WO2016073157A1

WO2016073157A1 - Anti-ang2 antibodies and methods of use thereof

Info

Publication number: WO2016073157A1
Application number: PCT/US2015/055672
Authority: WO
Inventors: Seth HARRIS; Patrick Koenig; Chingwei Vivian LEE; Sarah SANOWAR; Ping Wu; Germaine Fuh
Original assignee: Genentech, Inc.; F. Hoffmann-La Roche Ag
Priority date: 2014-11-06
Filing date: 2015-10-15
Publication date: 2016-05-12

Abstract

The invention provides anti-Ang2 antibodies, including dual-specific antibodies that specifically bind Ang2 and VEGF, and methods of using the same for treatment of pathological disorders associated with angiogenesis. The invention also provides methods of identifying improved antibody variants (e.g., with enhanced binding affinity, expression, and/or stability).

Description

ANTI-ANG2 ANTIBODIES AND METHODS OF USE THEREOF

SEQUENCE LISTING

The instant application contains a Sequence Listing which has been submitted electronically in ASCI I format and is hereby incorporated by reference in its entirety. Said ASCI I copy, created on

October 7, 2015 is named 50474_086WO3_Sequence_Listing_1 0_7_15_ST25 and is 101 ,858 bytes in size.

FIELD OF THE INVENTION

The invention relates to anti-Angiopoietin-2 (Ang2) antibodies, and methods of using the same, including for treatment of disorders associated with pathological angiogenesis. The invention also relates to methods of identifying antibody variants (e.g., with enhanced binding affinity, expression, and/or stability). BACKGROUND

Angiogenesis is a tightly-regulated process through which new blood vessels form from preexisting blood vessels. Although angiogenesis is important during development to ensure adequate blood circulation, many disorders are associated with pathological angiogenesis, such as certain ocular disorders and cell proliferative disorders. For example, in age-related macular degeneration (AMD), a condition affecting more than 1 million persons in the United States of America, choroidal

neovascularization (CNV) and vascular permeability lead to the death of photoreceptor cells and subsequent vision loss. Vascular endothelial growth factor (VEGF) is a clinically validated driver of CNV and neutralization of VEGF, for example using an anti-VEGF blocking antibody, leads to stable or improved vision by the majority of treated patients.

The growth factors Angiopoietin-1 (Ang1 ) and Angiopoietin-2 (Ang2) and their receptor tyrosine kinase Tie2 are involved in vascular stability. Signaling of Ang1 through Tie2 can stabilize vascular structures, while the binding of Ang2 to Tie2 in many cases inhibits these stabilizing signals, especially in the absence of VEGF. However, when present in conjunction with VEGF, Ang2 can promote neovascularization.

SUMMARY

The present invention relates to anti-Angiopoietin-2 (Ang2) antibodies, and methods of using the same. The present invention also relates to methods of identifying antibody variants (e.g., with enhanced binding affinity, expression, and/or stability).

In one aspect, the invention features an isolated antibody that specifically binds angiopoietin-2

(Ang2), wherein the antibody binds to an epitope on Ang2 comprising one or more amino acid residues selected from the group consisting of Cys433, Cys435, Met440, Leu441 , Cys450, and Gly451 of Ang2. In some embodiments, the epitope further comprises one or more additional amino acid residues selected from the group consisting of Phe469, Tyr475, and Ser480 of Ang2. In some embodiments, the epitope further comprises one or more additional amino acid residues selected from the group consisting of Lys432, Ile434, Asp448, Ala449, Pro452, and Tyr476 of Ang2. In some embodiments, the epitope consists of amino acid residues Lys432, Cys433, Ile434, Cys435, Met440, Leu441 , Asp448, Ala449, Cys450, Gly451 , Pro452, Phe469, Tyr475, Tyr476, and Ser480 of Ang2.

In another aspect, the invention features an isolated antibody that specifically binds Ang2, wherein the antibody comprises a paratope comprising one or more amino acid residues selected from the group consisting of light chain variable region amino acid residues Gln27; Phe27a; Leu28, Met28, or Ala28; Ser29; Ser30; Phe31 ; Ser67; Gly68; Gly91 ; Leu92; Leu93, Lys93, or Val93; Ser94 or Pro94; and Leu96 and the heavy chain variable region amino acid residues Trp33; His35, Tyr35, or Asp35; Tyr58, Ile58, Trp58, or Leu58; Phe97; Phe98; Leu99 or Ala99; and TyM OOa. In some embodiments, the paratope consists of light chain variable region amino acid residues Gln27; Phe27a; Leu28, Met28, or Ala28; Ser29; Ser30; Phe31 ; Ser67; Gly68; Gly91 ; Leu92; Leu93, Lys93, or Val93; Ser94 or Pro94; and Leu96 and the heavy chain variable region amino acid residues Trp33; His35, Tyr35, or Asp35; Tyr58, Ile58, Trp58, or Leu58; Phe97; Phe98; Leu99 or Ala99; and Tyrl OOa.

In another aspect, the invention features an isolated antibody that specifically binds Ang2, wherein the antibody comprises the following six hypervariable regions (HVRs) : (i) an HVR-L1 comprising the amino acid sequence of RASQFX₁SX₂FGX₃X₄ (SEQ ID NO: 26), wherein X^ is Leu, Met, or Ala, X₂ is Ser, Lys, or Thr, X₃ is Val or Leu, and X₄ is Ala, Ser, Met, Gly, Thr, or Asn; (ii) an HVR-L2 comprising the amino acid sequence of GX₁X₂X₃LX₄X5 (SEQ ID NO: 27), wherein X^ is Ala, Ser, or Gly, X₂ is Arg, Ser, Leu, or Lys, X₃ is Ser, Ala, or Gly, X₄ is Tyr, Val, Ala, or Glu, and X₅ is Ser, Gly, or Gin; (iii) an HVR-L3 comprising the amino acid sequence of X₁QX₂X₃X₄X₅X₆LT (SEQ ID NO: 28), wherein X^ is His, Gin, Phe, Trp, Tyr, or Met, X₂ is Gly, Met, or Phe, X₃ is Leu, Pro, or Ser, X₄ is Val, Leu, lie, Gly, Lys, or Arg, X₅ is Ser, His, Leu, or Pro, and X₆ is Pro or Asp; (iv) an HVR-H1 comprising the amino acid sequence of

is Tyr or Ala, X₂ is Trp or Pro, X₃ is lie, Met, or Gin, and X₄ is His, Tyr, Trp, or Asp; (v) an HVR-H2 comprising the amino acid sequence of

X₁X₂X₃X₄X₅X₆GX₇X₈X₉YADSVKG (SEQ ID NO: 30), wherein X^ is Gly or Lys, X₂ is lie, Gly, or Trp, X₃ is Thr, Val, or Asp, X₄ is Pro, Leu, or Glu, X₅ is Asp, Ala, or Leu, X₆ is Gly or Glu, X₇ is Tyr or Ala, X₈ is Thr, Glu, His, or Asp; and X₉ is Tyr, Leu, Trp, lie, or Lys; and (vi) an HVR-H3 comprising the amino acid sequence of

is Phe, Tyr, or Met, X₂ is Val or Thr, X₃ is Phe or Pro, X₄ is Phe or Pro, X₅ is Leu or Ala, X₆ is Tyr or Trp; X₇ is Ala, Thr, Val, or Ser, and X₈ is Met, Tyr, or Trp. In some embodiments, the antibody comprises the following six HVRs: (i) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (ii) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (iii) an HVR-L3 comprising the amino acid sequence of QQGLLSPLT (SEQ ID NO: 9) ; (iv) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (v) an HVR-H2 comprising the amino acid sequence of G ITPAGG YTYYADSVKG (SEQ ID NO: 6) ; and (vi) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7). In some embodiments, the antibody comprises the following six HVRs: (i) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVG (SEQ ID NO: 32) ; (ii) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (iii) an HVR-L3 comprising the amino acid sequence of

WQGLLSPLT (SEQ ID NO: 33) ; (iv) an HVR-H1 comprising the amino acid sequence of DYWIY (SEQ ID NO: 34) ; (v) an HVR-H2 comprising the amino acid sequence of G ITPAGG YE YYADSVKG (SEQ ID NO: 35) ; and (vi) an HVR-H3 comprising the amino acid sequence of FVFFLPYVMDY (SEQ ID NO: 36). In some embodiments, the antibody comprises the following six HVRs: (i) an HVR-L1 comprising the amino acid sequence of RASOFLSSFGVA (SEQ ID NO: 2) ; (ii) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (iii) an HVR-L3 comprising the amino acid sequence of HQGLKSPLT (SEQ ID NO: 37) ; (iv) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (v) an HVR- H2 comprising the amino acid sequence of GITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and (vi) an HVR- H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7). In some embodiments, the antibody comprises the following six HVRs: (i) an HVR-L1 comprising the amino acid sequence of RASOFLSSFGVA (SEQ ID NO: 2) ; (ii) an HVR-L2 comprising the amino acid sequence of GARSLYS (SEQ ID NO: 39) ; (iii) an HVR-L3 comprising the amino acid sequence of HQGLVSPLT (SEQ ID NO: 40) ; (iv) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (v) an HVR-H2 comprising the amino acid sequence of GITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and (vi) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7). In some embodiments, the antibody further comprises the following heavy chain variable region framework regions (FRs) : (i) an FR- H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFPIS (SEQ ID NO: 41 ) ; (ii) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (iii) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and (iv) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44). In some embodiments, the antibody further comprises the following heavy chain variable region framework regions (FRs) : (i) an FR-H1 comprising the amino acid sequence of

EVQLVESGGGLVQPGGSLRLSCAASGFTIM (SEQ ID NO: 45) ; (ii) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (iii) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and (iv) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44).

In another aspect, the invention features an isolated antibody that specifically binds Ang2, wherein the antibody comprises (a) a light chain variable region (V_L) having at least 95% sequence identity to the amino acid sequence of SEQ ID NO: 1 0, 1 1 , 12, 13, 14, 15, 16, 17, 46, 48, 51 , 78, or 79; (b) a heavy chain variable region (V_H) having at least 95% sequence identity to the amino acid sequence of SEQ ID NO: 18, 47, 49, or 50; or (c) a light chain variable region as in (a) and a heavy chain variable region as in (b). In some embodiments, the antibody comprises a V_H sequence of SEQ ID NO: 49. In some embodiments, the antibody comprises a V_L sequence of SEQ ID NO: 51 .

In another aspect, the invention features an isolated antibody that specifically binds Ang2 and vascular endothelial growth factor (VEGF), wherein the antibody binds to a region within amino acid residues 432-480 of human Ang2 polypeptide (SEQ ID NO: 1 ).

In another aspect, the invention features an isolated antibody that specifically binds Ang2 and

VEGF, wherein the antibody binds to an epitope on Ang2 comprising one or more amino acid residues selected from the group consisting of Lys432, Cys433, Ile434, Cys435, Met440, Leu441 , Asp448, Ala449, Cys450, Gly451 , Pro452, Phe469, Tyr475, Tyr476, and Ser480 of Ang2. In some embodiments, the epitope comprises three or more amino acid residues selected from the group consisting of Lys432, Cys433, Ile434, Cys435, Met440, Leu441 , Asp448, Ala449, Cys450, Gly451 , Pro452, Phe469, Tyr475, Tyr476, and Ser480 of Ang2. In some embodiments, the epitope consists of amino acid residues Lys432, Cys433, Ile434, Cys435, Met440, Leu441 , Asp448, Ala449, Cys450, Gly451 , Pro452, Phe469, Tyr475, Tyr476, and Ser480 of Ang2.

In another aspect, the invention features an isolated antibody that specifically binds Ang2 and VEGF, wherein the antibody binds to an epitope on VEGF comprising one or more amino acid residues selected from the group consisting of Phe17, Met18, Tyr21 , Gln22, Tyr25, Lys48, Asn62, Asp63, Glu64, Gly65, Leu66, Met81 , Ile83, Lys84, Pro85, His86, Gln87, Gly88, Gln89, His90, Ile91 , Lys101 , Glu103, Cys104, and Pro106 of human VEGF. In some embodiments, the epitope comprises amino acid residues Phe17, Tyr21 , and Tyr25 of human VEGF. In some embodiments, the epitope comprises amino acid residues Phe17, Ile81 , and Gln89 of human VEGF. In some embodiments, the epitope consists of

Phe17, Met18, Tyr21 , Gln22, Tyr25, Lys48, Asn62, Asp63, Glu64, Gly65, Leu66, Met81 , Ile83, Lys84, Pro85, His86, Gln87, Gly88, Gln89, His90, Ile91 , Lys101 , Glu103, Cys104, and Pro106 of human VEGF.

In another aspect, the invention features an isolated antibody that specifically binds Ang2 and VEGF, wherein the antibody comprises a paratope that binds to Ang2, wherein the paratope comprises one or more amino acid residues selected from the group consisting of light chain variable region amino acid residues Gln27; Phe27a; Leu28, Met28, or Ala28; Ser29; Ser30; Phe31 ; Ser67; Gly68; Gly91 ; Leu92; Leu93, Lys93, or Val93; Ser94 or Pro94; and Leu96 and the heavy chain variable region amino acid residues Trp33; His35, Tyr35, or Asp35; Tyr58, Ile58, Trp58, or Leu58; Phe97; Phe98; Leu99 or Ala99; and Tyr1 00a. In some embodiments, the paratope consists of light chain variable region amino acid residues Gln27; Phe27a; Leu28, Met28, or Ala28; Ser29; Ser30; Phe31 ; Ser67; Gly68; Gly91 ;

Leu92; Leu93, Lys93, or Val93; Ser94 or Pro94; and Leu96 and the heavy chain variable region amino acid residues Trp33; His35, Tyr35, or Asp35; Tyr58, Ile58, Trp58, or Leu58; Phe97; Phe98; Leu99 or Ala99; and Tyr1 00a.

In another aspect, the invention features an isolated antibody that specifically binds Ang2 and VEGF, wherein the antibody comprises a paratope that binds to VEGF, wherein the paratope comprises one or more amino acid residues selected from the group consisting of light chain variable region amino acid residues Leu28, Met28 or Ala28; Ser29; Phe31 ; Tyr49; Ser53; and Leu92 and the heavy chain variable region amino acid residues Ser30, Gly30, or Met30; Asp31 ; Tyr32 or Ala32; Trp33; Ile51 ; Thr52; Pro52a or Glu52a; Ala53 or Asp53; Gly54; Gly55; Tyr56 or Ala56; Phe95 or Met95; Val96 or Thr96; Phe97; Phe98; Leu99 or Ala99; and TyM OOa. In some embodiments, the paratope consists of light chain variable region amino acid residues Leu28, Met28, or Ala28; Ser29; Phe31 ; Tyr49; Ser53; and Leu92 and the heavy chain variable region amino acid residues Ser30, Gly30, or Met30; Asp31 ; Tyr32 or Ala32; Trp33; Ile51 ; Thr52; Pro52a or Glu52a; Ala53 or Asp53; Gly54; Gly55; Tyr56 or Ala56; Phe95 or Met95; Val96 or Thr96; Phe97; Phe98; Leu99 or Ala99; and Tyr100a.

VEGF, wherein the antibody comprises a paratope that binds to VEGF and Ang2, wherein the paratope comprises one or more amino acid residues selected from the group consisting of light chain variable region amino acid residues Ser30, Phe31 , and Leu92 and the heavy chain variable region amino acid residues Leu99 or Ala99 and Pro100. In some embodiments, the paratope consists of the light chain variable region amino acid residues Ser30, Phe31 , and Leu92 and the heavy chain variable region amino acid residues Leu99 or Ala99 and Prol OO.

In another aspect, the invention features an isolated antibody that specifically binds Ang2 and VEGF, wherein the antibody comprises the following six HVRs: (i) an HVR-L1 comprising the amino acid sequence of RASQFX₁SX₂FGX₃X₄ (SEQ ID NO: 26), wherein X^ is Leu, Met, or Ala, X₂ is Ser, Lys, or Thr, X₃ is Val or Leu, and X₄ is Ala, Ser, Met, Gly, Thr, or Asn; (ii) an HVR-L2 comprising the amino acid sequence of GX₁X₂X₃LX₄X5 (SEQ ID NO: 27), wherein X^ is Ala, Ser, or Gly, X₂ is Arg, Ser, Leu, or Lys, X₃ is Ser, Ala, or Gly, X₄ is Tyr, Val, Ala, or Glu, and X₅ is Ser, Gly, or Gin; (iii) an HVR-L3 comprising the amino acid sequence of X₁QX₂X₃X₄X₅X₆LT (SEQ ID NO: 28), wherein X^ is His, Gin, Phe, Trp, Tyr, or Met, X₂ is Gly, Met, or Phe, X₃ is Leu, Pro, or Ser, X₄ is Val, Leu, lie, Gly, Lys, or Arg, X₅ is Ser, His, Leu, or Pro, and X₆ is Pro or Asp; (iv) an HVR-H1 comprising the amino acid sequence of DX_!X₂X₃X₄ (SEQ ID NO: 29), wherein X^ is Tyr or Ala, X₂ is Trp or Pro, X₃ is lie, Met, or Gin, and X₄ is His, Tyr, Trp, or Asp; (v) an HVR-H2 comprising the amino acid sequence of

(SEQ ID NO: 30), wherein X^ is Gly or Lys, X₂ is lie, Gly, or Trp, X₃ is Thr, Val, or Asp, X₄ is Pro, Leu, or Glu, X₅ is Asp, Ala, or Leu, X₆ is Gly or Glu, X₇ is Tyr or Ala, X₈ is Thr, Glu, His, or Asp; and X₉ is Tyr, Leu, Trp, lie, or Lys; and (vi) an HVR-H3 comprising the amino acid sequence of

(SEQ ID NO: 31 ), wherein X^ is Phe, Tyr, or Met, X₂ is Val or Thr, X₃ is Phe or Pro, X₄ is Phe or Pro, X₅ is Leu or Ala, X₆ is Tyr or Trp; X₇ is Ala, Thr, Val, or Ser, and X₈ is Met, Tyr, or Trp. In some embodiments, the antibody comprises the following six HVRs: (i) an HVR-L1 comprising the amino acid sequence of

RASQFLSSFGVA (SEQ ID NO: 2) ; (ii) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (iii) an HVR-L3 comprising the amino acid sequence of QQGLLSPLT (SEQ ID NO: 9) ; (iv) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (v) an HVR-H2 comprising the amino acid sequence of GITPAGGYTYYADSVKG (SEQ ID NO: 6) ; and (vi) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7). In some embodiments, the antibody comprises the following six HVRs: (i) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVG (SEQ ID NO: 32) ; (ii) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (iii) an HVR-L3 comprising the amino acid sequence of WQGLLSPLT (SEQ ID NO: 33) ; (iv) an HVR-H1 comprising the amino acid sequence of DYWIY (SEQ ID NO: 34) ; (v) an HVR-H2 comprising the amino acid sequence of GITPAGGYEYYADSVKG (SEQ ID NO: 35) ; and (vi) an HVR-H3 comprising the amino acid sequence of FVFFLPYVMDY (SEQ ID NO: 36). In some embodiments, the antibody comprises the following six HVRs: (i) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (ii) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (iii) an HVR-L3 comprising the amino acid sequence of HQGLKSPLT (SEQ ID NO: 37) ; (iv) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (v) an HVR-H2 comprising the amino acid sequence of GITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and (vi) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7). In some embodiments, the antibody comprises the following six HVRs: (i) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (ii) an HVR-L2 comprising the amino acid sequence of GARSLYS (SEQ ID NO: 39) ; (iii) an HVR-L3 comprising the amino acid sequence of HQGLVSPLT (SEQ ID NO: 40) ; (iv) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (v) an HVR-H2 comprising the amino acid sequence of GITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and (vi) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7). In some embodiments, the antibody further comprises the following heavy chain variable region FRs: (i) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFPIS (SEQ ID NO: 41 ) ; (ii) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (iii) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and (iv) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44). In some embodiments, the antibody further comprises the following heavy chain variable region FRs: (i) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFTIM (SEQ ID NO: 45) ; (ii) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (iii) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and (iv) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44).

VEGF, wherein the antibody comprises (a) a V_L having at least 95% sequence identity to the amino acid sequence of SEQ ID NO: 10, 1 1 , 12, 13, 14, 15, 16, 17, 46, 48, 51 , 78, or 79; (b) a V_H having at least 95% sequence identity to the amino acid sequence of SEQ ID NO: 18, 47, 49, or 50; or (c) a light chain variable region as in (a) and a heavy chain variable region as in (b). In some embodiments, the antibody comprises a V_H sequence of SEQ ID NO: 49. In some embodiments, the antibody comprises a V_L sequence of SEQ ID NO: 51 .

In another aspect, the invention features an isolated antibody that competes for binding to Ang2 with any one of the preceding antibodies.

In another aspect, the invention features an isolated antibody that competes for binding to Ang2 and VEGF with any one of the preceding antibodies.

In another aspect, the invention features an isolated antibody that binds to the same epitope as any one of the preceding antibodies.

In some embodiments, any one of the preceding antibodies can bind VEGF with a Kd of about 15 nM or lower and Ang2 with a Kd of about 15 nM or lower. In some embodiments, any one of the preceding antibodies can bind VEGF with a Kd of about 10 nM or lower and Ang2 with a Kd of about 10 nM or lower. In some embodiments, any one of the preceding antibodies can bind VEGF with a Kd of about 5 nM or lower and Ang2 with a Kd of about 5 nM or lower. In some embodiments, any one of the preceding antibodies can bind VEGF with a Kd of lower than 1 nM and Ang2 with a Kd of lower than 1 nM. In some embodiments, any one of the preceding antibodies can bind VEGF with a Kd of lower than 0.5 nM and Ang2 with a Kd of lower than 0.5 nM. In some embodiments, any one of the preceding antibodies can bind VEGF with a Kd of lower than 0.25 nM and Ang2 with a Kd of lower than 0.25 nM.

In some embodiments, any one of the preceding antibodies can inhibit or block binding of Ang2 or VEGF to its receptor. In some embodiments, the antibody inhibits or blocks binding of Ang2 or VEGF to its receptor with a IC50 value of 8 nM or lower. In some embodiments, the antibody inhibits or blocks binding of Ang2 to its receptor with an IC50 value of 50 pM to 2 nM. In some embodiments, the antibody inhibits or blocks binding of Ang2 to its receptor with an IC50 value of 75 pM. In some embodiments, the antibody inhibits or blocks binding of VEGF to its receptor with an IC50 value of 50 pM to 2 nM. In some embodiments, the antibody inhibits or blocks binding of VEGF to its receptor with an IC50 value of 85 nM.

In some embodiments, any one of the preceding antibodies can bind to Ang2 with 50-fold greater affinity than to Ang1 . In some embodiments, the antibody can bind to Ang2 with 75-fold greater affinity than to Ang1 . In some embodiments, the antibody can bind to Ang2 with 100-fold greater affinity than to Ang1 .

In some embodiments, any one of the preceding antibodies can be a dual-specific antibody. In some embodiments, any one of the preceding antibodies can be a monoclonal antibody. In some embodiments, any one of the preceding antibodies can be an IgG antibody. In some embodiments, any one of the preceding antibodies can be an antibody fragment that specifically binds VEGF and Ang2. In some embodiments, the antibody fragment is selected from the group consisting of Fab, single chain variable fragment (scFv), Fv, Fab', Fab'-SH, F(ab')₂, and diabody. In some embodiments, the antibody fragment is a Fab. In some embodiments, any one of the preceding antibodies, at least a portion of the framework sequence can be a human consensus framework sequence. In some embodiments, any one of the preceding antibodies can be a chimeric, humanized, or fully human antibody.

In another aspect, the invention features a method of producing the antibody of any of the antibodies described herein, the method comprising culturing a host cell that comprises any of the preceding vectors (e.g., expression vectors) and recovering the antibody. In some embodiments, the host cell is prokaryotic. In certain embodiments, the host cell is Escherichia coli. In other embodiments, the host cell is eukaryotic. In certain embodiments, the host cell is a 293 cell, a CHO cell, a yeast cell, or a plant cell.

In another aspect, the invention features a method of reducing or inhibiting angiogenesis in a subject having a disorder associated with pathological angiogenesis, comprising administering to the subject an effective amount of any one of the preceding antibodies, thereby reducing or inhibiting angiogenesis in the subject. In some embodiments, the disorder associated with pathological angiogenesis is an ocular disorder or a cell proliferative disorder. In certain embodiments, the disorder associated with pathological angiogenesis is an ocular disorder. In some embodiments, the ocular disorder is selected from the group consisting of retinopathy including proliferative diabetic retinopathy, choroidal neovascularization (CNV), age-related macular degeneration (AMD), diabetic and other ischemia-related retinopathies, diabetic macular edema (DME), pathologic myopia, von Hippel-Lindau disease, histoplasmosis of the eye, retinal vein occlusion (including central (CRVO) and branched (BRVO) forms), corneal neovascularization, retinal neovascularization, retinopathy of prematurity (ROP), familial exudative vitreoretinopathy (FEVR), Coats' disease, Norrie Disease, Osteoporosis-Pseudoglioma Syndrome (OPPG), subconjunctival hemorrhage, and hypertensive retinopathy. In certain embodiments, the ocular disorder is AMD.

In another aspect, the invention features a method for treating a disorder associated with pathological angiogenesis, the method comprising administering an effective amount of any one of the preceding antibodies to a subject in need of such treatment. In some embodiments, the disorder associated with pathological angiogenesis is an ocular disorder or a cell proliferative disorder. In certain embodiments, the disorder associated with pathological angiogenesis is an ocular disorder. In some embodiments, the ocular disorder is selected from the group consisting of retinopathy including proliferative diabetic retinopathy, choroidal neovascularization (CNV), age-related macular degeneration (AMD), diabetic and other ischemia-related retinopathies, diabetic macular edema (DME), pathologic myopia, von Hippel-Lindau disease, histoplasmosis of the eye, retinal vein occlusion (including central (CRVO) and branched (BRVO) forms), corneal neovascularization, retinal neovascularization, retinopathy of prematurity (ROP), familial exudative vitreoretinopathy (FEVR), Coats' disease, Norrie Disease, Osteoporosis-Pseudoglioma Syndrome (OPPG), subconjunctival hemorrhage, and hypertensive retinopathy. In some embodiments, the ocular disorder is AMD.

In some embodiments of the preceding methods of reducing or inhibiting angiogenesis in a subject having a disorder associated with pathological angiogenesis or for treating a disorder associated with pathological angiogenesis, the method comprises administering to the subject an effective amount of a second agent, wherein the second agent is selected from the group consisting of another antibody, a chemotherapeutic agent, a cytotoxic agent, an anti-angiogenic agent, an immunosuppressive agent, a prodrug, a cytokine, a cytokine antagonist, cytotoxic radiotherapy, a corticosteroid, an anti-emetic, a cancer vaccine, an analgesic, and a growth-inhibitory agent.

In certain embodiments of the preceding methods, the antibody or antigen-binding fragment thereof is administered intravitreally, by eye drop, subcutaneously, intravenously, intramuscularly, topically, orally, transdermal^, intraperitoneally, intraorbitally, by implantation, by inhalation, intrathecally, intraventricularly, or intranasally. In some embodiments, the administration is intravitreally.

In some embodiments of the preceding methods, the subject is human.

In another aspect, the invention features a pharmaceutical composition comprising any one of the preceding antibodies. In some embodiments, the pharmaceutical composition is used for treating a disorder associated with pathological angiogenesis in a mammal. In some embodiments, the disorder associated with pathological angiogenesis is an ocular disorder or a cell proliferative disorder. In certain embodiments, the disorder associated with pathological angiogenesis is an ocular disorder. In some embodiments, the ocular disorder is selected from the group consisting of retinopathy including proliferative diabetic retinopathy, choroidal neovascularization (CNV), age-related macular degeneration (AMD), diabetic and other ischemia-related retinopathies, diabetic macular edema (DME), pathologic myopia, von Hippel-Lindau disease, histoplasmosis of the eye, retinal vein occlusion (including central

(CRVO) and branched (BRVO) forms), corneal neovascularization, retinal neovascularization, retinopathy of prematurity (ROP), familial exudative vitreoretinopathy (FEVR), Coats' disease, Norrie Disease, Osteoporosis-Pseudoglioma Syndrome (OPPG), subconjunctival hemorrhage, and hypertensive retinopathy. In some embodiments, the ocular disorder is AMD.

In another aspect, the invention features a method of identifying an amino acid residue alteration that confers enhanced binding of an antibody to a target molecule, the method comprising: (a) providing a display library comprising nucleic acids encoding candidate antibody variants, wherein each candidate antibody variant comprises an amino acid residue alteration in each HVR of the heavy chain variable region (V_H) or the light chain variable region (V_L) compared to a reference antibody; (b) sorting the display library based on binding of the candidate antibody variants to the target molecule to form a sorted library, wherein the sorted library comprises candidate antibody variants with enhanced binding to the target molecule compared to the reference antibody; and (c) comparing the frequency at which each amino acid residue alteration is present in the display library and in the sorted library as determined by massively parallel sequencing, thereby determining whether each amino acid residue alteration is enriched in the sorted library compared to the display library, whereby the amino acid residue alteration is identified as conferring enhanced binding to the target molecule if it is enriched in the sorted library compared to the display library. In some embodiments, the method further comprises determining the frequency at which each amino acid alteration is present in the display library and the sorted library by massively parallel sequencing following step (b). In some embodiments, step (c) further comprises comparing the frequency at which a pair comprising a first amino acid residue alteration and a second amino acid residue alteration is present in the display library and in the sorted library, thereby determining whether the pair is enriched, depleted, or neutral in the sorted library compared to the display library. In some embodiments, the amino acid residue alteration is enriched at least 2-fold in the sorted library compared to the display library. In some embodiments, the amino acid residue alteration is enriched at least 4-fold in the sorted library compared to the display library. In some embodiments, the antibody is a dual specific antibody.

In another aspect, the invention features a method of identifying an amino acid residue alteration that allows enhanced binding of a dual specific antibody to both a first epitope and a second epitope, the method comprising: (a) providing a display library comprising nucleic acids encoding candidate antibody variants, wherein each candidate antibody variant comprises an amino acid residue alteration in each HVR of the V_H or the V_L compared to a reference dual specific antibody; (b) sorting the display library based on binding of the candidate antibody variants to the first epitope to form a first sorted library, wherein the first sorted library comprises candidate antibody variants with enhanced binding to the first epitope compared to the reference dual specific antibody; (c) sorting the display library based on binding of the candidate antibody variants to the second epitope to form a second sorted library, wherein the second sorted library comprises candidate antibody variants with enhanced binding to the second epitope compared to the reference dual specific antibody; and (d) comparing the frequency at which each amino acid residue alteration is present in the display library, the first sorted library, and the second sorted library as determined by massively parallel sequencing, thereby determining whether each amino acid residue alteration is enriched, depleted, or neutral in the first sorted library and the second sorted library compared to the display library, whereby the amino acid residue alteration is identified as allowing enhanced binding of the dual specific antibody to both the first epitope and the second epitope if the amino acid residue alteration is enriched in both the first sorted library and the second sorted library compared to the display library or is enriched in one of either the first sorted library or the second sorted library and is neutral in the other sorted library. In some embodiments, the method further comprises determining the frequency at which each amino acid residue alteration is present in the display library, the first sorted library, and the second sorted library by massively parallel sequencing following step (c). In some embodiments, step (d) further comprises comparing the frequency at which a pair comprising a first amino acid residue alteration and a second amino acid residue alteration is present in the display library and in the first sorted library, the second sorted library, or both, thereby determining whether the pair is enriched, depleted, or neutral in the first sorted library, second sorted library, or both, compared to the display library. In some embodiments, the amino acid residue alteration is enriched at least 2-fold in the first sorted library or second sorted library compared to the display library. In some embodiments, the amino acid residue is enriched at least 4-fold in the sorted library, first sorted library, or second sorted library compared to the display library. In some embodiments, the first epitope and the second epitope are from the same target molecule. In some embodiments, the first epitope is from a first target molecule and the second epitope is from a second target molecule. In some embodiments, the first target molecule and the second target molecule are cytokines. In some embodiments, the first target molecule is VEGF, and the second target molecule is selected from the group consisting of Ang2, Ang1 , PDGF-B, PDGF-C, Stromal-derived growth factor- 1 , placental growth factor (PIGF), factor D, and complement factor 1 . In some embodiments, the first target molecule is VEGF and the second target molecule is Ang2.

In some embodiments of any of the preceding methods of identifying an amino acid residue alteration that confers enhanced binding of an antibody to a target molecule or that allows enhanced binding of a dual specific antibody to both a first epitope and a second epitope, the display library comprises candidate antibody variants having amino acid residue alterations at every position in each HVR of the V_H or V_L. In some embodiments, the display library comprises amino acid residue alterations in only the V_H or the V_L of the candidate antibody variants. In some embodiments, the display library comprises amino acid residue alterations in the V_H and the V_L of the candidate antibody variants. In some embodiments, the display library comprises a V_H library and a V_L library, wherein the V_H library comprises candidate antibody variants with an amino acid residue alteration in each HVR of the V_H, and the V_L library comprises candidate antibody variants with an amino acid residue alteration in each HVR of the V_L. In some embodiments, the display library is selected from the group consisting of a phage display library, a bacterial display library, a yeast display library, a mammalian display library, a ribosome display library, and an mRNA display library. In some embodiments, the display library is a phage display library. In some embodiments, the amino acid residue alteration is encoded by a degenerate codon set. In some embodiments, the degenerate codon set is an NNK or an NNS codon set, wherein N is A, C, G, or T; K^"is G or T; and S is C or G. In some embodiments, the degenerate codon set is an NNK codon set.

In some embodiments of any of the preceding methods of identifying an amino acid residue alteration that confers enhanced binding of an antibody to a target molecule, the sorting of step (b) comprises contacting the display library with an immobilized target molecule. In some embodiments, the sorting of step (b) comprises contacting the display library with a soluble target molecule. In some embodiments, the antibody is a monoclonal antibody. In some embodiments, the antibody is an IgG antibody. In some embodiments, the antibody is an antibody fragment. In some embodiments, the antibody fragment is selected from the group consisting of Fab, scFv, Fv, Fab', Fab'-SH, F(ab')₂, and diabody. In some embodiments, the antibody fragment is a Fab.

In some embodiments of any of the preceding methods of identifying an amino acid residue alteration that allows enhanced binding of a dual specific antibody to both a first epitope and a second epitope, the sorting of step (c) comprises contacting the display library with an immobilized first epitope or second epitope. In some embodiments, the sorting of step (c) comprises contacting the display library with a soluble first epitope or second epitope. In some embodiments, the dual specific antibody is a monoclonal antibody. In some embodiments, the dual specific antibody is an IgG antibody. In some embodiments, the dual specific antibody is an antibody fragment. In some embodiments, the antibody fragment is selected from the group consisting of Fab, scFv, Fv, Fab', Fab'-SH, F(ab')₂, and diabody. In some embodiments, the antibody fragment is a Fab.

In some embodiments of any of the preceding methods of identifying an amino acid residue alteration that confers enhanced binding of an antibody to a target molecule or that allows enhanced binding of a dual specific antibody to both a first epitope and a second epitope, the display library comprises at least 1 x 10⁶ candidate antibody variants. In some embodiments, the display library comprises at least 1 .5 x 10⁷ candidate antibody variants. In some embodiments, the display library comprises at least 2.5 x 10⁷ candidate antibody variants.

In some embodiments of any of the preceding methods of identifying an amino acid residue alteration that confers enhanced binding of an antibody to a target molecule or that allows enhanced binding of a dual specific antibody to both a first epitope and a second epitope, the massively parallel sequencing comprises deep sequencing, ultra-deep sequencing, and/or next-generation sequencing. In some embodiments, the massively parallel sequencing comprises determining the sequence of at least 500,000 reads. In some embodiments, the massively parallel sequencing comprises determining the sequence of at least 1 ,000,000 reads.

In some embodiments of any of the preceding methods of identifying an amino acid residue alteration that confers enhanced binding of an antibody to a target molecule or that allows enhanced binding of a dual specific antibody to both a first epitope and a second epitope, the method further comprises generating an antibody that comprises an amino acid residue alteration identified by the steps of the method.

In another aspect, the invention features a method of generating a dual specific antibody that binds a first epitope with a Kd of lower than 1 nM and a second epitope with a Kd of lower than 1 nM, the method comprising:(a) providing a dual specific antibody that binds the first epitope with a Kd of greater than 1 nM and the second epitope with a Kd of greater than 1 nM; (b) identifying one or more amino acid residue alterations that allows enhanced binding of the dual specific antibody to both the first epitope and the second epitope according to any of the preceding methods, wherein the one or more amino acid residue alterations allows binding the first epitope with a Kd of lower than 1 nM and the second epitope with a Kd of lower than 1 nM; and (c) altering the amino acid sequence of the dual specific antibody based on the results of step (b), thereby generating a dual affinity antibody that binds a first epitope with a Kd of lower than 1 nM and a second epitope with a Kd of lower than 1 nM.

Brief Description of the Drawings

FIGURE 1 is a table showing the degenerate oligonucleotides used for library mutagenesis of the light chain HVRs of the anti-VEGF G6 antibody.

FIGURE 2A is a graph showing the display level of the indicated libraries on the surface of phage as determined by an enzyme-linked immunosorbancy assay (ELISA).

FIGURE 2B is a graph showing the VEGF binding for the indicated phage-displayed libraries as determined by ELISA.

FIGURE 3A is a graph showing the phage IC₅₀ against human Ang2 (Fc.hAng2.RBD) for the indicated DAF clones as determined by competitive ELISA.

FIGURE 3B is a graph showing the phage IC₅₀ against human VEGF₁₀₉ for the indicated DAF clones as determined by competitive ELISA.

FIGURE 3C is a table showing the phage IC₅₀ against Fc.hAng2.RBD (abbreviated "hAng2Fc"), hVEGF₁₀₉ ("hVEGF"), murine Fc.mAng2.RBD ("mAng2Fc"), and human Fc.hAngl .RBD ("hAngl Fc") for the indicated DAF clones as determined by competitive ELISA.

FIGURE 4A is a sequence alignment of the light chain variable domains of the indicated DAF clones compared to the anti-VEGF antibody G6. HVR sequences are delimited by the denoted boxes for each of the DAFs. Residues shown in white text in shaded boxes indicate residues that are different between G6 and the DAF clone.

FIGURE 4B is a sequence alignment of the heavy chain variable domains of the indicated DAF clones compared to the anti-VEGF antibody G6. HVR sequences are delimited by the denoted boxes for each of the DAFs.

FIGURE 5 is a graph summarizing the results of ELISA experiments to determine binding to the indicated antigen. Bovine serum albumin (BSA) serves as a negative control.

FIGURE 6 is a graph showing the results of a receptor-blocking competitive ELISA experiment comparing the Tie2.Fc blocking activity of the anti-Ang2 antibody Ab536 (Amgen) to the DAF clones 5A1 and 5A12 in human IgG format.

FIGURE 7 is a table showing the degenerate oligonucleotides used for library mutagenesis for the affinity maturation of clone 5A12. Soft and limited strategies of randomization were utilized, which allow for wild-type and homologous amino acids based on natural antibodies or -50% of wildtype and 50% of all other amino acids (Bostrom et al. Methods Mol. Biol. 525:353-376, 2009).

FIGURE 8A is a graph showing the phage IC₅₀ against hAng2his8 for the indicated affinity- matured clones compared to 5A12 (labeled 5A12 amber) as determined by competitive ELISA. Affinity- matured clones were selected by screening of libraries based on the phagemid vector encoding 5A12 with an amber stop (TAG) for lower levels of display when expressed in the E. coli suppressor strain XL1 .

FIGURE 8B a graph showing the phage IC₅₀ against hVEGF₁₀₉ for the indicated affinity-matured clones compared to 5A12 (labeled 5A12 amber) as determined by competitive ELISA.

FIGURE 8C is a table showing phage IC₅₀ against hAng2his8 (abbreviated "hAng2") and hVEGF₁₀₉ (abbreviated "hVEGF") for the indicated affinity-matured clones as determined by competitive ELISA.

FIGURE 9A is a sequence alignment of the light chain variable domains of the indicated affinity- matured clones compared to the anti-Ang2/anti-VEGF DAF 5A12. HVR sequences are delimited by the denoted boxes for each of the DAFs. Residues shown in white text in shaded boxes indicate residues that are different between 5A12 and the affinity-matured clone.

FIGURE 9B is a sequence alignment of the heavy chain variable domains of the indicated affinity- matured clones compared to the anti-Ang2/anti-VEGF DAF 5A12. HVR sequences are delimited by the denoted boxes for each of the DAFs.

FIGURE 1 0 is a graph showing the results of a baculovirus ELISA experiment as a measure of non-specific binding and increased risk for fast clearance for the indicated affinity-matured 5A12 variant clones expressed as IgG compared to the control antibodies rituximab and R5D, which are known to have acceptable and fast clearance, respectively (see, e.g., Hotzel et al. Mabs 6:753-760, 2012).

FIGURE 1 1 is a graph showing the results of a receptor-blocking competitive ELISA experiment against Tie2.Fc using the affinity-matured clone 5A12 4.2 expressed as a Fab.

FIGURE 12 is a graph showing the results of a receptor-blocking competitive ELISA experiment against the VEGF receptor Fit comparing 5A12 with affinity-matured variants 5A12 3.4 and 5A12 4.2.

FIGURES 13A and 13B are renderings of the crystal structures of the 5A12 4.2 Fab complexed either to Ang2 (Figure 13A) or VEGF (Figure 13B). The light chain (LC) is shown in light blue and the heavy chain (HC) is shown in dark blue.

FIGURES 13C and 13D are alternative views of the crystal structures of 5A12 4.2 displayed in

Figures 13A and 13B showing the residues of the paratope of the antibody that contacts Ang2 (Figure 13C, shown in left panel in orange) and the paratope of the antibody that contacts VEGF (Figure 13D, shown in right panel in red). The paratopes shown indicate residues that are 4 A away from Ang2 or VEGF, respectively. The light chain is shown in light blue and the heavy chain is shown in dark blue.

FIGURE 14A is a rendering of the crystal structure of Ang2 showing the epitopic region bound by

5A12 4.2 (dotted black lines) and the epitopic region bound by bound by Tie2 (dotted red lines). SC, shape complementarity.

FIGURE 14B is a rendering of the crystal structure of VEGF showing the epitopic region bound by 5A12 4.2 (dotted black lines) and the epitopic region bound by bound by G6 (dotted red lines).

FIGURES 14C and 14D are renderings of superimposed crystal structures of the anti-VEGF antibody G6 alone ("Apo"), G6 bound to VEGF ("G6:VEGF"), 5A12 4.2 bound to Ang2 ("5A12 4.2:Ang2"), and 5A12 4.2 bound to VEGF ("5A12 4.2:VEGF"). The positions of the indicated HVRs (H1 -H3 and L1 - L3) are shown.

FIGURES 15A and 15B is a series of renderings of the crystal structure of 5A12 4.2 bound to either Ang2 (left panel) or VEGF (right panel). The Ang2 binding by 5A12 4.2 does not substantially involve the HVR-H2 loop (yellow), while the HVR-H2 loop is involved in VEGF binding.

FIGURES 16A and 16B are heatmaps showing the enrichment ratios (ER) for all 1040 mutations in the heavy (left) and light chain (right) HVRs obtained from Ang2 (Figure 16A) and VEGF panning (Figure 16B). The line plot shows whether each position is solvent-exposed or buried as a free Fab as in the crystal structure of the 5A12 4.2 Fab in the Ang2-bound (orange line) and in the VEGF-bound form (gray line).

FIGURES 16C and 1 6D are graphs showing a comparison of the enrichment ratios obtained for single mutations of the 5A12 4.2 heavy chain variable region using a 3NNK or a 1 NNK library design panned against VEGF (Figure 16C) or Ang2 (Figure 16D).

FIGURES 17A-17B is a series of renderings of the crystal structure of 5A12 4.2 bound to either

Ang2 (Figures 17A) or VEGF (Figures 1 7B). Figures 17A shows the functional paratope for Ang2 binding, and Figure 17B shows the functional paratope for VEGF binding determined from the deep mutagenesis scanning data. The mean enrichment of all mutations at every mutated HVR position is color-coded (positions in blue indicate that mutations are depleted, indicating positions which do not tolerate mutations, while positions shown in red are those at which mutations are enriched in average) . For every position the enrichment of mutations were calculated by the following equation : log2(FrqMut_Sort/FrqMut_NoSort) , where FrqMut_Sort = 1 - FrqWT_Sort and FrqMut_Noso_rt = 1 - FrqWT_NoSort, FrqWTsort and FrqWT_UnSort is the frequency of the wild type amino acid at a given position in the phage- sorted or the unsorted library, respectively..

FIGURE 1 8 is a graph showing the distributions of mutations at different HVR position classes

(surface exposed, buried residues, and residues which are in close contact with antigen ("contact positions")) which were determined using the structural information obtained from the crystal structure of 5A12 4.2 bound to Ang2 or to VEGF. The distributions of mutations at these different position classes are shown as a violin plot. The black dots represent the enrichment ratio of individual mutations. Based on the crystal structures, residues of 5A12 4.2 which are in contact with the antigens were defined as residues which are within a 5A radius of an antigen atom. Antigen atoms are all atoms which form the amino acid chain of the antigen, excluding water, bound ions, and the like.

FIGURES 19A and 19B show heatmaps of the log2-fold enrichment ratio for mutations at the indicated V_H positions from the Ang2-panned (Figure 19A) or VEGF-panned (Figure 19B) HC-3NN K library compared to the frequency of the same mutation in the unsorted HC-3NNK library. Black boxes indicate the wild-type 5A12 4.2 sequence, while grey boxes indicate mutations which have not been found in the sorted sample.

FIGURES 20A and 20B are graphs showing a comparison of the log2-fold enrichment ratio for mutations from the VEGF-panned HC-3NNK library with the corresponding enrichment ratio from the Ang2-panned HC-3NN K library (Figure 20A) or for the VEGF-panned and Ang2-panned LC-3NNK libraries (Figure 20B). Selected mutations which have been tested experimentally or which are otherwise described herein are indicated.

FIGURE 21 is a series of graphs showing comparisons between the log2-fold enrichment ratio calculated for single mutations from the deep sequencing data set with the log2-fold IC50 change observed in a phage competition assay. 25 single mutants (15 in the heavy chain and 10 in the light chain) were selected from the mutagenesis data set and phage competition ELISA was performed against Ang2 (red) or VEGF (green). For the heavy chain, the IC50 fold change was compared against enrichment ratios derived from 3NNK and 1 NNK libraries.

FIGURES 22A-22B is a series of graphs (Circos plots) showing the highest-enriched position pairs as calculated from deep sequencing data obtained from the Ang2-panned HC-3NNK (Figure 22A) and LC-3NNK (Figure 22B) libraries. For this visualization, the enrichment of mutation pairs at the same positions were summed. The circular segments represent positions in or near HVRs while the ribbons connecting two positions represent a position pair. A wider ribbon shows that the sum of enrichment at this position is larger. At the end of each ribbon the amino acids which form mutation pairs at the positions connected by the ribbon are listed. The histogram at the outer layer of the Circos plot shows the Ca-Ca distance between two mutation pairs. The ribbons of the position pairs are highlighted in pink when they contain fold-stabilizing mutations. Ribbons of the position pairs are highlighted in orange when they contain affinity-improving mutation pairs. The light chain numbering uses the Chothia numbering scheme, while the heavy chain numbering uses the Kabat numbering scheme.

FIGURES 22C-22D is a series of graphs (scatter plots) showing the correlation between the log2- fold enrichment ratio of a given mutation in the dataset obtained from the Ang2-panned 3NNK libraries and the partner potentiation score. The melting temperature (T_m) of selected variants is shown compared to the parental Fab 5A12 4.2.

FIGURES 22E-22F are graphs showing the fold change in affinity (Kd) for Ang2 (Figure 22E) and VEGF (Figure 22F) as measured by BIACORE® surface plasmon resonance (SPR) for selected double mutation pairs and their constituent individual mutations.

FIGURE 23A is a rendering of the structure of 5A12 4.2 showing positions (spheres) in the heavy (red) or light chain (blue) of 5A12 4.2 which have been used to generate higher affinity variants based on deep mutagenesis scanning.

FIGURE 23B is a table showing dual-specific affinity matured variants of 5A12 4.2 with sub- nanomolar affinity for VEGF and Ang2 as measured by BIACORE® SPR in three independent experiments. The table also indicates the mutations that each clone contains as well as the melting temperature (T_m) compared to the parent Fab 5A12 4.2, if available.

FIGURE 23C is a series of graphs showing the results of in vitro receptor-blocking assays comparing the blocking of Ang2 binding to Tie2 (left panel) and VEGF to Flt-1 (right panel) by 5A12 4.2, the higher-affinity variants T.30M, T.28P, and T.28P-VR, and two control antibodies, G5.5 and G6.31 , which are monospecific for Ang2 and VEGF, respectively. CTL indicates the negative control antibody which does not interact with VEGF or Ang2.

FIGURE 23D is a table showing results from the in vitro receptor-blocking assays described in Figure 23C.

FIGURE 24 is a table comparing the affinity for Ang2, VEGF, and Ang1 of the indicated affinity matured clones derived from 5A12 4.2 obtained by phage display (5A12 4.2.16.2) or by deep mutagenesis scanning using 3NNK libraries (T.30M, T.28P, and T.28P-VR). The table also indicates the IC50 of the clones for inhibiting HUVEC migration.

FIGURE 25A is a schematic showing the phage panning strategy for affinity maturation of 5A12

4.2.

FIGURE 25B is a sequence alignment showing light chain variable region amino acid residue sequences of selected clones obtained from affinity maturation of 5A12 4.2. HVR sequences are delimited by the denoted boxes for each of the DAFs.

FIGURE 25C is a sequence alignment showing heavy chain variable region amino acid residue sequences of selected clones obtained from affinity maturation of 5A12 4.2. HVR sequences are delimited by the denoted boxes for each of the DAFs.

FIGURE 25D is a table showing the phage IC₅₀ against hVEGF₁₀₉ and hAng2his8 for the indicated clones obtained from affinity maturation of 5A12 4.2.

FIGURE 26A is a graph showing the results of a receptor-blocking ELISA experiment against the

VEGF receptor Fit comparing 5A12 4.2 with selected affinity matured clones 5A12 4.2.5 (#5), 5A12 4.2.9 (#9), 5A12 4.2.16 (#16), and 5A12 4.2.28 (#28). The anti-VEGF antibodies AVASTIN®, LUCENTIS®, and G6.31 were used as positive controls.

FIGURE 26B is a graph showing the results of a receptor-blocking ELISA experiment against the Ang2 receptor Tie2. G5.5, a high affinity anti-Ang2 antibody, was used as a positive control. FIGURE 26C is a table summarizing the results of the experiments shown in Figures 26A and

26B.

FIGURE 27A is a sequence alignment showing light chain variable region amino acid residue sequences of selected clones obtained from affinity maturation of 5A12 4.2.16. HVR sequences according to the Kabat definition are underlined.

FIGURE 27B is a sequence alignment showing heavy chain variable region amino acid residue sequences of selected clones obtained from affinity maturation of 5A12 4.2.16. HVR sequences are delimited by the denoted boxes for each of the DAFs.

FIGURE 27C is a table showing the phage IC₅₀ against hVEGF₁₀₉ and hAng2his8 for the indicated clones obtained from affinity maturation of 5A12 4.2.1 6.

FIGURE 28 is a table showing the Kd of the indicated clones obtained from affinity maturation of 5A12 4.2.16 as determined by a BIACORE® SPR assay. G6.31 is an anti-VEGF antibody used as a positive control. G5.5 is an anti-Ang2 antibody used as a positive control. Asterisk (^*) indicates that the off-rate constant observed was at or near the detection limit of the instrument (BIACORE® 3000) ;

therefore, a K_off of 5x10⁶ was used and reported Kd are upper limits.

FIGURE 29A is a sequence alignment showing light chain variable region amino acid residue sequences of selected clones, including 5A12 4.2, 5A12 4.2.16.2, T.28P, T.30M, and T.28P-VR. HVR sequences are delimited by the denoted boxes for each of the DAFs.

FIGURE 29B is a sequence alignment showing heavy chain variable region amino acid residue sequences of selected clones, including 5A12 4.2, 5A12 4.2.16.2, T.28P, T.30M, and T.28P-VR. HVR sequences are delimited by the denoted boxes for each of the DAFs.

FIGURES 30A and 30B are graphs showing laser-induced neovascular lesion area measured in rat eyes treated with different antibody formats against VEGF and Ang2. An anti-ragweed antibody served as a negative control. An anti-VEGF (G6.31 ) Fab and a combination of G6.31 and an anti-Ang2 (G5.5) Fab served as positive controls. "Bispecific Fab'2" is a bi-specific F(ab')₂ format of G6.31 and

G5.5 linked by a single cysteine residue. The T.28P-VR variant was tested either as a monomeric Fab or as a dimeric F(ab')₂.

Detailed Description of Embodiments of the Invention

I. Definitions

The term "about" as used herein refers to the usual error range for the respective value readily known to the skilled person in this technical field. Reference to "about" a value or parameter herein includes (and describes) embodiments that are directed to that value or parameter per se.

An "acceptor human framework" for the purposes herein is a framework comprising the amino acid sequence of a light chain variable domain (V_L) framework or a heavy chain variable domain (V_H) framework derived from a human immunoglobulin framework or a human consensus framework, as defined below. An acceptor human framework "derived from" a human immunoglobulin framework or a human consensus framework may comprise the same amino acid sequence thereof, or it may contain amino acid sequence changes. In some embodiments, the number of amino acid changes are 10 or less, 9 or less, 8 or less, 7 or less, 6 or less, 5 or less, 4 or less, 3 or less, or 2 or less. In some embodiments, the V_L acceptor human framework is identical in sequence to the V_L human immunoglobulin framework sequence or human consensus framework sequence.

"Affinity" refers to the strength of the sum total of noncovalent interactions between a single binding site of a molecule (e.g., an antibody) and its binding partner (e.g., an antigen). Unless indicated otherwise, as used herein, "binding affinity" refers to intrinsic binding affinity which reflects a 1 :1 interaction between members of a binding pair (e.g., antibody and antigen). The affinity of a molecule X for its partner Y can generally be represented by the dissociation constant (Kd). Affinity can be measured by common methods known in the art, including those described herein. Specific illustrative and exemplary embodiments for measuring binding affinity are described in the following.

An "affinity-matured" antibody is one with one or more alterations in one or more HVRs thereof which result in an improvement in the affinity of the antibody for antigen, compared to a parent antibody which does not possess those alteration(s). Preferred affinity-matured antibodies will have nanomolar or even picomolar affinities for the target antigen. Affinity-matured antibodies are produced by procedures known in the art. Marks et al. Bio/Technology 10:779-83 (1992) describes affinity maturation by V_H and V|_ domain shuffling. Random mutagenesis of HVR and/or framework residues is described by: Barbas et al., Proc. Nat. Acad. Sci. USA 91 :3809-13 (1 994) ; Schier et al. Gene 169:147-55 (1995) ; Yelton et al. J. Immunol. 1 55:1 994-2004 (1995) ; Jackson et al., J. Immunol. 154(7) :331 0-19 (1 995) ; and Hawkins et al. J. Mol. Biol. 226:889-96 (1992).

The term "Angiopoietin-2" or "Ang2" as used herein, refers to any native Ang2 from any vertebrate source, including mammals such as primates (e.g. humans) and rodents (e.g., mice and rats), unless otherwise indicated. The term encompasses "full-length," unprocessed Ang2 as well as any form of Ang2 that results from processing in the cell. The term also encompasses naturally occurring variants of Ang2, e.g., splice variants or allelic variants. Additional information on the human Ang2 gene can be found under NCBI Gene ID No. 285. The amino acid sequence of an exemplary human Ang2 is shown in SEQ ID NO: 19. The amino acid sequence of an exemplary full-length human Ang2 can be found, e.g., under NCBI Accession No. NP_001 1 12359 or UniProt Accession No. 015123. Ang2 is a ligand for the Tie2 receptor. Ang2 has also been found to bind integrins, for example, integrin beta 2 (Bezuidenhout et al. Inflammation 32(6) :393-401 , 2009).

The terms "anti-Ang2 antibody," an "antibody that binds to Ang2," and "antibody that specifically binds Ang2" refer to an antibody that is capable of binding Ang2 with sufficient affinity such that the antibody is useful as a diagnostic and/or therapeutic agent in targeting Ang2. In one embodiment, the extent of binding of an anti-Ang2 antibody to an unrelated, non-Ang2 protein is less than about 10% of the binding of the antibody to Ang2 as measured, e.g., by a radioimmunoassay (RIA). In certain

embodiments, an antibody that binds to Ang2 has a dissociation constant (Kd) of < 1 μΜ, < 100 nM, < 1 0 nM, < 1 nM, < 0.1 nM, < 0.01 nM, or < 0.001 nM (e.g. 10^"8 M or less, e.g. from 10^"8 M to 10^"13 M, e.g., from 10^"9 M to 10^"13 M). In certain embodiments, an anti-Ang2 antibody binds to an epitope of Ang2 that is conserved among Ang2 from different species.

The terms "anti-Ang2/anti-VEGF" antibody or "antibody that specifically binds to Ang2 and VEGF" or similar terms as used herein refer to a dual-specific antibody (e.g., a Fab or an IgG) that specifically binds Ang2 and VEGF. The term "vascular endothelial growth factor" or "VEGF" as used herein refers to the 165-amino acid human vascular endothelial cell growth factor and related 121 -, 189-, and 206- amino acid human vascular endothelial cell growth factors, as described by Leung et al. Science, 246:1306 (1 989), and Houck et al. Mol. Endocrin., 5:1806 (1 991 ), together with the naturally occurring allelic and processed forms thereof. The term "VEGF" also refers to VEGFs from non-human species such as mouse, rat or primate. Sometimes the VEGF from a specific species are indicated by terms such as hVEGF for human VEGF, mVEGF for murine VEGF, and etc. The term "VEGF" is also used to refer to truncated forms of the polypeptide comprising amino acids 8 to 109 or 1 to 109 of the 165-amino acid human vascular endothelial cell growth factor. Reference to any such forms of VEGF may be identified in the present application, e.g., by "VEGF₁₀₉," "VEGF (8-1 09)," "VEGF (1 -109)" or "VEGF₁₆₅." The amino acid positions for a "truncated" native VEGF are numbered as indicated in the native VEGF sequence. For example, amino acid position 17 (methionine) in truncated native VEGF is also position 17 (methionine) in native VEGF. The truncated native VEGF has binding affinity for the KDR and Flt-1 receptors comparable to native VEGF. The term "VEGF variant" as used herein refers to a VEGF polypeptide which includes one or more amino acid mutations in the native VEGF sequence. Optionally, the one or more amino acid mutations include amino acid substitution(s). For purposes of shorthand designation of VEGF variants described herein, it is noted that numbers refer to the amino acid residue position along the amino acid sequence of the putative native VEGF (provided in Leung et al., supra and Houck et al., supra.). Unless specified otherwise, the term "VEGF" as used herein indicates VEGF-A.

The terms "anti-VEGF antibody," an "antibody that binds to VEGF," and "antibody that specifically binds VEGF" refer to an antibody that is capable of binding VEGF with sufficient affinity such that the antibody is useful as a diagnostic and/or therapeutic agent in targeting VEGF. In one embodiment, the extent of binding of an anti-VEGF antibody to an unrelated, non-VEGF protein is less than about 1 0% of the binding of the antibody to VEGF as measured, e.g., by a radioimmunoassay (RIA). In certain embodiments, an antibody that binds to VEGF has a dissociation constant (Kd) of < 1 μΜ, < 100 nM, < 1 0 nM, < 1 nM, < 0.1 nM, < 0.01 nM, or < 0.001 nM (e.g. 10^"8 M or less, e.g. from 10^"8 M to 10^"13 M, e.g., from 10^"9 M to 10^"13 M). In certain embodiments, an anti-VEGF antibody binds to an epitope of VEGF that is conserved among VEGF from different species.

The term "antibody" herein is used in the broadest sense and encompasses various antibody structures, including but not limited to monoclonal antibodies, polyclonal antibodies, multispecific antibodies (e.g., bispecific antibodies), and antibody fragments so long as they exhibit the desired antigen-binding activity.

An "antibody that binds to the same epitope" as a reference antibody refers to an antibody that contacts an overlapping set of amino acid residues of the antigen as compared to the reference antibody or blocks binding of the reference antibody to its antigen in a competition assay by 50% or more. The amino acid residues of an antibody that contact an antigen can be determined, for example, by determining the crystal structure of the antibody in complex with the antigen or by performing hydrogen/deuterium exchange. In some embodiments, residues of an antibody that are within 5 A of the antigen are considered to contact the antigen. In some embodiments, an antibody that binds to the same epitope as a reference antibody blocks binding of the reference antibody to its antigen in a competition assay by 50% or more, and conversely, the reference antibody blocks binding of the antibody to its antigen in a competition assay by 50% or more. An exemplary competition assay is provided herein.

"Antibody fragments" comprise a portion of an intact antibody, preferably the antigen binding or variable region of the intact antibody. Examples of antibody fragments include Fab, Fab', F(ab')₂, and Fv fragments; diabodies; linear antibodies (see U.S. Patent No. 5,641 ,870, Example 2; Zapata et al. Protein Eng. 8(10) : 1057-1062 (1995)) ; single-chain antibody molecules; and multispecific antibodies formed from antibody fragments.

Papain digestion of antibodies produces two identical antigen-binding fragments, called "Fab" fragments, and a residual "Fc" fragment, a designation reflecting the ability to crystallize readily. The Fab fragment consists of an entire L chain along with the variable region domain of the H chain (V_H), and the first constant domain of one heavy chain (C_H1 ). Pepsin treatment of an antibody yields a single large F(ab')2 fragment which roughly corresponds to two disulfide linked Fab fragments having divalent antigen-binding activity and is still capable of cross-linking antigen. Fab' fragments differ from Fab fragments by having additional few residues at the carboxy terminus of the C_H1 domain including one or more cysteines from the antibody hinge region. Fab'-SH is the designation herein for Fab' in which the cysteine residue(s) of the constant domains bear a free thiol group. F(ab')₂ antibody fragments originally were produced as pairs of Fab' fragments which have hinge cysteines between them . Other chemical couplings of antibody fragments are also known.

The term "Fc region" herein is used to define a C-terminal region of an immunoglobulin heavy chain that contains at least a portion of the constant region. The term includes native sequence Fc regions and variant Fc regions. In one embodiment, a human IgG heavy chain Fc region extends from Cys226, or from Pro230, to the carboxyl-terminus of the heavy chain. However, the C-terminal lysine (Lys447) of the Fc region may or may not be present. Unless otherwise specified herein, numbering of amino acid residues in the Fc region or constant region is according to the EU numbering system , also called the EU index, as described in Kabat et al., Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, MD (1991 ).

"Fv" consists of a dimer of one heavy- and one light-chain variable region domain in tight, non- covalent association. From the folding of these two domains emanate six hypervariable loops (3 loops each from the H and L chain) that contribute the amino acid residues for antigen binding and confer antigen binding specificity to the antibody. However, even a single variable domain (or half of an Fv comprising only three Hs specific for an antigen) has the ability to recognize and bind antigen, although often at a lower affinity than the entire binding site.

"Single-chain Fv" also abbreviated as "sFv" or "scFv" are antibody fragments that comprise the V_H and V_L antibody domains connected into a single polypeptide chain. Preferably, the sFv polypeptide further comprises a polypeptide linker between the V_H and V_L domains which enables the sFv to form the desired structure for antigen binding. For a review of sFv, see Pluckthun in The Pharmacology of Monoclonal Antibodies, vol. 1 13, Rosenburg and Moore eds., Springer-Verlag, New York, pp. 269-315 (1994). The term "diabodies" refers to small antibody fragments prepared by constructing sFv fragments (see preceding paragraph) with short linkers (about 5-10 residues) between the V_H and V_L domains such that inter-chain but not intra-chain pairing of the V domains is achieved, resulting in a bivalent fragment, i.e., fragment having two antigen-binding sites. Bispecific diabodies are heterodimers of two "crossover" sFv fragments in which the V_H and V_L domains of the two antibodies are present on different polypeptide chains. Diabodies are described more fully in, for example, EP 404,097; WO 93/1 1 161 ; and Hollinger et al. Proc. Natl. Acad. Sci. USA, 90:6444-6448 (1993).

A "blocking" antibody or an "antagonist" antibody is one which inhibits or reduces biological activity of the antigen it binds. Certain blocking antibodies or antagonist antibodies substantially or completely inhibit the biological activity of the antigen.

The "class" of an antibody refers to the type of constant domain or constant region possessed by its heavy chain. There are five major classes of antibodies: IgA, IgD, IgE, IgG, and IgM, and several of these may be further divided into subclasses (isotypes), e.g., IgG_! , lgG₂, lgG₃, lgG₄, IgA^ and lgA₂. The heavy chain constant domains that correspond to the different classes of immunoglobulins are called α, δ, ε, γ, and μ, respectively.

Antibody "effector functions" refer to those biological activities attributable to the Fc region (a native sequence Fc region or amino acid sequence variant Fc region) of an antibody, and vary with the antibody isotype. Examples of antibody effector functions include: C1 q binding and complement dependent cytotoxicity; Fc receptor binding; antibody-dependent cell-mediated cytotoxicity (ADCC) ; phagocytosis; down regulation of cell surface receptors (e.g., B cell receptor) ; and B cell activation.

"Antibody-dependent cell-mediated cytotoxicity" or "ADCC" refers to a form of cytotoxicity in which secreted Ig bound onto Fc receptors (FcRs) present on certain cytotoxic cells (e.g., Natural Killer (N K) cells, neutrophils, and macrophages) enable these cytotoxic effector cells to bind specifically to an antigen-bearing target cell and subsequently kill the target cell with cytotoxins. The antibodies "arm" the cytotoxic cells and are absolutely required for such killing. The primary cells for mediating ADCC, N K cells, express FcyRI II only, whereas monocytes express FcyRI, FcyRII, and FcyRII I. FcR expression on hematopoietic cells is summarized in Table 3 on page 464 of Ravetch and Kinet, Annu. Rev. Immunol. 9:457-92 (1 991 ). To assess ADCC activity of a molecule of interest, an in vitro ADCC assay, such as that described in US Patent No. 5,500,362 or 5,821 ,337 can be performed. Useful effector cells for such assays include peripheral blood mononuclear cells (PBMC) and Natural Killer (NK) cells. Alternatively, or additionally, ADCC activity of the molecule of interest can be assessed in vivo, e.g., in a animal model such as that disclosed in Clynes et al. Proc. Natl. Acad. Sci. USA 95:652-656 (1998).

"Fc receptor" or "FcR" describes a receptor that binds to the Fc region of an antibody. The preferred FcR is a native sequence human FcR. Moreover, a preferred FcR is one which binds an IgG antibody (a gamma receptor) and includes receptors of the FcyRI, FcyRI I, and FcyRI II subclasses, including allelic variants and alternatively spliced forms of these receptors. FcyRI I receptors include FcyRI IA (an "activating receptor") and FcyRI IB (an "inhibiting receptor"), which have similar amino acid sequences that differ primarily in the cytoplasmic domains thereof. Activating receptor FcyRIIA contains an immunoreceptor tyrosine-based activation motif (ITAM) in its cytoplasmic domain. Inhibiting receptor FcyRIIB contains an immunoreceptor tyrosine-based inhibition motif (ITIM) in its cytoplasmic domain (see review M. in Daeron, Annu. Rev. Immunol. 15:203-234 (1997)). FcRs are reviewed in Ravetch and Kinet, (Annu. Rev. Immunol. 9:457-492 (1991 )) ; Capel et al., (Immunomethods 4:25-34 (1994)) ; and de Haas et al., (J. Lab. Clin. Med. 126:330-41 (1995)). Other FcRs, including those to be identified in the future, are encompassed by the term "FcR" herein. The term also includes the neonatal receptor, FcRn, which is responsible for the transfer of maternal IgGs to the fetus (Guyer et al., J. Immunol. 1 17:587 (1976) and Kim et al., J. Immunol. 24:249 (1994)).

"Human effector cells" are leukocytes which express one or more FcRs and perform effector functions. Preferably, the cells express at least FcyRI II and perform ADCC effector function. Examples of human leukocytes which mediate ADCC include peripheral blood mononuclear cells (PBMC), natural killer (NK) cells, monocytes, cytotoxic T cells, and neutrophils; with PBMCs and NK cells being preferred. The effector cells can be isolated from a native source, e.g., from blood.

"Complement dependent cytotoxicity" or "CDC" refers to the lysis of a target cell in the presence of complement. Activation of the classical complement pathway is initiated by the binding of the first component of the complement system (C1 q) to antibodies (of the appropriate subclass) which are bound to their cognate antigen. To assess complement activation, a CDC assay, e.g., as described in Gazzano- Santoro et al., J. Immunol. Methods 202:163 (1996), can be performed.

An "epitope" is the portion of the antigen to which the antibody specifically binds. For a polypeptide antigen, the epitope is generally a peptide portion of about 4-15 amino acid residues.

The terms "full-length antibody," "intact antibody," and "whole antibody" are used herein interchangeably to refer to an antibody having a structure substantially similar to a native antibody structure or having heavy chains that contain an Fc region as defined herein.

A "human antibody" is one which possesses an amino acid sequence which corresponds to that of an antibody produced by a human and/or has been made using any of the techniques for making human antibodies. This definition of a human antibody specifically excludes a humanized antibody comprising non-human antigen-binding residues.

A "human consensus framework" is a framework which represents the most commonly occurring amino acid residues in a selection of human immunoglobulin V_L or V_H framework sequences. Generally, the selection of human immunoglobulin V_L or V_H sequences is from a subgroup of variable domain sequences. Generally, the subgroup of sequences is a subgroup as in Kabat et al., Sequences of Proteins of Immunological Interest, Fifth Edition, N IH Publication 91 -3242, Bethesda MD (1991 ), vols. 1 -3. In one embodiment, for the V_L, the subgroup is subgroup kappa I as in Kabat et al., supra. In one embodiment, for the V_H, the subgroup is subgroup II I as in Kabat et al., supra.

"Humanized" forms of non-human (e.g., rodent) antibodies are chimeric antibodies that contain minimal sequence derived from the non-human antibody. For the most part, humanized antibodies are human immunoglobulins (recipient antibody) in which residues from a hypervariable region of the recipient are replaced by residues from a hypervariable region of a non-human species (donor antibody) such as mouse, rat, rabbit or non-human primate having the desired antibody specificity, affinity, and capability. In some instances, framework region (FR) residues of the human immunoglobulin are replaced by corresponding non-human residues. Furthermore, humanized antibodies can comprise residues that are not found in the recipient antibody or in the donor antibody. These modifications are made to further refine antibody performance. In general, the humanized antibody will comprise substantially all of at least one, and typically two, variable domains, in which all or substantially all of the hypervariable loops correspond to those of a non-human immunoglobulin and all or substantially all of the FRs are those of a human immunoglobulin sequence. The humanized antibody optionally also will comprise at least a portion of an immunoglobulin constant region (Fc), typically that of a human immunoglobulin. For further details, see Jones et al., Nature 321 :522-525 (1986) ; Riechmann et al., Nature 332:323-329 (1988) ; and Presta, Curr. Op. Struct. Biol. 2:593-596 (1992).

An "immunoconjugate" is an antibody conjugated to one or more heterologous molecule(s), including but not limited to a cytotoxic agent.

The term an "isolated antibody" when used to describe the various antibodies disclosed herein, means an antibody that has been identified and separated and/or recovered from a cell or cell culture from which it was expressed. Contaminant components of its natural environment are materials that would typically interfere with diagnostic or therapeutic uses for the polypeptide, and can include enzymes, hormones, and other proteinaceous or non-proteinaceous solutes. In some embodiments, an antibody is purified to greater than 95% or 99% purity as determined by, for example, electrophoretic (e.g., SDS- PAGE, isoelectric focusing (IEF), capillary electrophoresis) or chromatographic (e.g., ion exchange or reverse phase HPLC). For a review of methods for assessment of antibody purity, see, e.g., Flatman et al., J. Chromatogr. 6848:79-87 (2007). In preferred embodiments, the antibody will be purified (1 ) to a degree sufficient to obtain at least 1 5 residues of N-terminal or internal amino acid sequence by use of a spinning cup sequenator, or (2) to homogeneity by SDS-PAGE under non-reducing or reducing conditions using Coomassie blue or, preferably, silver stain. Isolated antibody includes antibodies in situ within recombinant cells, because at least one component of the polypeptide natural environment will not be present. Ordinarily, however, isolated polypeptide will be prepared by at least one purification step.

The term "monoclonal antibody" as used herein refers to an antibody from a population of substantially homogeneous antibodies, i.e., the individual antibodies comprising the population are substantially similar and bind the same epitope(s), except for possible variants that may arise during production of the monoclonal antibody, such variants generally being present in minor amounts. Such monoclonal antibody typically includes an antibody comprising a variable region that binds a target, wherein the antibody was obtained by a process that includes the selection of the antibody from a plurality of antibodies. For example, the selection process can be the selection of a unique clone from a plurality of clones, such as a pool of hybridoma clones, phage clones or recombinant DNA clones. It should be understood that the selected antibody can be further altered, for example, to improve affinity for the target, to humanize the antibody, to improve its production in cell culture, to reduce its immunogenicity in vivo, to create a multispecific antibody, etc., and that an antibody comprising the altered variable region sequence is also a monoclonal antibody of this invention. In addition to their specificity, the monoclonal antibody preparations are advantageous in that they are typically uncontaminated by other

immunoglobulins. The modifier "monoclonal" indicates the character of the antibody as being obtained from a substantially homogeneous population of antibodies, and is not to be construed as requiring production of the antibody by any particular method. For example, the monoclonal antibodies to be used in accordance with the present invention may be made by a variety of techniques, including the hybridoma method (e.g., Kohler et al., Nature, 256:495 (1975) ; Harlow et al., Antibodies: A Laboratory Manual, (Cold Spring Harbor Laboratory Press, 2nd ed. 1988) ; Hammerling et al., in: Monoclonal Antibodies and T-Cell Hybridomas 563-681 , (Elsevier, N.Y., 1981 ), recombinant DNA methods (see, e.g., U.S. Patent No. 4,81 6,567), phage display technologies (see, e.g., Clackson et al., Nature, 352:624-628 (1991 ) ; Marks et al., J. Mol. Biol., 222:581 -597 (1991 ) ; Sidhu et al., J. Mol. Biol. 338(2) :299-310 (2004) ; Lee et al., J. Mol. Biol. 340(5) :1073-1093 (2004) ; Fellouse, Proc. Nat. Acad. Sci. USA 101 (34) :12467- 12472 (2004) ; and Lee et al. J. Immunol. Methods 284(1 -2) :1 19-132 (2004) and technologies for producing human or human-like antibodies from animals that have parts or all of the human

immunoglobulin loci or genes encoding human immunoglobulin sequences (see, e.g., W098/24893, WO/9634096, WO/9633735, and WO/91 10741 , Jakobovits et al., Proc. Natl. Acad. Sci. USA, 90:2551 (1993) ; Jakobovits et al., Nature, 362:255-258 (1993) ; Bruggemann et al., Year in Immuno., 7:33 (1993) ; U.S. Patent Nos. 5,545,806, 5,569,825, 5,591 ,669 (all of GenPharm) ; 5,545,807; WO 97/17852, U.S. Patent Nos. 5,545,807; 5,545,806; 5,569,825; 5,625,126; 5,633,425; and 5,661 ,016, and Marks et al., Bio/Technology, 10: 779-783 (1992) ; Lonberg et al., Nature, 368: 856-859 (1 994) ; Morrison, Nature, 368: 812-813 (1994) ; Fishwild et al., Nature Biotechnology, 14: 845-851 (1996) ; Neuberger, Nature

Biotechnology, 14: 826 (1996) ; and Lonberg and Huszar, Intern. Rev. Immunol., 13: 65-93 (1995).

The monoclonal antibodies herein specifically include chimeric, humanized, fully human, and affinity-matured antibodies. Chimeric antibodies are antibodies in which a portion of the heavy and/or light chain is identical with or homologous to corresponding sequences in antibodies derived from a particular species or belonging to a particular antibody class or subclass, while the remainder of the chain(s) is identical with or homologous to corresponding sequences in antibodies derived from another species or belonging to another antibody class or subclass, as well as fragments of such antibodies, so long as they exhibit the desired biological activity (U.S. Pat. No. 4,816,567; and Morrison et al., Proc. Natl. Acad. Sci. USA, 81 :6851 -6855 (1984)). Chimeric antibodies of interest herein include "primatized" antibodies comprising variable domain antigen-binding sequences derived from a non-human primate (e.g. Old World Monkey, Ape etc) and human constant region sequences.

The term "multispecific antibody" is used in the broadest sense and specifically covers an antibody comprising a heavy chain variable domain (V_H) and a light chain variable domain (V_L), where the V_HV_L unit has polyepitopic specificity (i.e., is capable of binding to two different epitopes on one biological molecule or each epitope on a different biological molecule). Such multispecific antibodies include, but are not limited to, full-length antibodies, antibodies having two or more V_L and V_H domains, antibody fragments such as Fab, Fv, dsFv, scFv, diabodies, bispecific diabodies and triabodies, antibody fragments that have been linked covalently or non-covalently. "Polyepitopic specificity" refers to the ability to specifically bind to two or more different epitopes on the same or different target(s). "Dual specificity" or "bispecificity" refers to the ability to specifically bind to two different epitopes on the same or different target(s). However, in contrast to bispecific antibodies, dual-specific antibodies have two antigen-binding arms that are identical in amino acid sequence and each Fab arm is capable of recognizing two antigens. Dual-specificity allows the antibodies to interact with high affinity with two different antigens as a single Fab or IgG molecule. According to one embodiment, the multispecific antibody in an lgG1 form binds to each epitope with an affinity of 5μΜ to 0.001 pM, 3μΜ to 0.001 pM, 1 μΜ to 0.001 pM, 0.5μΜ to 0.001 pM or 0.1 μΜ to 0.001 pM. "Monospecific" refers to the ability to bind only one epitope.

A "naked antibody" refers to an antibody that is not conjugated to a heterologous moiety (e.g., a cytotoxic moiety) or radiolabel. The naked antibody may be present in a pharmaceutical composition.

By "paratope" is meant the part of an antibody which selectively binds the epitope of an antigen.

The paratope is typically a region of about 15-22 amino acid residues of the antibody's Fv region and may contain amino acids from the antibody's V_H and V_L chains.

With regard to the binding of a antibody to a target molecule, the term "specific binding" or "specifically binds to" or is "specific for" a particular polypeptide or an epitope on a particular polypeptide target means binding that is measurably different from a non-specific interaction. Specific binding can be measured, for example, by determining binding of a molecule compared to binding of a control molecule. For example, specific binding can be determined by competition with a control molecule that is similar to the target, for example, an excess of non-labeled target. In this case, specific binding is indicated if the binding of the labeled target to a probe is competitively inhibited by excess unlabeled target. The term "specific binding" or "specifically binds to" or is "specific for" a particular polypeptide or an epitope on a particular polypeptide target as used herein can be exhibited, for example, by a molecule having a Kd for the target of 10^"4 M or lower, alternatively 10^"5 M or lower, alternatively 10^"6 M or lower, alternatively 10^"7 M or lower, alternatively 10^"8 M or lower, alternatively 1 0^"9 M or lower, alternatively 1 0^"10 M or lower, alternatively 1 0^{"1 1} M or lower, alternatively 10^"12 M or lower or a Kd in the range of 10^"4 M to 10^"6 M or 10^"6 M to 10^"10 M or 10^"7 M to 10^"9 M. As will be appreciated by the skilled artisan, affinity and Kd values are inversely related. A high affinity for an antigen is measured by a low Kd value. In one embodiment, the term "specific binding" refers to binding where a molecule binds to a particular polypeptide or epitope on a particular polypeptide without substantially binding to any other polypeptide or polypeptide epitope.

The term "variable" refers to the fact that certain segments of the variable domains differ extensively in sequence among antibodies. The variable or "V" domain mediates antigen binding and defines specificity of a particular antibody for its particular antigen. However, the variability is not evenly distributed across the 1 10-amino acid span of the variable domains. Instead, the V regions consist of relatively invariant stretches called framework regions (FRs) of 15-30 amino acids separated by shorter regions of extreme variability called "hypervariable regions" that are each 9-12 amino acids long. The term "hypervariable region" or "HVR" when used herein refers to the amino acid residues of an antibody which are responsible for antigen-binding. The hypervariable region generally comprises amino acid residues from e.g., around about residues 24-35 (L1 ), 50-58 (L2) and 89-97 (L3) in the V_L, and around about residues 26-35 (H1 ), 49-65 (H2) and 93-102 (H3) in the V_H (in one embodiment, H1 is around about residues 31 -35) ; Kabat et al. Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, MD. (1991 )) and/or those residues from a "hypervariable loop" (e.g., residues 26-32 (L1 ), 50-52 (L2), and 91 -96 (L3) in the V_L, and 26-32 (H1 ), 53-55 (H2), and 96- 101 (H3) in the V_H; Chothia and Lesk, J. Mol. Biol. 196:901 -917 (1 987). The variable domains of native heavy and light chains each comprise four FRs, largely adopting a beta-sheet configuration, connected by three hypervariable regions, which form loops connecting, and in some cases forming part of, the beta- sheet structure. The hypervariable regions in each chain are held together in close proximity by the FRs and, with the hypervariable regions from the other chain, contribute to the formation of the antigen-binding site of antibodies (see Kabat et al. Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, MD. (1991 )). Accordingly, the HVR and FR sequences generally appear in the following sequence in V_H (or V_L) : FR1 -H1 (L1 )-FR2-H2(L2)-FR3-H3(L3)-FR4. The constant domains are not involved directly in binding an antibody to an antigen, but exhibit various effector functions, such as participation of the antibody in antibody dependent cellular cytotoxicity (ADCC).

The term "variable domain residue numbering as in Kabat" or "amino acid position numbering as in Kabat," and variations thereof, refers to the numbering system used for heavy chain variable domains or light chain variable domains of the compilation of antibodies in Kabat et al., supra. Using this numbering system, the actual linear amino acid sequence may contain fewer or additional amino acids corresponding to a shortening of, or insertion into, a FR or HVR of the variable domain. For example, a heavy chain variable domain may include a single amino acid insert (residue 52a according to Kabat) after residue 52 of H2 and inserted residues (e.g. residues 82a, 82b, and 82c, etc. according to Kabat) after heavy chain FR residue 82. The Kabat numbering of residues may be determined for a given antibody by alignment at regions of homology of the sequence of the antibody with a "standard" Kabat numbered sequence.

The Kabat numbering system is generally used when referring to a residue in the variable domain (approximately residues 1 -107 of the light chain and residues 1 -1 13 of the heavy chain) (e.g, Kabat et al., Sequences of Immunological Interest. 5th Ed. Public Health Service, National Institutes of Health,

Bethesda, Md. (1991 )). The "EU numbering system" or "EU index" is generally used when referring to a residue in an immunoglobulin heavy chain constant region (e.g., the EU index reported in Kabat et al., supra). The "EU index as in Kabat" refers to the residue numbering of the human lgG1 EU antibody. Unless stated otherwise herein, references to residue numbers in the variable domain of antibodies means residue numbering by the Kabat numbering system . Unless stated otherwise herein, references to residue numbers in the constant domain of antibodies means residue numbering by the EU numbering system (e.g., see United States Provisional Application No. 60/640,323, Figures for EU numbering).

As used herein, "administering" is meant a method of giving a dosage of a compound (e.g., an anti-Ang2 antibody of the invention, a nucleic acid encoding an anti-Ang2 antibody of the invention, a dual-specific anti-Ang2 and anti-VEGF antibody of the invention, or a nucleic acid encoding a bispecific anti-Ang2 and anti-VEGF antibody of the invention) or a composition (e.g., a pharmaceutical composition, e.g., a pharmaceutical composition including an anti-Ang2 antibody of the invention or including a bispecific anti-Ang2 and anti-VEGF antibody of the invention) to a subject. The compositions utilized in the methods described herein can be administered, for example, intravitreally (e.g., by intravitreal injection), by eye drop, intramuscularly, intravenously, intradermal^, percutaneously, intraarterially, intraperitoneally, intralesionally, intracranially, intraarticularly, intraprostatically, intrapleural^, intratracheally, intrathecal^, intranasally, intravaginally, intrarectally, topically, intratumorally, peritoneally, subcutaneously, subconjunctival^, intravesicularly, mucosally, intrapericardially, intraumbilically, intraocularly, intraorbitally, orally, topically, transdermal^, by inhalation, by injection, by implantation, by infusion, by continuous infusion, by localized perfusion bathing target cells directly, by catheter, by lavage, in cremes, or in lipid compositions. The compositions utilized in the methods described herein can also be administered systemically or locally. The method of administration can vary depending on various factors (e.g., the compound or composition being administered and the severity of the condition, disease, or disorder being treated).

A "disorder" is any condition that would benefit from treatment with the antibody. For example, a disorder may involve abnormal or pathological angiogenesis (e.g., excessive, inappropriate, or uncontrolled angiogenesis) or vascular permeability. A disorder can be chronic or acute.

"Angiogenesis" refers to the process through which new blood vessels form from pre-existing blood vessels. Angiogenesis is distinct from vasculogenesis, which is the de novo formation of endothelial cells from mesoderm cell precursors. Disorders associated with pathological angiogenesis can be treated by compositions and methods of the invention. These disorders include both nonneoplastic disorders and cell proliferative disorders. Cell proliferative disorders include but are not limited those described below. Non-neoplastic disorders include but are not limited to ocular conditions (non- limiting ocular conditions include, for example, retinopathy including proliferative diabetic retinopathy, choroidal neovascularization (CNV), age-related macular degeneration (AMD), diabetic and other ischemia-related retinopathies, diabetic macular edema (DME), pathologic myopia, von Hippel-Lindau disease, histoplasmosis of the eye, retinal vein occlusion (including central (CRVO) and branched (BRVO) forms), corneal neovascularization, retinal neovascularization, retinopathy of prematurity (ROP), familial exudative vitreoretinopathy (FEVR), Coats' disease, Norrie Disease, Osteoporosis-Pseudoglioma Syndrome (OPPG), subconjunctival hemorrhage, and hypertensive retinopathy), undesired or aberrant hypertrophy, arthritis, rheumatoid arthritis (RA), psoriasis, psoriatic plaques, sarcoidosis, atherosclerosis, atherosclerotic plaques, vascular restenosis, arteriovenous malformations (AVM), meningioma, hemangioma, angiofibroma, thyroid hyperplasias (including Grave's disease), corneal and other tissue transplantation, chronic inflammation, lung inflammation, acute lung injury/ARDS, sepsis, primary pulmonary hypertension, malignant pulmonary effusions, cerebral edema (e.g., associated with acute stroke/closed head injury/ trauma), synovial inflammation, pannus formation in RA, myositis ossificans, hypertropic bone formation, osteoarthritis (OA), refractory ascites, polycystic ovarian disease, endometriosis, 3rd spacing of fluid diseases (pancreatitis, compartment syndrome, burns, bowel disease), uterine fibroids, premature labor, chronic inflammation such as IBD (Crohn's disease and ulcerative colitis), renal allograft rejection, inflammatory bowel disease, nephrotic syndrome, undesired or aberrant tissue mass growth (non-cancer), hemophilic joints, hypertrophic scars, inhibition of hair growth, Osler- Weber syndrome, pyogenic granuloma retrolental fibroplasias, scleroderma, trachoma, vascular adhesions, synovitis, dermatitis, preeclampsia, ascites, pericardial effusion (such as that associated with pericarditis), and pleural effusion.

An "anti-angiogenesis agent" or "angiogenesis inhibitor" refers to a small molecular weight substance, a polynucleotide, a polypeptide, an isolated protein, a recombinant protein, an antibody, or conjugates or fusion proteins thereof, that inhibits angiogenesis, vasculogenesis, or undesirable vascular permeability, either directly or indirectly. It should be understood that the anti-angiogenesis agent includes those agents that bind and block the angiogenic activity of the angiogenic factor or its receptor. For example, an anti-angiogenesis agent is an antibody or other antagonist to an angiogenic agent as defined above, e.g., antibodies to VEGF-A or to the VEGF-A receptor (e.g., KDR receptor or Flt-1 receptor), anti-PDGFR inhibitors such as GLEEVEC™ (Imatinib Mesylate). Anti-angiogenesis agents also include native angiogenesis inhibitors, e.g., angiostatin, endostatin, etc. See, for example, Klagsbrun and D'Amore, Annu. Rev. Physiol., 53:217-39 (1991 ) ; Streit and Detmar, Oncogene, 22:3172-31 79 (2003) (e.g., Table 3 listing anti-angiogenic therapy in malignant melanoma) ; Ferrara & Alitalo, Nature Medicine 5(12) :1359-1364 (1999) ; Tonini et al., Oncogene, 22:6549-6556 (2003) (e.g., Table 2 listing known anti- angiogenic factors) ; and, Sato Int. J. Clin. Oncol., 8:200-206 (2003) (e.g., Table 1 lists anti-angiogenic agents used in clinical trials).

The terms "cancer" and "cancerous" refer to or describe the physiological condition in mammals that is typically characterized by unregulated cell growth/proliferation. Examples of cancer include but are not limited to, carcinoma, lymphoma, blastoma, sarcoma, and leukemia. More particular examples of such cancers include squamous cell cancer, small-cell lung cancer, pituitary cancer, esophageal cancer, astrocytoma, soft tissue sarcoma, non-small cell lung cancer, adenocarcinoma of the lung, squamous carcinoma of the lung, cancer of the peritoneum, hepatocellular cancer, gastrointestinal cancer, pancreatic cancer, glioblastoma, cervical cancer, ovarian cancer, liver cancer, bladder cancer, hepatoma, breast cancer, colon cancer, colorectal cancer, endometrial or uterine carcinoma, salivary gland carcinoma, kidney cancer, liver cancer, prostate cancer, vulval cancer, thyroid cancer, hepatocellular carcinoma, brain cancer, endometrial cancer, testis cancer, cholangiocarcinoma, gallbladder carcinoma, gastric cancer, melanoma, and various types of head and neck cancer.

The terms "cell proliferative disorder" and "proliferative disorder" refer to disorders that are associated with some degree of abnormal cell proliferation. In one embodiment, the cell proliferative disorder is cancer.

The terms "cancer", "cancerous", "cell proliferative disorder", "proliferative disorder" and "tumor" are not mutually exclusive as referred to herein.

A "chemotherapeutic agent" is a chemical compound useful in the treatment of cancer. Examples of chemotherapeutic agents include alkylating agents such as thiotepa and cyclosphosphamide

(CYTOXAN®) ; alkyl sulfonates such as busulfan, improsulfan and piposulfan; aziridines such as benzodopa, carboquone, meturedopa, and uredopa; ethylenimines and methylamelamines including altretamine, triethylenemelamine, triethylenephosphoramide, triethylenethiophosphoramide and trimethylomelamine; acetogenins (especially bullatacin and bullatacinone) ; delta-9-tetrahydrocannabinol (dronabinol, MARINOL®) ; beta-lapachone; lapachol; colchicines; betulinic acid; a camptothecin (including the synthetic analogue topotecan (HYCAMTIN®), CPT-1 1 (irinotecan, CAMPTOSAR®),

acetylcamptothecin, scopolectin, and 9-aminocamptothecin) ; bryostatin; callystatin; CC-1065 (including its adozelesin, carzelesin and bizelesin synthetic analogues) ; podophyllotoxin; podophyllinic acid; teniposide; cryptophycins (particularly cryptophycin 1 and cryptophycin 8) ; dolastatin; duocarmycin (including the synthetic analogues, KW-2189 and CB1 -TM1 ) ; eleutherobin; pancratistatin; a sarcodictyin; spongistatin; nitrogen mustards such as chlorambucil, chlornaphazine, chlorophosphamide, estramustine, ifosfamide, mechlorethamine, mechlorethamine oxide hydrochloride, melphalan, novembichin, phenesterine, prednimustine, trofosfamide, uracil mustard; nitrosoureas such as carmustine, chlorozotocin, fotemustine, lomustine, nimustine, and ranimnustine; antibiotics such as the enediyne antibiotics (e.g., calicheamicin, especially calicheamicin gammal I and calicheamicin omegall (see, e.g., Nicolaou et al., Angew. Chem Intl. Ed. Engl., 33: 183-1 86 (1994)) ; CDP323, an oral alpha-4 integrin inhibitor; dynemicin, including dynemicin A; an esperamicin; as well as neocarzinostatin chromophore and related chromoprotein enediyne antibiotic chromophores), aclacinomysins, actinomycin, authramycin, azaserine, bleomycins, cactinomycin, carabicin, caminomycin, carzinophilin, chromomycins, dactinomycin, daunorubicin, detorubicin, 6-diazo-5-oxo-L-norleucine, doxorubicin (including ADRIAMYCIN®, morpholino-doxorubicin, cyanomorpholino-doxorubicin, 2-pyrrolino-doxorubicin, doxorubicin HCI liposome injection (DOXIL®), liposomal doxorubicin TLC D-99 (MYOCET®), peglylated liposomal doxorubicin (CAELYX®), and deoxydoxorubicin), epirubicin, esorubicin, idarubicin, marcellomycin, mitomycins such as mitomycin C, mycophenolic acid, nogalamycin, olivomycins, peplomycin, porfiromycin, puromycin, quelamycin, rodorubicin, streptonigrin, streptozocin, tubercidin, ubenimex, zinostatin, zorubicin; anti-metabolites such as methotrexate, gemcitabine (GEMZAR®), tegafur (UFTORAL®), capecitabine (XELODA®), an epothilone, and 5-fluorouracil (5-FU) ; combretastatin; folic acid analogues such as denopterin, methotrexate, pteropterin, trimetrexate; purine analogs such as fludarabine, 6-mercaptopurine, thiamiprine, thioguanine; pyrimidine analogs such as ancitabine, azacitidine, 6-azauridine, carmofur, cytarabine, dideoxyuridine, doxifluridine, enocitabine, floxuridine; androgens such as calusterone, dromostanolone propionate, epitiostanol, mepitiostane, testolactone; anti-adrenals such as

aminoglutethimide, mitotane, trilostane; folic acid replenisher such as frolinic acid; aceglatone;

aldophosphamide glycoside; aminolevulinic acid; eniluracil; amsacrine; bestrabucil; bisantrene;

edatraxate; defofamine; demecolcine; diaziquone; elformithine; elliptinium acetate; an epothilone;

etoglucid; gallium nitrate; hydroxyurea; lentinan; lonidainine; maytansinoids such as maytansine and ansamitocins; mitoguazone; mitoxantrone; mopidanmol; nitraerine; pentostatin; phenamet; pirarubicin; losoxantrone; 2-ethylhydrazide; procarbazine; PSK® polysaccharide complex (JHS Natural Products, Eugene, Oreg.) ; razoxane; rhizoxin; sizofuran; spirogermanium ; tenuazonic acid; triaziquone; 2,2', 2'- trichlorotriethylamine; trichothecenes (especially T-2 toxin, verracurin A, roridin A and anguidine) ;

urethan; vindesine (ELDISINE®, FILDESIN®) ; dacarbazine; mannomustine; mitobronitol; mitolactol; pipobroman; gacytosine; arabinoside ("Ara-C") ; thiotepa; taxoid, e.g., paclitaxel (TAXOL®, Bristol-Myers Squibb Oncology, Princeton, N.J.), albumin-engineered nanoparticle formulation of paclitaxel

(ABRAXAN E™), and docetaxel (TAXOTERE®, Rhome-Poulene Rorer, Antony, France) ; chloranbucil; 6- thioguanine; mercaptopurine; methotrexate; platinum agents such as cisplatin, oxaliplatin (e.g.,

ELOXATIN®), and carboplatin; vincas, which prevent tubulin polymerization from forming microtubules, including vinblastine (VELBAN®), vincristine (ONCOVIN®), vindesine (ELDISINE®, FILDESIN®), and vinorelbine (NAVELBIN E®) ; etoposide (VP-16) ; ifosfamide; mitoxantrone; leucovorin; novantrone;

edatrexate; daunomycin; aminopterin; ibandronate; topoisomerase inhibitor RFS 2000;

difluoromethylornithine (DMFO) ; retinoids such as retinoic acid, including bexarotene (TARGRETIN®) ; bisphosphonates such as clodronate (for example, BONEFOS® or OSTAC®), etidronate (DIDROCAL®), NE-58095, zoledronic acid/zoledronate (ZOMETA®), alendronate (FOSAMAX®), pamidronate

(AREDIA®), tiludronate (SKELID®), or risedronate (ACTONEL®) ; troxacitabine (a 1 ,3-dioxolane nucleoside cytosine analog) ; antisense oligonucleotides, particularly those that inhibit expression of genes in signaling pathways implicated in aberrant cell proliferation, such as, for example, PKC-alpha, Raf, H-Ras, and epidermal growth factor receptor (EGF-R) (e.g., erlotinib (Tarceva™)) ; and VEGF-A that reduce cell proliferation; vaccines such as THERATOPE® vaccine and gene therapy vaccines, for example, ALLOVECTIN® vaccine, LEUVECTIN® vaccine, and VAXID® vaccine; topoisomerase 1 inhibitor (e.g., LURTOTECAN®) ; rm RH (e.g., ABARELIX®) ; BAY439006 (sorafenib; Bayer) ; SU-1 1248 (sunitinib, SUTENT®, Pfizer) ; perifosine, COX-2 inhibitor (e.g. celecoxib or etoricoxib), proteosome inhibitor (e.g. PS341 ) ; bortezomib (VELCADE®) ; CCI-779; tipifarnib (R1 1577) ; orafenib, ABT510; Bcl-2 inhibitor such as oblimersen sodium (GENASENSE®) ; pixantrone; EGFR inhibitors; tyrosine kinase inhibitors; serine-threonine kinase inhibitors such as rapamycin (sirolimus, RAPAMUNE®) ;

farnesyltransferase inhibitors such as lonafarnib (SCH 6636, SARASAR™) ; and pharmaceutically acceptable salts, acids or derivatives of any of the above; as well as combinations of two or more of the above such as CHOP, an abbreviation for a combined therapy of cyclophosphamide, doxorubicin, vincristine, and prednisolone; and FOLFOX, an abbreviation for a treatment regimen with oxaliplatin (ELOXATIN™) combined with 5-FU and leucovorin, and pharmaceutically acceptable salts, acids or derivatives of any of the above; as well as combinations of two or more of the above.

Chemotherapeutic agents as defined herein include "anti-hormonal agents" or "endocrine therapeutics" which act to regulate, reduce, block, or inhibit the effects of hormones that can promote the growth of cancer. They may be hormones themselves, including, but not limited to: anti-estrogens with mixed agonist/antagonist profile, including, tamoxifen (NOLVADEX®), 4-hydroxytamoxifen, toremifene (FARESTON®), idoxifene, droloxifene, raloxifene (EVISTA®), trioxifene, keoxifene, and selective estrogen receptor modulators (SERMs) such as SERM3; pure anti-estrogens without agonist properties, such as fulvestrant (FASLODEX®), and EM800 (such agents may block estrogen receptor (ER) dimerization, inhibit DNA binding, increase ER turnover, and/or suppress ER levels) ; aromatase inhibitors, including steroidal aromatase inhibitors such as formestane and exemestane (AROMAS IN®), and nonsteroidal aromatase inhibitors such as anastrazole (AR I M ID EX®), letrozole (FEMARA®) and aminoglutethimide, and other aromatase inhibitors include vorozole (RIVISOR®), megestrol acetate

(MEGASE®), fadrozole, and 4(5)-imidazoles; lutenizing hormone-releasing hormone agonists, including leuprolide (LU PRON® and ELIGARD®), goserelin, buserelin, and tripterelin; sex steroids, including progestines such as megestrol acetate and medroxyprogesterone acetate, estrogens such as

diethylstilbestrol and premarin, and androgens/retinoids such as fluoxymesterone, all transretionic acid and fenretinide; onapristone; anti-progesterones; estrogen receptor down-regulators (ERDs) ; anti- androgens such as flutamide, nilutamide and bicalutamide; and pharmaceutically acceptable salts, acids or derivatives of any of the above; as well as combinations of two or more of the above.

The term "cytotoxic agent" as used herein refers to a substance that inhibits or prevents a cellular function and/or causes cell death or destruction. Cytotoxic agents include, but are not limited to, radioactive isotopes (e.g., At^{21 1} , I¹³¹ , I¹²⁵, Y⁹⁰, Re¹⁸⁶, Re¹⁸⁸, Sm¹⁵³, Bi²¹², P³², Pb²¹² and radioactive isotopes of Lu) ; chemotherapeutic agents or drugs (e.g., methotrexate, adriamicin, vinca alkaloids (vincristine, vinblastine, etoposide), doxorubicin, melphalan, mitomycin C, chlorambucil, daunorubicin or other intercalating agents) ; growth inhibitory agents; enzymes and fragments thereof such as nucleolytic enzymes; antibiotics; toxins such as small molecule toxins or enzymatically active toxins of bacterial, fungal, plant or animal origin, including fragments and/or variants thereof; and the various antitumor or anticancer agents disclosed herein.

An "effective amount" of an agent, e.g., a pharmaceutical formulation, refers to an amount effective, at dosages and for periods of time necessary, to achieve the desired therapeutic or prophylactic result.

A "growth inhibitory agent" when used herein refers to a compound or composition which inhibits growth of a cell either in vitro or in vivo. Thus, the growth inhibitory agent may be one which significantly reduces the percentage of cells in S phase. Examples of growth inhibitory agents include agents that block cell cycle progression (at a place other than S phase), such as agents that induce G1 arrest and M- phase arrest. Classical M-phase blockers include the vincas (vincristine and vinblastine), taxanes, and topoisomerase II inhibitors such as doxorubicin, epirubicin, daunorubicin, etoposide, and bleomycin. Those agents that arrest G1 also spill over into S-phase arrest, for example, DNA alkylating agents such as tamoxifen, prednisone, dacarbazine, mechlorethamine, cisplatin, methotrexate, 5-fluorouracil, and ara- C. Further information can be found in Mendelsohn and Israel, eds., The Molecular Basis of Cancer, Chapter 1 , entitled "Cell cycle regulation, oncogenes, and antineoplastic drugs" by Murakami et al. (W.B. Saunders, Philadelphia, 1995), e.g., p. 13. The taxanes (paclitaxel and docetaxel) are anticancer drugs both derived from the yew tree. Docetaxel (TAXOTERE®, Rhone-Poulenc Rorer), derived from the European yew, is a semisynthetic analogue of paclitaxel (TAXOL®, Bristol-Myers Squibb). Paclitaxel and docetaxel promote the assembly of microtubules from tubulin dimers and stabilize microtubules by preventing depolymerization, which results in the inhibition of mitosis in cells.

As used herein, "codon set" refers to a set of different nucleotide triplet sequences used to encode desired variant amino acids. A set of oligonucleotides can be synthesized, for example, by solid phase synthesis, including sequences that represent all possible combinations of nucleotide triplets provided by the codon set and that will encode the desired group of amino acids. A standard form of codon designation is that of the IUB code, which is known in the art and described herein. A codon set typically is represented by 3 capital letters in italics, e.g., NNK, NNS, XYZ, DVK, and the like (e.g., NNK codon refers to N = A/T/G/C at positions 1 and 2 in the codon and K= G/T at equimolar ratio in position 3 to encode all 20 natural amino acids). Synthesis of oligonucleotides with selected nucleotide

"degeneracy" at certain positions is well known in that art, for example the TRIM approach (Knappek et al., J. Mol. Biol. 296:57-86, 1999) ; Garrard and Henner, Gene 128:103, 1 993). Such sets of

oligonucleotides having certain codon sets can be synthesized using commercial nucleic acid synthesizers (available from , for example, Applied Biosystems, Foster City, CA), or can be obtained commercially (for example, from Life Technologies, Rockville, MD). Therefore, a set of oligonucleotides synthesized having a particular codon set will typically include a plurality of oligonucleotides with different sequences, the differences established by the codon set within the overall sequence. Oligonucleotides, as used according to the invention, have sequences that allow for hybridization to a variable domain nucleic acid template and also can, but do not necessarily, include restriction enzyme sites useful for, for example, cloning purposes.

An "isolated nucleic acid" refers to a nucleic acid molecule that has been separated from a component of its natural environment. An isolated nucleic acid includes a nucleic acid molecule contained in cells that ordinarily contain the nucleic acid molecule, but the nucleic acid molecule is present extrachromosomally or at a chromosomal location that is different from its natural chromosomal location.

The term "control sequences" refers to DNA sequences necessary for the expression of an operably linked coding sequence in a particular host organism. The control sequences that are suitable for prokaryotes, for example, include a promoter, optionally an operator sequence, and a ribosome binding site. Eukaryotic cells are known to utilize promoters, polyadenylation signals, and enhancers.

The terms "host cell," "host cell line," and "host cell culture" are used interchangeably and refer to cells into which exogenous nucleic acid has been introduced, including the progeny of such cells. Host cells include "transformants" and "transformed cells," which include the primary transformed cell and progeny derived therefrom without regard to the number of passages. Progeny may not be completely identical in nucleic acid content to a parent cell, but may contain mutations. Mutant progeny that have the same function or biological activity as screened or selected for in the originally transformed cell are included herein.

Nucleic acid is "operably linked" when it is placed into a functional relationship with another nucleic acid sequence. For example, DNA for a presequence or secretory leader is operably linked to DNA for a polypeptide if it is expressed as a preprotein that participates in the secretion of the polypeptide; a promoter or enhancer is operably linked to a coding sequence if it affects the transcription of the sequence; or a ribosome binding site is operably linked to a coding sequence if it is positioned so as to facilitate translation. Generally, "operably linked" means that the DNA sequences being linked are contiguous, and, in the case of a secretory leader, contiguous and in reading phase. However, enhancers do not have to be contiguous. Linking is accomplished by ligation at convenient restriction sites. If such sites do not exist, the synthetic oligonucleotide adaptors or linkers are used in accordance with conventional practice.

"Percent (%) amino acid sequence identity" with respect to the polypeptide sequences identified herein is defined as the percentage of amino acid residues in a candidate sequence that are identical with the amino acid residues in the polypeptide being compared, after aligning the sequences and introducing gaps, if necessary, to achieve the maximum percent sequence identity, and not considering any conservative substitutions as part of the sequence identity. Alignment for purposes of determining percent amino acid sequence identity can be achieved in various ways that are within the skill in the art, for instance, using publicly available computer software such as BLAST, BLAST-2, ALIGN or Megalign (DNASTAR) software. Those skilled in the art can determine appropriate parameters for measuring alignment, including any algorithms needed to achieve maximal alignment over the full-length of the sequences being compared. For purposes herein, however, % amino acid sequence identity values are generated using the sequence comparison computer program ALIGN-2. The ALIGN-2 sequence comparison computer program was authored by Genentech, Inc. and the source code has been filed with user documentation in the U.S. Copyright Office, Washington D.C., 20559, where it is registered under U.S. Copyright Registration No. TXU510087. The ALIGN-2 program is publicly available through Genentech, Inc., South San Francisco, California. The ALIGN-2 program should be compiled for use on a UN IX operating system, preferably digital UN IX V4.0D. All sequence comparison parameters are set by the ALIGN-2 program and do not vary.

In situations where ALIGN-2 is employed for amino acid sequence comparisons, the % amino acid sequence identity of a given amino acid sequence A to, with, or against a given amino acid sequence B (which can alternatively be phrased as a given amino acid sequence A that has or comprises a certain % amino acid sequence identity to, with, or against a given amino acid sequence B) is calculated as follows:

100 times the fraction X/Y where X is the number of amino acid residues scored as identical matches by the sequence alignment program ALIGN-2 in that program's alignment of A and B, and where Y is the total number of amino acid residues in B. It will be appreciated that where the length of amino acid sequence A is not equal to the length of amino acid sequence B, the % amino acid sequence identity of A to B will not equal the % amino acid sequence identity of B to A. Unless specifically stated otherwise, all % amino acid sequence identity values used herein are obtained as described in the immediately preceding paragraph using the ALIGN-2 computer program .

The amino acid sequences described herein are contiguous amino acid sequences unless otherwise specified.

The term "package insert" is used to refer to instructions customarily included in commercial packages of therapeutic products, that contain information about the indications, usage, dosage, administration, combination therapy, contraindications and/or warnings concerning the use of such therapeutic products.

The term "pharmaceutical composition" refers to a preparation which is in such form as to permit the biological activity of an active ingredient contained therein to be effective, and which contains no additional components which are unacceptably toxic to a subject to which the formulation would be administered.

A "pharmaceutically acceptable carrier" refers to an ingredient in a pharmaceutical formulation, other than an active ingredient, which is nontoxic to a subject., A pharmaceutically acceptable carrier includes, but is not limited to, a buffer, excipient, stabilizer, or preservative.

The term "prodrug" as used in this application refers to a precursor or derivative form of a pharmaceutically active substance that is less cytotoxic to tumor cells compared to the parent drug and is capable of being enzymatically activated or converted into the more active parent form . See, e.g., Wilman, "Prodrugs in Cancer Chemotherapy" Biochemical Society Transactions, 14, pp. 375-382, 615th Meeting Belfast (1986) and Stella et al., "Prodrugs: A Chemical Approach to Targeted Drug Delivery," Directed Drug Delivery, Borchardt et al., (ed.), pp. 247-267, Humana Press (1985) . The prodrugs of this invention include, but are not limited to, phosphate-containing prodrugs, thiophosphate-containing prodrugs, sulfate-containing prodrugs, peptide-containing prodrugs, D-amino acid-modified prodrugs, glycosylated prodrugs, β-lactam-containing prodrugs, optionally substituted phenoxyacetamide-containing prodrugs or optionally substituted phenylacetamide-containing prodrugs, 5-fluorocytosine and other 5- fluorouridine prodrugs which can be converted into the more active cytotoxic free drug. Examples of cytotoxic drugs that can be derivatized into a prodrug form for use in this invention include, but are not limited to, those chemotherapeutic agents described above.

The term "precancerous" refers to a condition or a growth that typically precedes or develops into a cancer.

By "reduce or inhibit" is meant the ability to cause an overall decrease preferably of 20% or greater, more preferably of 50% or greater, and most preferably of 75%, 85%, 90%, 95%, or greater. Reduce or inhibit can refer to the symptoms of the disorder being treated, the presence or size of metastases, the size of the primary tumor, or the size or number of the blood vessels in angiogenic disorders.

A "subject" is a vertebrate, preferably a mammal, more preferably a human. Mammals include, but are not limited to, farm animals (such as cows, and sheep), sport animals, pets (such as cats, dogs and horses), primates (e.g., humans and non-human primates such as monkeys), and rodents (e.g., mice and rats).

The term "therapeutically effective amount" refers to an amount of an antibody or antibody fragment to treat a disease or disorder in a subject. In the case of an allergic, inflammatory, or autoimmune disease (e.g., asthma, arthritis, etc.), the therapeutically effective amount of the antibody or antibody fragment (e.g., an anti-Ang2 antibody, including bispecific anti-Ang2 antibodies that bind to Ang2 and a second biological molecule, e.g., VEGF, such as anti-Ang2/Anti-VEGF antibodies) may ameliorate or treat the disease, or prevent, reduce, ameliorate, or treat symptoms associated with the disease. In the case of a proliferative disease (e.g., a cancerous tumor), the therapeutically effective amount of the antibody or antibody fragment may reduce the number of cancer cells; reduce the primary tumor size; inhibit (i.e., slow to some extent and preferably stop) cancer cell infiltration into peripheral organs; inhibit (i.e., slow to some extent and preferably stop) tumor metastasis; inhibit, to some extent, tumor growth; and/or relieve to some extent one or more of the symptoms associated with the disorder. To the extent the antibody or antibody fragment may prevent growth and/or kill existing cancer cells, it may be cytostatic and/or cytotoxic. For cancer therapy, efficacy in vivo can, for example, be measured by assessing the duration of survival, time to disease progression (TTP), duration of disease free survival (DFS), duration of progression free survival (PFS), the response rates (RR), duration of response, and/or quality of life.

As used herein, "treatment" (and grammatical variations thereof such as "treat" or "treating") refers to clinical intervention in an attempt to alter the natural course of the individual being treated, and can be performed either for prophylaxis or during the course of clinical pathology. Desirable effects of treatment include, but are not limited to, preventing occurrence or recurrence of disease, alleviation of symptoms, diminishment of any direct or indirect pathological consequences of the disease, preventing metastasis, decreasing the rate of disease progression, amelioration or palliation of the disease state, and remission or improved prognosis. In some embodiments, antibodies of the invention are used to delay development of a disease or to slow the progression of a disease.

"Tumor", as used herein, refers to all neoplastic cell growth and proliferation, whether malignant or benign, and all pre-cancerous and cancerous cells and tissues. The term "vector," as used herein, is intended to refer to a nucleic acid molecule capable of transporting another nucleic acid to which it has been linked. One type of vector is a "plasmid", which refers to a circular double stranded DNA loop into which additional DNA segments may be ligated.

Another type of vector is a phage vector. Another type of vector is a viral vector, wherein additional DNA segments may be ligated into the viral genome. Certain vectors are capable of autonomous replication in a host cell into which they are introduced (e.g., bacterial vectors having a bacterial origin of replication and episomal mammalian vectors). Other vectors (e.g., non-episomal mammalian vectors) can be integrated into the genome of a host cell upon introduction into the host cell, and thereby are replicated along with the host genome. Moreover, certain vectors are capable of directing the expression of genes to which they are operatively linked. Such vectors are referred to herein as "recombinant expression vectors" (or simply, "recombinant vectors" or "expression vectors"). In general, expression vectors of utility in recombinant DNA techniques are often in the form of plasmids. In the present specification, "plasmid" and "vector" may be used interchangeably.

As used herein, "library" refers to a plurality of antibody or antibody fragment sequences, or the nucleic acids that encode these sequences, the sequences being different in the variant amino acid(s) or combinations thereof that are introduced into these sequences.

A "mutation" is a deletion, insertion, or substitution of a nucleotide(s) relative to a reference nucleotide sequence, such as a wild type sequence.

"Phage display" is a technique by which variant polypeptides are displayed as fusion proteins to at least a portion of coat protein on the surface of phage, e.g., filamentous phage, particles. A utility of phage display lies in the fact that large libraries of randomized protein variants can be rapidly and efficiently sorted for those sequences that bind to a target antigen with high affinity. Display of peptide and protein libraries on phage has been used for screening millions of polypeptides for those with specific binding properties. Polyvalent phage display methods have been used for displaying small random peptides and small proteins through fusions to either gene m or gene VI I I of filamentous phage. See, for example, Wells et al., Curr. Opin. Struct. Biol., 3:355-362, 1 992, and references cited therein. In monovalent phage display, the members of a protein or peptide library are each fused to a gene I I I protein or a portion thereof, and expressed at low levels in the presence of wild type gene II I protein so that phage particles display one copy (or none) of the fusion proteins. Avidity effects are reduced relative to polyvalent phage so that sorting is on the basis of intrinsic ligand affinity, and phagemid vectors are used, which simplify DNA manipulations. See, for example, Lowman et al., Methods: A companion to Methods in Enzymology, 3:205-0216, 1991 .

A "variant" or "mutant" of a starting or reference polypeptide (e.g., a reference antibody or its variable domain(s)/HVR(s)), is a polypeptide that 1 ) has an amino acid sequence different from that of the starting or reference polypeptide and 2) was derived from the starting or reference polypeptide through either natural or artificial (man-made) mutagenesis. Such variants include, for example, deletions from , and/or insertions into and/or substitutions of, residues within the amino acid sequence of the polypeptide of interest, referred to herein as "amino acid residue alterations." Thus, a variant HVR refers to a HVR comprising a variant sequence with respect to a starting or reference polypeptide sequence (such as that of a source antibody or antigen binding fragment). An amino acid residue alteration, in this context, refers to an amino acid different from the amino acid at the corresponding position in a starting or reference polypeptide sequence (such as that of a reference antibody or fragment thereof). Any combination of deletion, insertion, and substitution may be made to arrive at the final variant or mutant construct, provided that the final construct possesses the desired functional characteristics. The amino acid changes also may alter post-translational processes of the polypeptide, such as changing the number or position of glycosylation sites.

A "wild type" or "reference" sequence or the sequence of a "wild type" or "reference" protein/polypeptide, such as an HVR or variable domain of a reference antibody, may be the reference sequence from which variant polypeptides are derived through the introduction of mutations. In general, the "wild type" sequence for a given protein is the sequence that is most common in nature. Similarly, a "wild type" gene sequence is the sequence for that gene which is most commonly found in nature.

Mutations may be introduced into a "wild type" gene (and thus the protein it encodes) either through natural processes or through man-induced means. The products of such processes are "variant" or "mutant" forms of the original "wild type" protein or gene.

A "reference antibody," as used herein, refers to an antibody or fragment thereof whose antigen- binding sequence serves as the template sequence upon which diversification according to the criteria described herein is performed. An antigen-binding sequence generally includes an antibody variable region, preferably at least one HVR, preferably including framework regions.

By "massively parallel sequencing" or "massive parallel sequencing," also known in the art as "next-generation sequencing," or "second generation sequencing," is meant any high-throughput nucleic acid sequencing approach. These approaches typically involve parallel sequencing of a large number (e.g., thousands, millions, or billions) of spatially separated, clonally amplified DNA templates or single DNA molecules. See, for example, Metzker, Nature Reviews Genetics 1 1 : 31 -36, 2010.

"Enriched," as used herein, means that an entity (e.g., an amino acid residue alteration) is present at a higher frequency in a sorted library as compared to a corresponding reference library (e.g., an unsorted library, or a library that has been sorted for a different or non-relevent antigen). In contrast, "depleted" means that an entity (for example, an amino acid residue alteration) is present at a lower frequency in a sorted library as compared to a corresponding reference library (e.g., an unsorted library, or a library that has been sorted for a different or non-relevent antigen). The term "neutral," when used in reference to methods of identifying amino acid residue variants, means that an entity is neither enriched nor depleted, in other words, it is present at approximately the same frequency in a sorted library as compared to a corresponding reference library (e.g., an unsorted library, or a library that has been sorted for a different or non-relevent antigen). II. Compositions and Methods

The invention provides novel antibodies that bind to Ang2 (including novel bispecific anti-Ang2 antibodies that bind to Ang2 and a second biological molecule, e.g., VEGF, such as anti-Ang2/Anti-VEGF antibodies). Antibodies of the invention are useful, for example, for reducing angiogenesis and for treating or delaying the progression of a disorder associated with pathological angiogenesis (e.g., ocular disorders or cell proliferative disorders). A. Exemplary Anti-Ang2 and Anti-Ang2/Anti- VEGF Antibodies

The invention provides anti-Ang2 antibodies useful for, e.g., treatment of conditions involving pathological angiogenesis (e.g., ocular disorders and cell proliferative disorders). In some instances, the invention provides bispecific anti-Ang2 antibodies that bind to Ang2 and a second biological molecule, e.g., VEGF. In some instances, the invention provides anti-Ang2 antibodies that are bispecific (e.g., anti- Ang2/anti-VEGF antibodies). In some instances, anti-Ang2/anti-VEGF antibodies are dual-specific.

In one example, the anti-Ang2 antibodies bind to an epitope on Ang2 including one or more amino acid residues (e.g., 1 , 2, 3, 4, 5, or 6 amino acid residues) selected from the group consisting of Cys433, Cys435, Met440, Leu441 , Cys450, and Gly451 of Ang2. For example, in some instances, the anti-Ang2 antibodies bind to an epitope on Ang2 including Cys433, Cys435, Met440, Leu441 , Cys450, or Gly451 .

In some instances the anti-Ang2 antibodies bind to an epitope on Ang2 including Cys433 and Cys435; Cys433 and Met440; Cys433 and Leu441 ; Cys433 and Cys450; Cys433 and Gly451 ; Cys435 and Met440; Cys435 and Leu441 , Cys435 and Cys450, Cys435 and Gly451 ; Met440 and Leu441 ;

Met440 and Cys450; Met440 and Gly451 ; Leu441 and Cys450; Leu441 and Gly451 ; or Cys450 and Gly451 . In some instances, the anti-Ang2 antibodies bind to an epitope on Ang2 including three or more, four or more, five or more, or all six residues. In some instances, the anti-Ang2 antibodies bind to an epitope on Ang2 including Cys433, Cys435, Met440, Leu441 , and Cys450. In some instances, the anti- Ang2 antibodies bind to an epitope on Ang2 including Cys433, Cys435, Met440, Leu441 , Cys450, and Gly451 .

In some instances, any of the above anti-Ang2 antibodies may bind to an epitope on Ang2 that further includes one or more additional amino acid residues (e.g., 1 , 2, or 3 amino acid residues) selected from the group consisting of Phe469, Tyr475, and Ser480 of Ang2. For example, in some instances, the anti-Ang2 antibodies bind to an epitope on Ang2 that further includes Phe469, Tyr475, or Ser480. In some instances, the anti-Ang2 antibodies bind to an epitope on Ang2 further including Phe469 and Tyr475; Phe469 and Ser480; or Tyr475 and Ser480. In some instances, the anti-Ang2 antibodies bind to an epitope on Ang2 further including Phe469, Tyr475, and Ser480.

In some instances, any of the above anti-Ang2 antibodies may bind to an epitope on Ang2 that further includes one or more amino acid residues selected from the group consisting of Cys433, Cys435, Met440, Leu441 , Cys450, and Gly451 of Ang2, the epitope on Ang2 further includes one or more additional amino acid residues (e.g., 1 , 2, 3, 4, 5, or 6 amino acid residues) selected from the group consisting of Lys432, Ile434, Asp448, Ala449, Pro452, and Tyr476 of Ang2. For instance, in some instances, the anti-Ang2 antibodies bind to an epitope on Ang2 that further includes Lys432, Ile434, Asp448, Ala449, Pro452, or Tyr476. In another example, the anti-Ang2 antibodies bind to an epitope on Ang2 that consists of amino acid residues Lys432, Cys433, Ile434, Cys435, Met440, Leu441 , Asp448, Ala449, Cys450, Gly451 , Pro452, Phe469, Tyr475, Tyr476, and Ser480 of Ang2.

In some instances, any of the above anti-Ang2 antibodies includes a paratope that includes one or more amino acid residues (e.g., 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acid residues) selected from the group consisting of light chain variable region amino acid residues Gln27; Phe27a; Leu28, Met28, or Ala28; Ser29; Ser30; Phe31 ; Ser67; Gly68; Gly91 ; Leu92; Leu93, Lys93, or Val93; Ser94 or Pro94; and Leu96 and the heavy chain variable region amino acid residues Trp33; His35, Tyr35, or Asp35; Tyr58, Ile58, Trp58, or Leu58; Phe97; Phe98; Leu99 or Ala99; and Tyrl OOa.

For example, in some instances, the anti-Ang2 antibody includes a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Leu93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2 antibody includes a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti- Ang2 antibody includes a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Val93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2 antibody includes a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Met28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2 antibody includes a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Ala28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2 antibody includes a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Pro94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2 antibody includes a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, Asp35, Tyr58, Phe97, Phe98, Leu99, or Tyr1 00a. In other instances, the anti-Ang2 antibody includes a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Ile58, Phe97,

Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2 antibody includes a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Trp58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2 antibody includes a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Leu58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2 antibody includes a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Ala99, or Tyrl OOa. In yet other instance, the anti-Ang2 antibody includes a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Leu93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, Tyr35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa.

For example, in some instances, these antibodies have a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Leu93, Ser94, and Leu96. In other instances, these antibodies have a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, and Leu96. In other instances, these antibodies have a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Met28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, and Leu96. In other instances, these antibodies have a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Ala28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, and Leu96. In other instances, these antibodies have a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Pro94, and Leu96. In yet other instances these antibodies have a paratope that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Val93, Ser94, and Leu96.

In yet other instances, these antibodies have a paratope that includes heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, and Tyrl OOa. In yet another instance, these antibodies have a paratope that includes heavy chain variable region amino acid residues Trp33, Tyr35, Tyr58, Phe97, Phe98, Leu99, and Tyrl OOa. In yet other instances, these antibodies have a paratope that includes heavy chain variable region amino acid residues Trp33, Asp35, Tyr58, Phe97, Phe98, Leu99, and Tyr1 00a. In some instances, these antibodies have a paratope that includes heavy chain variable region amino acid residues Trp33, His35, Ile58, Phe97, Phe98, Leu99, and Tyrl OOa. In some instances, these antibodies have a paratope that includes heavy chain variable region amino acid residues Trp33, His35, Trp58, Phe97, Phe98, Leu99, and Tyrl OOa. In other instances, these antibodies have a paratope that includes heavy chain variable region amino acid residues Trp33, His35, Leu58, Phe97, Phe98, Leu99, and Tyrl OOa. In some instances, these antibodies have a paratope that includes heavy chain variable region amino acid residues Trp33, His35, Trp58, Phe97, Phe98, Ala99, and Tyr100a. It is to be understood that any of the preceding antibodies may include a paratope that consists of the listed amino acid residues.

In some instances, any of the above anti-Ang2 antibodies includes at least one, two, three, four, five, or six HVRs selected from (a) an HVR-L1 comprising the amino acid sequence of

RASQFX₁SX₂FGX₃X₄ (SEQ ID NO: 26), wherein is Leu, Met, or Ala, X₂ is Ser, Lys, or Thr, X₃ is Val or Leu, and X₄ is Ala, Ser, Met, Gly, Thr, or Asn; (b) an HVR-L2 comprising the amino acid sequence of

GX₁X₂X₃LX₄X5 (SEQ ID NO: 27), wherein is Ala, Ser, or Gly, X₂ is Arg, Ser, Leu, or Lys, X₃ is Ser, Ala, or Gly, X₄ is Tyr, Val, Ala, or Glu, and X₅ is Ser, Gly, or Gin; (c) an HVR-L3 comprising the amino acid sequence of X^X^X^sXeLT (SEQ ID NO: 28), wherein is His, Gin, Phe, Trp, Tyr, or Met, X₂ is Gly, Met, or Phe, X₃ is Leu, Pro, or Ser, X₄ is Val, Leu, lie, Gly, Lys, or Arg, X₅ is Ser, His, Leu, or Pro, and X₆ is Pro or Asp; (d) an HVR-H1 comprising the amino acid sequence of DX_!X₂X₃X₄ (SEQ ID NO: 29), wherein X^ is Tyr or Ala, X₂ is Trp or Pro, X₃ is lie, Met, or Gin, and X₄ is His, Tyr, Trp, or Asp; (e) an HVR-H2 comprising the amino acid sequence of X₁X₂X₃X4X5X6GX7X8X9YADSVKG (SEQ ID NO: 30), wherein X^ is Gly or Lys, X₂ is lie, Gly, or Trp, X₃ is Thr, Val, or Asp, X₄ is Pro, Leu, or Glu, X₅ is Asp, Ala, or Leu, X₆ is Gly or Glu, X₇ is Tyr or Ala, X₈ is Thr, Glu, His, or Asp; and X₉ is Tyr, Leu, Trp, lie, or Lys; and/or (f) an HVR-H3 comprising the amino acid sequence of X₁X₂X3X4X5PX₆X₇X₈DY (SEQ ID NO: 31 ), wherein X^ is Phe, Tyr, or Met, X₂ is Val or Thr, X₃ is Phe or Pro, X₄ is Phe or Pro, X₅ is Leu or Ala, X₆ is Tyr or Trp; X₇ is Ala, Thr, Val, or Ser, and X₈ is Met, Tyr, or Trp.

For example, in some instances the anti-Ang2 antibodies include (a) an HVR-L1 comprising the amino acid sequence of RASQFX₁SX₂FGX₃X₄ (SEQ ID NO: 26), wherein X^ is Leu, Met, or Ala, X₂ is Ser, Lys, or Thr, X₃ is Val or Leu, and X₄ is Ala, Ser, Met, Gly, Thr, or Asn ; (b) an HVR-L2 comprising the amino acid sequence of GX₁X₂X₃LX₄X₅ (SEQ ID NO: 27), wherein X^ is Ala, Ser, or Gly, X₂ is Arg, Ser, Leu, or Lys, X₃ is Ser, Ala, or Gly, X₄ is Tyr, Val, Ala, or Glu, and X₅ is Ser, Gly, or Gin; (c) an HVR-L3 comprising the amino acid sequence of X^X^X^XeLT (SEQ ID NO: 28), wherein X^ is His, Gin, Phe, Trp, Tyr, or Met, X₂ is Gly, Met, or Phe, X₃ is Leu, Pro, or Ser, X₄ is Val, Leu, lie, Gly, Lys, or Arg, X₅ is Ser, His, Leu, or Pro, and X₆ is Pro or Asp; (d) an HVR-H1 comprising the amino acid sequence of

is Tyr or Ala, X₂ is Trp or Pro, X₃ is lie, Met, or Gin, and X₄ is His, Tyr, Trp, or Asp; (e) an HVR-H2 comprising the amino acid sequence of

X₁X₂X₃X₄X₅X₆GX₇X₈X₉YADSVKG (SEQ ID NO: 30), wherein X^ is Gly or Lys, X₂ is lie, Gly, or Trp, X₃ is Thr, Val, or Asp, X₄ is Pro, Leu, or Glu, X₅ is Asp, Ala, or Leu, X₆ is Gly or Glu, X₇ is Tyr or Ala, X₈ is Thr, Glu, His, or Asp; and X₉ is Tyr, Leu, Trp, lie, or Lys; and (f) an HVR-H3 comprising the amino acid sequence of

is Phe, Tyr, or Met, X₂ is Val or Thr, X₃ is Phe or Pro, X₄ is Phe or Pro, X₅ is Leu or Ala, X₆ is Tyr or Trp; X₇ is Ala, Thr, Val, or Ser, and X₈ is Met, Tyr, or Trp.

In some instances, the anti-Ang2 antibodies include at least one, two, three, four, five, or six HVRs selected from (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (b) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (c) an HVR-L3 comprising the amino acid sequence of QQGLLSPLT (SEQ ID NO: 9) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (e) an HVR-H2 comprising the amino acid sequence of G ITPAGG YTYYADSVKG (SEQ ID NO: 6) ; and/or (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7). For example, the anti-Ang2 antibodies may include (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (b) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (c) an HVR-L3 comprising the amino acid sequence of QQGLLSPLT (SEQ ID NO: 9) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (e) an HVR-H2 comprising the amino acid sequence of G ITPAGG YTYYADSVKG (SEQ ID NO: 6) ; and (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7).

In other instances, the anti-Ang2 antibodies include at least one, two, three, four, five, or six HVRs selected from (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVG (SEQ ID NO: 32) ; (b) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (c) an HVR- L3 comprising the amino acid sequence of WQGLLSPLT (SEQ ID NO: 33) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIY (SEQ ID NO: 34) ; (e) an HVR-H2 comprising the amino acid sequence of G ITPAGGYEYYADSVKG (SEQ ID NO: 35) ; and/or (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYVMDY (SEQ ID NO: 36). For example, the anti-Ang2 antibodies may include (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVG (SEQ ID NO: 32) ; (b) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (c) an HVR-L3 comprising the amino acid sequence of WQGLLSPLT (SEQ ID NO: 33) ; (d) an HVR-H 1 comprising the amino acid sequence of DYW IY (SEQ ID NO: 34) ; (e) an HVR-H2 comprising the amino acid sequence of

GITPAGGYEYYADSVKG (SEQ ID NO: 35) ; and (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYVMDY (SEQ ID NO: 36).

In some instances, any of the preceding anti-Ang2 antibodies includes at least one, two, three, or four heavy chain variable region FRs selected from (a) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFTIS (SEQ ID NO: 52) ; (b) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (c) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and/or (d) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44). For example, in some instances, the anti- Ang2 antibodies include (a) an FR-H1 comprising the amino acid sequence of

EVQLVESGGGLVQPGGSLRLSCAASGFPIS (SEQ ID NO: 41 ) ; (b) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (c) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and (d) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44).

In yet other instances, the anti-Ang2 antibodies include at least one, two, three, four, five, or six

HVRs selected from (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (b) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (c) an HVR-L3 comprising the amino acid sequence of HQGLKSPLT (SEQ ID NO: 37) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (e) an HVR-H2 comprising the amino acid sequence of GITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and/or (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7). For example, the anti-Ang2 antibodies may include (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (b) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (c) an HVR-L3 comprising the amino acid sequence of HQGLKSPLT (SEQ ID NO: 37) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (e) an HVR-H2 comprising the amino acid sequence of GITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7).

In yet other instances, the anti-Ang2 antibodies include at least one, two, three, four, five, or six HVRs selected from (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (b) an HVR-L2 comprising the amino acid sequence of GARSLYS (SEQ ID NO: 39) ; (c) an HVR- L3 comprising the amino acid sequence of HQGLVSPLT (SEQ ID NO: 40) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (e) an HVR-H2 comprising the amino acid sequence of GITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and/or (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7). For example, the anti-Ang2 antibodies may include (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (b) an HVR-L2 comprising the amino acid sequence of GARSLYS (SEQ ID NO: 39) ; (c) an HVR-L3 comprising the amino acid sequence of HQGLVSPLT (SEQ ID NO: 40) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (e) an HVR-H2 comprising the amino acid sequence of GITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7).

In some instances, any of the preceding anti-Ang2 antibodies includes at least one, two, three, or four heavy chain variable region FRs selected from (a) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFPIS (SEQ ID NO: 41 ) ; (b) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (c) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and/or (d) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44). For example, in some instances, the anti- Ang2 antibodies include (a) an FR-H1 comprising the amino acid sequence of

In other instances, any of the preceding anti-Ang2 antibodies includes at least one, two, three, or four heavy chain variable region FRs selected from (a) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFTIM (SEQ ID NO: 45) ; (b) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (c) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and/or (d) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44). For example, in some instances, the anti- Ang2 antibodies include (a) an FR-H1 comprising the amino acid sequence of

EVQLVESGGGLVQPGGSLRLSCAASGFTIM (SEQ ID NO: 45) ; (b) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (c) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and (d) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44).

The anti-Ang2 antibodies may also include a light chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 10, 1 1 , 12, 13, 14, 15, 16, 1 7, 46, 48, 51 , 78, or 79. In some instances, for example, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 1 1 . In some instances, for example, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 46. In other instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 48. In yet other instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 51 . In yet other instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 78. In yet other instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 79.

The anti-Ang2 antibodies may also include a light chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 71 , 72, 73, 74, 75, 76, or 77. In some instances, for example, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 71 . In some instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 72. In some instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 73. In some instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 74. In some instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 75. In some instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 76. In some instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 77. Any of the preceding antibodies may have a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 49.

The anti-Ang2 antibodies may also include a heavy chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 18, 47, 49, or 50. In some instances, for example, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 18. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 47. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 49. In yet other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 50.

The anti-Ang2 antibodies may also include a heavy chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 53, 54, 55, 56, 57, 58, 59, 60, 61 , 62, 63, 64, 65, 66, 67, 68, 89 or 70. In some instances, for example, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 53. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 54. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 55. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 56. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 57. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 58. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 59. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 60. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 61 . In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 62. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 63. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 64. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 65. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 66. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 67. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 68. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 69. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 70. Any of the preceding antibodies may include a light chain variable region having at least 80% sequence identity to SEQ ID NO: 48.

For example, the anti-Ang2 antibodies may also include (a) a light chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NOs: 10, 1 1 , 12, 13, 14, 15, 16, 17, 46, 48, 51 , 78, or 79 and (b) a heavy chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 18, 47, 49, or 50. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 18 and a light chain variable region having the sequence of SEQ ID NO: 1 1 , such as possessed by the anti-Ang2 antibody 5A12 4.2. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 46, such as possessed by the anti-Ang2 antibody 5A12 4.2.16.2 (also referred to as DAF16.2). In other instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 49 and a light chain variable region having the sequence of SEQ ID NO: 48, such as possessed by the anti- Ang2 antibody T.28P. In other instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 49 and a light chain variable region having the sequence of SEQ ID NO: 51 , such as possessed by the anti-Ang2 antibody T.28P-VR. In yet other instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 50 and a light chain variable region having the sequence of SEQ ID NO: 48, such as possessed by the anti-Ang2 antibody T.30M. In yet other instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 49 and a light chain variable region having the sequence of SEQ ID NO: 78, such as possessed by the anti-Ang2 antibody "5A12.4". In yet other instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 49 and a light chain variable region having the sequence of SEQ ID NO: 79, such as possessed by the anti-Ang2 antibody "5A12.5".

In another example, the anti-Ang2 antibodies may also include (a) a light chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 71 , 72, 73, 74, 75, 76, or 77 and (b) a heavy chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 49, 53, 54, 55, 56, 57, 58, 59, 60, 61 , 62, 63, 64, 65, 66, 67, 68, 89, or 70. For example, in some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 53 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 54 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 55 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 56 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 57 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 58 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 59 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 60 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 61 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 62 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 63 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 64 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 65 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 66 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 67 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 68 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 69 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 70 and a light chain variable region having the sequence of SEQ ID NO: 48.

In yet additional examples, in some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 71 . In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 72. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 73. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 74. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 75. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 76. In some instances, the anti-Ang2 antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 77.

Antibodies of the invention may also specifically bind both Ang2 and VEGF. In one example, the anti-Ang2/anti-VEGF antibodies bind an epitope on Ang2 including one or more amino acid residues (e.g., 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, or 15 amino acid residues) selected from the group consisting of Lys432, Cys433, Ile434, Cys435, Met440, Leu441 , Asp448, Ala449, Cys450, Gly451 , Pro452, Phe469, Tyr475, Tyr476, and Ser480 of Ang2. For example, in some instances, the anti- Ang2/anti-VEGF antibodies bind an epitope on Ang2 that includes Lys432, Cys433, Ile434, Cys435,

Met440, Leu441 , Asp448, Ala449, Cys450, Gly451 , Pro452, Phe469, Tyr475, Tyr476, or Ser480 of Ang2. In some instances the anti-Ang2/anti-VEGF antibodies bind an epitope on Ang2 that consists of Lys432, Cys433, Ile434, Cys435, Met440, Leu441 , Asp448, Ala449, Cys450, Gly451 , Pro452, Phe469, Tyr475, Tyr476, and Ser480 of Ang2.

In one example, the anti-Ang2/anti-VEGF antibodies bind an epitope on VEGF including one or more amino acid residues (e.g. 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 1 5, 16, 1 7, 18, 19, 20, 21 , 22, 23, 24, or 25 amino acid residues) selected from the group consisting of Phe17, Met18, Tyr21 , Gln22, Tyr25, Lys48, Asn62, Asp63, Glu64, Gly65, Leu66, Met81 , Ile83, Lys84, Pro85, His86, Gln87, Gly88, Gln89, His90, Ile91 , Lys101 , Glu103, Cys104, and Pro106 of human VEGF. For example, in some instances, the anti-Ang2/anti-VEGF antibodies bind an epitope on VEGF that includes Phe17, Met18, Tyr21 , Gln22, Tyr25, Lys48, Asn62, Asp63, Glu64, Gly65, Leu66, Met81 , Ile83, Lys84, Pro85, His86, Gln87, Gly88, Gln89, His90, Ile91 , Lys101 , Glu1 03, Cys104, or Pro106 of human VEGF. In other instances, the anti-Ang2/anti-VEGF antibodies bind an epitope on VEGF that includes Phe17, Tyr21 , and Tyr25 of human VEGF. In other instances, the anti-Ang2/anti-VEGF antibodies bind an epitope on VEGF that includes Phe17, Ile81 , and Gln89 of human VEGF. In other instances, the anti-Ang2/anti-VEGF antibodies bind an epitope on VEGF that includes Phe17, Met18, Tyr21 , Gln22, Tyr25, Lys48, Asn62, Asp63, Glu64, Gly65, Leu66, Met81 , Ile83, Lys84, Pro85, His86, Gln87, Gly88, Gln89, His90, Ile91 , Lys101 , Glu103, Cys104, or Pro106 of human VEGF. In other instances, the anti-Ang2/anti-VEGF antibodies bind an epitope on VEGF that consists of Phe17, Met18, Tyr21 , Gln22, Tyr25, Lys48, Asn62, Asp63, Glu64, Gly65, Leu66, Met81 , Ile83, Lys84, Pro85, His86, Gln87, Gly88, Gln89, His90, Ile91 , Lys101 , Glu103, Cys104, and Pro106 of human VEGF.

In some instances, any of the above anti-Ang2/anti-VEGF antibodies includes a paratope that binds Ang2 that includes one or more amino acid residues (e.g., 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 15, 16, 1 7, 18, 1 9, or 20 amino acid residues) selected from the group consisting of light chain variable region amino acid residues Gln27; Phe27a; Leu28, Met28, or Ala28; Ser29; Ser30; Phe31 ; Ser67; Gly68; Gly91 ; Leu92; Leu93, Lys93, or Val93; Ser94 or Pro94; and Leu96 and the heavy chain variable region amino acid residues Trp33; His35, Tyr35 or Asp35; Tyr58, Ile58, Trp58, or Leu58; Phe97; Phe98; Leu99 or Ala99; and Tyrl OOa.

For example, in some instances, the anti-Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Leu93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances the anti- Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Val93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Met28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Ala28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, or Tyr100a. In other instances, the anti-Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Pro94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, Asp35, Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Ile58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Trp58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Leu58, Phe97, Phe98, Leu99, or Tyrl OOa. In other instances, the anti-Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Ala99, or Tyrl OOa. In yet other instances, the anti-Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Leu93, Ser94, or Leu96 or the heavy chain variable region amino acid residues Trp33, Tyr35,

Tyr58, Phe97, Phe98, Leu99, or Tyrl OOa. It is to be understood that any of the preceding antibodies may include a paratope that binds to Ang2 that consists of the listed amino acid residues.

For example, in some instances, the anti-Ang2/anti-VEGF antibody includes a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Leu93, Ser94, and Leu96. In other instances, these antibodies include a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, and Leu96. In other instances, these antibodies include a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Met28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, and Leu96. In other instances, these antibodies include a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Ala28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Ser94, and Leu96. In other instances, these antibodies include a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Lys93, Pro94, and Leu96. In yet other instances, these antibodies include a paratope that binds Ang2 that includes light chain variable region amino acid residues Gln27, Phe27a, Leu28, Ser29, Ser30, Phe31 , Ser67, Gly68, Gly91 , Leu92, Val93, Ser94, and Leu96.

In yet other instances, these antibodies include a paratope that binds Ang2 that includes heavy chain variable region amino acid residues Trp33, His35, Tyr58, Phe97, Phe98, Leu99, and Tyrl OOa. In yet another instance, these antibodies include a paratope that binds Ang2 that includes heavy chain variable region amino acid residues Trp33, Tyr35, Tyr58, Phe97, Phe98, Leu99, and Tyrl OOa. In yet other instances, these antibodies include a paratope that binds Ang2 that includes heavy chain variable region amino acid residues Trp33, Asp35, Tyr58, Phe97, Phe98, Leu99, and Tyrl OOa. In some instances, these antibodies include a paratope that binds Ang2 that includes heavy chain variable region amino acid residues Trp33, His35, Ile58, Phe97, Phe98, Leu99, and Tyrl OOa. In some instances, these antibodies include a paratope that binds Ang2 that includes heavy chain variable region amino acid residues Trp33, His35, Trp58, Phe97, Phe98, Leu99, and Tyrl OOa. In other instances, these antibodies include a paratope that binds Ang2 that includes heavy chain variable region amino acid residues Trp33, His35, Leu58, Phe97, Phe98, Leu99, and Tyrl OOa. In some instances, these antibodies include a paratope that binds Ang2 that includes heavy chain variable region amino acid residues Trp33, His35, Trp58, Phe97, Phe98, Ala99, and Tyrl OOa.

In one example, the anti-Ang2/anti-VEGF antibodies have a paratope that binds to VEGF and includes one or more amino acid residues (e.g., 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 , 22, or 23 amino acid residues) selected from the group consisting of light chain variable region amino acid residues Leu28, Met28, or Ala28; Ser29; Phe31 ; Tyr49; Ser53; and Leu92 and the heavy chain variable region amino acid residues Ser30, Gly 30, or Met30; Asp31 ; Tyr32 or Ala32; Trp33; Ile51 ; Thr52; Pro52a or Glu52a; Ala53 or Asp53; Gly54; Gly55; Tyr56 or Ala56; Phe95 or Met95; Val96 or Thr96; Phe97; Phe98; Leu99 or Ala99; and TyM OOa. In one instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, or Leu92 or the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, or TyM OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, or Leu92 or the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Asp53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, or TyM OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, or Leu92 or the heavy chain variable region amino acid residues Met30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Asp53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, or Tyrl OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Met28, Ser29, Phe31 , Tyr49, Ser53, or Leu92 or the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, or TyM 00a. In yet another instance, the anti- Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Ala28, Ser29, Phe31 , Tyr49, Ser53, or Leu92 or the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, or Tyrl OOa. In one instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, or Leu92 or the heavy chain variable region amino acid residues Gly30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, or Tyrl OOa. In one instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, or Leu92 or the heavy chain variable region amino acid residues Ser30, Asp31 , Ala32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, or Tyr100a. In one instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, or Leu92 or the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Glu52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, or Tyrl OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, or Leu92 or the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Ala56, Phe95, Val96, Phe97, Phe98, Leu99, or Tyrl OOa. In one instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, or Leu92 or the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Met95, Val96, Phe97, Phe98, Leu99, or Tyrl OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, or Leu92 or the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Thr96, Phe97, Phe98, Leu99, or Tyr100a. In one instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, or Leu92 or the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Ala99, or Tyrl OOa.

In some instances, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, and Leu92. In other instances, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Met28, Ser29, Phe31 , Tyr49, Ser53, and Leu92. In yet other instances, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Ala28, Ser29, Phe31 , Tyr49, Ser53, and Leu92. In some instances, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Asp53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In other instances, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes heavy chain variable region amino acid residues Met30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Asp53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In some instances, the anti- Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes heavy chain variable region amino acid residues Gly30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Asp53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In some instances, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes heavy chain variable region amino acid residues Ser30, Asp31 , Ala32, Trp33, Ile51 , Thr52, Pro52a, Asp53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In some instances, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Glu52a, Asp53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In some instances, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Asp53, Gly54, Gly55, Ala56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyr100a. In some instances, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Asp53, Gly54, Gly55, Tyr56, Met95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In some instances, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Asp53, Gly54, Gly55, Tyr56, Phe95, Thr96, Phe97, Phe98, Leu99, and Tyrl OOa. In some instances, the anti- Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Asp53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Ala99, and Tyrl OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and TyM OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Asp53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and TyM OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Met30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Asp53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In another instance, the anti- Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Met28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55,

Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Ala28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Gly30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Ser30, Asp31 , Ala32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Glu52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Ala56, Phe95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In another instance, the anti- Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Met95, Val96, Phe97, Phe98, Leu99, and Tyrl OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Thr96, Phe97, Phe98, Leu99, and TyM OOa. In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that includes light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Ala99, and Tyr100a. It is to be understood that in any of the preceding antibodies, the paratope that binds to VEGF may consist of the listed amino residues.

In another instance, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF that consists of of light chain variable region amino acid residues Leu28, Ser29, Phe31 , Tyr49, Ser53, and Leu92 and the heavy chain variable region amino acid residues Ser30, Asp31 , Tyr32, Trp33, Ile51 , Thr52, Pro52a, Ala53, Gly54, Gly55, Tyr56, Phe95, Val96, Phe97, Phe98, Leu99, and TyM OOa.

In another example, the anti-Ang2/anti-VEGF antibodies have a paratope that binds to VEGF and Ang2 and includes one or more amino acid residues (e.g., 1 , 2, 3, 4, or 5 amino acid residues) selected from the group consisting of light chain variable region amino acid residues Ser30, Phe31 , and Leu92 and the heavy chain variable region amino acid residues Leu99 or Ala99 and Pro100. For example, in some instances, the anti-Ang2/anti-VEGF antibodies include a paratope that binds to VEGF and Ang2 that includes light chain variable region amino acid residues Ser30, Phe31 , or Leu92 or the heavy chain variable region amino acid residues Leu99 or Pro1 00. For example, in some instances, the anti- Ang2/anti-VEGF antibodies include a paratope that binds to VEGF and Ang2 that includes light chain variable region amino acid residues Ser30, Phe31 , or Leu92 or the heavy chain variable region amino acid residues Ala99 or Prol OO. In another instance, the anti-Ang2/anti-VEGF antibodies includes a paratope that binds to VEGF and Ang2 that includes light chain variable region amino acid residues Ser30, Phe31 , and Leu92 and the heavy chain variable region amino acid residues Leu99 and Pro100. In another instance, the anti-Ang2/anti-VEGF antibodies includes a paratope that binds to VEGF and Ang2 that includes light chain variable region amino acid residues Ser30, Phe31 , and Leu92 and the heavy chain variable region amino acid residues Ala99 and Prol OO. In another instance, the anti- Ang2/anti-VEGF antibodies includes a paratope that binds to VEGF and Ang2 that consists of light chain variable region amino acid residues Ser30, Phe31 , and Leu92 and the heavy chain variable region amino acid residues Leu99 and Prol OO. In another instance, the anti-Ang2/anti-VEGF antibodies includes a paratope that binds to VEGF and Ang2 that consists of light chain variable region amino acid residues Ser30, Phe31 , and Leu92 and the heavy chain variable region amino acid residues Ala99 and Pro100.

The anti-Ang2/anti-VEGF antibodies of the invention may include at least one, two, three, four, five, or six HVRs selected from (a) an HVR-L1 comprising the amino acid sequence of

RASQFX₁SX₂FGX₃X₄ (SEQ ID NO: 26), wherein is Leu, Met, or Ala, X₂ is Ser, Lys, or Thr, X₃ is Val or Leu, and X₄ is Ala, Ser, Met, Gly, Thr, or Asn; (b) an HVR-L2 comprising the amino acid sequence of GX₁X₂X₃LX₄X5 (SEQ ID NO: 27), wherein is Ala, Ser, or Gly, X₂ is Arg, Ser, Leu, or Lys, X₃ is Ser, Ala, or Gly, X₄ is Tyr, Val, Ala, or Glu, and X₅ is Ser, Gly, or Gin; (c) an HVR-L3 comprising the amino acid sequence of X^X^X^sXeLT (SEQ ID NO: 28), wherein is His, Gin, Phe, Trp, Tyr, or Met, X₂ is Gly, Met, or Phe, X₃ is Leu, Pro, or Ser, X₄ is Val, Leu, lie, Gly, Lys, or Arg, X₅ is Ser, His, Leu, or Pro, and X₆ is Pro or Asp; (d) an HVR-H1 comprising the amino acid sequence of DX₁X₂X₃X₄ (SEQ ID NO: 29), wherein X^ is Tyr or Ala, X₂ is Trp or Pro, X₃ is lie, Met, or Gin, and X₄ is His, Tyr, Trp, or Asp; (e) an HVR-H2 comprising the amino acid sequence of X₁X₂X₃X4X5X6GX7X8X9YADSVKG (SEQ ID NO: 30), wherein X^ is Gly or Lys, X₂ is lie, Gly, or Trp, X₃ is Thr, Val, or Asp, X₄ is Pro, Leu, or Glu, X₅ is Asp, Ala, or Leu, X₆ is Gly or Glu, X₇ is Tyr or Ala, X₈ is Thr, Glu, His, or Asp; and X₉ is Tyr, Leu, Trp, lie, or Lys; and/or (f) an HVR-H3 comprising the amino acid sequence of X₁X₂X3X4X5PX₆X7X₈DY (SEQ ID NO: 31 ), wherein X^ is Phe, Tyr, or Met, X₂ is Val or Thr, X₃ is Phe or Pro, X₄ is Phe or Pro, X₅ is Leu or Ala, X₆ is Tyr or Trp; X₇ is Ala, Thr, Val, or Ser, and X₈ is Met, Tyr, or Trp.

For example, in some instances the anti-Ang2/anti-VEGF antibodies include (a) an HVR-L1 comprising the amino acid sequence of RASQFX₁SX₂FGX₃X₄ (SEQ ID NO: 26), wherein X^ is Leu, Met, or Ala, X₂ is Ser, Lys, or Thr, X₃ is Val or Leu, and X₄ is Ala, Ser, Met, Gly, Thr, or Asn; (b) an HVR-L2 comprising the amino acid sequence of GX₁X₂X₃LX₄X₅ (SEQ ID NO: 27), wherein X^ is Ala, Ser, or Gly, X₂ is Arg, Ser, Leu, or Lys, X₃ is Ser, Ala, or Gly, X₄ is Tyr, Val, Ala, or Glu, and X₅ is Ser, Gly, or Gin; (c) an HVR-L3 comprising the amino acid sequence of X₁QX₂X₃X₄X₅X₆LT (SEQ ID NO: 28), wherein X^ is His, Gin, Phe, Trp, Tyr, or Met, X₂ is Gly, Met, or Phe, X₃ is Leu, Pro, or Ser, X₄ is Val, Leu, lie, Gly, Lys, or Arg, X₅ is Ser, His, Leu, or Pro, and X₆ is Pro or Asp; (d) an HVR-H1 comprising the amino acid sequence of

X1X2X3X4X5X6GX7X8X9YADSVKG (SEQ ID NO: 30), wherein X, is Gly or Lys, X₂ is lie, Gly, or Trp, X₃ is Thr, Val, or Asp, X₄ is Pro, Leu, or Glu, X₅ is Asp, Ala, or Leu, X₆ is Gly or Glu, X₇ is Tyr or Ala, X₈ is Thr, Glu, His, or Asp; and X₉ is Tyr, Leu, Trp, lie, or Lys; and (f) an HVR-H3 comprising the amino acid sequence of X₁X₂X₃X4X₅PX₆X₇X₈DY (SEQ ID NO: 31 ), wherein X^ is Phe, Tyr, or Met, X₂ is Val or Thr, X₃ is Phe or Pro, X₄ is Phe or Pro, X₅ is Leu or Ala, X₆ is Tyr or Trp; X₇ is Ala, Thr, Val, or Ser, and X₈ is Met, Tyr, or Trp.

In some instances, the anti-Ang2/anti-VEGF antibodies include at least one, two, three, four, five, or six HVRs selected from (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (b) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (c) an HVR-L3 comprising the amino acid sequence of QQGLLSPLT (SEQ ID NO: 9) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (e) an HVR-H2 comprising the amino acid sequence of G ITPAGGYTYYADSVKG (SEQ ID NO: 6) ; and/or (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7). For example, the anti-Ang2/anti-VEGF antibodies may include (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (b) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (c) an HVR-L3 comprising the amino acid sequence of QQGLLSPLT (SEQ ID NO: 9) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (e) an HVR-H2 comprising the amino acid sequence of G ITPAGGYTYYADSVKG (SEQ ID NO: 6) ; and (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7).

In other instances, the anti-Ang2/anti-VEGF antibodies include at least one, two, three, four, five, or six HVRs selected from (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVG (SEQ ID NO: 32) ; (b) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (c) an HVR-L3 comprising the amino acid sequence of WQGLLSPLT (SEQ ID NO: 33) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIY (SEQ ID NO: 34) ; (e) an HVR-H2 comprising the amino acid sequence of G ITPAGGYEYYADSVKG (SEQ ID NO: 35) ; and/or (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYVMDY (SEQ ID NO: 36). For example, the anti-Ang2/anti-VEGF antibodies may include (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVG (SEQ ID NO: 32) ; (b) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (c) an HVR-L3 comprising the amino acid sequence of WQGLLSPLT (SEQ ID NO: 33) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIY (SEQ ID NO: 34) ; (e) an HVR-H2 comprising the amino acid sequence of GITPAGGYEYYADSVKG (SEQ ID NO: 35) ; and (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYVMDY (SEQ ID NO: 36).

In some instances, any of the preceding anti-Ang2/anti-VEGF antibodies includes at least one, two, three, or four heavy chain variable region FRs selected from (a) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFTIS (SEQ ID NO: 52) ; (b) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (c) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and/or (d) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44). For example, in some instances, the anti-Ang2/anti-VEGF antibodies include (a) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFPIS (SEQ ID NO: 41 ) ; (b) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (c) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and (d) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44).

In yet other instances, the anti-Ang2/anti-VEGF antibodies include at least one, two, three, four, five, or six HVRs selected from (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (b) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (c) an HVR-L3 comprising the amino acid sequence of HQGLKSPLT (SEQ ID NO: 37) ; (d) an HVR-H 1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (e) an HVR-H2 comprising the amino acid sequence of G ITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and/or (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7). For example, the anti-Ang2/anti-VEGF antibodies may include (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (b) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ; (c) an HVR-L3 comprising the amino acid sequence of HQGLKSPLT (SEQ ID NO: 37) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (e) an HVR-H2 comprising the amino acid sequence of G ITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7).

In yet other instances, the anti-Ang2/anti-VEGF antibodies include at least one, two, three, four, five, or six HVRs selected from (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (b) an HVR-L2 comprising the amino acid sequence of GARSLYS (SEQ ID NO: 39) ; (c) an HVR-L3 comprising the amino acid sequence of HQGLVSPLT (SEQ ID NO: 40) ; (d) an HVR-H 1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (e) an HVR-H2 comprising the amino acid sequence of G ITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and/or (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7). For example, the anti-Ang2/anti-VEGF antibodies may include (a) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ; (b) an HVR-L2 comprising the amino acid sequence of GARSLYS (SEQ ID NO: 39) ; (c) an HVR-L3 comprising the amino acid sequence of HQGLVSPLT (SEQ ID NO: 40) ; (d) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ; (e) an HVR-H2 comprising the amino acid sequence of G ITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and (f) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7).

In some instances, any of the preceding anti-Ang2/anti-VEGF antibodies includes at least one, two, three, or four heavy chain variable region FRs selected from (a) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFPIS (SEQ ID NO: 41 ) ; (b) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (c) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and/or (d) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44). For example, in some instances, the anti-Ang2/anti-VEGF antibodies include (a) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFPIS (SEQ ID NO: 41 ) ; (b) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (c) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and (d) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44).

In other instances, any of the preceding anti-Ang2/anti-VEGF antibodies includes at least one, two, three, or four heavy chain variable region FRs selected from (a) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFTIM (SEQ ID NO: 45) ; (b) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (c) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and/or (d) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44). For example, in some instances, the anti-Ang2/anti-VEGF antibodies include (a) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFTIM (SEQ ID NO: 45) ; (b) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ; (c) an FR-H3 comprising the amino acid sequence of RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and (d) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44).

The anti-Ang2/anti-VEGF antibodies may also include a light chain variable region having at least

80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 10, 1 1 , 12, 13, 14, 15, 16, 17, 46, 48, 51 , 78, or 79. In some instances, for example, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 1 1 . In some instances, for example, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 46. In other instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 48. In yet other instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 51 . In yet other instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 78. In yet other instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 79.

The anti-Ang2/anti-VEGF antibodies may also include a light chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 71 , 72, 73, 74, 75, 76, or 77. In some instances, for example, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 71 . In some instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 72. In some instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 73. In some instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 74. In some instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 75. In some instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 76. In some instances, the antibody includes a light chain variable region having at least 80% sequence identity to SEQ ID NO: 77. Any of the preceding antibodies may have a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 49.

The anti-Ang2/anti-VEGF antibodies may also include a heavy chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 18, 47, 49, or 50. In some instances, for example, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 18. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 47. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 49. In yet other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 50.

The anti-Ang2/anti-VEGF antibodies may also include a heavy chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 53, 54, 55, 56, 57, 58, 59, 60, 61 , 62, 63, 64, 65, 66, 67, 68, 89 or 70. In some instances, for example, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 53. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 54. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 55. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 56. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 57. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 58. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 59. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 60. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 61 . In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 62. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 63. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 64. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 65. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 66. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 67. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 68. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 69. In other instances, the antibody includes a heavy chain variable region having at least 80% sequence identity to SEQ ID NO: 70. Any of the preceding antibodies may include a light chain variable region having at least 80% sequence identity to SEQ ID NO: 48.

For example, the anti-Ang2/anti-VEGF antibodies may also include (a) a light chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NOs: 10, 1 1 , 12, 13, 14, 15, 16, 17, 46, 48, 51 , 78, or 79 and (b) a heavy chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%, at least 90% (e.g., 91 %, 92%, 93%, 94%, or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 18, 47, 49, or 50. In some instances, the anti-Ang2/anti- VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 18 and a light chain variable region having the sequence of SEQ ID NO: 1 1 , such as possessed by the anti- Ang2/anti-VEGF antibody 5A12 4.2. In some instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 46, such as possessed by the anti-Ang2/anti-VEGF antibody 5A12 4.2.16.2 (also referred to as DAF16.2). In other instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 49 and a light chain variable region having the sequence of SEQ ID NO: 48, such as possessed by the anti-Ang2/anti-VEGF antibody T.28P. In other instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 49 and a light chain variable region having the sequence of SEQ ID NO: 51 , such as possessed by the anti-Ang2/anti-VEGF antibody T.28P-VR. In yet other instances, the anti- Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 50 and a light chain variable region having the sequence of SEQ ID NO: 48, such as possessed by the anti- Ang2/anti-VEGF antibody T.30M. In yet other instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 49 and a light chain variable region having the sequence of SEQ ID NO: 78, such as possessed by the anti-Ang2/anti-VEGF antibody "5A12.4". In yet other instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 49 and a light chain variable region having the sequence of SEQ ID NO: 79, such as possessed by the the anti-Ang2/anti-VEGF antibody "5A12.5". In another example, the anti-Ang2/anti-VEGF antibodies may also include (a) a light chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 71 , 72, 73, 74, 75, 76, or 77 and (b) a heavy chain variable region having at least 80% (e.g., at least 81 %, 82%, 83%, 84%, 85%, 86%, 87%, 88%, or 89%), at least 90% (e.g., 91 %, 92%, 93%, or 94%), or at least 95% (e.g., 96%, 97%, 98%, or 99%) sequence identity to, or the sequence of, SEQ ID NO: 49, 53, 54, 55, 56, 57, 58, 59, 60, 61 , 62, 63, 64, 65, 66, 67, 68, 89, or 70.

For example, in some instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 53 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 54 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 55 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 56 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 57 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 58 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 59 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 60 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 61 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 62 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 63 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 64 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 65 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 66 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 67 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti^■VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 68 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 69 and a light chain variable region having the sequence of SEQ ID NO: 48. In some instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 70 and a light chain variable region having the sequence of SEQ ID NO: 48.

In yet additional examples, in some instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 71 . In some instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 72. In some instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 73. In some instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 74. In some instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 75. In some instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 76. In some instances, the anti-Ang2/anti-VEGF antibodies include a heavy chain variable region having the sequence of SEQ ID NO: 47 and a light chain variable region having the sequence of SEQ ID NO: 77.

In a further aspect of the invention, an anti-Ang2 or anti-Ang2/anti-VEGF antibody according to any of the above embodiments is a monoclonal antibody, comprising a chimeric, humanized, or human antibody. In one embodiment, an anti-Ang2 antibody is an antibody fragment, for example, a Fv, Fab, Fab', scFv, diabody, or F(ab')₂ fragment. In another embodiment, the antibody is a full-length antibody, e.g., an intact IgG antibody (e.g., an intact lgG1 antibody) or other antibody class or isotype as defined herein.

In a further aspect, an anti-Ang2 or anti-Ang2/anti-VEGF antibody according to any of the above embodiments may incorporate any of the features, singly or in combination, as described in Sections 1 -7 below.

1. Antibody Affinity

In certain embodiments, an antibody provided herein has a dissociation constant (Kd) of < 1 μΜ, < 100 nM, < 1 0 nM, < 1 nM, < 0.1 nM, < 0.01 nM, or < 0.001 nM (e.g., 10^"8 M or less, e.g., from 10^"8 M to 10^"13 M, e.g., from 10^"9 M to 10^"13 M). In certain embodiments, a bispecific anti-Ang2/anti-VEGF antibody (e.g., a dual-specific anti-Ang2/anti-VEGF antibody) provided herein has a dissociation constant (Kd) of < 1 μΜ, < 1 00 nM, < 10 nM, < 1 nM, < 0.1 nM, < 0.01 nM, or < 0.001 nM (e.g., 10^"8 M or less, e.g., from 10^"8 M to 10^"13 M, e.g., from 10^"9 M to 10^"13 M).

In one embodiment, Kd is measured by a radiolabeled antigen binding assay (RIA). In one embodiment, an RIA is performed with the Fab version of an antibody of interest and its antigen. For example, solution binding affinity of Fabs for antigen is measured by equilibrating Fab with a minimal concentration of (¹²⁵l)-labeled antigen in the presence of a titration series of unlabeled antigen, then capturing bound antigen with an anti-Fab antibody-coated plate (see, e.g., Chen et al., J. Mol. Biol.

293:865-881 (1 999)). To establish conditions for the assay, M ICROTITER^® multi-well plates (Thermo Scientific) are coated overnight with 5 pg/ml of a capturing anti-Fab antibody (Cappel Labs) in 50 mM sodium carbonate (pH 9.6), and subsequently blocked with 2% (w/v) bovine serum albumin in PBS for two to five hours at room temperature (approximately 23Ό). In a non-adsorbent plate (Nunc #269620), 100 pM or 26 pM [¹²⁵l]-antigen are mixed with serial dilutions of a Fab of interest (e.g., consistent with assessment of the anti-VEGF antibody, Fab-12, in Presta et al., Cancer Res. 57:4593-4599 (1997)). The Fab of interest is then incubated overnight; however, the incubation may continue for a longer period (e.g., about 65 hours) to ensure that equilibrium is reached. Thereafter, the mixtures are transferred to the capture plate for incubation at room temperature (e.g., for one hour). The solution is then removed and the plate washed eight times with 0.1 % polysorbate 20 (TWEEN-20®) in PBS. When the plates have dried, 150 μΙ/well of scintillant (MICROSCINT-20™; Packard) is added, and the plates are counted on a TOPCOUNT™ gamma counter (Packard) for ten minutes. Concentrations of each Fab that give less than or equal to 20% of maximal binding are chosen for use in competitive binding assays.

According to another embodiment, Kd is measured using a BIACORE^® surface plasmon resonance assay. For example, an assay using a BIACORE^®-2000 or a BIACORE ^®-3000 (BIAcore, Inc., Piscataway, NJ) is performed at 25 Ό with immobilized antigen CM5 chips at -10 response units (RU). In one embodiment, carboxymethylated dextran biosensor chips (CM5, BIACORE, Inc.) are activated with A/-ethyl-/V- (3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC) and /V-hydroxysuccinimide (NHS) according to the supplier's instructions. Antigen is diluted with 10 mM sodium acetate, pH 4.8, to 5 pg/ml (-0.2 μΜ) before injection at a flow rate of 5 μΙ/minute to achieve approximately 10 response units (RU) of coupled protein. Following the injection of antigen, 1 M ethanolamine is injected to block unreacted groups. For kinetics measurements, two-fold serial dilutions of Fab (0.78 nM to 500 nM) are injected in PBS with 0.05% polysorbate 20 (TWEEN-20®) surfactant (PBST) at 25 Ό at a flow rate of approximately 25 μΙ/min. Association rates (k_on) and dissociation rates (k_off) are calculated using a simple one-to-one Langmuir binding model (BIACORE ^® Evaluation Software version 3.2) by simultaneously fitting the association and dissociation sensorgrams. The equilibrium dissociation constant (Kd) is calculated as the ratio k_off/k_on- See, for example, Chen et al., J. Mol. Biol. 293:865-881 (1999). If the on-rate exceeds 10⁶M^"V¹ by the surface plasmon resonance assay above, then the on-rate can be determined by using a fluorescent quenching technique that measures the increase or decrease in fluorescence emission intensity (excitation = 295 nm ; emission = 340 nm , 16-nm band-pass) at 25 Ό of a 20 nM anti-antigen antibody (Fab form) in PBS, pH 7.2, in the presence of increasing concentrations of antigen as measured in a spectrometer, such as a stop-flow equipped spectrophometer (Aviv Instruments) or a 8000-series SLM-AMINCO™ spectrophotometer (ThermoSpectronic) with a stirred cuvette.

In some embodiments, an anti-Ang2/anti-VEGF antibody of the invention binds VEGF with a Kd of about 10 nM or lower and Ang2 with a Kd of about 10 nM or lower. In other embodiments, an anti- Ang2/anti-VEGF antibody of the invention binds VEGF with a Kd of about 5 nM or lower and Ang2 with a Kd of about 5 nM or lower. In yet other embodiments, an anti-Ang2/anti-VEGF antibody of the invention binds VEGF with a Kd of about 1 nM or lower and Ang2 with a Kd of about 1 nM or lower. In some embodiments, an anti-Ang2/anti-VEGF antibody of the invention binds VEGF with a Kd of about 0.5 nM or lower and Ang2 with a Kd of about 0.5 nM or lower. In other embodiments, an anti-Ang2/anti-VEGF antibody of the invention binds VEGF with a Kd of about 0.25 nM or lower and Ang2 with a Kd of about 0.25 nM or lower. In still further embodiments, an anti-Ang2/anti-VEGF antibody of the invention binds VEGF with a Kd of about 0.1 nM or lower and Ang2 with a Kd of about 0.1 nM or lower. In some embodiments, an anti-Ang2/anti-VEGF antibody of the invention binds VEGF with a Kd of about 50 pM or lower and Ang2 with a Kd of about 50 pM or lower.

In some embodiments, an anti-Ang2 antibody or anti-Ang2/anti-VEGF antibody of the invention binds to Ang2 with 10-fold greater affinity than to Ang1 . In other embodiments, an anti-Ang2 antibody or anti-Ang2/anti-VEGF antibody of the invention binds to Ang2 with 25-fold greater affinity than to Ang1 . In some embodiments, an anti-Ang2 antibody or anti-Ang2/anti-VEGF antibody of the invention binds to Ang2 with 50-fold greater affinity than to Ang1 . In some embodiments, an anti-Ang2 antibody or anti- Ang2/anti-VEGF antibody of the invention binds to Ang2 with 100-fold greater affinity than to Ang1 . In some embodiments, an anti-Ang2 antibody or anti-Ang2/anti-VEGF antibody of the invention binds to Ang2 with 100-fold greater affinity than to Ang1 .

2. Antibody Fragments

In certain embodiments, an antibody provided herein is an antibody fragment. Antibody fragments include, but are not limited to, Fab, Fab', Fab'-SH, F(ab')₂, Fv, and scFv fragments, and other fragments described below. For a review of certain antibody fragments, see Hudson et al. Nat. Med. 9:129-134 (2003). For a review of scFv fragments, see, e.g., Pluckthun, in The Pharmacology of Monoclonal Antibodies, vol. 1 13, Rosenburg and Moore eds., (Springer-Verlag, New York), pp. 269-315 (1994) ; see also WO 93/16185; and U.S. Patent Nos. 5,571 ,894 and 5,587,458. For discussion of Fab and F(ab')₂ fragments comprising salvage receptor binding epitope residues and having increased in vivo half-life, see U.S. Patent No. 5,869,046.

Diabodies are antibody fragments with two antigen-binding sites that may be bivalent or bispecific. See, for example, EP 404,097; WO 1993/01 161 ; Hudson et al. Nat. Med. 9:129-134 (2003) ; and Hollinger et al. Proc. Natl. Acad. Sci. USA 90: 6444-6448 (1993). Triabodies and tetrabodies are also described in Hudson et al. Nat. Med. 9:129-134 (2003).

Single-domain antibodies are antibody fragments comprising all or a portion of the heavy chain variable domain or all or a portion of the light chain variable domain of an antibody. In certain embodiments, a single-domain antibody is a human single-domain antibody (Domantis, Inc., Waltham , MA; see, e.g., U.S. Patent No. 6,248,51 6 B1 ).

Antibody fragments can be made by various techniques, including but not limited to proteolytic digestion of an intact antibody as well as production by recombinant host cells (e.g., E. coli or phage), as described herein.

In some embodiments, an antibody fragment provided herein is a bispecific antibody fragment (e.g., a dual-specific antibody fragment). 3. Chimeric and Humanized Antibodies

In certain embodiments, an antibody provided herein is a chimeric antibody. Certain chimeric antibodies are described, e.g., in U.S. Patent No. 4,81 6,567; and Morrison et al. Proc. Natl. Acad. Sci. USA, 81 :6851 -6855 (1984)). In one example, a chimeric antibody comprises a non-human variable region (e.g., a variable region derived from a mouse, rat, hamster, rabbit, or non-human primate, such as a monkey) and a human constant region. In a further example, a chimeric antibody is a "class switched" antibody in which the class or subclass has been changed from that of the parent antibody. Chimeric antibodies include antigen-binding fragments thereof.

In certain embodiments, a chimeric antibody is a humanized antibody. Typically, a non-human antibody is humanized to reduce immunogenicity to humans, while retaining the specificity and affinity of the parental non-human antibody. Generally, a humanized antibody comprises one or more variable domains in which HVRs (or portions thereof) are derived from a non-human antibody, and FRs (or portions thereof) are derived from human antibody sequences. A humanized antibody optionally will also comprise at least a portion of a human constant region. In some embodiments, some FR residues in a humanized antibody are substituted with corresponding residues from a non-human antibody (e.g., the antibody from which the HVR residues are derived), e.g., to restore or improve antibody specificity or affinity.

Humanized antibodies and methods of making them are reviewed, e.g., in Almagro and

Fransson, Front. Biosci. 13:1619-1633 (2008), and are further described, e.g., in Riechmann et al., Nature 332:323-329 (1988) ; Queen et al., Proc. Nat'l Acad. Sci. USA 86:1 0029-10033 (1 989) ; US Patent Nos. 5, 821 ,337, 7,527,791 , 6,982,321 , and 7,087,409; Kashmiri et al., Methods 36:25-34 (2005) (describing specificity determining region (SDR) grafting) ; Padlan, Mol. Immunol. 28:489-498 (1991 ) (describing "resurfacing") ; Dall'Acqua et al., Methods 36:43-60 (2005) (describing "FR shuffling") ; and Osbourn et al., Methods 36:61 -68 (2005) and Klimka et al., Br. J. Cancer, 83:252-260 (2000) (describing the "guided selection" approach to FR shuffling).

Human framework regions that may be used for humanization include but are not limited to: framework regions selected using the "best-fit" method (see, e.g., Sims et al. J. Immunol. 151 :2296 (1993)) ; framework regions derived from the consensus sequence of human antibodies of a particular subgroup of light or heavy chain variable regions (see, e.g., Carter et al. Proc. Natl. Acad. Sci. USA, 89:4285 (1992) ; and Presta et al. J. Immunol., 1 51 :2623 (1993)) ; human mature (somatically mutated) framework regions or human germline framework regions (see, e.g., Almagro and Fransson, Front. Biosci. 13:1619-1633 (2008)) ; and framework regions derived from screening FR libraries (see, e.g., Baca et al., J. Biol. Chem. 272:10678-10684 (1997) and Rosok et al., J. Biol. Chem. 271 :2261 1 -2261 8 (1996)). 4. Human Antibodies

In certain embodiments, an antibody provided herein is a human antibody. Human antibodies can be produced using various techniques known in the art. Human antibodies are described generally in van Dijk and van de Winkel, Curr. Opin. Pharmacol. 5: 368-74 (2001 ) and Lonberg, Curr. Opin. Immunol. 20:450-459 (2008). Human antibodies may be generated by isolating Fv clone variable domain sequences selected from human-derived phage display libraries. Such variable domain sequences may then be combined with a desired human constant domain. Techniques for selecting human antibodies from antibody libraries are described below.

5. Library-Derived Antibodies

Antibodies of the invention may be isolated by screening combinatorial libraries for antibodies with the desired activity or activities. For example, a variety of methods are known in the art for generating phage display libraries and screening such libraries for antibodies possessing the desired binding characteristics. Such methods are reviewed, e.g., in Hoogenboom et al. in Methods in Molecular Biology 178:1 -37 (O'Brien et al., ed., Human Press, Totowa, NJ, 2001 ) and further described, e.g., in the McCafferty et al., Nature 348:552-554; Clackson et al., Nature 352: 624-628 (1991 ) ; Marks et al., J. Mol. Biol. 222: 581 -597 (1992) ; Marks and Bradbury, in Methods in Molecular Biology 248:161 -1 75 (Lo, ed., Human Press, Totowa, NJ, 2003) ; Sidhu et al., J. Mol. Biol. 338(2) : 299-310 (2004) ; Lee et al., J. Mol. Biol. 340(5) : 1073-1093 (2004) ; Fellouse, Proc. Natl. Acad. Sci. USA 101 (34) : 12467-12472 (2004) ; and Lee et al., J. Immunol. Methods 284(1 -2) : 1 19-132(2004).

In certain phage display methods, repertoires of V_H and V_L genes are separately cloned by polymerase chain reaction (PCR) and recombined randomly in phage libraries, which can then be screened for antigen-binding phage as described in Winter et al., Ann. Rev. Immunol., 12: 433-455 (1994). Phage typically display antibody fragments, either as single-chain Fv (scFv) fragments or as Fab fragments. Libraries from immunized sources provide high-affinity antibodies to the immunogen without the requirement of constructing hybridomas. Alternatively, the naive repertoire can be cloned (e.g., from human) to provide a single source of antibodies to a wide range of non-self and also self antigens without any immunization as described by Griffiths et al., EMBO J, 12: 725-734 (1993). Finally, naive libraries can also be made synthetically by cloning unrearranged V-gene segments from stem cells, and using

PCR primers containing random sequence to encode the highly variable CDR3 regions and to accomplish rearrangement in vitro, as described by Hoogenboom and Winter, J. Mol. Biol., 227: 381 -388 (1 992). Patent publications describing human antibody phage libraries include, for example: US Patent No. 5,750,373, and US Patent Publication Nos. 2005/0079574, 2005/01 19455, 2005/0266000,

2007/01 17126, 2007/0160598, 2007/0237764, 2007/0292936, and 2009/0002360.

Antibodies or antibody fragments isolated from human antibody libraries are considered human antibodies or human antibody fragments herein.

6. Multispecific Antibodies

In any one of the above aspects, the anti-Ang2 antibody provided herein may be a multispecific antibody, for example, a bispecific antibody. Multispecific antibodies are monoclonal antibodies that have binding specificities for at least two different sites. In certain embodiments, one of the binding specificities is for Ang2 and the other is for any other antigen, for example, VEGF. In certain embodiments, bispecific antibodies may bind to two different epitopes of Ang2. Bispecific antibodies can be prepared as full- length antibodies or antibody fragments. In some instances, the bispecific antibody is an anti-Ang2/anti- VEGF dual-specific antibody.

Techniques for making multispecific antibodies include, but are not limited to, recombinant co- expression of two immunoglobulin heavy chain-light chain pairs having different specificities (see Milstein et al. Nature. 305: 537, 1983, WO 93/08829, and Traunecker et al. EMBO J. 10: 3655, 1991 ), and "knob- in-hole" engineering (see, e.g., U.S. Patent No. 5,731 ,1 68). Multi-specific antibodies may also be made by engineering electrostatic steering effects for making antibody Fc-heterodimeric molecules (WO 2009/089004A1 ) ; cross-linking two or more antibodies or fragments (see, e.g., US Patent No. 4,676,980, and Brennan et al. Science. 229: 81 , 1985) ; using leucine zippers to produce bispecific antibodies (see, e.g., Kostelny et al. J. Immunol. 148(5) : 1547-1553, 1 992) ; using "diabody" technology for making bispecific antibody fragments (see, e.g., Hollinger et al. Proc. Natl. Acad. Sci. USA., 90: 6444-6448, 1993) ; and using single-chain Fv (sFv) dimers (see, e.g. Gruber et al. J. Immunol. 152: 5368, 1994) ; and preparing trispecific antibodies as described, e.g., in Tutt et al. J. Immunol. 147: 60, 1991 .

Engineered antibodies with three or more functional antigen binding sites, including Octopus antibodies," are also included herein (see, e.g. US 2006/0025576 A1 ).

The antibody or fragment herein also includes a "Dual action Fab" or "DAF" comprising an antigen binding site that binds to Ang2 as well as another, different antigen, e.g., VEGF (see, e.g., US 2008/0069820). 7. Antibody Variants

In certain embodiments, amino acid sequence variants of the anti-Ang2 antibodies of the invention (e.g., bispecific anti-Ang2 antibodies of the invention that bind to Ang2 and a second biological molecule, e.g., VEGF, such as dual-specific anti-Ang2/anti-VEGF antibodies of the invention) are contemplated. For example, it may be desirable to improve the binding affinity and/or other biological properties of the antibody. Amino acid sequence variants of an antibody may be prepared by introducing appropriate modifications into the nucleotide sequence encoding the antibody, or by peptide synthesis. Such modifications include, for example, deletions from , and/or insertions into and/or substitutions of residues within the amino acid sequences of the antibody. Any combination of deletion, insertion, and substitution can be made to arrive at the final construct, provided that the final construct possesses the desired characteristics, for example, antigen-binding. a. Substitution, Insertion, and Deletion Variants

In certain embodiments, antibody variants having one or more amino acid substitutions are provided. Sites of interest for substitutional mutagenesis include the HVRs and FRs. Conservative substitutions are shown in Table 1 under the heading of "preferred substitutions." More substantial changes are provided in Table 1 under the heading of "exemplary substitutions," and as further described below in reference to amino acid side chain classes. Amino acid substitutions may be introduced into an antibody of interest and the products screened for a desired activity, for example, retained/improved antigen binding, decreased immunogenicity, or reduced ADCC or CDC. Table 1. Exemplary and Preferred Amino Acid Substitutions

Amino acids may be grouped according to common side-chain properties:

(1 ) hydrophobic: Norleucine, Met, Ala, Val, Leu, lie;

(2) neutral hydrophilic: Cys, Ser, Thr, Asn, Gin ;

(3) acidic: Asp, Glu ;

(4) basic: H is, Lys, Arg ;

(5) residues that influence chain orientation : Gly, Pro;

(6) aromatic: Trp, Tyr, Phe.

Non-conservative substitutions will entail exchanging a member of one of these classes for another class.

One type of substitutional variant involves substituting one or more hypervariable region residues of a parent antibody (e.g. a humanized or human antibody) . Generally, the resulting variant(s) selected for further study will have modifications (e.g., improvements) in certain biological properties (e.g. , increased affinity, reduced immunogenicity) relative to the parent antibody and/or will have substantially retained certain biological properties of the parent antibody. An exemplary substitutional variant is an affinity-matured antibody, which may be conveniently generated, e.g., using phage display-based affinity maturation techniques such as those described herein. Briefly, one or more HVR residues are mutated and the variant antibodies displayed on phage and screened for a particular biological activity (e.g.

binding affinity).

Alterations (e.g., substitutions) may be made in HVRs, e.g., to improve antibody affinity. Such alterations may be made in HVR "hotspots," i.e., residues encoded by codons that undergo mutation at high frequency during the somatic maturation process (see, e.g., Chowdhury, Methods Mol. Biol.

207:179-196 (2008)), and/or residues that contact antigen, with the resulting variant V_H or V_L being tested for binding affinity. Affinity maturation by constructing and reselecting from secondary libraries has been described, e.g., in Hoogenboom et al. in Methods in Molecular Biology 178:1 -37 (O'Brien et al., ed., Human Press, Totowa, NJ, (2001 ).) In some embodiments of affinity maturation, diversity is introduced into the variable genes chosen for maturation by any of a variety of methods (e.g., error-prone PCR, chain shuffling, or oligonucleotide-directed mutagenesis). A secondary library is then created. The library is then screened to identify any antibody variants with the desired affinity. Another method to introduce diversity involves HVR-directed approaches, in which several HVR residues (e.g., 4-6 residues at a time) are randomized. HVR residues involved in antigen binding may be specifically identified, e.g., using alanine scanning mutagenesis or modeling. HVR-H3 and HVR-L3 in particular are often targeted.

In certain embodiments, substitutions, insertions, or deletions may occur within one or more

HVRs so long as such alterations do not substantially reduce the ability of the antibody to bind antigen. For example, conservative alterations (e.g., conservative substitutions as provided herein) that do not substantially reduce binding affinity may be made in HVRs. Such alterations may, for example, be outside of antigen contacting residues in the HVRs. In certain embodiments of the variant V_H and V_L sequences provided above, each HVR either is unaltered, or contains no more than one, two or three amino acid substitutions.

A useful method for identification of residues or regions of an antibody that may be targeted for mutagenesis is called "alanine scanning mutagenesis" as described by Cunningham and Wells (1989) Science, 244:1081 -1085. In this method, a residue or group of target residues (e.g., charged residues such as Arg, Asp, His, Lys, and Glu) are identified and replaced by a neutral or negatively charged amino acid (e.g., alanine or polyalanine) to determine whether the interaction of the antibody with antigen is affected. Further substitutions may be introduced at the amino acid locations demonstrating functional sensitivity to the initial substitutions. Alternatively, or additionally, a crystal structure of an antigen- antibody complex to identify contact points between the antibody and antigen. Such contact residues and neighboring residues may be targeted or eliminated as candidates for substitution. Variants may be screened to determine whether they contain the desired properties.

Amino acid sequence insertions include amino- and/or carboxyl-terminal fusions ranging in length from one residue to polypeptides containing a hundred or more residues, as well as intrasequence insertions of single or multiple amino acid residues. Examples of terminal insertions include an antibody with an N-terminal methionyl residue. Other insertional variants of the antibody molecule include the fusion to the N- or C-terminus of the antibody to an enzyme (e.g. for ADEPT) or a polypeptide which increases the serum half-life of the antibody. b. Glycosylation variants

In certain embodiments, anti-Ang2 antibodies of the invention (e.g., bispecific anti-Ang2 antibodies of the invention that bind to Ang2 and a second biological molecule, e.g., VEGF, such as dual- specific anti-Ang2/anti-VEGF antibodies of the invention) can be altered to increase or decrease the extent to which the antibody is glycosylated. Addition or deletion of glycosylation sites to an anti-Ang2 antibody of the invention may be conveniently accomplished by altering the amino acid sequence such that one or more glycosylation sites is created or removed.

Where the antibody comprises an Fc region, the carbohydrate attached thereto may be altered. Native antibodies produced by mammalian cells typically comprise a branched, biantennary

oligosaccharide that is generally attached by an N-linkage to Asn297 of the CH2 domain of the Fc region. See, e.g., Wright et al. TIBTECH 15:26-32 (1997). The oligosaccharide may include various

carbohydrates, e.g., mannose, N-acetyl glucosamine (GlcNAc), galactose, and sialic acid, as well as a fucose attached to a GlcNAc in the "stem" of the biantennary oligosaccharide structure. In some embodiments, modifications of the oligosaccharide in an antibody of the invention may be made in order to create antibody variants with certain improved properties. c. Fc region variants

In certain embodiments, one or more amino acid modifications may be introduced into the Fc region of an anti-Ang2 antibody of the invention (e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as an anti-Ang2/anti-VEGF antibody of the invention), thereby generating an Fc region variant. The Fc region variant may comprise a human Fc region sequence (e.g., a human lgG1 , lgG2, lgG3 or lgG4 Fc region) comprising an amino acid modification (e.g., a substitution) at one or more amino acid positions.

In certain embodiments, the invention contemplates an anti-Ang2 antibody variant that possesses some but not all effector functions, which make it a desirable candidate for applications in which the half life of the antibody in vivo is important yet certain effector functions (such as complement and ADCC) are unnecessary or deleterious. In vitro and/or in vivo cytotoxicity assays can be conducted to confirm the reduction/depletion of CDC and/or ADCC activities. For example, Fc receptor (FcR) binding assays can be conducted to ensure that the antibody lacks FcyR binding (hence likely lacking ADCC activity), but retains FcRn binding ability. The primary cells for mediating ADCC, N K cells, express FcyRII I only, whereas monocytes express FcyRI, FcyRII and FcyRI I I. FcR expression on hematopoietic cells is summarized in Table 3 on page 464 of Ravetch and Kinet, Annu. Rev. Immunol. 9:457-492 (1991 ). Non- limiting examples of in vitro assays to assess ADCC activity of a molecule of interest is described in U .S. Patent No. 5,500,362 (see, e.g. Hellstrom , I. et al. Proc. Nat'l Acad. Sci. USA 83:7059-7063 (1 986)) and Hellstrom, I et al., Proc. Nat'l Acad. Sci. USA 82:1499-1502 (1985) ; 5,821 ,337 (see Bruggemann, M. et al., J. Exp. Med. 166:1351 -1361 (1987)). Alternatively, non-radioactive assays methods may be employed (see, for example, ACTI™ non-radioactive cytotoxicity assay for flow cytometry (CellTechnology, Inc. Mountain View, CA; and CYTOTOX 96 non-radioactive cytotoxicity assay (Promega, Madison, Wl). Useful effector cells for such assays include peripheral blood mononuclear cells (PBMC) and Natural Killer (NK) cells. Alternatively, or additionally, ADCC activity of the molecule of interest may be assessed in vivo, e.g., in a animal model such as that disclosed in Clynes et al. Proc. Nat'l Acad. Sci. USA 95:652-656 (1 998). C1 q binding assays may also be carried out to confirm that the antibody is unable to bind C1 q and hence lacks CDC activity. See, e.g., C1 q and C3c binding ELISA in WO 2006/029879 and WO 2005/100402. To assess complement activation, a CDC assay may be performed (see, for example, Gazzano-Santoro et al. J. Immunol. Methods 202:163 (1996) ; Cragg, M.S. et al. Blood. 101 :1045-1052 (2003) ; and Cragg, M.S. and M.J. Glennie Blood. 103:2738-2743 (2004)). FcRn binding and in vivo clearance/half life determinations can also be performed using methods known in the art (see, e.g., Petkova, S.B. et al. Int'l. Immunol. 18(12) :1759- 769 (2006)).

Antibodies with reduced effector function include those with substitution of one or more of Fc region residues 238, 265, 269, 270, 297, 327 and 329 (U.S. Patent Nos. 6,737,056 and 8,219,149). Such Fc mutants include Fc mutants with substitutions at two or more of amino acid positions 265, 269, 270, 297 and 327, including the so-called "DANA" Fc mutant with substitution of residues 265 and 297 to alanine (US Patent No. 7,332,581 and 8,219,149).

Certain antibody variants with improved or diminished binding to FcRs are described. (See, e.g., U.S. Patent No. 6,737,056; WO 2004/056312, and Shields et al., J. Biol. Chem. 9(2) : 6591 -6604 (2001 ).) In certain embodiments, an antibody variant comprises an Fc region with one or more amino acid substitutions which improve ADCC, e.g., substitutions at positions 298, 333, and/or 334 of the Fc region (EU numbering of residues).

In some embodiments, alterations are made in the Fc region that result in altered (i.e., either improved or diminished) C1 q binding and/or Complement Dependent Cytotoxicity (CDC), e.g., as described in US Patent No. 6,194,551 , WO 99/51642, and Idusogie et al. J. Immunol. 164: 4178-4184 (2000).

Antibodies with increased half lives and improved binding to the neonatal Fc receptor (FcRn), which is responsible for the transfer of maternal IgGs to the fetus (Guyer et al., J. Immunol. 1 17:587 (1976) and Kim et al., J. Immunol. 24:249 (1994)), are described in US2005/0014934A1 (Hinton et al.). Those antibodies comprise an Fc region with one or more substitutions therein which improve binding of the Fc region to FcRn. Such Fc variants include those with substitutions at one or more of Fc region residues: 238, 256, 265, 272, 286, 303, 305, 307, 31 1 , 312, 317, 340, 356, 360, 362, 376, 378, 380, 382, 413, 424 or 434, e.g., substitution of Fc region residue 434 (US Patent No. 7,371 ,826).

See also Duncan & Winter, Nature 322:738-40 (1 988) ; U.S. Patent No. 5,648,260; U.S. Patent No. 5,624,821 ; and WO 94/29351 concerning other examples of Fc region variants. d. Cysteine engineered antibody variants

In certain embodiments, it may be desirable to create cysteine engineered antibodies, e.g., "thioMAbs," in which one or more residues of an antibody are substituted with cysteine residues. In particular embodiments, the substituted residues occur at accessible sites of the antibody. By substituting those residues with cysteine, reactive thiol groups are thereby positioned at accessible sites of the antibody and may be used to conjugate the antibody to other moieties, such as drug moieties or linker-drug moieties, to create an immunoconjugate, as described further herein. In certain embodiments, any one or more of the following residues may be substituted with cysteine: V205 (Kabat numbering) of the light chain; A1 18 (EU numbering) of the heavy chain; and S400 (EU numbering) of the heavy chain Fc region. Cysteine engineered antibodies may be generated as described, for example, in U.S. Patent No. 7,521 ,541 . e. Antibody derivatives

In certain embodiments, an anti-Ang2 antibody of the invention (e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as a dual-specific anti-Ang2/anti-VEGF antibody of the invention) provided herein may be further modified to contain additional nonproteinaceous moieties that are known in the art and readily available. The moieties suitable for derivatization of the antibody include but are not limited to water soluble polymers. Non-limiting examples of water soluble polymers include, but are not limited to, polyethylene glycol (PEG), copolymers of ethylene glycol/propylene glycol, carboxymethylcellulose, dextran, polyvinyl alcohol, polyvinyl pyrrolidone, poly-1 , 3-dioxolane, poly-1 ,3,6-trioxane, ethylene/maleic anhydride copolymer, polyaminoacids (either homopolymers or random copolymers), and dextran or poly(n-vinyl pyrrolidone)polyethylene glycol, propropylene glycol homopolymers, prolypropylene oxide/ethylene oxide co-polymers, polyoxyethylated polyols (e.g., glycerol), polyvinyl alcohol, and mixtures thereof.

Polyethylene glycol propionaldehyde may have advantages in manufacturing due to its stability in water. The polymer may be of any molecular weight, and may be branched or unbranched. The number of polymers attached to the antibody may vary, and if more than one polymer are attached, they can be the same or different molecules. In general, the number and/or type of polymers used for derivatization can be determined based on considerations including, but not limited to, the particular properties or functions of the antibody to be improved, whether the antibody derivative will be used in a therapy under defined conditions, etc.

In another embodiment, conjugates of an antibody and nonproteinaceous moiety that may be selectively heated by exposure to radiation are provided. In one embodiment, the nonproteinaceous moiety is a carbon nanotube (Kam et al., Proc. Natl. Acad. Sci. USA 102: 1 1600-1 1605 (2005)). The radiation may be of any wavelength, and includes, but is not limited to, wavelengths that do not harm ordinary cells, but which heat the nonproteinaceous moiety to a temperature at which cells proximal to the antibody-nonproteinaceous moiety are killed.

B. Recombinant Methods and Compositions

Anti-Ang2 antibodies of the invention (e.g., bispecific anti-Ang2 antibodies of the invention that bind to Ang2 and a second biological molecule, e.g., VEGF, such as anti-Ang2/anti-VEGF antibodies of the invention) may be produced using recombinant methods and compositions, for example, as described in U.S. Patent No. 4,81 6,567. In one embodiment, isolated nucleic acid encoding an anti-Ang2 or anti- Ang2/anti-VEGF antibody described herein is provided. Such nucleic acid may encode an amino acid sequence comprising the V_L and/or an amino acid sequence comprising the V_H of the antibody (e.g., the light and/or heavy chains of the antibody). In a further embodiment, one or more vectors (e.g., expression vectors) comprising such nucleic acid are provided. In a further embodiment, a host cell comprising such nucleic acid is provided. In one such embodiment, a host cell comprises (e.g., has been transformed with) : (1 ) a vector comprising a nucleic acid that encodes an amino acid sequence comprising the V_L of the antibody and an amino acid sequence comprising the V_H of the antibody, or (2) a first vector comprising a nucleic acid that encodes an amino acid sequence comprising the V_L of the antibody and a second vector comprising a nucleic acid that encodes an amino acid sequence comprising the V_H of the antibody. In one embodiment, the host cell is eukaryotic, e.g. a Chinese Hamster Ovary (CHO) cell or lymphoid cell (e.g., YO, NSO, Sp20 cell). In one embodiment, a method of making an anti- Ang2 antibody (e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as an anti-Ang2/Anti-VEGF antibody of the invention), wherein the method comprises culturing a host cell comprising a nucleic acid encoding the antibody, as provided above, under conditions suitable for expression of the antibody, and optionally recovering the antibody from the host cell (or host cell culture medium).

For recombinant production of an anti-Ang2 antibody (e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as an anti-Ang2/anti- VEGF antibody of the invention), nucleic acid encoding an antibody, e.g., as described above, is isolated and inserted into one or more vectors for further cloning and/or expression in a host cell. Such nucleic acid may be readily isolated and sequenced using conventional procedures (e.g., by using

oligonucleotide probes that are capable of binding specifically to genes encoding the heavy and light chains of the antibody).

Suitable host cells for cloning or expression of antibody-encoding vectors include prokaryotic or eukaryotic cells described herein. For example, antibodies may be produced in bacteria, in particular when glycosylation and Fc effector function are not needed. For expression of antibody fragments and polypeptides in bacteria, see, e.g., U.S. Patent Nos. 5,648,237, 5,789,199, and 5,840,523. (See also

Charlton, Methods in Molecular Biology, Vol. 248 (B.K.C. Lo, ed., Humana Press, Totowa, NJ, 2003), pp. 245-254, describing expression of antibody fragments in E. coli.) After expression, the antibody may be isolated from the bacterial cell paste in a soluble fraction and can be further purified.

In addition to prokaryotes, eukaryotic microbes such as filamentous fungi or yeast are suitable cloning or expression hosts for antibody-encoding vectors, including fungi and yeast strains whose glycosylation pathways have been "humanized," resulting in the production of an antibody with a partially or fully human glycosylation pattern. See Gerngross, Nat. Biotech. 22:1409-1414 (2004), and Li et al., Nat. Biotech. 24:210-215 (2006).

Suitable host cells for the expression of glycosylated antibody are also derived from multicellular organisms (invertebrates and vertebrates). Examples of invertebrate cells include plant and insect cells. Numerous baculoviral strains have been identified which may be used in conjunction with insect cells, particularly for transfection of Spodoptera frugiperda cells.

Plant cell cultures can also be utilized as hosts. See, e.g., US Patent Nos. 5,959,1 77, 6,040,498, 6,420,548, 7,125,978, and 6,41 7,429 (describing PLANTIBODIES™ technology for producing antibodies in transgenic plants). Vertebrate cells may also be used as hosts. For example, mammalian cell lines that are adapted to grow in suspension may be useful. Other examples of useful mammalian host cell lines are monkey kidney CV1 line transformed by SV40 (COS-7) ; human embryonic kidney line (293 or 293 cells as described, e.g., in Graham et al., J. Gen Virol. 36:59 (1977)) ; baby hamster kidney cells (BHK) ; mouse Sertoli cells (TM4 cells as described, e.g., in Mather, Biol. Reprod. 23:243-251 (1980)) ; monkey kidney cells (CV1 ) ; African green monkey kidney cells (VERO-76) ; human cervical carcinoma cells (HELA) ; canine kidney cells (MDCK; buffalo rat liver cells (BRL 3A) ; human lung cells (W138) ; human liver cells (Hep G2) ; mouse mammary tumor (MMT 060562) ; TRI cells, as described, e.g., in Mather et al., Annals N. Y. Acad. Sci. 383:44-68 (1982) ; MRC 5 cells; and FS4 cells. Other useful mammalian host cell lines include Chinese hamster ovary (CHO) cells, including DHFR^" CHO cells (Urlaub et al., Proc. Natl. Acad. Sci. USA 77:4216 (1980)) ; and myeloma cell lines such as Y0, NS0 and Sp2/0. For a review of certain mammalian host cell lines suitable for antibody production, see, e.g., Yazaki and Wu, Methods in Molecular Biology, Vol. 248 (B.K.C. Lo, ed., Humana Press, Totowa, NJ), pp. 255-268 (2003). C. Assays

Anti-Ang2 antibodies of the invention (e.g., bispecific anti-Ang2 antibodies of the invention that bind to Ang2 and a second biological molecule, e.g., VEGF, such as anti-Ang2/anti-VEGF antibodies of the invention) provided herein may be identified, screened for, or characterized for their physical/chemical properties and/or biological activities by various assays known in the art.

7. Binding assays and other assays

In one aspect, an anti-Ang2 antibody of the invention (e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as an anti-Ang2/anti- VEGF antibody of the invention) is tested for its antigen binding activity, for example, by known methods such as ELISA, Western blot, etc.

In another aspect, competition assays may be used to identify an antibody that competes with an anti-Ang2 antibody of the invention for binding to Ang2. In certain embodiments, such a competing antibody binds to the same epitope (e.g., a linear or a conformational epitope) that is bound by an anti- Ang2 antibody of the invention. Detailed exemplary methods for mapping an epitope to which an antibody binds are provided in Morris (1 996) "Epitope Mapping Protocols," in Methods in Molecular Biology vol. 66 (Humana Press, Totowa, NJ).

In an exemplary competition assay, immobilized Ang2 is incubated in a solution comprising a first labeled antibody that binds to Ang2 and a second unlabeled antibody that is being tested for its ability to compete with the first antibody for binding to Ang2. The second antibody may be present in a hybridoma supernatant. As a control, immobilized Ang2 is incubated in a solution comprising the first labeled antibody but not the second unlabeled antibody. After incubation under conditions permissive for binding of the first antibody to Ang2, excess unbound antibody is removed, and the amount of label associated with immobilized Ang2 is measured. If the amount of label associated with immobilized Ang2 is substantially reduced in the test sample relative to the control sample, then that indicates that the second antibody is competing with the first antibody for binding to Ang2. See, e.g., Harlow and Lane (1988) Antibodies: A Laboratory Manual. Ch.14 (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY). A competition assay as described above can be also be used to test whether an antibody competes for binding to VEGF with an antibody of the invention, by substituting immobilized Ang2 in the assay for immobilized VEGF.

2. Activity assays

In one aspect, assays are provided for identifying anti-Ang2 antibodies of the invention (e.g., bispecific anti-Ang2 antibodies of the invention that bind to Ang2 and a second biological molecule, e.g., VEGF, such as anti-Ang2/anti-VEGF antibodies of the invention) or fragments thereof having biological activity. Biological activity may include, for example, binding to Ang2 (e.g., Ang2 in the blood stream), or a peptide fragment thereof, either in vivo, in vitro, or ex vivo. In other embodiments, biological activity may include blocking or neutralizing Ang2, or preventing Ang2 from binding to a ligand, for example, a receptor, such as Tie2. In some aspects, biological activity may include the ability to bind to Ang2 and VEGF (e.g., Ang2 and VEGF in the bloodstream), either in vivo, in vitro (e.g., Ang2 and VEGF displayed on phage or purified Ang2 or VEGF), or ex vivo. In some embodiments, binding to Ang2 and VEGF is not simultaneous, e.g., an antibody able to bind to both Ang2 and VEGF typically only binds to one molecule of Ang2 or VEGF at a time. In certain embodiments, biological activity may include blocking or neutralizing Ang2 and VEGF, or preventing Ang2 and VEGF from binding to a ligand, for example, a receptor such as Tie2, KDR, or Flt-1 . Antibodies having such biological activity in vivo and/or in vitro are provided. In certain embodiments, an antibody of the invention is tested for such biological activity, as described in detail in the Examples herein below.

D. Immunoconjugates

The invention also provides immunoconjugates comprising an anti-Ang2 antibody of the invention (e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as an anti-Ang2/anti-VEGF antibody of the invention) conjugated to one or more cytotoxic agents, such as chemotherapeutic agents or drugs, growth inhibitory agents, toxins (e.g., protein toxins, enzymatically active toxins of bacterial, fungal, plant, or animal origin, or fragments thereof), or radioactive isotopes.

In one embodiment, an immunoconjugate is an antibody-drug conjugate (ADC) in which an antibody is conjugated to one or more drugs, including but not limited to a maytansinoid (see U.S. Patent Nos. 5,208,020, 5,416,064 and European Patent EP 0 425 235 B1 ) ; an auristatin such as

monomethylauristatin drug moieties DE and DF (MMAE and MMAF) (see U.S. Patent Nos. 5,635,483 and 5,780,588, and 7,498,298) ; a dolastatin ; a calicheamicin or derivative thereof (see U.S. Patent Nos. 5,712,374, 5,714,586, 5,739,1 1 6, 5,767,285, 5,770,701 , 5,770,710, 5,773,001 , and 5,877,296; Hinman et al., Cancer Res. 53:3336-3342 (1 993) ; and Lode et al., Cancer Res. 58:2925-2928 (1998)) ; an anthracycline such as daunomycin or doxorubicin (see Kratz et al., Current Med. Chem. 13:477-523 (2006) ; Jeffrey et al., Bioorganic & Med. Chem. Letters 16:358-362 (2006) ; Torgov et al., Bioconj. Chem. 16:717-721 (2005) ; Nagy et al., Proc. Natl. Acad. Sci. USA 97:829-834 (2000) ; Dubowchik et al., Bioorg. & Med. Chem. Letters 12:1 529-1532 (2002) ; King et al., J. Med. Chem. 45:4336-4343 (2002) ; and U.S. Patent No. 6,630,579) ; methotrexate; vindesine; a taxane such as docetaxel, paclitaxel, larotaxel, tesetaxel, and ortataxel; a trichothecene; and CC1065.

In another embodiment, an immunoconjugate comprises an anti-Ang2 antibody as described herein (e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as an anti-Ang2/anti-VEGF antibody of the invention) conjugated to an enzymatically active toxin or fragment thereof, including but not limited to diphtheria A chain, nonbinding active fragments of diphtheria toxin, exotoxin A chain (from Pseudomonas aeruginosa), ricin A chain, abrin A chain, modeccin A chain, alpha-sarcin, Aleurites fordii proteins, dianthin proteins, Phytolaca americana proteins (PAPI, PAPII, and PAP-S), momordica charantia inhibitor, curcin, crotin, sapaonaria officinalis inhibitor, gelonin, mitogellin, restrictocin, phenomycin, enomycin, and the tricothecenes.

In another embodiment, an immunoconjugate comprises an anti-Ang2 antibody as described herein (e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as an anti-Ang2/anti-VEGF antibody of the invention) conjugated to a radioactive atom to form a radioconjugate. A variety of radioactive isotopes are available for the production of radioconjugates. Examples include At^{21 1} , I¹³¹ , I¹²⁵, Y⁹⁰, Re¹⁸⁶, Re¹⁸⁸, Sm¹⁵³, Bi²¹², P³², Pb²¹² and radioactive isotopes of Lu. When the radioconjugate is used for detection, it may comprise a radioactive atom for scintigraphic studies, for example tc99m or 1123, or a spin label for nuclear magnetic resonance (NMR) imaging (also known as magnetic resonance imaging, mri), such as iodine-123 again, iodine-131 , indium-1 1 1 , fluorine-19, carbon-13, nitrogen-15, oxygen-1 7, gadolinium, manganese or iron.

Conjugates of an antibody and cytotoxic agent may be made using a variety of bifunctional protein coupling agents such as N-succinimidyl-3-(2-pyridyldithio) propionate (SPDP), succinimidyl-4-(N- maleimidomethyl) cyclohexane-1 -carboxylate (SMCC), iminothiolane (IT), bifunctional derivatives of imidoesters (such as dimethyl adipimidate HCI), active esters (such as disuccinimidyl suberate), aldehydes (such as glutaraldehyde), bis-azido compounds (such as bis (p-azidobenzoyl) hexanediamine), bis-diazonium derivatives (such as bis-(p-diazoniumbenzoyl)-ethylenediamine), diisocyanates (such as toluene 2,6-diisocyanate), and bis-active fluorine compounds (such as 1 ,5-difluoro-2,4-dinitrobenzene) . For example, a ricin immunotoxin can be prepared as described in Vitetta et al., Science 238:1098 (1987). Carbon-14-labeled 1 -isothiocyanatobenzyl-3-methyldiethylene triaminepentaacetic acid (MX- DTPA) is an exemplary chelating agent for conjugation of radionucleotide to the antibody. See

W094/1 1026. The linker may be a "cleavable linker" facilitating release of a cytotoxic drug in the cell. For example, an acid-labile linker, peptidase-sensitive linker, photolabile linker, dimethyl linker or disulfide-containing linker (Chari et al., Cancer Res. 52:127-131 (1992) ; U.S. Patent No. 5,208,020) may be used.

The immunuoconjugates or ADCs herein expressly contemplate, but are not limited to such conjugates prepared with cross-linker reagents including, but not limited to, BMPS, EMCS, GMBS, HBVS, LC-SMCC, MBS, MPBH, SBAP, SIA, SIAB, SMCC, SMPB, SM PH, sulfo-EMCS, sulfo-GMBS, sulfo- KMUS, sulfo-MBS, sulfo-SIAB, sulfo-SMCC, and sulfo-SMPB, and SVSB (succinimidyl-(4- vinylsulfone)benzoate) which are commercially available (e.g., from Pierce Biotechnology, Inc., Rockford, IL, U.S.A). E. Methods and Compositions for Diagnostics and Detection

In certain embodiments, any of the anti-Ang2 antibodies of the invention (e.g., bispecific anti- Ang2 antibodies of the invention that bind to Ang2 and a second biological molecule, e.g., VEGF, such as dual-specific anti-Ang2/anti-VEGF antibodies of the invention) is useful for detecting the presence of Ang2 in a biological sample. In some embodiments, a bispecific anti-Ang2 antibody that binds Ang2 and VEGF is useful for detecting Ang2 and for detecting VEGF. The term "detecting" as used herein encompasses quantitative or qualitative detection. In certain embodiments, a biological sample comprises a cell or tissue.

In one instance, an anti-Ang2 antibody for use in a method of diagnosis or detection is provided. In a further aspect, a method of detecting the presence of Ang2 in a biological sample is provided. In certain embodiments, the method comprises contacting the biological sample with an anti-Ang2 antibody as described herein under conditions permissive for binding of the anti-Ang2 antibody to Ang2, and detecting whether a complex is formed between the anti-Ang2 antibody and Ang2. Such method may be an in vitro or in vivo method. In another embodiment, an anti-Ang2/anti-VEGF antibody for use in a method of diagnosis or detection is provided. In a further aspect, a method of detecting the presence of Ang2 and/or VEGF in a biological sample is provided. In certain embodiments, the method comprises contacting the biological sample with an anti-Ang2/anti-VEGF antibody as described herein under conditions permissive for binding of the anti-Ang2/anti-VEGF antibody to Ang2 and/or VEGF, and detecting whether a complex is formed between the anti-Ang2/anti-VEGF antibody and Ang2 and/or VEGF. Such method may be an in vitro or in vivo method.

In certain embodiments, labeled anti-Ang2 antibodies of the invention (e.g., bispecific anti-Ang2 antibodies of the invention that bind to Ang2 and a second biological molecule, e.g., VEGF, such as anti- Ang2/anti-VEGF antibodies of the invention) are provided. Labels include, but are not limited to, labels or moieties that are detected directly (such as fluorescent, chromophoric, electron-dense,

chemiluminescent, and radioactive labels), as well as moieties, such as enzymes or ligands, that are detected indirectly, e.g., through an enzymatic reaction or molecular interaction. Exemplary labels include, but are not limited to, the radioisotopes ³²P, ¹⁴C, ¹²⁵l, ³H, and ¹³¹1, fluorophores such as rare earth chelates or fluorescein and its derivatives, rhodamine and its derivatives, dansyl, umbelliferone, luceriferases, e.g., firefly luciferase and bacterial luciferase (U.S. Patent No. 4,737,456), luciferin, 2,3- dihydrophthalazinediones, horseradish peroxidase (H RP), alkaline phosphatase, β-galactosidase, glucoamylase, lysozyme, saccharide oxidases, e.g., glucose oxidase, galactose oxidase, and glucose-6- phosphate dehydrogenase, heterocyclic oxidases such as uricase and xanthine oxidase, coupled with an enzyme that employs hydrogen peroxide to oxidize a dye precursor such as HRP, lactoperoxidase, or microperoxidase, biotin/avidin, spin labels, bacteriophage labels, stable free radicals, and the like.

F. Pharmaceutical Compositions

Pharmaceutical compositions of an anti-Ang2 antibody of the invention (e.g., a bispecific anti- Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as a dual-specific anti-Ang2/anti-VEGF antibody of the invention) are prepared by mixing such an antibody having the desired degree of purity with one or more optional pharmaceutically acceptable carriers (Remington's Pharmaceutical Sciences 16th edition, Osol, A. Ed. (1980)), in the form of lyophilized compositions or aqueous solutions. Pharmaceutically acceptable carriers are generally nontoxic to recipients at the dosages and concentrations employed, and include, but are not limited to: buffers such as phosphate, citrate, and other organic acids; antioxidants including ascorbic acid and methionine; preservatives (such as octadecyldimethylbenzyl ammonium chloride; hexamethonium chloride;

benzalkonium chloride; benzethonium chloride; phenol, butyl or benzyl alcohol; alkyl parabens such as methyl or propyl paraben; catechol; resorcinol; cyclohexanol; 3-pentanol ; and m-cresol) ; low molecular weight (less than about 10 residues) polypeptides; proteins, such as serum albumin, gelatin, or immunoglobulins; hydrophilic polymers such as polyvinylpyrrolidone; amino acids such as glycine, glutamine, asparagine, histidine, arginine, or lysine; monosaccharides, disaccharides, and other carbohydrates including glucose, mannose, or dextrins; chelating agents such as EDTA; sugars such as sucrose, mannitol, trehalose or sorbitol; salt-forming counter-ions such as sodium ; metal complexes (e.g. Zn-protein complexes) ; and/or non-ionic surfactants such as polyethylene glycol (PEG). Exemplary pharmaceutically acceptable carriers herein further include insterstitial drug dispersion agents such as soluble neutral-active hyaluronidase glycoproteins (sHASEGP), for example, human soluble PH-20 hyaluronidase glycoproteins, such as rHuPH20 (HYLEN EX^®, Baxter International, Inc.). Certain exemplary sHASEGPs and methods of use, including rHuPH20, are described in US Patent Publication Nos. 2005/0260186 and 2006/0104968. In one aspect, a sHASEGP is combined with one or more additional glycosaminoglycanases such as chondroitinases.

Exemplary lyophilized antibody compositions are described in US Patent No. 6,267,958.

Aqueous antibody compositions include those described in US Patent No. 6,171 ,586 and

WO2006/044908, the latter compositions including a histidine-acetate buffer.

The composition herein may also contain more than one active ingredients as necessary for the particular indication being treated, preferably those with complementary activities that do not adversely affect each other. For example, it may be desirable to further provide an additional therapeutic agent (e.g., a chemotherapeutic agent, a cytotoxic agent, a growth inhibitory agent, and/or an anti-hormonal agent, such as those recited herein above). Such active ingredients are suitably present in combination in amounts that are effective for the purpose intended.

Active ingredients may be entrapped in microcapsules prepared, for example, by coacervation techniques or by interfacial polymerization, for example, hydroxymethylcellulose or gelatin-microcapsules and poly-(methylmethacylate) microcapsules, respectively, in colloidal drug delivery systems (for example, liposomes, albumin microspheres, microemulsions, nano-particles and nanocapsules) or in macroemulsions. Such techniques are disclosed in Remington's Pharmaceutical Sciences 16th edition, Osol, A. Ed. (1980).

Sustained-release preparations may be prepared. Suitable examples of sustained-release preparations include semipermeable matrices of solid hydrophobic polymers containing the antibody, which matrices are in the form of shaped articles, for example, films, or microcapsules.

The compositions to be used for in vivo administration are generally sterile. Sterility may be readily accomplished, e.g., by filtration through sterile filtration membranes. G. Therapeutic Methods and Compositions

Any of the anti-Ang2 antibodies of the invention (e.g., bispecific anti-Ang2 antibodies of the invention that bind to Ang2 and a second biological molecule, e.g., VEGF, such as anti-Ang2/anti-VEGF antibodies of the invention) may be used in therapeutic methods.

In one aspect, an anti-Ang2 antibody (e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as an anti-Ang2/anti-VEGF antibody of the invention) for use as a medicament is provided. In further aspects, an anti-Ang2 antibody for use in treating or delaying progression of a disorder associated with angiogenesis (e.g., pathological angiogenesis) is provided. In certain embodiments, an anti-Ang2 antibody for use in a method of treatment is provided. In certain embodiments, the invention provides an anti-Ang2 antibody for use in a method of treating an individual having a disorder associated with pathological angiogenesis comprising administering to the individual an effective amount of the anti-Ang2 antibody. In one such embodiment, the method further comprises administering to the individual an effective amount of at least one additional agent, for example, as described below.

In another aspect, the invention provides a method of reducing or inhibiting angiogenesis in a subject having a disorder associated with pathological angiogenesis, comprising administering to the subject an effective amount of an anti-Ang2 antibody of the (e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as an anti-Ang2/anti- VEGF antibody of the invention) thereby reducing or inhibiting angiogenesis in the subject. In certain embodiments, the disorder associated with pathological angiogenesis is an ocular disorder (e.g.

neovascular age-related macular degeneration) or a cell proliferative disorder. In one such embodiment, the method further comprises administering to the individual an effective amount of at least one additional agent, for example, as described below.

In another aspect, the invention provides a method of treating a tumor, a cancer, or a cell proliferative disorder, the method comprising administering an effective amount of an anti-Ang2 antibody of the invention (e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as an anti-Ang2/anti-VEGF antibody of the invention). In one such embodiment, the method further comprises administering to the individual an effective amount of at least one additional agent, for example, as described below.

In a further aspect, the invention provides for the use of an anti-Ang2 antibody of the invention

(e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g., VEGF, such as an anti-Ang2/anti-VEGF antibody of the invention) for the manufacture or preparation of a medicament. In one embodiment, the medicament is for treatment of a disorder associated with pathological angiogenesis (e.g., an ocular disorder or a cell proliferative disorder). In a further embodiment, the medicament is for use in a method of treating a disorder associated with pathological angiogenesis (e.g., an ocular disorder or a cell proliferative disorder) comprising

administering to an individual having a disorder associated with pathological angiogenesis an effective amount of the medicament. In one such embodiment, the method further comprises administering to the individual an effective amount of at least one additional agent, for example, as described below. An "individual" according to any of the above embodiments may be a human. In a further aspect, the invention provides pharmaceutical compositions comprising any of the anti-Ang2 antibodies of the invention (e.g. , bispecific anti-Ang2 antibodies of the invention that bind to Ang2 and a second biological molecule, e.g. , VEG F, such as an anti-Ang2/anti-VEG F antibody of the invention) provided herein, e.g., for use in any of the above therapeutic methods. In one embodiment, a pharmaceutical composition comprises any of the anti-Ang2 antibodies provided herein (e.g ., bispecific anti-Ang2 antibodies of the invention that bind to Ang2 and a second biological molecule, e.g., VEG F, such as anti-Ang2/anti-VEGF antibodies of the invention) and a pharmaceutically acceptable carrier. In another embodiment, a pharmaceutical composition comprises any of the anti-Ang2 antibodies provided herein (e.g ., bispecific anti-Ang2 antibodies of the invention that bind to Ang2 and a second biological molecule, e.g. , VEG F, such as anti-Ang2/anti-VEG F antibodies of the invention) and at least one additional agent, for example, as described herein.

Antibodies of the invention can be used either alone or in combination with other agents in a therapy. For instance, an anti-Ang2 antibody of the invention (e.g., a bispecific anti-Ang2 antibody of the invention that binds to Ang2 and a second biological molecule, e.g. , VEG F, such as an anti-Ang2/anti- VEGF antibody of the invention) may be co-administered with at least one additional agent. In certain embodiments, an additional agent is a chemotherapeutic agent, a cytotoxic agent, an anti-angiogenic agent, an immunosuppressive agent, a prodrug, a cytokine, a cytokine antagonist, cytotoxic radiotherapy, a corticosteroid, an anti-emetic, a cancer vaccine, an analgesic, a growth-inhibitory agent, an apoptotic agent, anti-tubulin agent, or other agent, such as a epidermal growth factor receptor (EG FR) antagonist (e.g., a tyrosine kinase inhibitor) , H ER1 /EG FR inhibitor (e.g., erlotinib (TARCEVA™), platelet derived growth factor inhibitor (e.g., G LEEVEC™ (Imatinib Mesylate)) , a COX-2 inhibitor (e.g., celecoxib) , interferon, cytokine, an antibody other than the anti-Ang2 antibody of the invention, such as an antibody that bind to one or more of the following targets Ang2, ErbB2, ErbB3, ErbB4, PDG FR-beta, BlyS, APR IL, BCMA, VEG F, or VEGF receptor(s) , TRAI L/Apo2, PD-1 , PD-L1 , PD-L2, or another bioactive or organic chemical agent.

Such combination therapies noted above encompass combined adm inistration (where two or more agents are included in the same or separate compositions) , and separate administration, in which case, administration of the antibody of the invention can occur prior to, simultaneously, and/or following, administration of the additional agent or agents. In one embodiment, administration of the anti-Ang2 antibody and administration of an additional agent occur within about one month, or within about one, two or three weeks, or within about one, two, three, four, five, or six days, of each other. Anti-Ang2 antibodies of the invention (e.g. , bispecific anti-Ang2 antibodies of the invention that bind to Ang2 and a second biological molecule, e.g. , VEGF, such as anti-Ang2/anti-VEGF antibodies of the invention) can also be used in combination with radiation therapy.

An anti-Ang2 antibody of the invention (e.g., a bispecific anti-Ang2 antibody of the invention that bind to Ang2 and a second biological molecule, e.g. , VEG F, such as an anti-Ang2/anti-VEGF antibody of the invention) , and/or any additional agent, can be adm inistered by any suitable means, including, intravitreally (e.g. , intravitreal injection or intravitreal implant) , by eye drop, intramuscularly, intravenously, intradermally, percutaneously, intraarterially, intraperitoneally, intralesionally, intracranially, intraarticularly, intraprostatically, intrapleural^, intratracheal^, intrathecally, intranasally, intravaginally, intrarectally, topically, intratumorally, peritoneally, subcutaneously, subconjunctival^, intravesicularly, mucosally, intrapericardially, intraumbilically, intraocularly, intraorbitally, orally, topically, transdermal^, by inhalation, by injection, by implantation, by infusion, by continuous infusion, by localized perfusion bathing target cells directly, by catheter, by lavage, in cremes, or in lipid compositions. The compositions utilized in the methods described herein can also be administered systemically or locally. The method of administration can vary depending on various factors (e.g., the compound or composition being administered and the severity of the condition, disease, or disorder being treated).

Dosing can be by any suitable route, for example, by injections, such as intravitreal, intravenous or subcutaneous injections, depending in part on whether the administration is brief or chronic. Various dosing schedules including but not limited to single or multiple administrations over various time-points, bolus administration, and pulse infusion are contemplated herein.

Antibodies of the invention would be formulated, dosed, and administered in a fashion consistent with good medical practice. Factors for consideration in this context include the particular disorder being treated, the particular mammal being treated, the clinical condition of the individual patient, the cause of the disorder, the site of delivery of the agent, the method of administration, the scheduling of

administration, and other factors known to medical practitioners. The antibody need not be, but is optionally formulated with one or more agents currently used to prevent or treat the disorder in question. The effective amount of such other agents depends on the amount of antibody present in the

composition, the type of disorder or treatment, and other factors discussed above. These are generally used in the same dosages and with administration routes as described herein, or about from 1 to 99% of the dosages described herein, or in any dosage and by any route that is empirically/clinically determined to be appropriate.

For the prevention or treatment of disease, the appropriate dosage of an antibody of the invention (when used alone or in combination with one or more other additional agents) will depend on the type of disease to be treated, the type of antibody, the severity and course of the disease, whether the antibody is administered for preventive or therapeutic purposes, previous therapy, the patient's clinical history and response to the antibody, and the discretion of the attending physician. The antibody is suitably administered to the patient at one time or over a series of treatments.

As a general proposition, the therapeutically effective amount of the antibodies of the invention administered to human will be in the range of about 0.01 to about 100 mg/kg of patient body weight whether by one or more administrations. In some embodiments, the antibody used is about 0.01 to about 45 mg/kg, about 0.01 to about 40 mg/kg, about 0.01 to about 35 mg/kg, about 0.01 to about 30 mg/kg, about 0.01 to about 25 mg/kg, about 0.01 to about 20 mg/kg, about 0.01 to about 15 mg/kg, about 0.01 to about 10 mg/kg, about 0.01 to about 5 mg/kg, or about 0.01 to about 1 mg/kg administered daily, for example. In one embodiment, an anti-Ang2 antibody described herein (e.g., a bispecific anti-Ang2 antibody of the invention that bind to Ang2 and a second biological molecule, e.g., VEGF, such as an anti-Ang2/Anti-VEGF antibody of the invention) is administered to a human at a dose of about 100 mg, about 200 mg, about 300 mg, about 400 mg, about 500 mg, about 600 mg, about 700 mg, about 800 mg, about 900 mg, about 1000 mg, about 1 1 00 mg, about 1200 mg, about 1300 mg or about 1400 mg on day 1 of 21 -day cycles. The dose may be administered as a single dose or as multiple doses (e.g., 2 or 3 doses), such as infusions. For repeated administrations over several days or longer, depending on the condition, the treatment would generally be sustained until a desired suppression of disease symptoms occurs. One exemplary dosage of the antibody would be in the range from about 0.05 mg/kg to about 1 0 mg/kg. Thus, one or more doses of about 0.5 mg/kg, 2.0 mg/kg, 4.0 mg/kg, or 10 mg/kg (or any combination thereof) may be administered to the patient. Such doses may be administered intermittently, for example, every week or every three weeks (e.g., such that the patient receives from about two to about twenty, or, for example, about six doses of the anti-Ang2 antibody). An initial higher loading dose, followed by one or more lower doses may be administered. The progress of this therapy is easily monitored by conventional techniques and assays.

In some embodiments, the methods may further comprise an additional therapy. The additional therapy may be radiation therapy, surgery, chemotherapy, gene therapy, DNA therapy, viral therapy, RNA therapy, immunotherapy, bone marrow transplantation, nanotherapy, monoclonal antibody therapy, or a combination of the foregoing. The additional therapy may be in the form of adjuvant or neoadjuvant therapy. In some embodiments, the additional therapy is the administration of small molecule enzymatic inhibitor or anti-metastatic agent. In some embodiments, the additional therapy is the administration of side-effect limiting agents (e.g., agents intended to lessen the occurrence and/or severity of side effects of treatment, such as anti-nausea agents, etc.). In some embodiments, the additional therapy is radiation therapy. In some embodiments, the additional therapy is surgery. In some embodiments, the additional therapy is a combination of radiation therapy and surgery. In some embodiments, the additional therapy is gamma irradiation. In some embodiments, the additional therapy may be a separate administration of one or more of the therapeutic agents described above.

H. Articles of Manufacture

In another aspect of the invention, an article of manufacture containing materials useful for the treatment, prevention and/or diagnosis of the disorders described above is provided. The article of manufacture comprises a container and a label or package insert on or associated with the container. Suitable containers include, for example, bottles, vials, syringes, IV solution bags, etc. The containers may be formed from a variety of materials such as glass or plastic. The container holds a composition which is by itself or combined with another composition effective for treating, preventing and/or diagnosing the condition and may have a sterile access port (for example the container may be an intravenous solution bag or a vial having a stopper pierceable by a hypodermic injection needle). At least one active agent in the composition is an antibody of the invention. The label or package insert indicates that the composition is used for treating the condition of choice. Moreover, the article of manufacture may comprise (a) a first container with a composition contained therein, wherein the composition comprises an antibody of the invention; and (b) a second container with a composition contained therein, wherein the composition comprises a further cytotoxic or otherwise therapeutic agent. The article of manufacture in this embodiment of the invention may further comprise a package insert indicating that the compositions can be used to treat a particular condition. Alternatively, or additionally, the article of manufacture may further comprise a second (or third) container comprising a pharmaceutically-acceptable buffer, such as bacteriostatic water for injection (BWFI), phosphate-buffered saline, Ringer's solution and dextrose solution. It may further include other materials desirable from a commercial and user standpoint, including other buffers, diluents, filters, needles, and syringes. /. Methods of Improving Antibodies

The invention provides methods of improving antibodies and identifying antibody variants. In some embodiments, the methods involve identifying one or more amino acid residue alterations that confers enhanced binding of an antibody to a target molecule. In other embodiments, the invention provides methods of identifying one or more amino acid residue alterations that allows enhanced binding of a dual specific antibody to both a first epitope and a second epitope. In other embodiments, the invention provides methods of improving the thermal stability, functional expression, and/or protein folding of an antibody.

For example, the invention provides a method of identifying an amino acid residue alteration that confers enhanced binding of an antibody to a target molecule that involve one or more (e.g., 1 , 2, or 3) of the following steps: (a) providing a display library that includes nucleic acids encoding candidate antibody variants, wherein each candidate antibody variant includes an amino acid residue alteration in each HVR of the V_H or the V_L compared to a reference antibody; (b) sorting the display library based on binding of the candidate antibody variants to the target molecule to form a sorted library, wherein the sorted library comprises candidate antibody variants with enhanced binding to the target molecule compared to the reference antibody; and/or (c) comparing the frequency at which each amino acid residue alteration is present in the display library and in the sorted library as determined by massively parallel sequencing, thereby determining whether each amino acid residue alteration is enriched in the sorted library compared to the display library, whereby the amino acid residue alteration is identified as conferring enhanced binding to the target molecule if it is enriched in the sorted library compared to the display library. In some instances, the method further includes determining the frequency at which each amino acid alteration is present in the display library and the sorted library by massively parallel sequencing following step (b). In some instances, step (c) further includes comparing the frequency at which a pair of amino acid residue alterations is present in the display library and in the sorted library, thereby determining whether the pair is enriched, depleted, or neutral in the sorted library compared to the display library. In some embodiments, the antibody is a dual specific antibody. In some instances of the preceding method, the target molecule is a polypeptide. In some embodiments, the target molecule is a cytokine. In some embodiments, the target molecule is VEGF or Ang2.

In some embodiments, an amino acid residue alteration or a pair of amino acid residue alterations may be enriched at least 2-fold in the sorted library compared to the display library. For example, an amino acid residue alteration or a pair of amino acid residue alterations may be enriched 1 .25-fold, 1 .5- fold, 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, 1 1 -fold, 12-fold, 14-fold, 16-fold, or more in the sorted library compared to the display library.

The invention also provides a method of identifying an amino acid residue alteration that allows enhanced binding of a dual specific antibody to both a first epitope and a second epitope that involves one or more (e.g., 1 , 2, 3, or 4), of the following steps: (a) providing a display library that includes nucleic acids encoding candidate antibody variants, wherein each candidate antibody variant includes an amino acid residue alteration in each HVR of the V_H or the V_L compared to a reference dual specific antibody; (b) sorting the display library based on binding of the candidate antibody variants to the first epitope to form a first sorted library, wherein the first sorted library includes candidate antibody variants with enhanced binding to the first epitope compared to the reference dual specific antibody; (c) sorting the display library based on binding of the candidate antibody variants to the second epitope to form a second sorted library, wherein the second sorted library includes candidate antibody variants with enhanced binding to the second epitope compared to the reference dual specific antibody; and (d) comparing the frequency at which each amino acid residue alteration is present in the display library, the first sorted library, and the second sorted library as determined by massively parallel sequencing, thereby determining whether each amino acid residue alteration is enriched, depleted, or neutral in the first sorted library and the second sorted library compared to the display library, whereby the amino acid residue alteration is identified as allowing enhanced binding of the dual specific antibody to both the first epitope and the second epitope if the amino acid residue alteration is enriched in both the first sorted library and the second sorted library compared to the display library or is enriched in one of either the first sorted library or the second sorted library and is neutral in the other sorted library. In some instances, the method further includes determining the frequency at which each amino acid residue alteration is present in the display library, the first sorted library, and the second sorted library by massively parallel sequencing following step (c). In some instances, step (d) further includes comparing the frequency at which a pair comprising a first amino acid residue alteration and a second amino acid residue alteration is present in the display library and in the first sorted library, the second sorted library, or both, thereby determining whether the pair is enriched, depleted, or neutral in the first sorted library, second sorted library, or both, compared to the display library. In some embodiments, the first epitope and the second epitope are from the same target molecule. In other embodiments, the first epitope and the second epitope are from different target molecules. In some embodiments, the first target molecule and/or the second target molecule are cytokines. In some embodiments, the first target molecule is VEGF and the second target molecule is Ang2.

In some embodiments, an amino acid residue alteration or a pair of amino acid residue alterations may be enriched at least 2-fold the first sorted library or second sorted library compared to the display library. For example, an amino acid residue alteration or a pair of amino acid residue alterations may be enriched 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, 1 1 -fold, 12-fold, 14-fold, 1 6-fold, or more the first sorted library or second sorted library compared to the display library.

In any of the preceding methods, the display library may include candidate antibody variants having one or more amino acid residue alterations at one or more positions in the antibody, for example, in the constant region and/or in the V_H and/or in the V_L, for example, in one or more HVRs and/or FRs. An antibody variant may include as many amino acid residue alterations as desired, for example, 1 , 2, 3, 4, 5, 6, 7, 8, 9, 1 0, 1 1 , 12, 13, 14, 15, 20, 25, or more amino acid residue alterations. In some instances, the display library may include candidate antibody variants having amino acid residue alterations at every position in each HVR of the V_H and/or V_L. For example, in some instances, the display library may include candidate antibody variants having amino acid residue alterations at every position in each HVR of the V_H. In other instances, the display library may include candidate antibody variants having amino acid residue alterations at every position in each HVR of the V_L. In some instances, the display library may include candidate antibody variants having amino acid residue alterations at every position in each HVR of the V_H and the V_L. In other instances, the display library may include antibody variants having amino acid residue alterations at a subset of positions in each HVR of the V_H and/or V_L. For example, in some instances, the display library may include candidate antibody variants having amino acid residue alterations at a subset of positions in each HVR of the V_H. In other instances, the display library may include candidate antibody variants having amino acid residue alterations at a subset of positions in each HVR of the V|_. In other instances, the display library may include candidate antibody variants having amino acid residue alterations at a subset of positions in each HVR of the V_H and the V_L.

In some instances, the display library includes amino acid residue alterations in only the V_H or the V_L of the candidate antibody variants. For example, in some instances, the display library includes amino acid residue alterations in only the V_H of the candidate antibody variants. In other instances, the display library includes amino acid residue alterations in only the V_L of the candidate antibody variants. In some instances, the display library includes a V_H library and a V_L library, wherein the V_H library includes candidate antibody variants with an amino acid residue alteration in each HVR of the V_H, and the V_L library includes candidate antibody variants with an amino acid residue alteration in each HVR of the V_L.

A display library may include any suitable number of antibody variants, for example, from about 1 x 10³ to about 1 x 10¹² or more (e.g., about 1 x 10³, about 1 x 10⁴, about 1 x 10⁵, about 1 x 10⁶, about 1 x 10⁷, about 1 .5 x 10⁷, about 2.5 x 1 0⁷, about 1 x 10⁸, about 1 x 10⁹, about 1 x 10¹⁰, about 1 x 10^{1 1} , about 1 x 10¹² or more) antibody variants.

In any of the preceding methods, an amino acid residue alteration may be encoded by any suitable codon set. For example, in some instances, the amino acid residue alteration is encoded by a degenerate codon set. Methods of substituting an amino acid of choice into a template nucleic acid are well established in the art, some of which are described herein. See also U.S. Patent No. 7,985,840, which is incorporated herein by reference in its entirety. For example, libraries as described above or in the Examples section below can be created by amino acid substitution with variant amino acids using the Kunkel method. See, for example, Kunkel et al., Methods Enzymol. 154:367-382, 1987.

An amino acid residue variation may be encoded by any suitable codon set. A codon set is a set of different nucleotide triplet sequences used to encode desired variant amino acids. Codon sets can be represented using symbols to designate particular nucleotides or equimolar mixtures of nucleotides as shown in below according to the IUB code.

IUB CODES

G Guanine

A Adenine

TThymine

C Cytosine

R (A or G)

Y (C or T) M (A or C)

K (G or T)

S (C or G)

W (A or T)

H (A or C or T)

B (C or G or T)

\/ (A or C or G)

D (A or G or T)

Λ/ (A or C or G or T)

As an illustrative example, in the codon set DVK, D ean be nucleotides A or G or T;

be A or G or C; and Kcan be G or T. This codon set can present 18 different codons and can encode amino acids Ala, Trp, Tyr, Lys, Thr, Asn, Ser, Arg, Asp, Glu, Gly, and Cys.

Oligonucleotide or primer sets can be synthesized using standard methods. A set of oligonucleotides can be synthesized, for example, by solid phase synthesis, containing sequences that represent all possible combinations of nucleotide triplets provided by the codon set and that will encode the desired group of amino acids. Synthesis of oligonucleotides with selected nucleotide "degeneracy" at certain positions is well known in that art. Such sets of nucleotides having certain codon sets can be synthesized using commercial nucleic acid synthesizers (available from , for example, Applied

Biosystems, Foster City, Calif.), or can be obtained commercially (for example, from Life Technologies, Rockville, Md.). Therefore, a set of oligonucleotides synthesized having a particular codon set will typically include a plurality of oligonucleotides with different sequences, the differences established by the codon set within the overall sequence. Oligonucleotides, as used according to the invention, have sequences that allow for hybridization to a variable domain nucleic acid template and also can include restriction enzyme sites for cloning purposes.

In any of the preceding methods, a degenerate codon set may be used to encode amino acid residue alterations. In some instances, the degenerate codon set is an NNKor an NNS codon set, wherein N is A, C, G, or T; K^" is G or T; and S is C or G. In particular instances, the degenerate codon set is an NNK codon set.

It is to be understood that any suitable display approach known in the art or described herein may be used in conjunction with any of the preceding methods. For example, the methods may involve phage display, bacterial display, yeast display, mammalian display, ribosome display, and/or m RNA display. In any of the preceding methods, any suitable display library may be used. For example, the display library may be selected from the group consisting of a phage display library, a bacterial display library, a yeast display library, a mammalian display library, a ribosome display library, and an m RNA display library. In particular embodiments, the display library is a phage display library.

Fusion polypeptides of an antibody variable domain can be displayed on the surface of a cell, virus, phagemid, or other particle in a variety of formats. These formats include, for example, scFv, Fab, and multivalent forms of these fragments. The multivalent forms may be a dimer of ScFv, Fab, or Fab', herein referred to as (ScFv)₂, Fab₂ and F(ab')₂, respectively. Methods for displaying fusion polypeptides comprising antibody fragments, on the surface of bacteriophage, are well known in the art, for example as described in patent publication number WO 92/01047 and herein. Other patent publications, for example, WO 92/20791 ; WO 93/06213; WO 93/1 1236, and WO 93/1 9172, describe related methods. Other publications have shown the identification of antibodies with artificially rearranged V gene repertoires against a variety of antigens displayed on the surface of phage (see, e.g., Hoogenboom et al. J. Mol. Biol. 227: 381 -388, 1992; and as disclosed in WO 93/06213 and WO 93/1 1236).

In any of the preceding methods, the display library may be sorted (selected) and/or screened to identify, for example, high-affinity binders to an antigen. Sorting may be performed as described herein or by other approaches known in the art. See, for example, U .S. Patent 7,985,840. In some embodiments, sorting may involve contacting the display library with an immobilized antigen (e.g., target molecule or epitope thereof). In other embodiments, sorting may involve contacting the display library with a soluble antigen. Antibody variants that have been selected can be further screened to characterize the antibody variant in terms of binding affinity (e.g., by SPR), stability, folding, structure (e.g., by X-ray

crystallography), or other attributes.

Any of the preceding methods may involve massively parallel sequencing, for example, to determine the frequency that an amino acid residue alteration appears in a library following sorting (referred to as a sorted library) as compared to the the frequency that the amino acid residue alteration appears in an unsorted library. A wide variety of approaches for massively parallel sequencing are known in the art, and any suitable approach may be used in the methods of the invention. See, for example, Metzker, Nature Reviews Genetics 1 1 : 31 -36, 2010, which is incorporated by reference herein in its entirety. Exemplary approaches include massively parallel signature sequencing (MPSS), polony sequencing, pyrosequencing (454/Roche Diagnostics), ion semiconductor sequencing, single-molecule real-time sequencing, sequencing by synthesis, sequencing by ligation. Commercially-available massively parallel sequencing platforms are available from Roche Diagnostics and other companies. The sequencing may be deep sequencing, ultra-deep sequencing, and/or next-generation sequencing.

In any of the preceding methods, the method may involve determining the sequence of at least about 100,000 reads or more (e.g., 100,000 reads, 200,000 reads, 300,000 reads, 400,000 reads, 500,000 reads, 600,000 reads, 700,000 reads, 800,000 reads, 900,000 reads, 1 ,000,000 reads, 2x1 0⁶ reads. 3x10⁶ reads, 4x10⁶ reads, 5x1 0⁶ reads, 6x10⁶ reads, 7x1 0⁶ reads, 8x10⁶ reads, 9x1 0⁶ reads, 10⁷ reads, 10⁸ reads, 10⁹ reads, 10¹⁰ reads, or more). The method may involve sequencing at any suitable depth.

It is to be understood that any of the preceding methods may be used to identify antibody variants with reduced properties (e.g., reduced binding affinity for a molecule). For instance, any of the preceding methods may be used to reduce binding affinity to a molecule for which the antibody has unwanted or off- target binding affinity, for example, an off-target protein. In some embodiments, the methods may be used to reduce binding affinity for the off-target protein while maintaining binding affinity to one or more target molecules. In such a method, the display library may be separately panned against both the target molecule and the off-target protein, and amino acid residue alterations that are depleted from the library panned against the off-target protein but are neutral or enriched in the library panned against the target molecule may be identified. Such amino acid residue alterations may be introduced into the reference antibody in order to reduce binding affinity to the off-target protein while maintaining binding affinity to the target molecule.

In any of the preceding methods, the antibody may be a monoclonal antibody. In any of the preceding methods, the antibody may be an IgG antibody. In any of the preceding methods, the antibody may be an antibody fragment. The antibody fragment may be selected from the group consisting of Fab, scFv, Fv, Fab', Fab'-SH, F(ab')₂, and diabody. In particular instances, the antibody fragment is a Fab.

In any of the preceding methods, the dual-specific antibody may be a monoclonal antibody. In any of the preceding methods, the dual-specific antibody may be an IgG antibody. In any of the preceding methods, the dual-specific antibody may be an antibody fragment. The antibody fragment may be selected from the group consisting of Fab, scFv, Fv, Fab', Fab'-SH, F(ab')₂, and diabody. In particular instances, the antibody fragment is a Fab.

Any of the preceding methods may further involve generating an antibody that has been identified by the steps of the method. The methods described above may be used with any of the antibodies described herein.

The invention also provides methods of generating a dual specific antibody that binds a first epitope with a Kd of lower than 1 nM and a second epitope with a Kd of lower than 1 nM that involves one or more (e.g., 1 , 2, or 3) of the following steps: (a) providing a dual specific antibody that binds the first epitope with a Kd of greater than 1 nM and the second epitope with a Kd of greater than 1 nM; (b) identifying one or more amino acid residue alterations that allows enhanced binding of the dual specific antibody to both the first epitope and the second epitope according to any of the preceding methods, wherein the one or more amino acid residue alterations allows binding the first epitope with a Kd of lower than 1 nM and the second epitope with a Kd of lower than 1 nM ; and (c) altering the amino acid sequence of the dual specific antibody based on the results of step (b), thereby generating a dual affinity antibody that binds a first epitope with a Kd of lower than 1 nM and a second epitope with a Kd of lower than 1 nM.

III. EXAMPLES

The following are examples of methods and compositions of the invention. It is understood that various other embodiments may be practiced, given the general description provided above. Example 1. Generation of Anti-Ang2/anti-VEGF Dual action Fabs (DAFs)

A. Library Construction

The cross-species VEGF-blocking antibody G6 (Lee et al. J. Molec. Biol. 340: 1073-1093, 2004; Fuh et al. J. Biol. Chem. 281 : 2265, 2006; Liang et al. J. Biol. Chem. 281 (2) : 951 -961 , 2006) was used as a starting template to recruit the second specificity towards Ang2 by mutation in its light chain HVRs. G6 has previously been demonstrated to block VEGF binding to its receptors VEGFR1 and VEGFR2.

Phage-displayed libraries were created using oligonucelotide-directed mutagenesis as described (Sidhu et al. J. Mol Biol. 338: 299-310, 2004). The library template vector G6 Fab-C was engineered to contain a single cysteine at the C-terminus of the heavy chain variable domain 1 upper hinge region to allow bivalent phage display of G6 Fab as previously described (Lee et al. J. Immun. Methods. 284(1 -2) : 1 19- 132, 2004). G6 Fab-C contained a stop codon (TAA) embedded in HVR-L1 , which was repaired during the mutagenesis reaction using degenerate oligonucleotides that annealed over the sequences encoding HVR-L1 libraries with length variations, HVR-L2 and HVR-L3. The library mutagenesis reactions were performed according to the method of Kunkel (Kunkel et al. Methods Enzymol. 154: 367-382, 1 987). The light chain HVR designs for the libraries are described in Figure 1 , which summarizes the degenerate codons used at each position for the different libraries. The mutagenesis products were pooled into one reaction per library and electroporated into E. coli SS320 cells and grown supplemented with K07 helper phage as described (Lee et al. J. Mol. Biol. 340(5) : 1073-1093, 2004). Approximately 10^{1 1} cells and approximately 5-1 0 μg DNA were used in each electroporation reaction. The number of transformants ranged from 10⁹-10¹⁰ transformants. Six libraries were generated. The display level and VEGF binding of individual library on the surface of phage was determined in ELISA binding assays (Figures 2A-2B).

B. Library Sorting and Screening

In order to select for clones that specifically bind to both Ang2 and VEGF, the libraries described above were sorted by binding selection (also referred to as panning) and then screened. The phage libraries were subjected to five rounds of binding selection for Ang2 using a purified fragment of human Ang2 (hAng2) containing the fibrinogen-like receptor-binding domain (RBD) fused to Fc (referred to herein as Fc.hAng2.RBD). A purified fragment of human Ang1 (hAngl ) containing the RBD fused to Fc (referred to herein as Fc. hAngl .RBD) was used starting in the third round to remove hAngl binders, and fifth round phage libraries were additionally panned against human VEGF (hVEGF₁₀₉), as described below.

NUNC 96-well MAXISORP® plates were coated overnight with Fc.hAng2.RBD (5 μg/ml) and blocked 1 h with alternating blocking agents such as 1 % bovine serum albumin (BSA) in phosphate buffered saline (PBS) or casein. Phage solutions of 10¹³ phage/ml were added to the coated immunoplates in the first selection cycle. The phage concentration was decreased in each round of selection. Following incubation of the phage solutions on the immunoplates to allow binding to the immobilized antigen, the plates were repeatedly washed with PBS, 0.5% TWEEN-20®. To increase the stringency, the incubation time was decreased (4 h for 1 ^st round, 3 h for 2^nd round, 3 h for 3^rd round, 2 h for 4^th round, 1 h for 5^th round) and the number of washes was increased in each round of selection. Bound phage was eluted with 0.1 M HCI for 30 minutes and the eluant was neutralized with 1 .0M Tris base. The recovery of phage per antigen-coated immunoplate well was calculated and compared to that of a blocked well without coated antigen to study the enrichment of phage clones displaying Fabs that specifically bound the target antigen. Eluted phage were amplified in E. coli and used for further rounds of selection. After two rounds of panning with Fc.hAng2.RBD, six libraries were divided in two pools (short lengths L1 -3 and long lengths L4-6). Starting from the third round, phage libraries were incubated with 1 μΜ Fc. hAngl .RBD for 2 hours before applying on Fc.hAng2.RBD coated plates to remove hAngl binders. At the fifth round, phage libraries were panned against both Fc.hAng2.RBD and hVEGF₁₀₉. There was over a thousand fold enrichment in both short and long length libraries at round 4 and 5. Random clones from round 4 and 5 were selected for screening and assayed using phage ELISA in which binding to hAngl , hAng2, and hVEGF were compared to binding of a non-relevant protein (BSA) to control for non-specific binding. The V_L regions of the positive clones were sequenced as described (Sidhu et al. J. Mol. Biol. 338: 299-31 0, 2004).

C. High-throughput characterization of hAng2/h VEGF binding clones

A high-throughput single spot competitive ELISA in a 96-well format (Sidhu et al. J. Mol Biol. 338:

299-310, 2004) was used to screen for high affinity clones for both Ang2 and VEGF (affinity screen), and to study the Tie2.Fc blocking profiles of these high affinity clones (receptor-blocking screen).

Briefly, MAXISORP® immunoplates were coated with 2 μg/ml hAng2.RBD.Fc overnight at 4°C and then blocked with 1 % (w/v) BSA for 1 h. Phagemid clones in E. co//^' XL1 -Blue were grown in 400 μΙ of 2YT broth supplemented with carbenicillin and M13-K07 helper phage; the cultures were grown with shaking overnight at 37°C in a 96-well format. Culture supernatants containing phage were diluted fivefold in PBST (PBS with 0.05% TWEEN-20® and 0.5% (w/v) BSA) with or without the addition of 50nM Fc.hAng2.RBD or

for the affinity screen. For the receptor-blocking screen, Fc.hAng2.RBD coated wells were incubated with or without Tie2.Fc before adding five-fold diluted phage supernatant. After incubation for 1 h at room temperature, the mixtures were transferred to the coated plates with Fc.hAng2.RBD and incubated for 10 min. The plate was washed with PBT (PBS with 0.05% TWEEN- 20®) and incubated for 30 min with anti-M13 antibody horseradish peroxidase (HRP) conjugate diluted 5000-fold to 1 nM in PBST. The plates were washed, developed with 3,3',5,5'-tetramethylbenzidine (TMB) substrate for approximately five minutes, quenched with 1 .0M H₃P0₄, and read spectrophotometrically at 450 nm.

In the affinity screen the ratio of the absorbance in the presence of solution-phase Fc.hAng2.RBD or hVEGF₁₀₉ to that in the absence of solution-phase Fc.hAng2.RBD or hVEGF₁₀₉ was used as an indication of the affinity to Ang2 or VEGF, respectively. A low ratio indicates that most of the Fab-phage were bound to solution-phase Fc.hAng2.RBD or hVEGF₁₀₉ in the initial incubation stage and, therefore, were unavailable for capture by immobilized Fc.hAng2.RBD. Similarly, for the blocking assay a low ratio indicated that the binding of a clone to Fc.hAng2.RBD is blocked by the interaction between

Fc.hAng2.RBD and Tie-2.Fc, indicating these clones are likely to be displaying blocking antibodies. Clones with a low ratio for both Fc.hAng2.RBD and hVEGF₁₀g (affinity screen) and Tie2.Fc blocking (receptor-blocking screen) were selected for further characterization.

Six unique clones were identified by sequencing and propagated from a single colony by growing in 25 ml of 2YT culture supplemented with carbenicillin and K07 helper phage overnight at 30°C. Phage purified by precipitation in PEG/NaCI were first diluted serially in PBST and tested for binding to an antigen-coated plate. The dilution that gave 50-70% saturating signal was used in the solution binding assay in which phage were first incubated with increasing concentration of antigen for two hours and then transferred to antigen-coated plates for 15 minutes to capture the unbound phage. IC₅₀ was calculated as the concentration of antigen in solution-binding stage that inhibited 50% of the phage from binding to immobilized antigen, as described previously (Lee et al. J. Mol. Biol. 340(5) : 1 073-1093, 2004). Phage IC₅₀ against hAng2, hAngl and hVEGF was determined for each of the six unique clones (Figures 3A- 3D). The V|_ and V_H amino acid sequences of each of the six unique clones identified by this method are shown in Figures 4A and 4B, respectively. D. Expression of Library Binders in IgG Format

The variable domains of the light and heavy chains of each of the six unique hAng2/hVEGF binding clones identified above were cloned into a vector previously designed for transient human IgG expression in mammalian cells (Lee et al. J. Mol. Biol. 340(5) : 1073-1093, 2004). Human IgG was purified with protein A affinity chromatography and screened by ELISA for binding their respective antigen(s). The dual-specific IgG antibodies that were produced using the variable domain of the light and heavy chains are specific to immobilized hAng2 and hVEGF, although clones 5A1 , 5B12, and 5C7 display some residual binding to hAngl (Figure 5). These results confirm the specificity of the variable domains of these clones for hAng2 and hVEGF.

E. Confirmation of receptor-blocking activity of anti-Ang2/anti-VEGF IgG antibodies

To confirm the Tie2 receptor-blocking activity of anti-Ang2/anti-VEGF IgG antibodies, a receptor-blocking assay was performed in a competitive ELISA format where a fragment of the receptor Tie2 fused to Fc (Tie2.Fc) was immobilized on MAXISORP® immunoplates at 2 μg/ml. The solution competition binding assay used 1 0nM biotinylated hAng2 which was equilibrated with serial dilutions of purified IgG proteins for 2 h. The unbound biotin-hAng2 was captured with immobilized Tie2.Fc on MAXISORP® plates for 15 min and detected with streptavidin-conjugated HRP. All isolated antibodies blocked hAng2 from binding to Tie2.Fc. For example, isolated antibodies 5A1 and 5A12 had better Tie2 blocking activity than the previously-characterized anti-Ang2 antibody Ab536 (Amgen) (Figure 6).

Example 2: Affinity Maturation of Anti-Ang2/anti-VEGF DAF 5A12

A. DAF Affinity Maturation Library Construction and Screening

For affinity improvement, phage libraries were constructed from the anti-Ang2/anti-VEGF DAF 5A12 (described in Example 1 ) in Fab-amber format for monovalent display with selected LC (HVR-L1 , L2, or L3) or HC (HVR-H2) residues mutated using either limited or soft randomization to allow either limited diversity based on natural amino acids or approximately 50% of wildtype with approximately 50% all other amino acids, respectively using degenerate oligonucleotides synthesized with 70-10-10-10 mixtures of nucleotide bases with the wild-type nucleotide in excess (Bostrom et al. Methods Mol. Biol. 5245: 353-376, 2009) (Figure 7). The resultant library DNA was electroporated into E. coli XLI cells yielding approximately 10⁹ transformants. Libraries were panned on immobilized hVEGF₁₀₉ or His-tagged human Ang2 (hAng2his8) with subsequent rounds using target in solution. A high-throughput single spot competitive ELISA was used to screen for high affinity clones for both hAng2 and hVEGF binding, as previously described (Sidhu et al. J. Mol Biol. 338: 299-310, 2004).

Selected clones were submitted for HVR sequencing and were ranked using phage IC₅₀ affinity,

Biacore affinity and receptor-blocking ability as well as tested for hAngl binding. Each of the affinity- matured 5A12 variant clones showed improvement in phage IC₅₀ affinity compared to 5A12 (Figures 8A- 8C). Variant clones were used as analyte in BIACORE® surface plasmon resonance (SPR)

measurements using a CM5 sensor chip immobilized with hVEGF_{1 0}9 or hAng2his8 at 25 °C to determine monovalent affinities. For example, clone 5A12 4.2 was measured using BIACORE® SPR to have a dissociation constand (Kd) of 5 nM for both Ang2 and VEGF, which is more than a magnitude lower than the affinity of anti-VEGF or anti-Ang2 antibodies which have been demonstrated to be efficacious in the clinic or in preclinical models (see, e.g., Rosenfeld et al. Opthamol. Clin. North Am. 19(3) : 361 -372, 2006). The V_L and V_H amino acid sequences of selected affinity-matured clones obtained by this method are shown in Figures 9A and 9B, respectively.

B. Antibody Characterization

The V|_ and V_H of selected affinity-matured clones were cloned into vectors previously designed for transient human IgG or Fab expression in mammalian cells (Lee et al. J. Mol. Biol. 340(5) : 1073-1093, 2004). Antibodies were purified with protein A affinity chromatography and screened for specificity by ELISA using their respective antigen(s) and by baculovirus ELISA as a measure of non-specific binding (Hotzel et al. MAbs. 4(6) : 753-760, 2012). Antibodies (IgG) of affinity-matured variants, including 5A12 4.2, showed only modest non-specific baculovirus particle binding as compared to control antibodies Rituxan and R5D, which show moderate and high levels of baculovirus particle binding, respectively (Figure 10).

To confirm Tie2.Fc blocking activity of the affinity-matured variants expressed in IgG format, a receptor-blocking competitive ELISA was performed. Tie2.Fc was immobilized on MAXISORP® immunoplates (2 μg/ml) and biotinylated hAng2 was equilibrated in solution with serial dilutions of purified antibodies prior to unbound biotin-hAng2 being captured with the immobilized Tie2.Fc and detected with streptavidin-conjugated HRP. Affinity-matured variants, including 5A12 4.2, showed improved blocking of the interaction between hAng2 and Tie2 (Figure 1 1 ).

A similar receptor-blocking competitive ELISA was performed to test for blocking of the VEGF- Flt-1 receptor interaction using capture on immobilized Flt-1 receptor with solution equilibration of biotinylated VEGF₁₆₅ with serial dilutions of purified antibodies (Bostrom et al. Science. 323(5921 ) : 1 610- 1614, 2009). Affinity-matured variants, for example 5A12 4.2, showed an approximate 4-fold improvement in VEGF blocking compared to 5A12 (Figure 12). Selected clones and improved variants purified as Fabs were also used as analyte in BIACORE® SPR measurements using a CM5 sensor chip immobilized with hVEGF₁₀₉ or hAng2his8 at 25Ό to determine monovalent affinities. Example 3: Structural analysis of anti-Ang2/anti-VEGF DAFs

A. Protein crystallization and structure determination

Purified protein samples were prepared for the Fab and antigen constructs. A C-terminal His- tagged receptor binding domain of Ang2 (residues 277-496) was expressed extracellularly in T. ni insect cells. The supernatant was loaded onto a Ni-NTA superflow column, washed with 50 mM Tris-CI pH 8.0 and 300 mM NaCI, and eluted with 250 mM imidazole in the same buffer. Fractions containing Ang2 were further purified over a SUPERDEX® S200 size exclusion column in 20 mM Tris-CI, pH 7.5 and 300 mM NaCI. Human VEGF (residues 8-109) was expressed, refolded, and purified as previously described (Christinger et al. Prot. Struct. Fund. Genet 26: 353-357, 1996). Fab 5A12 was expressed in E. coli. The cell paste was thawed into a buffer of PBS, 25 mM EDTA and 1 mM PMSF. The mixture was

homogenized and then passed twice through a microfluidizer. Cleared lysate (12,000 rpm , 60 min) was loaded onto a Protein G column equilibrated with PBS and eluted with 0.58% acetic acid. Peak fractions were loaded onto a SP-SEPHAROSE® column equilibrated with 20 mM MES, pH 5.5. The protein was eluted with a salt gradient of 0 to 0.25 M NaCI and peak fractions passed over a S200 column in 20 mM TrisCI, pH 7.5 and 250 mM NaCI (buffer A).

Complexes were prepared by mixing the respective components with purification over a size- exclusion chromatography column. Crystals of 5A12 4.2 in complex with Ang2 were grown by mixing a solution containing the two proteins with an equal volume of crystallization solution of 0.1 M Tris pH 8.5, 0.2 M Li₂S0₄, 25% polyethylene glycol (PEG) 3350 and 3% sucrose as an additive. These crystallization drops were incubated in hanging drop vapor diffusion format at 19°C. Crystals of 5A12 4.2 in complex with VEGF were grown by mixing solution containing the two proteins with equal volume of crystallization solution of 0.1 M Tris pH 8.0, 2 M ammonium sulfate, and 0.3 M non-detergent sulfo-betaine 195 as an additive. These crystallization drops were incubated in sitting drop vapor diffusion format at 19°C. Single crystals were extracted, cryoprotected by passage through a solution containing the respective crystallization mother liquors with 20% glycerol, and flash frozen in liquid nitrogen for data collection.

Data were collected at the Advanced Photon Source beamlines 21 -IDG and 21 -IDF. The structures were solved by molecular replacement using prior Ang2, VEGF, and unrelated Fab structures as search models. The coordinates were refined to the following statistics and reasonable geometry criteria: Fab5A12 4.2 : Ang2 complex crystal structure

resolution: 2.27 A

Spacegroup: C2

Unit cell : 181 .7 x 109.3 x 43.9 with angles 90°, 100.5°, 90°

R/Rfree: 18.1 1 /22.29

Ramachandran outliers: 1 .4%

Fab5A12 4.2 : VEGF complex crystal structure

Resolution: 2.75 A

Spacegroup: P21212

Unit cell : 87.4 x 313.9 x 51 .1 with angles 90°, 90°, 90°

R/Rfree: 20.21 /24.47

Ramachandran outliers: 1 .6%

To classify positions as buried or solvent-exposed, the crystal structures of the Fab molecules in the hAng2 and hVEGF bound state were analyzed using the program Arealmol of the CCP4 suit (Winn et al. Acta. Cryst. D67: 235-242, 201 1 ). The resulting solvent accessible area (SAA) was compared with the maximal solvent accessible area (mSAA) (Miller et al. J. Mol. Biol. 196(3) :641 -656, 1987). Residues which had a SAA/mSAA < 0.1 were classified as buried, and residues which had have a SAA/mSAA > 0.5 were classified as solvent-exposed. The figures were generated using Pymol. B. Structural basis for anti-Ang2/anti-VEGF DAF dual binding affinity

Using the methods described in Example 3A, the crystal structures of the 5A12 4.2 Fab in complex with the receptor-binding domains of Ang2 or VEGF were generated (Figures 13A-13D). Both antigens, Ang2 and VEGF, in the structures are highly similar to previously reported structures of Ang2 in complex with Tie2 (RMSD 0.567A over 1405 atoms, see, e.g., Barton et al. Nat. Struct. Mol. Biol. 13(6) : 524-532, 2006) and VEGF in complex with the G6 Fab (root-mean-square deviation (RMSD) 0.43A over 1 1 10 atoms, see, e.g., Fuh et al. J. Biol. Chem. 281 (10) : 6625-6631 , 2006). The epitope on Ang2 recognized by 5A12 4.2 largely overlaps with the Tie2 binding site of Ang2, potentially explaining the mechanism of Ang2 blockade (Figure 14A and 14B). The epitope of 5A12 4.2 on VEGF is nearly identical to the epitope of G6 as described, e.g., by Liang et al. J. Biol. Chem. 281 (2) : 951 -961 , 2006 (Figure 14C and 14D).

The crystal structures of the 5A12 4.2 Fab in complex with the receptor-binding domains of Ang2 or VEGF allowed a determination of the epitopes of Ang2 and VEGF that are bound by 5A12 4.2. Table 2 shows residues of the human Ang2 epitope that are contacted by the DAF 5A12 4.2 as determined by the crystal structure to have any atom within 5 A of 5A12 4.2. 5A12 4.2 buries a surface area of 638 A² on Ang2 (Figure 14A). Tables 3A and 3B show residues of the human VEGF epitope that have any atom within 5 A of 5A12 4.2 as determined by the crystal structure. The crystal structure of 5A12 4.2 bound to VEGF shows two copies of the Fab molecule bound to either end of the symmetrical VEGF dimer. The two copies of 5A12 4.2 bury surface areas of approximately 830 A² and 903 A², together masking approximately 1 733 A² of VEGF surface area.

Table 2: Epitope residues of hAng2 bound by 5A12 4.2

Table 3A: Epitope residues of VEGF bound by 5A12 4.2 Fab1

Table 3B: Epitope residues of VEGF bound by 5A12 4.2 Fab2

Distance Cutoff 5 A

Phe17

Met18

Tyr21

Gln22

Tyr25

Lys48

Asn62

Asp63

Residues Glu64

Gly65

Leu66

Met81

Ile83

Lys84

Pro85

His86 Distance Cutoff 5 A

Gln87

Gly88

Gln89

His90

Ile91

Cys104

Pro106

The crystal structures of the 5A12 4.2 Fab in complex with the receptor-binding domains of human Ang2 or human VEGF allowed a determination of the amino acid residues of 5A12 4.2 that are involved in binding to Ang2 or VEGF. Table 4 shows the amino acid residues of the paratope of 5A12 4.2 that is involved in binding to Ang2. Table 5 shows the amino acid residues of the paratope of 5A12 4.2 that is involved in binding to VEGF.

Table 4: Paratope residues of 5A12 4.2 that bind hAng2

Table 5: Paratope residues of 5A12 4.2 that bind hVEGF

Five HVR positions of 5A12 4.2 are involved in binding both VEGF and Ang2 (HC-Leu99, HC-

Tyr100a, LC-Ser30, LC-Phe31 , and LC-Leu92). Moreover, when mapping the two paratopes on the surface of the antigen binding region, it is evident that the spatial organization of the two paratopes is highly intertwined (Figures 13C and 13D). In order to accommodate the binding of two structurally diverse epitopes with essentially the same binding site, 5A12.4.2 utilizes a large degree of plasticity similar as previously described for a Her2/VEGF DAF (Bostrom et al. PloS One 6(4) :e17887, 201 1 ). A comparison between the HVR loop conformation of 5A12 4.2 in the VEGF bound state (5A12_VEGF) with 5A12 4.2 in the Ang2 bound state (5A12_Ang2) shows that most of the HVR-loops adopt the same conformations, with two notable exceptions: HVR-H3 and HVR-H2 (Figure 14C). HVR-H2, in particular, undergoes drastic conformational change in the VEGF-bound state compared to Ang2 bound state; HC-Ala53, for example, moves by 14 A (Figure 14D). When overlaying the structures of 5A12 4.2_VEGF nd 5A12 4.2_Ang2 with the parental G6 in the VEGF bound state (G6_VEGF) or in the unbound state (G6_unbound) , it is evident that conformation of HVR-H2 and HVR-H3 in 5A12 4.2 VEGF and G6VEGF are identical while the conformation of the two loops which are used to bind Ang2 resemble G6_unb0und- Thus, the plasticity of the 5A12 4.2 binding site was inherited from the parental antibody G6 and not added when recruiting dual-specificity. In addition to the large conformational changes of the HVR-H2 and HVR-H3 region, some side chains in other HVR loops use different rotamers during binding of the two antigens thereby providing additional plasticity without the requirement for rearrangement of the main chain backbone. Ang2 binding by 5A12 4.2 does not substantially involve the HVR-H2 loop, while the HVR-H2 loop is involved in VEGF binding (Figures 15A and 15B). Example 4: Deep mutagenesis scanning of the Anti-Ang2/anti-VEGF DAF 5A12 4.2 to generate DAFs with sub-nanomolar affinity for both antigens

A. Deep mutagenesis scanning to identify substitutions compatible with enhanced dual affinity The intertwined paratopes of the anti-Ang2/anti-VEGF DAF 5A12 4.2 for Ang2 and VEGF and structural plasticity of the 5A12 4.2 binding site, as determined in Example 3, suggested that it could be difficult to identify mutations that are compatible with improving the binding towards both antigens. To overcome this challenge, we used deep mutagenesis scanning, which combined combinatorial library selection and deep sequencing using next-generation sequencing technology (also known as high- throughput sequencing or massive parallel sequencing), to assess the effect of all possible individual mutations of 5A12 4.2 HVRs on binding both antigens. As a comparison, 5A12 4.2 was also affinity matured in parallel using phage display of randomized libraries and selection of binding variants without deep sequencing analysis.

We used a library design in which all three HVRs of either the heavy or light chain of 5A12 4.2 were mutated simultaneously (Table 6). Each clone in this library carried three

mutations (N = A or T or G or C; K= G or T), one in each of the three HVRs. We refer to this library design as triple- Λ/Λ// (3ΝΝΚ). This library design strategy allowed us to estimate the effect of single as well as double mutations (i.e., mutation pairs) on antibody binding function, which enabled epitasis analysis. Mutation pairs with positive cooperativity can be an important driver for the fitness (e.g., the binding affinity, expression, and/or stability) of a protein as they can, for example, allow overcoming local fitness minima and promote stability. This approach overcame a potential drawback of previous deep mutagenesis scanning approaches that involve mutating each HVR separately, which limits the resulting information to the effect of single mutations on binding function. As a control to determine if the 3NNK library design allowed for the correct fitness assessment of single mutations, we also generated a library where each HVR of the 5A12 4.2 heavy chain was mutated separately (referred to as 1 NN K) (Table 6). Table 6: Library design overview

Enrichment ratios (ER) as a measure of antibody fitness for all mutations at all randomized positions were calculated by dividing the frequency of a mutation in sorted pools with the frequency in the unsorted libraries. This gave a comprehensive overview of the effect of single mutations (Figures 1 6A- 16B). To determine if it was possible to estimate the fitness score of single mutations from the 3NN K library design, we compared the enrichment ratios obtained from the 3NNK library design with the ratios obtained from the 1 NNK library design. Enrichment ratios obtained with the two different library designs correlated well (using a linear model, the R² for the Ang2-panned samples is 0.73, and the R² for the VEGF-panned samples is 0.82), demonstrating the validity of this approach (Figures 16C-16D).

B. Comparison of structural and deep mutagenesis scanning data

We compared the structural data from Example 3 with the mutagenesis data obtained from the 3NNK libraries. The ER of 1092 mutations for the two binding functions of 5A12 4.2 are plotted as a heatmap (Figures 16A-16B). Interestingly, most of the 52 HVR positions exhibited distinct profiles of enrichment (red) and depletion (blue) of the mutations with VEGF or Ang2 binding selection, indicating that the two binding functions involve differential usage of these HVR positions. The functional paratopes, i.e., the HVR positions especially important for antigen interaction and therefore largely not tolerant to mutation, were identified by calculating the average of the enrichment ratios of all mutations at every position. The functional paratopes for both antigens were well within the structural paratopes for each antigen (compare Figures 13C-13D and 17A-17B). In contrast to the extensively overlapping structural paratopes for its two antigens (see Figures 13C-13D), the two functional paratopes had some degree of spatial separation (Figures 17A-17B). Hotspot positions (positions which do not tolerate mutation) for Ang2 binding were located mostly on the light chain (e.g., HVR-L1 , HVR-L2 and HVR-L3) with the exception of HC-Y100a, while hotspot positions for VEGF binding were located primarily on the heavy chain (e.g., HVR-H3) (Figure 18).

Structurally, the HVR-H2 loop of 5A12 is engaged with VEGF, whereas it is largely not involved in binding Ang2 (see Example 3). The differences in the different binding modes were reflected in the mutational data: the data obtained from the Ang2-panned sample suggested that CDR-H2 does tolerate mutations at several positions without effecting Ang2 binding, while the data from the VEGF-panned sample showed that mutations at several positions were not compatible with VEGF binding (Figures 15A- 15B and 19A-19B). For example, three residues in HVR-H2, HC-Gly50, HC-Gly54, and HC-Gly55, did not tolerate mutations in the VEGF-panned sample (Figure 1 9D). It has been previously suggested that these residues are required for function of the parental G6 antibody, as they might provide the flexibility for HVR-H2 to adopt the unusual conformation for VEGF binding (Fuh et al. J. Biol. Chem. 281 (10) :6625- 6631 ) (see also Figure 18).

In some instances, there are also similarities in the observed enrichment patterns for a few HVR positions between the Ang2- and VEGF-sorted samples, suggesting antigen-independent effects at these positions. Positions where both samples show mutation-sensitive results could in principle play a common structural role. For example, HC-Arg94 and HC-Ala49 are structurally buried at the base of HVR loops and do not make contact with either antigen, yet both tolerate few if any mutations (Figures 16A- 16B). HC-Arg94 forms a salt bridge with position Asp101 , which is important for CDR-H3 conformation (Morea et al. J. Mol. Biol. 275-269-294, 1998). Positions where both samples show similar toleration to most mutations (i.e., maintained binding for both antigens, for example, HC-Thr28 and LC-Ser56) tend to be surface-exposed and structurally distant from the antigen-binding area; their ability to tolerate mutations indicates relatively little involvement in antigen binding and structural restraint.

The enrichment pattern further provides insight into the mechanism of affinity maturation towards the single antigens. Several contact residues (less than 5 A from an antigen atom) were identified with either single mutations (e.g., for Ang2 binding: LC-S30aT; for VEGF binding: LC-A53D, HC-Y1 OOaW) or with a series of mutations (e.g., for Ang2 binding: LC-L93V, I,R, and K; for VEGF binding: LC-F31 H,S, T, and N) with high enrichment, suggesting that these mutations optimize the contact between the antigen and the antibody (Figures 16A-16B). The optimization of existing polar and non-polar antigen-antibody contacts is a well known mechanism for driving affinity maturation of antibodies. Further, some buried positions that are not in direct contact with the antigens contain selected substitutions that are highly enriched (e.g., for Ang2 binding LC-A34M, LC-Q89HW; for VEGF binding HC-F95Y, A100bV), demonstrating structural alteration of buried, non-contact sites in antibody affinity maturation.

Therefore, the mutagenesis data obtained using deep sequencing of 3NNK libraries were consistent with the structural data, further validating the deep mutagenesis scanning approach.

C. Using deep scanning mutagenesis to identify residues for affinity maturation

For further guided affinity maturation of 5A12 4.2, we used the deep mutagenesis scanning results to identify substitutions which either increased the affinity towards both antigens or that improved the binding towards one antigen without decreasing the affinity toward the other antigen. An example of the latter is mutation HC-A53D, which was enriched in the VEGF-panned sample but was neither enriched nor depleted in the Ang2-panned sample (see circled positions in Figures 19C and 19D). A comparison of the enrichment of each mutation obtained in VEGF panning versus the enrichment obtained in Ang2 panning in the two 3NNK libraries suggested that several residues, including several light chain residues, could be further optimized to increase dual affinity to both VEGF and Ang2 (Figures 20A-20B).

We also determined the position of mutations which show strong enrichment. Based on the structural data, HVR positions could be classified as exposed or buried residues (see Example 3). For this analysis, mutations which were enriched 4-fold were considered strongly enriched. While solvent- accessible exposed residues were more strongly enriched than buried residues on average, more buried mutations showed strong enrichment than solvent-exposed mutations (Figure 18). This suggested that higher affinity gains could be achieved by optimizing buried positions.

To determine how the enrichment data correlated with antigen binding affinity, we selected 25 substitutions which showed increased enrichment to at least one of the antigens (Ang2 binding: HC-28P, HC-30M, HC-35Q, HC-35Y, HC-35W, HC-52D, HC-53D, HC-54E, HC-56H, HC-95Y, HC-97P, HC- 100aW, HC-100bT, HC-100bS, LC-28V, LC-30aT, LC-33L, LC-34M, LC-34N, LC-54N, LC-54T, LC-56W, LC-89W, LC-89H, LC-92S, LC-93K; VEGF binding: HC-28P, HC-30M, HC-35Q, HC-35Y, HC-35W, HC- 51 T, HC-52D, HC-53D, HC-54E, HC-56H , HC-95Y, HC-97P, HC-100aW, HC-100bT, HC-100bV, HC- 100bS, LC-28V, LC-30aT, LC-33L, LC-34M, LC-34N, LC-54N, LC-54T, LC-56W, LC-89W, LC-89H, LC- 92S, LC-93K).

We generated clones harboring each single point mutation in 5A12 4.2. To obtain an affinity estimate, the IC50 values of these clones for both antigens were determined in a phage competition assay (Figure 21 ). The results show that the ER metric serves as an excellent classifier to distinguish affinity improving mutations from mutations which have no effect on affinity or reduce the affinity (receiver operator characteristics, area under the curve 0.93). While the relationship between fold enrichment and the fold IC50 change was not linear, the fold enrichment ratios obtained from either the 3NN K or the 1 NNK libraries classified most of the mutants correctly as mutations which increase, decrease, or have neutral fitness.

D. Enrichment analysis in combination with an cooperativity analysis can be used to identify mutation pairs that enhance binding synergistically

The 3NNK library design, in which all three HVRs of the heavy or light chain were mutated simultaneously, enabled identification of pairs of mutations located in different HVRs which increased the affinity of the DAF towards one or even both antigens. Without wishing to be bound by theory, mutation pairs could be enriched for several reasons. First, the pair could contain one mutation that by itself has a strong impact on affinity. This increases the likelihood that the second mutation in the pair is a

"hitchhiker," which does not contribute significantly to improved affinity. Second, the pair could contain a fold-stabilizing mutation. For instance, the nature of phage display allows for the enrichment of fold- stabilizing mutations which have little or no impact on antigen binding. Third, the pair could contain mutations which contribute to antigen binding at least additively or in some cases synergistically.

We calculated enrichment ratios for double mutations from all four 3NNK datasets. The enrichment ratios of double mutations were determined by calculating the enrichment ratio of all clones which carried NNK mutations at two given positions, ignoring the third NNK mutation. We calculated ER for 170,204 out of 400,428 possible mutation pairs. To filter out sampling effects, mutation pairs which had less than 10 sequence counts either in the sorted or unsorted sample were removed from the analysis. One surprising pattern was that a few distinct mutations were identified in several top enriched mutation pairs, as shown in Figures 22A and 22B. For example, in the dataset obtained from the Ang2- panned HC-3NN K library, the HC-F97P mutation formed strongly enriched mutation pairs with several distinct mutations located in the HVR-H1 and HVR-H2 loop. Further, the majority of the "mutation partners" of HC-F97P were not located in close spatial proximity to HC-F97P based on an analysis of the structure of 5A12_Ang2. A similar pattern was observed for mutations LC-L93K and HC-F98P.

Without being bound to theory, two scenarios could explain the frequent presence of a particular mutation in enriched mutation pairs. In the first scenario, the mutation provides a specific improvement of the antigen binding interface and the resulting higher affinity is further improved or not diminished by various mutation partners, while in the second scenario, the mutation exhibits an advantage in fitness other than affinity during phage selection. For HC-F97P, we suspect the latter reason, because its mutation partners are mostly spatially distant, and, because when 5A12 4.2 HC-F97P affinity was tested in phage competition ELISA, no change was detected compared to 5A12 4.2. The strong selection of HC-F97P could therefore be caused by improved functional folding in E. co//^' for clones carrying HC-F97P, resulting in improved functional phage display and higher enrichment during panning.

We used a modified cooperativity analysis to predict if HC-F97P and other mutations showing a similar behavior improved the stability of the 5A12 4.2 fold (see Araya et al. Proc. Natl. Acad. Sci. USA 109(42) : 16858-16863, 2012). Using a multiplicative cooperativity model, we calculated partner potentiation scores for a large subset of all possible mutation pairs from the data sets obtained from Ang2 panning. Indeed, HC-F97P, HC-F98P and LC-L93K had the highest partner potentiation score (>20) in the Ang2-panned dataset (Figures 22C and 22D). We confirmed that all three single variants showed an elevated melting temperature compared with 5A12 4.2 using differential scanning fluorimetry. Therefore, the partner potentiation score can be used to avoid selecting mutation pairs which contain a mutation more likely to be involved in protein folding and/or stability than in binding affinity. Alternatively, if the engineering goal is enhanced protein folding and/or stability (e.g., thermal stability), the partner potentiation score can be used to identify candidate mutation pairs for further testing.

In addition to enrichment and partner potentiation score, we also applied a third measure to identify affinity-improving mutation pairs: the spatial distance between the two mutations. Using distance as an additional criterion, we sought to identify pairs which had strong cooperativity in terms of their binding function, and thus pairs that would typically not be selected from the single mutation analysis. Without wishing to be bound by theory, strong binding cooperativity could involve structural changes in conformation and/or dynamics that could be more likely to occur between adjacent residues (see, e.g., Skinner et al. Proc. Natl. Acad. Sci. USA 93(20) :10753-10757, 1 996). All tested mutations had a spatial Ca-Ca distance of 5 A or less.

Based on the preceding information, we developed a method to identify mutation pairs that act at least additively to improve dual affinity. We then calculated a partner potentiation score (Araya et al. Proc. Natl. Acad. Sci. USA 109(42) : 16858-1 6863, 201 ) for all mutation pairs in the dataset to identify fold-stabilizing mutations in silico. Mutation pairs which contained a mutation that exhibited a large partner potentiation score were then removed from the double mutation dataset. Since mutation pairs from the Ang2 panning of the heavy and light chain library showed the strongest enriched mutation pairs, three heavy chain mutation pairs (HC-W33P/I51 G, HC-S30M/I51 W, HC-H35D/G50K) and four light chain mutation pairs (LC-Q89H/V33L, LC-Q89Y/V33L, LC-Q89Y/A34M, LC-Q89W/A34G and LC-Q89H/A34S) from Ang2 panning were selected for further testing (Table 7).

Table 7: Enrichment of Mutation Pairs from Ang2 Panning of 3NNK Libraries

These mutation pair variants and their respective single mutation variants were generated for determination of binding affinity by BIACORE® SPR (Figures 22E and 22F and Table 10). All selected double mutations showed an increase in Ang2 affinity over 5A12 4.2, confirming the validity of this approach to identify double mutations with improved binding function. Surprisingly, among the selected double mutations was HC-H35D/G50K, a pair that showed a strong increase in Ang2 binding in the context of the double mutant, while the individual single mutations had a strong negative impact on Ang2 binding function. Three of the selected pairs (LC-Q89Y/V33L, LC-Q89W/A34G, and LC-Q89H/A34S) also showed an improvement in VEGF binding. Each of these mutation pairs involved two adjacent positions at the stem of the HVR-L1 and HVR-L3 loops, suggesting that improved packing in this region had a strong influence on dual specificity.

Therefore, the deep mutagenesis scanning approach using 3NNK libraries identified mutation pairs that improved binding affinity of an antibody for its antigen. In the context of dual specific antibodies, this approach also identified mutation pairs compatible with improved binding affinity for one or both antigens.

E. Generation and characterization of 5 A 12 4.2 variants with sub-nanomolar affinity to both antigens based on deep mutagenesis scanning data

To affinity mature 5A12 4.2 to sub-nanomolar affinity for both VEGF and Ang2, we selected several amino acid alterations in the heavy and light chain of 5A12 identified by the deep mutagenesis scanning described above. Based on their enrichment, 14 mutations at 1 0 heavy chain HVR positions (T28P, S30M, H35Y/W, T52D, A53D, G54E, T57E, F97P, Y100aW, A100bT/V/S) as well as 12 mutations at seven light chain HVR positions (S30aT, V33L, A34M/N, S52R/K, Q89H/W, L92S, L93K/V/I) were chosen for further testing (Figures 20A-20B), either as single mutations or in combination with other mutations, as described below. The mutation pairs described above in Section D of this Example were also characterized as part of this analysis. These variants were generated and expressed in Fab format for further characterization. These variants were also compared with clones 5A12 4.2.16 and 5A12 4.2.16.2 (also referred to as "DAF16.2"), which were obtained by affinity maturation of 5A12 4.2 using phage display and selection without use of deep sequencing-based enrichment data (see Section F. ll below).

The affinity for Ang2 and VEGF for single and double mutations identified by deep sequencing as determined by phage IC50 is shown in Table 8. The affinity for Ang2 and VEGF for single and double mutations identified by deep sequencing as determined by BIACORE® SPR is shown in Table 9. The affinity as determined by BIACORE® SPR for Ang2 and VEGF for double mutation pairs identified as described above in Section D is shown in Table 10 (note that Table 10 shows results from two separate experiments for selected clones for Ang2 binding affinity). To generate higher affinity clones, a stepwise approach was chosen. First, several two- and three-mutation combinations of single mutations were generated and tested for the Ang2 and VEGF affinity (Table 8 and 9).

Table 8: Phage IC50 Values for 5A12 4.2 Affinity Matured Variant Clones

LC.A34N 221 0.01 0.37 18.65

LC.Q89H 0.138 14.28 0.6 11.50

LC.L92S 27.68 0.07 5.06 1.36

LC.L93K 0.48 4.10 4.21 1.64

LC.Q89W 2.55 0.77 6.2 1.11

HC.T52D 39.5 0.05 0.08 86.25

HC.A53D 0.05 39.40 67 0.10

HC.G54E 24.2 0.08 0.43 16.05

HC.F95Y 11.9 0.17 160 0.04

HC.YIOOaW 0.21 9.38 115 0.06

HC.AIOObT 0.49 4.02 69 0.10

HC.AIOObV 0.1 19.70 137 0.05

HC.AIOObS 1.2 1.64 71.4 0.10

HC.F97P 160 0.01 6.85 1.01

LC.Q89H/LC.L93K("D") 3.71 0.53 0.1 69.00

LC.Q89H/HC.A100bV 5.2 0.38 22.4 0.31

LC.L93K/HC.A100bV 11.3 0.17 6.18 1.12

LC.Q89H/LC.L93K/HC.A1 OObV 51.3 0.04 0.2 34.50

HC.A53D/LC.L93K 0.015 131.33 21 0.33

HC.A53D/LC.Q89H 0.01 197.00 1 6.90

HC.A53D/LC.Q89H/LC.L93K 0.03 65.67 0.04 172.50

("T")

Table 9: BIACORE® SPR Kd Values for selected 5A124.2 Affinity Matured Clones

LC.Q89H/LC.L93K ("D") 0.675 0.259 n/a

HC.A53D/LC.Q89H/LC.L93K ("T") 0.21 0.5 71 .7 n/a, not available

Table 10: BIACORE® SPR Kd Values (nM) of mutation pairs identified by deep mutagenesis scanning

The two highest dual affinity variants LC-Q89H/LC-L93K (referred to as "D" for simplicity) and HC-A53D/LC-Q89H/LC-L93K (referred to as "T" for simplicity) were then further dual affinity matured by the addition of several further mutations. Table 1 1 shows phage IC50 values and the fold difference relative to 5A12 4.2 for these variants. Table 12 shows the binding affinity (Kd) as determined by BIACORE® SPR for Ang2 and VEGF for selected variants. Table 11 : Phage IC50 values of combination variants based on variants D or T

Table 12: BIACORE® SPR Kd Values (nM) of clones derived from T or D (values ± standard deviation)

The improved variants T.28P, T.30M, and T.28P-VR further carried the HVR-H2 A53D mutation and an additional mutation in HVR-L3 (LC-L93K or L93V) (see Figures 29A-29B). Except from LC-L93K, which slightly increased the melting temperature, we did not include any additional fold-stabilizing mutants from the cooperativity analysis, as they were not compatible with dual specificity. These improved 5A12 4.2 variants showed improved dual specificity, having sub-nanomolar affinity for Ang2 as well as VEGF as measured by SPR (BIACORE®) (Figure 23B and Table 12). These improved variants were tested for their ability to block the ligand-receptor interaction using in vitro receptor blocking assays. Each of these affinity improved variants showed improved blocking activity over the parental antibody 5A12 4.2. For example, the T.28P variant a similar blocking activity as the high affinity anti-VEGF antibody G6.31 and an in-house generated high affinity anti-Ang2 antibody G5.5 (Figures 23C-23D). These variants also had better dual-specificity and increased potency in blocking VEGF-induced migration in a cell-based assay (HUVEC migration assay) than an antibody clone obtained by phage display and selection without use of deep mutagenesis scanning (5A12 4.2.16.2) (Figure 24).

Dual targeting strategies for disorders associated with pathological angiogenesis (e.g., ocular diseases such as AMD as well as cell proliferative disorders such as cancer) promise treatments with improved efficacy. Designing such dual targeting therapies holds several challenges, including the design of high affinity blockers against two targets and determination of the most efficacious ratio of the two blocking antibodies for administration. Additionally, for some conditions, including ocular disorders such as AMD, the amount of therapeutic antibody that can be delivered (e.g., by intravitreal injection) is limited. These challenges can be overcome using high affinity DAFs, which combine dual-targeting activity within a small antibody fragment. The results presented in this Example demonstrated the ability of the deep scanning mutagenesis approach described herein to identify variants that allowed engineering, for the first time, an antibody with a dual-specific binding site having sub-nanomolar affinity against two structurally-distinct antigens.

The optimized library design as well as the deep mutagenesis scanning approach described herein can also be applied for the improvement of monospecific antibodies. For example, as

demonstrated above, the library design facilitates the identification of fold-stabilizing mutations in the HVRs which are compatible with antigen binding, allowing the optimization of HVRs for improved antibody thermostability. The identification of affinity-improving mutation pairs by the methods described herein can allow, for example, achieving a high monospecific affinity of an antibody in cases where the single mutation space has been already extensively been explored unsuccessfully.

F. Materials and Methods

I. 5A 12 4.2 Library Construction, Phage Display, and Panning

For additional affinity improvement of 5A12 4.2, phage libraries were constructed from variant 5A12 4.2 in Fab-amber format for monovalent Fab phage display with either light chain HVR residues (i.e., HVR-L1 , HVR-L2, and HVR-L3) or heavy chain HVR residues (i.e., HVR-H 1 , HVR-H2 and HVR-H3) residues randomized using the A/A/ degenerate codon that encodes for all 20 amino acids with 32 codons (see, e.g., Brenner et al. Proc. Natl. Acad. Sci. USA 89(12) : 5381 -5383, 1992). Libraries were designed to allow either one NNK mutation in each of the three light chain or heavy chain HVRs (3NNK) or only one NNK mutation in a single heavy chain HVR (1 NNK). The resultant library DNA was electroporated into E. coli XLI cells, yielding approximately 10⁹ transformants.

Libraries were panned on immobilized hVEGF_{1 0}9 or hAng2his8 with subsequent rounds using biotinylated targets (Ang2 or VEGF) in solution and captured on streptavidin-coated plates. See Figure 25A for a schematic of the panning strategy. //. Isolation of affinity-matured variant 5A 12 4.2. 16.2 by phage display and selection

To directly screen for and isolate high affinity clones from the phage libraries, a high-throughput single-spot competitive ELISA was used to screen for high affinity clones for both hAng2 and hVEGF binding (Sidhu et al. J. Mol. Biol. 338(2) : 299-310). Selected clones were submitted for standard Sanger sequencing and clones with unique sequences were ranked by their apparent affinities using phage IC50 (Figures 25B-25D). Selected top variants (clones 5, 9, and 16) were expressed and purified as Fab for further testing in receptor-blocking ELISA experiments (Figures 26A-26C). Based on the results of the phage IC50 and the receptor blocking assays, clone 16 (5A12 4.2.1 6) was selected as a template antibody for further affinity maturation.

A second round of affinity maturation was conducted using NNK libraries based on 5A12 4.2.16 in Fab format for monovalent phage display as described above for 5A12 4.2, allowing one NNK mutation in each of the three heavy chain variable region HVRs. Panning and screening of the 5A12 4.2.16 heavy chain A/A/ libraries was performed as described above for 5A12 4.2 NN K libraries. Selected clones were sequenced by Sanger sequencing and phage IC50 affinities were determined (Figures 27A-27C).

Selected top variants (clones 2, 4, 5, 8) were expressed and purified as Fab for affinity measurements by BIACORE® using Fabs as analytes with a CM5 sensor chip immobilized with hVEGF₁₀₉ or hAng2his8 at 25 Ό to determine monovalent affinities (Figure 1 9). Clone 2 (5A12 4.2.16.2) showed best dual affinity for both VEGF and Ang2, with a Kd below 0.003 nM for VEGF and a Kd of 1 .51 nM for Ang2 (Figure 28). DAF 5A12 4.2.16.2 (also referred to as DAF16.2) was chosen as a lead variant based on affinity measurements by BIACORE® SPR and a VEGF-cell based assay potency described below. ///. Deep sequencing of DAF 5A 124.2 Affinity Maturation Libraries

For deep sequencing, phagemid double stranded DNA was isolated from E. coli XL-l cells carrying phagemid vectors from selected rounds of DAF 5A12 4.2 NN K library panning described in the in Section F. I above. The unsorted library and Round 3 of the Ang2 sorted libraries and Round 2 of the VEGF sorted libraries were sequenced. These rounds were selected to allow for both a plate-based panning round, which typically removes non-functional clones from the generated library, as well as a solution-based off rate selection, which typically enriches for clones with high affinity. For VEGF, only one round of plate panning was performed, while for Ang2, two rounds were performed because the observed enrichment was comparatively low for the first round.

Purified DNA was used as template for a limited cycle PCR-based amplification of VL and VH regions using PHUSION® DNA polymerase (New England Biolabs). PCR products were purified by agarose gel extraction and clean-up (Qiagen Gel Extraction Kit). Eluted amplicon DNA was used as the basis for deep sequencing library preparation with standard lllumina library preparation methods, using a TRUSEQ™ DNA Sample Prep kit (lllumina). Adapter-ligated libraries were subjected to a single cycle of PCR and sequenced on the lllumina M ISEQ®, using paired-end sequencing with an insert size of 200bp or 300bp as appropriate to cover the entire length of the amplicon.

IV. Deep Sequencing Analysis of DAF 5A 12 4.2 Affinity Maturation Libraries

Sequencing data were analyzed using the statistical programming language R (see, e.g., R Core Team , R: A language and environment for statistical computing, 2013) and the ShortRead package (see Morgan et al. Bioinformatics 25(19) : 2607-2608). Quality control (QC) was performed on identified HVR sequences, where each HVR sequence was checked for the correct length and was allowed to carry only up to one NNK mutation and no ηοη-Λ/Λ/Κ^" mutation. The obtained sequence counts for all samples before and after quality control are listed below in Table 13. Position weight matrices were generated by calculating the frequency of all mutations of every randomized position. Enrichment ratios for each mutation were calculated by dividing the frequency of a given mutation at a given position in the sorted sample with the frequency of the very same mutation in the unsorted sample, as described previously (Fowler et al. Nature Methods 7(9) : 741 -746, 2010). The log2 enrichment ratio for all randomized positions is shown in Figures 16A-16B.

Table 13: Sequence Counts for Deep Sequencing

Ang2 HC-3NNK 4838445 2740029

VEGF HC-3NNK 4721 529 2398256

unsorted HC-1 NNK HVR-H1 279535 83312 168/168

Ang2 HC-1 NNK HVR-H1 1441 03 92573

VEGF HC-1 NNK HVR-H1 250058 1 15401

unsorted HC-1 NNK HVR-H2 5521 89 168783 231 /231

Ang2 HC-1 NNK HVR-H2 287387 188498

VEGF HC-1 NNK HVR-H2 491958 185947

unsorted HC-1 NNK HVR-H3 1581 86 66989 210/210

Ang2 HC-1 NNK HVR-H3 390433 256998

VEGF HC-1 NNK HVR-H3 467303 229836

The enrichment ratios of double mutations were calculated from the 3NNK library by calculating the enrichment ratio of all clones which carry NNK mutations at two given positions, ignoring the third NNK mutation. In order to filter out sampling effects, mutation pairs which had less than 10 sequence counts either in the sorted or unsorted sample were removed from the analysis.

To test the performance of the ER as a classifier for identifying affinity-improving mutations, receiver operator characteristics were used. The performance of single variants in a phage competition ELISA was used as a gold standard (IC₅₀ mutant IC₅o _Wiidt_yPe > -5 for affinity improving mutations, IC₅₀ mutant IC₅₀ wiidtype < 1 -5 for mutations which do not improve the affinity).

Four different models describing cooperativity were tested (Mani et al. Proc. Natl. Acad. Sci. USA

105(9) : 3461 -3466, 2008) to determine which model is the best representative for the data.

The multiplicative model

EnrichAB = EnrichA x EnrichB

was chosen based on a simple linear regression.

The cooperativity used here is thus defined as:

Cooperativity = EnrichAB - EnrichAx EnrichB

The partner potentiation score to identify fold-stabilizing mutations are calculated as described previously (Ref) by calculating first the partner normalized cooperativity score for all available mutation pairs:

_ EnrichAB - EnrichA x EnrichB

a^→b EnrichB

The partner potentiation score is the mean of the partner normalized cooperativity scores of a given mutation mutation A:

P„ _ i=l

n

The data were plotted using ggplot2 (Wickam, ggplot2: elegant graphics for data analysis (Springer New York, 2009) and Circos (Krzywinski et al. Genome Research 19(9) :1639-1645, 2009). V. Phage IC50 determination of deep sequencing variants

Variants based on the deep sequencing information were generated by Kunkel mutagenesis on a 5A12.4.2 single strand Fab amber phagemid template. Clones were sequenced to confirm the mutations.

To determine the phage IC50 for deep sequencing variants, we performed phage solution competition ELISAs. First, the phage were propagated and purified. 10 μΙ of XL-1 bacteria infected with one of the clones for 30 min at 37Ό was plated on a carbenicillin plate. A colony was picked and grown in 2 ml of 2YT media containing 50 μg/ml carbenicillin at 37°C for 3-4 h. Helper phage K07 was added to the culture at a final concentration of 10¹⁰ plaque-forming units (pfu)/ml for another 1 h at 37°C. 20 ml of 2YT media with 50 μg/ml carbenicillin and 50 μg/ml kanamycin were added to the culture for growth overnight at 37°C. The phage was purified as described above.

Second, the concentration of purified phage that would be optimal for use in the following competition ELISA assay was determined (i.e., approximately 90% of maximal binding capacity on the coated plate). 96-well NUNC MAXISORP® plates were coated with 2 μg/ml hVEGF₁₀₉ or hAng2his8 in PBS at 4Ό overnight. The wells were blocked by adding 65 μΙ 1 % BSA for 30 min followed by 40 μΙ of 1 % TWEEN-20® for another 30 minutes. Next, the wells were washed with PBS with 0.05% TWEEN-20® 5 times. Various dilutions of phages down to 0.1 optical density (O.D.)/ml in ELISA buffer (PBS with 0.5% BSA and 0.05% TWEEN-20®) were added to the wells for 15 min at room temperature. The wells were then washed with PBS with 0.05% TWEEN-20® at least three times. 75 μΙ of HRP-conjugated anti-M13 antibody (Amersham , 1 :5000 dilution in ELISA buffer) per well was added and incubated at room temperature for 30 min. The wells were washed again with PBS with 0.05% TWEEN-20® at least five times. Next, 100 μΙ/well of a 1 :1 ratio of 3,3',5,5'-tetramethylbenzidine (TMB) Peroxidase substrate and Peroxidase Solution B (H₂0₂) (Kirkegaard-Perry Laboratories (Gaithersburg, MD)) was added to the well and incubated for 5 min at room temperature. The reaction was stopped by adding 100 μΙ of 1 M phosphoric acid (H₃P0₄) to each well and allowed to incubate for 5 min at room temperature. The optical density of the color in each well was determined using a standard ELISA plate reader at 450 nm . The dilutions of phage were plotted against the O.D. readings.

Third, a competition ELISA was performed. 96-well NUNC MAXISORP® plates were coated with 2 μg/ml hVEGF₁₀₉ or hAng2his8 in PBS at 4°C overnight. The wells were blocked by adding 65 μΙ of 1 % BSA for 30 min followed by 40 μΙ of 1 % TWEEN-20® for another 30 min. The wells were washed 5 times with PBS with 0.05% TWEEN-20®. Based on the binding assay above, 50 μΙ of the dilution of phage that resulted in about 90% of maximum binding to the coated plate was incubated with 50 μΙ of various concentrations of hVEGF₁₀₉ or hAng2his8 (0.1 to 10OOnM) in ELISA buffer solution for 2 h at room temperature in a well. The unbound phage was assayed by transferring 75 μΙ of the well mixture to a second 96-well plate pre-coated with hVEGF₁₀₉ or hAng2his8 and incubating at room temperature for 15 min. The wells of the second plate were washed with PBS with 0.5% TWEEN-20® at least three times. 75 μΙ of HRP-conjugated anti-M13 antibody (1 :5000 dilution with ELISA buffer) per well was added and incubated at room temperature for 30 min. The wells were washed again with PBS with 0.05% TWEEN- 20® at least five times. Next, 100 μΙ/well of a 1 :1 ratio of TMB Peroxidase substrate and Peroxidase Solution B (H₂0₂) (Kirkegaard-Perry Laboratories) was added to the well and incubated for 5 min at room temperature. The reaction was stopped by adding 100 μΙ of 1 M phosphoric acid (H₃P0₄) to each well and allowed to incubate for 5 min at room temperature. The O.D. of the color in each well was determined using a standard ELISA plate reader at 450 nm . The concentrations of competitor hVEGF₁₀₉ or hAng2his8 were plotted against the O.D. readings. The phage IC50 is the concentration of hVEGF₁₀₉ or hAng2his that inhibits 50% of the phage.

VI. Expression and purification of Fab constructs of deep sequencing variants

Based on phage IC50, we chose variants with high affinity to both hVEGF and hAng2 to make Fab proteins for BIACORE® SPR affinity determination. The amber phagemid of selected variants were transformed into E coli non-amber-suppressor strain 34B8 cells. Single colonies were picked and grown in complete CRAP medium with 25 μ9/ιτιΙ carbenicillin at 30°C for at least 22 h. The expressed Fab proteins were purified using a protein G high trap column with standard purification protocol (Amersham Pharmacia). VII. Fab affinity determination by BIA CORE® SPR

To determine the binding affinity of selected Fab variants for VEGF, SPR measurement with a BIACORE® T200 instrument was performed. Briefly, a series S sensor chip CM5 was activated with 1 - ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) and N-hydroxysuccinimide (NHS) reagents according to the supplier's instructions, and hVEGF₁₀₉ was coupled to achieve 50-80 response units (RU), then following by blocking un-reacted groups with 1 M ethanolamine. 3-fold serial dilutions of Fab in HBS-P buffer (0.01 M HEPES pH 7.4, 0.15 M NaCI, 0.005% surfactant P20) from low (0.02 nM) to high (50 nM) were injected (flow rate: 30μΙ/Γη ίη) . The binding responses on hVEGF were corrected by subtracting of RU from a blank flow cell. The sensorgram was recorded and subject to reference and buffer subtraction before evaluating by BIACORE® T200 Evaluation Software (version 2.0). Association rates (k_on) and dissociation rates (k_off) were calculated using a simple one-to-one Langmuir binding model. The equilibrium dissociation constant (Kd) was calculated as the ratio of k_off/k_on.

To determine the binding affinity of selected Fab variants for Ang2 or Ang1 , SPR measurement with a BIACORE® T200 instrument was performed. Briefly, a series S sensor chip CM5 was activated with EDC and NHS reagents according to the supplier's instructions, and anti-human Fc was coupled to achieve 10000 response units (RU), then following by blocking un-reacted groups with 1 M ethanolamine. For kinetics measurements, approximately 5 nM of human Fc fusion of Ang2 or Ang1 protein was first injected at 10 μΙ/min flow rate to capture approximately 100 RU at 2 different flow cells (FC), except for FC1 (which served as a reference). Next, 3-fold serial dilutions of Fab in HBS-P buffer (0.01 M H EPES pH 7.4, 0.15M NaCI, 0.005% surfactant P20) from low (0.02 nM) to high (50 nM) were injected (flow rate: 30 μΙ/min). The binding responses on hAng2 or hAngl flow cell were corrected by subtracting of RU from a blank flow cell. The sensorgram was recorded and subject to reference and buffer subtraction before evaluating by BIACORE® T200 Evaluation Software (version 2.0). Association rates (k_on) and dissociation rates (k_off) were calculated using a simple one-to-one Langmuir binding model. The equilibrium dissociation constant (Kd) was calculated as the ratio of k_off/k_on. VIII. Differential Scanning Fluorimetry

The melting temperature (T_m) was determined using differential scanning fluorimetry. A 100 μ9/ιηΙ purified Fab solution was mixed 1 :500 with SYPRO® Orange dye. The sample was slowly heated in a qPCR machine from 20 °C to 100°C while the fluorescence was measured at 590 nm at 0.2 °C temperature increments. The T_m is the maximum of the derivative of the obtained fluorescence melting curve.

IX. HUVEC Migration Assays

HUVEC migration assays were performed by pre-coating Falcon 24-multiwell insert systems with mouse laminin overnight and adding HUVEC cells that had been starved overnight, harvested, and reuspended in assay medium . Cells (5x10⁴) were added to the upper chamber and 20 ng/mL VEGF was added to the lower chamber to stimulate cell migration in the presence or absence of antibodies for 16 h. After fixing and scraping the upper membrane face, cells on the lower face were fixed with methanol and stained with SYTOX® Green and images were acquired using an inverted fluorescent microscope. Cell numbers were analyzed with ImageJ software and IC50 was calculated. HUVEC IC50 migration values indicate the cell activity potency of anti-VEGF antibodies on the VEGF-induced migration of these cells.

Example 5: DAF T.28P-VR shows in vivo efficacy in a preclinical model

The in vivo efficacy of the dual affinity improved DAF T.28P-VR was tested in a laser-induced choroidal neovascularization (CNV) rat model. A comparison of the size of the area in which vascular sprouting occurred after the laser-induced lesion as well as after administration of the antibodies demonstrated that dual targeting of Ang2 and VEGF resulted in reduced vascular sprouting compared to treatment with either a high-affinity anti-Ang2 Fab or a high-affinity anti-VEGF Fab alone.

Two different formats of T.28P-VR were generated. One format was a standard dual-specific Fab format as described above, while the second was a dimeric F(ab')₂ format in which two monomeric DAF molecules were linked together by a single cysteine bond. While the monomeric DAF format was used to test if the efficacy of co-administered mono-specific anti-Ang2 and anti-VEGF can be reached by administration of the DAF, the dimeric DAF allowed determination of whether additional avidity led to improved efficacy. T.28P-VR was compared with either co-administered anti-Ang2 and anti-VEGF Fabs or a bi-specific F(ab')₂ format in which the anti-VEGF and the anti-Ang2 Fabs were linked together by a single cysteine bond. Monomeric T.28P-VR DAF showed the same efficacy as the two co-administered high affinity monospecifc antibodies (Figure 30). However, the dimeric DAF showed a significantly greater reduction of vessel sprouting compared to the co-administered Fab or the bi-specific Fab molecule (Figure 30). This efficacy study demonstrates the feasibility of using one high affinity DAF molecule to block the activity of two soluble factors in vivo as a therapy for disorders associated with pathological angiogenesis. Materials and Methods

The laser-induced rat choroidal neovascularization model was performed using laser

photocoagulation in Brown Norway rats under anesthesia. Pupils were dilated with 1 % tropicamide and then 6 laser spots were delivered unilaterally using an OCULIGHT® CL photocoagulator with lesions approximately 2 disc diameters from the optic nerve head with evident rupture of Bruch's membrane and lack of sub-retinal hemorrhaging. Buprenorphine was administered immediately prior to photocoagulation and q12h for 48 h. Experimental designs varied in the amount and dosage frequently of intravitreal injections following laser coagulation. Treatment groups of n=6 rats of anti-VEGF, anti-VEGF/Ang2, and non-targeting anti-gD or ragweed IgG as negative controls were included and 5 μΙ volumes were delivered via 30G Hamilton needles. At time of harvest, animals were injected with 1 m l_ of 25 mg/mL fluroscein dextran via tail vein and then animals were euthanized. The eyes were harvested and fixed with 4% paraformaldehyde in PBS for an hour prior to removal or corea, lens, and retina to flat-mount the entire retinal pigment epithelium with choroid and with eye-cup. The CNV was imaged with a Zeiss Axio Image M2 microscope and captured and processed with analysis using N IH ImageJ software. The CNV was isolated with freehand selection and the resultant pixel amount converted to area by a pixel/length ratio. Statistical analysis used the "all pairs, Tukey HSD" method within JMP9 software.

Example 6: Anti-Ang2/anti-VEGF DAF T.28P affinity maturation by phage display and deep scanning mutagenesis to generate variants with reduced off-target Ang1 binding. A. DAF 5A 12 4.2 T28P Affinity Maturation Library Construction and Panning

Affinity maturated variant T.28P had a detectable off-target Ang1 binding affinity. For additional affinity improvement to VEGF and Ang2, and reduction of affinity for Ang1 , phage libraries were constructed from variant T.28P in Fab-amber format for monovalent Fab phage display with either LC (L1 , L2, and L3) or HC (H1 , H2, and H3) HVR residues mutated using the A/A/ degenerate codon. Libraries were designed to allow one NNK mutation in each of the three LC or HC HVRs (3NNK). The resulting library DNA was electroporated into E. coliXLI cells, yielding approximately 1 0⁹ transformants. Libraries were panned on immobilized hVEGF₁₀₉, hAng2his, or hAngl .Fc, with subsequent rounds using biotinylated targets (Ang2 or VEGF) in solution and captured on streptavidin-coated plates or target directly immobilized on plates (hAngl ). B. Deep sequencing of DAF 5A 124.2 T28P Affinity Maturation Libraries

For deep sequencing, phagemid double strand DNA was isolated from E. coli XLI cells carrying phagemid vectors from selected rounds of DAF T.28P NNK library panning. Purified DNA was used as template for a limited cycle PCR-based amplification of V_L and V_H regions. PCR products were purified by agarose gel extraction and clean-up (Qiagen Gel Extraction Kit). Eluted amplicon DNA was used as basis for deep sequencing library preparation with standard lllumina library prep methods, using

TRUSEQ™ DNA Sample Prep (lllumina). Adapter-ligated libraries were subjected to a single cycle of PCR and sequenced on the lllumina M ISEQ®, paired-end 200bp or 300bp as appropriate to cover the entire length of the amplicon. C. Deep Sequencing Analysis of DAF 5A 12 4.2 T28P Affinity Maturation Libraries

Sequencing data were analyzed using the statistical programming language R and the

ShortRead package. Quality control was performed on identified HVR sequences, where each HVR sequence was checked for the correct length and was allowed to carry only up to one NNK mutation and no non-WMC mutation. The obtained sequence counts for all samples before and after quality control are listed in Table 14. Position weight matrices were generated by calculating the frequency of all mutations at every randomized position. Enrichment ratios for all mutations were calculated by dividing the frequency of a given mutation at a given position in the sorted sample with the frequency of the very same mutation in the unsorted sample, as described previously (Fowler et al. Nature Methods 7(9) :741 - 746, 201 0). To identify single mutations which interfered with Ang1 binding while having a minimal impact on Ang2 or VEGF binding, the following filter was applied to all combinations of Ang1 , Ang2 and VEGF panned samples: lOg 2ERx, Angl < -1 Λ lOg 2ERx, Ang 2 > -0.5 Λ log lERx, VEGF > 0 where ERx.y is the enrichment of mutation x in panning against antigen y. Since we panned each antigen using two different panning conditions (see Table 14), there are 8 different combinations of how pannings against different antigens can be combined and filtered. Mutations which passed the filter more than once in the different combinations were selected for further characterization.

Table 14: Overview of number of sequence reads obtained from MISEQ® sequencing of the different samples and their sequence counts after QC

D. Clone generation and purification

Variant plasmids based on the deep sequencing information were generated by synthesizing the selected DNA inserts containing the single muations and subcloning either the VL or VH into a human kappa or human Fab.flag vector, respectively. Clone plasmids (LC and HC) were transiently transfected into Expi293 cells in a 30ml_ volume and resultant supernatants were harvested after 7 days of growth. Clone antibody fragments (Fab) were purified using FLAG™ tag-based affinity chromatography.

E. Characterization of deep sequencing variants with single mutation affinity by BIACORE® SPR To determine the binding affinity of selected Fab variants, SPR measurement with a BIACORE® T200 instrument was used. Briefly, series S sensor chip CM5 was activated with EDC and NHS reagents according to the supplier's instructions, and hVEGF was coupled to achieve 50-80 response units (RU), then following by blocking un-reacted groups with 1 M ethanolamine.

Using a single-cycle kinetic method, 5-fold serial dilutions of Fab in HBS-P buffer (0.01 M H EPES pH 7.4, 0.15M NaCI, 0.005% surfactant P20) from low (0.08 nM) to high (50 nM) were injected (flow rate: 30 μΙ/min) over the hVEGF-coated flow cells. The binding responses (RU) on hVEGF were corrected by subtracting of RU from a blank flow cell. The sensorgram was recorded and subject to reference and buffer subtraction before evaluating by BIACORE® T200 Evaluation Software (version 2.0). Association rates (k_on) and dissociation rates (k_off) were calculated using a simple one-to-one Langmuir binding model. The equilibrium dissociation constant (Kd) was calculated as the ratio k_off/k_on.

F. Binding kinetic determination to human Ang2 and Ang1

To determine the binding affinity of selected Fab variants, SPR measurement with a BIACORE® T200 instrument was used. Briefly, series S sensor chip CM5 was activated with EDC and NHS reagents according to the supplier's instructions, and anti-human Fc was coupled to achieve 10000 response units (RU), then following by blocking un-reacted groups with 1 M ethanolamine.

For kinetic measurements, approximately 5 nM of human Fc fusion of Ang2 or Ang1 protein was first injected at 10 μΙ/min flow rate to capture approximately 100 RU on 2 different flow cells (FC), except for FC1 (reference). 5-fold serial dilutions of Fab in HBS-P buffer (0.01 M H EPES pH 7.4, 0.15 M NaCI, 0.005% surfactant P20) from low (1 .6 nM) to high (1000 nM) were injected (flow rate: 30 μΙ/min). The binding responses (RU) on hAng2 or hAngl flow cells were corrected by subtracting of RU from a blank flow cell. The sensorgram was recorded and subject to reference and buffer subtraction before evaluating by BIACORE®T200 Evaluation Software (version 2.0). Association rates (k_on) and dissociation rates (k_off) were calculated using a simple one-to-one Langmuir binding model. The equilibrium dissociation constant (Kd) was calculated as the ratio k /k

E. Binding affinity of clones identified by deep mutagenesis scanning

Using the methods described above, single mutation variants of T.28P were identified by deep mutagenesis scanning based on panning against VEGF, Ang2, and Ang1 . The resulting variants maintained dual affinity for VEGF and Ang2, while having markedly reduced affinity to Ang1 (Table 15).

The SPR measurements indicated that for each of the single mutation variants, the Kd was greater than

1 μΜ. Table 15: Affinity to VEGF, Ang2, and Ang1 of single mutation variants of T.28P as determined by Biacore.

Other Embodiments

Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, the descriptions and examples should not be construed as limiting the scope of the invention. The disclosures of all patent and scientific literature cited herein are expressly incorporated in their entirety by reference.

Claims

WHAT IS CLAIMED IS:

1 . An isolated antibody that specifically binds angiopoietin 2 (Ang2), wherein the antibody binds to an epitope on Ang2 comprising one or more amino acid residues selected from the group consisting of Cys433, Cys435, Met440, Leu441 , Cys450, and Gly451 of Ang2.

2. The isolated antibody of claim 1 , wherein the epitope further comprises one or more additional amino acid residues selected from the group consisting of Phe469, Tyr475, and Ser480 of Ang2.

3. The isolated antibody of claim 1 or 2, wherein the epitope further comprises one or more additional amino acid residues selected from the group consisting of Lys432, Ile434, Asp448, Ala449, Pro452, and Tyr476 of Ang2.

4. The isolated antibody of claim 3, wherein the epitope consists of amino acid residues Lys432, Cys433, Ile434, Cys435, Met440, Leu441 , Asp448, Ala449, Cys450, Gly451 , Pro452, Phe469, Tyr475, Tyr476, and Ser480 of Ang2.

5. An isolated antibody that specifically binds Ang2, wherein the antibody comprises a paratope comprising one or more amino acid residues selected from the group consisting of light chain variable region amino acid residues Gln27; Phe27a; Leu28, Met28, or Ala28; Ser29; Ser30; Phe31 ; Ser67; Gly68; Gly91 ; Leu92; Leu93, Lys93, or Val93; Ser94 or Pro94; and Leu96 and the heavy chain variable region amino acid residues Trp33; His35, Tyr35, or Asp35; Tyr58, Ile58, Trp58, or Leu58; Phe97; Phe98; Leu99 or Ala99; and Tyrl OOa.

6. The isolated antibody of claim 5, wherein the paratope consists of light chain variable region amino acid residues Gln27; Phe27a; Leu28, Met28, or Ala28; Ser29; Ser30; Phe31 ; Ser67; Gly68; Gly91 ; Leu92; Leu93, Lys93, or Val93; Ser94 or Pro94; and Leu96 and the heavy chain variable region amino acid residues Trp33; His35, Tyr35, or Asp35; Tyr58, Ile58, Trp58, or Leu58; Phe97; Phe98; Leu99 or Ala99; and Tyrl OOa.

7. An isolated antibody that specifically binds Ang2, wherein the antibody comprises the following six hypervariable regions (HVRs) :

(i) an HVR-L1 comprising the amino acid sequence of RASQFX₁SX₂FGX₃X₄ (SEQ ID NO: 26), wherein is Leu, Met, or Ala, X₂ is Ser, Lys, or Thr, X₃ is Val or Leu, and X₄ is Ala, Ser, Met, Gly, Thr, or Asn;

(ii) an HVR-L2 comprising the amino acid sequence of GX₁X2X₃LX₄X5 (SEQ ID NO: 27), wherein is Ala, Ser, or Gly, X₂ is Arg, Ser, Leu, or Lys, X₃ is Ser, Ala, or Gly, X₄ is Tyr, Val, Ala, or Glu, and X₅ is Ser, Gly, or Gin; (iii) an HVR-L3 comprising the amino acid sequence of X₁QX₂X₃X₄X₅X₆LT (SEQ ID NO: 28), wherein XT is His, Gin, Phe, Trp, Tyr, or Met, X₂ is Gly, Met, or Phe, X₃ is Leu, Pro, or Ser, X₄ is Val, Leu, lie, Gly, Lys, or Arg, X₅ is Ser, His, Leu, or Pro, and X₆ is Pro or Asp;

(iv) an HVR-H1 comprising the amino acid sequence of DX₁X₂X₃X₄ (SEQ ID NO: 29), wherein is Tyr or Ala, X₂ is Trp or Pro, X₃ is lie, Met, or Gin, and X₄ is His, Tyr, Trp, or Asp;

(v) an HVR-H2 comprising the amino acid sequence of X₁X₂X₃X₄X₅X₆GX₇X₈X₉YADSVKG (SEQ ID NO: 30), wherein is Gly or Lys, X₂ is lie, Gly, or Trp, X₃ is Thr, Val, or Asp, X₄ is Pro, Leu, or Glu, X₅ is Asp, Ala, or Leu, X₆ is Gly or Glu, X₇ is Tyr or Ala, X₈ is Thr, Glu, His, or Asp; and X₉ is Tyr, Leu, Trp, lie, or Lys; and

(vi) an HVR-H3 comprising the amino acid sequence of X₁X₂X₃X₄X₅PX₆X7X₈DY (SEQ ID NO: 31 ), wherein is Phe, Tyr, or Met, X₂ is Val or Thr, X₃ is Phe or Pro, X₄ is Phe or Pro, X₅ is Leu or Ala, X₆ is Tyr or Trp; X₇ is Ala, Thr, Val, or Ser, and X₈ is Met, Tyr, or Trp.

8. The isolated antibody of claim 7, wherein the antibody comprises the following six HVRs:

(i) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVA (SEQ ID NO: 2) ;

(ii) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ;

(iii) an HVR-L3 comprising the amino acid sequence of QQGLLSPLT (SEQ ID NO: 9) ;

(iv) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ;

(v) an HVR-H2 comprising the amino acid sequence of G ITPAGG YTYYADS VKG (SEQ ID NO: 6) ; and

(vi) an HVR-H3 comprising the amino acid sequence of FVFFLPYAMDY (SEQ ID NO: 7).

9. The isolated antibody of claim 7, wherein the antibody comprises the following six HVRs:

(i) an HVR-L1 comprising the amino acid sequence of RASQFLSSFGVG (SEQ ID NO: 32) ;

(ii) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ;

(iii) an HVR-L3 comprising the amino acid sequence of WQGLLSPLT (SEQ ID NO: 33) ;

(iv) an HVR-H1 comprising the amino acid sequence of DYWIY (SEQ ID NO: 34) ;

(v) an HVR-H2 comprising the amino acid sequence of G ITPAGG YE YYADS VKG (SEQ ID NO: 35) ; and

(vi) an HVR-H3 comprising the amino acid sequence of FVFFLPYVMDY (SEQ ID NO: 36).

10. The isolated antibody of claim 7, wherein the antibody comprises the following six HVRs:

(ii) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ;

(iii) an HVR-L3 comprising the amino acid sequence of HQGLKSPLT (SEQ ID NO: 37) ;

(iv) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ;

(v) an HVR-H2 comprising the amino acid sequence of GITPDGGYTYYADSVKG (SEQ ID NO: 38) ; and

1 1 . The isolated antibody of claim 7, wherein the antibody comprises the following six HVRs:

(ii) an HVR-L2 comprising the amino acid sequence of GARSLYS (SEQ ID NO: 39) ;

(iii) an HVR-L3 comprising the amino acid sequence of HQGLVSPLT (SEQ ID NO: 40) ;

(iv) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ;

12. The isolated antibody of any one of claims 7, 10, and 1 1 , wherein the antibody further comprises the following heavy chain variable region framework regions (FRs) :

(i) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFPIS (SEQ ID NO: 41 ) ;

(ii) an FR-H2 comprising the amino acid sequence of WVRQAPGKGLEWVA (SEQ ID NO: 42) ;

(iii) an FR-H3 comprising the amino acid sequence of

RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and

(iv) an FR-H4 comprising the amino acid sequence of WGQGTLVTVSS (SEQ ID NO: 44).

13. The isolated antibody of claim 7 or 10, wherein the antibody further comprises the following heavy chain variable region framework regions (FRs) :

(i) an FR-H1 comprising the amino acid sequence of EVQLVESGGGLVQPGGSLRLSCAASGFTIM (SEQ ID NO: 45) ;

(iii) an FR-H3 comprising the amino acid sequence of

RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and

14. An isolated antibody that specifically binds Ang2, wherein the antibody comprises (a) a light chain variable region (V_L) having at least 95% sequence identity to the amino acid sequence of SEQ ID NO: 10, 1 1 , 12, 13, 14, 15, 1 6, 17, 46, 48, 51 , 78, or 79; (b) a heavy chain variable region (V_H) having at least 95% sequence identity to the amino acid sequence of SEQ ID NO: 18, 47, 49, or 50; or (c) a light chain variable region as in (a) and a heavy chain variable region as in (b).

15. The isolated antibody of claim 14, comprising a V_H sequence of SEQ ID NO: 49.

16. The isolated antibody of claim 14, comprising a V_L sequence of SEQ ID NO: 51

17. An isolated antibody that specifically binds Ang2 and vascular endothelial growth factor (VEGF), wherein the antibody binds to a region within amino acid residues 432-480 of human Ang2 polypeptide (SEQ ID NO: 1 ).

18. An isolated antibody that specifically binds Ang2 and VEGF, wherein the antibody binds to an epitope on Ang2 comprising one or more amino acid residues selected from the group consisting of Lys432, Cys433, Ile434, Cys435, Met440, Leu441 , Asp448, Ala449, Cys450, Gly451 , Pro452, Phe469, Tyr475, Tyr476, and Ser480 of Ang2.

19. The isolated antibody of claim 18, wherein the epitope comprises three or more amino acid residues selected from the group consisting of Lys432, Cys433, Ile434, Cys435, Met440, Leu441 , Asp448, Ala449, Cys450, Gly451 , Pro452, Phe469, Tyr475, Tyr476, and Ser480 of Ang2.

20. The isolated antibody of claim 19, wherein the epitope consists of amino acid residues Lys432, Cys433, Ile434, Cys435, Met440, Leu441 , Asp448, Ala449, Cys450, Gly451 , Pro452, Phe469, Tyr475, Tyr476, and Ser480 of Ang2.

21 . An isolated antibody that specifically binds Ang2 and VEGF, wherein the antibody binds to an epitope on VEGF comprising one or more amino acid residues selected from the group consisting of Phe17, Met18, Tyr21 , Gln22, Tyr25, Lys48, Asn62, Asp63, Glu64, Gly65, Leu66, Met81 , Ile83, Lys84, Pro85, His86, Gln87, Gly88, Gln89, His90, Ile91 , Lys1 01 , Glu103, Cys104, and Pro106 of human VEGF.

22. The isolated antibody of claim 21 , wherein the epitope comprises amino acid residues Phe17, Tyr21 , and Tyr25 of human VEGF.

23. The isolated antibody of claim 21 , wherein the epitope comprises amino acid residues Phe17, Ile81 , and Gln89 of human VEGF.

24. The isolated antibody of claim 22 or 23, wherein the epitope consists of Phe17, Met18, Tyr21 , Gln22, Tyr25, Lys48, Asn62, Asp63, Glu64, Gly65, Leu66, Met81 , Ile83, Lys84, Pro85, His86, Gln87, Gly88, Gln89, His90, Ile91 , Lys101 , Glu103, Cys104, and Pro106 of human VEGF.

25. An isolated antibody that specifically binds Ang2 and VEGF, wherein the antibody comprises a paratope that binds to Ang2, wherein the paratope comprises one or more amino acid residues selected from the group consisting of light chain variable region amino acid residues Gln27; Phe27a; Leu28, Met28, or Ala28; Ser29; Ser30; Phe31 ; Ser67; Gly68; Gly91 ; Leu92; Leu93, Lys93, or Val93; Ser94 or Pro94; and Leu96 and the heavy chain variable region amino acid residues Trp33; His35, Tyr35, or Asp35; Tyr58, Ile58, Trp58, or Leu58; Phe97; Phe98; Leu99 or Ala99; and Tyr1 00a.

26. The isolated antibody of claim 25, wherein the paratope consists of light chain variable region amino acid residues Gln27; Phe27a; Leu28, Met28, or Ala28; Ser29; Ser30; Phe31 ; Ser67; Gly68; Gly91 ; Leu92; Leu93, Lys93, or Val93; Ser94 or Pro94; and Leu96 and the heavy chain variable region amino acid residues Trp33; His35, Tyr35, or Asp35; Tyr58, Ile58, Trp58, or Leu58; Phe97; Phe98; Leu99 or Ala99; and TyM OOa.

27. An isolated antibody that specifically binds Ang2 and VEGF, wherein the antibody comprises a paratope that binds to VEGF, wherein the paratope comprises one or more amino acid residues selected from the group consisting of light chain variable region amino acid residues Leu28, Met28, or Ala28; Ser29; Phe31 ; Tyr49; Ser53; and Leu92 and the heavy chain variable region amino acid residues Ser30, Gly30, or Met30; Asp31 ; Tyr32 or Ala32; Trp33; Ile51 ; Thr52; Pro52a or Glu52a; Ala53 or Asp53; Gly54; Gly55; Tyr56 or Ala56; Phe95 or Met95; Val96 or Thr96; Phe97; Phe98; Leu99 or Ala99; and Tyr100a.

28. The isolated antibody of claim 27, wherein the paratope consists of light chain variable region amino acid residues Leu28, Met28, or Ala28; Ser29; Phe31 ; Tyr49; Ser53; and Leu92 and the heavy chain variable region amino acid residues Ser30, Gly30, or Met30; Asp31 ; Tyr32 or Ala32; Trp33; Ile51 ; Thr52; Pro52a or Glu52a; Ala53 or Asp53; Gly54; Gly55; Tyr56 or Ala56; Phe95 or Met95; Val96 or Thr96; Phe97; Phe98; Leu99 or Ala99; and Tyr1 00a.

29. An isolated antibody that specifically binds Ang2 and VEGF, wherein the antibody comprises a paratope that binds to VEGF and Ang2, wherein the paratope comprises one or more amino acid residues selected from the group consisting of the light chain variable region amino acid residues Ser30, Phe31 , and Leu92 and the heavy chain variable region amino acid residues Leu99 or Ala99 and Pro100.

30. The isolated antibody of claim 29, wherein the paratope consists of the light chain variable region amino acid residues Ser30, Phe31 , and Leu92 and the heavy chain variable region amino acid residues Leu99 or Ala99 and Pro100.

31 . An isolated antibody that specifically binds Ang2 and VEGF, wherein the antibody comprises the following six HVRs:

(ii) an HVR-L2 comprising the amino acid sequence of GX₁X2X₃LX₄X5 (SEQ ID NO: 27), wherein is Ala, Ser, or Gly, X₂ is Arg, Ser, Leu, or Lys, X₃ is Ser, Ala, or Gly, X₄ is Tyr, Val, Ala, or Glu, and X₅ is Ser, Gly, or Gin;

(iii) an HVR-L3 comprising the amino acid sequence of X₁QX₂X₃X₄X₅X₆LT (SEQ ID NO: 28), wherein XT is His, Gin, Phe, Trp, Tyr, or Met, X₂ is Gly, Met, or Phe, X₃ is Leu, Pro, or Ser, X₄ is Val, Leu, lie, Gly, Lys, or Arg, X₅ is Ser, His, Leu, or Pro, and X₆ is Pro or Asp; (iv) an HVR-H1 comprising the amino acid sequence of DX₁X₂X₃X₄ (SEQ ID NO: 29), is Tyr or Ala, X₂ is Trp or Pro, X₃ is lie, Met, or Gin, and X₄ is His, Tyr, Trp, or Asp;

32. The isolated antibody of claim 31 , wherein the antibody comprises the following six HVRs:

(ii) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ;

(iv) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ;

33. The isolated antibody of claim 31 , wherein the antibody comprises the following six HVRs:

(ii) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ;

(iv) an HVR-H1 comprising the amino acid sequence of DYWIY (SEQ ID NO: 34) ;

34. The isolated antibody of claim 31 , wherein the antibody comprises the following six HVRs:

(ii) an HVR-L2 comprising the amino acid sequence of GASSLYS (SEQ ID NO: 8) ;

(iv) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ;

35. The isolated antibody of claim 31 , wherein the antibody comprises the following six HVRs:

(ii) an HVR-L2 comprising the amino acid sequence of GARSLYS (SEQ ID NO: 39) ; (iii) an HVR-L3 comprising the amino acid sequence of HQGLVSPLT (SEQ ID NO: 40) ;

(iv) an HVR-H1 comprising the amino acid sequence of DYWIH (SEQ ID NO: 5) ;

36. The isolated antibody of any one of claims 31 , 34, or 35, wherein the antibody further comprises the following heavy chain variable region FRs:

(iii) an FR-H3 comprising the amino acid sequence of

RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and

37. The isolated antibody of claim 31 or 34, wherein the antibody further comprises the following heavy chain variable region FRs:

(iii) an FR-H3 comprising the amino acid sequence of

RFTISADTSKNTAYLQMNSLRAEDTAVYYCAR (SEQ ID NO: 43) ; and

38. An isolated antibody that specifically binds Ang2 and VEGF, wherein the antibody comprises (a) a V|_ having at least 95% sequence identity to the amino acid sequence of SEQ ID NO: 10, 1 1 , 12, 13, 14, 15, 1 6, 17, 46, 48, 51 , 78, or 79; (b) a V_H having at least 95% sequence identity to the amino acid sequence of SEQ ID NO: 1 8, 47, 49, or 50; or (c) a light chain variable region as in (a) and a heavy chain variable region as in (b).

39. The isolated antibody of claim 38, comprising a V_H sequence of SEQ ID NO: 49.

40. The isolated antibody of claim 38, comprising a V_L sequence of SEQ ID NO: 51 .

41 . An isolated antibody that competes for binding to Ang2 with the antibody of any one of claims 1 -

40.

42. An isolated antibody that competes for binding to Ang2 and VEGF with the antibody of any one of claims 17-40.

43. An isolated antibody that binds to the same epitope as the antibody of any one of claims 1 -40.

44. The isolated antibody of any one of claims 1 -43, wherein the antibody binds VEGF with a Kd of about 15 nM or lower and Ang2 with a Kd of about 15 nM or lower.

45. The isolated antibody of claim 44, wherein the antibody binds VEGF with a Kd of about 5 nM or lower and Ang2 with a Kd of about 5 nM or lower.

46. The isolated antibody of claim 45, wherein the antibody binds VEGF with a Kd of lower than 1 nM and Ang2 with a Kd of lower than 1 nM.

47. The isolated antibody of claim 46, wherein the antibody binds VEGF with a Kd of lower than 0.5 nM and Ang2 with a Kd of lower than 0.5 nM.

48. The isolated antibody of claim 47, wherein the antibody binds VEGF with a Kd of lower than 0.25 nM and Ang2 with a Kd of lower than 0.25 nM.

49. The isolated antibody of any one of claims 1 -48, wherein the antibody inhibits or blocks binding of Ang2 or VEGF to its receptor.

50. The isolated antibody of claim 49, wherein the antibody inhibits or blocks binding of Ang2 or VEGF to its receptor with an IC50 value of 8 nM or lower.

51 . The isolated antibody of claim 50, wherein the antibody inhibits or blocks binding of Ang2 to its receptor with an IC50 value of 50 pM to 2 nM.

52. The isolated antibody of claim 51 , wherein the antibody inhibits or blocks binding of Ang2 to its receptor with an IC50 value of 75 pM.

53. The isolated antibody of any one of claims 50-52, wherein the antibody inhibits or blocks binding of VEGF to its receptor with an IC50 value of 50 pM to 2 nM.

54. The isolated antibody of claim 53, wherein the antibody inhibits or blocks binding of VEGF to its receptor with an IC50 value of 85 nM.

55. The isolated antibody of any one of claims 1 -54, wherein the antibody binds to Ang2 with 50-fold greater affinity than to Ang1 .

56. The isolated antibody of claim 55, wherein the antibody binds to Ang2 with 75-fold greater affinity than to Ang1 .

57. The isolated antibody of claim 56, wherein the antibody binds to Ang2 with 100-fold greater affinity than to Ang1 .

58. The isolated antibody of any one of claims 1 -57, wherein the antibody is a dual-specific antibody.

59. The isolated antibody of any one of claims 1 -58, wherein the antibody is a monoclonal antibody.

60. The isolated antibody of any one of claims 1 -58, wherein the antibody is an IgG antibody.

61 . The isolated antibody of any one of claims 1 -58, wherein the antibody is an antibody fragment that specifically binds VEGF and Ang2.

62. The isolated antibody of claim 61 , wherein the antibody fragment is selected from the group consisting of Fab, single chain variable fragment (scFv), Fv, Fab', Fab'-SH, F(ab')₂, and diabody.

63. The isolated antibody of claim 62, wherein the antibody fragment is a Fab.

64. The isolated antibody of any one of claims 1 -58, wherein at least a portion of the framework sequence is a human consensus framework sequence.

65. The isolated antibody of any one of claims 1 -58, wherein the antibody is a chimeric, humanized, or fully human antibody.

66. A polynucleotide encoding an isolated antibody of any one of claims 1 -65.

67. A vector comprising the polynucleotide of claim 66.

68. A host cell comprising the vector of claim 67.

69. A method of producing the antibody of any one of claims 1 -65, the method comprising culturing a host cell that comprises the vector of claim 67 and recovering the antibody.

70. The method of claim 69, wherein the host cell is prokaryotic.

71 . The method of claim 70, wherein the host cell is Escherichia coli.

72. The method of claim 69, wherein the host cell is eukaryotic.

73. The method of claim 72, wherein the host cell is a 293 cell, a CHO cell, a yeast cell, or a plant cell.

74. A method of reducing or inhibiting angiogenesis in a subject having a disorder associated with pathological angiogenesis, comprising administering to the subject an effective amount of the antibody of any one of claims 1 -65, thereby reducing or inhibiting angiogenesis in the subject.

75. The method of claim 74, wherein the disorder associated with pathological angiogenesis is an ocular disorder or a cell proliferative disorder.

76. The method of claim 75, wherein the disorder associated with pathological angiogenesis is an ocular disorder.

77. The method of claim 76, wherein the ocular disorder is selected from the group consisting of retinopathy including proliferative diabetic retinopathy, choroidal neovascularization (CNV), age-related macular degeneration (AMD), diabetic and other ischemia-related retinopathies, diabetic macular edema (DME), pathologic myopia, von Hippel-Lindau disease, histoplasmosis of the eye, retinal vein occlusion (including central (CRVO) and branched (BRVO) forms), corneal neovascularization, retinal

neovascularization, retinopathy of prematurity (ROP), familial exudative vitreoretinopathy (FEVR), Coats' disease, Norrie Disease, Osteoporosis-Pseudoglioma Syndrome (OPPG), subconjunctival hemorrhage, and hypertensive retinopathy.

78. The method of claim 77, wherein the ocular disorder is AMD.

79. A method for treating a disorder associated with pathological angiogenesis, the method comprising administering an effective amount of the antibody of any one of claims 1 -65 to a subject in need of such treatment.

80. The method of claim 79, wherein the disorder associated with pathological angiogenesis is an ocular disorder or a cell proliferative disorder.

81 . The method of claim 80, wherein the disorder associated with pathological angiogenesis is an ocular disorder.

82. The method of claim 81 , wherein the ocular disorder is selected from the group consisting of retinopathy including proliferative diabetic retinopathy, choroidal neovascularization (CNV), age-related macular degeneration (AMD), diabetic and other ischemia-related retinopathies, diabetic macular edema (DME), pathologic myopia, von Hippel-Lindau disease, histoplasmosis of the eye, retinal vein occlusion (including central (CRVO) and branched (BRVO) forms), corneal neovascularization, retinal

83. The method of claim 82, wherein the ocular disorder is AMD

84. The method of any one of claims 74-83, further comprising administering to the subject an effective amount of a second agent, wherein the second agent is selected from the group consisting of another antibody, a chemotherapeutic agent, a cytotoxic agent, an anti-angiogenic agent, an

immunosuppressive agent, a prodrug, a cytokine, a cytokine antagonist, cytotoxic radiotherapy, a corticosteroid, an anti-emetic, a cancer vaccine, an analgesic, and a growth-inhibitory agent.

85. The method of any one of claims 74-84, wherein the antibody or antigen-binding fragment thereof is administered intravitreally, by eye drop, subcutaneously, intravenously, intramuscularly, topically, orally, transdermal^, intraperitoneally, intraorbitally, by implantation, by inhalation, intrathecal^,

intraventricularly, or intranasally.

86. The method of claim 85, wherein the administration is intravitreally.

87. The method of any one of claims 74-86, wherein the subject is human.

88. A pharmaceutical composition comprising the antibody of any one of claims 1 -65.

89. The pharmaceutical composition of claim 88, wherein the pharmaceutical composition is used for treating a disorder associated with pathological angiogenesis in a mammal.

90. The pharmaceutical composition of claim 89, wherein the disorder associated with pathological angiogenesis is an ocular disorder or a cell proliferative disorder.

91 . The pharmaceutical composition of claim 90, wherein the disorder associated with pathological angiogenesis is an ocular disorder.

92. The pharmaceutical composition of claim 91 , wherein the ocular disorder is selected from the group consisting of retinopathy including proliferative diabetic retinopathy, choroidal neovascularization (CNV), age-related macular degeneration (AMD), diabetic and other ischemia-related retinopathies, diabetic macular edema (DME), pathologic myopia, von Hippel-Lindau disease, histoplasmosis of the eye, retinal vein occlusion (including central (CRVO) and branched (BRVO) forms), corneal

neovascularization, retinal neovascularization, retinopathy of prematurity (ROP), familial exudative vitreoretinopathy (FEVR), Coats' disease, Norrie Disease, Osteoporosis-Pseudoglioma Syndrome (OPPG), subconjunctival hemorrhage, and hypertensive retinopathy.

93. The pharmaceutical composition of claim 92, wherein the ocular disorder is AMD.

94. A method of identifying an amino acid residue alteration that confers enhanced binding of an antibody to a target molecule, the method comprising:

(a) providing a display library comprising nucleic acids encoding candidate antibody variants, wherein each candidate antibody variant comprises an amino acid residue alteration in each HVR of the heavy chain variable region (V_H) or the light chain variable region (V_L) compared to a reference antibody;

(b) sorting the display library based on binding of the candidate antibody variants to the target molecule to form a sorted library, wherein the sorted library comprises candidate antibody variants with enhanced binding to the target molecule compared to the reference antibody; and

(c) comparing the frequency at which each amino acid residue alteration is present in the display library and in the sorted library as determined by massively parallel sequencing, thereby determining whether each amino acid residue alteration is enriched in the sorted library compared to the display library,

whereby the amino acid residue alteration is identified as conferring enhanced binding to the target molecule if it is enriched in the sorted library compared to the display library.

95. The method of claim 94, further comprising determining the frequency at which each amino acid alteration is present in the display library and the sorted library by massively parallel sequencing following step (b).

96. The method of claim 94 or 95, wherein step (c) further comprises comparing the frequency at which a pair comprising a first amino acid residue alteration and a second amino acid residue alteration is present in the display library and in the sorted library, thereby determining whether the pair is enriched, depleted, or neutral in the sorted library compared to the display library.

97. The method of any one of claims 94-96, wherein the antibody is a dual specific antibody.

98. A method of identifying an amino acid residue alteration that allows enhanced binding of a dual specific antibody to both a first epitope and a second epitope, the method comprising:

(a) providing a display library comprising nucleic acids encoding candidate antibody variants, wherein each candidate antibody variant comprises an amino acid residue alteration in each HVR of the V_H or the V_L compared to a reference dual specific antibody;

(b) sorting the display library based on binding of the candidate antibody variants to the first epitope to form a first sorted library, wherein the first sorted library comprises candidate antibody variants with enhanced binding to the first epitope compared to the reference dual specific antibody;

(c) sorting the display library based on binding of the candidate antibody variants to the second epitope to form a second sorted library, wherein the second sorted library comprises candidate antibody variants with enhanced binding to the second epitope compared to the reference dual specific antibody; and

(d) comparing the frequency at which each amino acid residue alteration is present in the display library, the first sorted library, and the second sorted library as determined by massively parallel sequencing, thereby determining whether each amino acid residue alteration is enriched, depleted, or neutral in the first sorted library and the second sorted library compared to the display library,

whereby the amino acid residue alteration is identified as allowing enhanced binding of the dual specific antibody to both the first epitope and the second epitope if the amino acid residue alteration is enriched in both the first sorted library and the second sorted library compared to the display library or is enriched in one of either the first sorted library or the second sorted library and is neutral in the other sorted library.

99. The method of claim 98, further comprising determining the frequency at which each amino acid residue alteration is present in the display library, the first sorted library, and the second sorted library by massively parallel sequencing following step (c).

100. The method of claim 98 or 99, wherein step (d) further comprises comparing the frequency at which a pair comprising a first amino acid residue alteration and a second amino acid residue alteration is present in the display library and in the first sorted library, the second sorted library, or both, thereby determining whether the pair is enriched, depleted, or neutral in the first sorted library, second sorted library, or both, compared to the display library.

101 . The method of any one of claims 98-100, wherein the first epitope and the second epitope are from the same target molecule.

102. The method of any one of claims 98-101 , wherein the first epitope is from a first target molecule and the second epitope is from a second target molecule.

103. The method of claim 102, wherein the first target molecule and the second target molecule are cytokines.

104. The method of claim 103, wherein the first target molecule is VEGF and the second target molecule is selected from the group consisting of Ang2, Ang1 , PDGF-B, PDGF-C, Stromal-derived growth factor- 1 , placental growth factor (PIGF), factor D, and complement factor 1 .

105. The method of claim 104, wherein the first target molecule is VEGF and the second target molecule is Ang2.

106. The method of any one of claims 94-105, wherein the display library comprises candidate antibody variants having amino acid residue alterations at every position in each HVR of the V_H or V_L.

107. The method of any one of claims 94-106, wherein the display library comprises amino acid residue alterations in only the V_H or the V_L of the candidate antibody variants.

108. The method of any one of claims 94-106, wherein the display library comprises amino acid residue alterations in the V_H and the V_L of the candidate antibody variants.

109. The method of any one of claims 94-106, wherein the display library comprises a V_H library and a V|_ library, wherein the V_H library comprises candidate antibody variants with an amino acid residue alteration in each HVR of the V_H, and the V_L library comprises candidate antibody variants with an amino acid residue alteration in each HVR of the V_L.

1 10. The method of any one of claims 94-109, wherein the display library is selected from the group consisting of a phage display library, a bacterial display library, a yeast display library, a mammalian display library, a ribosome display library, and an m RNA display library.

1 1 1 . The method of claim 1 10, wherein the display library is a phage display library.

1 12. The method of any one of claims 94-1 1 1 , wherein the amino acid residue alteration is encoded by a degenerate codon set.

1 13. The method of claim 1 12, wherein the degenerate codon set is an NNK or an NNS codon set, wherein N is A, C, G, or T; K^"is G or T; and S is C or G.

1 14. The method of claim 1 13, wherein the degenerate codon set is an NNK codon set.

1 15. The method of any one of claims 94-1 14, wherein the sorting of step (b) or (c) comprises contacting the display library with an immobilized target molecule or epitope.

1 16. The method of any one of claims 94-1 14, wherein the sorting of step (b) or (c) comprises contacting the display library with a soluble target molecule or epitope.

1 17. The method of any one of claims 94-1 16, wherein the display library comprises at least 1 x 10⁶ candidate antibody variants.

1 18. The method of claim 1 17, wherein the display library comprises at least 1 .5 x 10⁷ candidate antibody variants.

1 19. The method of claim 1 18, wherein the display library comprises at least 2.5 x 10⁷ candidate antibody variants.

120. The method of any one of claims 94-1 19, wherein the massively parallel sequencing comprises deep sequencing, ultra-deep sequencing, and/or next-generation sequencing.

121 . The method of any one of claims 94-120, wherein the massively parallel sequencing comprises determining the sequence of at least 500,000 reads.

122. The method of claim 121 , wherein the massively parallel sequencing comprises determining the sequence of at least 1 ,000,000 reads.

123. The method of any one of claims 94-122, wherein the antibody or dual specific antibody is a monoclonal antibody.

124. The method of any one of claims 94-123, wherein the antibody or dual specific antibody is an IgG antibody.

125. The method of any one of claims 94-123, wherein the antibody or dual specific antibody is an antibody fragment.

126. The method of claim 125, wherein the antibody fragment is selected from the group consisting of Fab, scFv, Fv, Fab', Fab'-SH, F(ab')₂, and diabody.

127. The method of claim 126, wherein the antibody fragment is a Fab.

128. The method of any one of claims 94-127, wherein the method further comprises generating an antibody that comprises an amino acid residue alteration identified by the steps of the method.

129. A method of generating a dual specific antibody that binds a first epitope with a Kd of lower than 1 nM and a second epitope with a Kd of lower than 1 nM, the method comprising:

(a) providing a dual specific antibody that binds the first epitope with a Kd of greater than 1 nM and the second epitope with a Kd of greater than 1 nM ;

(b) identifying one or more amino acid residue alterations that allows enhanced binding of the dual specific antibody to both the first epitope and the second epitope according to the method of any one of claims 98-127, wherein the one or more amino acid residue alterations allows binding the first epitope with a Kd of lower than 1 nM and the second epitope with a Kd of lower than 1 nM ; and

(c) altering the amino acid sequence of the dual specific antibody based on the results of step (b), thereby generating a dual affinity antibody that binds a first epitope with a Kd of lower than 1 nM and a second epitope with a Kd of lower than 1 nM.