene (Flp-Deleter strain). Germline transmission by coat colour was identified by the presence of white, strain BALB/c, offspring (G1). Actual germline transmission was confirmed by PCR genotyping using primers specific for the targeted allele (data not shown).

(293) Existence of Redox Forms of IL-33 In Vivo.

(294) Models of Alternaria alternata induced airway inflammation in mice have been previously described (Kouzaki et al. J. Immunol. 2011, 186: 4375-4387; Bartemes et al J Immunol, 2012, 188: 1503-1513). Male or female wildtype or humanized IL-33 mice (6-10 weeks) were anaesthetized briefly with isofluorane and administered either 25 μg of Alternaria alternata (ALT) extract (Greer, Lenoir, N.C.) or vehicle intranasally in a total volume of 50 μl. At multiple timepoints after ALT challenge, mice were terminally anaesthetized with pentobarbital sodium prior to bronchoalveolar lavage (BAL). Bronchoalveolar lavage fluid (BALF) was collected by lavage (0.3 ml, 0.3 ml & 0.4 ml) via tracheal cannula. BALF was centrifuged and supernatant was analysed for presence of redox forms of IL-33 using assays described above. All work was carried out to UK Home Office ethical and husbandry standards under the authority of an appropriate project license.

(295) FIG. 28 shows analysis of BALF from humanized IL-33 mice collected at varying timepoints following ALT intranasal challenge, using a combination of multiple ELISA assays. (A) Millipore, (B) R&D systems and (C) IL330425/sST2-biotin assays were used to measure IL-33 in the presence or absence of sST2 (left hand graphs). Signals in the presence of sST2 (signal from the reduced IL-33 fraction eliminated) were compared with a disulphide bonded IL-33 standard to quantify the levels of disulphide bonded IL-33. The reduced IL-33 signal was calculated as the difference in signal between IL-33 measurements in the presence and absence of ST2, quantified against a reduced IL-33 standard. Estimations for reduced IL-33 are shown on the right hand graphs. All assays show that the released IL-33 was predominantly in its reduced form with a maximum between 5 and 30 min, followed by a rapid decline becoming undetectable by 120 minutes. Conversely, IL-33-DSB gradually increased from time 0, peaking at 30-120 minutes and disappearing by 24 hours. These data are consistent with a model where IL-33 is released in reduced form and then rapidly oxidised in vivo to IL33-DSB.

(296) FIG. 29 shows analysis of BALF from wild type BALB/c mice collected at varying timepoints following ALT intranasal challenge. FIG. 29A Mouse IL-33 ELISA (R&D systems) was used to measure IL-33 in the presence or absence of sST2 (media-treated mouse IL-33 used as standard curve). FIG. 29B Signals in the presence of sST2 (signal from the reduced IL-33 fraction eliminated) were compared with a media-treated mouse IL-33 standard to quantify the levels of oxidised IL-33. The reduced IL-33 signal was calculated as the difference in signal between IL-33 measurements in the presence and absence of ST2, quantified against a reduced mouse IL-33 standard. Data show that the released IL-33 was predominantly in its reduced form peaking at 15 minutes, followed by a rapid decline becoming undetectable by 120 minutes. Conversely, IL-33-DSB gradually increased from time 0, peaking at 45-60 minutes and disappearing by 24 hours. These data are consistent with a model where IL-33 is released in reduced form and then rapidly oxidised in vivo to IL33-DSB.

Example 5 Characterisation of Anti-IL-33 Antibodies

(297) H338L293 Causes a Conformational Change

(298) Monoclonal antibody H338L293 (SEQ ID NOs 182 and 187), the generation of which is described in examples 2 and 3, is an allosteric modulator of IL-33. IL-33 can significantly change conformation into a disulfide bonded form (as described in Example 4). The following experiments demonstrate that H338L293 mAb appears to destabilize the IL-33 molecule, promoting its unfolding and accelerating conversion to the disulfide bonded form. Reagents and protein modifications used herein were as described in previous examples.

(299) Sypro Orange Assay

(300) Sypro orange binds nonspecifically to hydrophobic surfaces, and water strongly quenches the fluorescence of Sypro Orange. When a protein unfolds, the exposed hydrophobic surfaces bind the dye, resulting in an increase in fluorescence. 5 uM of antibody was incubated with 20 uM redIL-33 at 25° C. in the presence of 8×SYPRO orange dye diluted from a 5000× stock (Life technologies S-6650) in 1×DPBS. Fluorescence (excitation 490 nm and emission 575 nm) was measured every minute using a Chromo4 Real Time detector (Bio-rad). We observed that incubating redIL-33 with H338L293 antibody, and not redIL-33 or antibody alone, there was an increase in the fluorescence signal indicative of protein unfolding.

(301) FIG. 30A shows relative fluorescence units at 100 minutes following incubation of 5 uM antibody with 20 uM redIL33 at 25° C. in the presence of 8×SYPRO orange dye. In the presence of redIL-33 H338L293, but not IL330004 or the control mAb, increased the fluorescent signal indicative of protein unfolding.

(302) FIG. 30B shows relative fluorescence units over time following incubation of varying concentrations of H338L293 with 20 uM redIL33 at 25° C. in the presence of 8×SYPRO orange dye. Fluorescent signal increased with increasing antibody concentration.

(303) SDS-PAGE Electrophoresis

(304) To determine if H338L293 could influence disulfide bonding of IL-33, IL-33 was monitored in the presence of H338L293 by comparing reduced and non-reduced SDS-PAGE analysis. 100 ug/ml recombinant human IL-33.sup.112-270 (BK349) was incubated in PBS/0.1% BSA containing either 1.5 mg/ml H338L293 mAb, NIP228 mAb or without addition of a mAb. Samples were incubated for 20 h at 37 C in a standard tissue culture incubator. Samples containing 1 ug of IL-33 were analysed by SDS-PAGE on Novex 12% Bis-Tris mini NuPAGE gels (Invitrogen) under reducing and non-reducing conditions. Following SDS-PAGE gels were washed 3×5 min in ddH20, incubated in EzBlue (Sigma G1041, a coomassie brilliant blue G-250 based protein stain) for 1 h, and destained in ddH20 until background of gel was clear. All gel staining steps performed at room temperature on a rocking platform. Gels were visualised using Epsom digital scanner.

(305) FIG. 30C shows SDS-PAGE analysis of IL-33. Preincubation of IL-33 with H338L293, but not control mAb or no mAb, increased the presence of the faster migrating, disulfide bonded form of IL-33 under non-reducing conditions.

(306) Inhibition of NFkB Signalling in Huvec by IgG

(307) NFkB signaling in human umbilical vein endothelial cells in response to IL-33 was assessed by nuclear translocation of the p65/RelA NFkB subunit, detected by immunofluorecence staining as described in Example 1. Cells were stimulated with varying IL-33 concentrations in the presence of multiple concentrations of test antibody H338L293 for either 30 minutes or 6 hours.

(308) FIG. 31 shows the effect of H338L293 on IL-33 stimulated NFkB translocation on HUVECs. These results show that, as seen previously, H338L293 did not inhibit nuclear translocation of p65/RelA NFkB in IL-33-stimulated Huvecs 30 minutes following stimulation. However, after 6 hours, inhibition is seen. The results are consistent with failure of H338L293 to inhibit IL-33 binding to ST2 directly but ability to convert IL-33 to a non-ST2 binding form within hours.

(309) Inhibition of IL-33 Binding to ST2 by Purified IgG

(310) The ability of H338L293 to inhibit the binding of FLAG® His tagged IL-33 to the ST2 receptor was assessed in a biochemical HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) competition assay as described in Example 1. Two conditions were tested. Firstly the antibody and IL-33 were added simultaneously to the assay exactly as in Example 1. As previously, purified IgG preparations were not able to inhibit the IL-33:ST2 interaction at concentrations tested. Secondly, H338L293 was preincubated with the IL-33FLAG® His for 18 hours prior to addition to the assay. In this case concentration-dependent inhibition of IL-33:ST2 binding was observed. Taken together these data are consistent with H338L293 converting IL-33 to a non-ST2 binding form over time.

(311) FIG. 32 shows the inhibition of the FRET signal, produced by human IL-33 binding to human ST2 with increasing concentrations of H338L293, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding. These results show that H338L293 inhibits IL-33 binding to ST2 only after prolonged pre-incubation with the ligand.

(312) Epitope Mapping

(313) Epitope mapping of H338L293 IgG was attempted in order to clarify the mode of IL-33 binding to this IgG. Size Exclusion Chromatography (SEC) experiments were performed in order to observe the formation of IL-33:IgG complexes. A BioSep-SES-S 2000 column (300×7.4 mm, s/n 550331-4) was equilibrated with Dulbecco's PBS at 0.5 mL min.sup.−1 on an Agilent HP1100 HPLC. Peaks were detected using the 280 nm signal from a Diode Array Detector (DAD). These studies confirmed that antibody-antigen complex formation was fairly slow taking at least several hours. Once sufficient time for full complex formation had been allowed, trypsin was added to pre-formed IL-33:IgG complexes, followed by SEC analysis. 36 min trypsin digest led to an increase in the retention time of the main peak to 14.1 minutes (intermediate between the untreated complex peak elution time (13.6 min) and the intact H338L292 IgG elution (14.4 min)). Mass spectrometry methods were then used to identify the minimal H338L293 IgG epitope. The Shimadzu MALDI-TOF MS observed masses from captured 14.1 min peak were 3,209 and 4,485.3 Da. ABI4800 MALDI-TOF MS observed masses were 3,208.9 Da peak at high intensity with a secondary 4,486.4 Da peak also present. The observed precursor ion mass of 3206-3208 Da and ABI4800 MS/MS fragmentation analysis of the 3,206 Da precursor ions matched the predicted tryptic IL-33 fragment MLMVTLSPTKDFWLHANNKEHSVELHK. This lies within the overall primary sequence of r Human IL-33-Flag His10 (SEQ ID NO. 627) as shown below:—

(314) TABLE-US-00021 MSITGISPITEYLASLSTYNDQSITFALEDESYEIYVEDLKKDEKKDKVL LSYYESQHPSNESGDGVDGKMLMVTLSPTKDFWLHANNKEHSVELHKCEK PLPDQAFFVLHNMHSNCVSFECKTDPGVFIGVKDNHLALIKVDSSENLCT ENILFKLSETNPAFLYKVVGAADYKDDDDKAAHHHHHHHHHH

(315) The identified peptide together with a truncate (LSPTKDFWLHANNKEHSVELHK) and scrambled variants of both, were then chemically synthesised and used in confirmatory T100 Biacore (GE Healthcare) binding studies. Protein G′ (Sigma Aldrich, P4689) was covalently coupled to the surface of a CM5 sensor chip (GE Healthcare) using standard amine coupling reagents according to manufacturer's instructions. The protein G′ surface was used to capture H338L293 or ST2-Fc via the Fc domain to provide a surface density of approximately 290 RU per cycle. IL33 peptides prepared in HBS-EP+ buffer, at a range of concentrations were passed over the sensor chip surface. The surface was regenerated using two 10 mM Glycine washes of pH 1.7 and pH 1.5 between each injection of antibody. The resulting sensorgrams were evaluated using Biacore T100 evaluation software 2.0.3 (GE Healthcare) and fitted to a 1:1 Langmuir binding model, to provide relative binding data.

(316) The full length synthetic epitope peptide bound strongly to H338L293 IgG but not the IL-33 receptor (ST2-Fc). Both the full length and the truncated synthetic peptide bound strongly to H338L293 but the scrambled full length and truncated versions did not. This is strong evidence that IL-33 fragment identified by peptide excision is not an artefact and represents the core of the H338L293 epitope. 0.625-20 nM of the truncated peptide was flowed over H338L293 IgG to estimate affinity. Good quality 1:1 fits were obtained giving a K.sub.D value of 2.36 nM.

(317) FIG. 33 shows epitope mapping of H338L293. The top panel shows SEC analysis of IL33:IgG complexes with H338L293 pre and post digestion with trypsin. The lower panel shows the truncate peptide that was determined to bind strongly to H338L293 coloured black within the IL-33 structure described by Lingel et al 2009.

Example 6 Cys→Ser IL-33 Mutants

(318) To understand the role of the four free cysteines of human IL-33 in its conversion to the disulphide bonded form, we generated a full panel of all possible Cys-to-Ser mutants. Most of these mutant IL-33 molecules showed similar initial activity through ST2 compared with wild type IL-33. Following incubation in media mutants did not display faster gel migration, consistent with lack of ability to form 2 disulfide bonds. However, loss of potency after media treatment varied between mutants.

(319) Generation of IL-33 Cysteine to Serine Mutant Panel

(320) cDNA molecules encoding the mature component of Human IL-33 (112-270); accession number (Swiss-Prot) O95760, and a series of variants with 1, 2, 3 or 4 cysteine residues mutated to serine in all combinations (15 in total) were synthesized by primer extension PCR and cloned into pJexpress404 (DNA 2.0). The wild-type (WT) and mutant IL-33 coding sequences were modified to contain a 10×his, Avitag, and Factor-Xa protease cleavage site (MHHHHHHHHHHAAGLNDIFEAQKIEWHEAAIEGR) at the N-terminus of the proteins.

(321) DNA encoding the IL-33 mutants was used to transform E. coli BL21 Gold cells. Transformed cells were cultured at 37° C. until they reached an OD600 nm of 0.3 to 0.5. Cultures were then grown at 18° C. until they reached an OD600 nm of 0.6 to 0.8, when protein expression was induced by addition of 100 mM IPTG. Cultures were continued at 18° C. for a further 20 hours before cells were harvested by centrifugation and stored at −80° C.

(322) Cells were resuspended in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM Imidazole, 5 mM BetaMercaptoethanol containing complete protease inhibitor tablets (Roche, 11697498001), 2.5 u/ml Benzonase nuclease (merck Millipore, 70746-3) and 1 mg/ml recombinant lysozyme. Resuspended cells were lysed using a Constant Systems cell disruptor at 25 kpsi and clarified by centrifugation at 75,000×g for 2 hours at 4° C. IL33 was purified from the supernatant by Nickel affinity chromatography in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM Imidazole, 5 mM BetaMercaptoethanol, eluting in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 250 mM Imidazole, 5 mM BetaMercaptoethanol. IL33 was further purified by size exclusion chromatography using a Superdex 75 10/300 GL column (GE Healthcare) in Phosphate Buffered Saline pH 7.4. Peak fractions were analysed by SDS PAGE. Fractions containing pure IL33 were pooled and the concentration measured by Nanodrop A280 measurement. Final samples were analysed by SDS PAGE and intact mass spectrometry. Protein was snap frozen in liquid nitrogen.

(323) TABLE-US-00022 TABLE 17 IL-33 mutant sequences Position Position Position Position Mutant 208 227 232 259 AA.112-270 IL33-01 C C C C SEQ ID NO: 632 IL33-02 S C C C SEQ ID NO: 634 IL33-03 C S C C SEQ ID NO: 635 IL33-04 C C S C SEQ ID NO: 636 IL33-05 C C C S SEQ ID NO: 637 IL33-06 S S C C SEQ ID NO: 638 IL33-07 C C S S SEQ ID NO: 639 IL33-08 S C S C SEQ ID NO: 640 IL33-09 C S C S SEQ ID NO: 641 IL33-10 C S S C SEQ ID NO: 642 IL33-11 S C C S SEQ ID NO: 643 IL33-12 S S S C SEQ ID NO: 644 IL33-13 S S C S SEQ ID NO: 645 IL33-14 S C S S SEQ ID NO: 646 IL33-15 C S S S SEQ ID NO: 647 IL33-16 S S S S SEQ ID NO: 648
Activity of IL-33 Cys.fwdarw.Ser Mutants

(324) To check protein integrity, activity of untreated wild type IL-33 (IL33-01) and the IL-33 mutants were measured in a ST2-dependent signaling assay. NFκB signaling in Human umbilical vein endothelial cells (Huvecs) in response to IL-33 was assessed by nuclear translocation of the p65/RelA NFkB subunit detected by immunofluorecence staining as described in Example 1. To investigate loss of activity after cell culture media treatment, IL-33 proteins 01-16 were incubated overnight in Iscoves Modified Dulbeccos Medium (IMDM) and assessed in comparison with the untreated proteins.

(325) TABLE-US-00023 TABLE 18 Activity of IL-33 mutants in HUVEC NFkB translocation assay Loss of activity IC50 (nM) after IMDM Name Sequence Untreated treatment SEQUENCE IL33-01 (WT) CCCC 0.13 +++ SEQ ID NO: 632 IL33-02 SCCC 0.19 ++ SEQ ID NO: 634 IL33-03 CSCC 0.16 ++ SEQ ID NO: 635 IL33-04 CCSC 0.17 + SEQ ID NO: 636 IL33-05 CCCS 0.10 ++ SEQ ID NO: 637 IL33-06 SSCC 0.28 + SEQ ID NO: 638 IL33-07 CCSS 0.22 + SEQ ID NO: 639 IL33-08 SCSC 0.21 − SEQ ID NO: 640 IL33-09 CSCS 0.20 ++ SEQ ID NO: 641 IL33-10 CSSC 0.61 + SEQ ID NO: 642 IL33-11 SCCS 0.49 + SEQ ID NO: 643 IL33-12 SSSC 0.39 − SEQ ID NO: 644 IL33-13 SSCS 0.08 + SEQ ID NO: 645 IL33-14 SCSS 0.14 − SEQ ID NO: 646 IL33-15 CSSS 0.12 − SEQ ID NO: 647 IL33-16 SSSS 0.17 − SEQ ID NO: 648

(326) FIG. 34 shows activity of the IL-33 mutants before and after treatment for 18 hours with IMDM. Wild type IL-33 (IL33-01) that has been pre-treated with culture media completely lost detectable activity. All mutants displayed less potency loss than WT. Some mutants were completely protected from potency loss.

(327) Human Mast Cell Cytokine Release

(328) To see if mutants were more potent at stimulating downstream responses at longer timepoints in vitro, an overnight mast cell IL-6 production assay was used to measure activity of human IL-33 wild type and selected mutants. Assay methods are described in Example 2. Data are exemplified by IL33-11.

(329) FIG. 35A shows that IL33-11 has greater potency than IL-33 WT at stimulating human mast cell IL-6 production. IL33-01 (WT) and IL33-11 without prior treatment were used to stimulate IL-6 production from human cord blood derived mast cells at varying concentrations, wherein the x-axis is the concentration of IL-33 in molar concentration and the y-axis is the level of IL-6 detected in the supernatants after 18 hours.

(330) In Vivo Potency of Mutant IL-33

(331) Female BALB/c mice (6-8 weeks) were anaesthetized briefly with isofluorane and administered either 0.1-10 ug of wild type human IL-33 (IL33-01, SEQ ID NO: 632), IL33-11 (SEQ ID NO: 643) or vehicle intranasally in a total volume of 50 μl. 24 hours after challenge, mice were terminally anaesthetized with pentobarbital sodium prior to bronchoalveolar lavage (BAL). BALF was collected and analysed as described in Example 4.

(332) FIG. 35B shows that intranasal administration of IL33-11 double mutant required only a tenth as much protein for an equivalent ST2-dependent IL-5 response compared to native IL-33. This is consistent with prolonged activity of the mutant in contrast to the more rapid inactivation of the wild type IL-33 in the mouse lung environment.

(333) NMR Analysis of IL33-11

(334) To investigate conformational differences between IL33-11 and wild type human IL-33 protein (IL33-01), NMR analysis was performed.

(335) Production of .sup.15N-IL-33 Proteins

(336) DNA encoding wild type IL-33 with an N-terminal 6His tag and TEV protease cleavage site (SEQ ID. 633) was used to transform E. coli BL21 Gold cells. Transformed cells were cultured at 37° C. in M9 minimal media supplemented with 5 g/L of .sup.15N-IsoGro™ powder until they reached an OD600 nm of 0.6 to 0.8, when protein expression was induced by addition of 100 mM IPTG. Cultures were continued at 18° C. for a further 20 hours before cells were harvested by centrifugation and stored at −80° C. Cells were resuspended in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM Imidazole, 5 mM BetaMercaptoethanol containing Complete protease inhibitor tablets (Roche, 11697498001), 2.5 U/ml Benzonase nuclease (merck Millipore, 70746-3) and 1 mg/ml recombinant lysozyme. Resuspended cells were lysed using a Constant Systems cell disruptor at 25 kpsi and clarified by centrifugation at 75,000×g for 2 hours at 4° C. IL-33 was purified from the supernatant by Nickel affinity chromatography in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM Imidazole, 5 mM BetaMercaptoethanol, eluting in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 250 mM Imidazole, 5 mM BetaMercaptoethanol. Eluted protein was incubated with TEV protease and dialysed into 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM Imidazole, 5 mM BetaMercaptoethanol at 4° C. De-tagged protein was separated from uncleaved IL-33 by Nickel affinity chromatography in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM Imidazole, 5 mM BetaMercaptoethanol. IL-33 was further purified by size exclusion chromatography using a HiLoad 16/60 Superdex 75 column (GE healthcare) in 20 mM Sodium phosphate pH 6.5, 100 mM NaCl, 5 mM BetaMercaptoethanol, using an AKTAxpress FPLC system (GE healthcare). Peak fractions were analysed by SDS PAGE. Fractions containing pure IL-33 were pooled and the concentration measured by Nanodrop A280 measurement. Protein was concentrated using an Amicon 10,000 molecular weight cut-off spin concentrator to a final concentration of 1.8 mg/ml (100 μM) for NMR analysis.

(337) NMR analysis

(338) NMR spectra were recorded at 298 K on a Bruker Avance 600 MHz spectrometer running Topspin 2.3 equipped with a 5 mm TCI Cryoprobe with Z-axis gradients. The .sup.15N-labelled IL33 WT sample was prepared as described with the addition of 5% deuterium oxide to allow sample locking. The exemplified .sup.1H-.sup.15N correlation spectra were acquired employing the sofast HMQC pulse sequence (Schanda, P; Brutscher, B; Very fast two-dimensional NMR spectroscopy for real-time investigation of dynamic events in proteins on the time scale of seconds, J. Am. Chem. Soc. (2005) 127, 8014-5) with (F2×F1) 1024×64 complex points (in states-TPPI mode), 9615×1460 Hz sweep width, 53.4 ms×43.8 ms acquisition times.

(339) FIG. 36 shows overlay of the .sup.1H-.sup.15N HMQC spectra for 0.1 mM .sup.15N-labeled IL33-01 and IL33-11 plotted in black and red, respectively. Assignment for relevant residues are indicated. Data show peak shifts around C208 and C259 as expected. However, there are more more peak shifts than expected from T185 to A196 which might indicate a conformation change.

Example 7 Isolation and Identification of Anti-IL-33 Antibodies Using IL33-11

(340) Cys.fwdarw.Ser mutant IL-33 proteins stabilise IL-33 in its reduced form and have different conformations to wild type (as described in Example 6). Mutant proteins may provide availability of different antibody epitopes or greater longevity/stability of epitopes and may therefore be useful for isolating neutralizing IL-33 antibodies, in particular to the reduced form of IL-33. This example uses oxidation-resistant mutant IL33-11 protein to isolate IL-33 antibodies by phage display.

(341) Recombinant Proteins

(342) N-terminal tagged His10/Avitag IL33-01 (WT, SEQ ID NO. 632), N-terminal tagged

(343) His10/Avitag IL33-11 (C2085, C259S, SEQ ID 643) and N-terminal tagged His10/Avitag cyno IL-33 (SEQ ID 649), were generated by transforming E. coli BL21(DE3) cells. Transformed cells were cultured in autoinduction media (Overnight Express™ Autoinduction System 1, Merck Millipore, 71300-4) at 37° C. for 18 hours before cells were harvested by centrifugation and stored at −20° C. Cells were resuspended in BugBuster (Merck Millipore, 70921-5), containing complete protease inhibitor cocktail tablets (Roche, 11697498001), 2.5 u/ml Benzonase nuclease (merck Millipore, 70746-3) and 1 mg/ml recombinant lysozyme. Cell lysate was clarified by centrifugation at 75,000×g for 2 hours at 4° C. IL-33 proteins were purified from the supernatant by Nickel affinity chromatography in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM Imidazole, eluting in 50 mM Sodium phosphate, pH 8.0, 300 mM NaCl, 250 mM Imidazole. IL-33 was further purified by size exclusion chromatography using a Superdex 75 10/300 GL column in Phosphate Buffered Saline pH 7.4. Peak fractions were analysed by SDS PAGE. Fractions containing pure IL-33 were pooled and the concentration measured by Nanodrop A280 measurement. Final samples were analysed by SDS PAGE.

(344) The human ST2 vector described in Example 1 was modified to contain human ST2 ECD with a C-terminal Flag-His tag (SEQ ID NO 650).

(345) TABLE-US-00024 TABLE 19 Reagents Catalogue Number/ Reagent Supplier Designation SEQ His10/Avi Human IL33-01 In house PS-937 SEQ ID NO. 632 His10/Avi Human IL33-11 In house PS-938 SEQ ID NO. 643 His10/Avi Cyno IL-33 In House CCH 3.sup.rd Apr. SEQ ID 2014 NO. 649 IL-4Rα Flag ®His In house 020629080 Bovine insulin - biotin Sigma I2258 Human ST2 Flag His In house BK282 SEQ ID NO. 650 Human ST2.Fc R&D Systems
Protein Modifications

(346) Proteins containing the Avitag sequence motif (GLNDIFEAQKIEWHE) were biotinylated using the biotin ligase (BirA) enzyme (Avidty, Bulk BirA) following the manufacturer's protocol. All IgGs and modified proteins without Avitag used herein were biotinylated via free amines using EZ link Sulfo-NHS-LC-Biotin (Thermo/Pierce, 21335) as described in Example 1.

(347) Selections

(348) Selections were performed essentially as described in Example 1 but using the IL33-11 C2085, C259S mutant protein. In brief, the scFv-phage particles were incubated with biotinylated recombinant IL-33-11 in solution (biotinylated via Avi tag). Particles were incubated with 100 nM biotinylated recombinant IL33-11 for 2 hours. ScFv bound to antigen were then captured on streptavidin-coated paramagnetic beads (Dynabeads®, M-280) following manufacturer's recommendations. Unbound phage was washed away in a series of wash cycles using PBS-Tween. The phage particles retained on the antigen were eluted, infected into bacteria and rescued for the next round of selection. Two more rounds of selections were carried out in the presence of decreasing concentrations of biotinylated IL33-11 (50 nM and 25 nM).

(349) Inhibition of IL-33 Binding to ST2 by Unpurified scFv

(350) A representative number of individual clones from the selection outputs after two or three rounds of selection described above were grown up in 96-well plates. ScFv were expressed in the bacterial periplasm (Kipriyanov, et al. J Immunol Methods 200(1-2): 69-77 (1997)) and screened for their inhibitory activity in a homogeneous FRET (fluorescence resonance energy transfer) HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) based IL-33:ST2-binding assay. Essentially methods were similar to those described in Example 1. In this assay, samples competed with FLAG® His-tagged human ST2 for binding to biotinylated human IL33-01 (IL33-01, SEQ ID No. 632) (wild type) or biotinylated human IL33-11 (IL33-11, SEQ ID No. 643).

(351) Unpurified anti-IL-33 antibody samples were tested for inhibition of biotinylated IL-33 binding FLAG® His-tagged ST2 by adding 5 microlitres of each dilution of antibody test sample to a 384 well low volume assay plate (Costar, 3673). Next, a solution containing 4 nM human FLAG® His-tagged ST2 and 5 nM anti-FLAG® XL665 detection (Cisbio International, 61FG2XLB) was prepared and 2.5 microlitres of the mix added to the assay plate. This was followed by the addition of 2.5 microlitres of a solution containing 2.4 nM biotinylated human IL33-01 or IL33-11 combined with 1.5 nM streptavidin cryptate detection (Cisbio International, 610SAKLB). All dilutions were performed in assay buffer containing 0.8 M potassium fluoride (VWR, 26820.236) and 0.1% bovine serum albumin (BSA, PAA, K05-013) in Dulbeccos PBS (Invitrogen, 14190185). Assay plates were incubated for 1 hour at room temperature and time resolved fluorescence was read at 620 nm and 665 nm emission wavelengths using an EnVision plate reader (Perkin Elmer). Plates were incubated for a further 16 hour (overnight) at 4 degrees Celsius and time resolved fluorescence read again. The negative control (non-specific binding) was defined by replacing biotinylated human IL33-01 or IL33-11 combined with streptavidin cryptate detection with streptavidin cryptate detection only. Data were analysed as described in Example 1.

(352) Inhibition of IL-33 Binding to ST2 by Purified scFv

(353) Single chain Fv clones which showed an inhibitory effect on IL-33:ST2 interaction as unpurified periplasmic extracts at both time points were subjected to DNA sequencing (Osbourn, et al. Immunotechnology 2(3):181-96 (1996); Vaughan, et al. Nat Biotechnol 14(3):309-14 (1996)). Unique scFv were expressed again in bacteria and purified by affinity chromatography (as described in WO01/66754). The potencies of these samples were determined by competing a dilution series of the purified preparation against FLAG® His-tagged human ST2 for binding to biotinylated IL33-01 or biotinylated IL33-11 as described above. Assay plates were incubated for 1 hour at room temperature (1 hour incubation), or assay plates were incubated for 1 hour at room temperature followed by 16 hour at 4 degrees Celsius (overnight incubation). Purified scFv preparations that were capable of inhibiting the IL-33:ST2 interaction at both timepoints were selected for conversion to IgG format.

(354) FIG. 37A: shows the inhibition of the FRET signal after 1 hour incubation, produced by human IL-33-01 binding to human ST2 with increasing concentrations of IL-33 scFv antibody 33v20064, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(355) FIG. 37B: shows the inhibition of the FRET signal after 1 hour incubation, produced by IL33-11 binding to human ST2 with increasing concentrations of IL-33 scFv antibody 33v20064, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(356) FIG. 37C: shows the inhibition of the FRET signal after overnight incubation, produced by human IL-33-01 binding to human ST2 with increasing concentrations of IL-33 scFv antibody 33v20064, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(357) FIG. 37D: shows the inhibition of the FRET signal after overnight incubation, produced by IL33-11 binding to human ST2 with increasing concentrations of IL-33 scFv antibody 33v20064, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(358) Reformatting of scFv to IgG1

(359) Purified scFv preparations that were capable of inhibiting the IL-33:ST2 interaction were converted to whole immunoglobulin G1 (IgG1) antibody format as described in Example 1. Antibodies that inhibited to a similar or greater extent then IL330004 (Example 1, SEQ ID Nos. 12 and 17) were taken forward for further analysis. Such antibodies are exemplified by 33v20064. SEQ ID NOs. corresponding to the various regions of antibody 33v20064 are shown in Table 20.

(360) TABLE-US-00025 TABLE 20 Anti-IL-33 Antibody Sequences VH VL VH CDRs VL CDRs IgG1 Sequence Sequence 1, 2, 3 1, 2, 3 33v20064 SEQ ID SEQ ID SEQ ID SEQ ID NO: 272 NO: 277 NO: 273 NO: 278 SEQ ID SEQ ID NO: 274 NO: 279 SEQ ID SEQ ID NO: 275 NO: 280
Inhibition of IL-33 Binding to ST2 by Purified IgG

(361) The ability of anti-IL-33 antibodies to inhibit the binding of biotinylated IL-33-01 to the FLAG®-His tagged ST2 receptor was assessed in a biochemical HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) competition assay, the principles of which are described above. Activity of purified IgG preparations were determined by competing a dilution series of the purified IgG against human FLAG®-His tagged ST2 for binding to human biotinylated human IL-33-01 (SEQ ID No. 632).

(362) FIG. 38A: shows the inhibition of the FRET signal after 1 hour incubation, produced by human IL-33 binding to human ST2 with increasing concentrations of 33v20064 IgG1 antibody, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(363) FIG. 38B: shows the inhibition of the FRET signal after overnight incubation, produced by human IL-33 binding to human ST2 with increasing concentrations of 33v20064 IgG1 antibody, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(364) Inhibition of IL-6 Production in Huvec by IgG

(365) 33v20064 was assessed for inhibition of IL-33 stimulated IL-6 production in HUVECs, the methods of which are described in Example 2. His-Avi human IL-33 wild type (IL33-01, SEQ ID No. 632) (30 ng/mL) or His-Avi mutant IL-33 (IL33-11, SEQ ID No. 643) (30 ng/mL) were used to stimulate HUVECs in the presence of varying concentrations of test antibodies.

(366) FIG. 39A: shows the inhibition of IL-6 production from IL33-01 (WT) stimulated HUVECs by antibody 33v20064 compared with IL330004 and anti-NIP IgG1 negative control antibody, NIP228, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent maximal response. 33v20064 showed partial inhibition of the response to WT IL-33 at high antibody concentrations, whereas IL330004 showed no effect.

(367) FIG. 39B: shows the inhibition of IL-6 production from IL33-11 stimulated HUVECs by antibody 33v20064 compared with IL330004 and anti-NIP IgG1 negative control antibody, NIP228, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent maximal response. 33v20064 showed a more complete inhibition of IL-6 production stimulated by IL33-11 mutant compared with IL330004.

(368) Cross-Reactivity of Anti-IL-33 Antibodies

(369) Cross-reactivity of anti IL-33 antibody 33v20064 was determined using a homogeneous FRET (fluorescence resonance energy transfer) HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) based IL-33:mAb-binding assay. In this assay, samples competed with biotinylated human IL-33-01 (SEQ ID No. 632) for binding to DyLight labelled 33v20064 IgG.

(370) TABLE-US-00026 TABLE 21 FRET Assay Reagents Catalogue Number/ Reagent Supplier Designation SEQUENCE Human IL-33 Flag ®His In house PS-582 SEQ ID No. 627 Mouse IL-33 Flag ®His In house BK-265 SEQ ID No. 628 Cynomolgus IL-33 In house PS-368 SEQ ID Flag ®His No. 629 B7H3 Avi-His In house DBPur125 Human IL-1 alpha R&D Systems 201-LB/CF Human IL-1 beta R&D Systems 200-LA/CF

(371) Human, cyno and mouse IL-33 FLAG® His (described in Example 1) were tested for inhibition of human IL-33 binding to DyLight labelled 33v20064 by adding 5 microlitres of each dilution of IL-33 sample to a 384 well low volume assay plate (Costar, 3673). Next, a solution containing 20 nM DyLight labelled 33v20064 was prepared and 2.5 microlitres added to the assay plate (labelled using kit (Innova Biosciences, 326-0010) as per manufacturer's instructions). This was followed by the addition of 2.5 microlitres of a solution containing 1.2 nM biotinylated human IL-33-01 combined with 1.5 nM streptavidin cryptate detection (Cisbio International, 610SAKLB). All dilutions were performed in assay buffer containing 0.8 M potassium fluoride (VWR, 26820.236) and 0.1% bovine serum albumin (BSA, PAA, K05-013) in Dulbeccos PBS (Invitrogen, 14190185). Assay plates were incubated for 1 hour at room temperature and time resolved fluorescence was read at 620 nm and 665 nm emission wavelengths using an EnVision plate reader (Perkin Elmer). Data were analysed by calculating the 665/620 nm ratio followed by the % Delta F values for each sample. The 665/620 nm ratio was used to correct for sample interference using Equation 1. The % Delta F for each sample was then calculated using Equation 2. The negative control (non-specific binding) was defined by replacing biotinylated human IL-33 combined with streptavidin cryptate detection with streptavidin cryptate detection only. The % Delta F values were subsequently used to calculate % specific binding as described in Equation 3. IC.sub.50 values were determined using GraphPad Prism software by curve fitting using a four-parameter logistic equation (Equation 4). These results demonstrated that 33v20064 cross reacts with cynomolgus IL-33. 33v20064 but did not show competition with mouse IL-33.

(372) FIG. 40A: shows the inhibition of the FRET signal, produced by biotinylated human IL-33-01 binding to DyLight labelled 33v20064, with increasing concentrations of test proteins, wherein the x-axis is the concentration of test sample in molar concentration and the y-axis is percent specific binding. Inhibition of the FRET signal was observed with human and cynomolgus, but not mouse, IL-33.

(373) Selectivity of Anti-IL-33 Antibodies

(374) Selectivity of anti IL-33 antibody 33v20064 was determined using a homogeneous FRET (fluorescence resonance energy transfer) HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) based IL-33:mAb-binding assay. In this assay, samples competed with biotinylated His-Avi human IL-33 (IL33-01, SEQ ID No. 632) for binding to wild type DyLight labelled 33v20064 IgG. Human IL-1 alpha and human IL-1 beta were tested for inhibition of biotinylated IL-33-01 binding DyLight labelled 33v20064 as described above. These results demonstrated that 33v20064 did not show competition with human IL-1 alpha or IL-1 beta.

(375) FIG. 40B: shows the inhibition of the FRET signal, produced by biotinylated human IL-33-01 binding to DyLight labelled 33v20064, with increasing concentrations of test proteins, wherein the x-axis is the concentration of test sample in molar concentration and the y-axis is percent specific binding. Inhibition of the FRET signal was not observed with with human IL-1 alpha or IL-1 beta.

Example 8 Optimization of Anti-IL-33 Ab 33v20064

(376) Germlining Framework Regions of 33v20064

(377) The amino acid sequences of the V.sub.H and V.sub.L domains of the parent antibody 33v20064 were aligned to the known human germline sequences in the IMGT database (Lefranc, M. P. et al. Nucl. Acids Res. 2009. 37(Database issue): D1006-D1012), and the closest germline was identified by sequence similarity. For the V.sub.H domains of the 33v20064 antibody lineage this was IGHV3-23*01. For the V.sub.L domains it was IGLV3-1. Germlining was carried out on 33v20064 prior to the affinity maturation process. Without considering the Vernier residues (Foote 1992), which were left unchanged, there were 6 residues in the frameworks of the V.sub.L domains of 33v20064 which differed from germline, 5 of which were reverted to the closest germline sequence using the Kunkel mutagenesis method (Clackson, T. and Lowman, H. B. Phage Display—A Practical Approach, 2004. Oxford University Press) with the appropriate mutagenic primers. The product of this germlining was 33_640001. Sequence ID Numbers are described in Table 22.

(378) TABLE-US-00027 TABLE 22 Antibody 33v20064 Germline Sequences VH VL VH CDRs VL CDRs IgG1 Sequence Sequence 1, 2, 3 1, 2, 3 33_640001 SEQ ID SEQ ID SEQ ID SEQ ID NO: 282 NO: 287 NO: 283 NO: 288 SEQ ID SEQ ID NO: 284 NO: 289 SEQ ID SEQ ID NO: 285 NO: 290
Inhibition of IL-33 Binding to ST2 by Purified scFv

(379) Activity of 33_640001 was compared with its non-germlined parent, 33v20064. The ability of scFv antibodies to inhibit the binding of biotinylated His Avi human IL-33 (IL33-01, SEQ ID No. 632) to the FLAG®-His tagged ST2 receptor was assessed in a biochemical HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) competition assay as described in Example 7.

(380) FIG. 41A: shows the inhibition of the FRET signal after 1 hour incubation, produced by human IL-33 binding to human ST2 with increasing concentrations of scFv antibodies, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding. 33_640001 had equivalent activity to its non-germlined parent.

(381) FIG. 41B: shows the inhibition of the FRET signal after overnight incubation, produced by human IL-33 binding to human ST2 with increasing concentrations of scFv antibodies, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding. 33_640001 had equivalent activity to its non-germlined parent.

(382) Affinity Maturation

(383) 33v20064 was optimised using a targeted mutagenesis approach and affinity-based phage display selections. Large scFv-phage libraries derived from the germlined parent (33_640001) were created by oligonucleotide-directed mutagenesis of the variable heavy (V.sub.H) complementarity determining region 3 (CDR3) and light (V.sub.L) chain CDR3 using standard molecular biology techniques as described (Clackson 2004). For the V.sub.H CDR3, the two Vernier positions preceding the Kabat-defined CDRs (i.e. V.sub.H positions 93 and 94) were also included for potential optimisation in the targeted mutagenesis approach. The libraries were subjected to affinity-based phage display selections in order to select variants with higher affinity for human IL-33. These selections were carried out by either alternating the biotinylated His-Avi human IL-33 wild type (IL33-01, SEQ ID NO. 632) and biotinylated His Avi mutant IL-33 (IL33-11, SEQ ID NO. 643) antigens in sequential rounds, or with the biotinylated IL33-11 antigen only in all rounds. The selections were performed essentially as described previously (Thompson 1996). In brief, the scFv-phage particles were incubated with the recombinant biotinylated antigen in solution. ScFv-phage bound to antigen were then captured on streptavidin-coated paramagnetic beads (Dynabeads® M-280) following the manufacturer's recommendations. The selected scFv-phage particles were then rescued as described previously (Osbourn, J. K., et al. Immunotechnology, 1996. 2(3): p. 181-96), and the selection process was repeated in the presence of decreasing concentrations of biotinylated antigen—typically from 50 nM to 10 pM over five rounds of selection.

(384) 33v20064 was also optimised using ribosome display technology essentially as described by Hanes et al. (Hanes, J., et al. Methods in Enzymology, 2000. 328: p. 404-30). The parent scFv clone 33v20064 was used as a template for library construction and conversion to a ribosome display format for subsequent selections. On the DNA level, a T7 promoter was added at the 5′-end for efficient transcription to mRNA. On the mRNA level, the construct contained a prokaryotic ribosome-binding site (Shine-Dalgarno sequence). At the 3′-end of the single chain, the stop codon was removed and a portion of M13 bacteriophage gIII (gene III) was added to act as a spacer between the nascent scFv polypeptide and the ribosome (Hanes 2000).

(385) A ribosome display library derived from the parent (33v20064) scFv construct was created by random mutagenesis using the Diversify™ PCR (polymerase chain reaction) Random Mutagenesis Kit (BD Biosciences) following the manufacturer's recommendations. The conditions for this error-prone PCR (EP) were chosen to introduce on average 8.1 nucleotide changes per 1000 basepairs (according to the manufacturer). The resulting EP library was then used in affinity-based ribosome display selections (Hanes 2000). The scFvs were expressed in vitro using the RiboMAX™ Large Scale RNA Production System (T7) (Promega) following the manufacturer's protocol and an E. coli-based prokaryotic cell-free translation system. The produced scFv antibody-ribosome-mRNA (ARM) complexes were incubated in solution with biotinylated human IL-33 antigen, with either alternating the biotinylated IL33-01 and biotinylated IL33-11 antigens in sequential rounds, or with the biotinylated IL-33-11 antigen only in all rounds. The specifically bound tertiary complexes (IL-33:ARM) were captured on streptavidin-coated paramagnetic beads (Dynabeads® M-280) following the manufacturer's recommendations (Dynal) whilst unbound ARMs were washed away. The mRNA encoding the bound scFvs were then recovered by reverse transcription-PCR (RT-PCR). The selection process was repeated on the obtained population for further rounds of selections with decreasing concentrations of biotinylated human IL-33 (100 nM to 100 pM over 5 rounds), in order to enrich and thereby select clones with higher affinity for IL-33. The outputs from selection rounds 3, 4 and 5 were sub-cloned into pCantab6 (McCafferty, J., et al. Appl Biochem Biotechnol, 1994. 47(2-3): p. 157.), and improved clones were identified as described below.

(386) Inhibition of IL-33 Binding to mAb by Unpurified scFv

(387) A representative number of individual clones from the selection outputs were grown up in 96-well plates. ScFv were expressed in the bacterial periplasm (Kipriyanov, et al. J Immunol Methods 200(1-2): 69-77 (1997)) and screened for their inhibitory activity in a homogeneous FRET (fluorescence resonance energy transfer) HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) based IL-33:mAb-binding assay. In this assay, samples competed with DyLight labelled 33v20064 IgG for binding to biotinylated His Avi IL-33-01 (SEQ ID NO. 632) or biotinylated His Avi cynomolgus IL-33 (SEQ ID NO. 649). Such epitope competition assays are based on the principle that a test antibody sample, which recognizes a similar epitope to the labelled anti IL-33 IgG, will compete with the labelled IgG for binding to biotinylated IL-33 resulting in a reduction in assay signal.

(388) Unpurified anti-IL-33 antibody samples were tested for inhibition of biotinylated His Avi IL33-01 (human) or biotinylated His Avi cynomolgus IL-33 binding DyLight labelled 33v20064 by adding 5 microlitres of sample to a 384 well low volume assay plate (Costar, 3673). Next, a solution containing 2.4 nM DyLight labelled 33v20064 was prepared for the human IL-33 assay and 6 nM DyLight labelled 33v20064 was prepared for the cynomolgus assay and 2.5 microlitres added to the assay plates (labelled using kit (Innova Biosciences, 326-0010) as per manufacturer's instructions). This was followed by the addition of 2.5 microlitres of a solution containing 0.8 nM biotinylated human IL-33-01 combined with 0.75 nM streptavidin cryptate detection (Cisbio International, 610SAKLB) for the human assay or a solution containing 4 nM biotinylated cynomolgus IL-33 combined with 1.5 nM streptavidin cryptate detection (Cisbio International, 610SAKLB) for the cynomolgus assay. All dilutions were performed in assay buffer containing 0.8 M potassium fluoride (VWR, 26820.236) and 0.1% bovine serum albumin (BSA, PAA, K05-013) in Dulbeccos PBS (Invitrogen, 14190185). Assay plates were incubated for 1 hour at room temperature and time resolved fluorescence was read at 620 nm and 665 nm emission wavelengths using an EnVision plate reader (Perkin Elmer). Data were analysed by calculating the 665/620 nm ratio followed by the % Delta F values for each sample. The 665/620 nm ratio was used to correct for sample interference using Equation 1. The % Delta F for each sample was then calculated using Equation 2. The negative control (non-specific binding) was defined by replacing biotinylated IL-33 combined with streptavidin cryptate detection with streptavidin cryptate detection only. The % Delta F values were subsequently used to calculate % specific binding as described in Equation 3.

(389) As the epitope competition assay reached its limit of sensitivity an assay using an intermediate optimised mAb 33_640027 was used for testing unpurified scFv samples. The assay was essentially as described for the 33v20064 competition assay with the following modifications: 20 nM DyLight labelled 33_640027 was prepared and 2.5 microlitres added to the assay plates. This was followed by the addition of 2.5 microlitres of a solution containing 0.32 nM biotinylated human IL-33-01 combined with 0.75 nM streptavidin cryptate detection (Cisbio International, 610SAKLB) for the human assay or a solution containing 0.8 nM biotinylated cynomolgus IL-33 combined with 1.5 nM streptavidin cryptate detection (Cisbio International, 610SAKLB) for the cynomolgus assay.

(390) Inhibition of IL-33 Binding to mAb by Purified scFv

(391) Single chain Fv clones which showed an inhibitory effect on IL-33:mAb interaction as unpurified periplasmic extracts were subjected to DNA sequencing (Osbourn, et al. Immunotechnology 2(3):181-96 (1996); Vaughan, et al. Nat Biotechnol 14(3):309-14 (1996).). Unique scFv were expressed again in bacteria and purified by affinity chromatography (as described in WO01/66754). Purified anti-IL-33 antibody samples were tested for potency of inhibition by competing a dilution series of the purified preparation against DyLight labelled 33v20064 IgG or DyLight labelled 33_640027 IgG for binding to biotinylated His Avi IL33-01, biotinylated His Avi IL33-11 or biotinylated His Avi cynomolgus IL-33. Methods are as described in the previous section.

(392) TABLE-US-00028 TABLE 23 Activity of scFv antibodies in epitope competition assays Potency IC50 (nM) Assay 33v20064 33_640027 33_640047 33_640050 33v20064 IL33-01 12 0.4 0.4 0.2 IL33-11 26 0.6 0.5 0.4 Cyno IL-33 103 8.2 2.7 1.4 33_640027 IL33-01 1498 3.8 2.3 2.5 IL33-11 1467 2.9 Not determined 2.4 Cyno IL-33 3433 38 8.8 4.1
Reformatting of scFv to IgG1

(393) Single chain Fv clones with desirable properties from the IL-33:mAb binding assays were converted to whole immunoglobulin G1 (IgG1) antibody format as described in Example 1. These include antibodies 33_640027 (derived from the EP library selections), and 33_640047, 33_640050 (derived from the V.sub.H CDR3 block mutagenesis library selections) SEQ ID NOs corresponding to the various regions of these antibodies are shown in Table 24

(394) TABLE-US-00029 TABLE 24 Sequences of IL33 antibodies VH VL VH CDRs VL CDRs IgG1 Sequence Sequence 1, 2, 3 1, 2, 3 33_640027 SEQ ID SEQ ID SEQ ID SEQ ID NO: 292 NO: 297 NO: 293 NO: 298 SEQ ID SEQ ID NO: 294 NO: 299 SEQ ID SEQ ID NO: 295 NO: 300 33_640047 SEQ ID SEQ ID SEQ ID SEQ ID NO: 312 NO: 317 NO: 313 NO: 318 SEQ ID SEQ ID NO: 314 NO: 319 SEQ ID SEQ ID NO: 315 NO: 320 33_640050 SEQ ID SEQ ID SEQ ID SEQ ID NO: 302 NO: 307 NO: 303 NO: 308 SEQ ID SEQ ID NO: 304 NO: 309 SEQ ID SEQ ID NO: 305 NO: 310
Inhibition of IL-33 Binding to mAb by Purified IgG

(395) The ability of anti-IL-33 antibodies to inhibit the binding of biotinylated His Avi IL33-01, biotinylated His Avi IL33-11 or biotinylated His Avi cynomolgus IL-33 to the DyLight labelled 33v20064 IgG or DyLight labelled 33_640027 IgG was assessed in a biochemical HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) competition assay as described. IgGs with desirable properties from the IL-33:mAb binding assays were selected for further analysis.

(396) Inhibition of IL-8 Production in Huvec by IgG

(397) A cytokine release assay was used to assess the inhibition of IL-33 induced IL-8 production from human umbilical vein endothelial cells (Huvec) by anti-IL-33 antibodies. Cells were exposed to IL-33 in the presence or absence of test antibody or ST2.Fc (R&D systems) essentially as described in Example 2 with minor modifications. Test solutions of purified IgG (in duplicate) were diluted to the desired concentration in complete culture media. N-terminal His Avi IL-33 (IL33-01, SEQ ID NO 632) was prepared in complete culture media mixed with appropriate test antibody to give a final IL-33 concentration of 2 ng/mL. All samples were incubated for 30 minutes at room temperature, prior to transfer of IL-33/antibody mixture to the assay plate. Following 18-24 hour incubation, IL-8 was measured in cell supernatants by ELISA (R&D Systems, DY208) adapted for europium readout as described in Example 2. Data were analyzed using Graphpad Prism software. IC.sub.50 values were determined by curve fitting using a four-parameter logistic equation. IC.sub.50 values were calculated and are summarized in Table 25 below.

(398) FIG. 42A shows HUVECs stimulated with IL33-01 in the presence of 33v20064, 33_640050, human ST2-Fc or control mAb, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is a percentage of the maximum response (IL-8 production).

(399) Neutralization of Mammalian Full Length IL-33

(400) Full length IL-33 is also active (Cayrol et al., Proc Natl Acad Sci USA 106(22):9021-6 (2009); Hayakawa et al., Biochem Biophys Res Commun 387(1):218-22 (2009); Talabot-Ayer et al., J Biol Chem. 284(29):19420-6 (2009)).

(401) To evaluate the ability of antibodies to neutralize full length IL-33, HEK293-EBNA cells expressing full length (FL) HuIL-33 (and mock-transfected controls) were harvested 24 hours following transfection with accutase (PAA, #L11-007). Cells were diluted to 1×10.sup.8/mL with PBS and homogenized for 30 seconds using a tissue homogenizer. Cell debris was removed by centrifugation. HUVECs were stimulated with cell lysates at varying concentrations. Stimulation of cytokine production was only observed with full length IL-33-transfected cell lysate and not with mock transfected cell lysate. A 1:1000 concentration of lysate that stimulated a sub-maximal cytokine release (approx EC.sub.50) was selected for antibody neutralization studies. Experiments were performed as described above. IC.sub.50 values were calculated and are summarized in Table 25 below.

(402) FIG. 42B shows HUVECs stimulated with full length IL-33 cell lysate in the presence of 33v20064, 33_640050, human ST2-Fc or control mAb, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is a percentage of the maximum response (IL-8 production).

(403) TABLE-US-00030 TABLE 25 IC50 values in the Huvec IL-8 assay Geomean IC50 (95% CI) (nM) vs. full length IL-33 Antibody vs. His Avi IL-33 cell lysate 33v20064 Partial inhibition 185 33_640050 0.7 (0.3-1.5) 0.32 ST2.Fc 0.3 (0.1-0.9) 0.67
Prevention of Disulphide Bonded Form of IL-33 by IgG

(404) IL-33-01 (0.14 nM or 3 ng/mL) was incubated in IMDM or PBS, both containing 1% BSA in the presence of absence of antibodies (25 ug/mL) or human ST2-Fc (105 ug/mL) for 0-24 hours at 37° C., 5% CO.sub.2. At various time points aliquots were removed and added to a pre-chilled plate containing PBS or sST2 (final concentration 10.5 ug/mL). sST2 was spiked into control mAb and untreated samples at harvesting to stop the IL-33 oxidation reaction continuing. Samples were aliquoted into pre-frozen 96 well plates and stored at −80° C. The human IL-33 ELISA was performed according to manufacturer's instructions (R&D Systems, Cat #DY3625, Lot #1362797) with the substitution of DELFIA detection system (Perkin Elmer) in place of streptavidin-HRP and onwards. Briefly, black 96 well Maxisorp plates were coated with 50 uL per well of capture antibody overnight at room temperature. Plates were washed 3× with 300 uL 0.05% Tween-20 in PBS and blocked with 150 uL 1% BSA in PBS for 1 hour at room temperature. Plates were washed 3× and 50 uL per well of samples or standards were added to the plate for 2 hours at room temperature with shaking (400 rpm). Plates were washed 3× and 50 uL per well of detection antibody was added to the plate for 2 hours at room temperature with shaking (400 rpm). Plates were washed as previously stated and 50 uL per well of streptavidin-europium diluted 1 in 1000 in DELFIA Assay Buffer was added to the plate for 40 minutes at room temperature, protected from light, with shaking (400 rpm). Plates were washed 7× with 300 uL per well of DELFIA Wash Buffer. 50 uL per well of DELFIA Enhancement Solution (pre-warmed to room temperature) was added to the plate. After 10 minutes incubation at room temperature protected from the light, fluorescence was measured using an EnVision plate reader (PerkinElmer). Standards and data interpolation were performed in Microsoft Excel with subsequent analysis performed in GraphPad Prism software.

(405) As discussed in Example 4, FIG. 24A, this ELISA detects predominantly disulphide bonded IL-33 (IL33-DSB) within the range of IL-33 concentrations measured in this experiment. The ELISA is used here to monitor the conversion of IL-33 to its disulphide bonded form in the presence of test antibodies.

(406) FIG. 43 shows conversion of IL33-01 to its disulphide bonded form (IL33-DSB) during incubation in IMDM (FIG. 43A) or PBS (FIG. 43B) in the presence or absence of test antibodies, wherein the x-axis is time in hours and the y-axis is the concentration of IL-33-DSB. IL330004 and 33v20064 slow the rate of IL-33 conversion to IL33-DSB. 33_640050 and ST2.Fc prevent conversion to IL-33-DSB over the time course tested.

(407) Recombination of Beneficial Mutations and Further Optimization

(408) With the aim of generating further affinity improvements, beneficial mutations identified from previous selection and screening cascades were recombined in a number of different ways, either by a simple additive approach or via a recombination library approach with further selections.

(409) Sequence analyses suggested that there were two single-point mutational ‘hotspots’ which were prevalent in many of the lead antibody sequences; I98M in V.sub.H CDR3 and Q50R in VL CDR2 (Kabat numbering). These two mutations were grafted onto the 33_640001 construct to generate a new antibody, 33_640036, using standard molecular biology techniques. In a further recombination, the V.sub.H of 33_640047 was paired with the V.sub.L of 33_640036 to generate antibody 33_640117. These are examples of sequence recombination using an additive approach. SEQ ID NOs are shown in Table 26.

(410) TABLE-US-00031 TABLE 26 Sequences of IL33 antibodies VH VL VH CDRs VL CDRs IgG1 Sequence Sequence 1, 2, 3 1, 2, 3 33_640036 SEQ ID SEQ ID SEQ ID SEQ ID NO: 352 NO: 357 NO: 353 NO: 358 SEQ ID SEQ ID NO: 354 NO: 359 SEQ ID SEQ ID NO: 355 NO: 360 33_640117 SEQ ID SEQ ID SEQ ID SEQ ID NO: 362 NO: 367 NO: 363 NO: 368 SEQ ID SEQ ID NO: 364 NO: 369 SEQ ID SEQ ID NO: 365 NO: 370

(411) In addition, selection outputs from block mutagenesis libraries covering the V.sub.H CDR3 and V.sub.L CDR3 regions had shown affinity improvements and good sequence diversity and were thus recombined using a population cloning approach. Round 3 selection outputs were recombined to form libraries in which clones contained randomly paired, individually randomised V.sub.H CDR3 and VL CDR3 sequences. These recombined V.sub.H CDR3/V.sub.L CDR3 libraries were then used in ribosome display selections with either alternating the biotinylated His Avi human IL-33 wild type (IL33-01, SEQ ID NO. 632) and biotinylated His Avi mutant IL-33 (IL33-11, SEQ ID NO. 643) antigens in sequential rounds, or with the biotinylated His Avi IL33-11 antigen only in all rounds. The selections were performed essentially as described for the individual CDR3 libraries, in the presence of decreasing concentrations of biotinylated antigen—from 50 nM to 30 pM over five rounds of selection.

(412) Crude scFv-containing periplasmic extracts were prepared of a representative number of individual scFv's from the selection outputs of the recombined V.sub.H CDR3/V.sub.L CDR3 libraries. The ability of anti-IL-33 antibodies to inhibit the binding of biotinylated His Avi IL33-01 or biotinylated His Avi cynomolgus IL-33 to the DyLight labelled 33v20064 IgG or DyLight labelled 33_640027 IgG was assessed in a biochemical HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) competition assay as described. ScFv variants which showed a significantly improved inhibitory effect when compared to parent scFv and leads generated pre-recombination, were subjected to DNA sequencing, and unique recombined variants were produced as purified scFv and tested as described in the previous section.

(413) Optimised antibodies obtained from these recombined libraries are exemplified by 33_640076, 33_640081, 33_640082, 33_640084, 33_640086 and 33_640087. The SEQ ID NOs. of these antibodies are shown in Table 27.

(414) TABLE-US-00032 TABLE 27 Sequences of IL33 antibodies VH VL VH CDRs VL CDRs IgG1 Sequence Sequence 1, 2, 3 1, 2, 3 33_640076 SEQ ID SEQ ID SEQ ID SEQ ID NO: 372 NO: 377 NO: 373 NO: 378 SEQ ID SEQ ID NO: 374 NO: 379 SEQ ID SEQ ID NO: 375 NO: 380 33_640081 SEQ ID SEQ ID SEQ ID SEQ ID NO: 382 NO: 387 NO: 383 NO: 388 SEQ ID SEQ ID NO: 384 NO: 389 SEQ ID SEQ ID NO: 385 NO: 390 33_640082 SEQ ID SEQ ID SEQ ID SEQ ID NO: 392 NO: 397 NO: 393 NO: 398 SEQ ID SEQ ID NO: 394 NO: 399 SEQ ID SEQ ID NO: 395 NO: 400 33_640084 SEQ ID SEQ ID SEQ ID SEQ ID NO: 402 NO: 407 NO: 403 NO: 408 SEQ ID SEQ ID NO: 404 NO: 409 SEQ ID SEQ ID NO: 405 NO: 410 33_640086 SEQ ID SEQ ID SEQ ID SEQ ID NO: 412 NO: 417 NO: 413 NO: 418 SEQ ID SEQ ID NO: 414 NO: 419 SEQ ID SEQ ID NO: 415 NO: 420 33_640087 SEQ ID SEQ ID SEQ ID SEQ ID NO: 422 NO: 427 NO: 423 NO: 428 SEQ ID SEQ ID NO: 424 NO: 429 SEQ ID SEQ ID NO: 425 NO: 430

(415) Additional spontaneous mutations were introduced into the variable regions of these antibodies in scFv format during the ribosome display selection procedures as a result of repeated rounds of PCR amplifications. These events add to the sequence diversity of the outputs but are often undesirable when they occur in the framework regions. Hence the spontaneous mutations which occurred in the framework regions of 33_640076, 33_640081, 33_640082, 33_640084, 33_640086 and 33_640087 were reverted back to germline on the IgG constructs as described in Example 3 using standard molecular biology techniques. Such spontaneous mutations which occurred in any of the CDRs or in the Vernier residues adjacent to the CDRs (e.g. V.sub.H positions 27, 28, 29, 30, 93 and 94 by Kabat numbering) were left unchanged. As an additional strategy to increase affinity, a previously identified ‘hotspot’ (the Q50R mutation in V.sub.L CDR2) was also grafted onto the constructs at the same time. The antibodies resulting from these germlining and hotspot grafting modifications were named 33_640076-1, 33_640081-A, 33_640082-2, 33_640084-2, 33_640086-2 and 33_640087-2, corresponding to their parental antibodies of 33_640076, 33_640081, 33_640082, 33_640084, 33_640086 and 33_640087 respectively. The SEQ ID NO. of these antibodies are shown in Table 28.

(416) TABLE-US-00033 TABLE 28 Sequences of IL33 antibodies VH VL VH CDRs VL CDRs IgG1 Sequence Sequence 1, 2, 3 1, 2, 3 33_640076-1 SEQ ID SEQ ID SEQ ID SEQ ID NO: 432 NO: 437 NO: 433 NO: 438 SEQ ID SEQ ID NO: 434 NO: 439 SEQ ID SEQ ID NO: 435 NO: 440 33_640081-A SEQ ID SEQ ID SEQ ID SEQ ID NO: 442 NO: 447 NO: 443 NO: 448 SEQ ID SEQ ID NO: 444 NO: 449 SEQ ID SEQ ID NO: 445 NO: 450 33_640082-2 SEQ ID SEQ ID SEQ ID SEQ ID NO: 452 NO: 457 NO: 453 NO: 458 SEQ ID SEQ ID NO: 454 NO: 459 SEQ ID SEQ ID NO: 455 NO: 460 33_640084-2 SEQ ID SEQ ID SEQ ID SEQ ID NO: 462 NO: 467 NO: 463 NO: 468 SEQ ID SEQ ID NO: 464 NO: 469 SEQ ID SEQ ID NO: 465 NO: 470 33_640086-2 SEQ ID SEQ ID SEQ ID SEQ ID NO: 472 NO: 477 NO: 473 NO: 478 SEQ ID SEQ ID NO: 474 NO: 479 SEQ ID SEQ ID NO: 475 NO: 480 33_640087-2 SEQ ID SEQ ID SEQ ID SEQ ID NO: 482 NO: 487 NO: 483 NO: 488 SEQ ID SEQ ID NO: 484 NO: 489 SEQ ID SEQ ID NO: 485 NO: 490
Optimisation of Additional CDRs

(417) As a further strategy to increase affinity, additional CDRs were optimised. Large scFv-phage libraries derived from the germlined parent (33_640001) were created by oligonucleotide-directed mutagenesis of the variable heavy (V.sub.H) CDR1 and CDR2 and light (V.sub.L) chain CDR1 and CDR2 using standard molecular biology techniques as described (Clackson 2004). Selections and screening were performed as described for V.sub.HV.sub.L CDR3 libraries. The most improved antibody variants were obtained from the V.sub.H CDR2 library. These are exemplified by 33_640166, 33_640169, 33_640170. The SEQ ID NOs. of these antibodies are shown in Table 29.

(418) TABLE-US-00034 TABLE 29 Sequences of IL33 antibodies VH VL VH CDRs VL CDRs IgG1 Sequence Sequence 1, 2, 3 1, 2, 3 33_640166 SEQ ID SEQ ID SEQ ID SEQ ID NO: 322 NO: 327 NO: 323 NO: 328 SEQ ID SEQ ID NO: 324 NO: 329 SEQ ID SEQ ID NO: 325 NO: 330 33_640169 SEQ ID SEQ ID SEQ ID SEQ ID NO: 332 NO: 337 NO: 333 NO: 338 SEQ ID SEQ ID NO: 334 NO: 339 SEQ ID SEQ ID NO: 335 NO: 340 33_640170 SEQ ID SEQ ID SEQ ID SEQ ID NO: 342 NO: 347 NO: 343 NO: 348 SEQ ID SEQ ID NO: 344 NO: 349 SEQ ID SEQ ID NO: 345 NO: 350

(419) As an additive strategy to achieve further improvements in affinity, the V.sub.H CDR2 sequences of 33_640166, 33_640169 and 33_640170 were grafted onto the IgG constructs of 33_640076-1, 33_640082-2, 33_640086-2 and 33_640087-2, using standard molecular biology methods. Antibodies resulting from these recombinations were exemplified by 33_640076-4, 33_640082-4, 33_640082-6, 33_640082-7, 33_640086-6 and 33_640087-7. The sequence origins and SEQ ID NO are shown in Table 30.

(420) TABLE-US-00035 TABLE 30 Sequence origin and SEQ ID numbers of IL33 antibodies VH VL VH CDRs VL CDRs IgG1 Origin Sequence Sequence 1, 2, 3 1, 2, 3 33_640076-4 Grafting of the SEQ ID SEQ ID SEQ ID SEQ ID V.sub.H CDR2 NO: 492 NO: 497 NO: 493 NO: 498 sequence of SEQ ID SEQ ID 33_640170 onto NO: 494 NO: 499 33_640076-1 SEQ ID SEQ ID NO: 495 NO: 500 33_640082-4 Grafting of the SEQ ID SEQ ID SEQ ID SEQ ID V.sub.H CDR2 NO: 502 NO: 507 NO: 503 NO: 508 sequence of SEQ ID SEQ ID 33_640166 onto NO: 504 NO: 509 33_640082-2 SEQ ID SEQ ID NO: 505 NO: 510 33_640082-6 Grafting of the SEQ ID SEQ ID SEQ ID SEQ ID V.sub.H CDR2 NO: 512 NO: 517 NO: 513 NO: 518 sequence of SEQ ID SEQ ID 33_640169 onto NO: 514 NO: 519 33_640082-2 SEQ ID SEQ ID NO: 515 NO: 520 33_640082-7 Grafting of the SEQ ID SEQ ID SEQ ID SEQ ID V.sub.H CDR2 NO: 522 NO: 527 NO: 523 NO: 528 sequence of SEQ ID SEQ ID 33_640170 onto NO: 524 NO: 529 33_640082-2 SEQ ID SEQ ID NO: 525 NO: 530 33_640086-6 Grafting of the SEQ ID SEQ ID SEQ ID SEQ ID V.sub.H CDR2 NO: 532 NO: 537 NO: 533 NO: 538 sequence of SEQ ID SEQ ID 33_640169 onto NO: 534 NO: 539 33_640086-2 SEQ ID SEQ ID NO: 535 NO: 540 33_640087-7 Grafting of the SEQ ID SEQ ID SEQ ID SEQ ID V.sub.H CDR2 NO: 542 NO: 547 NO: 543 NO: 548 sequence of SEQ ID SEQ ID 33_640170 onto NO: 544 NO: 549 33_640087-2 SEQ ID SEQ ID NO: 545 NO: 550

(421) The recombination of beneficial V.sub.H CDR3/V.sub.L CDR3 and V.sub.H CDR2 sequences was also carried out using a population cloning and selection approach. The round 3 selection outputs from the block mutagenesis libraries covering the V.sub.H CDR2 region was recombined with the round 3 selection outputs of the recombined V.sub.H CDR3/V.sub.L CDR3 library in a population cloning approach using standard molecular biology techniques. Selection outputs comprising of large numbers of scFv variants were recombined to form libraries in which clones contained randomly paired sequences derived from the V.sub.H CDR3/V.sub.L CDR3 and V.sub.H CDR2 selections. Selections were performed as described for V.sub.HV.sub.L CDR3 libraries in the presence of decreasing concentrations of biotinylated antigen—typically from 3 nM to 3 pM over five rounds of selection. Crude scFv-containing periplasmic extracts from a representative number of individual scFv's from the selection outputs were screened in biochemical HTRF® assays as described for V.sub.HV.sub.L CDR3 libraries. ScFv variants which showed a significantly improved inhibitory effect when compared to parent scFv and leads generated pre-recombination, were subjected to DNA sequencing.

(422) As the epitope competition assay utilising 33_640027 reached its limit of sensitivity, an assay using 33_640117 was used for testing purified scFv samples. The assay was essentially as described for the 33v20064 competition assay with the following modifications: 2.5 nM DyLight labelled 33_640117 was prepared and 2.5 microlitres added to the assay plates. This was followed by the addition of 2.5 microlitres of a solution containing 0.12 nM biotinylated human IL-33-01 combined with 0.75 nM streptavidin cryptate detection (Cisbio International, 610SAKLB) for the human assay or a solution containing 0.24 nM biotinylated cynomolgus IL-33 combined with 1.5 nM streptavidin cryptate detection (Cisbio International, 610SAKLB) for the cynomolgus assay. Fluorescence was read after one hour and overnight incubation. The most potent samples at both time points were taken forward for reformatting to IgG.

(423) Unique recombined variants were evaluated as purified scFv and the most active scFv's were then selected and converted to IgG1 format as described in Example 1. Antibodies obtained from these recombined libraries are exemplified by 33_640201 and 33_640237. The spontaneous mutations which were introduced into the framework regions of 33_640201 and 33_640237 during ribosome display selections were reverted back to germline sequences as described previously in this section, and their germlined counterparts were named 33_640201-2 and 33_640237-2 respectively. The SEQ IDs of these antibodies are shown in Table 31.

(424) TABLE-US-00036 TABLE 31 Sequences of IL33 antibodies VH VL VH CDRs VL CDRs IgG1 Sequence Sequence 1, 2, 3 1, 2, 3 33_640201 SEQ ID SEQ ID SEQ ID SEQ ID NO: 552 NO: 557 NO: 553 NO: 558 SEQ ID SEQ ID NO: 554 NO: 559 SEQ ID SEQ ID NO: 555 NO: 560 33_640237 SEQ ID SEQ ID SEQ ID SEQ ID NO: 562 NO: 567 NO: 563 NO: 568 SEQ ID SEQ ID NO: 564 NO: 569 SEQ ID SEQ ID NO: 565 NO: 570 33_640201-2 SEQ ID SEQ ID SEQ ID SEQ ID NO: 572 NO: 577 NO: 573 NO: 578 SEQ ID SEQ ID NO: 574 NO: 579 SEQ ID SEQ ID NO: 575 NO: 580 33_640237-2 SEQ ID SEQ ID SEQ ID SEQ ID NO: 582 NO: 587 NO: 583 NO: 588 SEQ ID SEQ ID NO: 584 NO: 589 SEQ ID SEQ ID NO: 585 NO: 590

(425) Data for antibodies optimised for V.sub.H CDR3/V.sub.L CDR3 and V.sub.H CDR2 by rational recombination or population approaches are exemplified by 33_640082-6, 33_640087-7, 33_640201 and 33_640237 in FIGS. 44 and 45.

(426) Inhibition of IL-33 Binding to mAb by Purified IgG

(427) The ability of anti-IL-33 antibodies to inhibit the binding of biotinylated His Avi human IL-33 or cynomolgus His Avi IL-33 to the DyLight labelled 33_640117 IgG was assessed in a biochemical HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) competition assays as described above.

(428) FIG. 44A shows inhibition of the FRET signal after 1 hour incubation, produced by biotinylated human IL-33 (IL33-01) binding to DyLight-labeled 33_640117 IgG with increasing concentrations of antibodies 33v20064, 33_640050, 33_640082-6, 33_640087-7, 33_640201 and 33_640237, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(429) FIG. 44B shows inhibition of the FRET signal after overnight incubation, produced by biotinylated human IL-33 (IL33-01) binding to DyLight-labeled 33_640117 IgG with increasing concentrations of antibodies 33v20064, 33_640050, 33_640082-6, 33_640087-7, 33_640201 and 33_640237, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(430) FIG. 44C shows inhibition of the FRET signal after 1 hour incubation, produced by biotinylated cynomolgus His Avi IL-33 binding to DyLight-labeled 33_640117 IgG with increasing concentrations of antibodies 33v20064, 33_640050, 33_640082-6, 33_640087-7, 33_640201 and 33_640237, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(431) TABLE-US-00037 TABLE 32 IC50 values of IgG1 antibodies in 117 epitope competition assays IC50 (nM) 117-WT 117-WT 117-cyno Antibody 1 hour Overnight 1 hour 33v20064 IgG 150 689 788 33_640050 IgG 0.3 1.5 4.1 33_640076-4 IgG 0.1 0.1 0.6 33-640081 IgG 0.3 0.6 10 33_640082-6 IgG 0.1 0.1 0.2 33_640082-7 IgG 0.1 0.1 0.2 33_640084-2 IgG 0.3 0.3 0.8 33_640086-6 IgG 0.1 0.1 0.2 33_640087-7 IgG 0.1 0.1 0.2 33_640201 IgG 0.2 0.2 0.3 33_640237 IgG 0.1 0.1 1.8
Inhibition of IL-8 Production in Huvec by IgG

(432) IgGs were tested in a Huvec IL-8 production assay. Cells were exposed to N-terminal His Avi IL-33 (IL33-01, SEQ ID NO 632) or mammalian full length IL-33 cell lysate (FL-IL33 lysate) in the presence or absence of test antibody as previously described. IC.sub.50 values were calculated and are summarized in Table 33 below.

(433) FIG. 45A shows HUVECs stimulated with IL33-01 in the presence of test antibodies 33_640050, 33_640082-6, 33_640087-7, 33_640201 and 33_640237, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is a percentage of the maximum response (IL-8 production).

(434) FIG. 45B shows HUVECs stimulated with full length IL-33 cell lysate in the presence of test antibodies 33_640050, 33_640082-6, 33_640087-7, 33_640201 and 33_640237, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is a percentage of the maximum response (IL-8 production).

(435) TABLE-US-00038 TABLE 33 IC50 values in the Huvec IL-8 assay IC50 (nM) vs. full length IL-33 Antibody vs. His Avi IL-33 cell lysate 33_640076 -4 0.032 0.095 33_640082-6 0.030 0.097 33_640082-7 0.031 0.095 33_640084-2 0.036 0.132 33_640086-6 0.026 0.081 33_640087-7 0.025 0.073 33_640201 0.046 0.101 33_640237 0.056 0.091
Germlining IGLJ Sequence

(436) The amino acid sequences of the V.sub.L framework regions of antibodies 33_640076-4, 33_640081-A, 33_640082-6, 33_640082-7, 33_640084-2, 33_640086-6, 33_640087-7, 33_640201-2 and 33_640237-2 were aligned to known human IGLJ germline sequences in the IMGT database (Lefranc, M. P. et al. Nucl. Acids Res. 2009. 37(Database issue): D1006-D1012), and the closest germline was identified by sequence similarity. For all of these antibodies this was IGLJ2, which has a single amino acid difference to the antibodies at position 104 of the V.sub.L region (Kabat numbering). This residue was reverted to germline as described in Example 3 using standard molecular biology methods. The resulting antibodies were named 33_640076-4B, 33_640081-AB, 33_640082-6B, 33_640082-7B, 33_640084-2B, 33_640086-6B, 33_640087-7B, 33_640201-2B and 33_640237-2B corresponding to their parental lineage of 33_640076-4, 33_640081-A, 33_640082-6, 33_640082-7, 33_640084-2, 33_640086-6, 33_640087-7, 33_640201-2 and 33_640237-2 respectively. The SEQ IDs of the V.sub.H and V.sub.L regions of these antibodies are shown in Table 34.

(437) TABLE-US-00039 TABLE 34 Sequences of germlined IL33 antibodies IgG1 VH Sequence VL Sequence 33_640076-4B SEQ ID NO: 592 SEQ ID NO: 594 33_640081-AB SEQ ID NO: 596 SEQ ID NO: 598 33_640082-6B SEQ ID NO: 600 SEQ ID NO: 602 33_640082-7B SEQ ID NO: 604 SEQ ID NO: 606 33_640084-2B SEQ ID NO: 608 SEQ ID NO: 610 33_640086-6B SEQ ID NO: 612 SEQ ID NO: 614 33_640087-7B SEQ ID NO: 616 SEQ ID NO: 618 33_640201-2B SEQ ID NO: 620 SEQ ID NO: 622 33_640237-2B SEQ ID NO: 624 SEQ ID NO: 626

Example 9 In Vivo Airway Inflammation Model

(438) Cloning, Expression and Purification of IL-33 Cytokine Trap

(439) Protein sequences for mouse IL-1RAcP and mouse ST2 were obtained from Swiss Prot (accession numbers Q61730 and P14719 respectively). Mouse IL-33 cytokine trap was designed based on Economides et al 2003 and consisted of amino acids 1-359 Q61730 and amino acids 27-332 P14719 fused to the Fc portion of human IgG1. Protein sequences were codon optimised (Geneart) and cloned into pDEST12.2 OriP proteins were secreted from cells into the media utilizing the native signal peptides from IL-1RAcP. For expression in CHO cells, the gateway linker was removed by overlapping primer PCR. Trap expression vector was transfected into CHO-transient mammalian cells. Mouse IL-33 trap was expressed and secreted into the medium. Harvests were pooled and filtered prior to purification using Protein A chromatography. Culture supernatants were loaded onto a 5 ml Hitrap Protein A column (GE Healthcare) and washed with 1×DPBS, bound Trap was eluted from the column using 0.1 M Sodium Citrate (pH 3.0) and neutralized by the addition of Tris-HCl (pH 9.0). The eluted material was further purified by SEC in 1×DPBS using a 5200 16:600 Superdex column (GE healthcare) and the concentration determined spectrophotometrically using an extinction coefficient based on the amino acid sequence (Mach et al., Anal. Biochem. 200(1):74-80 (1992)).

(440) Humanized IL-33 Mice

(441) Methods for generation of humanized IL-33 mice have been previously described in Example 4. The humanized mice are used in models of airways and/or allergic inflammation to assess the effect of anti-human IL-33 antibodies.

(442) In Vivo Airway Inflammation Model

(443) Models of Alternaria alternata (ALT) induced airway inflammation in mice have been previously described (Kouzaki et al. J. Immunol. 2011, 186: 4375-4387; Bartemes et al J Immunol, 2012, 188: 1503-1513). Endogenous IL-33 is released rapidly following ALT exposure and drives IL-33-dependent IL-5 production and eosinphilia in the lung. Male or female wildtype or humanized IL-33 mice (6-10 weeks) were anaesthetized briefly with isofluorane and administered either 25 μg of ALT extract (Greer, Lenoir, N.C.) or vehicle intranasally in a total volume of 50 μl. Mice were treated intraperitoneally or intranasally with test substances: IL330004 IgG (SEQ ID Nos. 12 and 17), H338L293 IgG (SEQ ID Nos. 182 and 187), mouse IL-33 Trap, 33_640050 (SEQ ID nos. 302 and 307), isotype control IgG (NIP228) or vehicle (PBS, 10 ml/kg) at 24 hours prior (for intraperitoneal treatment) or 2 hours prior (for intranasal treatment) to intranasal challenge with ALT. At 24 hours after challenge, mice were terminally anaesthetised with pentobarbital sodium prior to exsanguination and bronchoalveolar lavage (BAL). Bronchoalveolar lavage fluid (BALF) was collected by lavage (0.3 ml, 0.3 ml & 0.4 ml) via tracheal cannula. BALF was centrifuged, cells counted (total cells by FACS (FacsCALIBER, BD)) and supernatant was analysed for cytokines by ELISA (Meso Scale Discovery, Rockville, Md.). Differential cell counts (200 cells/slide) were performed on cytospin preparations stained with Diff-Quik (Fisher Scientific, UK). All work was carried out to UK Home Office ethical and husbandry standards under the authority of an appropriate project license.

(444) FIG. 46 shows that H338L293 dose-dependently inhibits ALT-induced BAL IL-5 and eosinophilia in wild type BALB/c mice. Test substances were dosed intranasally (10, 30 or 100 mg/kg as indicated in brackets) at −2 hours prior to challenge with 25 ug of ALT. BALF was harvested at 24 hours post ALT challenge and analysed for presence of IL-5 (FIG. 46A) and eosinophils (FIG. 46B). Significant effect of test substances was determined using one-way ANOVA with Bonferroni's multiple comparisons test. ***p<0.001, ˜˜p<0.01 compared to control mAb (n=4-8). Mouse IL-33 Trap was used as a positive control.

(445) FIG. 47 shows that H338L293 (30 mg/kg) and mouse IL-33 Trap (10 mg/kg), but not IL330004 (30 mg/kg), inhibit ALT-induced BAL IL-5 in humanized IL-33 mice. Test substances were dosed intranasally at −2 hours prior to challenge with 25 ug of ALT. BALF was harvested at 24 hours post ALT challenge and analysed for presence of IL-5. Significant effect of test substances was determined using one-way ANOVA with Bonferroni's multiple comparisons test. ***p<0.001, **p<0.01 (n=4).

(446) FIG. 48 shows that 33_640050 dose dependently inhibits Alternaria-induced BAL IL-5 in humanized IL-33 mice. Test substances were dosed intraperitoneally (0.3, 3 or 30 mg/kg as indicated in brackets) at −24 hours prior to challenge with 25 ug of Alternaria. BALF was harvested at 24 hours post ALT challenge and analysed for presence of IL-5. Significant effect of test substances was determined using one-way ANOVA with Bonferroni's multiple comparisons test. ***p<0.001, **p<0.01 (n=4-5).

Example 10 Characterization of Anti-IL-33 Antibodies

(447) Inhibition of IL-33 Binding to ST2 by Purified IgG

(448) The ability of anti-IL-33 antibodies to inhibit the binding of biotinylated IL33-01 to the FLAG®-His tagged ST2 receptor was assessed in a biochemical HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) competition assay, the principles of which are described above. Activity of purified IgG preparations were determined by competing a dilution series of the purified IgG against human FLAG®-His tagged ST2 for binding to human biotinylated human IL33-01 (SEQ ID No. 632).

(449) FIG. 49A: shows the inhibition of the FRET signal after overnight incubation, produced by human IL-33 binding to human ST2 with increasing concentrations of antibodies 33v20064, 33_640087-7, 33_640087-7B, 33_640050 and 33_640237-2B, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is percent specific binding.

(450) Inhibition of IL-8 Production in Huvec by IgG

(451) IgGs were tested in a Huvec IL-8 production assay. Cells were exposed to N-terminal His Avi IL-33 (IL33-01, SEQ ID NO 632) in the presence or absence of test antibody as previously described. IC.sub.50 values were calculated and are summarized in Table 35 below. Data shows that germlining of the IGLJ Sequence did not have any effect on antibody potency.

(452) FIG. 49B shows HUVECs stimulated with IL33-01 in the presence of test antibodies 33_640087-7, 33_640087-7B, 33_640237-2 and 33_640237-2B, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is a percentage of the maximum response (IL-8 production).

(453) TABLE-US-00040 TABLE 35 IC50 values in the Huvec IL-8 assay Antibody IC50 (nM) vs. His Avi IL-33 33_640087-7 0.041 33_640087-7B 0.046 33_640237-2 0.105 33_640237-2B 0.067
Selectivity and Cross-Reactivity of Anti-IL-33 Antibodies

(454) Selectivity and cross-reactivity of germlined anti-IL-33 antibodies was determined using a homogeneous FRET (fluorescence resonance energy transfer) HTRF® (Homogeneous Time-Resolved Fluorescence, Cisbio International) based IL-33:mAb-binding assay. In this assay, samples competed with biotinylated human IL-33-01 (SEQ ID No. 632) for binding to DyLight labelled 33_640087-7B IgG (SEQ ID Nos. 618 and 618) or 33_640237-2B IgG (SEQ ID Nos. 624 and 626).

(455) Human, cyno and mouse IL-33 FLAG® His (described in Example 1 and Table 21), Human IL-1 alpha and IL-1 beta (R&D Systems) (Table 21) or rat IL-33 (GenScript) were tested for inhibition of human IL-33 binding to DyLight650 labelled 33_640087-7B or DyLight650 labelled 33_640237-2B by adding 5 microlitres of each dilution of sample to a 384 well low volume assay plate (Costar, 3673). Next, a solution containing 1.2 nM DyLight650 labelled 33_640087-7B or 33_640237-2B was prepared and 2.5 microlitres added to the assay plate (labelled using kit (Innova Biosciences, 326-0010) as per manufacturer's instructions). This was followed by the addition of 2.5 microlitres of a solution containing 0.12 nM biotinylated human IL-33-01 combined with 0.75 nM streptavidin cryptate detection (Cisbio International, 610SAKLB). All dilutions were performed in assay buffer containing 0.8 M potassium fluoride (VWR, 26820.236) and 0.1% bovine serum albumin (BSA, PAA, K05-013) in Dulbeccos PBS (Invitrogen, 14190185). Assay plates were incubated for 4 hours at room temperature followed by 18 hours at 4 degrees Celsius and time resolved fluorescence was read at 620 nm and 665 nm emission wavelengths using an EnVision plate reader (Perkin Elmer). Data were analysed by calculating the 665/620 nm ratio followed by the % Delta F values for each sample. The 665/620 nm ratio was used to correct for sample interference using Equation 1. The % Delta F for each sample was then calculated using Equation 2. The negative control (non-specific binding) was defined by replacing biotinylated human IL-33 combined with streptavidin cryptate detection with streptavidin cryptate detection only. The % Delta F values were subsequently used to calculate % specific binding as described in Equation 3. IC.sub.50 values were determined using GraphPad Prism software by curve fitting using a four-parameter logistic equation (Equation 4). These results demonstrated that 33_640087-7B and 33_640237-2B cross react with cynomolgus IL-33 but not mouse IL-33, rat IL-33, human IL-1 alpha or human IL-1 beta.

(456) FIG. 50A: shows the inhibition of the FRET signal, produced by biotinylated human IL-33-01 binding to DyLight labelled 33_640087-7B (SEQ ID Nos. 618 and 618), with increasing concentrations of test proteins, wherein the x-axis is the concentration of test sample in molar concentration and the y-axis is percent specific binding. Inhibition of the FRET signal was observed with human and cynomolgus, but not mouse or rat IL-33, human IL-1 alpha or human IL-1 beta.

(457) FIG. 50B: shows the inhibition of the FRET signal, produced by biotinylated human IL-33-01 binding to DyLight labelled 33_640237-2B (SEQ ID Nos. 624 and 626), with increasing concentrations of test proteins, wherein the x-axis is the concentration of test sample in molar concentration and the y-axis is percent specific binding. Inhibition of the FRET signal was observed with human and cynomolgus, but not mouse or rat IL-33, human IL-1 alpha or human IL-1 beta.

(458) Neutralisation of Endogenous IL-33 in the HUVEC IL-8 Assay

(459) In order to determine whether antibodies were able to neutralize endogenous IL-33, human lung tissue was used to provide a source of endogenous IL-33 protein. The study was approved by the NRES East of England (Cambridge East) Research Ethics Committee (reference number 08/H0304/505) and tissue was donated with the informed consent of patients. Non-cancerous adjacent tissue from lung cancer patients and from lung transplant surgeries were supplied in Aqix RS-I medium (Aqix Ltd) on ice by Papworth Hospital NHS Trust Research Tissue Bank. Tissue was diluted with 400 mg/mL in PBS and homogenized for 30 seconds using a tissue homogenizer. Cell debris was removed by centrifugation. HUVECs were stimulated with lung lysates at varying concentrations. An EC50 concentration of lysate that stimulated IL-8 release was selected for antibody neutralization studies. Cells were exposed to lung lysate in the presence or absence of test antibody as previously described. sST2 inhibited the IL-8 response by a maximum of approximately 70%, suggesting that most, but not all, of the IL-8 production was driven by endogenous IL-33 within the lung lysate. 33_640050 and 33_640087-7B IgG inhibited the IL-8 response to a similar extent as sST2, demonstrating their ability to bind and neutralize endogenous IL-33. 33_640050 IgG neutralized the lung lysate with an IC50 of 0.032 nM. 33_640087-7B neutralized the lung lysate with an IC.sub.50 of 0.013 nM. sST2 neutralized the lung lysate with an IC50 of 0.019 nM

(460) FIG. 51 shows HUVECs stimulated with human lung lysate in the presence of test antibodies 33_640050 and 33_640087-7B in comparison with sST2, wherein the x-axis is the concentration of antibody in molar concentration and the y-axis is a percentage of the maximum response (IL-8 production). sST2 inhibited the IL-8 response by a maximum of approximately 70%, suggesting that most, but not all, of the IL-8 production was driven by endogenous IL-33 within the lung lysate. Both antibodies inhibited the IL-8 response to a similar extent as sST2, demonstrating their ability to bind and neutralize endogenous IL-33.

(461) In Vivo Airway Inflammation Model

(462) Methods for generation of humanized IL-33 mice have been previously described in Example 4. Humanized mice were used in a model of Alternaria alternata (ALT) induced airway inflammation as described in Example 9 to assess the effect of 33_640087-7B. Male or female wild type or humanized IL-33 mice (6-10 weeks) were anaesthetized briefly with isofluorane and administered either 25 μg of ALT extract (Greer, Lenoir, N.C.) or vehicle intranasally in a total volume of 50 μl. Mice were treated intraperitoneally with test substances: 33_640087-7B IgG (SEQ ID Nos. 618 and 618), isotype control IgG (NIP228) or vehicle (PBS, 10 ml/kg) at 24 hours prior to intranasal challenge with ALT. At 24 hours after challenge, mice were terminally anaesthetised with pentobarbital sodium prior to exsanguination and bronchoalveolar lavage (BAL). Bronchoalveolar lavage fluid (BALF) was collected by lavage (0.3 ml, 0.3 ml & 0.4 ml) via tracheal cannula. BALF was centrifuged and supernatant was analysed for cytokines by ELISA (Meso Scale Discovery, Rockville, Md.). All work was carried out to UK Home Office ethical and husbandry standards under the authority of an appropriate project license.

(463) FIG. 52 shows that 33_640087-7B dose dependently inhibits Alternaria-induced BAL IL-5 in humanized IL-33 mice. Test substances were dosed intraperitoneally (0.1, 1, 3 or 10 mg/kg as indicated in brackets) at −24 hours prior to challenge with 25 ug of Alternaria. BALF was harvested at 24 hours post ALT challenge and analysed for presence of IL-5. Significant effect of test substances was determined using one-way ANOVA with Bonferroni's multiple comparisons test. ***p<0.001, **p<0.01 (n=5-6).

Example 11 Affinity of Anti-IL-33 Antibodies

(464) The affinity of the anti-IL-33 antibody fragment (Fab) for recombinant human IL33 was determined using real time interaction monitoring by BIACORE™ and at equilibrium using KinExA™ for 33_640087-7B. For both methodologies the human IL33 protein was purified by SEC-HPLC to ensure quality of the antigen and also of the Fab for the biacore analysis.

(465) Biacore Affinity Analysis

(466) Fab fragments were generated by papain cleavage from full length IgG1 and purified by SEC. The affinity of the antibody fragment (Fab) was measured using the Biacore T100 at 25° C. Streptavidin was covalently immobilised to a C1 chip surface using standard amine coupling techniques at a concentration of 4 m/ml in 10 mM Sodium acetate pH 4.5. Typical final streptavidin surface densities in the range 115-170 RUs were achieved. Recombinant, enzymatically biotinylated human IL-33 (produced in-house) was titrated onto the streptavidin chip surface at 4 m/ml in HBS-EP+ buffer to enable ˜30 RUs of Fab binding at saturation (Rmax). This low level of analyte binding ensured minimal mass transport effects.

(467) The IL-33 Fab was serially diluted from 5 nM to 78 pM in HBS-EP+ buffer and flowed over the chip at 50 μl/min, with 3 minutes association and up to 30 minutes dissociation. Multiple buffer only injections were made under the same conditions to allow for double reference subtraction of the final sensorgram sets, which were analysed using the BiaEval software (version 2.0.1). The chip surface was fully regenerated with pulses of 3M MgCl.sub.2.

(468) The affinity of ST2-Flag-His10 (SEQ. ID no. 650) expressed in HEK-EBNA cells for human IL-33 was determined by BIACORE™ using the same methods described above.

(469) TABLE-US-00041 TABLE 36 Biacore Affinity Results for anti-IL-33 Fab k.sub.a k.sub.d K.sub.D Analyte Fit setting (M.sup.−1 s.sup.−1) (s.sup.−1) (pM) R.sub.max Chi.sup.2 33_640001 Fab Rmax global 9.76E+6 3.26E−2 3340 27.0 0.022 33_640050 Fab Rmax local 3.08E+7 1.05E−4 3.4 48.9 0.022 33_640087-7B Fab Rmax local 2.20E+7 9.42E−6 0.43 30.7-32.8 0.010 ST2-FH monomer Rmax local 1.52E+7 4.35E−5 2.9 23.4 0.022
KinExA Affinity Analysis

(470) In order to confirm the high affinity found with the SPR assay we turned to Kinetic Exclusion Assays (KinExA). KinExA is increasingly finding favour for resolving higher-affinity protein:protein interactions, especially those in the pM to sub-pM ranges where surface based biosensor techniques reach their practical limits (Rathanaswami P, Roalstad S, Roskos L, Qiaojuan J S, Lackie S, Babcook J. Demonstration of an in vivo generated sub picomolar affinity fully human monoclonal antibody to interleukin-8. Biochemical and Biophysical Research Communications. 2005; 334: 1004-1013).

(471) The affinity of the antibody 33_640087-7B was measured by kinetic exclusion assays performed on the KinExA 3200. The sampling beads were prepared by mixing 200 mg of dry UltraLink Biosupport Azlactone beads with 110 μg of IL-33 (as mentioned previously) in 2.5 ml of 50 mM sodium hydrogen bicarbonate pH 8.4 at room temperature for 2 hours with constant agitation. The beads were rinsed and blocked with 10 mg/ml BSA in 1M Tris pH 8.7. Prior to use, the beads were resuspended into D-PBS, 0.02% sodium azide. 33_640087-7B/IL-33 equilibrium mixtures were prepared in sample buffer composed of 1 mg/ml BSA, 0.02% sodium azide in DPBS (Dulbecco's PBS). Two different IgG concentrations were used with varying IL-33 concentrations, 5 pM of 33_640087-7B with IL-33 serially diluted from 125 pM to 61 fM and 500 fM 33_640087-7B with IL-33 serially diluted from 62.5 pM to 15 fM, both were carried out with zero IL-33 controls. The fluorescent secondary detection reagent was Alexa Fluor 647 goat anti-human-Fc diluted in 1 mg/ml BSA, 0.02% sodium azide, 0.1% Tween 20 in DPBS. The samples were run on the KinExA whilst housed in a temperature controlled cabinet set at 25° C. The data was analysed using the KinExA Pro software version 4.1.11.

(472) KinExA assays indicate that 33_640087-7B has a K.sub.D of <142 fM (femtomolar) for human IL-33 (Table 37).

(473) TABLE-US-00042 TABLE 37 KinExA Affinity Results for 33_640087-7B IgG Upper 95% Lower 95% Estimated confidence interval confidence interval Analyte K.sub.D (pM) K.sub.D (pM) K.sub.D (pM) 33_640087-7B 0.03 0.142 undefined

Example 12 Activity of Oxidized IL-33

(474) In Vivo Pilot Study to Explore Activity of Oxidized IL-33

(475) In Example 4 (see also Cohen, E. S. et al. Oxidation of the alarmin IL-33 regulates ST2-dependent inflammation. Nat. Commun. 6:8327 doi: 10.1038/ncomms9327 (2015)) we describe the discovery of an oxidized, disulphide bonded form of IL-33 (DSB IL-33) and showed that this form does not bind ST2. To investigate if oxidized IL-33 had an alternative activity independent of ST2, ST2-deficient mice were treated intraperitoneally or intranasally with repeated doses of human IL-33 for 2, 4 or 6 weeks. Histological analysis was performed on multiple tissues.

(476) ST2-deficient mice were generated as previously described (Townsend, M. J., Fallon, P. G., Matthews, D. J., John, H. E., and McKenzie, A. N. J. (2000). T1/ST2-deficient mice demonstrate the importance of T1/ST2 in developing primary T helper cell type 2 responses. J. Exp. Med. 191, 1069-1076). Female ST2-deficient mice (12 weeks) were anaesthetized briefly with isofluorane and administered either 10 μg of N-terminal His Avi IL-33 (IL33-01, SEQ ID NO 632; lot number CCH168, endotoxin levels 0.03 EU/mg), or vehicle (PBS) intranasally in a total volume of 50 μl. Alternatively, the IL-33 or vehicle was administered by i.p. injection. This process was repeated 3×weekly for a total 18 treatments. Mice that received only 2 or 4 weeks treatment with IL-33 were administered PBS for the first 4 or 2 weeks of dosing, prior to receiving IL-33.

(477) At 24 hours after the last treatment, mice were terminally anaesthetised with pentobarbital sodium prior to exsanguination and bronchoalveolar lavage (BAL). Blood was collected by cardiac bleed using an EDTA flushed syringe. Blood haematology was done using Sysmex XTVet haematology analyser. The remaining blood was centrifuged and plasma extracted. Bronchoalveolar lavage fluid (BALF) was collected by lavage (0.3 ml, 0.3 ml & 0.4 ml) via tracheal cannula. BAL cells were counted (BAL total cells by flowcytometer (MACSquant, Miltenye Biotec) and BALF was centrifuged to separate supernatant, that was analysed for cytokines by ELISA (Meso Scale Discovery, Rockville, Md.). Differential cell counts (200 cells/slide) were performed on cytospin preparations stained with Diff-Quik (Fisher Scientific, UK). Following PBS lavage lungs were inflated with 10% neutral buffered formalin (NBF) via a tracheal infusion to maintain lung architecture and immerse-fixed in NBF for 24-48 hours. Fixed lung samples were then cut transversely into 4 equal cross-sections before being processed through a series of alcohols, xylene and into paraffin wax. Finally the lung cross-sections were then embedded into paraffin wax blocks. 4 μm histological sections were cut and stained with haematoxylin and eosin (H&E) for analysis and inflammation scoring assessment. All work was carried out to UK Home Office ethical and husbandry standards under the authority of an appropriate project license.

(478) To investigate IL-33 exposure via intraperitoneal or intranasal routes, human IL-33 was measured in BALF and plasma using Millipore human IL-33 assay (Cat #HTH17MAG-14K lot 2159117) as described in Example 4. Ability of sST2-Fc to reduce the assay signal was used to determine the presence of ST2-binding (reduced) vs non-ST2-binding (oxidized) IL-33. Following intranasal administrations, IL-33 was consistently detected in BALF and plasma at 2, 4 and 6 week endpoints. The majority of the IL-33 detected was oxidised. Following a single intraperitoneal administration, IL-33 was detected transiently in the plasma 5 hours after dosing but was undetectable by 24 hours. Following repeated IL-33 administrations, human IL-33 was not consistently detected in BALF or plasma at the 2, 4 or 6 week endpoints. These data indicated that the best systemic exposure of oxidized IL-33 is achieved through intranasal dosing.

(479) FIG. 53A. Experimental design of in vivo pilot study. ST2-deficient mice were treated intraperitoneally or intranasally with repeated administration of human IL-33 or vehicle (PBS) for 6 weeks (n=3-4 per group). Tissues, BALF and serum were collected 24 hours following the final IL-33 administration.

(480) FIG. 53B. Analysis of human IL-33 exposure in BAL fluid following repeated administration of human IL-33 to BALB/c mice. Human IL-33 was measured in BALF from mice treated as described in FIG. 53A, wherein the x-axis shows the treatment group and the y-axis shows the human IL-33 assay signal in arbitrary units. IL-33 was only detected in BALF after intra-nasal dosing. The IL-33 detected was found to be predominantly oxidized (non-ST2 binding).

(481) FIG. 53C Analysis of IL-33 exposure in plasma following a single intraperitoneal administration of human IL-33 (10 ug). Human IL-33 was measured in plasma at 2, 5, 24 and 48 hours after administration, wherein the x-axis shows the time point of analysis and the y-axis shows the human IL-33 assay signal in arbitrary units. IL-33 was detected transiently in the plasma 5 hours after dosing but was undetectable by 24-48 hours.

(482) FIG. 53D Analysis of IL-33 exposure in plasma following repeated administration of human IL-33 to BALB/c mice. Human IL-33 was measured in plasma from mice treated as described in FIG. 53A, wherein the x-axis shows the treatment group and the y-axis shows the human IL-33 assay signal in arbitrary units. IL-33 was only detected in plasma after intra-nasal dosing. The IL-33 detected was found to be predominantly oxidized (non-ST2 binding).

(483) Histological analysis was performed on multiple tissues. No relevant abnormalities in human IL-33 treated mice compared to controls in liver, brain, spleen, skin, stomach, lymph node or heart. In the lung, increased lymphocytic perivascular inflammation was present in IL-33 treated mice compared to controls only in the intranasal group and only after 6 weeks of treatment. This is consistent with the highest exposure to oxidized IL-33 (FIG. 53). In conclusion, the treatment of ST2 KO mice with IL-33 increases the presence of perivascular lymphocytic infiltrate in lungs of IL-33 treated mice compared to controls. This pathology may be mediated through oxidized IL-33.

(484) FIG. 54A shows representative H&E stained paraffin sections of lung tissue from mice administered PBS intranasally for 6 weeks (n=3)

(485) FIG. 54B shows representative H&E stained paraffin sections of lung tissue from mice administered IL-33 intranasally for 6 weeks (n=4).

(486) Pathway Analysis of IL-33 Treated Mouse Lung

(487) To gain insight into the pathways modulated by oxidized IL-33 in the mouse lung leading to the inflammatory response observed, microarray analysis was performed on lung tissue from 6 week PBS treated versus 6 week IL-33-treated animals.

(488) Lung tissues from 7 ST2KO mice (3 dosed with PBS and 4 dosed with IL33-01 as described above) were collected and directly placed into 350 uL of RLT buffer (Qiagen #79216). Tissue was then disrupted using a Qiagen TissueLyser (Qiagen #85300) according to manufacturer's protocol and RNA was purified using the RNeasy Fibrous Tissue kit (Qiagen #74704). Purified RNA from this kit was then concentrated using RNeasy Micro Columns (Qiagen #74004) according to manufacturer's protocol. RNA was then amplified to single stranded DNA using Affymetrix's GeneChip WT Plus Reagent kit (Affymetrix #902513) and hybridised onto Mouse Transcriptome 1.0 (MTA1.0) genechips (Affymetrix #900720), washed in Affymetrix Fluidics Station and scanned on the Affymetrix Genechip Scanner 3000 7G. Data was then processed in Affymetrix Expression console. Data were analysed in Microsoft Excel and sorted for genes where at least 2 out of 4 IL33-treated mice had greater than ±1.2-fold change in signal from the control group average. The sorted gene list were converted to KEGG IDs using Biological Database Network (BioDB net v2.1) and analysed in KEGG pathway analysis (www.kegg.jp; KEGG Mapper v2.5). The same genes that were analysed in KEGG pathway were also analysed using Ingenuity Pathway Analysis (IPA) (Qiagen). IPA analysis suggested pathways relating to cell cycle appeared modulated. Example genes are listed in Table 38.

(489) TABLE-US-00043 TABLE 38 Genes modulated in ST2-deficient mice by intranasal IL-33 treatment Gene Symbols Upregulated genes relating to cell cycle HSPA1A, CEBP, CDKN1A, (>1.2 average fold change) JUNB, AHCY, SOX2, MYC, ID2, HRAS, EGR1, WT1, JUND, COX4I1 Downregulated genes relating to cell cycle IL1A, HGF, ERG, ZEB1, (>−1.2 average fold change) P10, DMTF1, ARHGAP18, PLA2R1, NSMCE2, CAV1
Signalling of DSB IL-33 in Huvecs

(490) To ascertain whether a response to oxidized (DSB) human IL-33 could be observed on human cells, stimulation of human cells in vitro was explored. The mouse microarray analysis indicated activation of pathways relating to cell cycle and therefore p38 MAP Kinase and JAK-STAT signaling were investigated. Human umbilical vein endothelial cells (Huvecs) were cultured according to manufacturer's instructions and stimulated with reduced or DSB IL-33. Nuclear translocation of p-p38 MAPK or p-STAT5 was detected by immunofluorecence staining. Imaging and quantification of the nuclear staining intensity was performed on ArrayScan VTi HCS Reader (Cellomics). The assay was essentially the same as that described for NFkB p65/RelA nuclear translocation but with the following modifications.

(491) For p-p38 MAPK assay, Huvecs were seeded at 1×10.sup.4/75 μl well in culture media [EBM-2 (Lonza, #CC-3156) with EGM-2 SingleQuot Kit Suppl. & Growth Factors (Lonza, #CC-4176)] into 96-well black walled, clear flat-bottomed Collagen I coated plates (Greiner #655956) and incubated at 37° C., 5% CO.sub.2 for 18-24 hours. Test samples of reduced or DSB IL-33 (in duplicate) were diluted to the desired concentration in complete culture media in 96 well U-bottom polypropylene plates (Greiner, 650201) and 75 uL added to the Huvec plates to initiate the stimulation. Following 15 or 30 minute assay incubation at 37° C., cells were fixed for 15 minutes in 3.7% formaldehyde solution (by addition of 50 uL of 16% solution that had been pre-warmed to 37° C.). Fixative was aspirated and cells were washed twice with 100 μL/well of PBS. Cells were stained for p-p38 with Phospho-p38 antibody (Cell signalling #9211S) at 1:250 dilution. Briefly, cells were permeabilised for 15 minutes at room temperature, blocked for 15 minutes and stained for 1 hour with primary antibody solution in a volume of 50 μL. Plates were washed×2 in blocking buffer and stained for 1 hour at room temperature with secondary antibody solution (DyLight 488-labelled goat anti-rabbit IgG; ThermoFisher Scientific #35552 at 1:400 dilution) and Hoechst nuclear stain (ThermoFisher Scientific #62249 at 1:10000 dilution). Plates were washed ×2 in PBS. Cells were stored in a final volume of 150 μL/well PBS and covered with a black, light-blocking seal (Perkin Elmer, #6005189) before reading on ArrayScan VTi HCS Reader. The intensity of nuclear staining was calculated using a suitable algorithm. Data were analysed using Graphpad Prism software.

(492) For pSTAT5 assay, Huvecs were seeded at 1×10.sup.4/750 μl/well in culture media [EBM-2 (Lonza, #CC-3156) with EGM-2 SingleQuot Kit Suppl. & Growth Factors (Lonza, #CC-4176)[into 96-well black walled, clear flat-bottomed Collagen I coated plates (Greiner, #655956) and incubated at 37° C., 5% CO.sub.2 for 18-24 hours. Following this complete media was aspirated, the cells were washed 2× in 100 uL PBS/well, the PBS was aspirated and 75 uL of Starve media [EBM-2 (Lonza, #CC-3156) with pen/strep] was added to each well. Cells were then were incubated at at 37° C., 5% CO.sub.2 for 18 hours. Test samples of reduced or DSB IL-33 (in duplicate) were diluted to the desired concentration in starve media in 96 well U-bottom polypropylene plates (Greiner, 650201) and 75 uL added to the Huvec plates to initiate the stimulation. Following 15 or 30 minute assay incubation at 37° C., cells were fixed for 15 minutes in 3.7% formaldehyde solution (by addition of 50 uL of 16% solution that had been pre-warmed to 37° C.). Fixative was aspirated and cells were washed twice with 100 μL/well of PBS. Cells were stained for p-STAT5 with Phospho-STAT5 rabbit antibody C71E5 (Cell Signalling #9314S) at 1:250 dilution, which was detected as described above.

(493) Reduced IL-33 triggered p-p38 MAPK signaling, which was lost upon oxidation to the DSB form (FIG. 55A), similar to that previously described for NFkB signaling (Examples 4-6). However DSB IL-33, but not reduced IL-33, triggered p-STAT5 signalling (FIG. 55B). Thus a clear switch in signaling pathway activation was observed upon conversion of human IL-33 from the reduced to the DSB form, suggesting that DSB IL-33 may have activity distinct of known IL-33 pathways.

(494) To confirm the result from the nuclear translocation assays, IL-33 signalling was determined by Western blot analysis. Huvecs were stimulated as above with reduced or DSB IL-33 (3 ng/mL) for 15 minutes. Cells were then washed twice in ice cold PBS and lysed with 250 uL RIPA Buffer (Pierce #89901) containing HALT protease inhibitors (Pierce #78430). Samples were subjected to SDS-PAGE under reducing conditions. Samples were mixed 3:1 with 4× NuPAGE gel loading buffer (Invitrogen) and denatured at 90° C. for 3 minutes. Reduced samples contained 2% beta-mercaptoethanol. Samples were run on NuPAGE Novex 4-12% Bis-Tris mini gels (Invitrogen) with MOPS running buffer (Invitrogen) according to manufacturer's instructions. Proteins were transferred to Nitrocellulose membranes (Invitrogen cat. no. IB3010-02) and detected by western blotting with rabbit phospho-p38 MAPK antibody (Cell signalling #9211S), rabbit phospho-STAT5 antibody C71E5 (Cell Signalling #9314S) or rabbit p-JAK2 antibody (Cell Signalling #3771S). Primary antibodies were detected with anti-rabbit-HRP (Cell Signalling #7074) and visualized using ECL reagent (Thermo Scientific #34096).

(495) FIG. 55A shows p-p38 MAPK nuclear translocation activity in Huvecs in response to reduced IL-33 or DSB IL-33 (IL33-01 pre-treated with IMDM media), wherein the x-axis shows the IL-33 concentration and the y-axis shows the nuclear translocation signal in arbitrary units. A concentration dependent signal was observed for reduced but not oxidized IL-33.

(496) FIG. 55B shows p-STAT5 nuclear translocation in Huvecs in response to reduced IL-33 or DSB IL-33 (IL33-01 pre-treated with IMDM media), wherein the x-axis shows the IL-33 concentration and the y-axis shows the nuclear translocation signal in arbitrary units. A concentration dependent signal was observed for DSB but not reduced IL-33.

(497) FIG. 55C. Western blot analysis of for p-p38 MAPK, p-JAK2 and p-STAT5 in Huvecs stimulated for 15 minutes with reduced IL-33 or DSB IL-33 (IL33-01 pre-treated with IMDM media). Activation of p-p38 MAPK was detected following stimulation with reduced but not DSB IL-33. Activation of p-JAK2 and p-STAT5 was detected following stimulation with DSB IL-33 but not reduced IL-33.

(498) Signalling of DSB IL-33 is Mediated Via the Receptor for Advanced Glycation End Products (RAGE)

(499) To gain insight into the pathways modulated by DSB IL-33 in Huvecs, Huvecs were cultured as described previously, plated at 1×10.sup.6 cells/well in a 6 well tissue culture treated plate (Nunc 140675). Following overnight incubation, cells were stimulated with DSB IL-33 for 2 or 6 hours. Cells were collected in 350 uL of RLT buffer (Qiagen #79216). RNA was purified using the RNeasy Micro Kit (Qiagen #74004) according to manufacturer's protocol. RNA was then amplified to single stranded DNA using Affymetrix's GeneChip WT Plus Reagent kit (Affymetrix #902513) and hybridised onto Human Genome U133A 2.0 (U133A 2.0) genechips (Affymetrix #900469), washed in Affymetrix Fluidics Station and scanned on the Affymetrix Genechip Scanner 3000 7G. Data was then processed in Affymetrix Expression console and sorted for genes with greater than ±1.8-fold change in signal from the untreated control. Very few gene expression changes were observed (Table 39). Nevertheless, the limited gene panel were analyzed using Ingenuity Pathway Analysis (IPA) (Qiagen), which suggested EIF2 signaling pathway at 2 hours and AGER signaling at 6 hours. Potentially these suggested scavenging/receptor for advanced glycation endproducts (RAGE) pathway activation.

(500) TABLE-US-00044 TABLE 39 Genes modulated in DSB-IL-33 stimulated Huvecs. Gene Symbols Timepoint Regulated genes >±1.8 fold change 2 hours EIF3F, RPL14, RPL38, RPL27A, RPL37A, RPS10, RPS27L 6 hours IFIT1, S100A8

(501) The receptor for advanced glycation end-products (RAGE) is a multi-ligand receptor that belongs to the immunoglobulin superfamily, and recognizes a variety of ligands, including high-mobility group box 1 (HMGB-1), 5100 family of proteins, advanced glycation end-products (AGE) and β-sheet fibrillar materials. It is thought to be involved in oxidative stress and has been linked to the pathogenesis of numerous diseases.

(502) To assess whether DSB directly interacted with RAGE, an ELISA format was used to explore RAGE binding to reduced IL-33 versus DSB IL-33 (FIG. 56A). Reduced or DSB N-terminal His Avi IL-33 (IL33-01, SEQ ID NO 632) were biotinylated as described in Example 7. Streptavidin plates (Thermo Scientific, AB-1226) were coated with biotinylated antigen at 50 μg/ml in PBS and incubated at room temperature for 1 hour. Plates were washed 3× with PBS-T (PBS+1% (v/v) Tween-20) and blocked with 300 μl/well blocking buffer (PBS with 1% BSA (Sigma, A9576)) for 1 hour. Plates were washed 3× with PBS-T. RAGE-Fc (R&D Systems #1145-RG) was diluted in blocking buffer, added to the IL-33-coated or control (no IL-33) wells and incubated at room temperature for 1 hour. RAGE-Fc was detected with anti-human IgG HRP (Sigma, A0170) diluted 1:5000 in blocking buffer, 50 μl/well for 1 hour at room temperature. Plates were washed 3× with PBS-T and developed with TMB, 50 μl/well (Sigma, T0440). The reaction was quenched with 50 μl/well 0.1M H2504 before reading on an EnVision™ plate reader, or similar equipment, at 450 nm.

(503) To further confirm the interaction of DSB IL-33 with RAGE, the ability of RAGE-Fc or anti-RAGE antibodies to inhibit the ST2-independent pSTAT5 signalling in Huvecs was evaluated. To this purpose, varying concentrations of DSB IL-33 (IMDM-treated IL33-01) were used to stimulate Huvecs according to the protocol described above in the presence or absence of RAGE-Fc (R&D Systems #1145-RG), ST2-Fc (R&D Systems #523-ST), anti-RAGE mAb (from WO 2008137552) or control reagents. Neutralisation of DSB IL-33 with RAGE-Fc (FIG. 56B), or neutralization of the receptor with anti-RAGE mAb (FIG. 56C) were able to completely inhibit the pSTAT5 signal.

(504) FIG. 56A. Binding of RAGE-Fc to reduced IL-33 or DSB plate surface by ELISA, wherein the x-axis shows the RAGE-Fc concentration and the y-axis shows the absorbance@450 nM. Data showed increased binding of RAGE to DSB IL-33 compared with reduced IL-33.

(505) FIG. 56B shows the pSTAT5 response to DSB IL-33 in Huvec in the presence of RAGE-Fc (50 ug/mL), ST2-Fc (50 ug/mL) or anti-NIP IgG1 negative control antibody, NIP228 (50 ug/mL). pSTAT5 signalling was completely inhibited by RAGE-Fc but not ST2-Fc or NIP228.

(506) FIG. 56C shows the pSTAT5 response to DSB IL-33 in Huvec in the presence of anti-RAGE mAb, m4F4 (10 ug/mL), or mouse IgG1 negative control antibody (10 ug/mL). pSTAT5 signalling was completely inhibited by m4F4 but not control mAb.

(507) Prevention of DSB IL-33 Activity with Anti-IL-33 Antibodies

(508) As described in Example 8, antibodies that bind IL-33 may prevent oxidation of IL-33 to the DSB form (FIG. 43A). The ability of IL-33 antibodies to prevent pSTAT5 signaling in Huvecs was evaluated.

(509) Fixed concentrations of 33_640087-7B (SEQ ID Nos 616 and 618), Anti-ST2 (from WO 2013/173761 Ab2; SEQ ID 85 and SEQ ID 19) and isotype control mAbs were prepared in IMDM and then combined (100 uL with 100 uL) with a titration of WT IL-33 (also prepared in IMDM) in 96 well U-bottom plates. Plates were incubated at 37° C. and 5% CO.sub.2 overnight. 75 uL from each well of these ‘preincubation treatment’ plates was added to ‘starved’ cells prepared as described above for pSTAT5 assay and incubated at 37° C. for 15 mins. Cells were then washed twice in ice cold PBS and 100 uL Lysis Buffer from the eBioscience Phospho-STAT5A/B Instant One ELISA (eBioscience #85-86112-11) was added to each well. pSTAT5 activity in cell lysates was then measured according to manufacturers instructions.

(510) FIG. 57 shows the pSTAT5 response in Huvecs to IL-33 treated with IMDM in the presence of 33_640087-7B (10 ug/mL) or Anti-ST2 mAb, Ab2, (10 ug/mL), wherein the x-axis is the concentration of IL-33 and the y axis is the pSTAT5 signal. pSTAT5 signalling was completely inhibited by 33_640087-7B but not anti-ST2, confirming that this response is ST2-independent.

(511) Anti-IL-33 Antibodies Inhibit RAGE-Dependent Response in Epithelial Cells

(512) RAGE is expressed highly in lung epithelial cells. Lung epithelial cell lines were evaluated for DSB IL-33 dependent responses. To this purpose, A549 cells were cultured in F12K Media (Gibco #21127022) supplemented with 1% Penicillin/Streptomycin and 10% FBS. Cells were harvested with 0.5% Trypsin-EDTA (Gibco, #15400-054), washed and seeded in 96 well plates at 1×10.sup.5/cells/well in culture media. Cells were then were incubated at at 37° C., 5% CO.sub.2 for 24 hours. The following day complete media was removed, cells were washed twice in PBS and media replaced with ‘starve’ media (F12K media with 1% Pen/Strep) and plates incubated for 24 hours at 37° C. and 5% CO.sub.2.

(513) Fixed concentrations of 33_640087-7B (SEQ ID Nos 616 and 618), Anti-ST2 (WO 2013/173761 Ab2 (SEQ ID 85 and SEQ ID 19)), anti-RAGE m4F4 (from WO 2008137552) and isotype control mAbs were prepared in IMDM and then combined (100 uL with 100 uL) with WT IL-33 (also prepared in IMDM) in 96 well U-bottom plates. Both cell and treatment plates were incubated at 37° C. and 5% CO.sub.2 overnight. 75 uL from each well of these ‘preincubation treatment’ plates was added to ‘starved’ cells prepared as described above and plates incubated for 24 hours at 37° C. and 5% CO.sub.2. The 96 well transwell system (Corning #CLS3422-48EA) is then set up by adding a 96 well transwell plate to a low binding 96 well receiver plate. 235 uL of complete media (F12K media supplemented with 10% FBS and 1% Pen/Strep) was added to the bottom chamber of the transwell system. Each well of the 96 well treated A549 cells were then washed in PBS, trypsinised to detach, centrifuged at 1000 rpm for 5 min, resuspended in 75 uL ‘starve’ media and added to the top chamber of the transwell system. The transwell plate was then incubated at 37° C., 5% CO.sub.2 for 16 hours. Media was then removed from both top and bottom chambers and cells removed from the bottom chamber using 235 uL trypsin. 100 uL of the trypsin/cell suspension was then added to 100 uL of Cell Titer Glo (Promega #G7571). A titration of fresh A549 cells is prepared in trypsin and added 50:50 to Cell Titer Glo to create a standard curve of cell number. Plates were then incubated and read according to manufacturer's instructions.

(514) FIG. 58A shows the migration of A549 cells after treatment with IL33-01 incubated in the presence of 33_640087-7B (10 ug/mL), Anti-ST2 mAb, Ab2, (10 ug/mL), or anti-RAGE mAb 4F4, wherein the x-axis shows the cell pre-treatment condition and the y axis is the number of cells migrated. Data demonstrate that pre-treatment of A549 cells with DSB IL-33 reduces subsequent cell migration. This inhibition of migration was reversed by anti-RAGE mAb and 33_640087-7B but not by anti-ST2.

(515) FIG. 58B shows the migration of A549 cells after treatment with DSB IL33-01 incubated in the presence of 33_640087-7B (10 ug/mL) or Anti-ST2 mAb, Ab2, (10 ug/mL, wherein the x-axis shows the cell pre-treatment condition and the y axis is the number of cells migrated. Data demonstrate that pre-treatment of A549 cells with DSB IL-33 reduces subsequent cell migration. This inhibition of migration was not reversed by 33_640087-7B or anti-ST2

(516) Together, these data confirm that 33_640087-7B inhibits DSB-IL_33 activity by preventing the conversion of reduced IL-33 to DSB IL-33 rather than neutralizing DSB IL-33 directly, consistent with its ability to bind only the reduced, ST2-active form of IL-33.