ANTIBODIES AGAINST AREG AND ITS USE

Abstract

Provided are anti-AREG antibodies or immunoreactive fragments thereof for the treatment, diagnosis or prophylaxis of fibrotic diseases, including but not limited to renal fibrosis, hepatic fibrosis, pulmonary fibrosis, in particular, IPF. Polynucleotides or nucleic acid molecules encoding the antibodies, expression vectors, host cells and methods for making the antibodies are also provided. The anti-AREG antibodies specifically bind to AREG and block the function of AREG, through binding residues that locate in the EGF like domain.

Claims

1. An isolated anti-AREG antibody or fragment thereof, having the ability of inhibiting fibrosis, preferably, the fibrosis is renal fibrosis, hepatic fibrosis, pulmonary fibrosis, more preferably, IPF.

2. The anti-AREG antibody or fragment thereof of claim 1, which is capable of binding to AREG, preferably, binding to human AREG.

3. The anti-AREG antibody or fragment thereof of claim 1, which is a human anti-AREG antibody, or a murine anti-AREG antibody, or a humanized anti-AREG antibody, or a chimeric anti-AREG antibody, preferably, is a human monoclonal antibody (mAb), murine mAb, humanized mAb, or chimeric mAb, or which preferably is Fab fragment or F(ab).sub.2 fragment.

4. The anti-AREG antibody or fragment thereof of claim 1, which binds to AREG with high affinity, preferably, with a dissociation constant (KD) of less than about 10 nM, preferably, less than 1 nM, 0.1 nM, or 0.01 nM, preferably, in the range of 1×10.sup.−8-1×10.sup.−11, more preferably, in the range of 1×10.sup.−9-1×10.sup.−11.

5. The anti-AREG antibody or fragment thereof of claim 1, which is capable of binding to soluble forms of AREG, preferably, is capable of binding to EGF-like domain of soluble forms of AREG.

6. The anti-AREG antibody or fragment thereof of claim 1, which is capable of binding to residues 101-184 of the human pro-AREG, and/or residues 171-184 of the human pro-AREG, and/or residues 94-177 of the murine pro-AREG, and/or residues 135-177 of the murine pro-AREG.

7. The anti-AREG antibody or fragment thereof of claim 1, which is capable of binding at least one, two, three, four or five amino acids within residues 101-184 of pro-AREG shown by any one of SEQ ID NOs: 123-132, preferably, within residues 142-184 of human pro-AREG shown by any one of SEQ ID NOs: 123-132.

8. The anti-AREG antibody or fragment thereof of claim 1, which is capable of interacting with Glu149 and/or His164 of human pro-AREG.

9. The anti-AREG antibody or fragment thereof of claim 1, which comprises a heavy chain variable region comprising heavy chain complementarity determining regions HCDR1, HCDR2, and HCDR3, and a light chain variable region comprising light chain complementarity determining regions LCDR1, LCDR2, and LCDR3, wherein: HCDR1, HCDR2, and HCDR3 are selected from the group consisting of: (1) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 2, HCDR3 shown by SEQ ID NO: 3; (2) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 2, HCDR3 shown by SEQ ID NO: 4; (3) HCDR1 shown by SEQ ID NO: 5, HCDR2 shown by SEQ ID NO: 2, HCDR3 shown by SEQ ID NO: 6; (4) HCDR1 shown by SEQ ID NO: 7, HCDR2 shown by SEQ ID NO: 8, HCDR3 shown by SEQ ID NO: 9; (5) HCDR1 shown by SEQ ID NO: 7, HCDR2 shown by SEQ ID NO: 10, HCDR3 shown by SEQ ID NO: 9; (6) HCDR1 shown by SEQ ID NO: 7, HCDR2 shown by SEQ ID NO: 8, HCDR3 shown by SEQ ID NO: 11; (7) HCDR1 shown by SEQ ID NO: 7, HCDR2 shown by SEQ ID NO: 8, HCDR3 shown by SEQ ID NO: 12; (8) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 13, HCDR3 shown by SEQ ID NO: 14; (9) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 15, HCDR3 shown by SEQ ID NO: 16; (10) HCDR1 shown by SEQ ID NO: 17, HCDR2 shown by SEQ ID NO: 18, HCDR3 shown by SEQ ID NO: 19; (11) HCDR1 shown by SEQ ID NO: 17, HCDR2 shown by SEQ ID NO: 18, HCDR3 shown by SEQ ID NO: 20; (12) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 13, HCDR3 shown by SEQ ID NO: 136; and (13) HCDR1, HCDR2, HCDR3 as shown in (1)-(12), but at least one of which includes one, two, three, four or five amino acids addition, deletion, conservative amino acid substitution or the combinations thereof; and LCDR1, LCDR2, and LCDR3 are selected from the group consisting of: (1) LCDR1 shown by SEQ ID NO: 21, LCDR2 shown by SEQ ID NO: 22, LCDR3 shown by SEQ ID NO: 23; (2) LCDR1 shown by SEQ ID NO: 21, LCDR2 shown by SEQ ID NO: 22, LCDR3 shown by SEQ ID NO: 24; (3) LCDR1 shown by SEQ ID NO: 25, LCDR2 shown by SEQ ID NO: 26, LCDR3 shown by SEQ ID NO: 27; (4) LCDR1 shown by SEQ ID NO: 28, LCDR2 shown by SEQ ID NO: 29, LCDR3 shown by SEQ ID NO: 30; (5) LCDR1 shown by SEQ ID NO: 31, LCDR2 shown by SEQ ID NO: 32, LCDR3 shown by SEQ ID NO: 30; (6) LCDR1 shown by SEQ ID NO: 33, LCDR2 shown by SEQ ID NO: 34, LCDR3 shown by SEQ ID NO: 30; (7) LCDR1 shown by SEQ ID NO: 35, LCDR2 shown by SEQ ID NO: 34, LCDR3 shown by SEQ ID NO: 30; (8) LCDR1 shown by SEQ ID NO: 36, LCDR2 shown by SEQ ID NO: 37, LCDR3 shown by SEQ ID NO: 38; (9) LCDR1 shown by SEQ ID NO: 39, LCDR2 shown by SEQ ID NO: 40, LCDR3 shown by SEQ ID NO: 38; (10) LCDR1 shown by SEQ ID NO: 41, LCDR2 shown by SEQ ID NO: 42, LCDR3 shown by SEQ ID NO: 38; (11) LCDR1 shown by SEQ ID NO: 43, LCDR2 shown by SEQ ID NO: 44, LCDR3 shown by SEQ ID NO: 38; (12) LCDR1 shown by SEQ ID NO: 39, LCDR2 shown by SEQ ID NO: 40, LCDR3 shown by SEQ ID NO: 38; (13) LCDR1 shown by SEQ ID NO: 45, LCDR2 shown by SEQ ID NO: 42, LCDR3 shown by SEQ ID NO: 46; (14) LCDR1 shown by SEQ ID NO: 47, LCDR2 shown by SEQ ID NO: 44, LCDR3 shown by SEQ ID NO: 46; (15) LCDR1 shown by SEQ ID NO: 48, LCDR2 shown by SEQ ID NO: 37, LCDR3 shown by SEQ ID NO: 49; (16) LCDR1 shown by SEQ ID NO: 50, LCDR2 shown by SEQ ID NO: 40, LCDR3 shown by SEQ ID NO: 51; (17) LCDR1 shown by SEQ ID NO: 50, LCDR2 shown by SEQ ID NO: 40, LCDR3 shown by SEQ ID NO: 52; (18) LCDR1 shown by SEQ ID NO: 50, LCDR2 shown by SEQ ID NO: 40, LCDR3 shown by SEQ ID NO: 53; (19) LCDR1 shown by SEQ ID NO: 54, LCDR2 shown by SEQ ID NO: 42, LCDR3 shown by SEQ ID NO: 55; (20) LCDR1 shown by SEQ ID NO: 56, LCDR2 shown by SEQ ID NO: 44, LCDR3 shown by SEQ ID NO: 55; and (21) LCDR1, LCDR2, LCDR3 as shown in (1)-(20), but at least one of which includes one, two, three, four or five amino acids addition, deletion, conservative amino acid substitution or the combinations thereof.

10. The anti-AREG antibody or fragment thereof of claim 1, which comprises a heavy chain variable region comprising heavy chain complementarity determining regions HCDR1, HCDR2, and HCDR3, and a light chain variable region comprising light chain complementarity determining regions LCDR1, LCDR2, and LCDR3, wherein: HCDR1, HCDR2, HCDR3, LCDR1, LCDR2, and LCDR3 are selected from the group consisting of: (1) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 2, HCDR3 shown by SEQ ID NO: 3, LCDR1 shown by SEQ ID NO: 21, LCDR2 shown by SEQ ID NO: 22, LCDR3 shown by SEQ ID NO: 23; (2) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 2, HCDR3 shown by SEQ ID NO: 4, LCDR1 shown by SEQ ID NO: 21, LCDR2 shown by SEQ ID NO: 22, LCDR3 shown by SEQ ID NO: 24; (3) HCDR1 shown by SEQ ID NO: 5, HCDR2 shown by SEQ ID NO: 2, HCDR3 shown by SEQ ID NO: 6, LCDR1 shown by SEQ ID NO: 25, LCDR2 shown by SEQ ID NO: 26, LCDR3 shown by SEQ ID NO: 27; (4) HCDR1 shown by SEQ ID NO: 7, HCDR2 shown by SEQ ID NO: 8, HCDR3 shown by SEQ ID NO: 9, LCDR1 shown by SEQ ID NO: 28, LCDR2 shown by SEQ ID NO: 29, LCDR3 shown by SEQ ID NO: 30; (5) HCDR1 shown by SEQ ID NO: 7, HCDR2 shown by SEQ ID NO: 10, HCDR3 shown by SEQ ID NO: 9, LCDR1 shown by SEQ ID NO: 31, LCDR2 shown by SEQ ID NO: 32, LCDR3 shown by SEQ ID NO: 30; (6) HCDR1 shown by SEQ ID NO: 7, HCDR2 shown by SEQ ID NO: 8, HCDR3 shown by SEQ ID NO: 11, LCDR1 shown by SEQ ID NO: 33, LCDR2 shown by SEQ ID NO: 34, LCDR3 shown by SEQ ID NO: 30; (7) HCDR1 shown by SEQ ID NO: 7, HCDR2 shown by SEQ ID NO: 8, HCDR3 shown by SEQ ID NO: 12, LCDR1 shown by SEQ ID NO: 35, LCDR2 shown by SEQ ID NO: 34, LCDR3 shown by SEQ ID NO: 30; (8) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 13, HCDR3 shown by SEQ ID NO: 14, LCDR1 shown by SEQ ID NO: 36, LCDR2 shown by SEQ ID NO: 37, LCDR3 shown by SEQ ID NO: 38; (9) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 13, HCDR3 shown by SEQ ID NO: 136, LCDR1 shown by SEQ ID NO: 39, LCDR2 shown by SEQ ID NO: 40, LCDR3 shown by SEQ ID NO: 38; (10) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 13, HCDR3 shown by SEQ ID NO: 136, LCDR1 shown by SEQ ID NO: 41, LCDR2 shown by SEQ ID NO: 42, LCDR3 shown by SEQ ID NO: 38; (11) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 13, HCDR3 shown by SEQ ID NO: 136, LCDR1 shown by SEQ ID NO: 43, LCDR2 shown by SEQ ID NO: 44, LCDR3 shown by SEQ ID NO: 38; (12) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 15, HCDR3 shown by SEQ ID NO: 16, LCDR1 shown by SEQ ID NO: 39, LCDR2 shown by SEQ ID NO: 40, LCDR3 shown by SEQ ID NO: 38; (13) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 15, HCDR3 shown by SEQ ID NO: 16, LCDR1 shown by SEQ ID NO: 45, LCDR2 shown by SEQ ID NO: 42, LCDR3 shown by SEQ ID NO: 46; (14) HCDR1 shown by SEQ ID NO: 1, HCDR2 shown by SEQ ID NO: 15, HCDR3 shown by SEQ ID NO: 16, LCDR1 shown by SEQ ID NO: 47, LCDR2 shown by SEQ ID NO: 44, LCDR3 SHOWN BY SEQ ID NO: 46; (15) HCDR1 shown by SEQ ID NO: 17, HCDR2 shown by SEQ ID NO: 18, HCDR3 shown by SEQ ID NO: 19, LCDR1 shown by SEQ ID NO: 48, LCDR2 shown by SEQ ID NO: 37, LCDR3 shown by SEQ ID NO: 49; (16) HCDR1 shown by SEQ ID NO: 17, HCDR2 shown by SEQ ID NO: 18, HCDR3 shown by SEQ ID NO: 20, LCDR1 shown by SEQ ID NO: 50, LCDR2 shown by SEQ ID NO: 40, LCDR3 shown by SEQ ID NO: 51; (17) HCDR1 shown by SEQ ID NO: 17, HCDR2 shown by SEQ ID NO: 18, HCDR3 shown by SEQ ID NO: 20, LCDR1 shown by SEQ ID NO: 50, LCDR2 shown by SEQ ID NO: 40, LCDR3 shown by SEQ ID NO: 52; (18) HCDR1 shown by SEQ ID NO: 17, HCDR2 shown by SEQ ID NO: 18, HCDR3 shown by SEQ ID NO: 20, LCDR1 shown by SEQ ID NO: 50, LCDR2 shown by SEQ ID NO: 40, LCDR3 shown by SEQ ID NO: 53; (19) HCDR1 shown by SEQ ID NO: 17, HCDR2 shown by SEQ ID NO: 18, HCDR3 shown by SEQ ID NO: 20, LCDR1 shown by SEQ ID NO: 54, LCDR2 shown by SEQ ID NO: 42, LCDR3 shown by SEQ ID NO: 55; (20) HCDR1 shown by SEQ ID NO: 17, HCDR2 shown by SEQ ID NO: 18, HCDR3 shown by SEQ ID NO: 20, LCDR1 shown by SEQ ID NO: 56, LCDR2 shown by SEQ ID NO: 44, LCDR3 shown by SEQ ID NO: 55; and (21) HCDR1, HCDR2, HCDR3, LCDR1, LCDR2, LCDR3 as shown in (1)-(20), but at least one of which includes one, two, three, four or five amino acids addition, deletion, conservative amino acid substitution or the combinations thereof.

11. The anti-AREG antibody or fragment thereof of claim 1, which comprises a heavy chain variable region, and a light chain variable region, wherein the heavy chain variable region has the amino acid sequence selected from the group consisting of SEQ ID NOs: 57-69, and an amino acid sequence having at least 95% sequence identity to any one of SEQ ID NOs: 57-69, and retaining the activity of epitope-binding, wherein the light chain variable region has the amino acid sequence selected from the group consisting of SEQ ID NOs: 70-89, and an amino acid sequence having at least 95% sequence identity to any one of SEQ ID NOs: 70-89, and retaining the activity of epitope-binding.

12. The anti-AREG antibody or fragment thereof of claim 1, which comprises a heavy chain variable region, and a light chain variable region, wherein the heavy chain variable region and the light chain variable region have the amino acid sequences selected from the group consisting of: (1) SEQ ID NO: 57 and SEQ ID NO: 70; (2) SEQ ID NO: 58 and SEQ ID NO: 71; (3) SEQ ID NO: 59 and SEQ ID NO: 72; (4) SEQ ID NO: 60 and SEQ ID NO: 73; (5) SEQ ID NO: 61 and SEQ ID NO: 74; (6) SEQ ID NO: 62 and SEQ ID NO: 75; (7) SEQ ID NO: 63 and SEQ ID NO: 76; (8) SEQ ID NO: 64 and SEQ ID NO: 77; (9) SEQ ID NO: 65 and SEQ ID NO: 78; (10) SEQ ID NO: 66 and SEQ ID NO: 79; (11) SEQ ID NO: 66 and SEQ ID NO: 80; (12) SEQ ID NO: 66 and SEQ ID NO: 81; (13) SEQ ID NO: 67 and SEQ ID NO: 79; (14) SEQ ID NO: 67 and SEQ ID NO: 82; (15) SEQ ID NO: 67 and SEQ ID NO: 83; (16) SEQ ID NO: 68 and SEQ ID NO: 84; (17) SEQ ID NO: 69 and SEQ ID NO: 85; (18) SEQ ID NO: 69 and SEQ ID NO: 86; (19) SEQ ID NO: 69 and SEQ ID NO: 87; (20) SEQ ID NO: 69 and SEQ ID NO: 88; (21) SEQ ID NO: 69 and SEQ ID NO: 89; and (22) two amino acid sequences having at least 95% sequence identity to any one of (1)-(21) respectively, and retaining the activity of epitope-binding.

13. The anti-AREG antibody or fragment thereof of claim 1, which is an isotype of IgG, IgM, IgA, IgE or IgD, preferably, an isotype of IgG1, IgG2, IgG3, or IgG4.

14. An isolated polynucleotide or a nucleic acid encoding the anti-AREG antibody or fragment thereof according to claim 1, wherein the polynucleotide or nucleic acid encodes the entire heavy chain variable region, or the entire light chain variable region, or the both on the same polynucleotide or on separate polynucleotides; or wherein the polynucleotide or nucleic acid encodes portions of heavy chain variable region, or the light chain variable region, or the both on the same polynucleotide or on separate polynucleotides.

15. The isolated polynucleotide or a nucleic acid of claim 14, which comprises the DNA sequence encoding the heavy chain variable region shown by any one of sequences SEQ ID NOs: 90, 92, 94, 96, 98, 100, 102, 104, 106, 108, 112, 115, and 117, and/or the DNA sequence encoding the light chain variable region shown by any one of sequences SEQ ID NOs: 91, 93, 95, 97, 99, 101, 103, 105, 107, 109, 110, 111, 113, 114, 116, 118, 119, 120, 121, and 122.

16. An isolated cell, or vector comprising one or more polynucleotide encoding the anti-AREG antibody or fragment thereof according to claim 1.

17. A composition comprising the anti-AREG antibody or fragment thereof according to claim 1, and a pharmaceutical acceptable carrier.

18. A method for treating a disorder, in which AREG is overexpressed, upregulated or activated, in a subject, comprising administering to the subject the anti-AREG antibody or fragment thereof according to claim 1, wherein the disorder is a fibrotic disease including but not limited to renal fibrosis, hepatic fibrosis, pulmonary fibrosis, in particular, IPF.

19. A method for diagnosing a disorder, in which AREG is overexpressed, upregulated or activated, comprising exposing a sample from a subject suspected of suffering from the disorder to the anti-AREG antibody or fragment thereof according to claim 1, and determining binding of the anti-AREG antibody or fragment thereof to the sample, wherein the disorder is a fibrotic disease, including but not limited to renal fibrosis, hepatic fibrosis, pulmonary fibrosis, in particular, IPF.

20. An isolated AREG protein, having an amino acid sequence shown in any one of SEQ ID NOs: 123-132, or an amino acid sequence that is at least 85% identical to any one of SEQ ID NOs: 123-132.

21. The isolated AREG protein of claim 20, which is an epitope for producing the anti-AREG antibody or fragment wherein the AREG antibody or fragment has the ability of inhibiting fibrosis, preferably, the fibrosis is renal fibrosis, hepatic fibrosis, pulmonary fibrosis, more preferably, IPF.

22. The isolated AREG protein of claim 20, which has the amino acid Glu149 and/or His164.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0122] FIG. 1 shows the binding of E1H3L4 and P7 to hAREG, mAREG and hAREG-C18.

[0123] FIG. 2 shows inhibition activities of anti-AREG mAbs against EGFR phosphorylation in hEGFR-expressing epidermoid carcinoma cells.

[0124] FIG. 3 shows five hAREG-EGFd variants generated by changing each amino acid at five different sites of hAREG-EGFd to the counterpart amino acid of mAREG-EGFd.

[0125] FIG. 4 shows the scheme of generating a mouse line in which Cdc42 gene is specifically deleted in AT2 cells. The mice in which the exon2 of the Cdc42 gene is specifically deleted in AT2 cells are named as Cdc42 AT2 null mice.

[0126] FIG. 5 shows that loss of Cdc42 in AT2 cells leads to progressive lung fibrosis in PNX-treated mice.

[0127] FIG. 6 shows that the anti-AREG antibody (P7) is effective for treating lung fibrosis in the IPF-like lung fibrosis mouse model.

[0128] FIG. 7 shows that the anti-AREG antibody (E1H3L4) treatment could accelerate the recovery of mice in the bleomycin-induced lung fibrosis mouse model.

[0129] FIG. 8 shows that the anti-AREG antibody (E1H3L4) is effective for treating lung fibrosis in the IPF-like lung fibrosis mouse model.

[0130] FIG. 9 shows that the anti-AREG antibody hu9C12v4 significantly prolongs the life expectancy of fibrosis mice in the IPF-like lung fibrosis mouse model.

DESCRIPTION OF PARTICULAR EMBODIMENTS OF THE INVENTION

[0131] The descriptions of particular embodiments and examples are provided by way of illustration and not by way of limitation. Those skilled in the art will readily recognize a variety of noncritical parameters that could be changed or modified to yield essentially similar results.

EXAMPLES

Example 1. Generation of Human mAbs Against AREG from a Phage Library

[0132] 1. Preparation of Soluble AREG Proteins or Peptides for Library Selection, and Screening

[0133] The DNA sequences encoding three forms of AREG proteins (listed below) were cloned into a prokaryotic expression vector(pETDuet), and expressed as a fusion protein carrying N-terminal tags (His.sub.6, thioredoxin(TRX), a HRV 3C Protease cleavage site, and an Avi tag). The proteins were expressed in Escherichia coli(TransB) by IPTG induction, and purified from the supernatant of cell lysate using Ni-NTA beads, cleaved by HRV 3C Protease, then was biotinylated using BirA enzyme.

[0134] The three AREG proteins are: [0135] hAREG: human AREG comprising residues 101-184 of the human pro-AREG with an N-terminal AVI tag (GLNDIFEAQKIEWHE). The amino acid sequence amino acid sequence of hAREG residues 101-184 is:

TABLE-US-00002 (SEQ ID NO: 129) SVRVEQVVKPPQNKTESENTSDKPKRKKKGGKNGKNRRNRKKKNPCNA EFQNFCIHGECKYIEHLEAVTCKCQQEYFGERCGEK  [0136] mAREG: mouse AREG comprising residues 94-177 of the mouse pro-AREG with an N-terminal AVI tag. The amino acid sequence is:

TABLE-US-00003 (SEQ ID NO: 130) GLNDIFEAQKIEWHEGGGGSGGSVRVEQVIKPKKNKTEGEKSTEKPKR KKKGGKNGKGRRNKKKKNPCTAKFQNFCIHGECRYIENLEVVTCNCHQ DYFGERCGEK [0137] mAREG-EGFd: EGF-like domain of mouse AREG comprises residues 135-177 of the mouse Pro-AREG with an N-terminal AVI tag. The amino acid sequence of mAreg residues 135-177 is:

TABLE-US-00004 (SEQ ID NO: 131) KKNPCTAKFQNFCIHGECRYIENLEVVTCNCHQDYFGERCGEK.

[0138] In addition, a biotinylated peptide, C18 was synthesized(Scilight-peptide, Beijing, China). C18 comprises 14 amino acids of the C-terminus of human AREG (residues 171-184) and a linker (residues GSSG) at the N-terminus. The sequence of C18 is GSSGKCQQEYFG ERCGEK (SEQ ID NO:132).

[0139] 2. Selection and Further Characterization of Antibodies from Phage Display Antibody Library

[0140] Phage Display Antibody Library

[0141] A human non-immune scFv(Single-chain variable fragment) antibody library was constructed from peripheral blood mononuclear cells (PBMCs) of 93 healthy donors. The library has a size of a total of 1.1×10.sup.10 members (Li et al., 2017).

[0142] Selection and Screening of Phage Antibody Library

[0143] Phage particles expressing scFv on their surface (phage-ScFv) were prepared from the library and used for selection of scFvs against the target antigens including the biotinylated AREG proteins and peptides. The antigens were captured on streptavidin-conjugated magnetic M-280 Dynabeads® (Life Technologies) and then incubated with 5×10.sup.12 phage particles prepared from the library, respectively. For each soluble AREG protein or peptide (antigen, Ab), two rounds of selection were performed. For obtaining cross-reactive human mAbs recognizing both hAREG and mAREG, the hAREG and mAREG were used respectively in the 1.sup.st and 2.sup.nd rounds of selection. After the 2.sup.nd round of selection, about 400 phage-Ab clones were screened for cross-binding activity to both hAREG and mAREG using ELISA, and clones with cross-binding activity or high binding affinity to hAREG were selected for sequencing analysis to identify clones with different antibody sequences, including variable regions of heavy (VH) and light (VL) chain. Some of the phage-Abs were subsequently converted into human IgG1 (hIgG1) or mouse IgG2a format and analyzed for binding to both hAREG and mAREG using enzyme-linked immunosorbent assay (ELISA) or Biacore.

[0144] 3. Preparation of Full-Length Antibody

[0145] The VH and VL coding sequence of a scFv was separately sub-cloned into antibody heavy chain (HC) expression vector(plasmid) and light chain (LC) expression vector(plasmid). To make full-length antibodies, 293F cells were transiently co-transfected with the two expression plasmids (HC+LC plasmids) at a 1:1 ratio. Six days after transfection, the cell culture supernatant was harvested for purification of antibodies by Protein A affinity chromatography.

[0146] 4. ELISA Assay

[0147] Streptavidin (Sigma, 4 g/mL) in phosphate buffered saline (PBS) was coated in U-bottom 96-well plate (Nunc, MaxiSorp™), 100 μL per well, at 4° C. overnight or 37° C. for 1 hour.

[0148] About 0.5 g/mL of AREG protein or peptide at 100 μL per well were then captured onto the plates by incubation at 30° C. for 1 hour. For phage-scFv based ELISA, serial diluted phage-scFvs in PBS containing 2% nonfat milk were added to each well at 100 μL per well. Specific bound phage-scFvs were detected by adding HRP-conjugated mouse anti-M13 antibody (GE Healthcare) and incubated for 30 mins at 30° C. In between each incubation step, the ELISA plate was washed for 6 times with PBST solution (0.05% Tween20 containing PBS) at 300 L per well. Followed by HRP-conjugated antibody incubation, the ELISA signal was developed by incubating with TMB substrate (Sigma) for 5-10 mins at 30° C. and then the reaction was stopped with 2M H.sub.2SO.sub.4 at 50 L per well. The absorbance at 450 nm with the correction wavelength set at 630 nm was read by a microplate reader (Bio-Rad). For IgG based ELISA, the method was basically the same as described above for phage-scFvs except the bound antibodies were detected by HRP-conjugated mouse anti-Fc secondary antibody (Thermo Fisher Scientific).

[0149] 5. E1L2 Antibody Engineering for Improved Affinity and Solubility

[0150] To improve the affinity of E1L2 antibody, the VH-CDR3 and VL-CDR3 of E1L2 were engineered separately. For VH-CDR3, a phage display sub-library with random mutagenesis for the HCDR3 of E1L2 was constructed. Antibody sub-library selection and screening were done similarly as described above for screening of antibody library against AREG. To obtain high-affinity hits, competitive elution with E1L2 full-length mAb was used. Subsequently, single clones were picked and rescued to produce phage-scFvs in the bacterial culture supernatant to screen for binding to hAREG. Only hits with higher binding affinities than E1L2 were retained. For VL-CDR3, it was engineered with specific amino acid mutations based on structural modeling. To improve solubility, the engineered VL CDRs of E1L2 were grafted to human IGLV1-44*01 germline.

[0151] 6. SPR Measurement of Affinity of Human mAbs

[0152] To evaluate affinities of human mAbs, SPR measurement was performed using Biacore T200 instrument. mAbs were captured on anti-human Fc CM5 biosensor chip surface, the EGF domain of hAREG (aa142-184) or mAREG (aa135-177) fused with mFc tag (hAREG-EGFd-mFc or mAREG-EGFd-mFc) were examined for binding to the mAbs. The fusion proteins in serial dilutions were injected over antibody-bound surface, followed by a dissociation phase. Association rates (Ka) and dissociation rates (Kd) were calculated using the one-to-one Langmuir binding model (BIA Evaluation Software, GE Life Sciences). The equilibrium dissociation constant (KD) was calculated as the ratio kd/ka.

[0153] 7. Results

[0154] Generation of Human mAbs, E1L2 and P7, Against AREG from a Phage Library

[0155] By using the above described phage antibody library selection and ELSIA screening, we identified C1, E1L2, P5, P6, P7 and P10 anti-AREG human mAbs. Specifically, by using E. coli expressed biotinylated hAREG and mAREG proteins in the 1.sup.st and 2.sup.nd round of the library selection, respectively, we identified C1 antibody. By using the biotinylated hAREG and biotinylated mAREG-EGFd (expressed in E. coli) in the 1.sup.st and 2.sup.nd round, respectively, we identified E1L2 antibody. By using hAREG derived C18 peptide as the target for library selection for two rounds, we identified P5, P6, P7, and P10; E1L2 was also screened out from this library selection. Among these antibodies, E1L2 and P7 antibodies were selected for further characterization based on their binding specificity and affinity to both hAREG and mAREG.

[0156] Creation of E1L2-Derived Antibody E1H3L4 with Improved Affinity and Solubility

[0157] The binding affinity of E1L2 was further improved by VH-CDR3 and VL-CDR3 engineering. The solubility of the engineered E1L2 was improved by grafting its VL-CDRs to human IGLV1-44*01 germline, thus, resulting in an antibody, E1H3L4. Comparing to E1L2, E1H3L4 has three amino-acid changes in the VH-CDR3 and four amino-acid changes in the VL-CDR3. Specifically, amino acids at positions 100-100, (Kabat system)correct in VH-CDR3 of E1H3L4 are SYNN, while they are GYDY in E1L2 antibody; amino acids at positions 93-95a (Kabat system) in VL-CDR3 of E1H3L4 are KNNK, while they are SGLN in E1L2 antibody. The CDRs of E1L2, E1H3L4 and P7 were listed in Table 1. The nucleotide sequences and the amino acid sequences of VH and VL of E1L2, E1H3L4 and P7 were listed in Table 2.

TABLE-US-00005 TABLE 1 CDRs of E1L2, E1H3L4, and P7 HCDR1 HCDR2 HCDR3 (SEQ ID No.) (SEQ ID No.) (SEQ ID No.) E1L2 SYAMS(1) AISGSGGSTYYA PTSRYSYGYDY DSVKG(2) (3) E1H3L4 SYAMS(1) AISGSGGSTYYA PTSRYSYSYNN DSVKG(2) (4) P7 SHAMS(5) AISGSGGSTYYA VDTKFDP(6) DSVKG(2) LCDR1 LCDR2 LCDR3 (SEQ ID No.) (SEQ ID No.) (SEQ ID No.) E1L2 TGNSNNVGDQGA RNNNRPS(22) STWDSGLNSVV V(21) (23) E1H3L4 TGNSNNVGDQGA RNNNRPS(22) STWDKNNKSVV V(21) (24) P7 SGSSSNIGSNTV SNNQRPS(26) EVWDDSLNGPV N(25) (27)
The differences between E1L2 and E1H3L4 were underlined.
CDRs are defined using Kabat system.

TABLE-US-00006 TABLE 2 Name VH VH VL VL of the (nucleotide (amino acid (nucleotide (amino acid Antibody sequence) sequence) sequence) sequence) E1L2 SEQ ID SEQ ID SEQ ID SEQ ID NO: 90 NO: 57 NO: 91 NO: 70 E1H3L4 SEQ ID SEQ ID SEQ ID SEQ ID NO: 92 NO: 58 NO: 93 NO: 71 P7 SEQ ID SEQ ID SEQ ID SEQ ID NO: 94 NO: 59 NO: 95 NO: 72

[0158] 8. Further Characterization of E1H3L4 and P7 mAbs

[0159] Comparing the binding of E1H3L4 and P7 to mAREG, E1H3L4 showed slightly stronger binding to mAREG than P7. Both of the two antibodies bound to the C-terminal peptide (C18, aa 171-184) within the EGF domain as expected since the C18 peptide was the target used in the library selection (FIG. 1).

Example 2. Generation of mAbs Against AREG Using Mouse Hybridoma Method and Humanization of the Mouse mAbs

[0160] 1. Preparation of Antigens for Immunization of Mice or SPR Analysis

[0161] Human AREG (hAREG) EGF-like domain fused with an Fc fragment of human IgG1 or mouse IgG2a was expressed in 293F as a fusion protein, named as hAREG-EGFd-hFc and hAREG-EGFd-mFc, respectively. 72 hours after transfection, the cell culture supernatant were harvested for purification of the Fc-fusion AREG proteins by Protein A affinity chromatography.

[0162] 2. Generation of Anti-hAREG EGF Domain Monoclonal Antibodies

[0163] Six week-old Balb/c mice (from Beijing Vital River Laboratory Animal Technology Co., Ltd.) were immunized by subcutaneously administration with 100 μl of 1:1 antigen/adjuvant emulsion containing 50 μg of hAREG-EGFd-mFc. For priming immunization, complete Freund's adjuvant (Sigma) was used. For boosting immunization, incomplete Freund's adjuvant (Sigma) was used. Boosting immunization was performed every two weeks. After the 3.sup.rd boosting immunization, the sera of the mice were evaluated for binding to biotinylated hAREG by ELISA one week after each immunization. Mice with high titers of anti-hAREG antibody were boosted intraperitoneally with 50 μg of hAREG-EGFd-mFc without adjuvant. Three days after boosting, spleenocytes were isolated and fused with SP2/0 cells, following standard hybridoma fusion methods.

[0164] The supernatant of hybridoma clones were examined for binding activity to biotinylated hAREG by ELISA. Clones with high binding activity were selected and expanded for subsequent subcloning, and the supernatant of the subclones were analyzed by ELISA and SPR. The SPR analysis was performed using Biacore T200 instrument (GE Life Sciences). Diluted supernatant was captured on anti-mFc CM5 biosensor chip, then 200 nM of hAREG EGF domain-hFc flowed in mobile phase. Subclones with high affinity was expanded for RNA extraction. Cells were resuspended in TRIzol (Life Technologies), and total RNA was extracted following the instruction manual. The cDNA of subclones was synthesized using PrimeScript™ RT Master Mix (TaKaRa). The VH and VL genes of each antibody were amplified using a set of PCR primers specific to mouse antibody variable genes. PCR products were cloned into a PCR sequencing vector for sequencing.

[0165] 3. Humanization of Anti-hAREG EGF-Like Monoclonal Antibodies

[0166] For humanization of the AREG mAbs, sequences of murine mAbs were searched for human germline IgG genes homologous to identify the human germline genes with high homology to the murine mAbs (9C12, 23H8 and 1H9), and then chosen them as templates for humanization. Humanization was carried out by complementarity-determining region (CDR)-grating, specifically by grafting CDRs of murine mAb onto human acceptor framework of the selected human germline gene templates. This humanization process was also guided by the simulated 3D structure of each antibody and human framework residues back mutation to murine residues in order to maintain the overall antibody and CDR loop structures as well as AREG binding affinity.

[0167] 4. hu9C12v1 Antibody Sub-Library Construction and Selection for Improving Affinity

[0168] To improve the affinity of antibody, two phage display sub-libraries with random mutagenesis for the HCDR3 and LCDR1 of hu9C12v1 were constructed through NNK degenerate codons. Antibody sub-library selection and screening were done similarly as described above for screening of antibody library against AREG and affinity improvement of E1L2 antibody. Only hits with higher binding affinities than hu9C12v1 were retained after the screening.

[0169] 5. SPR Measurement Affinity of the mAbs

[0170] To evaluate affinities of different mouse hybridoma mAbs or their humanized and engineered variants, SPR measurement was performed using Biacore T200 instrument. mAbs were captured on anti-human Fc CM5 biosensor chip surface, hAREG-EGFd-mFc, mAREG-EGFd-mFc or hAREG-98aa (purchased from PeproTech, cat #100-55B) in serial dilutions were injected over antibody-bound surface, followed by a dissociation phase. Association rates (Ka) and dissociation rates (Kd) were calculated using the one-to-one Langmuir binding model (BIA Evaluation Software, GE Life Sciences). The equilibrium dissociation constant (KD) was calculated as the ratio kd/ka.

[0171] 6. Results

[0172] Generation of Anti-hAREG EGF Domain Monoclonal Antibodies

[0173] Anti-hAREG mAbs were generated based on conventional hybridoma fusion technology. MAbs with high binding activities in ELISA and SPR assay were selected for further characterization. Through screening thousands of hybridoma clones, we identified a panel of mAbs with high binding affinity to hAREG. Three top mAbs, 9C12, 23H8 and 1H9, were selected for further analysis based on their unique sequences, binding affinity and high yield of recombinant antibody production. These antibodies were made as mouse antibodies of mIgG1 or mIgG2a isotype or chimeric antibodies (mouse variable region grafted onto human IgG1 constant regions) by recombinant expression.

[0174] Humanization of 9C12 or Creation of Humanized Antibody Variants with Improved Binding Affinity to hAREG or mAREG and Improved Physicochemical Properties

[0175] CDR-grafting and structural modeling were used to generate the first version of the humanized 9C12, hu9C12v1, which has comparable affinity to hAREG as the chimeric antibody, ch9C12 (having the variable regions of 9C12, and constant regions of human IgG1). To improve the affinity of hu9C12v1 for hAREG, two phage display sub-libraries with randomized mutations within its HCDR3 and LCDR3 regions were separately constructed. After stringent bio-panning selections, a small panel of affinity-improved antibodies was obtained. Based on the sequences of this panel of antibodies, three mAbs, hu9C12v4, hu9C12v5 and hu9C12v6, were created to improve the binding affinity to hAREG or mAREG, and to improve their physicochemical properties. Comparing to hu9C12v1, hu9C12v4 has one amino-acid difference in the VH-CDR2, six amino-acid differences in the VK-CDR1, and two amino-acid differences in the VK-CDR2; hu9C12v5 has five amino-acid differences in the VH-CDR3, five amino-acid differences in the VK-CDR1, and one amino-acid difference in the VK-CDR2; hu9C12v6 has two amino-acid differences in the VH-CDR3, five amino-acid differences in the VK-CDR1, and one amino-acid difference in the VK-CDR2. The CDRs of the three mAbs were compared to the murine antibody as shown in Table 3. The nucleotide sequences and the amino acid sequences of VH and VL of the three mAbs and murine antibody were listed in Table 4. The SPR-determined binding affinities of the three mAbs to hAREG or mAREG were listed in Tables 5-6.

TABLE-US-00007 TABLE 3 Comparison of CDRs among different versions of mAb 9C12 HCDR1 HCDR2 HCDR3 (SEQ ID No.) (SEQ ID No.) (SEQ ID No.) m9C12 SYPMS(7) TISTGGTYTYYP QGPIYYGNYYYA DSVKG(8) MDY(9) 9C12v1 SYPMS(7) TISTGGTYTYYP QGPIYYGNYYYA DSVKG(8) MDY(9) hu9C12v4 SYPMS(7) TISTGGRYTYYP QGPIYYGNYYYA DSVKG(10) MDY(9) hu9C12v5 SYPMS(7) TISTGGTYTYYP QGPILRKNYYYG DSVKG(8) MDV(11) hu9C12v6 SYPMS(7) TISTGGTYTYYP QGPIYYGNYYYG DSVKG(8) MDV(12) LCDR1 LCDR2 LCDR3 (SEQ ID No.) (SEQ ID No.) (SEQ ID No.) m9C12 RSSQSLVHSDGN KVSNRFS(29) SQSTHVPYT TYLH(28) (30) hu9C12v1 RSSQSLVHSDGN KVSNRFS(29) SQSTHVPYT TYLH(28) (30) hu9C12v4 RSSQSLVDGEDG KVSERFD(32) SQSTHVPYT TYLN(31) (30) hu9C12v5 RSSQSLVDGQDG KVSNRFD(34) SQSTHVPYT TYLH(33) (30) hu9C12v6 RSSQSLVNQEGE KVSNRFD(34) SQSTHVPYT TYLH(35) (30)
The differences between Abs were underlined.
CDRs are defined using Kabat system.

TABLE-US-00008 TABLE 4 Name VH VH VL VL of the (nucleotide (amino acid (nucleotide (amino acid Antibody sequence) sequence) sequence) sequence) m9C12 SEQ ID SEQ ID SEQ ID SEQ ID NO: 96  NO: 60 NO: 97  NO: 73 hu9C12v1 SEQ ID SEQ ID SEQ ID SEQ ID NO: 98  NO: 61 NO: 99  NO: 74 hu9C12v4 SEQ ID SEQ ID SEQ ID SEQ ID NO: 100 NO: 62 NO: 101 NO: 75 hu9C12v5 SEQ ID SEQ ID SEQ ID SEQ ID NO: 102 NO: 63 NO: 103 NO: 76 hu9C12v6 SEQ ID SEQ ID SEQ ID SEQ ID NO: 104 NO: 64 NO: 105 NO: 77

TABLE-US-00009 TABLE 5 hAREG-EGFd mAREG-EGFd mAbs K.sub.a (M.sup.−1, s.sup.−1) K.sub.off (s) K.sub.D (M) K.sub.a (M.sup.−1, s.sup.−1) K.sub.off (s) K.sub.D (M) ch9C12 5.87 × 10.sup.5  1.0 × 10.sup.−3 1.71 × 10.sup.−9  hu9C12v1 5.70 × 10.sup.5 5.72 × 10.sup.−4 1.00 × 10.sup.−9  hu9C12v4 7.45 × 10.sup.5 7.70 × 10.sup.−4 1.04 × 10.sup.−9  2.96 × 10.sup.5 2.67 × 10.sup.−3 9.01 × 10.sup.−8 hu9C12v5 4.83 × 10.sup.5 4.11 × 10.sup.−4 8.52 × 10.sup.−10 hu9C12v6 9.78 × 10.sup.5 3.89 × 10.sup.−4 3.98 × 10.sup.−10

TABLE-US-00010 TABLE 6 hAREG-98aa mAbs K.sub.a (M.sup.−1, s.sup.−1) K.sub.off (s) K.sub.D (M) hu9C12v4 8.04 × 10.sup.6 3.72 × 10.sup.−3 4.62 × 10.sup.−10 hu9C12v5 8.11 × 10.sup.6 2.83 × 10.sup.−3 3.49 × 10.sup.−10 hu9C12v6 2.43 × 10.sup.7 4.57 × 10.sup.−3 1.88 × 10.sup.−10

[0176] Humanization of 23H8

[0177] We used CDR-grafting and structural modeling to generate the humanized 23H8 mAbs. The human VH germline gene IGHV3-21 was used for VH-CDR grafting. The human VK germline genes IGKV7-3, IGKV1-39, and IGKV4-1 were used for VK-CDR grafting, and generated three versions of humanized 23H8 VK chains. By combining the humanized VH and the three humanized VKs, we generated mAbs hu23H8v1, hu23H8v2 and hu23H8v3, respectively. These three humanized mAbs had similar affinity to hAREG as the chimeric 23H8 (murine variable regions and human IgG1 constant regions), indicating that grafting of the VK-CDRs of 23H8 to the three different human VK germline backbones were all successful. A couple of additional mutations were introduced into the humanized mAbs to remove the potential undesired post-translational modifications or immunogenicity, and three more variants were generated, hu23H8v4, -v5 and -v6. The CDRs of these mAbs were compared to the murine antibody as shown in Table 7. The nucleotide sequences and the amino acid sequences of VH and VL CDRs of these mAbs were shown in Table 8. The SPR-determined binding affinities of these mAbs to hAREG were listed in Tables 9-10.

TABLE-US-00011 TABLE 7 Comparison of CDRs among different versions of mAb 23H8 HCDR1 HCDR2 HCDR3 (SEQ ID No.) (SEQ ID No.) (SEQ ID No.) m23H8 SYAMS(1) TISTGGSHTYYP HGYLLYDGYYEW DSVKG(13) YFDV(14) hu23H8v1 SYAMS(1) TISTGGSHTYYP HGYLLYDGYYEW DSVKG(13) YFDY(136) hu23H8v2 SYAMS(1) TISTGGSHTYYP HGYLLYDGYYEW DSVKG(13) YFDY(136) hu23H8v3 SYAMS(1) TISTGGSHTYYP HGYLLYDGYYEW DSVKG(13) YFDY(136) hu23H8v4 SYAMS(1) TISTGGSHTYYP HGYLLYEGYYEW ESVKG(15) YFDY(16) hu23H8v5 SYAMS(1) TISTGGSHTYYP HGYLLYEGYYEW ESVKG(15) YFDY(16) hu23H8v6 SYAMS(1) TISTGGSHTYYP HGYLLYEGYYEW ESVKG(15) YFDY(16) LCDR1 LCDR2 LCDR3 (SEQ ID No.) (SEQ ID No.) (SEQ ID No.) m23H8 KASQSVDYDGHS AASNLES(37) QQSTEDPPYT FLN(36) (38) hu23H8v1 RASESVDYDGHS AASNKDT(40) QQSTEDPPYT FIN(39) (38) hu23H8v2 RASQSVDYDGHS AASNLQS(42) QQSTEDPPYT FLN(41) (38) hu23H8v3 KSSQSVDYDGHS AASNRES(44) QQSTEDPPYT FLN(43) (38) hu23H8v4 RASESVDYDGHS AASNKDT(40) QQSTEDPPYT FIN(39) (38) hu23H8v5 RASQSVDYEGHS AASNLQS(42) QQSTENPPYT FLN(45) (46) hu23H8v6 KSSQSVDYEGHS AASNRES(44) QQSTENPPYT FLN(47) (46)
The differences between Abs were underlined.
CDRs are defined using Kabat system.

TABLE-US-00012 TABLE 8 Name VH VH VL VL of the (nucleotide (amino acid (nucleotide (amino acid Antibody sequence) sequence) sequence) sequence) 23H8 SEQ ID SEQ ID SEQ ID SEQ ID NO: 106 NO: 65 NO: 107 NO: 78 hu23H8v1 SEQ ID SEQ ID SEQ ID SEQ ID NO: 108 NO: 66 NO: 109 NO: 79 hu23H8v2 SEQ ID SEQ ID SEQ ID SEQ ID NO: 108 NO: 66 NO: 110 NO: 80 hu23H8v3 SEQ ID SEQ ID SEQ ID SEQ ID NO: 108 NO: 66 NO: 111 NO: 81 hu23H8v4 SEQ ID SEQ ID SEQ ID SEQ ID NO: 112 NO: 67 NO: 109 NO: 79 hu23H8v5 SEQ ID SEQ ID SEQ ID SEQ ID NO: 112 NO: 67 NO: 113 NO: 82 hu23H8v6 SEQ ID SEQ ID SEQ ID SEQ ID NO: 112 NO: 67 NO: 114 NO: 83

TABLE-US-00013 TABLE 9 hAREG-EGFd mAbs K.sub.a (M.sup.−1, s.sup.−1) K.sub.off (s) K.sub.D (M) ch23H8 4.18 × 10.sup.5 3.13 × 10.sup.−4 7.50 × 10.sup.−10 hu23H8V1 1.05 × 10.sup.6 2.14 × 10.sup.−4 2.05 × 10.sup.−10 hu23H8V2 1.02 × 10.sup.6 2.49 × 10.sup.−4 2.44 × 10.sup.−10 hu23H8V3 1.20 × 10.sup.6 2.13 × 10.sup.−4 1.78 × 10.sup.−10 hu23H8V4 8.21 × 10.sup.5 2.76 × 10.sup.−4 3.36 × 10.sup.−10 hu23H8V5 6.28 × 10.sup.5 3.48 × 10.sup.−4 5.65 × 10.sup.−10 hu23H8V6 8.03 × 10.sup.5 2.46 × 10.sup.−4 3.09 × 10.sup.−10

TABLE-US-00014 TABLE 10 hAREG-98aa mAbs K.sub.a (M.sup.−1, s.sup.−1) K.sub.off (s) K.sub.D (M) hu23H8V5 1.26 × 10.sup.7 8.63 × 10.sup.−4 6.96 × 10.sup.−10 hu23H8V6 1.11 × 10.sup.5 5.95 × 10.sup.−4 5.37 × 10.sup.−10

[0178] Humanization of 1H9

[0179] Similar to humanization of 23H8, the human VH germline gene IGHV3-21 was used for VH-CDR grafting; the human VK germline gene IGKV7-3, IGKV1-39, and IGKV4-1 were used for VK-CDR grafting. The CDRs of these mAbs were compared to the murine antibody as shown in Table 11. The nucleotide sequences and the amino acid sequences of VH and VL CDRs of these mAbs were shown in Table 12. The SPR-determined binding affinities of these mAbs to hAREG were listed in Table 13.

TABLE-US-00015 TABLE 11 Comparison of CDRs among different versions of mAb 1H9 HCDR1 HCDR2 HCDR3 (SEQ ID No.) (SEQ ID No.) (SEQ ID No.) m1H9 GYPMS(17) TISTGARHTYYP HEGLRRGKYHCI DSVKG(18) MDY(19) hu1H9v1-5 GYPMS(17) TISTGARHTYYP HEGLRRGKYHSI DSVKG(18) MDY(20) LCDR1 LCDR2 LCDR3 (SEQ ID No.) (SEQ ID No.) (SEQ ID No.) m1H9(15) KASQSIDYDGDS AASNLES(37) HQCNEDPYM FLN(48) (49) hu1H9v1(16) RASESVDYDGDS AASNKDT(40) HQSNEDPYM FIN(50) (51) hu1H9v2(17) RASESVDYDGDS AASNKDT(40) HQSNEDPYL FIN(50) (52) hu1H9v3(18) RASESVDYDGDS AASNKDT(40) HQSNEDPYV FIN(50) (53) hu1H9v4(19) RASQSIDYDGDS AASNLQS(42) QQSNEDPYV FLN(54) (55) hu1H9v5(20) KSSQSIDYDGDS AASNRES(44) QQSNEDPYV FLN(56) (55)
The differences between Abs were underlined.
CDRs are defined using Kabat system.

TABLE-US-00016 TABLE 12 Name VH VH VL VL of the (nucleotide (amino acid (nucleotide (amino acid Antibody sequence) sequence) sequence) sequence) 1H9 SEQ ID SEQ ID SEQ ID SEQ ID NO: 115 NO: 68 NO: 116 NO: 84 hu1H9v1 SEQ ID SEQ ID SEQ ID SEQ ID NO: 117 NO: 69 NO: 118 NO: 85 hu1H9v2 SEQ ID SEQ ID SEQ ID SEQ ID NO: 117 NO: 69 NO: 119 NO: 86 hu1H9v3 SEQ ID SEQ ID SEQ ID SEQ ID NO: 117 NO: 69 NO: 120 NO: 87 hu1H9v4 SEQ ID SEQ ID SEQ ID SEQ ID NO: 117 NO: 69 NO: 121 NO: 88 hu1H9v5 SEQ ID SEQ ID SEQ ID SEQ ID NO: 117 NO: 69 NO: 122 NO: 89

TABLE-US-00017 TABLE 13 hAREG-EGFd mAbs K.sub.a (M.sup.−1, s.sup.−1) K.sub.off (s) K.sub.D (M) ch1H9 3.58 × 10.sup.5 4.67 × 10.sup.−4 1.31 × 10.sup.−10 hu1H9v1 2.91 × 10.sup.5 2.33 × 10.sup.−4 7.99 × 10.sup.−10 hu1H9v2 3.01 × 10.sup.5 2.29 × 10.sup.−4 7.62 × 10.sup.−10 hu1H9v3 2.67 × 10.sup.5 2.59 × 10.sup.−4 9.69 × 10.sup.−10 hu1H9v4 2.43 × 10.sup.5 1.84 × 10.sup.−4 7.58 × 10.sup.−10 hu1H9v5 2.54 × 10.sup.5 2.06 × 10.sup.−4 8.11 × 10.sup.−10

Example 3. Activity Analysis of Anti-AREG mAbs

[0180] 1. Preparation of AREG Proteins for In Vitro Activity Analysis of Anti-hAREG Antibodies

[0181] Human or mouse EGFR extracellular domain (ECD) encoding cDNA fused with a His6 and an Avi tag at the C-terminus, was co-transfected with a plasmid encoding BirA-hFc for biotinylation in 293F cells. 72 hours after transfection, the cell culture supernatant was harvested for purification of the hEGFR ECD His6-Avi-biotin fusion protein or mEGFR ECD His6-Avi-biotin fusion protein (hEFGR-ECD, mEGFR-ECD) by Protein A affinity chromatography.

[0182] Human or mouse AREG EGF domain with four additional residues (DLLA) at the C-terminus, was expressed in 293F cells as an mFc-fusion protein. 72 hours after transfection, the cell culture supernatant was harvested for purification of the hAREG-EGFd-DLLA-mFc (hAREG-DLLA) or mAREG-EGFd-DLLA-mFc (mAREG-DLLA) fusion proteins by Protein A affinity chromatography.

[0183] 2. Inhibition of hAREG for Binding to EGFR Analyzed by Competition ELISA

[0184] Briefly, streptavidin (Sigma, 5 μg/mL) were coated in U-bottom 96-well plates, 100 nM biotinylated hEGFR-ECD or mEGFR-ECD in 100 μL per well were then captured onto the plates. Different antibodies at serial diluted concentrations were mixed with 5 nM hAREG-DLLA or 50 nM mAREG-DLLA protein and added to the ELISA plate. The binding of hAREG-DLLA to hEGFR-ECD or the binding of mAREG-DLLA to mEGFR-ECD was detected by HRP-conjugated mouse anti-mouse IgG Fc antibody (Thermo Fisher).

[0185] 3. Inhibition of EGFR Receptor Phosphorylation

[0186] A431 (a human epidermoid carcinoma cell line) cells were serum-starved for one hour, and were subsequently either treated with hAREG-DLLA (2.5 nM) alone or treated with the mixtures of hAREG and anti-AREG antibodies for one hour. Approximately 1-2×10.sup.5 cells/well in 6-well plates were used in each treatment. The treated cells were washed with PBS twice and then lysed on ice using RIPA buffer. The cell lysates were then subjected to SDS-PAGE and followed by Western blotting. The phosphorylated form of EGFR (tyrosine 1068) and total EGFR were detected using an anti-phosphotyrosine mAb (Abcam, EP774Y) and a rabbit polyclonal antibody (Cell Signaling technology, #2232), respectively. An anti-α-Tubulin mAb (clone B-5-1-2, Sigma-Aldrich) was used to detect alpha Tubulin expression in the cell lysates, and served as a loading control for Western blotting analysis.

[0187] 4. Epitope Mapping

[0188] To identify the epitopes of our anti-AREG mAbs, five amino acids that are different and have distinct physical properties between hAREG-EGFd and mAREG-EGFd were chosen for mutagenesis. Five hAREG-EGFd variants were generated by changing each amino acid at five different sites of hAREG-EGFd to the counterpart amino acid of mAREG-EGFd. In addition, two mAREG-EGFd variants were generated by changing each amino acid at two different sites of mAREG-EGFd to the counterpart amino acid of hAREG-EGFd. These variants were then examined for binding with anti-AREG mAbs using SPR (Biacore T200).

[0189] 5. Results

[0190] Anti-AREG mAbs Block AREG Binding to EGFR

[0191] It has been previously shown that the addition of four amino acids (DLLA) to the C-terminal of EGF domain of human AREG greatly improved the biological activity of recombinant expressed EGF domain (Thompson et al., 1996). To facilitate the competition ELISA assay, we used the hAREG-EGFd-DLLA-mFc (hAREG-DLLA) or mAREG-EGFd-DLLA-mFc (mAREG-DLLA) as the ligands for binding to EGFR in the assay. The results showed that E1H3L4, P7 and hu9C12v4 all competed with mAREG-DLLA for binding to mEGFR-ECD. hu9C12v4 showed the best activity among these three mAbs. hu9C12v4, hu9C12v6, hu23H8v5, hu23H8v6 and hu1H9v3 in their hIgG1 forms were also tested in the competition ELISA, and they all showed potent activity in competition with hAREG-DLLA for binding to hEGFR-ECD with subnanomolar IC50.

[0192] In addition, we also tested two previously reported antibodies, huPAR34 (U.S. patent application No. 2004/0210040) and AR558 (US20170002068A1) in human IgG1 forms. These two antibodies also showed potent activity in competition with hAREG-DLLA for binding to hEGFR-ECD.

[0193] Anti-AREG mAbs Inhibit EGFR Phosphorylation

[0194] Anti-AREG mAbs were tested for their inhibition activities against EGFR phosphorylation in hEGFR-expressing epidermoid carcinoma cells, A431. Low concentrations of the antibodies were sufficient for blocking AREG-induced phosphorylation of EGFR of A431 cells. 1.2 nM of 23H8 or 1H9 completely blocked hAREG-induced phosphorylation of EGFR. 9C12 showed relatively weaker blocking activity than 23H8 and 1H9 (FIG. 2).

[0195] Epitope Mapping

[0196] To identify the epitopes of our anti-AREG mAbs, five hAREG-EGFd variants were generated by changing each amino acid at five different sites of hAREG-EGFd to the counterpart amino acid of mAREG-EGFd (FIG. 3). The variants were then examined for binding with anti-AREG mAbs using SPR (Biacore T200). Two amino acids (Glu149 and His164) were identified as critical epitope residues for the binding of the mAb to hAREG. As revealed by Biacore analysis, for hu9C12v4, hu9C12v6, hu23H8 and hu1H9 mAbs, the hAREG-H164N variant completely lost binding activity to mAbs, hAREG-E149K variant had slightly reduced binding activity, and other three hAREG variants had no effect on the binding of the mAbs, demonstrating that His164 is the most critical epitope residue for the binding of our anti-AREG mAbs. For huPAR34, E149K and H164N variants had reduced binding activity, other three residue changes have no or minor effect. For AR558, the E149K variant completely lost binding activity, other four residue changes have no or minor effect on the binding of hAREG to AR558, and indicating Glu149 is the most critical epitope residue for AR558.

[0197] In addition, using two mAREG variants, we found that mAREG-K149E/N164H (using hAREG numbering) variant gained full binding affinity for the anti-hAREG antibodies that have no or very weak cross-reactivity to mAREG; the mAREG-N164H variant gained partial binding ability to the antibodies. These results indicate that the amino acids, Lys149 and Asn164 in mAREG, are residues responsible for lack of cross reactivity of the mAbs (hu9C12v6, hu23H8, hu1H9, huPAR34 and AR558) to mAREG.

Example 4. Animal Study

[0198] 1. Establishing Animal Models

[0199] Cdc42 AT2 Null Mice are Generated by Knocking Out Cdc42 Gene Specifically in Alveolar Type II Cells (AT2 Cells)

[0200] In order to specifically delete Cdc42 gene in AT2 cells, mice carrying a Spc-CreER knock-in allele are crossed with Cdc42 floxed (Cdc42.sup.flox/flox) mice (FIG. 4A). In Cdc42.sup.flox/flox mice, the exon 2 of Cdc42 gene, which contains the translation initiation exon of Cdc42 gene, is flanked by two loxp sites. In Spc-CreER; Cdc42.sup.flox/flox mice the exon 2 of Cdc42 gene, exon 2 of Cdc42 gene is specifically deleted in AT2 cells by Cre/loxp-mediated recombination after tamoxifen treatment (FIG. 4B). Spc-CreER; Cdc42.sup.flox/flox mice are named as Cdc42 AT2 null mice. The fragments of Cdc42 DNA sequence before and after deleting the exon2 of the Cdc42 gene are shown as follows. All these mice were maintained in the animal facility in specific pathogen free conditions.

[0201] The Cdc42 sequence before deleting the exon2 of the Cdc42 gene is shown in SEQ ID NO: 133. The Cdc42 sequence after deleting the exon2 of the Cdc42 gene is shown in SEQ ID NO: 134.

[0202] Lungs of Cdc42 AT2 Null Mice Develop Progressive Fibrotic Changes after PNX Treatment

[0203] Left lung lobe resection (peumonectomy, PNX) on Cdc42 AT2 null mice and control mice were performed. The lungs of Cdc42 AT2 null mice and control mice at different time points after PNX treatment were analyzed (FIG. 5A). We found that some Cdc42 AT2 null mice showed significant weight loss and increased respiration rates after post-PNX day 21. Indeed, fully 50% of PNX-treated Cdc42 AT2 null mice reached the predefined health-status criteria for endpoint euthanization by post-PNX day 60 (FIG. 5B), and more than 70% of PNX-treated Cdc42 AT2 null mice (n=33) reached their endpoints by post-PNX day 180 (FIG. 5B). H&E staining shows lungs of sham-treated and PNX-treated control mice do not shown fibrotic changes (FIG. 5C). H&E staining shows that the entire lung lobes of PNX-treated Cdc42 AT2 null mice at endpoints have dense fibrotic changes (FIG. 5D).

[0204] The lungs of Cdc42 AT2 null mice start to show fibrotic changes at post-PNX day 21. The Cdc42 AT2 null lungs have shown dense fibrotic changes at the edge of lungs (FIG. 5D). H&E staining shows that histological changes of the fibrotic region of Cdc42 AT2 null lungs recapitulate the histological changes of human IPF lungs.

[0205] Lungs collected from Control and Cdc42 AT2 null mice at post-PNX day 21 were stained with an anti-Collagen I antibody (FIG. 5E). Much stronger immunofluorescence signals for Collagen I are detected in the dense fibrotic regions of lungs of Cdc42 AT2 null mice as compared with control lungs. The area of dense Collagen I in lungs of Cdc42 AT2 null mice gradually increases from post-PNX day 21 to post-PNX day 60 (FIG. 5F). qPCR analysis showed that the Collagen I mRNA expression levels increased gradually from post-PNX day 21 to post-PNX day 60 in the lungs of Cdc42 AT2 null mice (FIG. 5G). Respiratory function analysis shows that the lung compliance gradually decreased in Cdc42 AT2 null mice from post-PNX day 21 to post-PNX day 60 (FIG. 5H). *P<0.05, ***P<0.001; ****P<0.0001, Student's t test.

[0206] This is the first mouse model that can highly mimic the pathogenesis and progression of IPF. Therefore, hereafter called IPF-like lung fibrosis mouse model. Using this animal model, we identified that AREG is a potential therapeutic target for pulmonary fibrosis.

[0207] Bleomycin-Induced Lung Fibrosis Mouse Model

[0208] Bleomycin-induced pulmonary fibrosis is a common experimental study model of human lung fibrosis. Wild type FVB/N mice (Charles River) in each group were intratracheally instilled with a single dose of BLM (1U/1 KG body weight, H20055883, Hai Zheng Pfizer Inc). The BLM-treated mice in all groups were closed monitored at different time points after the bleomycin administration.

[0209] This is an animal model that can recapitulate acute lung injury-induced lung fibrosis, such as post-pneumonia lung fibrosis or ILD (interstitial lung diseases). Bleomycin induces lung injury via oxidant-mediated DNA breaks, leading to alveolar epithelial cell death (1-3 days after injury) and acute inflammatory responses 3-9 days after injury). And lung fibrosis occurs in the lungs at 10-21 days post-injury.

[0210] We adopted our IPF-like mouse model and bleomycin-induced lung fibrosis mouse model to explore the therapeutic effects of our AREG antibodies. Furthermore, we also compared the therapeutic effects of two drugs, Nintedanib and Pirfenidone, in order to comprehensively evaluate the potential therapeutic effects of AREG antibodies with existing FDA approved drugs.

[0211] 2. The Animal Study Design and Analysis of Treatments of Pulmonary Fibrosis in the Mouse Models

[0212] 1) IPF-like lung fibrosis mouse model: Three-month-old male Cdc42 AT2 null mice with a similar body weight (˜30 g) were selected for the experiments. The mice were injected with tamoxifen intraperitoneally (dosage: 75 mg/kg) four times every other day. Two weeks after the last injection, the mice were treated with PNX. Post-PNX day 14 is the time-point when fibrosis starts. At post-PNX 14 days, PNX-treated mice were weighed and proceed to the treatments.

[0213] 2) Bleomycin-induced lung fibrosis mouse model: Three-month-old male FVB/N mice with a similar body weight (˜30 g) were selected for the bleomycin treatment. Specifically, an endotracheal cannula was inserted into the trachea of anesthetized mice before delivering bleomycin solution (dosage: 1U/kg). Then one day after the bleomycin delivery, mice were weighed and proceed to treatments.

[0214] 3) Treatment groups: Mice were divided into different groups: Control, anti-AREG antibody, Nintedanib, and Pirfenidone groups. All mice in each group are age matched and body weight matched. For control groups, mice were treated with an isotype-match control antibody. Control antibody or anti-AREG antibodies were given intraperitoneally at 10-15 mg/kg, every 5 days. In addition, mice in the control group were treated with 0.5% sodium methylcellulose sodium solution via oral gavage once a day. Mice in the Nintedanib group were treated with Nintedanib via oral gavage once a day (60 mg/kg). Mice in the Pirfenidone treatment group were treated with Pirfenidone (100 mg/kg) via oral gavage once a day. Mice in the Nintedanib group and Pirfenidone group were also treated with PBS solution intraperitoneally every five days.

[0215] 4) Animal study

[0216] a) The body weight of the mice in all groups was monitored every other day. The general health condition of mice in all groups is closely monitored twice a day.

[0217] b) The humane endpoint is defined by the loss of overall body weight (30% of their initial body weight).

[0218] Animal studies were conducted under the approved Institutional Animal Care and Use Committee protocols. Lung tissues were collected at the endpoint of the study. The hydroxyproline content in each mouse lung was measured by a hydroxyproline kit (Sigma, Cat #MAK008). Lung fibrosis scale was evaluated using the histology analysis. Lung tissues were fixed by 4% PFA, sectioned and proceed for H&E staining. The final histological fibrosis scores were assigned by analyzing various fields of the lung.

[0219] 3. Results

[0220] 1) Anti-AREG antibody (P7): Our results show the anti-AREG (P7) antibody can significantly slow the weight loss of the Cdc42 AT2 null and can prolong the survival time of Cdc42 AT2 null mice (FIG. 6A-6C). In addition, the anti-AREG antibody (P7) can significantly reduce the content of hydroxyproline in the lungs of the Cdc42 AT2 null mice (FIG. 6D). FIG. 6A shows outline scheme of the treatment and sampling procedure, FIG. 6B shows that the anti-AREG antibody (P7) can prolong the survival time of Cdc42 AT2 null mice, FIG. 6C shows that the anti-AREG antibody (P7) can significantly slow down the weight loss of the Cdc42 AT2 null mice, and FIG. 6D shows that the anti-AREG antibody (P7) can significantly reduce the content of hydroxyproline in the lungs of Cdc42 AT2 null mice as compared with mice treated with the blank antibody (*, P<0.05, Student's t test).

[0221] 2) Anti-AREG antibody (E1H3L4): Our results show that the anti-AREG antibody (E1H3L4) can accelerate the resolution of fibrosis and promote the weight recovery in the bleomycin-treated mice (FIG. 7B). FIG. 7A shows that the survival rates of mice treated with the blank antibody and mice treated with the anti-AREG antibody (E1H3L4) are not significantly different in the bleomycin-induced lung fibrosis mouse model. However, FIG. 7B shows that the mice in the anti-AREG antibody (E1H3L4) treatment group recovered better than the mice in the blank antibody treatment group.

[0222] In addition, the anti-AREG (E1H3L4) antibody treatment can significantly prolong the survival time of Cdc42 AT2 null mice (FIG. 8A-8B). The H&E staining analysis showed that the area of mouse lung fibrosis in the anti-AREG (E1H3L4) treatment group was significantly reduced in the lungs of Cdc42 AT2 null mice (FIG. 8C). FIG. 8A shows the outline scheme of the treatment and sampling procedure. FIG. 8B shows that the anti-AREG antibody (E1H3L4) can significantly prolong the survival time of Cdc42 AT2 null mice, and FIG. 8C shows that the lung fibrosis was significantly reduced in the mice of the anti-AREG antibody (E1H3L4) treatment group as compared with the mice in the control group through the H&E staining analysis.

[0223] 3) Anti-AREG antibody (hu9C12v4): Our results show that the anti-AREG (hu9C12v4) can significantly prolong the survival time of Cdc42 AT2 null mice (FIG. 9A-9B), whereas ninetadnib and pirfenidone do not significantly prolong the survival time of Cdc42 AT2 null mice. The H&E staining analysis showed that the area of mouse lung fibrosis in the anti-AREG antibody (hu9C12v4) treatment group was significantly reduced in the lungs of the Cdc42 AT2 null mice (FIG. 9C). Specifically, FIG. 9A shows the outline scheme of the treatment and sampling procedure, FIG. 9B shows that the anti-AREG antibody (hu9C12v4) treatment can significantly prolong the survival time of Cdc42 AT2 null mice, and FIG. 9C shows that the lung fibrosis was significantly reduced in the mice of the anti-AREG antibody (hu9C12v4) treatment group as compared with the mice in the control, Nintedanib, and Pirfenidone groups through the H&E staining analysis.

[0224] Taking together, these results demonstrate that our anti-AREG monoclonal antibodies are effective for treating lung fibrosis.

REFERENCE

[0225] 1 Barkauskas, C. E., and Noble, P. W. (2014). Cellular mechanisms of tissue fibrosis. 7. New insights into the cellular mechanisms of pulmonary fibrosis. American journal of physiology Cell physiology 306, C987-996. [0226] 2 Li, D., He, W., Liu, X., Zheng, S., Qi, Y., Li, H., Mao, F., Liu, J., Sun, Y., Pan, L., et al. (2017). A potent human neutralizing antibody Fc-dependently reduces established HBV infections. eLife 6: e26738. [0227] 3 Steele, M. P., and Schwartz, D. A. (2013). Molecular mechanisms in progressive idiopathic pulmonary fibrosis. Annual review of medicine 64, 265-276. [0228] 4 Thompson, S. A., Harris, A., Hoang, D., Ferrer, M., and Johnson, G. R. (1996). COOH-terminal extended recombinant amphiregulin with bioactivity comparable with naturally derived growth factor. The Journal of biological chemistry 271, 17927-17931.