BISPECIFIC ANTIBODY CONSTRUCTS AND METHODS OF USE

CROSS-REFERENCE TO RELATED APPLICATION

This application claims the benefit of priority of U.S. Provisional Appl. No. 62/580,079, filed November 1, 2017, the contents of which are incorporated by reference in their entirety herein.

FIELD OF THE INVENTION

The present invention relates to bispecific antibody constructs that specifically bind integrin alpha-V and integrin a5 and methods of making and using those constructs to treat cancer (e.g., prostate cancer) or other conditions associated with expression and activity of those integrins.

Integrin alpha-V and integrin a5 function in diverse tumor cell types to promote a range of behaviors essential to the survival, progression and metastatic behaviors of cancer cells. These include stem-progenitor function, anchorage to the surrounding matrix, facilitation of fibrin-invasion in metastatic niches, transduction of survival signals,

activation of membrane proteases including matrix-metalloproteases, response to specific environmental cues to home to distant organs, bone metastases, the induction of fibroblast- myofibroblast transitions, epithelial-mesenchymal transition, angiogenesis, and immune evasion via metalloprotease and TGF-beta signaling activation. Weis et al, Cold Spring Harbor Perspectives in Medicine 2011;l ;a006478; Schaffner et al, Cancers 2013;1 :27-47; Raab-Westphal et al, Cancers (Basel) 2017;9: 110; Sutherland et al, Cancers (Basel)

2012;4: 1106-45; Khan et al, Cell and Tissue Research 2016;365:65673; Hinz, Nature

Medicine 2013;19: 1567; Wilson et al., PLoS One 2011;6:e22842. These integrins also play critical roles in non-cancer pathologies such as fibrosis and pathological angiogenesis.

Weis et al., 2011; Hinz, 2013; Henderson et al, Nature Medicine

2013: 19: 10.1038/nm.3282; Asano et al., J. Immunol 2005;175:7708-18; Ray, Nature

Reviews Gastroenterology & Hepatology 2013: 11 :4; Avraamides et al, Nat Rev Cancer 2008;8:604-617. SUMMARY

An object of the present invention is to provide a bispecific antibody construct that simultaneously and specifically binds to integrin alpha-V and integrin a5, which are encoded by the distinct ITGA V and ITGA5 genes, respectively, in humans. In some embodiments, for example, where the antibody construct is conjugated to a

chemotherapeutic or cytotoxic agent, the antibody construct may specifically bind integrin alpha-V and integrin a5 without inhibiting the activity of either or both targets. In other embodiments, for example, where the antibody construct is unconjugated to a

chemotherapeutic or cytotoxic agent, the antibody construct may specifically bind integrin alpha-V and integrin a5 and inhibit their biological activities to a clinically beneficial extent.

Accordingly, in a first aspect, the invention features a bispecific antibody construct or a biologically active fragment thereof that includes (a) a first paratope that specifically binds an alpha-V integrin; and (b) a second paratope that specifically binds an a5 integrin. The epitope on the a5 integrin may be wholly contained within that protein. Alternatively, the epitope may reside at least partially within the βΐ integrin that, in living systems, forms a heterodimer with a5. Any type of bispecific antibody is contemplated; for non-limiting examples of formats, see FIG. 11. More information about bispecific antibodies can be found in a review by Brinkmann et al, mAbs 9(2): 182-212 (2017). See Fig. 2 of that review for an illustration of many bispecific antibody formats that have been used in the art and that could be utilized for the present bispecific antibody construct.

In one embodiment, the construct comprises an antigen binding fragment (Fab) portion that contains the first paratope and a single-chain variable fragment (scFv) portion that contains the second paratope. Either paratope may be from an anti-alpha-V antibody, such as abituzumab; the other paratope is from an anti-a5 antibody, such as volociximab. The bispecific construct may be bivalent, trivalent, tetravalent, or of higher valency.

The bispecific antibody constructs described herein can include, for example, one or more of the following complementarity determining regions (CDRs):

(a) within a light chain variable domain of the first paratope: RASQDISNYLA (PIVLCDRI ; SEQ ID NO: 1); YTSKIHS (PIVLCDR2; SEQ ID NO:2); QQGNTFPYT (PIVLCDR3; SEQ ID NO:3); or an amino acid sequence at least 80% identical to any of those three, or where there are zero, one or two amino acid deletions, additions, or substitutions in any of them;

(b) within a heavy chain variable domain of the first paratope: SFWMH

(PIVHCDRI ; SEQ ID NO:4); YINPRS GYTEYNEIFRD (PIVHCDR2; SEQ ID NO:5); FLGRGAMDY (PIVHCDR3; SEQ ID NO:6); or an amino acid sequence at least 80% identical to any of those three, or where there are zero, one or two amino acid deletions, additions, or substitutions in any of them;

(c) within a light chain variable domain of the second paratope: TASSSVSSNYLH (P2V1XDRI ; SEQ ID NO:7); STSNLAS (P ₂ VLCDR2; SEQ ID NO:8); HQYLRSPPT

(P2V1XDR3; SEQ ID NO: 9); or an amino acid sequence at least 80% identical to any of those three, or where there are zero, one, or two amino acid deletions, additions, or substitutions in any of them; or

(d) within a heavy chain variable domain of the second paratope: GFSLTDYGVH (P ₂ VHCDR1; SEQ ID NO: 10); VIWSDGSSTYNSALKS (P ₂ VHCDR2; SEQ ID NO: 11);

HGTYYGMTTTGDALDY (P ₂ VHCDR3; SEQ ID NO: 12); or an amino acid sequence at least 80% identical to any of those three (e.g., there may be zero, one or two amino acid deletions, additions, or substitutions in any of them).

Regarding the framework regions (FRs), any of the bispecific antibody constructs described herein, and in particular any construct having one or more of the CDRs described above, can include one or more of the following framework regions:

(a) within a light chain variable domain contributing to the first paratope:

DIQMTQSPSSLSASVGDRVTITC (PIVLFRI ; SEQ ID NO: 13); WYQQKPGKAPKLLIY (PIVLFR2; SEQ ID NO: 14); GVPSRFSGSGSGTDYTFTISSLQPEDIATYYC (PIVLFR3; SEQ ID NO: 15); FGQGTKVEIK (PIVLFR4; SEQ ID NO: 16); or an amino acid sequence at least 80% identical to any of those three (e.g., there may be zero, one or two amino acid deletions, additions, or substitutions in any of them);

(b) within a heavy chain variable domain contributing to the first paratope:

QVQLQQSGAELAEPGASVKMSCKASGYTFS (PIVHFRI; SEQ ID NO: 17);

WVRQAPGQGLEWIG (PIVHFR2; SEQ ID NO: 18); KATMTTDTSTSTAYMELSSLRSEDTAVYYCAS (PIVHFR3; SEQ ID NO: 19);

WGQGTTVTVSS (PIVHFR4; SEQ ID NO:20); or an amino acid sequence at least 80% identical to any of those four (e.g., there may be zero, one or two amino acid deletions, additions, or substitutions in any of them);

(c) within a light chain variable domain contributing to the second paratope:

QIVLTQSPAIMSASLGERVTMTC (P ₂ VLFR1 ; SEQ ID NO:21);

WYQQKPGSAPNLWIY (P ₂ VLFR2; SEQ ID NO:22);

GVPARFSGSGSGTSYSLTISSMEAEDAATYYC (P2VLFR3; SEQ ID NO:23);

FGGGTKLEIK (P ₂ VLFR4; SEQ ID NO:24); or an amino acid sequence at least 80% identical to any of those four (e.g., there may be zero, one or two amino acid deletions, additions, or substitutions in any of them); or

(d) within a heavy chain variable domain contributing to the second paratope:

QVQLKESGPGLVAPSQSLSITCTIS (P ₂ VHFR1; SEQ ID NO:25);

WVRQPP GKGLEWLV (P ₂ VHFR2; SEQ ID NO:26);

RMTIRKDNSKSQVFLIMNSLQTDDSAMYYCAR (P ₂ VHFR3; SEQ ID NO:27);

WGQGTSVTVSS (P ₂ VHFR4; SEQ ID NO:28); or an amino acid sequence at least 80% identical to any of those four (e.g., there may be zero, one or two amino acid deletions, additions, or substitutions in any of them).

In a second embodiment, the second paratope, which binds and may inhibit the activity (e.g., at least partially neutralize the activity) of a5 or α5β1, can include one or more of the following CDRs:

(a) within a light chain variable domain of the second paratope, RASQSVSSYLA (PIVLCDRI ; SEQ ID NO:33); DASNRAT (PIVLCDR2; SEQ ID NO:34); QQRSNWPLT (PIVLCDR3; SEQ ID NO:35); or an amino acid sequence at least 80% identical to any of those three, or where there are zero, one or two amino acid deletions, additions, or substitutions in any of them; and

(b) within a heavy chain variable domain of the second paratope, SSSYWG (PIVHCDRI; SEQ ID NO:36); SIYYSGRNYNNPSLKS (PIVHCDR2; SEQ ID NO:37); and HYYGSGSSYYYYDLD (PIVHCDR3; SEQ ID NO:38); or an amino acid sequence at least 80% identical to any of those three, or where there are zero, one or two amino acid deletions, additions, or substitutions in any of them.

In a third embodiment, the second paratope, which binds and may inhibit at least partially an activity of a5 (e.g., α5β1), can include one or more of the following CDRs:

(a) within a light chain variable domain of the second paratope: RASQSVSSYLA

(PIVLCDRI ; SEQ ID NO:33); DASNRAT (PIVLCDR2; SEQ ID NO:34); QQRSNWPLT

(PIVLCDR3; SEQ ID NO:35); or an amino acid sequence at least 80% identical to any of those three, or where there are zero, one or two amino acid deletions, additions, or substitutions in any of them; and

(b) within a heavy chain variable domain of the second paratope: SYAMH

(PIVHCDRI; SEQ ID NO:39); VISFDGSNKNYADSVKG (PIVHCDR2; SEQ ID NO:40); and EYWGTYYYGMDV (PIVHCDR3; SEQ ID NO:41); or an amino acid sequence at least 80% identical to any of those three, or where there are zero, one or two amino acid deletions, additions, or substitutions in any of them.

In a fourth embodiment, the second paratope, which binds and may inhibit at least partially an activity of a5 (e.g., α5β1), can include one or more of the following CDRs:

(a) within a light chain variable domain of the second paratope: RASQSVSSYLA (PIVLCDRI ; SEQ ID NO:33); DASNRAT (PIVLCDR2; SEQ ID NO:34); QQRSNWPLT (PIVLCDR3; SEQ ID NO:35); or an amino acid sequence at least 80% identical to any of those three, or where there are zero, one or two amino acid deletions, additions, or substitutions in any of them; and

(b) within a heavy chain variable domain of the second paratope: STYAMH (PIVHCDRI ; SEQ ID NO:42); VISYDGSNKYYADSVKG (PIVHCDR2; SEQ ID NO:43); and RESPPIYYYYGMDV (PIVHCDR3; SEQ ID NO:44); or an amino acid sequence at least 80% identical to any of those three, or where there are zero, one or two amino acid deletions, additions, or substitutions in any of them.

In the second, third, and fourth embodiments, the CDRs are as disclosed in U.S. Patent No. 8,039,596, the entire content of which is incorporated by reference herein.

An antibody construct of the invention can also have one or more of the following characteristics: it may further include an Fc region, such as one derived from a human IgA or human IgG Fc region; it may be a chimera (with one or all variable domain sequences derived from a non-human animal, and the Fc region derived from a human); it may be fully human, humanized, or de-immunized; it may be divalent, trivalent, or tetravalent; and/or it may include a detectable marker. For example, an antibody construct can include an Fc portion and be chimeric; can be human and tetravalent; can be humanized and include a detectable marker; or have any other compatible combination of the features just mentioned or set out below.

In some embodiments, the construct comprises a first polypeptide comprising a first light chain variable domain linked to a light chain constant region, and a second polypeptide chain comprising, in order from amino to carboxy terminus, a first antibody heavy chain variable domain linked to a heavy chain constant region linked to an scFv, where the scFv contains, in either order, a second heavy chain variable domain that is different from the first heavy chain variable domain and is linked to a second light chain variable domain that is different from the first light chain variable domain. The first light and heavy chain variable domains form a paratope that binds to either integrin a5 (e.g. , α5β1) or integrin alpha V. The second light and heavy chain variable domains form a second paratope that binds to the other integrin. An example of this construct with signal sequences (MELGLS WIFLLAILKGVQC; SEQ ID NO:29, and

MRVPAQLLGLLLLWLPGTRC, SEQ ID NO: 30) still attached to the respective amino termini is shown in Figs. 12A and 12B. The sequence of FIG.12A (SEQ ID NO:31) is configured to include, in addition to the signal sequence SEQ ID NO: 29, the variable region of the heavy chain of a first paratope, the constant region of that heavy chain, a linker, the heavy chain variable region of a second paratope, a linker, and the light chain variable region of the second paratope. As one of ordinary skill in the art would recognize, the heavy and light chain variable regions of the second paratope would cooperate as an scFv that is covalently attached to the carboxy terminus of the heavy chain constant region. The light chain (including signal sequence SEQ ID NO: 30) of the first paratope is shown in Fig. 12B (SEQ ID NO:32). In one embodiment, the invention includes a construct comprising two polypeptides, one of which is represented by the sequence shown in Fig. 12A (SEQ ID NO: 31), except lacking the signal sequence MELGLSWIFLLAILKGVQC (SEQ ID NO:29) portion of SEQ ID NO: 31 , and the other of which is represented by the sequence shown in Fig. 12B (SEQ ID NO: 32), except lacking the signal sequence MRVPAQLLGLLLLWLPGTRC (SEQ ID NO:30) portion of SEQ ID NO: 32.

In one aspect, the disclosure features a bispecific construct comprising an amino acid sequence that is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, or at least 95% identical to the amino acid sequence set forth in SEQ ID NO:63. In certain instances, the disclosure features a bispecific construct comprising an amino acid sequence set forth in SEQ ID NO:63 except having 1 to 20 (e.g., 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 13, 14, 15, 16, 17, 18, 19 or 20) amino acid substitutions (e.g., conservative amino acid substitutions). In certain instances, the substitutions are in the framework, constant, and/or linker region. In certain instances, the substitutions are in one or more of the CDRs.

In another aspect, the disclosure features a bispecific construct comprising an amino acid sequence that is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, or at least 95% identical to the amino acid sequence set forth in SEQ ID NO:64. In certain instances, the disclosure features a bispecific construct comprising an amino acid sequence set forth in SEQ ID NO:64 except having 1 to 20 (e.g., 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 13, 14, 15, 16, 17, 18, 19 or 20) amino acid substitutions (e.g., conservative amino acid substitutions). In certain instances, the substitutions are in the framework and/or constant region. In certain instances, the substitutions are in one or more of the CDRs.

In one aspect, the disclosure features a bispecific construct comprising : (1) an amino acid sequence that is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, or at least 95% identical to the amino acid sequence set forth in SEQ ID NO:63; and (2) an amino acid sequence that is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, or at least 95% identical to the amino acid sequence set forth in SEQ ID NO:64. In certain instances, the disclosure features a bispecific construct comprising an amino acid sequence set forth in SEQ ID NO:63 and/or 64 except having 1 to 20 (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 substitutions (e.g., conservative amino acid substitutions). In certain instances, the substitutions are in the framework, constant, and/or linker region. In certain instances, the substitutions are in one or more of the CDRs. In another aspect, the invention features nucleic acid that comprises a sequence or sequences that together encode all polypeptides of a bispecific antibody construct as described herein, or a portion thereof, or a set of nucleic acid molecules that together encode the polypeptides that make up the bispecific antibody construct. For example, the nucleic acid may encode a polypeptide comprising a heavy and/or light chain variable domain within the first or second paratope of any of the bispecific antibody constructs described herein. The nucleic acid may encode a polypeptide comprising a heavy chain variable domain of one of the paratopes, a heavy chain constant domain, and an scFv containing the heavy and light chain variable domains of the other paratope. As an example, the nucleic acid may encode a polypeptide comprising the amino acid sequence of Fig. 12 A, with or without the signal sequence shown in that figure (or with a different signal sequence). Any signal sequence known in the art as useful for expressing antibody sequences can be used. The nucleic acid preferably includes appropriate expression control sequences operably linked to the coding sequence.

In another aspect, the invention features an expression vector that includes the nucleic acid described above or elsewhere herein.

In another aspect, the invention features a host cell (e.g., a bacterial, yeast, insect, or mammalian cell) comprising the expression vector described above or elsewhere herein.

In another aspect, the invention features a protein expression system that includes the host cell described above or elsewhere herein.

In another aspect, the invention features a pharmaceutical composition that includes a bispecific antibody construct described herein and a pharmaceutically acceptable carrier. The composition can be formulated for administration to a patient through any route (e.g., it may be formulated for oral, nasal, intramuscular, intraperitoneal, or intravascular (e.g. , intravenous), administration).

In another aspect, the invention features a kit that includes a bispecific antibody construct described herein and instructions for use in treatment.

In another aspect, the invention features a method of making a bispecific antibody construct described herein. The method can include a step of culturing the host cell as described herein under conditions in which the nucleic acid encoding one or both polypeptide(s) making up the bispecific antibody are expressed.

In another aspect, the invention features a method of treating cancer. The method can include administering a therapeutically effective amount of a bispecific antibody construct described herein to a patient in need thereof (e.g. , a patient who has a cancer in which either integrin is implicated in the pathogenesis or progression of the disease, including of any of the following types of cancer: prostate, kidney, breast, melanoma, pancreatic, ovarian, colonic, cervical, head and neck, lung, gastric, endometrial, bone and brain metastases, soft-tissue sarcoma, osteosarcoma, hepatoma, basal cell carcinoma, glioblastoma, angiosarcoma, T-cell lymphoma, and multiple myeloma).

In another aspect, the invention features a method of treating a non-malignant condition in which either alpha v or alpha 5 integrin is expressed and implicated in the pathogenesis or progression of the disease. Such diseases include, e.g., non-malignant fibrotic diseases involving lung, liver, heart, kidney, bone marrow, or skin, such as scleroderma, chronic graft-vs-host disease, liver fibrosis, and interstitial lung disease. The general category of non-malignant diseases treatable with the present antibodies also includes, e.g., conditions involving pathological angiogenesis, such as proliferative retinopathy, wet macular degeneration, arteriovenous malformations, and cavernous hemangiomas. The method of treatment can include administering a therapeutically effective amount of a bispecific antibody construct described herein to a patient in identified need thereof.

As used in this specification and the appended claims, the singular forms "a," "an" and "the" include a plurality of like or unlike referents unless the context clearly dictates otherwise. For example, when we say the invention features "a method," we do not mean only one single method but rather a plurality of methods that may or may not share common steps. When we refer to "a host cell," that may be a single host cell or a plurality of cells that may be highly similar or different (e.g. , a mixture of different cell types).

Other features and advantages of the invention will be apparent from the following detailed description, and from the claims. All publications mentioned herein are hereby incorporated by reference in their entirety.

BRIEF DESCRIPTION OF THE DRAWINGS FIGs. 1A and IB are graphs summarizing the results obtained in migration and adhesions assays with PC-3 prostate cancer cells (FIG. 1A) and DU-145 prostate cancer cells (FIG. IB). The cells were assayed in the presence of serum-free medium (SFM, negative control) or human bone marrow-mesenchymal stromal cell conditioned media (hBM-MSC CM), a matched isotype control antibody (Iso) and antibodies that specifically bind integrin βΐ (βΐ), integrin β3 (β3), integrin alpha 5 (α5), integrin alpha 6 (a6), integrin alpha-V (aV), or both integrin a5 and integrin alpha-V (a5aV).

FIGs. 2A and 2B are a pair of bar graphs illustrating that human umbilical vein endothelial cell (HUVEC) migration induced by prostate cancer-hBM-MSC interaction is regulated by integrins a5 (FIG. 2A) and alpha-V (FIG. 2B).

FIGs. 3A and 3B include data concerning the purity and binding of an

ITGA5BlxAV bispecific antibody. FIG.3A is a photograph of a reduced SDS-PAGE gel showing the purity of the bispecific antibody. FIG. 3B is a table containing data concerning binding studies that demonstrate correlation of binding between anti-a5, anti- alpha-V, and bispecific anti-a5-alpha-V antibodies across a panel of cell lines (PC-3, C42B, DU-145, VCAP, and RWPE-1). The mean fluorescent intensities (MFI) for the a5, alpha- V, and bispecific ITGA5BlxAV stained cell populations for each cell line were calculated from the flow cytometry data. Then, a correlation coefficient between the sum of a5 and alpha-V MFIs and the bispecific ITGA5BlxAV was calculated.

FIG. 4 is a panel of bar graphs mapping adherent and migrated PC-3 (top) and DU- 145 cells (bottom) following antibody neutralization. Adhesion was assessed after one hour of co-culturing with hBM-MSC conditioned medium (CM) and migration was assessed after 24 hours. The a5 and alpha-V integrins were neutralized with 50 μg/ml of each of the indicated antibodies, and the bispecific antibody construct was found to be superior to either the anti-a5 or anti-alpha-V antibodies alone in inhibiting adhesion and migration. ***: P < 0.001. SFM = serum-free media. "Iso" = isotype control. FIG. 5 is a panel of bar graphs displaying the results of migration and adhesion assays of prostate cancer cells (PC-3 cells, shown in the top two graphs and DU-145 cells, shown in the bottom two graphs). The integrins a5 and alpha- V were neutralized with the monoclonal and bispecific antibody constructs indicated. The bispecific antibody

ITGa5 ixaV inhibited the adhesion and migration of both cell types more than the combined antibody treatment. The significance of the difference in the effect of the bispecific antibody compared to combination of monoclonal antibodies was calculated using one-way analysis of variance (ANOVA) with multiple comparisons between equivalent dose pairs amongst the two test groups. Significance shown is the p-value summary (GraphPad Prism). ****; p < 0.0001. SFM = serum-free medium. "Iso" = isotype control.

FIGs. 6A and 6B are a pair of bar graphs showing the results of cellular migration studies. Endothelial migration was induced by prostate cancer-hBM-MSC co-culture for 24 hours with integrin blockade on DU-145 cells (FIG. 6A) or on HUVEC endothelial cells (FIG. 6B). Neutralization was performed with 10 μg/mL of total antibody. The bispecific antibody was superior to the combined monoclonal antibodies in inhibiting endothelial cell migration to co-culture. ***: P < 0.001.

FIGs. 7A and 7B are a panel of bar graphs summarizing data from assays of non- prostate cancer cell migration (FIG. 7 A; 24 hours) and adhesion (FIG. 7B; one hour) to hBM-MSC CM (conditioned medium). Neutralization was performed with 50 μg/mL of total antibody. The bispecific antibody construct was superior to the combination of anti- α5β1 and anti-alpha-V monoclonal antibodies at inhibiting migration and adhesion of a wide range of non-prostate cancer cells. * indicates P < 0.05. ** indicates P< 0.01.

SFM = serum-free media. "Iso" = isotype control.

FIGs. 8A and 8B are a pair of bar graphs illustrating the results of an assay for cancer cell viability. The cells were exposed to the antibodies indicated, at the

concentrations indicated. IgG is an isotype control; ITGA5 is an integrin a5 monospecific antibody; ITGB1 is an integrin beta 1 monospecific antibody; ITGB3 is an integrin beta 3 monospecific antibody; ITGAVB3 is an integrin alpha-V-beta 3 antibody. FIG. 9 is a bar graph showing the results of an assay in which HUVEC or DU-145 cells were treated with the indicated doses of bevacizumab or the bispecific antibody construct ITGA5BlxAV (doses in μg/mL) prior to assaying HUVEC migration to DU- 145/hBM-MSC co-culture. Bispecific antibody treatment (1, 3 μg/mL) of either HUVEC or DU-145 was significantly more potent at reducing endothelial migration compared to bevacizumab. ****: P≤ 0.0001, ***: P < 0.001.

FIG. 10 is a schematic diagram showing a format of one embodiment of a bispecific antibody of the invention. Stable scFvs are attached to the constant region of an IgG, with the resulting construct being tetravalent.

FIG. 11 is a schematic diagram derived from Fig. 1 of Liu et al, Front. Immunol

(2017) doi.org/10.3389/fimmu.2017.00038, showing various formats of bispecific antibodies and scaffolds useful in the present invention; a bispecific antibody that specifically binds integrin alpha-V and integrin a5 can be engineered in any of these formats or assume the configuration of any of these scaffolds. (A) is a bispecific quadroma, which can be generated by somatic fusion of two hybridomas; (B-J) are bispecific formats developed by using a knob-in-hole (KiH) Fc heterodimerization strategy; (K) is a bispecific IgGl developed by controlled Fab-arm exchange; (L) is a bispecific Fc- fusion construct developed by electrostatic optimization; (M-O) are bispecific formats developed by strand exchange, insertion of cleavage motif and expressing two light chains with a single heavy chain; and (P-W) are other bispecific or multi-specific scaffolds.

Abbreviations: OAscFab-IgG represents a one-arm single-chain Fab-immunoglobulin gamma (IgG); dsFv-IgG represents a disulfide stabilized Fv-IgG; Cfae-IgGl represents a controlled Fab-arm exchanged IgGl ; scFv-Fc represents a charged pair single-chain Fv-Fc fusion; SEEDbody represents a strand-exchange engineered domain body; LUZ-Y represents a two-arm leucine zipper heterodimeric monoclonal antibody; κγ-body represents a kappa lambda body; BiTE represents a bispecific T-cell engager; BiKE/TriKE represents bispecific and trispecific killer cell engagers, respectively; DART represents dual-affinity retargeting molecules; mFc represents a monomericFc; and Fcab represents an Fc antigen binding domain. FIG. 12A is a representation of an amino acid sequence (SEQ ID NO: 47) including a signal sequence and the heavy chain of an IgG construct that binds alpha-V, linked to an scFv domain that binds α5β1. Sub-regions, including the signal sequence (which would be cleaved off when the antibody is secreted from a host cell), CDRs and FRs of the heavy chain variable domain targeting alpha-V, and an IgG heavy chain constant region, are marked. At the carboxy end of the heavy chain constant region is a linker sequence GGGGSGGGGS (SEQ ID NO:45), followed by an scFv targeting α5β1. The scFv sequence targeting α5β1 comprises a heavy chain variable region with CDRs and FRs identified, followed by another linker (GGGGS GGGGSGGGGS; SEQ ID NO:46), followed by a light chain variable region with CDRs and FRs identified.

FIG. 12B is a representation of an amino acid sequence (SEQ ID NO: 48) that includes a signal sequence and the variable light chain of a construct that binds alpha-V. Two polypeptide having the sequence shown in FIG. 12A and two polypeptides having the sequence shown in FIG. 12B (minus their respective signal sequences) would together form a tetravalent, bispecific, IgG-scFv antibody targeting the a5 integrin (α5β1) and the alpha-V integrin.

FIGs. 12C-1 and 12C-2 are a table delineating the hypervariable CDRs and the FRs of the sequences shown in FIGs. 12A and 12B, and variants thereof.

FIG. 13 is a representation of an expression vector (a plasmid) useful in expressing nucleic acid sequences encoding heavy and/or light chains of an antibody construct described herein.

FIG. 14 is an alignment of sequences including variants of CDRs in the variable regions of heavy and light chains targeting α5β1. The uppermost sequence of each chain represents a murine sequence and those beneath it represent five variants that are humanized antibodies, as disclosed in U.S. Patent No. 7,276,589, the entire content of which is hereby incorporated by reference herein. The uppermost VH sequence is assigned SEQ ID NO:49. The five VH sequences below it are assigned SEQ ID NOs:50-54, in order. The uppermost VL sequence is assigned SEQ ID NO:55. The five VH sequences below it are assigned SEQ ID NOs: 56-60, in order. FIG. 15A is a Western blot showing the effect of antibody treatment on expression of three proteins in PC-3 cancer cells.

FIG. 15B is a bar graph quantifying the level of integrin av expression in PC-3 cancer cells after antibody treatment, normalized to vehicle.

FIGs. 16A and 16B are Western blots showing the effect of antibody treatment on expression of four proteins in DU-145 and VCAP cancer cells.

FIG. 17A is a Western blot showing the effect of antibody treatment on expression of four proteins in C42B cancer cells.

FIG. 17B is pair of bar graphs showing the effect of antibody treatment on relative expression of integrin av and a5 in C42B cells.

FIGs. 18A and 18B are Western blots showing the effect of antibody treatment on expression of various proteins in HUVEC endothelial cells.

FIG. 19 is a Western blot showing the effect of antibody treatment on expression of various proteins in in activated fibroblasts.

FIG. 20A is a pair of nucleotide sequences, the upper one (SEQ ID NO:61)

encoding the polypeptide shown in FIG. 12A, including its signal sequence, and the lower one (SEQ ID NO:62) encoding the polypeptide shown in FIG 12B, including its signal sequence.

FIG. 20B is a pair of amino acid sequences, the upper one (SEQ ID NO:63) being the mature form (i.e., without the signal sequence) of the polypeptide shown in FIG. 12A, and the lower one (SEQ ID NO: 64) being the mature form of the polypeptide shown in

FIG. 12B.

DETAILED DESCRIPTION

The studies described below indicate cooperativity between the a5 integrin and the alpha- V integrin in a range of cancer-specific behaviors, including tumor-stromal interactions and the induction of an angiogenic response required for the survival and progression of these tumors. These two integrins are also implicated as contributing to many non-cancer pathologies related to fibrosis or angiogenesis. For example, alpha-V is critical to the generation of autocrine TGF-beta signaling in scleroderma fibroblasts and in the generation of myofibroblasts in liver fibrosis and interstitial lung disease (Henderson et al, 2013; Asano et al, 2005; Ray, 2013). Integrin alpha-V and α5 are master regulators of vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF)-driven angiogenesis, such as in angiogenesis pathologies including, e.g., proliferative retinopathy, wet macular degeneration, arteriovenous malformations, and cavernous hemangiomas.

As detailed below, a bispecific antibody containing an antigen-binding domain that binds to and neutralizes a5 integrin (or its obligate heterodimeric form α5β1) combined with a different antigen-binding domain that binds to and neutralizes alpha-V integrin was found to result in markedly enhanced anti-tumor activity compared to either single agent monospecific antibody, or a mixture of the two monospecific antibodies, in a range of tumor biology assays, as detailed below.

The central hypothesis pertinent to the invention described herein is that a bispecific antibody that simultaneously targets a5 (or α5β1) and alpha-V integrin would be significantly superior to combinations of individual a5 and alpha-V monospecific antibodies by virtue of enhanced binding precision to these integrins that require close physical proximity on the cell membrane for their optimal cooperativity, e.g. in focal adhesions. Additional advantages of a bispecific construct over combinations of monospecific antibodies likely include increased binding affinity as a result of the localization of the remaining binding site to the cell membrane following binding of the first, and increased binding specificity as a result of binding two or more cell surface antigens. Increased binding affinity and residence time when both antigens are engaged simultaneously by the two pharmacophores in a bispecific antibody is referred to as cross-arm binding efficiency. The spatial exclusion of one antibody by another, as well as asymmetric pharmacokinetic and pharmacodynamic properties, could account for suboptimal results with monospecific antibodies in combination. The virtues of a bispecific antibody may therefore translate to enhanced anti-tumor efficacy and reduced toxicity in vivo, with improved ease of administration as well as simplified pharmacokinetic and pharmacodynamic monitoring. The below data demonstrate the superiority of a prototypical bispecific a5 i+alpha-V antibody construct over combinations of monospecific α5/α5β1 and alpha-V antibodies in a range of tumor biology assays, including tumor-stromal interactions and induction of angiogenesis.

Integrin alpha-V: Integrins are heterodimeric integral membrane proteins

composed of an alpha chain and a beta chain. Integrin alpha-V (sometimes referred to as av, aV, av, or aV) undergoes post-translational cleavage to yield disulfide-linked heavy and light chains that combine with multiple integrin beta chains to form different integrins. The monoclonal antibodies intetumumab and abituzumab target this protein, which is found on some tumor cells (Elez et al , Annals of Oncology, 26(1): 132-140, 2015).

Integrin a5: Integrin a5 (sometimes referred to as alpha 5 or a5) is a protein that, in humans, is encoded by the ITGA5 gene. The a5 chain undergoes post-translational cleavage in the extracellular domain to yield disulfide-linked light and heavy chains that join with beta 1 (βΐ) to form a fibronectin receptor referred to as α5β1.

Configurations of the antibody constructs: An antibody construct of the present invention can be configured in various ways, provided it selectively or specifically binds integrin alpha-V and integrin a5. Examples of useful constructs are illustrated in Figures 10 and 11. Examples of antibody fragments that may be incorporated in the present antibody constructs include but are not limited to Fv, Fab, Fab', Fab'-SH, F(ab')2;

diabodies; linear antibodies; and single-chain Fv antibody molecules (scFv). Methods of generating nucleic acid constructs that encode a bispecific antibody construct or a part thereof and methods of expressing those constructs are known in the art. See, e.g. , Spiess et al , (Mol. Immunol. 67:95-106, 2015); Han et al , (Sci. Rep. 7: 8360, 2017); Fan et al. (J. Hematol. Oncol. 8: 130, 2015); Kontermann (Acta Pharmacologica Sinica 26(1): 1-9, 2005); and Kontermann and Brinkmann (Drug Discovery Today 20(7):838-847, 2015).

An "antibody construct that binds to the same epitope" as a reference antibody refers to an antibody construct that 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 construct to its antigen in a competition assay by 50% or more. The details of the competition assay to be used for this determination are set out herein.

"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 herein.

The antibody constructs of the invention will be capable of binding integrin alpha-V and integrin a5 with sufficient affinity to be useful as diagnostic and/or therapeutic agents. In some embodiments, the extent of binding of an antibody construct of the invention to an unrelated protein (i.e., a protein other than integrin alpha-V or integrin a5) is less than about 10% of the binding of the antibody to integrin alpha-V or integrin a5 as measured, e.g., by a radioimmunoassay (RIA). In certain embodiments, an antibody that binds to integrin alpha-V or integrin a5 has a dissociation constant (KD) of < 1 μΜ, < 100 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), where KD is determined using a surface plasmon resonance technique (e.g., with a Biacore™ instrument) in which the antigen is

immobilized, the antibody serves as analyte, and the following conditions are used: 37 °C, 0.05% Tween® 20 polyethylene glycol sorbitan monolaurate, 20 mM N-(2-acetamido)-2- aminoethanesulfonic acid (ACES buffer), 150 mM NaCl, pH 7.4. In certain embodiments, an anti-integrin alpha-V, anti-integrin a5 antibody binds to an epitope of integrin alpha-V or integrin a5 that is conserved among those integrins from different species (e.g., human and non-human primate or human and murine).

An "affinity matured" antibody refers to an antibody with one or more alterations in the variable region, e.g., in one or more hypervariable regions (HVRs), compared to a parent antibody which does not possess such alterations, such alterations resulting in an improvement in the affinity of the antibody for antigen.

As used herein, a "bispecific antibody construct" comprises or consists of a protein or complex of proteins having a sequence or sequences that specifically and simultaneously bind two molecular targets. In the context of the present invention, the first target is integrin alpha-V and the second target is integrin a5 (alone or when heterodimerized with βΐ). The protein may be further configured and modified as described herein (e.g. , modified to include a detectable label). The bispecific antibody constructs of the invention can be more simply referred to as antibodies, antibody constructs or constructs. The term "construct" is used, in part because the compositions can be constructed using genetic engineering and can be non-naturally occurring. Accordingly, any of the antibody constructs described herein, nucleic acids that encode them, related expression vectors and host cells can be non-naturally occurring.

"Framework" or "FR" refers to variable domain residues other than hypervariable region (HVR) residues. The FR of a variable domain generally consists of four FR domains: FR1, FR2, FR3, and FR4. Accordingly, the HVR and FR sequences generally appear in the following sequence in VH (or VL): FR1-H1(L1)-FR2-H2(L2)-FR3-H3(L3)- FR4.

The term "hypervariable region" or "HVR" as used herein refers to each of the regions of an antibody variable domain that is hypervariable in sequence and/or forms a structurally defined loop ("hypervariable loop") and/or contains the antigen-contacting residues ("antigen contacts"). Generally, antibodies comprise six HVRs (also known as "complementarity determining regions" or "CDRs"): three in the VH (HI, H2, H3), and three in the VL (LI, L2, L3). Exemplary HVRs herein include: (a) hypervariable loops occurring at amino acid residues 26-32 (LI), 50-52 (L2), 91-96 (L3), 26-32 (HI), 53-55 (H2), and 96-101 (H3) (Chothia and Lesk, J. Mol. Biol. 196:901-917, 1987); (b) CDRs occurring at amino acid residues 24-34 (LI), 50-56 (L2), 89-97 (L3), 31-35b (HI), 50-65 (H2), and 95-102 (H3) (Kabat et al, Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, Md. (1991)); (c) antigen contacts occurring at amino acid residues 27c-36 (LI), 46-55 (L2), 89-96 (L3), 30-35b (HI), 47-58 (H2), and 93-101 (H3) (MacCallum et al. J. Mol. Biol. 262:132-145, 1996); and (d) combinations of (a), (b), and/or (c), including HVR amino acid residues 46-56 (L2), 47-56 (L2), 48-56 (L2), 49-56 (L2), 26-35 (HI), 26-35b (HI), 49-65 (H2), 93-102 (H3), and 94-102 (H3).

The term "knob-into-hole" or "KnH" technology as used herein refers to the technology directing the pairing of two polypeptides together in vitro or in vivo by introducing a protuberance (knob) into one polypeptide and a cavity (hole) into the other polypeptide at an interface where they interact. For example, KnHs have been introduced in the Fc:Fc binding interfaces, CL:CHI interfaces, and VH/VL interfaces of antibodies (see, e.g., US 2011/0287009, US2007/0178552, WO 96/027011, WO 98/050431, and Zhu et al, 1997, Protein Science 6:781-788). In some embodiments, KnHs drive the pairing of two different heavy chains together during the manufacture of multispecific antibodies. For example, multispecific antibodies having KnH in their Fc regions can further comprise single variable domains linked to each Fc region, or further comprise different heavy chain variable domains that pair with similar or different light chain variable domains. KnH technology can be also be used to pair two different receptor extracellular domains together or any other polypeptide sequences that comprise different target recognition sequences (e.g., including affibodies, peptibodies and other Fc fusions).

The term "knob mutation" as used herein refers to a mutation that introduces a protuberance (knob) into a polypeptide at an interface in which the polypeptide interacts with another polypeptide. In some embodiments, the other polypeptide has a hole mutation.

The term "hole mutation" as used herein refers to a mutation that introduces a cavity (hole) into a polypeptide at an interface in which the polypeptide interacts with another polypeptide. In some embodiments, the other polypeptide has a knob mutation.

As is known in the art, naturally occurring immunoglobulins include a heavy chain and a light chain, each of which is further divided into a constant region and a variable region. Within the variable regions, there are three hypervariable regions, also known as "complementarity-determining regions" or CDRs, interspersed within four "framework" regions (FRs). It is the CDRs that are primarily responsible for binding a target molecule, which may be an antigen. The sequence within the antibody that binds the target molecule is the "paratope," and the sequence within the target molecule that is bound is the

"epitope." The CDRs of each chain are typically referred to as CDR1, CDR2, and CDR3, numbered sequentially from the N-terminus, and the variable regions of the heavy and light chains can be abbreviated VH and VL, respectively. As the constructs of the invention are bispecific, they will include two paratopes. Accordingly, the first CDR in the variable region of the light chain of the first paratope may be abbreviated as "PIVLCDRI ." The second CDR would be "PIVLCDR2." The first CDR in the variable region of the heavy chain of the second paratope would be "P2VHCDRI ," and so forth.

A "single chain Fv" or "scFv" denotes a binding entity in which the variable regions of the heavy (VH) and light (VL) chains of a conventional antibody have been engineered to form one chain. A linker sequence is typically inserted between the VH and VL regions of an scFv to facilitate proper folding and creation of an active paratope.

Sequences: With regard to useful sequences, the present constructs can include CDRs, FRs, and biologically active fragments or other variants thereof from previously generated or newly generated immunoglobulins. For example, the CDRs within a paratope that binds integrin alpha-V can include those described in U.S. Patent No. 8,562,986, the entire content of which is hereby incorporated by reference herein. The CDRs within a paratope that binds integrin a5 or α5β1 can include those described in U.S. Patent No. 7,276,589, the content of which, as noted above, is incorporated herein by reference.

"Percent (%) amino acid sequence identity" with respect to a reference polypeptide sequence is defined as the percentage of amino acid residues in a candidate sequence that are identical to corresponding amino acid residues in the reference polypeptide sequence, 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 aligning sequences, 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 from Genentech, Inc., South San Francisco, Calif, or may be compiled from the source code. The ALIGN-2 program should be compiled for use on a UNIX operating system, including digital UNIX 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.

A "human antibody" is one that possesses an amino acid sequence that corresponds to that of an antibody produced by a human or a human cell or derived from a non-human source that utilizes human antibody repertoires or other human antibody-encoding sequences. It need not be a naturally-occurring antibody. This definition of a human antibody specifically excludes a humanized antibody comprising non-human antigen- binding residues.

A "human consensus framework" is a framework that represents the most commonly occurring amino acid residues in a selection of human immunoglobulin VL or VH framework sequences. Generally, the selection of human immunoglobulin VL or VH 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, NIH Publication 91-3242, Bethesda Md. (1991), vols. 1-3. In some embodiments, for the VL, the subgroup is subgroup kappa I as in Kabat et al, supra. In some embodiments, for the VH, the subgroup is subgroup III as in Kabat et al, supra.

A "humanized" antibody refers to a chimeric antibody comprising amino acid residues from non-human HVRs and amino acid residues from human FRs. In certain embodiments, a humanized antibody will comprise substantially all of at least one, and typically two, variable domains, in which all or substantially all of the HVRs (e.g., CDRs) correspond to those of a non-human antibody, and all or substantially all of the FRs correspond to those of a human antibody. A humanized antibody optionally may comprise at least a portion of an antibody constant region derived from a human antibody. A

"humanized form" of an antibody, e.g., a non-human antibody, refers to an antibody that has undergone humanization. Immunoglobulins generated in a non-human animal (e.g. , a rodent or non-human primate) can be humanized and/or de-immunized according to standard methods. In some instances, the humanized and/or de-immunized antibody constructs will be chimeric, in that they will include sequences (e.g., variable domain sequences) found in the non-human immunoglobulins from which they were generated as well as sequences (e.g., constant domain sequences) found in human immunoglobulins. In other embodiments, the antibody constructs of the present invention can include CDRs and FRs from a fully human immunoglobulin.

To de-immunize an antibody, one can identify and remove human T-cell epitopes from the original immunoglobulin (e.g. , a mouse antibody), replacing them with alternate sequences that are not human T-cell epitopes. De-immunization differs from humanization, which replaces original immunoglobulin (e.g., mouse antibody) sequences with human consensus sequences. De-immunization techniques useful in making the present constructs can be found, for example, in WO 98/52976, WO 00/34317 and WO 02/69232.

Additional Moieties: The bispecific antibody constructs described herein can also include one or more additional moieties. For example, the constructs can include a detectable label and/or a moiety that extends the construct's circulating half life (e.g., modified human serum albumin). A "detectable label" is a composition detectable by spectroscopic, photochemical, biochemical, immunochemical, chemical, or other physical means. For example, a detectable label within a bispecific antibody construct of the present invention can be a fluorescent dye, an electron-dense reagent, a radioisotope, an enzyme (e.g., an enzyme commonly used in an ELISA), biotin, digoxigenin, a hapten, or any other protein or non-protein entity that can be made detectable. The radioisotope may be, for example, ¾, ¹⁴ C, ² P, ⁵ S, or ¹²⁵ I. In some cases, radioisotopes can be employed not only because they are detectable, but also because they can be toxic to cells in their vicinity. The detectable labels may be incorporated into the bispecific antibody constructs at a number of positions, and any method known in the art for conjugating the construct to the label may be employed. For example, one could use a method described by Hunter et al. {Nature, 144:945, 1962); David et al. {Biochemistry, 13: 1014, 1974); Pain et al. {J.

Immunol. Meth., 40:219, 1981); or Nygren {J. Histochem. and Cytochem., 30:407, 1982). The lifetime of radiolabeled antibody constructs may be extended by adding a substance that stabilizes the radiolabeled construct by, for example, protecting it from degradation. Any substance or combination of substances that stabilizes the construct may be used, including those disclosed in U. S. Patent No. 5,961,955.

An antibody construct may be isolated. An "isolated" antibody is one that has been separated from a component of its natural environment. 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 review of methods for assessment of antibody purity, see, e.g., Flatman et al, J. Chromatogr. B 848:79-87 (2007).

Nucleic Acids and Expression Vectors:

The nucleic acids and expression vectors described herein may be "isolated" by virtue of being separated from a component of their natural environment (insofar as that may exist). 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.

Host Cells and Expression Systems: Host cells are transfected or transformed with expression or cloning vectors described herein for bispecific antibody production and cultured in conventional nutrient media modified as appropriate for inducing promoters, selecting transformants, or amplifying the genes encoding the desired sequences. One of ordinary skill in the art can, without undue experimentation, select the culture conditions, such as medium, temperature, pH and the like. In general, principles, protocols, and practical techniques for maximizing the productivity of cell cultures can be found in Mammalian Cell Biotechnology: a Practical Approach, M. Butler, ed. (IRL Press, 1991) and Sambrook et al, supra.

Methods of eukaryotic cell transfection and prokaryotic cell transformation are known to the ordinarily skilled artisan, for example, those that use CaCh or CaPCn, those that are liposome-mediated, and electroporation. Depending on the host cell used, transformation is performed using standard techniques appropriate to such cells. The calcium treatment employing calcium chloride, as described in Sambrook et al, supra, or electroporation is generally used for prokaryotes. Infection with Agrobacterium

tumefaciens is used for transformation of certain plant cells, as described by Shaw et al. {Gene, 23:315, 1983) and WO 89/05859 published June 29, 1989. For mammalian cells without such cell walls, the calcium phosphate precipitation method of Graham and van der Eb {Virology, 52:456-457, 1978) can be employed. General aspects of mammalian cell host system transfections have been described in U.S. Patent No. 4,399,216.

Transformations into yeast are typically carried out according to the method of Van Solingen et al. {J. Bact, 130:946, 1977) and Hsiao et al. {Proc. Natl. Acad. Sci. USA, 76:3829, 1979). However, other methods for introducing DNA into cells, such as by nuclear microinjection, electroporation, bacterial protoplast fusion with intact cells, or poly cations, e.g., polybrene, poly ornithine, may also be used. For various techniques for transforming mammalian cells, see Keown et al. {Methods in Enzymology, 185:527-537, 1990) and Mansour et al. {Nature, 336:348-352, 1988).

Suitable host cells for cloning or expressing the DNA in the vectors herein include prokaryote, yeast, or higher eukaryote cells. Suitable prokaryotes include but are not limited to eubacteria, such as Gram-negative or Gram-positive organisms, for example, Enterobacteriaceae such as E. coli. Various E. coli strains are publicly available, such as E. coli K12 strain MM294 (ATCC 31,446); E. coli XI 776 (ATCC 31,537); E. coli strain

W3110 (ATCC 27,325) and K5 772 (ATCC 53,635). Other suitable prokaryotic host cells include Erwinia, Klebsiella, Proteus, Salmonella, e.g., Salmonella typhimurium, Serratia, e.g., Serratia marcescans, and Shigella, as well as Bacilli such as B. subtilis and B.

licheniformis (e.g., B. licheniformis 41P disclosed in DD 266,710 published April 12, 1989), Pseudomonas such as P. aeruginosa, and Streptomyces. These examples are illustrative rather than limiting. Strain W3110 is one particularly preferred host or parent host because it is a common host strain for recombinant DNA product fermentations.

Preferably, the host cell secretes minimal amounts of proteolytic enzymes. For example, strain W3110 may be modified to effect a genetic mutation in the genes encoding proteins endogenous to the host, with examples of such hosts including E. coli W3110 strain 1 A2, which has the complete genotype tonA; E. coli W3110 strain 9E4, which has the complete genotype tonA ptr3; E. coli W3110 strain 27C7 (ATCC 55,244), which has the complete genotype tonA ptr3 phoA El 5 (argF-lac)169 degP ompT kan ^r ; E. coli W3110 strain 37D6, which has the complete genotype tonA ptr3 phoA E15 (argF-lac)169 degP ompT rbs7 ilvG kan ^r ; E coli W3110 strain 40B4, which is strain 37D6 with a non-kanamycin resistant degP deletion mutation; and an E. coli strain having mutant periplasmic protease disclosed in U.S. Patent No. 4,946,783 issued August 7, 1990. Alternatively, in vitro methods of cloning, e.g., PCR or other nucleic acid polymerase reactions, are suitable.

In addition to prokaryotes, eukaryotic microbes such as filamentous fungi or yeast are suitable cloning or expression hosts for ADAM8-encoding vectors. Saccharomyces cerevisiae is a commonly used lower eukaryotic host microorganism. Others include Schizosaccharomyces pombe (Beach and Nurse, Nature, 290: 140, 1981); EP 139,383 published May 2, 1985); Kluyveromyces hosts (U.S. Patent No. 4,943,529; Fleer et al, Bio/Technology, 9: 968-975, 1991)) such as, e.g., K. lactis (MW98-8C, CBS683, CBS4574; Louvencourt et al. (J. Bacteriol, 737, 1983), K. fragilis (ATCC 12,424), K. bulgaricus (ATCC 16,045), K. wickeramii (ATCC 24,178), K. waltii (ATCC 56,500), K.

drosophilarum (ATCC 36,906; Vanden Berg et al. {Bio/Technology, 8: 135, 1990), K.

thermotolerans, md K. marxianus; yarrowia (EP 402,226); Pichia pastoris (EP 183,070; Sreekrishna et al. (J. Basic Microbiol., 28:265-278, 1988); Candida; Trichoderma reesia (EP 244,234); Neurospora crassa (Case et al, Proc. Natl. Acad. Sci. USA, 76:5259-5263, 1979); Schwanniomyces such as Schwanniomyces occidentalis (EP 394,538 published October 31, 1990); and filamentous fungi such as, e.g., Neurospora, Penicillium,

Tolypocladium (WO 91/00357 published January 10, 1991), and Aspergillus hosts such as A. nidulans (Ballance et al, Biochem. Biophys. Res. Commun., 112:284-289, 1983; Tilburn et al. (Gene, 26:205-221, 1983; Yelton et al, Proc. Natl. Acad. Sci. USA, 81: 1470-1474, 1984) and .4. niger (Kelly and Hynes, EMBO J. , 4:475-479, 1985). Methylotropic yeasts are suitable herein and include, but are not limited to, yeast capable of growth on methanol selected from the genera consisting of Hansenula, Candida, Kloeckera, Pichia,

Saccharomyces, Torulopsis, and Rhodotorula. A list of specific species that are exemplary of this class of yeasts may be found in C. Anthony, The Biochemistry of Methylotrophs, 269 (1982).

Suitable host cells for the expression of glycosylated antibodies are derived from multicellular organisms. Examples of invertebrate cells include insect cells such as Drosophila S2 and Spodoptera Sf9, as well as plant cells. Examples of useful mammalian host cell lines include Chinese hamster ovary (CHO) and COS cells. More specific examples include monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); human embryonic kidney line (293 or 293 cells subcloned for growth in suspension culture, Graham et al, J. Gen Virol, 36:59 (1977)); Chinese hamster ovary cells/-DHFR (CHO, Urlaub and Chasin, Proc. Natl. Acad. Sci. USA, 77:4216 (1980)); mouse Sertoli cells (TM4, Mather, Biol. Reprod., 23:243-251 (1980)); human lung cells (W138, ATCC CCL 75); human liver cells (Hep G2, HB 8065); and mouse mammary tumor (MMT 060562, ATCC CCL51). The selection of the appropriate host cell is deemed to be within the skill in the art.

The terms "host cell," "host cell line," and "host cell culture," which may be used interchangeably, refer to cells into which exogenous nucleic acid (e.g. , a nucleic acid molecule or expression vector, as described herein) 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, as they may contain mutations. Mutant progeny that have substantially the same function or biological activity as screened or selected for in the originally transformed cell are included herein.

Pharmaceutical Compositions and Kits:

As noted, the invention features kits that includes a bispecific antibody construct as described herein and instructions for use. The instructions may be in the form of a "package or packet insert," which refers to information and instructions customarily included in commercial packages of therapeutic products, that contains information about the indications, usage, dosage, administration, combination therapy, contraindications, and/or warnings concerning the use of such therapeutic products. The term "package insert" is also used to refer to instructions customarily included in commercial packages of diagnostic products that contain information about the intended use, test principle, preparation and handling of reagents, specimen collection and preparation, calibration of the assay and the assay procedure, performance and precision data such as sensitivity and specificity of the assay.

Methods of Treatment: The compositions described herein can be used to inhibit angiogenesis whenever it is harmful (e.g. , in the context of vascularizing a tumor). While the compositions described herein are not limited to those that achieve a clinical result through any particular cellular mechanism, it is expected that directly targeting integrins expressed by cancer cells or other cells within a tumor (e.g., stromal cells such as cancer- associated fibroblasts) will block the adhesion, survival, proliferation, invasion, and/or migration of cancer cells. The constructs may also enhance the function of the immune system by blocking immunosuppressive signaling in the cancer-associated

microenvironment.

"Patient response" or "response" (and grammatical variations thereof) can be assessed using any endpoint indicating a benefit to the patient, including, without limitation, (1) inhibition, to some extent, of disease progression, including slowing down and complete arrest; (2) reduction in the number of disease episodes and/or symptoms; (3) reduction in lesional size; (4) inhibition (i.e., reduction, slowing down or complete stopping) of disease cell infiltration into adjacent peripheral organs and/or tissues; (5) inhibition (i.e. reduction, slowing down or complete stopping) of disease spread; (6) increase in host immune surveillance that results in regression or ablation of the disease lesion; (7) relief, to some extent, of one or more symptoms associated with the disorder; (8) increase in the length of disease-free presentation following treatment; and/or (9) decreased mortality at a given point of time following treatment. When a patient response is elicited, the antibody construct and/or the formulation of which it is a part are being used/administered to achieve a clinically beneficial outcome.

An "individual," "patient," or "subject" is a mammal. Mammals include, but are not limited to, domesticated animals (e.g., cows, sheep, cats, dogs, pigs, and horses), primates (e.g., humans and non-human primates such as monkeys), rabbits, and rodents (e.g., mice and rats). In certain embodiments, the individual, patient, or subject is a human.

Binding Assays and Other Assays: In some embodiments, an antibody construct provided herein is tested for its antigen binding activity, e.g., by known methods such as ELISA or Western blotting.

In some embodiments, competition assays may be used to identify an antibody that competes with an anti-integrin alpha-V, anti-integrin a5 antibody construct as described herein for binding to the respective integrins. In some embodiments, competition assays may be used to identify an antibody that competes with a bispecific antibody construct described herein for binding to integrin alpha-V and integrin a5. In certain embodiments, such a competing antibody binds to one of the same epitopes (e.g., linear or conformational epitopes) bound by an antibody construct described herein. Detailed exemplary methods for mapping an epitope to which an antibody binds are provided in Morris (1996) "Epitope Mapping Protocols," in Methods in Molecular Biology vol. 66 (Humana Press, Totowa, N.J.).

In an exemplary competition assay, immobilized integrin alpha-V is incubated in a solution comprising a first labeled antibody that binds to integrin alpha-V (e.g., a bispecific antibody construct described herein) and a second unlabeled antibody that is being tested for its ability to compete with the first antibody for binding to integrin alpha-V. The second antibody may be present in a hybridoma supernatant or may be purified. As a control, immobilized integrin alpha-V 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 integrin alpha-V, excess unbound antibody is removed, and the amount of label associated with immobilized integrin alpha-V is measured. If the amount of label associated with immobilized integrin alpha-V 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 integrin alpha- V. See Harlow and Lane (1988) Antibodies: A Laboratory Manual ch. 14 (Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.).

In a further exemplary competition assay, immobilized integrin a5 is incubated in a solution comprising a first labeled antibody that binds to integrin a5 and a second unlabeled antibody that is being tested for its ability to compete with the first antibody for binding to integrin a5. The second antibody may be present in a hybridoma supernatant or may be purified. As a control, immobilized integrin a5 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 integrin a5, excess unbound antibody is removed, and the amount of label associated with immobilized integrin a5 is measured. If the amount of label associated with immobilized integrin a5 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 integrin a5.

A number of anti-integrin antibodies with specificity for α5, α5β1, or aV are known in the art. Some of these are described in the tables below. CDRs or variable domains or entire heavy /light chains from any prior art anti-a5 or α5β1 or βΐ antibody, or from any newly generated anti-a5 or α5β1 or βΐ antibody, could substitute for the anti-o^l antibody sequences utilized in the embodiment described in the Examples section. Likewise, CDRs or variable domains or entire heavy /light chains from any prior art anti-aV or ανβ3 antibody, or from any newly generated anti-aV or ανβ3 antibody, could substitute for the aV antibody sequences utilized in the embodiment described in the Examples section.

Table 1; Anti-AV, anti-AVB3, anti-A5, and anti-A5Bl Antibodies

Antigen

Specificity Sequence, Deposit Number, or Other Information Citation

MOR04624 VLK

diqmtqspsslsasvgdrvtitcrasqgissnlnwyqqkpgkapklliyaa

snlqsgpsrfsgsgsgtdftltisslq

pedfavyycqqysdqsytfgqgtkveikrt (SEQ ID NO:77)

VH

qvqlvesggglvqpggslrlscaasgftfssygmswvrqapgkglewvs

sisysdsnt yadsvkgrftisrdns

kntlylqmnslraedtavyycarglgdyghhhglsgifdywgqgtlvtvss

(SEQ ID NO:78)

MOR04055 VU3

dieltqppsvsvapgqtariscsgdsigeqyahwyqqkpgqapvlviydd

nk sgiperfsgsnsgntatltis

gtqaedeadyycgsytltntasvfgggtkltvlg (SEQ ID NO:79)

VH3

qvqlvesggglvqpggslrlscaasgftfsnyamnwvrqapgkglewvs

risysgsdtyyadsvkgrftisrdnskntlylqmnslraedtavyycarege

fgfmystlvfdswgqgtlvtvss (SEQ ID NO: 80)

MOR04971 VU3

dieltqppsvsvapgqtariscsgdsigeqyahwyqqkpgqapvlviydd

nk sgiperfsgsnsgntatltis

gtqaedeadyycssytyssdasvfgggtkltvlg (SEQ ID NO: 81)

VH3

qvqlvesggglvqpggslrlscaasgftfsnyamnwvrqapgkglewvs

aihdnghtyypdsvkgrftisrdnskntlylqmnslraedtavyycarege

fgfmystlvfdswgqgtlvtvss (SEQ ID NO: 82)

MOR04974 WO200713487 a5bl 6A2 VLK

diqmtqspsslsasvgdrvtitcrasqgissnlnwyqqkpgkapklliyaa snlqsgpsrfsgsgsgtdftltisslq

pedfatyy cqqy asprqtfgqgtkveikrt

(SEQ ID NO: 83)

VH

qvqlvesggglvqpggslrlscaasgftfssygmswvrqapgkglewvs girakqsgyatdyaapvkgrftisrctaskntlylqmnslraedtavyycarg lgdyghhhglsgifdywgqgtlvtvss (SEQ ID NO: 84)

MOR04975

VLK

diqmtqspsslsasvgdrvtitcrasqgissnlnwyqqkpgkapklliyaa snlqsgpsrfsgsgsgtdftltisslq

pedfatyy cqqy efgiqtfgqgtkveikrt

(SEQ ID NO: 85)

VH

qvqlvesggglvqpggslrlscaasgftfssygmswvrqapgkglewvs girakqsgyatdyaapvkgrftisrdnskntlylqmnslraedtavyycarg lgdyghhhglsgifdywgqgtlvtvss (SEQ ID NO: 86)

MOR04977

VLK

diqmtqspsslsasvgdrvtitcrasqgissnlnwyqqkpgkapklliyaa snlqsgpsrfsgsgsgtdftltisslq

pedfatyy cqqyssnpqtfgqgtkveikrt (SEQ ID NO: 87) VH

qvqlvesggglvqpggslrlscaasgftfssygmswvrqapgkglewvsf iepkwrggathyaasvkgrftisrdnskntlylqrnnslraedtavyycargl gdyghhhglsgifdy wgqgtlvtvss (SEQ ID NO: 88)

MOR04985

VLK

diqmtqspsslsasvgdrvtitcrasqgissnlnwyqqkpgkapklliyaa snlqsgpsrfsgsgsgtdftltisslq

pedfavyycqqysdqsytfgqgtkveikrt (SEQ ID NO: 89)

VH

qvqlvesggglvqpggslrlscaasgftfssygmswvrqapgkglewvs

girakqsgyatdyaapvkgrftisrdnskntlylqmnslraedtavyycarg

Igdyghhhglsgifdywgqgtlvtvss (SEQ ID NO:90)

CN104994874B

BETA 1 IgG; specific for mouse, rat, human beta 1 EP2938359A1

Mitjans et al., J.

Cell Science

108:2825-2838

Av hybridoma line 271-14D9.F8, IgGl, kappa (1995)

Mitjans et al., J.

Cell Science

108:2825-2838

Av hybridoma line 271-20A9; IgGl, kappa (1995) Mitjans et al., J.

Cell Science 108:2825-2838

Av hybridoma line 271-23G5; IgGl, kappa (1995)

World Wide Web abcam.com/inte grin-alpha-v-

Immunogen Clone EPR 16800; IgG; Binds to antibody- recombinant fragment within Human Integrin alpha V eprl6800-

Av aa 1-250; also mouse and rat abl79475.html

5 avB3-binding monoclonal abs (17E6, 10C4, 23C6, LM609, 69-6-5) were shown to bind to epitopes in the Kamata et al, B-propeller domain (ligand binding). 4 are shown to PLoS One 8(6):

AvB3 bind the thigh domain (AMF7, M9, P2W7, P3G8). e66096 (2013)

Table 2 Commercial Antibodies

Species

Antigen Supplier specificity Isotype Source

World Wide Web

sigmaaldrich.com/catalog/product/ mm/CBL497?lang=en&region=US f¾S

&utm_medium=referral&utm_sour ce=biocompare 1 &utm_campaign=b

Sigma Human IgG2b iocompare 2017

World Wide Web

lsbio.com/antibodies/anti-itga5-

«5 antibody-integrin-alpha-5-antibody-

LifeSpan cd49e-antibody-wb-western-ls- Biosciences Human IgG3 cl3705/14446?trid=247

World Wide Web

sigmaaldrich.com/catalog/product/ mm/MAB1956Z?utm_source=bioc

Sigma Human IgG3 ompare 1 &utm campaign=biocomp are 2017&utm medium=referral%

C2%A0%20%C2%A0%C2%A0

World Wide Web

lsbio.com/antibodies/anti-itga5- antibody-integrin-alpha-5-antibody- cd49e-antibody-flow-if-

LifeSpan immunofluorescence-ls- Biosciences Human IgG2b cl3678/14419?trid=247

World Wide Web

rndsy stems . com/ products/human- integrin-alpha5-cd49e-antibody- α.5 238307_mabl864?utm_source=bio compare&utm_medium=referral&u tm_campaign=product&utm_term=

R&D Human IgGl primary antibodies

World Wide Web

origene.com/catalog/antibodies/pri mary-antibodies/am32340pu- f¾S

n/integrin-alpha-5-itga5-mouse- monoclonal-antibody-clone-id-clb-

OriGene Human IgG2b 705

World Wide Web

bosterbio.com/anti-integrin-alpha-

5-picoband-trade-antibody-pb9254-

«5β1 boster.html?utm_source=listing%3

Human, Abiocompare&utm_medium=paid Mouse, %2Blisting%2Bppc&utm_campaig

Boster Rat n=product%2Blist%3A#details

World Wide Web

sigmaaldrich.com/catalog/product/ mm/MAB2514?lang=en&region=U

S&utm_medium=referral%A0%20

%AO%AO&utm_source=biocompar el&utm campaign=biocompare 20

Sigma Human XgG3 17

World Wide Web

sigmaaldrich.com/catalog/product/ mm/MAB 1969?lang=en&region=U

S&utm_medium=referral%A0%20

%AO%AO&utm_source=biocompar el&utm campaign=biocompare 20

Sigma Human ¾G1 17 World Wide Web

sigmaaldrich.com/catalog/product/ mm/MAB2575?lang=en&region=U οδβΐ S&utm_medium=referral%A0%20

%AO%AO&utm_source=biocompar

Human, el&utm campaign=biocompare 20

Sigma Rat igG2c 17

World Wide Web

οδβΐ Εηζο Life enzolifesciences.com/product.php?

Sciences Human IgG4k pid=ENZ-ABS385

World Wide Web antibodies- online, com/ antibody/228615/ anti- Integrin+alpha+5+beta+ 1 +antibody /?utm_source=partner&utm_mediu m=biocompare&utm_campaign=no

Human, n_sponsored&utm_content=primar

Ab Online Monkey IgG y oem&utm term=ABIN228615

World Wide Web

bdbiosciences.eom/us/p/5556507ut aSpl m_source=biocompare&utm_medi

BD Human, IgGl um=catalog&utm_campaign=bioco

Biosciences Pig Kappa mparecat

World Wide Web

sigmaaldrich.com/catalog/product/ mm/MABT207?lang=en&region=

US &utm_medium=referral% A0%2

0%AO%AO&utm_source=biocomp

Millipore IgGl arel&utm campaign=biocompare αν Sigma Human kappa 2017

World Wide Web

lsbio.com/antibodies/anti-itgav- antibody-integrin-alpha-v-antibody-

LifeSpan cd51 -antibody-flow- wb-western-ls- αν Biosciences Human IgGl cl3739/14480?trid=247

World Wide Web

abeam. com/Integrin-alpha-V- antibody-23C6-Low-endotoxin-

Azide-free- abl 85741.html?utm_source=bioco mpare&utm_medium=paid_referral &utm_term=l ry_3311_185741 &ut αν Abeam Human IgGl m campaign=editorial

World Wide Web

origene.com/catalog/antibodies/pri αν OriGene Human IgG2a mary-antibodies/am32337pu- n/cd51 -itgav-mouse-monoclonal- antibody-clone-id-nki-m9-former- paf2a

World Wide Web

STEMCELL IgG2a stemcell .com/ products/ anti-human- αν Technologies Human kappa cd51 -antibody-clone-nki-m9.html

World Wide Web

Enzo Life enzolifesciences.com/product.php? αν Sciences Human IgG l pid=ALX-803-304

The invention will be further illustrated with the following non-limiting examples. These examples disclose, among other things, that a bispecific antibody construct that simultaneously binds and inhibits integrin alpha-V and integrin α5β1 is more efficacious than a pair of distinct antibodies, one of which targets the alpha-V integrin and the other of which targets the α5β1 integrin.

EXAMPLES

EXAMPLE 1

A prototype bispecific, tetravalent antibody construct designated "ITGA5BlxAV" was designed with the format shown in FIG. 10, and was prepared by standard recombinant methods. The Fab portion was designed based on the variable domains of Abituzumab, and the scFv was designed based on the variable domains of Volociximab. Genes encoding the two amino acid sequences shown in FIG. 12A (heavy chain plus scFv) and FIG. 12B (light chain) were prepared and cloned into the expression vector via Hindlll/Notl. Two plasmids were isolated and subjected to transient expression with 293 cells as the host. The supernatant was collected and the antibody purified using a protein A affinity column.

The bispecific antibody demonstrated enhanced potency in a range of cancer cell adhesion and migration assays. In addition, in the C4-2b prostate cancer line, the bispecific antibody construct reduced cell viability more potently than the monospecific antibodies alone and induced effects equivalent to those induced by the combination of the two monospecific antibodies, but did so at one-fifth the dose of the combination. In studies of endothelial cell migration induced by tumor-stromal cell interactions, the bispecific antibody is particularly potent in blocking endothelial cell migration and is superior to bevacizumab. Taken together, the below data indicate that this novel bispecific antibody has unique and potentbiological activity in a range of experimental cancer systems.

Other Antibodies: Other antibodies were sourced as follows: monoclonal a5, alpha-V (Millipore MAB1956Z and MABT207), heterodimeric a5bl : (Alpha5/betal 7H5.4.2.8 hybridoma (ATCC No. PTA-7421)).

Cell migration assay: Cell migration assays were performed using a Boy den Chamber set up. A 24-well assay plate was prepared by adding 800 of chemoattractant underneath uncoated 8-micron pore transwell inserts (Corning 353097) and heated to 37°C. 25,000 cells (harvested in log-phase growth) were added from a homogenous cell suspension to the top of the well to a total volume of 200 μ The assay was stopped after 24 hours at 37°C by removal and washing of the membrane inserts with PBS and formalin (Fischer Scientific SF100-4) fixation of the cells for 15 minutes. Following another 3x PBS wash, cells were removed from the tops of the membranes using cotton tip swabs and were stained using calcein AM (ThermoFisher C1430). After drying, membranes were mounted onto microscope slides and cell counts were performed on 3-5 representative fields of a membrane using a 1 Ox or 20x objective.

Cell adhesion assay: A 96-well plate was plated with 50 xL of adherent substrate in triplicate and incubated at 37°C for 1 hour. Non-specific binding sites were blocked by incubating the wells with 1% BSA for 30 minutes at 37°C. After washrg the wells 3x with PBS, 10,000 cells were harvested in log-phase growth and incubated for 20 minutes on ice with the indicated dose of integrin-neutralizing antibody, added in 100 of serum-free culture media to each well from a homogenous cell suspension. The plate was incubated for 60-90 minutes at 37°C to allow establishment of adherence profiles. To remove nonadherent cells, the wells were gently washed 5x with PBS. Adherent cell counts were assayed using Cell Counting Kit 8 (CCK-8) reagent (Bimake).

Endothelial cell migration assay: hBM-MSCs (human bone-marrow derived mesenchymal stromal cells) were cultured at confluence for 24 hours in a 24-well format. 20,000 cancer cells were seeded into the wells and co-cultured overnight. Then, an assay for cell migration towards the co-culture was performed as described above, using

10,000 HUVECs (human umbilical vein endothelial cells). Cell viability assay: The effect of bispecific ITGA5V neutralization on cell viability was monitored by cell viability assay using Cell Counting Kit-8 (CCK-8)

(Bimake) cell viability reagent. Briefly, cells were incubated with designated ITG antibody concentration and then seeded in a 96-well microtiter plate with replications. Cell viability was measured with GloMax-Multi Microplate Reader (Promega) quantitatively by recording the absorbance at 450 nm. Percent cell viability (%) was calculated and shown as a ratio of absorbance in ITG antibody treated cells to absorbance in IgG control cells after subtracting the average absorbance of background.

Flow Cytometry: Cells with variable membrane integrin a5 and alpha-V expression were incubated with monoclonal a5 or alpha-V antibodies (Millipore MAB1956Z and

MABT207), the bispecific antibody construct ITGA5BlxAV, or isotype control, for 1 hour at room temperature followed by 1 hour of incubation with a secondary-conjugated antibody, before being analyzed by flow cytometry.

Statistics: The mean fluorescent intensities (MFI) for the a5, alpha-V, and bispecific ITGA5BlxAV stained cell populations for each cell line were calculated from the flow cytometry data. Then a correlation coefficient between the sum of a5 and alpha-V MFIs and the bispecific ITGA5BlxAV was calculated.

The results indicate that, although integrin a5 blockade is sufficient to abrogate migration in PC-3 prostate cancer cells, blockade of both integrin a5 and integrin alpha-V is required to abrogate migration in DU-145 prostate cancer cells. As shown in FIGs. 1 A and IB, there is heterogeneity in prostate cancer cell interactions with the hBM-MSC secretome. PC-3 (FIG. 1A) and DU-145 (FIG. IB) prostate cancer cells were pre-treated with 50 μg/mL of the indicated integrin neutralizing antibody or a matched isotype control (Iso), and then tested in either a migration assay (movement across an 8 μιτι pore polypropylene membrane for 24 hours) or adhesion assay (1 hour) to hBM-MSC conditioned media (CM). The antibodies targeting a5 and βΐ integrins inhibited migration and adhesion of PC-3 cells, while the other antibodies tested, including an antibody targeting alpha-V, had no statistically significant effect (FIG. 1A). With DU-145 cells, blocking both a5 integrin and alpha-V integrin was required to neutralize adhesion and migration (FIG. IB). Endothelial cell migration is the earliest required phase of angiogenesis that can be studied in vitro. Using the endothelial cell migration assay described above, the role of a5 and alpha-V integrins in mediating these responses was studied in experiments with co- cultured PC-3 and hBM-MSCs. As shown in in FIGs. 2A and 2B, the endothelial cells (HUVECs) exhibited a chemotactic response to PC-3-hBM-MSC co-cultures (see the fourth bar in each of FIGs. 2A and 2B) that was inhibited PC-3 a5 neutralization (FIG. 2A) or HUVEC alpha-V neutralization (FIG. 2B). Neutralization was performed by pre- treatment with 50 μg/mL of integrin neutralizing antibody or matched isotype control (Iso). Thus, endothelial cell migration induced by prostate cancer-hBM-MSC interaction is regulated by integrins a5 and alpha-V.

A bispecific antibody for α5β1 (obligate heterodimeric partners) and alpha-V (ITGA5BlxAV) was generated and its purity demonstrated by SDS-PAGE (FIG. 3A). To demonstrate binding specificity, flow cytometric binding studies were conducted with monoclonal a5, alpha-V, and bispecific ITGA5BlxAV across a panel of cell lines with variable integrin a5 and alpha-V expression density. The sum of integrin a5 and integrin alpha-V binding (MFI) was highly correlated with bispecific binding (correlation .9187; FIG. 3B).

The bispecific antibody construct ITGA5BlxAV was found to be superior to either an integrin a5 -neutralizing antibody or an integrin alpha-V neutralizing antibody in inhibiting prostate cancer cell adhesion and migration (see FIG. 4), and we have also found it to be superior to a combination (mixture) of integrin a5 and integrin alpha-V antibodies (see FIG. 5). See also FIGs. 6A and 6B, summarizing data showing that the bispecific antibody construct ITGA5BlxAV is superior to the combination of anti-integrin a5 and anti-integrin alpha-V antibodies in inhibiting endothelial cell migration in co-cultures of prostate cells and hBM-MSC cells.

The bispecific antibody was also found to be superior to the combination of two monospecific antibodies that target the same integrins (aV and alpha-V) in neutralizing adhesion and migration in other cancer cell types, including the glioblastoma cell line A- 172, the breast cancer cell line MDA-MB-231, the cervical cancer cell line C-33A, and the uterine cancer cell line SKUT1 (FIGs. 7A and 7B). The bispecific antibody construct was more effective in inhibiting prostate cancer cell survival than either an anti-integrin a5 antibody, an anti-integrin alpha-V antibody, or those two monospecific antibodies used in combination. C4-2B cells were incubated with the bispecific antibody construct or with the anti-integrin a5 antibody and/or the anti- integrin alpha-V antibody, as indicated in FIGs. 8A and 8B. The antibodies were incubated with the cells at 4°C for 30-60 minutes and were then seeded (in triplicate) into 96-well plates and maintained in a humidified 37°C incubator with 5% C02 for 24-48 hours before performing the cell viability assay. The bispecific antibody construct was shown to be potent at one-fifth the concentration of the combination of monospecific antibodies

targeting α5β1 and alpha-V integrin.

When the bispecific antibody construct was compared to bevacizumab, an antibody that targets VEGF, it was found that the bispecific antibody construct was superior to bevacizumab in inhibiting endothelial cell migration to prostate-cancer-hBM-MSC co- culture (FIG. 9). HUVEC or DU-145 cells were treated with the doses of the antibodies indicated in FIG. 9 prior to assaying HUVEC migration to DU145/hBM-MSC.

EXAMPLE 2

The ability of the bispecific antibody ITGA5BlxAV to block adhesion, migration, cell survival and induction of endothelial migration in co-culture with stromal cells was compared to the ability of monospecific antibodies targeting a5 integrin (or its obligate α5β1 heterodimer) or integrin alpha v (or the ανβ3 heterodimer), both alone and in combination, to block those functions, using α5/αν co-expressing cells from diverse tumor types including prostate, breast, glioma, cervix, and uterine cancer.

Results: While combined a5 and av neutralization with dual monospecific antibodies was superior to individual single agents in blocking adhesion, migration, and induction of endothelial chemotaxis across diverse tumor types that co-expressed a5 and av integrins, the bispecific antibody ITGA5BlxAV was significantly superior to the combined monospecific antibodies in these assays (FIGS. 7A and 7B). In addition, a significant reduction in cell survival was noted in selected tumor types with ITGA5BlxAV (FIGS. 8A and 8B). Strikingly, ITGA5BlxAV was significantly more potent than bevacizumab in the inhibition of endothelial migration induced by tumor-stromal cell interactions (FIG. 9).

Targeting tumor, stromal and endothelial cells simultaneously with ITGA5BlxAV represents a potentially effective therapeutic strategy for targeting diverse mechanisms of progressive disease in the tumor microenvironment.

EXAMPLE 3

The mechanism of action of the ITGA5BlxAV antibody was further explored by assessing the comparative fates of integrins on various cell types in vitro following treatment with the ITGA5BlxAV antibody, one of the individual monospecific antibodies, or a combination of the two monospecific antibodies.

PC-3 cells were treated with 10 μg/mL of a monospecific antibody, a combination of the monospecific antibodies, or ITGA5BlxAV for 20 minutes on ice before being seeded into 6-well plates. 48h later, cells were harvested for Western blot or flow cytometry. FIG. 15A shows that treatment of PC-3 cancer cells with ITGA5BlxAV bispecific antibody construct results in stronger depletion of the target integrins compared to control antibodies, even in combination. Maximal reduction of AV expression by ITGA5BlxAV was confirmed by flow cytometry, with the quantification of the latter experiment shown in FIG. 15B.

DU-145 and VCAP cancer cells were treated with 10 μg/mL of a monospecific antibody, a combination of the monospecific antibodies, or ITGA5BlxAV for 20 minutes on ice before being seeded into 6-well plates. 48h later, cells were harvested for Western blot. In both DU-145 and VCAP cell types, a strong adaptive upregulation of one or both integrins is seen with

monospecific antibodies alone and in combination; this upregulation is strongly mitigated with ITGA5BlxAV treatment (FIGs. 16A and 16B).

C42B cancer cells were treated with 10 μg/mL of a monospecific antibody, a combination of the monospecific antibodies, or ITGA5BlxAV for 20 minutes on ice before being seeded into 6-well plates. 48h later, cells were harvested for Western blot. In C42B cells, ITGA5BlxAV treatment maximally induces loss of integrin a5 expression and mitigates compensatory integrin aV upregulation (FIGs. 17A and 17B). HUVEC endothelial cells were treated with or without 10 μg/mL of ITGA5BlxAV (FIG. 18 A), or with 10 μg/mL of a monospecific antibody, a combination of the monospecific antibodies, or ITGA5BlxAV (FIG. 18B), for 20 minutes on ice before being seeded into 6-well plates. 48h later, cells were harvested for Western blot. The data in FIGs. 18A and 18B show that, in HUVEC endothelial cells, ITGA5BlxAV treatment has specific and potent impact on integrin av (and not integrin a5) depletion compared to monospecific and combination antibody therapy. These data suggest a mechanism of action relevant to control of pathological angiogenesis, in which integrin av is known to play an important role in cooperation with integrin a5. These data are consistent with the experiments described above showing the superior potency and efficacy of ITGA5BlxAV on endothelial migration induced by tumor-stromal interactions, compared to monospecific and combination anti-integrin antibody blockade and to bevacizumab therapy.

Activated fibroblasts were treated with or without 10 μg/mL of ITGA5BlxAV for 20 minutes on ice before being seeded into 6-well plates. 72h later, cells were harvested for Western blot. Given the known functions of the target integrins in activated fibroblasts exhibiting alpha- smooth muscle actin expression, the impact of the bispecific antibody on integrin fate, a-smooth muscle actin expression, signaling, and cell viability was assessed. The data in FIG. 19 demonstrate that treatment with ITGA5BlxAV results in a strong downregulation of integrin av and a5, loss of actin expression and blockade of Akt signaling in these cells, without loss of cell viability. These data suggest a possible reprogramming of activated fibroblasts by ITGA5BlxAV treatment. This would have implications for fibroblasts associated with cancer as well as other pathological fibrotic conditions in which av signaling has been implicated, including scleroderma and related pathophysiological states such as chronic graft-versus-host disease.