LIGANDS OF THE UROKINASE RECEPTOR AND THEIR USE IN TREATING, DETECTING, AND IMAGING CANCER

CROSS REFERENCE

This application claims the benefit of United States Provisional Application No.

62/635,509, filed February 26, 2018, which is incorporated herein by reference in its entirety.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

This invention was made with government support under contracts CA186478 and 1176631-1-PAECO awarded by the National Institutes of Health. The government has certain rights in the invention.

TECHNICAL FIELD

The present invention pertains generally to anti-cancer therapeutics and cancer diagnostic agents and methods of treating, detecting, and imaging cancer. In particular, the invention relates to engineered ligands that bind to the urokinase receptor and methods of treating, detecting, and imaging cancer using such ligands. BACKGROUND

Approximately 90% of cancer-related deaths are a result of cancers reaching the metastatic stage, where cells from the primary tumor begin to migrate through the vasculature and form new tumors throughout the body. Once cancers reach this stage, the available treatment options are scarce and largely ineffective, and as a result the median survival time for patients is on the order of 1-2 years. There is thus a dire need for therapies that specifically target the mechanisms of metastasis to reduce the mortality caused by this disease. SUMMARY

The invention relates to engineered ligands that bind to the urokinase receptor and methods of treating, detecting, and imaging cancer using such ligands.

In one aspect, the invention includes a ligand of a urokinase receptor (uPAR) comprising a fusion protein comprising a growth factor domain (GFD) of urokinase (uPA) linked to a somatomedin B (SMB) domain of vitronectin.

In certain embodiments, the ligand further comprises a linker connecting the GFD to the SMB domain. In some embodiments, the linker is about 40 to about 50 angstroms in length.

In another embodiment, the linker comprises a sequence selected from the group consisting of SEQ ID NO: l 1, SEQ ID NO: 14, and SEQ ID NO: 15, or a sequence displaying at least about 80-100% sequence identity thereto, including any percent identity within this range, such as 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% sequence i dentity thereto, wherein the ligand is capable of binding to and/or inhibiting the urokinase receptor. In certain embodiments, the linker comprises at least one amino acid substitution selected from the group consisting of S5N, S10G,

G12R, and G19S, and wherein positions of the amino acids are numbered relative to the reference linker sequence of SEQ ID NO: 11.

In certain embodiments, the GFD comprises at least one mutation comprising an amino acid substitution selected from the group consisting of R30W, K37G, K37R,

Q39R, E41K, E41R, and H42Y, and wherein positions of the amino acids are numbered relative to the reference murine GFD sequence of SEQ ID NO: 1. In one embodiment, the GFD comprises the R30W, K37R, Q39R, E41R, and H42Y amino acid substitutions.

In other embodiments, the GFD comprises at least one mutation comprising an amino acid substitution selected from the group consisting of L4W, Q6R, H29Q, K36G, Q40R, H41R, and K46R, and wherein positions of the amino acids are numbered relative to the reference human GFD sequence of SEQ ID NO:3. In one embodiment, the GFD comprises the L4W, Q6R, H29Q, K36G, Q40R, H41R, and K46R amino acid

substitutions. In certain embodiments, the SMB domain comprises at least one mutation comprising an amino acid substitution selected from the group consisting of DIN, E3G, MT4V, and K17E, and wherein positions of the amino acids are numbered relative to the reference murine SMB domain sequence of SEQ ID NO:5. In one embodiment, the SMB domain comprises the DIN, E3G, M14V, and K17E amino acid substitutions.

In other embodiments, the SMB domain comprises at least one mutation comprising an amino acid substitution selected from the group consisting of DI G, E3G, K18E, and P41S, and wherein positions of the amino acids are numbered relative to the reference human SMB domain sequence of SEQ ID NO:7. In one embodiment, the SMB domain comprises the DIG, E3G, K18E, and P41 S amino acid substitutions.

In certain embodiments, the ligand comprises a fusion protein comprising an amino acid sequence selected from the group consisting of SEQ ID NO: 12 and SEQ ID NO: 13, or a sequence displaying at least about 80-100% sequence identity thereto, including any percent identity within this range, such as 81 , 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% sequence identity thereto, wherein the ligand is capable of binding and/or inhibiting the urokinase receptor.

In certain embodiments, the ligand further comprises an immunoglobulin Fc domain covalently linked to the fusion protein. The immunoglobulin Fc domain may be derived from an IgG (e.g., IgGl, IgG2, IgG3, or lgG4), IgM, IgE, IgA or IgD, or a combination or hybrid thereof. In one embodiment, the Fc fragment is derived from an IgG2a immunoglobulin. In another embodiment, the Fc domain is derived from a human IgGl isotype. The immunoglobulin Fc domain may be linked for example, to the N- terminal or C -terminal end of the fusion protein. In certain embodiments, the ligand comprises a fusion protein comprising an amino acid sequence selected from the group consisting of SEQ ID NOS: 17-20; or a sequence displaying at least about 80-100% sequence identity thereto, including any percent identity within this range, such as 81 , 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% sequence identity thereto, wherein the ligand is capable of binding to and/or inhibiting the urokinase receptor. In certain embodiments, the ligand further comprises an anti-cancer therapeutic agent conjugated to the fusion protein. The anti-cancer therapeutic agent may include, but is not limited to, a cytotoxic agent, a drug, a toxin, a nuclease, a hormone, an

immunomodulator, a pro-apoptotic agent, an anti-angiogenic agent, a boron compound, a photoactive agent, and a radioisotope. In another embodiment, a composition for use in the treatment of cancer is provided comprising the ligand, wherein the anti-cancer therapeutic agent is conjugated to the fusion protein.

In another embodiment, the fusion protein further comprises a signal peptide (e.g., a urokinase signal peptide).

In another embodiment, the fusion protein further comprises a tag or detectable label.

In another aspect, the invention includes a composition for use in the treatment of cancer comprising a ligand of a urokinase receptor described herein. In certain embodiments, the composition further comprises a pharmaceutically acceptable excipient. In another embodiment, the composition further comprises one or more other anti-cancer therapeutic agents, such as, but not limited to, ch emotherapeuti c,

immunotherapeutic, or biologic therapeutic agents.

In another aspect, the invention includes a method for treating cancer comprising administering to a subject in need thereof a therapeutically effective amount of a ligand of a urokinase receptor described herein. Multiple cycles of treatment may be administered to a subj ect. In certain embodiments, the ligand is admini stered according to a daily dosing regimen or intermittently. Preferably, the ligand is administered for a time period sufficient to efFect at least a partial tumor response, and more preferably a complete tumor response in the subject.

A ligand of a urokinase receptor may be administered by any suitable mode of administration. In certain embodiments, the ligand is administered intravenously, subcutaneously, or intralesionally to a subject. In another embodiment, the ligand is administered locally at a site of a tumor or cancerous cells in the subject. In another embodiment, the method further comprises performing surgery, radiation therapy, chemotherapy, immunotherapy, or biologic therapy, or a combination thereof.

In another aspect, the invention includes a kit comprising a pharmaceutical composition compri sing a ligand of a urokinase receptor and instructions for treating cancer. The kit may further comprise means for delivering the composition to a subject.

In another aspect, the invention includes a urokinase receptor-targeted imaging agent comprising a ligand of the urokinase receptor described herein conjugated to a diagnostic agent. The diagnostic agent can be, for example, an isotopic label, a fluorescent label, a chemiluminescent label, a bioluminescent label, a paramagnetic ion, an enzyme, a contrast agent (e.g., ultrasound contrast agent, a magnetic resonance imaging (MRI) contrast agent, or a radiocontrast agent), or a photoactive agent.

Exemplary' fluorescent labels include fluorescein derivatives, rhodamine derivatives, coumarin derivatives, cyanine derivatives, acridine derivatives, squaraine derivatives, naphthalene derivatives, oxadiazol derivatives, anthracene derivatives, pyrene derivatives, oxazine derivatives, arylmethine derivatives, and tetrapyrrole derivatives. Alternatively, the fluorescent label may comprise a fluorescent protein, such as, but not limited to, a green fluorescent protein (GFP), enhanced green fluorescent protein (EGFP), yellow fluorescent protein (YFP), enhanced yellow fluorescent protein (EYFP), blue fluorescent protein (BFP), red fluorescent protein (RFP), TagRFP, Dronpa, Padron, mApple, mCherry, rsCherry, and rsCherryRev.

Isotopic labels may comprise radioactive isotopes (e.g., gamma-emitters, beta- emitters, and positron-emitters) or non-radioactive isotopes (e.g., stable trace isotopes), such as, but not limited to, ¾, ² H, ^{1 20} I, ¹²³ I, ¹²⁴ I, ^l25 I, ¹³¹ I, ³⁵ S, ⁿ C, ¹³ C, ¹⁴ C, ³² P , ¹⁵ N, ¹³ N, ^{1 10} In, ^m In, ¹⁷⁷ LU, ¹⁸ F, ⁵² Fe, ⁶² Cu, ⁶⁴ Cu, ⁶⁷ Cu, ⁶⁷ Ga, ⁶⁸ Ga, ⁸⁶ Y, ⁹⁰ Y, ⁸⁹ Zr, ^94m Tc, ⁹⁴ Tc, ^99m Tc, ¹⁵⁴ Gd, ¹⁵⁵ Gd, ¹⁵⁶ Gd, ^1S7 Gd, ^{1 58} Gd, ^ls O, ¹⁸⁶ Re, ¹⁸⁸ Re, ⁵¹ M, ^52m Mn, ⁵⁵ Co, ⁷² As, ⁷⁵ Br,

76 _Br , 8 m _{Rbj and} 83 _Sr

Exemplary paramagnetic ions include chromium (III), manganese (II), iron (III), iron (II), cobalt (II), nickel (II), copper (II), neodymium (III), samarium (III), ytterbium (III), gadolinium (III), vanadium (II), terbium (ill), dysprosium (III), holmium (ill) and erbium (IP).

In another embodiment, the invention includes a urokinase receptor-targeted imaging agent comprising a ligand of a urokinase receptor, described herein, conjugated to a contrast agent. For example, the contrast agent may be an ultrasound contrast agent (e.g., a microbubble), a magnetic resonance imaging (MRI) contrast agent, or a radi ocontrast agent.

In another aspect, the invention includes a method of detecting cancer, the method comprising: a) administering a detectably effective amount of a urokinase receptor- targeted imaging agent, described herein, to a patient suspected of having cancer, under conditions wherein the urokinase receptor-targeted imaging agent binds to urokinase receptors present on tumors or cancerous cells, if present, in the patient; and b) detecting the urokinase receptor-targeted imaging agent bound to the tumors or cancerous cells, if present, by imaging tissue of the patient.

In certain embodiments, imaging of tissue is performed using a method selected from the group consisting of ultrasound imaging (UI), positron emission tomography (PET), single photon emission computed tomography (SPECT), magnetic resonance imaging (MRI), computed tomography (CT), optical imaging (OI), photoacoustic imaging (PI), and fluorescence imaging.

The methods of the invention can be used for determining the prognosis of the patient. Detection of a precan cerous lesion indicates the patient is at risk of developing cancer. Detection of increased levels of the urokinase receptor on the surface of tumors or cancerous cells is associated with tumor growth and cancer progression.

In another aspect, the invention includes a method of imaging tissue of a patient suspected of having cancer, the method comprising: a) contacting tissue of the patient with a detectably effective amount of a urokinase receptor-targeted imaging agent described herein under conditions wherein the urokinase receptor-targeted imaging agent binds to urokinase receptors present on tumors and cancerous cells, if present in the tissue; and b) imaging the tissue of the patient, wherein detection of increased binding of the urokinase receptor-targeted imaging agent to the tissue of the patient compared to a control indicates that the patient has cancer. The tissue may he contacted with the urokinase receptor-targeted imaging agent either in vivo or in vitro.

In certain embodiments, imaging of tissue is performed using a method selected from the group consisting of ultrasound imaging (UI), positron emission tomography (PET), single photon emission computed tomography (SPECT), magnetic resonance imaging (MRI), computed tomography (CT), optical imaging (OI), photoacoustic imaging (PI), and fluorescence imaging.

In another aspect, the invention includes a method of monitoring progression of cancer in a patient, the method comprising: imaging tissue of the patient according to a method described herein, wherein a first image is obtained at a first time point and a second image is obtained later at a second time point, wherein detection of increased binding of the urokinase receptor-targeted imaging agent to the tissue of the patient at the second time point compared to the first time point indicates that the patient is worsening, and detecti on of decreased binding of the urokinase receptor-targeted imaging agent to the tissue of the patient at the second time point compared to the first time point indicates that the patient is improving. Increased binding of the urokinase receptor-targeted imaging agent to the tissue of the patient may be caused, for example, by growth of a tumor or the presence of more tumors or cancerous cells at the second time point, which can be determined by inspection of the images. Alternatively, decreased binding of the urokinase receptor-targeted imaging agent to the tissue of the patient may be caused, for example, by tumor shrinkage or the presence of fewer tumors or cancerous cells.

In another aspect, the invention includes a method for evaluating the effect of an agent for treating cancer in a patient, the method comprising: imaging tissue of the patient according to a method described herein before and after the patient is treated with the agent, wherein detection of increased binding of the urokinase receptor-targeted imaging agent to the tissue of the patient (e.g., from tumor growth or increase in number of tumors or cancer cells) after the patient is treated with the agent compared to before the patient is treated with the agent indicates that the patient is worsening, and decreased binding of the urokinase receptor-targeted imaging agent to the tissue of the patient (e.g., from reduction in tumor size or reduction in the number of cancer cells) after the subject is treated with the agent compared to before the patient is treated with the agent indicates that the patient is improving.

A method for monitoring the efficacy of a therapy for treating cancer in a patient, the method comprising: imaging tissue of the patient according to a method described herein before and after the subject undergoes said therapy, wherein detection of increased binding of the urokinase receptor-targeted imaging agent to the tissue of the patient (e.g., from tumor growth or increase in number of tumors or cancer cells) after the patient undergoes said therapy compared to before the patient undergoes said therapy indicates that the patient is worsening, and decreased binding of the urokinase receptor-targeted imaging agent to the tissue of the patient (e.g., from reduction in tumor size or reduction in the number of cancer cells) after the patient undergoes said therapy compared to before the patient undergoes said therapy indicates that the patient is improving.

In another aspect, the invention includes a method of treating a patient suspected of having cancer, the method comprising: a) receiving information regarding whether or not cancer was detected in the patient using a urokinase receptor-targeted imaging agent according to a method described herein; and b) administering anti-cancer therapy to the subject if cancer was detected in the patient. In certain embodiments, the anti -cancer therapy comprises surgery, radiation therapy, chemotherapy, hormonal therapy, immunotherapy, or biologic therapy, or any combination thereof.

In another embodiment, the invention includes a method of using a urokinase receptor-targeted imaging agent comprising a fluorescent label for fluorescence imaging of cancerous cells, the method comprising: a) contacting the cancerous cells with the urokinase receptor-targeted imaging agent comprising the fluorescent label, wherein cancerous cells expressing the urokinase receptor uptake the urokinase receptor-targeted imaging agent; b) illuminating the cell with light at a fluorescence excitation wavelength of the fluorescent label; and c) recording a fluorescence image of the cancerous cells by detecting fluorescence emitted by the fluorescent label of the urokinase receptor-targeted imaging agent. Fluorescence images may be visualized, for example, with a fluorescence microscope, a fiber-optic fluorescence imaging system, or a medical fluorescence imaging device, such as a miniaturized medical imaging system (e.g., a handheld microscope, a laparoscope, an endoscope, or a microendoscope).

In another embodiment, the invention includes a method of using a urokinase receptor-targeted imaging agent comprising a bioluminescent label for bioluminescence imaging of cancerous cells, the method comprising: a) contacting the cancerous cells with the urokinase receptor-targeted imaging agent comprising the bioluminescent label, wherein cancerous cells expressing the urokinase receptor uptake the urokinase receptor- targeted imaging agent; b) contacting the cancerous cells with a chemiluminescent substrate; and c) recording a bioluminescence image of the cancerous cells by detecting bioluminescence emitted from the bioluminescent label of the urokinase receptor-targeted imaging agent.

In another aspect, the invention includes a kit comprising an imaging agent comprising a ligand of a urokinase receptor and instructions for imaging cancer. The kit may further comprise means for delivering the composition to a subject.

In another aspect, the invention includes a recombinant polynucleotide compri sing a promoter operably linked to a polynucl eotide encoding a ligand of a urokinase receptor described herein. In another embodiment, the recombinant polynucleotide is provided by a vector such as a bacterial plasmid vector or a viral expression vector. Exemplary viral vectors include adenovirus, retrovirus (e.g., g- retrovirus and lenti virus), poxvirus, adeno-associated virus, baculovirus, or herpes simplex virus vectors. In another embodiment, the invention includes a host cell or host subject comprising the recombinant polynucleotide.

In another embodiment, the invention includes a method for producing a ligand of a urokinase receptor, the method comprising: a) transforming a host cell with a recombinant polynucleotide comprising a promoter operably linked to a polynucleotide encoding a ligand of a urokinase receptor described herein; b) culturing the transformed host cell under conditions whereby the ligand of the urokinase receptor is expressed; and c) isolating the ligand of the urokinase receptor from the host cell.

In another embodiment, the invention includes a method for producing a ligand of a urokinase receptor in a host subject, the method comprising introducing into the host subject a recombinant polynucleotide comprising a promoter operably linked to a nucleotide sequence encoding the ligand of the urokinase receptor, wherein the ligand of the urokinase receptor is expressed in the host subj ect in an amount sufficient to inhibit the urokinase receptor and/or have anti-tumor activity in the subject.

In another aspect, the invention includes a method for treating cancer compri sing administering to a subject in need thereof a therapeutically effective amount of a recombinant polynucleotide comprising a promoter operably linked to a nucleotide sequence encoding the ligand of the urokinase receptor.

These and other embodiments of the subject invention will readily occur to those of skill in the art in view of the disclosure herein.

BRIEF DESCRIPTION OF THE FIGURES

FIGS. 1A and IB show the strategy for developing a bispecific (i.e., biepitopic) ligand that binds different domains of uPAR. FIG. 1 A shows that uPAR is a key regulator of numerous biological processes that drive cancer growth and metastasis. Binding of the soluble ligand uPA to uPAR localizes ECM degradation at the cell surface and uPAR- driven cancer growth and metastasis via pathways 1 -7. uPAR also binds vitronectin in the absence of uPA (pathway 8), facilitating migration and metastasis. FIG. 1B shows our approach to target uPAR-driven cancer growth and metastasis: i) The growth factor domain (GFD) of uPA binds uPAR and blocks localized ECM degradation and pathways 3, 5, and 6 in panel A, but not pathways 2, 4, 7, and 8. ii) SMB binds uPAR and blocks pathways 4, 7, and 8, and is expected to block 2, but not 3, 5, and 6, and ECM

degradation iii) GFD and SMB simultaneously bind uPAR, and will be linked together to form a high affinity bispecific ligand that simultaneously blocks multiple uPAR-mediated processes driving cancer growth and metastasis.

FIG. 2 shows modeling the GDF-SMB fusion protein with a 4x(GGGGS) amino acid linker. The GFD domain of uPA and the SMB domain of vitronectin are shown genetically fused by a 20-amino acid linker comprising four GGGGS (SEQ ID NO: 9) repeats (black), and simultaneously bound to uPAR. Three different possible orientations of the l inker are shown from multiple views of the complex to indicate the presence or absence of potential steric clash between the linker and proteins. The figures were generated by editing data from PDB 3BT1 (Huai et al (2008) Nat. Struct. Mol Biol. 15, 422-423) using IJCSF chimera molecular viewing software (Pettersen et al. (2004) J. Comput. Chem. 25, 1605-1612).

FIG. 3 shows modeling the GDF-SMB fusion protein with a 5x(GGGGS) amino acid linker (SEQ ID NO: 11). The GFD domain of uPA and the SMB domain of vitronectin are shown genetically fused by a 25-amino acid linker comprising four GGGGS repeats (black), and simultaneously bound to uPAR. Three different possible orientations of the linker are shown from multiple views of the complex to indicate the presence or absence of potential steric clash between the linker and proteins. The figures were generated by editing data from PDB 3BT1 (Huai et al., supra) using UCSF chimera molecular viewing software (Pettersen et al., supra).

FIG. 4 shows equilibrium binding titrations of three different yeast-displayed mGFD-mSMB fusion protein designs to soluble muPAR. Three different mGFD-mSMB designs containing either a 3x, 4x, or 5x GGGGS linker (SEQ ID NO: 11) bridging the GFD and SMB domains were displayed on the surface of yeast and their binding affinities for uPAR were compared. Binding reactions were incubated at room

temperature for 24 hours and analyzed by flow cytometry. The mGFD-mSMB fusion protein with the 5x(GGGGS) linker (henceforth simply referred to as "mGFD-mSMB) had the highest affinity for muPAR, indicating the 5x(GGGGS) linker is the most optimal of the three tested.

FIG. 5 shows equilibrium binding titrations of yeast-displayed mGFD and mGFD-mSMB to soluble muPAR. Binding reactions were incubated at room temperature for 24 hours and analyzed by flow cytometry. The K _d values are 1.83 nM and 0.22 nM for mGFD and mGFD-mSMB, respectively. These data indicate the bispecific mGFD- mSMB fusion protein has approximately 8.3-fold higher affinity for muPAR compared to the wild type mGFD ligand.

FIGS. 6A and 6B show analyses of specific amino acids driving the enhanced affinity of the mGFD-mSMB fusion protein for muPAR. FIG. 6A shows a structural representation of the binding interaction between the SMB domain of mGFD-mSMB and muPAR. The mouse SMB structure has not been solved; therefore, the human proteins are shown here as an example. Amino acids Y27 and Y28 (grey sticks) in SMB form the specific binding interaction with uPAR (Deng, et al. (1996) J. Biol. Chem. 271 , 12716- 12723; Deng et al. (1996) J. Cell Biol. 134, 1563-1571; and Okumura et al. (2002) J.

Biol. Chem. 277, 9395-9404). The linker fusing GFD and SMB is shown in black. FIG. 6B shows amino acid mutations Y27A and Y28A in SMB abolishes its affinity for uPAR. As expected, these mutations effectively render the affinity of mGFD-mSMB

approximately equal to that of mGFD for muPAR. The figure was generated using data from the PDB file 3BT1 (Huai et al., supra).

FIG. 7 shows a structural alignment of the mouse GFD -uPAR complex with the human GFD-uPAR-SMB complex. The structure of human uPAR bound to hGFD and hSMB is shown aligned to the structure of mouse uPAR bound by mGFD. The figure was generated using data from PDB files 3BT1 (Huai et al., supra) and 3LAQ (Lin et al. (2010) J. Biol. Chem. 285, 10982-1092) and UCSF chimera molecular viewing software (Pettersen et al., supra).

FIG. 8 shows equilibrium binding titrations of yeast-displayed hGFD and the hGFD-hSMB fusion protein to soluble huPAR. Binding reactions were incubated at room temperature for 24 hours and analyzed by flow cytometry. The K _d values are 1.04 nM and 0.12 nM for hGFD and hGFD-hSMB, respectively. These data indicate the bispecific hGFD-hSMB fusion protein has approximately 8.7-fold higher affinity for huPAR compared to the wild type hGFD ligand.

FIGS. 9A and 9B show analyses of specific amino acids driving the enhanced affinity of the hGFD-hSMB fusion protein for uPAR. FIG. 9 A shows a structural representation of the binding interaction between the SMB domain of hGFD-hSMB and huPAR. Amino acids Y27 and Y28 (grey sticks) in SMB form the specific binding interaction with huPAR. The linker fusing hGFD and hSMB is shown in black. FIG. 9B shows that amino acid mutations Y27A and Y28A in SMB abolish its affinity for uPAR. As expected, these mutations effectively render the affinity of hGFD-hSMB equal to that of hGFD for huPAR. These data indicate the superior uPAR binding of hGFD-hSMB compared to hGFD alone is due to the additional interactions between amino acids Y27 and Y28 of SMB and uPAR. Binding reactions were incubated at room temperature for 24 hours and analyzed by flow cytometry.

FIGS. 10 A and 10B show the purification of soluble recombinant uPAR proteins. FIG. 10A shows chromatograms of huPAR, muPAR, and muPAR-FLAG purified by FPLC on a Superdex 75, 10/300 GL column. Proteins were diluted and run in PBS, pH 7.4 buffer at 0.4 mL/min. Collected fractions are highlighted in gray. FIG. 10B shows an analysi s of proteins collected after FPLC. Proteins were analyzed by 4-12%

polyacrylamide gel electrophoresis run at 140 V for 55 minutes, and subsequently stained with Coomassie blue. Lanes contain, from left to right: huPAR, muPAR, muPAR-FLAG, and Kaleidoscope protein ladder.

FIG. 11 shows amino acid sequence alignment of mGFD variants from 5 separate DNA libraries. 10 sequences from each of 5 separate mGFD DNA libraries are shown arranged by library from high-to-low mutation frequency. The wild-type mGFD sequence is shown at top, and mutations in the variants below are highlighted in color.

FIG. 12 shows sort progression of mGFD library 1. Four rounds of FACS were used to isolate high affinity variants from the first mGFD library. Equilibrium binding screens were used to perform all sorts. Selective pressure for improved variants was increased by reducing the concentration of muPAR incubated with the library between sorts 2 and 3, and by collecting a smaller percentage of the top binding variants in each successive sort. The gates used to select variants are shown, along with the percentage of the expressing population collected.

FIGS. 13A and 13B show analyses of mGFD library 1 after sort 4. FIG. 13A shows an amino acid sequence alignment of 20 mGFD variants after sort 4. The wild- type mGFD sequence is highlighted in light gray, and mutations are highlighted with colored squares. FIG. I3B shows equilibrium binding titrations of yeast-displayed mGFD and the bulk mGFD library after sort 4 to soluble muPAR. Binding reactions were incubated at room temperature for 24 hours and analyzed by flow cytometry. The K _d of mGFD is 3.89 nM, and the apparent K _d of the library is 1.52 nM. These data indicate the first round of directed evolution generated marginally improved mGFD variants. FIG. 14 shows amino acid sequence alignment of mGFD library 2 variants. DNA from the mGFD library 1 variants remaining after the 4 ^th sort was mutagenized using error-prone PCR for a subsequent round of directed evolution. Five separate DNA libraries were generated with varying mutation frequency, and 10 sequences from three of the libraries are shown here aligned to the wild-type mGFD sequence. Sequences are arranged by library (1, 3, and 5) from top to bottom, with libraries 1 and 5 having the highest and lowest mutation frequency, respectively. The wild-type mGFD sequence is shown at top, and mutations in the variants below are highlighted in color.

FIG. 15 shows sort progression of mGFD library 2. Six rounds of FACS were used to isolate high affinity variants from the second mGFD library. Sorts 1 -2 were performed using equilibrium binding screens with the indicated concentrations of soluble muPAR, and sorts 3-6 were performed using kinetic off rate screens. Selective pressure for improved variants was increased by reducing the concentration of muPAR incubated with the library between sorts 1 and 2, or by increasing the‘off" time in sorts 3-6, and by collecting a smaller percentage of the top binding variants in successive sorts. The gates used to select variants are shown, along with the percentage of the expressing population collected.

FIGS. 16A and 16B show sequence analyses of mGFD library 2 after sorts 5 and 6. Amino acid sequence alignment of (FIG. 16A) 35 mGFD variants isolated after sort 5, and (FIG. 16B) 10 mGFD variants isolated after sort 6. The wild-type mGFD sequence is shown at top, and mutations in the variants below are highlighted in color.

FIGS. 17A and 17B show analyses of the engineered *mGFD protein generated by directed evolution . FIG. 17A shows equilibrium binding titrations of yeast-displayed wild-type mGFD and the engineered *mGFD protein to soluble muPAR. Binding reactions were incubated at room temperature for 48 hours and analyzed by flow cytometry. The K _d values of mGFD and *mGFD are 2.11 nM and 0 009 nM, respectively. These data indicate directed evolution of the mGFD protein successfully generated a variant with 235 -fold higher affinity for muPAR. FIG 17B shows amino acid mutations in *mGFD mapped onto the wild-type mGFD-muPAR complex. mGFD and muPAR are shown in green and blue ribbon, respectively. Side chains of the five amino acids mutated in *mGFD are show as red spheres. The figure was generated using data from PDB file 3LAQ ²⁸ and UCSF chimera molecular viewing software (Pettersen et al., supra).

FIG. 18 shows amino acid sequence alignment of hGFD variants from 5 separate DNA libraries.10 sequences from each of 5 separate hGFD DNA libraries with varying levels of mutation frequency are shown arrange from high to low mutation frequency.

The wild-type hGFD sequence is shown at top, and mutations in the variants below are highlighted in color.

FIG. 19 shows sort progression of hGFD library 1. Four rounds of FACS were used to isolate high affinity variants from the first hGFD library. Equilibrium binding screens were used to perform all sorts. Selective pressure for improved variants was increased by reducing the concentration of huPAR incubated with the library between sorts, and by collecting a smaller percentage of the top binding variants in each successive sort. The gates used to select variants are shown, along with the percentage of the expressing population collected.

FIGS. 20A and 20B show analyses of hGFD library 1 variants remaining after the 4 ^th sort. FIG. 20A shows an amino acid sequence alignment of 20 hGFD library 1 variants isolated after the 4 ^th sort. The wild-type hGFD sequence is shown at top, and mutations in the variants below are highlighted in color. FIG. 20B shows equilibrium binding titrations of yeast-displayed hGFD variants and wt hGFD to soluble huPAR. Binding reactions were incubated at room temperature for 24 hours and analyzed by flow cytometry. The K _d values are: wt mGFD = 0.43 nM, mGFDK46R = 0.32 nM, mGFDDMG = 0.2 nM, mGFDD14G,K48R = ^: 0.13 nM. These data indicate the first round of directed evolution generated marginally improved hGFD variants.

FIG. 21 shows sort progression of hGFD library 2. Six rounds of FACS were used to isolate high affinity variants from the second hGFD library. For the first sort, two libraries were sorted separately and the collected variants were combined for subsequent sorts. Sorts 1 -4 were performed using equilibrium binding screens with the indicated concentrations of soluble huPAR, and sorts 5-6 were performed using kinetic off rate screens. Selective pressure for improved variants was increased by reducing the concentration of huPAR incubated with the library between sorts, or by increasing the "off time in sorts 5-6, and by collecting a smaller percentage of the top binding variants in successive sorts. The gates used to select variants are shown, along with the percentage of the expressing population collected.

FIGS. 22A and 22B show sequence analyses of the hGFD library 2 after sorts 4 and 6. Amino acid sequence alignments of (FIG. 22A) 10 hGFD variants isolated after sort 4, and (FIG. 22B) 40 hGFD variants isolated after sort 6. The wild-type hGFD sequence is shown at top, and mutations in the variants below are highlighted in color.

FIGS. 23 A and 23B show analyses of hGFD Library 2 variants after sort 6. FIG. 23 A shows an amino acid sequence alignment of seven hGFD variants isolated after sort 6 to wild type hGFD. FIG. 23B shows kinetic dissociation of soluble huPAR from the yeast-displayed hGFD library 2 variants listed in panel A as a function of incubation time. Samples were incubated at room temperature for the indicated time and analyzed by flow cytometry. The hGFD L2.S6.15 variant (henceforth referred to as *hGFD, FIG. 45) displayed marginally slower dissociation kinetics than all other variants.

FIGS. 24 A and 24B show an analysis of the engineered *hGFD protein generated by directed evolution. FIG. 24A show equilibrium binding titrations of yeast-displayed wild-type hGFD and the engineered *hGFD protein to soluble huPAR. Binding reactions were incubated at room temperature for 48 hours and analyzed by flow cytometry. The K _d values of hGFD and *hGFD are 0.843 nM and 0.028 nM, respectively. These data indicate directed evolution of the hGFD protein successfully generated a variant with 30- fold higher affinity for huPAR. FIG. 24B shows amino acid mutations in *hGFD mapped onto the w ^r i ld-type hGFD-huPAR structure. Side chains of mutated amino acids in *hGFD are show as spheres. The figure wasgenerated using data from PDB file 3BT1 ¹⁹ and UCSF chimera molecular viewing software.

FIG. 25 shows an amino acid sequence alignment of linker-mSMB variants from 5 separate DNA libraries. 10 sequences from each of 5 separate linker-mSMB DNA libraries with varying levels of mutation frequency are shown arranged by library from high to low mutation frequency. The wild-type linker-mSMB sequence is shown at top, and mutations in the variants below are highlighted in color. Libraries were generated in the context of the mGFD-mSMB fusion protein. The sequence of the adjoining mGFD region was not mutated, and is omitted here for clarity.

FIG. 26 shows a sort progression of the Hnker-mS MB library. Six rounds of FACS were used to isolate high affinity variants from the linker-mSMB library. Sorts 1-2 were performed using equilibrium binding screens with the indicated concentrations of soluble muPAR, and sorts 3-6 were performed using kinetic off rate screens. Selective pressure for improved variants was increased by reducing the concentration of muPAR incubated with the library between sorts 1 and 2, or by increasing the "off" time between sorts 3-6, and by collecting a smaller percentage of the top binding variants in successive sorts. The gates used to select variants are shown, along with the percentage of the expressing population collected.

FIG. 27 shows a sequence analysis of the linker-mSMB library after sort 6. The amino acid sequence alignment of 10 linker-mSMB variants isolated after sort 6 is shown. The wild-type linker-mSMB sequence is shown at the top, and mutations in the variants below are highlighted in color.

FIG. 28 shows equilibrium binding titrations of the yeast-displayed mGFD- mSMB protein and the evolved mGFD-*mSMB protein to soluble muPAR. Binding reactions were incubated at room temperature for 24 hours and analyzed by flow cytometry. The K _d values of mGFD-mSMB and the evolved mGFD-*mSMB are 0.751 nM and 0.126 nM, respectively. These data indicate directed evolution of the linker- mSMB region of the mGFD-mSMB fusion protein successfully generated a variant with 6-fold higher affinity for muPAR.

FIG. 29 shows amino acid mutations in mGFD-*mSMB mapped onto the human GFD-SMB-uPAR structure. The structure of the murine SMB has not been solved. Therefore, mutations in the linker and SMB domain of the mGFD-*mSMB protein are shown here mapped onto the human GFD-SMB-uPAR complex. GFD and SMB are shown fused by a 25 -amino acid linker (black). The seven amino acid mutations in the mGFD-*mSMB protein are shown as dark spheres. The figure was generated using data from PDB files 3BT1 and UCSF chimera molecular viewing software. FIG. 30 shows equilibrium binding titrations of a yeast-displayed hGFD-hSMB and hGFD ^F25A -hSMB to soluble muPAR. A single F25A amino acid mutation in the growth factor-like binding loop of hGFD (hGFD ^Fi5A ) reduces the overall affinity of the hGFD-hSMB fusion protein for huPAR by over two orders of magnitude (K _d = 15 nM). Binding reactions were incubated at room temperature for 24 hours and analyzed by flow cytometry.

FIG. 31 shows amino acid sequence alignment of linker-hSMB variants from 1 of 5 separate DNA libraries. 5 separate linker-hSMB DNA libraries with varying levels of mutation frequency were generated using error-prone PCR. Sequences of 10 variants from the library with the highest mutation frequency are shown. Libraries were generated in the context of the hGFD ^F25A -hSMB fusion protein. The sequence of the adj oining hGFD ^F2:,A region was not mutated, and is omitted here for clarity. The wild-type linker- hSMB sequence is shown at top, and mutations in the variants below are highlighted in color.

FIG. 32 shows sort progression of the linker-hSMB library. Five rounds of FACS were used to isolate high affinity variants from the linker-hSMB library. All sorts were performed using equilibrium binding screens with the indicated concentrations of soluble huPAR. Selective pressure for improved variants was increased by sequentially reducing the concentration of muPAR incubated with the library between sorts, and by collecting a smaller percentage of the top binding variants in successive sorts. The gates used to select vari ants are shown, along with the percentage of the expressing population collected.

FIGS. 33A and 33B show analysis of linker-hSMB library variants remaining after the 5 ^th sort. FIG. 33 A shows an amino acid sequence alignment of 10 linker-hSMB library' variants isolated after the 5 ^th sort. The wild-type linker-hSMB sequence is shown at top, and mutations in the variants below are highlighted in color. FIG. 33B shows equilibrium binding titrations of the two yeast-displayed linker-hSMB variants identified after the 5 ^th sort to soluble huPAR. Binding reactions were incubated at room temperature for 24 hours and analyzed by flow cytometry. The K _d values were essentially equal for both variants, 0.49 nM for L1.S6.5, and 0.42 nM for L1.S6.1 (henceforth referred to as hGFD ^F25A -*hSMB). FIG. 34 shows equilibrium binding titrations of the yeast-displayed hGFD ^F25A - hSMB and the evolved hGFD ^F25A -*hSMB protein to soluble huPAR. Binding reactions were incubated at room temperature for 24 hours and analyzed by flow cytometry. The K _d values of hGFD ^F25A -hSMB and the evolved hGFD ^F25A -*hSMB are 7.10 nM and 0.68 nM, respectively. These data indicate directed evolution of the linker-hSMB region of the hGFD-hSMB fusion protein successfully generated a variant with 10-fold higher affinity for huPAR.

FIG. 35 shows amino acid mutations in hGFD-*hSMB mapped onto the wild type human GFD-SMB-uPAR structure. hGFD and hSMB are shown fused by a 25-amino acid linker (black). Side-chains of the 5 amino acid mutations in hGFD-*hSMB are shown as dark spheres (absent in the case of mutations to glycine). Three additional amino acids (ALA) belonging to the signal peptide of SMB were mistakenly included at the ML-terminus of the SMB domain in the library. These were later removed and the affinity for uPAR was unchanged (data not shown). Therefore, for future consistency these amino acids are omitted from the sequence here; the 5x(GGGGS) linker (SEQ ID NO: 11) ends at S73, and the SMB domain begins at D74 The figure was generated using data from PDB files 3BT1 ¹⁹ and UCSF chimera molecular viewing software.

FIG. 36 shows amino acid mutations in the engineered *hGFD-*hSMB (hGS) and *mGFD-*mSMB (mGS) bispecific uPAR ligands mapped onto the wild type human GFD-SMB-uPAR structure. Proteins domains in the mGS-uPAR structure (left) and hGS-uPAR structure (right) are colored as in FIGS. 29 and 35, respectively. Side-chains of amino acid mutations are shown as dark spheres (absent in the case of mutations to glycine). The figure was generated using data from PDB files 3BT1 ¹⁹ and UCSF chimera molecular viewing software.

FIGS. 37A-37D show the purification of soluble recombinant Fc-fusion proteins by FPLC. Chromatograms are shown of mGFD-Fc (FIG. 37A), mGS-Fc (FIG. 37B), hGFD-Fc (FIG. 37C), and hGS-Fc (FIG. 37D) purified by FPLC on a Superdex 200 increase, 10/300 GL column. Proteins were suspended and run in PBS, pH 7.4 buffer at 0.4 mL/min. Collected fractions are highlighted in gray. FIGA. 38A and 38B show analyses of Fc-fusion proteins collected after purification by FPLC. Proteins were analyzed by 4-12% polyacrylamide gel

electrophoresis run at 140 V for 55 minutes, and subsequently stained with SimplyBlue Safestain. Lanes contain, from left to right: (FIG. 38 A) mGFD-Fc, mGS-Fc, Novex Sharp protein ladder, and (FIG. 38 B) hGFD-Fc, hGS-Fc, Kaleidoscope protein ladder.

FIGS. 39A and 39B show the purified Fc-fusion proteins inhibit uPA-uPAR binding. The GFD domain of uPA was displayed on the surface of yeast cells as a fusion to the aga2p coat protein, and binding to 2 nM uPAR was measured in the presence of increasing concentrations of (FIG. 39 A) mGFD-Fc and mGS-Fc, and (FIG. 39B) hGFD- Fc and hGS-Fc. Binding reactions were incubated at room temperature for 24 hours and analyzed by flow cytometry. Accurate icso values for the Fc-fusion proteins cannot be derived from these results, given the uPAR concentration used in the assay is orders of magnitude above the K _d of the mGS-Fc and hGS-Fc proteins, and thus binding is concentration-dependent and not Ka-dependent. These data simply verify the purified Fc- fusion proteins are active and capable of inhibiting the uPA-uPAR binding interaction.

FIG. 40 shows uPAR-specific tumor localization of hGFD-Fc and hGS-Fc proteins. The uPAR-positive human breast (MDA-MB-231) tumors, and uPAR-negative human embryonic kidney (HEK 293 T) tumors were implanted into the left and right shoulders of athymic nude mice, respectively. Once tumors reached 5-10 mm in diameter, 1.5 nmol of hGFD-Fc-680 or hGS-Fc-680 were inj ected via tail vein, and protein localization was tracked via whole-body fluorescence imaging at the indicated time points. Images are of mice injected with PBS (left), hGFD-Fc-680 (middle), and hGS-Fc-680 (right). Radiant efficiency scale: min = ^: 6E8, max = 3.8E9.

FIG. 41 shows the quantification of uPAR-specific tumor localization of hGFD- Fc and hGS-Fc proteins in vivo. The signal for uPAR-positive tumor fluorescence from images shown in FIG. 41 is plotted here divided by the uPAR-negative tumor fluorescence. A signal of 1 (dashed line) indicates zero uPAR-positive tumor specificity.

FIG. 42 shows the average uPAR-specific tumor localization of hGFD-Fc and hGS-Fc proteins in vivo. Mice were inj ected with hGFD-Fc-680 or hGS-Fc-680 (3 each) as in FIG. 41, and uPAR-positive and uPAR-negative tumor fluorescence was measured each day for 22 days. The signal for uPAR-positive tumor fluorescence is plotted here divided by the uPAR-negative tumor fluorescence. A signal of 1 (dashed line) indicates zero uPAR-positive tumor specificity.

FIGS. 43 A and 43 B show uPAR-specific tumor localization of hGFD (FIG. 43 A) and hGS (FIG. 43B) monomer proteins. The uPAR-positive human breast (MDA-MB- 231) tumors, and uPAR-negative human embryonic kidney (HEK 293 T) tumors were implanted into the left and right shoulders of athymic nude mice, respectively. 1.5 nmol of hGFD-680 or hGS-680 were inj ected via tail vein, and protein localization was tracked via whole-body fluorescence imaging at the indicated time points. Radiant efficiency scale: min = 2.6E8, max = 1.6E9.

FIGS. 44A-44C show internalization of hGFD-Fc and hGS-Fc proteins into human breast cancer (MDA-MB-231) cells in vitro. FIG. 44A shows cell fluorescence from surface-bound Alexa 488-labeled proteins is quenched by an anti -488 antibody. Cells were incubated with 2 nM hGFD-Fc-488 or hGS-Fc-488 at 4°C for 1 hr, washed, and incubated with an anti-Alexa-488 antibody (a488-Ab) for an additional 30 minutes where indicated prior to analysis by flow cytometry. Given that incubation at 4°C precludes internalization, the cell fluorescence measured in this case is caused by the Alexa 488-labeled proteins binding to cell-surface uPAR. FIG. 44B shows internalization of hGFD-Fc-488 and hGS-Fc-488 proteins. Cells were incubated in culture flasks at 37°C with the indicated concentrations of 488-labeled proteins in the presence or absence of a 100-fold molar excess of unlabeled protein for 15 hours. Cells were then harvested, washed, and incubated at 4°C with a488-Ab for an additional 30 minutes to quench surface-bound protein fluorescence prior to analysis by flow cytometry. The hGS-Fc protein demonstrated significantly higher internalization compared to hGFD-Fc (** p < 0.005, *** p < 0.0001). FIG. 44C shows internalization of hGFD-Fc and hGS-Fc-488 over time. Cells were incubated with 2 nM hGFD-Fc-488 or hGS-Fc-488 for the indicated times and analyzed as described in panel (FIG. 44B).

FIG. 45 shows sequence alignments of wild type and engineered proteins. The wild type sequences of the murine and human GFD domains of urokinase (mGFD and hGFD, respectively), and the linker-SMB domains of vitronectin (linker-mSMB and linker-hSMB, respectively) were aligned to the corresponding engineered proteins generated by directed evolution. Only the linker- SMB domains of mGFD-mSMB and hGFD ^t2:,A -hSMB proteins are shown for clarity. Amino acid mutations in the engineered proteins are highlighted in gray. Three additional amino acids (ALA) belonging to the signal peptide of human SMB were mistakenly included at the NIL-terminus in the linker-hSMB library. These were later removed and the affinity for uPAR was unchanged (data not shown). Therefore, for future consistency these amino acids are omitted from the sequence here; the [GlyrSer]? linker ends at S73, and the hSMB domain begins at D74.

DETAILED DESCRIPTION

The practice of the present invention will employ, unless otherwise indicated, conventional methods of medicine, pharmacology, chemistry, biochemistry, molecular biology and recombinant DNA techniques, and immunol ogy, within the skill of the art. Such techniques are explained fully in the literature. See, e.g., Wolbarst et al. Medical Imaging: Essentials for Physicians (Wiley-Blackwell, 2013); Handbook of Experimental Immunology, Vols. I-IV (D.M Weir and C.C. Blackwell eds., Blackwell Scientific Publications); A.L. Lehninger, Biochemistry (Worth Publishers, Inc., current addition); Sambrook et al., Molecular Cloning: A Laboratory Manual (3 ^rd Edition, 2001); Methods In Enzymology (S. Colowick and N. Kaplan eds., Academic Press, Inc.); Fmoc Solid Phase Peptide Synthesis: A Practical Approach (W. C. Chan and Peter D. White eds., Oxford University Press, 1 ^st edition, 2000) ; N. Leo Benoiton, Chemistry of Peptide Synthesis (CRC Press; 1 ^st edition, 2005); Peptide Synthesis and. Applications (Methods in Molecular Biology, John Howl ed., Humana Press, 1 ^st ed., 2005); and Pharmaceutical Formulation Development of Peptides and Proteins (The Taylor & Francis Series in Pharmaceutical Sciences, Lars Hovgaard, Sven Frokjaer, and Marco van de Weert eds., CRC Press; 1 ^st edition, 1999).

All publications, patents and patent applications cited herein, whether supra or infra, are hereby incorporated by reference in their entireties. I. DEFINITIONS

In describing the present invention, the following terms will be employed, and are intended to be defined as indicated below.

It must be noted that, as used in this specification and the appended claims, the singular forms "a," "an" and "the" include plural referents unless the content clearly dictates otherwise. Thus, for example, reference to "a ligand" includes a mixture of two or more ligands, and the like.

The term "about," particularly in reference to a given quantity, is meant to encompass deviations of plus or minus five percent.

The terms "protein," "peptide," "oligopeptide" and "polypeptide" refer to any compound comprising naturally occurring or synthetic amino acid polymers or amino acid-like molecules including but not limited to compounds comprising amino and/or imino molecules. No particular size is implied by use of the terms "protein," "peptide," "oligopeptide" or "polypeptide" and these terms are used interchangeably. Included within the definition are, for example, polypeptides containing one or more analogs of an amino acid (including, for example, unnatural amino acids, etc.), polypeptides with substituted linkages, as well as other modifications known in the art, both naturally occurring and non-naturally occurring (e.g., synthetic). Thus, synthetic oligopeptides, dimers, mul timers (e.g., tandem repeats, linearly-linked peptides), cyclized, branched molecules and the like, are included within the definition. The terms also include molecules comprising one or more peptoids (e.g., N-substituted glycine residues) and other synthetic amino acids or peptides. (See, e.g., U.S. Patent Nos. 5,831,005;

5,877,278; and 5,977,301; Nguyen et al. (2000) Chem. Biol. 7(7):463-473; and Simon et al. (1992) Proc. Natl. Acad. Sci. USA 89(20):9367-9371 for descriptions of peptoids). Non-limiting lengths of peptides suitable for use in the present invention includes peptides of 3 to 5 residues in length, 6 to 10 residues in length (or any integer therebetween), 11 to 20 residues in length (or any integer therebetween), 21 to 75 residues in length (or any integer therebetween), 75 to 100 (or any integer therebetween), or polypeptides of greater than 100 residues in length. Typically, polypeptides useful in this invention can have a maximum length suitable for the intended application. Preferably, the polypeptide is between about 3 and 100 residues in length. Generally, one skilled in art can easily select the maximum length in view of the teachings herein.

Further, proteins, peptides, and polypeptides, as described herein, for example synthetic proteins, peptides, and, polypeptides may include additional molecules such as localization sequences, tags, labels, or other chemical moieties. Such moieties may further enhance inhibition of the urokinase receptor, facilitate purification of ligands, and/or detection of the l igands.

Thus, references to polypeptides or peptides also include derivatives of the amino acid sequences of the invention including one or more non-naturally occurring amino acids. A first polypeptide or peptide is "derived from" a second polypeptide or peptide if it is (i) encoded by a first polynucleotide derived from a second polynucleotide encoding the second polypeptide or peptide, or (ii) displays sequence identity to the second polypeptide or peptide as described herein. Sequence (or percent) identity can be determined as described below. Preferably, derivatives exhibit at least about 50% percent identity, more preferably at least about 80%, and even more preferably between about 85% and 99% (or any value therebetween) to the sequence from which they were derived. Such derivatives can include post-expression modifications of the polypeptide or peptide, for example, glycosylation, acetylation, phosphorylation, and the like.

Amino acid derivatives can also include modifications to the native sequence, such as deletions, additions and substitutions (generally conservative in nature), so long as the polypeptide or peptide maintains the desired activity (e.g., inhibits activity of a urokinase receptor). These modifications may be deliberate, as through site-directed mutagenesis, or may be accidental, such as through mutations of hosts that produce the proteins or errors due to PCR amplification. Furthermore, modifications may be made that have one or more of the following effects: increasing affinity and/or specificity for a urokinase receptor and facilitating cell processing. Ligands described herein can be made recombinantly, synthetically, or in tissue culture.

A urokinase (uPA) polynucleotide, nucleic acid, oligonucleotide, protein, polypeptide, or peptide refers to a molecule derived from any source. The molecule need not be physically derived from an organism, but may be synthetically or recombinantly produced. A number of uPA nucleic acid and protein sequences are known.

Representative uPa growth factor domain (GFD) sequences are presented in SEQ ID NOS: 1-4 and additional representative sequences are listed in the National Center for Biotechnology Information (NCBI) database. See, for example, NCBI entries: Accession Nos. NG_011904, NM 001145031 , NM_001319191, NM_002658, XM O 11539866, NM_0Q8873, XM_017315919, NM_013085, NM_001163593, XM__014180296, NM 74147, NM__001082011, NM__001082011 , NM__001252350, XM_020711866, NM_001201525, XM_021480008, and XM__009306977; all of which sequences (as entered by the date of filing of this application) are herein incorporated by reference.

Any of these sequences or a variant thereof comprising a sequence having at least about 80-100% sequence identity thereto, including any percent identity within this range, such as 81 , 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99% sequence identity thereto, can be used to construct a ligand of a urokinase receptor (e.g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain), or a nucleic acid encoding a ligand of a urokinase receptor, as described herein.

A vitronectin polynucleotide, nucleic acid, oligonucleotide, protein, polypeptide, or peptide refers to a molecule derived from any source. The molecule need not be physically derived from an organism, but may be synthetically or recombinantly produced. A number of vitronectin nucleic acid and protein sequences are known.

Representative vitronectin somatomedin B (SMB) domain sequences are presented in SEQ ID NOS:5-8 and additional representative sequences are listed in the National Center for Biotechnology Information (NCBI) database. See, for example, NCBI entries: Accession Nos. NM__000638, NM__01 l707, NM 214104, NM_001082292,

XM_008309129, XM_013942476, XM_012825212, XMJ303968478, XM__Q06Q95517, XM_ 004385445, XM 004635422, XM 003469567, XM 011360077, XM 003770025, XM_003996497, XM_004543802, XM_007259609, XM_022595435, XM_848947,

XM 022210765, XM 003912492, XM 008066715, XM 021704164, XM 012449385, XM_021177950, XM 021213035, and XM O 19982376; all of which sequences (as entered by the date of filing of this application) are herein incorporated by reference.

Any of these sequences or a variant thereof comprising a sequence having at least about 80-100% sequence identity thereto, including any percent identity within this range, such as 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99% sequence identity thereto, can be used to construct a ligand of a urokinase receptor (e.g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain), or a nucleic acid encoding a ligand of a urokinase receptor, as described herein.

The terms "fusion protein," "fusion polypeptide," or "fusion peptide" as used herein refer to a fusion comprising a growth factor domain (GFD) of urokinase (uPA) in combination with a somatomedin B (SMB) domain of vitronectin as part of a single continuous chain of amino acids, which chain does not occur in nature. The GFD and SMB domains may be connected directly to each other by peptide bonds or may be separated by intervening amino acid sequences (i.e., a linker). Fusion proteins may also contain an immunoglobulin Fc domain as well as other sequences exogenous to the GFD and SMB domains. For example, a fusion protein may also include targeting, localization, or tag sequences. In addition, a fusion protein may be conjugated to an anti- cancer therapeutic agent or a diagnostic agent.

By "fragment" is intended a mol ecule consi sting of only a part of the intact full- length sequence and structure. The fragment can include a C -terminal deletion an N- terminal deletion, and/or an internal deletion of the polypeptide. Active fragments of a particular protein or polypeptide will generally include at least about 5-14 contiguous amino acid residues of the full length molecule, but may include at least about 15-25 contiguous amino acid residues of the full length molecule, and can include at least about 20-50 or more contiguous amino acid residues of the full length molecule, or any integer between 5 amino acids and the full length sequence, provided that the fragment in question retains biological activity, such as anti -tumor activity or inhibitory activity (e.g., the ability to inhibit a urokinase receptor), as defined herein.

"Substantially purified" generally refers to isolation of a substance (compound, polynucleotide, protein, polypeptide, peptide composition) such that the substance comprises the majority percent of the sample in which it resides. Typically in a sample, a substantially purified component comprises 50%, preferably 80%-85%, more preferably 90-95% of the sample. Techniques for purifying polynucleotides and polypeptides of interest are well-known in the art and include, for example, ion-exchange

chromatography, affinity chromatography and sedimentation according to density.

By "isolated" is meant, when referring to a polypeptide or peptide, that the indicated molecule is separate and discrete from the whole organism with which the molecule is found in nature or is present in the substantial absence of other biological macro molecules of the same type. The term "isolated" with respect to a polynucleotide is a nucleic acid molecul e devoid, in whole or part, of sequences normally associated with it in nature; or a sequence, as it exists in nature, but having heterologous sequences in association therewith; or a molecule disassociated from the chromosome.

A ligand is said to "interact" with a receptor if it binds specifically (e.g., in a iock- and-key type mechanism), non-specifically or in some combination of specific and non specific binding. A first ligand "interacts preferentially" with a receptor if it binds (non- specifically and/or specifically) to the receptor with greater affinity and/or greater specificity than it binds to other proteins (e.g., binds to urokinase receptor to a greater degree than to other proteins). The term "affinity" refers to the strength of binding and can be expressed quantitatively as a dissociation constant (Kd). It is to be understood that specific binding does not necessarily require interaction between specific amino acid residues and/or motifs of each peptide. For example, in certain embodiments, the ligands described herein interact preferentially with a urokinase receptor but, nonetheless, may be capable of binding other proteins at a weak, yet detectable, level (e.g., 10% or less of the binding shown to the urokinase receptor). Typically, weak binding, or background binding, is readily discernible from the preferential interaction with the protein or receptor of interest, e.g., by use of appropriate controls.

The term "antagonist" as used herein refers to any molecule that inhibits urokinase receptor activity. For example, the antagonist may comprise a fusion protein comprising a urokinase GFD linked to a vitronectin SM B domain. Inhibition of the urokinase receptor may be complete or partial (i.e., all activity, some activity, or most activity is blocked by an inhibitor). For example, an antagonist may reduce the urokinase receptor activity by 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%, or any amount in between as compared to native or control levels. As used herein, the terms "detection agent", "diagnostic agent", and "detectable label" are used interchangeably and refer to a molecule or substance capable of detection, including, but not limited to, fluorescers, chemiluminescers, chromophores,

bioluminescent proteins, enzymes, enzyme substrates, enzyme cofactors, enzyme inhibitors, isotopic labels, semiconductor nanoparticles, dyes, metal ions, metal sols, ligands (e.g., biotin, streptavidin or haptens) and the like. The term "fluorescer” refers to a substance or a portion thereof which is capable of exhibiting fluorescence in the detectable range. Particular examples of labels which may be used in the practice of the invention include, but are not limited to, SYBR green, SYBR gold, a CAL Fluor dye such as CAL Fluor Gold 540, CAL Fluor Orange 560, CAL Fluor Red 590, CAL Fluor Red 610, and CAL Fluor Red 635, a Quasar dye such as Quasar 570, Quasar 670, and Quasar 705, an Alexa Fluor such as Alexa Fluor 350, Alexa Fluor 488, Alexa Fluor 546, Alexa Fluor 555, Alexa Fluor 594, Alexa Fluor 647, and Alexa Fluor 784, a cyanine dye such as Cy 3, Cy3.5, Cy5, Cy5.5, and Cy7, fluorescein, 2', 4', 5', 7'-tetrachloro-4-7- dichlorofluorescein (TET), carboxyfluorescein (FAM), 6-carboxy-4',5'-dichloro-2',7'- dimethoxyfluorescein (JOE), hexachlorofluorescein (FIEX), rhodamine, carboxy-X- rhodamine (ROX), tetramethyl rhodamine (TAMRA), FITC, dansyl, umbelliferone, dimethyl acridinium ester (DMAE), Texas red, luminol, and quantum dots, enzymes such as alkaline phosphatase (AP), beta-lactamase, chloramphenicol acetyltransferase (CAT), adenosine deaminase (ADA), aminoglycoside phosphotransferase (neo ^r , 0418 dihydrofolate reductase (DPIFR), hygromycin-B-phosphotransferase (HPH), thymidine kinase (TK), b-galactosidase (lacZ), and xanthine guanine phosphoribosyltransferase (XGPRT), beta-glucuronidase (gus), placental alkaline phosphatase (PLAP), and secreted embryonic alkaline phosphatase (SEAP). Enzyme tags are used with their cognate substrate. The terms also include chemiluminescent labels such as luminol, isoluminol, acridinium esters, and peroxy oxalate and bioluminescent proteins such as firefly luciferase, bacterial luciferase, Renilla luciferase, and aequorin. The terms also include isotopic labels, including radioactive and non-radioactive isotopes, such as, ³ H, ² H, ^12lJ I, ^m I, ¹²⁴ I, ¹²⁵ l, ¹³¹ 1, ³⁵ S, ^U C, ¹³ C, ¹⁴ C, ³² P , ¹⁵ N, ¹³ N, ¹¹⁰ In, In, ¹⁷⁷ Lu, ¹⁸ F, ⁵² Fe, ⁶² Cu, ⁶⁴ Cu, ⁶⁷ Cu, ⁶⁷ Ga, ⁶⁸ Ga, ⁸⁶ Y, ⁹⁰ Y, ⁸⁹ Zr, ^94m Tc, ⁹⁴ Tc, ^99m Tc, ¹⁵⁴ Gd, ¹⁵⁵ Gd, ¹⁵⁶ Gd, ¹⁵⁷ Gd, ¹⁵⁸ Gd, ¹⁵ 0, ¹⁸⁶ Re, ¹⁸⁸ Re, ⁵¹ M,’ ^2m Mn, ⁵⁵ Co, ² As, ⁷ ¾r, ⁷⁶ Br, ^82m Rb, and ⁸³ Sr. The terms also include color-coded microspheres of known fluorescent light intensities (see e g., microspheres with xMAP technology produced by Luminex (Austin, TX); microspheres containing quantum dot nanocrystals, for example, containing different ratios and combinations of quantum dot colors (e.g., Qdot nanocrystals produced by Life Technologies (Carlsbad, CA); glass coated metal nanoparticles (see e.g., SERS nanotags produced by Nanoplex Technologies, Inc. (Mountain View, CA); barcode materials (see e.g., sub-micron sized striped metallic rods such as Nanobarcodes produced by Nanoplex Technologies, Inc.), encoded microparticles with colored bar codes (see e.g., CellCard produced by Vitra Bioscience, vitrabio.com), glass microparticles with digital holographic code images (see e.g., CyVera microbeads produced by Illumina (San Diego, CA), near infrared (NIR) probes, and nanoshells. The terms also include contrast agents such as ultrasound contrast agents (e.g. SonoVue microbubbles comprising sulfur hexafluoride, Optison microbubbles comprising an albumin shell and octafluoropropane gas core, Levovist microbubbles comprising a lipid/galactose shell and an air core, Perflexane lipid microspheres comprising perfluorocarbon microbubbles, and Perflutren lipid

microspheres comprising octafluoropropane encapsulated in an outer lipid shell), magnetic resonance imaging (MRI) contrast agents (e.g., gadodiamide, gadobenic acid, gadopentetic acid, gadoteridol, gadofosveset, gadoversetamide, gadoxetic acid), and radiocontrast agents, such as for computed tomography (CT), radiography, or fluoroscopy (e.g., diatrizoic acid, metrizoic acid, iodamide, iotalamic acid, ioxitalamic acid, ioglicic acid, acetrizoic acid, iocarmic acid, methiodal, diodone, metrizamide, iohexol, ioxaglic acid, iopamidol, iopromide, iotrolan, i over sol, iopentol, iodixanol, iomeprol, iobitridol, ioxilan, iodoxamic acid, iotroxic acid, ioglycamic acid, adipiodone, iobenzamic acid, iopanoic acid, iocetamic acid, sodium iopodate, tyropanoic acid, and calcium iopodate).

"Pharmaceutically acceptable excipient or carrier" refers to an excipient that may optionally be included in the compositions of the invention and that causes no significant adverse toxicological effects to the patient. "Pharmaceutically acceptable salt" includes, but is not limited to, amino acid salts, salts prepared with inorganic acids, such as chloride, sulfate, phosphate, diphosphate, bromide, and nitrate salts, or salts prepared from the corresponding inorganic acid form of any of the preceding, e.g., hydrochloride, etc., or salts prepared with an organic acid, such as malate, maleate, fumarate, tartrate, succinate, ethyl succinate, citrate, acetate, lactate, methanesulfonate, benzoate, ascorbate, para-toluenesulfonate, palmoate, salicylate and stearate, as well as estolate, gluceptate and lactobionate salts. Similarly salts containing pharmaceutically acceptable cations include, but are not limited to, sodium, potassium, calcium, aluminum, lithium, and ammonium (including substituted ammonium).

The terms "tumor," "cancer" and "neoplasia" are used interchangeably and refer to a cell or population of cells whose growth, proliferation or survival is greater than growth, proliferation or survival of a normal counterpart cell, e g. a cell proliferative, hyperproliferative or differentiative disorder. Typically, the growth is uncontrolled. The term "malignancy" refers to invasion of nearby tissue. The term "metastasis" or a secondary', recurring or recurrent tumor, cancer or neoplasia refers to spread or dissemination of a tumor, cancer or neoplasia to other sites, locations or regions within the subj ect, in which the sites, locations or regions are distinct from the primary tumor or cancer. Neoplasia, tumors and cancers include benign, malignant, metastatic and non- metastatic types, and include any stage (I, II, III, IV or V) or grade (Gl, G2, G3, etc.) of neoplasia, tumor, or cancer, or a neoplasia, tumor, cancer or metastasis that is progressing, worsening, stabilized or in remission. In particular, the terms "tumor," "cancer" and "neoplasia" include carcinomas, such as squamous cell carcinoma, adenocarcinoma, adenosquamous carcinoma, anaplastic carcinoma, large cell carcinoma, and small cell carcinoma. These terms include, but are not limited to, breast cancer, prostate cancer, lung cancer, ovarian cancer, testicular cancer, colon cancer, rectal cancer, pancreatic cancer, gastrointestinal cancer, hepatic cancer, endometrial cancer, leukemia, lympho a, adrenal cancer, thyroid cancer, pituitary cancer, adrenocortical cancer, renal cancer, brain cancer (e.g., glioblastoma and astrocytoma), skin cancer (e.g., basal-cell cancer, squamous-cell cancer, and melanoma), head cancer, neck cancer, oral cavity cancer, tongue cancer, and esophageal cancer.

An "effective amount" of a ligand of a urokinase receptor (e.g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain), or a nucleic acid encoding such a ligand is an amount sufficient to effect beneficial or desired results, such as an amount that binds to and/or inhibits a urokinase receptor. An effective amount can be administered in one or more administrations, applications or dosages.

By "anti-tumor activity" is intended a reduction in the rate of cell proliferation, and hence a decline in growth rate of an existing tumor or in a tumor that arises during therapy, and/or destruction of existing neoplastic (tumor) cells or newly formed neoplastic cells, and hence a decrease in the overall size of a tumor during therapy. Such activity can be assessed using animal models.

By "therapeutically effective dose or amount" of a ligand of a urokinase receptor (e g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain), or a nucleic acid encoding such a ligand is intended an amount that, when administered as described herein, brings about a positive therapeutic response, such as anti-tumor activity. The exact amount required will vary' from subject to subject, depending on the species, age, and general condition of the subject, the severity of the condition being treated, the particular drug or drugs employed, mode of administration, and the like. An appropriate "effective" amount in any individual case may be determined by one of ordinary skill in the art using routine experimentation, based upon the information provided herein.

The term "tumor response" as used herein means a reduction or elimination of all measurable lesions. The criteria for tumor response are based on the WHO Reporting Criteria [WHO Offset Publication, 48-World Health Organization, Geneva, Switzerland, (1979)]. Ideally, all uni- or bidimensionally measurable lesions should be measured at each assessment. When multiple lesions are present in any organ, such measurements may not be possible and, under such circumstances, up to 6 representative lesions should be selected, if available. The term "complete response" (CR) as used herein means a complete

disappearance of all clinically detectable malignant disease, determined by 2 assessments at least 4 weeks apart.

The term "partial response" (PR) as used herein means a 50% or greater reduction from baseline in the sum of the products of the longest perpendicular diameters of all measurable disease without progression of evaluable disease and without evidence of any new lesions as determined by at least two consecutive assessments at least four weeks apart. Assessments should show a partial decrease in the size of lytic lesions, recalcifications of lytic lesions, or decreased density of blastic lesions.

By "subject" is meant any member of the subphylum chordata, including, without limitation, humans and other primates, including non-human primates such as chimpanzees and other apes and monkey species; farm animals such as cattle, sheep, pigs, goats and horses; domestic mammals such as dogs and cats; laboratory animals including rodents such as mice, rats and guinea pigs; birds, including domestic, wild and game birds such as chickens, turkeys and other gallinaceous birds, ducks, geese, and the like.

"Homology" refers to the percent identity between two polynucleotide or two polypeptide molecules. Two nucleic acid, or two polypeptide sequences are

"substantially homologous" to each other when the sequences exhibit at least about 50% sequence identity, preferably at least about 75% sequence identity, more preferably at least about 80% 85% sequence identity, more preferably at least about 90% sequence identity, and most preferably at least about 95% 98% sequence identity over a defined length of the molecules. As used herein, substantially homologous also refers to sequences showing complete identity to the specified sequence.

In general, "identity" refers to an exact nucleotide to nucleotide or amino acid to amino acid correspondence of two polynucleotides or polypeptide sequences, respectively. Percent identity can be determined by a direct comparison of the sequence information between two molecules by aligning the sequences, counting the exact number of matches between the two aligned sequences, dividing by the length of the shorter sequence, and multiplying the result by 100. Readily available computer programs can be used to aid in the analysis, such as ALIGN, Dayhoff, M O. in Atlas of Protein Sequence and Structure M.O. Dayhoff ed., 5 Suppl. 3:353 358, National biomedical Research Foundation, Washington, DC, which adapts the local homology algorithm of Smith and Waterman Advances in Appl. Math. 2:482 489, 1981 for peptide analysis. Programs for determining nucleotide sequence identity are available in the Wisconsin Sequence Analysis Package, Version 8 (available from Genetics Computer Group, Madison, WI) for example, the BESTFIT, FASTA and GAP programs, which also rely on the Smith and Waterman algorithm. These programs are readily utilized with the default parameters recommended by the manufacturer and described in the Wisconsin Sequence Analysis Package referred to above. For example, percent identity of a particular nucleotide sequence to a reference sequence can be determined using the homology algorithm of Smith and Waterman with a default scoring table and a gap penalty of six nucleotide positions.

Another method of establi shing percent identity in the context of the present invention is to use the MPSRCH package of programs copyrighted by the University of Edinburgh, developed by John F. Collins and Shane S. Sturrok, and distributed by IntelliGenetics, Inc. (Mountain View, CA) From this suite of packages, the Smith Waterman algorithm can be employed where default parameters are used for the scoring table (for example, gap open penalty of 12, gap extension penalty of one, and a gap of six). From the data generated the "Match" value reflects "sequence identity." Other suitable programs for calculating the percent identity or similarity between sequences are generally known in the art, for example, another alignment program is BLAST, used with default parameters. For example, BLASTN and BLASTP can be used using the following default parameters: genetic code = standard; filter = none; strand = both; cutoff = 60; expect = 10; Matrix = BLOSUM62; Descriptions = 50 sequences; sort by = HIGH SCORE; Databases = non-redundant, GenBank + EMBL + DDBJ + PDB + GenBank CDS translations + Swiss protein + Spupdate + PIR. Details of these programs are readily available.

Alternatively, homology can be determined by hybridization of polynucleotides under conditions which form stable duplexes between homologous regions, followed by digestion with single stranded specific nuclease(s), and size determination of the digested fragments. DNA sequences that are substantially homologous can be identified in a Southern hybridization experiment under, for example, stringent conditions, as defined for that particular system. Defining appropriate hybridization conditions is within the skill of the art. See, e.g., Sambrook et al., supra, DNA Cloning, supra, Nucleic Acid Hybridization, supra.

"Recombinant" as used herein to describe a nucleic acid molecule means a polynucleotide of genomic, cDNA, viral, semisynthetic, or synthetic origin which, by virtue of its origin or manipulation, is not associated with all or a portion of the polynucleotide with which it is associated in nature. The term "recombinant" as used with respect to a protein or polypeptide means a polypeptide produced by expression of a recombinant polynucleotide. In general, the gene of interest is cloned and then expressed in transformed organisms, as described further below. The host organism expresses the foreign gene to produce the protein under expression conditions

The term "transformation" refers to the insertion of an exogenous polynucleotide into a host cell, irrespective of the method used for the insertion. For example, direct uptake, transduction or f-mating are included. The exogenous polynucleotide may be maintained as a non-integrated vector, for example, a plasmid, or alternatively, may be integrated into the host genome.

"Recombinant host cells," "host cells," "cells", "cell lines," "cell cultures, " and other such terms denoting microorganisms or higher eukaryotic cell lines cultured as unicellular entities refer to cells which can be, or have been, used as recipients for recombinant vector or other transferred DNA, and include the original progeny of the original cell which has been transfected.

A "coding sequence" or a sequence which "encodes" a selected polypeptide, is a nucleic acid molecule which is transcribed (in the case of DNA) and translated (in the case of mRNA) into a polypeptide in vivo when placed under the control of appropriate regulatory sequences (or "control elements"). The boundaries of the coding sequence can be determined by a start codon at the 5' (amino) terminus and a translation stop codon at the 3' (carboxy) terminus. A coding sequence can include, but is not limited to, cDNA from viral, prokaryotic or eukaryotic mRNA, genomic DNA sequences from viral or prokaryotic DNA, and even synthetic DNA sequences. A transcription termination sequence may be located 3' to the coding sequence.

Typical "control elements," include, but are not limited to, transcription promoters, transcription enhancer elements, transcription termination signals, polyadenylation sequences (located 3' to the translation stop codon), sequences for optimization of initiation of translation (located 5’ to the coding sequence), and translation termination sequences.

"Operably linked" refers to an arrangement of elements wherein the components so described are configured so as to perform their usual function. Thus, a given promoter operably linked to a coding sequence is capable of effecting the expression of the coding sequence when the proper enzymes are present. The promoter need not be contiguous with the coding sequence, so long as it functions to direct the expression thereof. Thus, for example, intervening untranslated yet transcribed sequences can be present between the promoter sequence and the coding sequence and the promoter sequence can still be considered "operably linked" to the coding sequence.

"Encoded by" refers to a nucleic acid sequence which codes for a polypeptide sequence, wherein the polypeptide sequence or a portion thereof contains an amino acid sequence of at least 3 to 5 amino acids, more preferably at least 8 to 10 amino acids, and even more preferably at least 15 to 20 amino acids from a polypeptide encoded by the nucleic acid sequence.

"Expression cassette" or "expression construct" refers to an assembly which is capable of di recting the expression of the sequence(s) or gene(s) of interest. An expression cassette generally includes control elements, as described above, such as a promoter which is operably linked to (so as to direct transcription of) the sequence(s) or gene(s) of interest, and often includes a polyadenyl tion sequence as well. Within certain embodiments of the invention, the expression cassette described herein may be contained within a plasmid construct. In addition to the components of the expression cassette, the plasmid construct may also include, one or more selectable markers, a signal which allows the plasmid construct to exist as single stranded DNA (e.g., a M13 origin of replication), at least one multiple cloning site, and a "mammalian" origin of replication (e.g., a SV40 or adenovirus origin of replication).

"Purified polynucleotide" refers to a polynucleotide of interest or fragment thereof which is essentially free, e.g., contains less than about 50%, preferably less than about 70%, and more preferably less than about at least 90%, of the protein with which the polynucleotide is naturally associated. Techniques for purifying polynucleotides of interest are well-known in the art and include, for example, disruption of the cell containing the polynucleotide with a chaotropic agent and separation of the

polynucleotide(s) and proteins by ion-exchange chromatography, affinity

chromatography and sedimentation according to density.

The term "transfection" is used to refer to the uptake of foreign DNA by a cell. A cell has been "transfected" when exogenous DNA has been introduced inside the cell membrane. A number of transfection techniques are generally known in the art. See, e.g., Graham et al. (1973) Virology, 52:456, Sambrook et al. (2001) Molecular Cloning, a laboratory^ manual, 3rd edition, Cold Spring Harbor Laboratories, New York, Davis et al. (1995) Basic Methods in Molecular Biology, 2nd edition, McGraw-Hill, and Chu et al. (1981) Gene 13 :197 Such techniques can be used to introduce one or more exogenous DNA moieties into suitable host cells. The term refers to both stable and transient uptake of the genetic material, and includes uptake of peptide- or antibody-linked DNAs.

A "vector" is capable of transferring nucleic acid sequences to target cells (e.g., viral vectors, non-viral vectors, particulate carriers, and liposomes). Typically, "vector construct," "expression vector," and "gene transfer vector," mean any nucleic acid construct capable of directing the expression of a nucleic acid of interest and which can transfer nucleic acid sequences to target cells. Thus, the term includes cloning and expression vehicles, as well as viral vectors.

The terms "variant," "analog" and "mutein" refer to biologically active derivatives of the reference molecule that retain desired activity, such as the ability to inhibit a urokinase receptor and having anti -tumor activity. In general, the terms "variant" and "analog" refer to compounds having a native polypeptide sequence and structure with one or more amino acid additions, substitutions (generally conservative in nature) and/or deletions, relative to the native molecule, so long as the modifications do not destroy biological activity and which are "substantially homologous" to the reference molecule as defined below. In general, the amino acid sequences of such analogs will have a high degree of sequence homology to the reference sequence, e.g., amino acid sequence homology of more than 50%, generally more than 60%-70%, even more particularly 80%-85% or more, such as at least 90%-95% or more, when the two sequences are aligned. Often, the analogs will include the same number of amino acids but will include substitutions, as explained herein. The term "mutein" further includes polypeptides having one or more amino acid-like molecules including but not limited to compounds comprising only amino and/or imino molecules, polypeptides containing one or more analogs of an amino acid (including, for example, unnatural amino acids, etc.), polypeptides with substituted linkages, as well as other modifications known in the art, both naturally occurring and non-naturally occurring (e.g., synthetic), cyclized, branched molecules and the like. The term also includes molecules comprising one or more N- substituted glycine residues (a "peptoid") and other synthetic amino acids or peptides. (See, e.g., U.S. Patent Nos. 5,831,005; 5,877,278; and 5,977,301; Nguyen et al., Chem. Biol. (2000) 7:463-473; and Simon et al„ Proc. Natl. Acad. Sci. USA (1992) 89:9367- 9371 for descriptions of peptoids). Methods for making polypeptide analogs and muteins are known in the art and are described further below.

As explained above, analogs generally include substitutions that are conservative in nature, i.e., those substitutions that take place within a family of amino acids that are related in their side chains. Specifically, amino acids are generally divided into four families: (1) acidic— aspartate and glutamate; (2) basic -- lysine, arginine, histidine; (3) non-polar— alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine, tryptophan; and (4) uncharged polar— glycine, asparagine, glutamine, cysteine, serine threonine, and tyrosine. Phenylalanine, tryptophan, and tyrosine are sometimes classified as aromatic amino acids. For example, it is reasonably predictable that an isolated replacement of leucine with isoleucine or valine, an aspartate with a glutamate, a threonine with a serine, or a similar conservative replacement of an amino acid with a structurally related amino acid, will not have a major effect on the biological activity. For example, the polypeptide of interest may include up to about 5-10 conservative or non-conservative amino acid substitutions, or even up to about 15-25 conservative or non-conservative amino acid substitutions, or any integer between 5-25, so long as the desired function of the molecule remains intact. One of skill in the art may readily determine regions of the mol ecule of interest that can tolerate change by reference to HoppAV oods and Kyte-Doolittle plots, well known in the art.

"Gene transfer" or "gene delivery" refers to methods or systems for reliably inserting DNA or RNA of interest into a host cell. Such methods can result in transient expression of non-integrated transferred DNA, extrachromosomal replication and expression of transferred replicons (e.g., epi somes), or integration of transferred genetic material into the genomic DNA of host cells. Gene delivery expression vectors include, but are not limited to, vectors derived from bacterial plasmid vectors, viral vectors, non- viral vectors, alphaviruses, pox viruses and vaccinia viruses.

The term "derived from" is used herein to identify the original source of a molecule but is not meant to limit the method by which the molecule is made which can be, for example, by chemical synthesis or recombinant means.

A polynucleotide "derived from" a designated sequence refers to a polynucleotide sequence which comprises a contiguous sequence of approximately at least about 6 nucleotides, preferably at least about 8 nucleotides, more preferably at least about 10-12 nucleotides, and even more preferably at least about 15-20 nucleotides corresponding, i.e., identical or complementary to, a region of the designated nucleotide sequence. The derived polynucleotide will not necessarily be derived physically from the nucleotide sequence of interest, but m ay be generated in any manner, including, but not limited to, chemical synthesis, replication, reverse transcription or transcription, which is based on the information provided by the sequence of bases in the region(s) from which the polynucleotide is derived. As such, it may represent either a sense or an anti sense orientation of the original polynucleotide. P. Modes of Carrying Out the Invention

Before describing the present invention in detail, it is to be understood that this invention is not limited to particular formulations or process parameters as such may, of course, vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments of the invention only, and is not intended to be limiting.

Although a number of methods and materials similar or equivalent to those described herein can be used in the practice of the present invention, the preferred materials and methods are described herein.

The present invention is based on the discovery that fusion proteins comprising a urokinase GFD linked to a vitronectin SMB domain can be used as ligands of the urokinase receptor for treatment of cancer. In particular, the inventors have engineered bispecific protein ligands that simultaneously bind to two distinct sites on the urokinase receptor with an affinity superior to the native ligands and have demonstrated that such ligands have anti -tumor activity (see Examples). In order to further an understanding of the invention, a more detailed discussion is provided below regarding ligands of the urokinase receptor and their use in treating cancer.

A. Ligands of the Urokinase Receptor

In one aspect, the invention provides ligands having anti-tumor activity that are capable of binding, preferably speci fically binding, to a urokinase receptor. The ligands comprise a fusion protein comprising a growth factor domain (GFD) of urokinase (uPA) linked to a somatomedin B (SMB) domain of vitronectin Inhibition of the urokinase receptor may be complete or partial (i.e., all activity, some activity, or most activity is blocked by an inhibitor). In certain embodiments, a ligand reduces urokinase receptor activity by 70% to 100%, or any amount in this range, such as 70%, 72%, 74%, 76%, 78%, 80%, 82%, 84%, 86%, 88%, 90%, 92%, 94%, 96%, 98%, or 100% as compared to native or control levels in certain embodiments, the ligand binds to a urokinase receptor with a dissociation constant (KD) of less than 0.2 nM, more preferably, less than 100 pM, and most preferably, less than 10 pM. Urokinase (uPA) nucleic acid and protein sequences may be derived from any source A number of uPA nucleic acid and protein sequences are known. Representative uPa growth factor domain (GFD) sequences are presented in SEQ ID NOS: 1-4 and additional representative sequences are listed in the National Center for Biotechnology Information (NCBI) database. See, for example, NCBI entries: Accession Nos.

NG_011904, NM_0Q 1145031, NM_001319191, NM_002658, XM_011539866, NM_008873, XM__017315919, NM__013085, NM__001163593, XM_014180296, NM_174147, NM_001082011, NM_001082011, NM_001252350, XM__020711866,

NM 001201525, XM 021480008, and XM 009306977; all of which sequences (as entered by the date of filing of this application) are herein incorporated by reference.

Any of these sequences or a variant thereof comprising a sequence having at least about 80-100% sequence identity thereto, including any percent identity within this range, such as 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99% sequence identity thereto, can be used to construct a ligand of a urokinase receptor (e.g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain), or a nucleic acid encoding a ligand of a urokinase receptor, as described herein. If desired, uPA polypeptides can contain other amino acid sequences, such as amino acid linkers or signal sequences, as well as ligands or tags useful in protein purification or detectable labels.

Vitronectin nucleic acid and protein sequences also may be derived from any source. A number of vitronectin nucleic acid and protein sequences are known.

Representative vitronectin somatomedin B (SMB) domain sequences are presented in SEQ ID NOS:5-8 and additional representative sequences are listed in the National Center for Biotechnology' Information (NCBI) database. See, for example, NCBI entries: Accession Nos. NM_000638, NM_011707, NM_214104, NM_001082292,

XM_ 008309129, XM_ 013942476, XM 012825212, XM 003968478, XM 006095517, XM_004385445, XM_004635422, XM_003469567, XM O 11360077, XM_003770025, XM 003996497, XM 004543802, XM 007259609, XM 022595435, XM 848947, XM_022210765, XM 003912492, XM_008066715, XM 021704164, XM O 12449385, XM_021177950, XM_ 021213035, and XM_019982376; all of which sequences (as entered by the date of filing of this application) are herein incorporated by reference. Any of these sequences or a variant thereof comprising a sequence having at least about 80-100% sequence identity thereto, including any percent identity within this range, such as 81 , 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99% sequence identity thereto, can be used to construct a ligand of a urokinase receptor (e.g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain), or a nucleic acid encoding a ligand of a urokinase receptor, as described herein. If desired, vitronectin polypeptides can contain other amino acid sequences, such as amino acid linkers or signal sequences, as well as ligands or tags useful in protein purification or detectable labels.

In certain embodiments, the ligand comprises engineered GFD and/or SMB domains that have been modified to improve affinity for the urokinase receptor, inhibition of the urokinase receptor, and/or anti-tumor activity. For example, the ligand may be engineered to comprise one or more mutations that improve its ability to bind to and/or inhibit the urokinase receptor and its efficacy in treating cancer.

In certain embodiments, the fusion protein comprises an engineered GFD comprising at least one amino acid substitution selected from the group consisting of R30W, K37G, K37R, Q39R, E41K, E41 R, and H42Y, wherein positions of the amino acids are numbered relative to the reference murine GFD sequence of SEQ ID NO: 1. In one embodiment, the GFD comprises the R30W, K37R, Q39R, E41R, and H42Y amino acid substitutions. Although the foregoing numbering is relative to reference murine GFD sequence of SEQ ID NO: 1, it is to be understood that the corresponding positions in other GFDs obtained from other species are also intended to be encompassed by the present invention.

In other embodiments, the fusion protein comprises an engineered GFD comprising at least one mutation comprising an amino acid substitution selected from the group consisting of L4W, Q6R, H29Q, K36G, Q40R, H41R, and K46R, and wherein positions of the amino acids are numbered relative to the reference human GFD sequence of SEQ ID NO:3. In one embodiment, the GFD comprises the L4W, Q6R, H29Q, K36G, Q40R, H41R, and K46R amino acid substitutions. Although the foregoing numbering is relative to reference human GFD sequence of SEQ ID NO:3, it is to be understood that the corresponding positions in other GFDs obtained from other species are also intended to be encompassed by the present invention.

In further embodiments, the fusion protein comprises an engineered SMB domain comprising at least one amino acid substitution selected from the group consisting of DIN, E3G, M14V, and KI E, wherein positions of the amino acids are numbered relative to the reference murine SMB domain sequence of SEQ ID NO:5. In one embodiment, the SMB domain comprises the DIN, E3G, M14V, and K17E amino acid substitutions. Although the foregoing numbering is relative to reference murine SMB domain sequence of SEQ ID NO:5, it is to be understood that the corresponding positions in other SMBs obtained from other species are also intended to be encompassed by the present invention.

In yet other embodiments, the fusion protein comprises an engineered SMB domain comprising at least one amino acid substitution selected from the group consisting of D1G, E3G, K18E, and P41 S, and wherein positions of the amino acids are numbered relative to the reference human SMB domain sequence of SEQ ID NO:7. In one embodiment, the SMB domain comprises the D1G, E3G, K18E, and P41 S amino acid substitutions. Although the foregoing numbering is relative to reference human SMB domain sequence of SEQ ID NO:7, it is to be understood that the corresponding positions in other SMBs obtained from other species are also intended to be encompassed by the present invention.

In certain embodiments, the fusion protein further comprises one or more linkers. Linkers are typically short peptide sequences of 2-30 amino acid residues, often composed of glycine and/or serine residues. Linker amino acid sequences will typically be short, e.g., 25 or fewer amino acids (i.e., 25, 24, 23, 22, 21, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1). Examples include short peptide sequences which facilitate cloning, poly-glycine linkers (Gly _n where n = 2, 3, 4, 5, 6, 7, 8, 9, 10 or more), histidine tags (His _n where n = 3, 4, 5, 6, 7, 8, 9, 10 or more), linkers composed of glycine and serine residues ([Gly-Gly-Gly-Gly-Ser] _n (SEQ ID NO: 9), wherein n = 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or more), GSAT, SEG, and Z-EGFR linkers.

Linkers may include restriction sites, which aid cloning and manipulation. Other suitable linker amino acid sequences will be apparent to those skilled in the art. (See e.g., Argos (1990) J. Mol Biol. 211(4):943-958; Crasto et al. (2000) Protein Eng 13:309-312;

George et al. (2002) Protein Eng. 15:871 -879; Arai et al. (2001) Protein Eng. 14:529-532; and the Registry of Standard Biological Parts (partsregistry.org/Protein domains/Linker).

In particular, the fusion protein may comprise a linker connecting the GFD to the SMB domain. In certain embodiments, the linker ranges in size from about 40 angstroms to about 50 angstroms in length, including any length within this range such as, 40, 41,

42, 43, 44, 45, 46, 47, 48, 49, or 50 angstroms in length. In another embodiment, the linker comprises a sequence selected from the group consisting of SEQ ID NO: l 1, SEQ ID NO: 14, and SEQ ID NO: 15, or a sequence displaying at least about 80-100% sequence identity thereto, including any percent identity within this range, such as 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% sequence identity thereto, wherein the ligand is capable of binding to tumor cells and/or inhibiting the urokinase receptor. In certain embodiments, the linker comprises at least one amino acid substitution selected from the group consisting of S5N, S10G, G12R, and G19S, and wherein positions of the amino acids are numbered relative to the reference linker sequence of SEQ ID NO: 11.

In certain embodiments, the ligand comprises a fusion protein comprising an amino acid sequence selected from the group consisting of SEQ ID NO: 12 and SEQ ID NO: 13, or a sequence displaying at least about 80-100% sequence identity thereto, including any percent identity within this range, such as 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% sequence identity thereto, wherein the ligand is capable of binding to tumor cells and/or inhibiting the urokinase receptor and has anti tumor activity. In certain embodiments, the N-terminus or the C -terminus of the fusion protein may be modified to enhance solubility, stability and/or improve delivery' to tumors. For example, the ends of the fusion protein may be modified by N-terminal acetylation and/or C-terminal amidation to reduce charge and increase stability.

In certain embodiments, the fusion protein further comprises an immunoglobulin Fc domain. The immunoglobulin Fc domain may be derived from an IgG (e.g., IgGl, IgG2, IgG3, or IgG4), IgM, IgE, IgA or IgD, or a combination or hybrid thereof. In one embodiment, the Fc fragment is derived from an IgGl or IgG2a isotype. In another embodiment, the Fc fragment is derived from a human immunoglobulin. The immunoglobulin Fc domain may be covalently linked directly to the N-terminal or C- terminal end of the fusion protein or connected indirectly through a linker to the fusion protein. In certain embodiments, the ligand comprises a fusion protein comprising an amino acid sequence selected from the group consisting of SEQ ID NOS: 17-20; or a sequence displaying at least about 80-100% sequence identity thereto, including any percent identity within this range, such as 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92,

93, 94, 95, 96, 97, 98, 99% sequence identity thereto, wherein the ligand is capable of binding to and/or inhibiting the urokinase receptor. In certain embodiments, the ligand targets tumors expressing the urokinase receptor in a mammal.

In certain embodiments, the fusion protein further comprises a tag sequence, which may be located, for example, at the N-terminus or C -terminus of the fusion protein. Exemplary tags that can be used in the practice of the invention include a His- tag, a Strep-tag, a TAP -tag, an S-tag, an SBP-tag, an Arg-tag, a calmodulin-binding peptide tag, a cellulose-binding domain tag, a DsbA tag, a c-myc tag, a glutathione S- transferase tag, a FLAG tag, a HAT -tag, a maltose-binding protein tag, a NusA tag, and a thioredoxin tag.

In certain embodiments, the fusion protein further comprises a signal peptide. Signal peptides are typically located at the N-terminus of the fusion protein and range in size from about 16 to 30 amino acids in length (i.e., 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or 30). Any signal peptide may be used that is capable of targeting the fusion protein to the cell membrane where the urokinase receptor is located.

Representative signal peptide sequences are listed in the Signal Peptide Database (proline.bic.nus.edu. sg/spdb/) and the Signal Peptide Website (signalpeptide.de/). In one embodiment, the signal peptide is a urokinase signal peptide.

In certain embodiments, the fusion protein further comprises a detectable label in order to facilitate detection of binding of the ligand to the urokinase receptor. Detectable labels suitable for use in the present invention include any composition detectable by spectroscopic, photochemical, biochemical, immunochemical, electrical, optical, or chemical means. Particular examples of labels that may be used with the invention include, but are not limited to radiolabels (e.g., Ή, ¹²⁵ 1, ³⁵ S, ¹⁴ C, or ³² P), phycoerythrin, Alexa dyes, fluorescein, YPet, CyPet, Cascade blue, allophycocyanin, Cy3, Cy5, Cy7, rhodamine, dansyl, umbelliferone, Texas red, luminol, acradimum esters, green fluorescent protein (GFP), enhanced green fluorescent protein (EGFP), yellow fluorescent protein (YFP), enhanced yellow fluorescent protein (EYFP), blue fluorescent protein (BFP), red fluorescent protein (RFP), Dronpa, Padron, mApple, mCherry, rsCherry, rsCherryRev, firefly luciferase, Renilla luciferase, biotin or other streptavidin- binding proteins, magnetic beads, electron dense reagents, NADPH, beta-galactosidase, horseradish peroxidase, glucose oxidase, alkaline phosphatase, chloramphenical acetyl transferase, and urease. Patents teaching the use of such labels include U.S. Pat. Nos. 3,817,837; 3,850,752; 3,939,350; 3,996,345; 4,277,437; 4,275,149; and 4,366,241.

The fusion protein may also be conjugated to a contrast agent suitable for imaging. Exemplary contrast agents include ultrasound contrast agents (e.g. SonoVue microbubbles comprising sulfur hexafluoride, Optison microbubbles comprising an albumin shell and octafluoropropane gas core, Levovist microbubbles comprising a lipid/galactose shell and an air core, Perflexane lipid microspheres comprising perfluorocarbon microbubbles, and Perflutren lipid microspheres comprising

octafluoropropane encapsulated in an outer lipid shell), magnetic resonance imaging (MRI) contrast agents (e.g., gadodi amide, gadobenic acid, gadopentetic acid, gadoteridol, gadofosveset, gadoversetamide, gadoxetic acid), and radiocontrast agents, such as for computed tomography (CT), radiography, or fluoroscopy (e.g., diatrizoic acid, metrizoic acid, iodamide, iotalamic acid, ioxitalamic acid, ioglicic acid, acetrizoic acid, iocarmic acid, methiodal, diodone, metrizamide, iohexol, ioxaglic acid, iopamidol, iopromide, iotrolan, ioversol, iopentol, iodixanol, iomeprol, iobitridol, ioxilan, iodoxamic acid, iotroxic acid, ioglycamic acid, adipiodone, iobenzamic acid, iopanoic acid, iocetamic acid, sodium iopodate, tyropanoic acid, and calcium iopodate). B. Production of Ligands

Ligands of the urokinase receptor (e.g., fusion proteins comprising a urokinase GFD linked to a vitronectin SMB domain) can be prepared in any suitable manner (e g., recombinant expression, purification from cell culture, chemical synthesis, etc.) and in various forms (e.g. tagged, labeled, li pi dated, ami dated, acetylated, etc.). Ligand fusion proteins may include naturally-occurring polypeptides, recombinantly produced polypeptides, synthetically produced polypeptides, or polypeptides produced by a combination of these methods. Means for preparing polypeptides are well understood in the art. Ligand fusion proteins are preferably prepared in substantially pure form (i.e. substantially free from other host cell or non-host cell proteins).

In one embodiment, the ligand fusion proteins are generated using recombinant techniques. One of skill in the art can readily determine nucleotide sequences that encode the desired fusion proteins using standard methodology and the teachings herein. Oligonucleotide probes can be devised based on the known sequences and used to probe genomic or cDNA libraries. The sequences can then be further isolated using standard techniques and, e.g., restriction enzymes employed to truncate the gene at desired portions of the full-length sequence. Similarly, sequences of interest can be isolated directly from cells and tissues containing the same, using known techniques, such as phenol extraction and the sequence further manipulated to produce the desired truncations. See, e.g., Sambrook et al., supra, for a description of techniques used to obtain and isolate DNA.

The sequences encoding fusion proteins can also be produced synthetically, for example, based on the known sequences. The nucleotide sequence can be designed with the appropriate codons for the particular amino acid sequence desired. The complete sequence is generally assembled from overlapping oligonucleotides prepared by standard methods and assembled into a complete coding sequence. See, e.g., Edge (1981) Nature 292:756; Nambair ei al. (1984) Science 223 : 1299; Jay et al (1984) J. Biol Chem.

259:631 1; Stemmer et al. (1995) Gene 164:49-53.

Recombinant techniques are readily used to clone sequences encoding peptides or polypeptides that can then be mutagenized in vitro by the replacem ent of the appropriate base pair(s) to result in the codon for the desired amino acid. Such a change can include as little as one base pair, effecting a change in a single amino acid, or can encompass several base pair changes. Alternatively, the mutations can be effected using a mismatched primer that hybridizes to the parent nucleotide sequence (generally cDNA corresponding to the RNA sequence), at a temperature below the melting temperature of the mismatched duplex. The primer can be made specific by keeping primer length and base composition within relatively narrow limits and by keeping the mutant base centrally located. See, e.g., Innis et al, (1990) PCR Applications: Protocols for

Functional Genomics; Zoller and Smith, Methods Enzymol. (1983) 100:468. Primer extension is effected using DNA polymerase, the product cloned and clones containing the mutated DNA, derived by segregation of the primer extended strand, selected.

Selection can be accomplished using the mutant primer as a hybridization probe. The technique is also applicable for generating multiple point mutations. See, e.g., Dalbie- McFarland et al. Proc. Natl Acad. Sci USA (1982) 79:6409.

Once coding sequences have been isolated and/or synthesized, they can be cloned into any suitable vector or replicon for expression. (See, also, Examples). As will be apparent from the teachings herein, a wide variety of vectors encoding modified polypeptides can be generated by creating expression constructs which operably link, in various combinations, polynucleotides encoding polypeptides having deletions or mutations therein.

Numerous cloning vectors are known to those of skill in the art, and the selection of an appropriate cloning vector is a matter of choice. Examples of recombinant DNA vectors for cloning and host cells which they can transform include the bacteriophage l (E. coli), pBR322 ( E . coli ), pACYC177 (E. coli), pKT230 (gram-negative bacteria), pGV1106 (gram-negative bacteria), pLAFRl (gram -negative bacteria), pME290 (non-E. coli gram-negative bacteria), pH VI 4 (E. coli and Bacillus subtilis), pBD9 {Bacillus), pIJ61 {Streptomyces), pUC6 {Streptomyc.es), YIp5 ( Saccharomyces ), YCpl9

{Saccharomyces) and bovine papilloma virus (mammalian cells). See, generally, DNA Cloning: Vols. I & II, supra, Sambrook et al. , supra, B. Perbal, supra. Insect cell expression systems, such as baculovirus systems, can also be used and are known to those of skill in the art and described in, e.g. , Summers and Smith, Texas Agricultural Experiment Station Bulletin No. 1555 (1987). Materials and methods for baculovirus/insect cell expression systems are commercially available in kit form from, inter alia, Invitrogen, San Diego CA ("MaxBac" kit).

Plant expression systems can also be used to produce the ligand fusion proteins described herein. Generally, such systems use virus-based vectors to transfect plant cells with heterologous genes. For a description of such systems, see, e.g., Porta et al., Mol. Biotech. (1996) 5:209-221; and Hackland et al., Arch. Virol. (1994) 139: 1-22.

Viral systems, such as a vaccinia-based infection/transfection system, as described in Tomei et al., J. Virol. (1993) 67:4017-4026 and Selby et al., J. Gen. Virol. (1993) 74: 1103-1 1 13, will also find use with the present invention. In this system, cells are first transfected in vitro with a vaccinia virus recombinant that encodes the bacteriophage T7 RNA polymerase. This polymerase displays exquisite specificity in that it only transcribes templates bearing T7 promoters. Following infection, cells are transfected with the DNA of interest, driven by a T7 promoter. The polymerase expressed in the cytoplasm from the vaccinia virus recombinant transcribes the transfected DNA into RNA that is then translated into protein by the host translational machinery. The method provides for high level, transient, cytoplasmic production of large quantities of RNA and its translation product(s).

The gene can be placed under the control of a promoter, ribosome binding site (for bacterial expression) and, optionally, an operator (collectively referred to herein as "control " elements), so that the DNA sequence encoding the desired polypeptide is transcribed into RNA in the host cell transformed by a vector containing this expression construction. The coding sequence may or may not contain a signal peptide or leader sequence. With the present invention, both the naturally occurring signal peptides or heterologous sequences can be used. Leader sequences can be removed by the host in post-translational processing. See , e.g., U.S. Patent Nos. 4,431 ,739; 4,425,437;

4,338,397. Such sequences include, but are not limited to, the TP A leader, as well as the honey bee mellitin signal sequence. Other regulatory sequences may also be desirable which allow for regulation of expression of the protein sequences relative to the growth of the host cell. Such regulatory sequences are known to those of skill in the art, and examples include those which cause the expression of a gene to be turned on or off in response to a chemical or physical stimulus, including the presence of a regulatory compound. Other types of regulatory elements may also be present in the vector, for example, enhancer sequences.

The control sequences and other regulatory sequences may be ligated to the coding sequence prior to insertion into a vector. Alternatively, the coding sequence can be cloned directly into an expression vector that already contains the control sequences and an appropriate restriction site.

In some cases, it may be necessary to modify the coding sequence so that it may be attached to the control sequences with the appropriate orientation; i.e., to maintain the proper reading frame. Mutants or analogs may be prepared by the deletion of a portion of the sequence encodi ng the protein, by insertion of a sequence, and/or by substitution of one or more nucleotides within the sequence. Techniques for modifying nucleotide sequences, such as site-directed mutagenesis, are well known to those skilled in the art. See, e.g., Sambrook et ah, supra; DNA Cloning, Vols. I and II, supra; Nucleic Acid Hybridization, supra.

The expression vector is then used to transform an appropriate host cell. A number of mammalian cell lines are known in the art and include immortalized cell lines available from the American Type Culture Collection (ATCC), such as, but not limited to, Chinese hamster ovary (CHO) cells, HeLa cells, baby hamster kidney (BHK) cells, monkey kidney cells (COS), human hepatocellular carcinoma cells (e.g., Hep G2), Vero293 cells, as well as others. Similarly, bacterial hosts such as E. coli, Bacillus sub tilis, and Streptococcus spp., will find use with the present expression constructs.

Yeast hosts useful in the present invention include inter alia, Saccharomyces cerevisiae, Candida albicans, Candida maltosa, Hamenula polymorpha, Kluyveromyces fragilis, Khtyveromyces lactis, Pichia guillerimondii, Pichia pastoris, Schizosaccharomyces pombe and Yarrow ia lipolytica. Insect cells for use with baculovirus expression vectors include, inter alia, Aedes aegypti, Autographa californica, Bombyx mori, Drosophila melanogaster, Spodoptera frugiperda , and Trichoplusia ni.

Depending on the expression system and host selected, the fusion proteins of the present invention are produced by growing host cells transformed by an expression vector described above under conditions whereby the protein of interest is expressed. The selection of the appropriate growth conditions is within the skill of the art.

In one embodiment, the transformed cells secrete the peptide or polypeptide product into the surrounding media. Certain regulatory sequences can be included in the vector to enhance secretion of the protein product, for example using a tissue

plasminogen activator (TPA) leader sequence, an interferon (g or a) signal sequence or other signal peptide sequences from known secretory proteins. The secreted peptide or polypeptide product can then be isolated by various techniques described herein, for example, using standard purification techniques such as but not limited to, hydroxyapatite resins, column chromatography, ion-exchange chromatography, size-exclusion chromatography, electrophoresis, HPLC, immunoadsorbent techniques, affinity chromatography, immunoprecipitation, and the like.

Alternatively, the transformed cells are disrupted, using chemical, physical or mechanical means, which lyse the cells yet keep the recombinant peptides or

polypeptides substantially intact. Intracellular proteins can also be obtained by removing components from the cell wall or membrane, e.g., by the use of detergents or organic solvents, such that leakage of the polypeptides occurs. Such methods are known to those of skill in the art and are described in, e.g., Protein Purification Applications: A Practical Approach, (Simon Roe, Ed., 2001).

For example, methods of disrupting cells for use with the present invention include but are not limited to: soni cation or ultrasonication; agitation; liquid or solid extrusion; heat treatment; freeze-thaw; desiccation; explosive decompression; osmotic shock; treatment with lytic enzymes including proteases such as trypsin, neuraminidase and lysozyme; alkali treatment; and the use of detergents and solvents such as bile salts, sodium dodecyl sulphate, Triton, NP40 and CHAPS. The particular technique used to disrupt the cells is largely a matter of choice and will depend on the cell type in which the polypeptide is expressed, culture conditions and any pre-treatment used.

Following disruption of the cells, cellular debris is removed, generally by centrifugation, and the intracellularly produced peptides or polypeptides are further purified, using standard purification techniques such as but not limited to, column chromatography, ion-exchange chromatography, size-exclusion chromatography, electrophoresis, HPLC, immunoadsorbent techniques, affinity chromatography, immunoprecipitation, and the like.

For example, one method for obtaining the fusion proteins of the present invention involves affinity purification, such as by immunoaffinity chromatography using antibodies (e.g., previously generated antibodies), or by lectin affinity chromatography. Particularly preferred lectin resins are those that recognize mannose moieties such as but not limited to resins derived from Galanthns nivalis agglutinin (GNA), Lens culinaris agglutinin (LCA or lentil lectin), Pisum sativum agglutinin (PSA or pea lectin), Narcissus pseudonarcissus agglutinin (NPA) and Allium ursinum agglutinin (AUA). The choice of a suitable affinity resin is within the skill in the art. After affinity purification, the peptides or polypeptides can be further purified using conventional techniques well known in the art, such as by any of the techniques described above.

Ligand fusion proteins can also be conveniently synthesized chemically, for example by any of several techniques that are known to those skilled in the peptide art. See, e.g., Fmoc Solid Phase Peptide Synthesis: A Practical Approach (W. C. Chan and Peter D. White eds., Oxford University Press, 1 ^st edition, 2000) ; N. Leo Benoiton, Chemistry of Peptide Synthesis (CRC Press; I ^st edition, 2005); Peptide Synthesis and Applications (Methods in Molecular Biology, John Howl ed., Humana Press, 1 ^st ed., 2005); and Pharmaceutical Formulation Development of Peptides and Proteins (The Taylor & Francis Series in Pharmaceutical Sciences, Lars Hovgaard, Sven Frokjaer, and Marco van de Weert eds., CRC Press; 1 ^st edition, 1999); herein incorporated by reference.

In general, these methods employ the sequential addition of one or more amino acids to a growing peptide chain. Normal ly, either the amino or carboxyl group of the first amino acid is protected by a suitable protecting group. The protected or derivatized amino acid can then be either attached to an inert solid support or utilized in solution by adding the next amino acid in the sequence having the complementary (amino or carboxyl) group suitably protected, under conditions that allow for the formation of an amide linkage. The protecting group is then removed from the newly added amino acid residue and the next amino acid (suitably protected) is then added, and so forth. After the desired amino acids have been linked in the proper sequence, any remaining protecting groups (and any solid support, if solid phase synthesis techniques are used) are removed sequentially or concurrently, to render the final peptide or polypeptide. By simple modification of this general procedure, it is possible to add more than one amino acid at a time to a growing chain, for example, by coupling (under conditions which do not racemize chiral centers) a protected tripeptide with a properly protected dipeptide to form, after deprotection, a pentapeptide. See, e.g., J. M. Stewart and J. D. Young, Solid Phase Peptide Synthesis (Pierce Chemical Co., Rockford, IL 1984) and G. Bar any and R. B. Merrifield, The Peptides: Analysis Synthesis Biology editors E. Gross and J.

Meienhofer, Vol. 2, (Academic Press, New York, 1980), pp. 3-254, for solid phase peptide synthesis techniques; and M. Bodansky, Principles of Peptide Synthesis.

(Springer-V erlag, Berlin 1984) and E. Gross and J. Meienhofer, Eds., The Peptides: Analysis Synthesis Biology. Vol. 1, for classical solution synthesis. These methods are typically used for relatively small polypeptides, i.e , up to about 50-100 amino acids in length, but are also applicable to larger polypeptides.

Typical protecting groups include t-butyl oxy carb onyl (Boc), 9- fluorenylmethoxycarbonyl (Fmoc) benzyloxycarbonyl (Cbz); p -tol uenesul fonyl (Tx); 2,4-dinitrophenyl; benzyl (Bzl); biphenylisopropyloxycarboxy-carbonyl, t- amyl oxy carb onyl , isobornyloxycarbonyl, o-bromobenzyloxycarbonyl, cyclohexyl, isopropyl, acetyl, o-nitrophenyl sulfonyl and the like.

Typical solid supports are cross-linked polymeric supports. These can include divinylbenzene cross-linked-styrene-based polymers, for example, divinylbenzene- hy droxymethyl styrene copolymers, divinylbenzene-chloromethylstyrene copolymers and divinylbenzene-benzhydrylaminopolystyrene copolymers. Ligand fusion proteins can also be chemically prepared by other methods such as by the method of simultaneous multiple peptide synthesis. See, e g., Houghten Proc.

Natl Acad. Sci. USA (1985) 82:5131-5135; U S. Patent No. 4,631 ,211.

C. Nucleic Acids Encoding Ligands

Nucleic acids encoding ligands of the urokinase receptor (e.g., fusion proteins comprising a urokinase GFD linked to a vitronectin SMB domain) can be used, for example, to treat cancer. Nucleic acids described herein can be inserted into an expression vector to create an expression cassette capable of producing the ligand fusion proteins in a suitable host cell. The ability of constructs to produce the ligand fusion proteins can be empirically determined (e.g., see Example 1 describing detection using immunofluorescent labeling).

Expression cassettes typically include control elements operably linked to the coding sequence, which allow for the expression of the gene in vivo in the subject species. For example, typical promoters for mammalian cell expression include the SV40 early promoter, a CMV promoter such as the CMV immediate early promoter, the mouse mammary tu or virus LTR promoter, the adenovirus major late promoter (Ad MLP), and the herpes simplex vims promoter, among others. Other nonviral promoters, such as a promoter derived from the murine metallothionein gene, will also find use for mammalian expression. Typically, transcription termination and polyadenylation sequences will also be present, located 3' to the translation stop codon. Preferably, a sequence for optimization of initiation of translation, located 5' to the coding sequence, is also present. Examples of transcription terminator/polyadenylati on signals include those derived from SV40, as described in Sambrook et al., supra , as well as a bovine growth hormone terminator sequence.

Enhancer elements ay also be used herein to increase expression levels of the mammalian constructs. Examples include the SV40 early gene enhancer, as described in Dijkema et al., EMPO J (1985) 4:761, the enhancer/ pro oter derived from the long terminal repeat (LTR) of the Rous Sarcoma Virus, as described in Gorman et al., Proc. Natl. Acad. Sci. USA (l982b) 79:6777 and elements derived from human CMV, as described in Boshart et al., Cell (1985) 41 :521, such as elements included in the CMV intron A sequence.

Once complete, the constructs encoding the ligand fusion proteins can be administered to a subject using standard gene delivery protocols. Methods for gene delivery are known in the art. See, e.g, U.S. Pat. Nos. 5,399,346, 5,580,859, 5,589,466. Genes can be delivered either directly to a vertebrate subject or, alternatively, delivered ex vivo, to cells derived from the subject and the cells reimplanted in the subject.

A number of viral based systems have been developed for gene transfer into mammalian cells. These include adenoviruses, retroviruses (g-retroviruses and lentiviruses), poxviruses, adeno-associated viruses, baculoviruses, and herpes simplex viruses (see e.g., Wamock et al. (2011) Methods Mol. Biol. 737: 1-25; Walther et al. (2000) Drugs 60(2):249-271; and Lundstrom (2003) Trends Biotechnol. 21 (3): 117-122; herein incorporated by reference).

For example, retroviruses provide a convenient platform for gene delivery systems. Selected sequences can be inserted into a vector and packaged in retroviral particles using techniques known in the art. The recombinant virus can then be isolated and delivered to cells of the subject either in vivo or ex vivo. A number of retroviral systems have been described (U.S. Pat. No. 5,219,740; Miller and Rosman (1989) BioTechniques 7:980-990; Miller, A. D. (1990) Human Gene Therapy 1 :5-14; Scarpa et al. (1991) Virology 180:849-852; Burns et al. (1993) Proc. Natl. Acad. Sci. USA

90:8033-8037; Boris-Lawrie and Temin (1993) Cur. Opin. Genet. Develop. 3: 102-109; and Ferry et al. (2011) Curr. Pharm. Des. 17(24):2516-2527). Lentiviruses are a class of retroviruses that are particularly useful for delivering polynucleotides to mammalian cells because they are able to infect both dividing and nondividing cells (see e.g., Lois et al (2002) Science 295:868-872; Durand et al. (2011) Viruses 3(2): 132-159; herein incorporated by reference).

A number of adenovirus vectors have also been described. Unlike retroviruses which integrate into the host genome, adenoviruses persist extrachromosomally thus minimizing the risks associated with insertional mutagenesis (Haj -Ahmad and Graham, J. Virol. (1986) 57:267-274; Bett et al, J. Virol. (1993) 67:5911-5921; Mittereder et al. Human Gene Therapy (1994) 5:717-729; Seth et al„ J. Virol. (1994) 68:933-940; Barr et al., Gene Therapy (1994) 1 :51-58; Berkner, K. L. BioTechniques (1988) 6:616-629; and Rich et al., Human Gene Therapy (1993) 4:461-476). Additionally, various adeno- associated virus (AAV) vector systems have been developed for gene delivery. AAV vectors can be readily constructed using techniques well known in the art.. See, e.g., U S. Pat. Nos. 5, 173,414 and 5, 139,941; International Publication Nos. WO 92/01070

(published 23 January 1992) and WO 93/03769 (published 4 March 1993); Lebkowski et al., Molec. Cell. Biol. (1988) 8:3988-3996; Vincent et al., Vaccines 90 (1990) (Cold Spring Harbor Laboratory Press); Carter, B. J. Current Opinion in Biotechnology (1992) 3:533-539; Muzyczka, N. Current Topics in Microbiol and Immunol. (1992) 158:97-129; Kotin, R. M. Human Gene Therapy (1994) 5:793-801; Shelling and Smith, Gene Therapy (1994) 1 : 165-169; and Zhou et al., J. Exp. Med. (1994) 179: 1867-1875.

Another vector system useful for delivering the polynucleotides of the present invention is the enterically administered recombinant poxvirus vaccines described by Small, Jr., P. A., et al. (U.S. Pat. No. 5,676,950, issued Oct. 14, 1997, herein incorporated by reference).

Additional viral vectors which will find use for delivering the nucleic acid molecules encoding the ligand fusion proteins of interest include those derived from the pox family of viruses, including vaccinia virus and avian poxvirus. By way of example, vaccinia vims recombinants expressing the ligand fusion proteins can be constructed as follows. The DNA encoding the particular ligand fusion protein coding sequence is first inserted into an appropriate vector so that it is adjacent to a vaccinia promoter and flanking vaccinia DNA sequences, such as the sequence encoding thymidine kinase (TK). This vector is then used to transfect cells which are simultaneously infected with vaccinia. Homologous recombination serves to insert the vaccinia promoter plus the gene encoding the coding sequences of interest into the viral genome. The resulting TK- recombinant can be selected by culturing the cells in the presence of 5- bromodeoxyuri dine and picking viral plaques resistant thereto.

Alternatively, avipoxviruses, such as the fowlpox and canarypox viruses, can also be used to deliver the genes. Recombinant avipox viruses, expressing immunogens from mammalian pathogens, are known to confer protective immunity when admini stered to non-avian species. The use of an avipox vector is particularly desirable in human and other mammalian species since members of the avipox genus can only productively replicate in susceptible avian species and therefore are not infective in mammalian cells. Methods for producing recombinant avipoxviruses are known in the art and employ genetic recombination, as described above with respect to the production of vaccinia viruses. See, e.g., WO 91/12882; WO 89/03429; and WO 92/03545.

Molecular conjugate vectors, such as the adenovirus chimeric vectors described in Michael et al., J. Biol. Chem. (1993) 268:6866-6869 and Wagner et al., Proc. Natl. Acad. Sci. USA (1992) 89:6099-6103, can also be used for gene delivery.

Members of the Alphavirus genus, such as, but not limited to, vectors derived from the Sindbis virus (SIN), Semliki Forest virus (SFV), and Venezuelan Equine Encephalitis virus (VEE), will also find use as viral vectors for delivering the

polynucleotides of the present invention. For a description of Sindbis-virus derived vectors useful for the practice of the instant methods, see, Dubensky et al. (1996) J. Virol. 70:508-519; and International Publication Nos. WO 95/07995, WO 96/17072; as well as, Dubensky, Jr., T. W., et al., U.S. Pat. No. 5,843,723, issued Dec. 1, 1998, and Dubensky, Jr., T. W., U.S. Patent No. 5,789,245, issued Aug. 4, 1998, both herein incorporated by reference. Particularly preferred are chimeric alphavirus vectors comprised of sequences derived from Sindbis virus and Venezuelan equine encephalitis virus. See, e.g., Perri et al. (2003) J. Virol . 77: 10394-10403 and International Publication Nos. WO 02/099035, WO 02/080982, WO 01/81609, and WO 00/61772; herein incorporated by reference in their entireties.

A vaccinia-based infection/transfection system can be conveniently used to provide for inducible, transient expression of the coding sequences of interest (for example, a ligand fusion protein expression cassette) in a host cell. In this system, cells are first infected in vitro with a vaccinia virus recombinant that encodes the

bacteriophage T7 RNA polymerase. This polymerase displays exquisite specificity in that it only transcribes templates bearing T7 promoters. Following infection, cells are transfected with the polynucleotide of interest, dri ven by a T7 promoter. The polymerase expressed in the cytoplasm from the vaccinia virus recombinant transcribes the transfected DNA into RNA which is then translated into protein by the host translational machinery. The method provides for high level, transient, cytoplasmic production of large quantities of RNA and its translation products. See, e.g., Elroy- Stein and Moss, Proc. Natl. Acad. Sci. USA (1990) 87:6743-6747; Fuerst et al., Proc. Natl. Acad. Sci. USA (1986) 83:8122-8126.

As an alternative approach to infection with vaccinia or avipox virus

recombinants, or to the delivery of genes using other viral vectors, an amplification system can be used that will lead to high level expression following introduction into host cells. Specifically, a T7 RNA polymerase promoter preceding the coding region for T7 RNA polymerase can be engineered. Translation of RNA derived from this template will generate T7 RNA polymerase which in turn will transcribe more template.

Concomitantly, there will be a cDNA whose expression is under the control of the T7 promoter. Thus, some of the T7 RNA polymerase generated from translati on of the amplification template RNA will lead to transcription of the desired gene. Because some T7 RNA polymerase is required to initiate the amplification, T7 RNA polymerase can be introduced into cells along with the template(s) to prime the transcription reaction. The polymerase can be introduced as a protein or on a plasmid encoding the RNA

polymerase. For a further discussion of T7 systems and their use for transforming cells, see, e.g., International Publication No. WO 94/26911; Studier and Moffatt, J. Mol. Biol. (1 86) 189: 113-130; Deng and Wolff, Gene (1994) 143 :245-249; Gao et al., Biochem. Biophys. Res. Commun. (1994) 200: 1201-1206; Gao and Huang, Nuc. Acids Res. (1993) 21 :2867-2872; Chen et al., Nuc. Acids Res (1994) 22:2114-2120; and U.S. Pat. No. 5, 135,855.

The synthetic expression cassette of interest can also be delivered without a viral vector. For example, the synthetic expression cassette can be packaged as DNA or RNA in liposomes prior to delivery' to the subject or to cells derived therefrom. Lipid encapsulation is generally accomplished using liposomes which are able to stably bind or entrap and retain nucleic acid. The ratio of condensed DNA to lipid preparation can vary but will generally be around 1 : 1 (mg DNA:micromoles lipid), or more of lipid. For a review of the use of liposomes as carriers for delivery of nucleic acids, see, Hug and Sleight, Biochim Biophys Acta (1991.) 1097: 1-17; Straubinger et al., in Methods of Enzymology (1983), Vol. 101, pp. 512-527.

Liposomal preparations for use in the present invention include cationic

(positively charged), anionic (negatively charged) and neutral preparations, with cationic liposomes particularly preferred. Cationic liposomes have been shown to mediate intracellular deliver _} ' of plasmid DNA (Feigner et al., Proc. Natl. Acad. Sci. USA (1987) 84:7413-7416); mRNA (Malone et al., Proc. Natl. Acad. Sci. USA (1989) 86:6077- 6081); and purified transcription factors (Debs et al., J. Biol. Chem. (1990) 265: 10189- 10192), in functional form.

Cationic liposomes are readily available. For example, N-[l-(2,3- di ol eyl oxy)propyl ] -N,N,N -trimethyl ammonium chloride (DOTMA) liposomes are available under the trademark Lipofectin, from GIBCO BRL, Grand Island, N.Y. (See, also, Feigner et al., Proc. Natl. Acad. Sci. USA (1987) 84:7413-7416). Other commercially available lipids include (DDAB/DOPE) and DOTAP/DOPE (Boerhinger). Other cationic liposomes can be prepared from readily available materials using techniques well known in the art. See, e.g., Szoka et al., Proc. Natl. Acad. Sci. USA (1978) 75:4194-4198; PCT Publication No. WO 90/11092 for a description of the synthesis of DO TAP (l,2-bis(oleoyloxy)-3-(trimethylammonio)propane) liposomes.

Similarly, anionic and neutral liposomes are readily available, such as, from Avanti Polar Lipids (Birmingham, AL), or can be easily prepared using readily available materials. Such materials include phosphatidyl choline, cholesterol, phosphatidyl ethanolamine, dioleoylphosphatidyl choline (DOPC), di ol eoylphosphati dyl glycerol (DOPG), dioleoylphoshatidyl ethanolamine (DOPE), among others. These materials can also be mixed with the DOTMA and DOTAP starting materials in appropriate ratios. Methods for making liposomes using these materials are well known in the art.

The liposomes can comprise multilammelar vesicles (MLVs), small unilamellar vesicles (SUVs), or large unilamellar vesicles (LUVs). The various liposome-nucleic acid complexes are prepared using methods known in the art. See, e.g., Straubinger et al., in METHODS OF IMMUNOLOGY (1983), Vol. 101 , pp. 512-527; Szoka et al., Proc. Natl. Acad. Sci. USA (1978) 75:4194-4198; Papahadjopoulos et al., Biochim. Biophys. Acta (1975) 394:483; Wilson et al., Cell (1979) 17:77); Deamer and Bangham, Biochim. Biophys. Acta (1976) 443 :629; Ostro et al., Biochem. Biophys. Res. Commun. (1977) 76:836; Fraley et al., Proc. Natl. Acad. Sci. USA (1979) 76:3348); Enoch and

Strittmatter, Proc. Natl. Acad. Sci. USA (1979) 76: 145); Fraley et al., J. Biol. Chem (1980) 255: 10431; Szoka and Papahadjopoulos, Proc. Natl. Acad. Sci. USA (1978) 75: 145; and Schaefer-Ridder et ah, Science (1982) 215: 166.

The DNA and/or peptide(s) can also be delivered in cochleate lipid compositions similar to those described by Papahadjopoulos et al., Biochem. Biophys. Acta (1975) 394:483-491. See, also, U S. Pat. Nos. 4,663,161 and 4,871,488.

The expression cassette of interest may also be encapsulated, adsorbed to, or associated with, particulate carriers. Examples of particulate carriers include those derived from polymethyl methacrylate polymers, as well as microparticles derived from poly(lactides) and poly(lactide-co-glycolides), known as PEG. See, e.g., Jeffery et al., Pharm. Res. (1993) 10:362-368; McGee J. P., et al., J Microencapsul. 14(2): 197-210, 1997; O'Hagan D. T., et ai., Vaccine 11(2): 149-54, 1993.

Furthermore, other particulate systems and polymers can be used for the in vivo or ex vivo delivery of the nucleic acid of interest. For example, polymers such as polylysine, polyarginine, polyornithine, spermine, spermidine, as well as conjugates of these molecules, are useful for transferring a nucleic acid of interest. Similarly, DEAE dextran-mediated transfection, calcium phosphate precipitation or precipitation using other insoluble inorganic salts, such as strontium phosphate, aluminum silicates including bentonite and kaolin, chromic oxide, magnesium silicate, talc, and the like, will find use with the present methods. See, e.g., Felgner, P. L., Advanced Drug Delivery Reviews (1990) 5: 163-187, for a review of delivery systems useful for gene transfer. Peptoids (Zuckerman, R. N., et al., U.S. Pat. No. 5,831 ,005, issued Nov. 3, 1998, herein incorporated by reference) may also be used for delivery of a construct of the present invention.

Additionally, biolistic delivery systems employing particulate carriers such as gold and tungsten, are especially useful for delivering synthetic expression cassettes of the present invention. The particles are coated with the synthetic expression cassette(s) to be delivered and accelerated to high velocity, generally under a reduced atmosphere, using a gun powder discharge from a "gene gun." For a description of such techniques, and apparatuses useful therefore, see, e.g., U.S. Pat. Nos. 4,945,050; 5,036,006;

5, 100,792; 5,179,022; 5,371,015; and 5,478,744. Also, needle-less injection systems can be used (Davis, H. L., et al, Vaccine 12: 1503-1509, 1994; Bioject, Inc., Portland, Oreg.). Recombinant vectors canying a synthetic expression cassette of the present invention are formulated into compositions for delivery' to a vertebrate subject. These compositions may either be prophylactic (e.g., to prevent cancer progression or metastasis) or therapeutic (e.g., to treat cancer). The compositions will comprise a "therapeutically effective amount" of the nucleic acid of interest such that an amount of the ligand (e.g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain) can be produced in vivo so that the urokinase receptor is inhibited in the individual to which it is administered. Preferably, the ligand is produced in a sufficient amount in the individual to have anti-tumor activity. The exact amount necessary will vary depending on the subject being treated; the age and general condition of the subject to be treated; the degree of protection desired; the severity of the condition being treated; the particular ligand fusion protein produced and its mode of administration, among other factors. An appropriate effective amount can be readily determined by one of skill in the art. Thus, a "therapeutically effective amount" will fall in a relatively broad range that can be determined through routine trials.

The compositions will generally include one or more "pharmaceutically acceptable excipients or vehicles" such as water, saline, glycerol, poly ethyleneglycol , hyaluronic acid, ethanol, etc. Additionally, auxiliary' substances, such as wetting or emulsifying agents, pH buffering substances, surfactants and the like, may be present in such vehicles. Certain facilitators of nucleic acid uptake and/or expression can also be included in the compositions or coadministered.

Once formulated, the compositions of the invention can be administered directly to the subject (e.g., as described above) or, alternatively, delivered ex vivo, to cells derived from the subject, using methods such as those described above. For example, methods for the ex vivo delivery and reimplantation of transformed cells into a subject are known in the art and can include, e.g., dextran-mediated transfection, calcium phosphate precipitation, polybrene mediated transfection, lipofectamine and LT-l mediated transfection, protoplast fusion, electroporation, encapsulation of the polynucleotide(s) in liposom es, and direct microinjection of the DNA into nuclei.

Direct delivery of synthetic expression cassette compositions in vivo will generally be accomplished with or without viral vectors, as described above, by injection using either a conventional syringe, needless devices such as Bioject or a gene gun, such as the Accell gene delivery system (PowderMed Ltd, Oxford, England).

D. Pharmaceutical Compositions

Ligands of the urokinase receptor (e.g., fusion proteins comprising a urokinase GFD linked to a vitronectin SMB domain or nucleic acids encoding them) can be formulated into pharmaceutical compositions optionally comprising one or more pharmaceutically acceptable excipients. Exemplary excipients include, without limitation, carbohydrates, inorganic salts, antimicrobial agents, antioxidants, surfactants, buffers, acids, bases, and combinations thereof. Excipients suitable for injectable compositions include water, alcohols, polyols, glycerine, vegetable oils, phospholipids, and surfactants. A carbohydrate such as a sugar, a derivatized sugar such as an alditol, aldonic acid, an esterified sugar, and/or a sugar polymer may be present as an excipient. Specifi c carbohydrate excipients include, for example: monosaccharides, such as fructose, maltose, galactose, glucose, D-mannose, sorbose, and the like; disaccharides, such as lactose, sucrose, trehalose, cell obi ose, and the like; polysaccharides, such as raffmose, melezitose, maltodextrins, dextrans, starches, and the like; and alditols, such as mannitol, xylitol, maltitol, lactitol, xylitol, sorbitol (glucitol), pyranosyl sorbitol, myoinositol, and the like. The excipient can also include an inorganic salt or buffer such as citric acid, sodium chloride, potassium chloride, sodium sulfate, potassium nitrate, sodium phosphate monobasic, sodium phosphate dibasic, and combinations thereof.

A composition of the invention can also include an antimicrobial agent for preventing or deterring microbial growth. Nonlimiting examples of antimicrobial agents suitable for the present invention include benzalkonium chloride, benzethonium chloride, benzyl alcohol, cetylpyridinium chloride, chlorobutanol, phenol, phenyl ethyl alcohol, phenylmer curie nitrate, thimersol, and combinations thereof.

An antioxidant can be present in the composition as well. Antioxidants are used to prevent oxidation, thereby preventing the deterioration of the ligand fusi on proteins or nucleic acids or other components of the preparation. Suitable antioxidants for use in the present invention include, for example, ascorbyl palmitate, butyl ated hydroxyanisole, butylated hydroxytoluene, hypophosphorous acid, monothioglycerol, propyl gallate, sodium bisulfite, sodium formaldehyde sulfoxylate, sodium metabisulfite, and combinations thereof.

A surfactant can be present as an excipient. Exemplary surfactants include: polysorbates, such as "Tween 20" and "Tween 80," and pluronics such as F68 and F88 (BASF, Mount Olive, New Jersey); sorbitan esters; lipids, such as phospholipids such as lecithin and other phosphatidylcholines, phosphatidylethanol amines (although preferably not in liposomal form), fatty acids and fatty esters; steroids, such as cholesterol; chelating agents, such as EDTA; and zinc and other such suitable cations.

Acids or bases can be present as an excipient in the composition. Nonlimiting examples of acids that can be used include those acids selected from the group consisting of hydrochloric acid, acetic acid, phosphoric acid, citric acid, malic acid, lactic acid, formic acid, trichloroacetic acid, nitric acid, perchloric acid, phosphoric acid, sulfuric acid, fumaric acid, and combinations thereof. Examples of suitable bases include, without limitation, bases selected from the group consisting of sodium hydroxide, sodium acetate, ammonium hydroxide, potassium hydroxide, ammonium acetate, potassium acetate, sodium phosphate, potassium phosphate, sodium citrate, sodium formate, sodium sulfate, potassium sulfate, potassium fumerate, and combinations thereof.

The amount of the ligand (e.g., when contained in a drug delivery system) in the composition will vary depending on a number of factors, but will optimally be a therapeutically effective dose when the composition is in a unit dosage form or container (e.g., a vial). A therapeutically effective dose can be determined experimentally by repeated administration of increasing amounts of the composition in order to determine which amount produces a clinically desired endpoint.

The amount of any individual excipient in the composition will vary depending on the nature and function of the excipient and particular needs of the composition.

Typically, the optimal amount of any indi vidual excipient is determined through routine experimentation, i.e., by preparing compositions containing varying amounts of the excipient (ranging from low to high), examining the stability and other parameters, and then determining the range at which optimal performance is attained with no significant adverse effects. Generally, however, the excipient(s) will be present in the composition in an amount of about 1 % to about 99% by weight, preferably from about 5% to about 98% by weight, more preferably from about 15 to about 95% by weight of the excipient, with concentrations less than 30% by weight most preferred. These foregoing pharmaceutical excipients along with other excipients are described in "Remington: The Science & Practice of Pharmacy", 19th ed., Williams & Williams, (1995), the

"Physician’s Desk Reference", 52nd ed., Medical Economics, Montvale, NJ (1998), and Kibbe, A.H., Handbook of Pharmaceutical Excipients, 3rd Edition, American

Pharmaceutical Association, Washington, D C., 2000.

The compositions encompass all types of formulations and in particular those that are suited for injection, e.g., powders or lyophilates that can be reconstituted with a solvent prior to use, as well as ready for injection solutions or suspensions, dry insoluble compositions for combination with a vehicle prior to use, and emul sions and liquid concentrates for dilution prior to administration. Examples of suitable diluents for reconstituting solid compositions prior to injection include bacteriostatic water for injection, dextrose 5% in water, phosphate buffered saline, Ringer's solution, saline, sterile water, deionized water, and combinations thereof. With respect to liquid pharmaceutical compositions, solutions and suspensions are envisioned. Additional preferred compositions include those for oral, ocular, or localized delivery.

The pharmaceutical preparations herein can also be housed in a syringe, an implantation device, or the like, depending upon the intended mode of delivery and use. Preferably, the compositions comprising one or more ligands of the urokinase receptor (e g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain or nucleic acids encoding them) described herein are in unit dosage form, meaning an amount of a conjugate or composition of the invention appropriate for a single dose, in a premeasured or pre-packaged form.

The compositions herein may optionally include one or more additional agents, such as drugs for treating cancer or other medications used to treat a subject for a condition or disease. Compounded preparations may include a ligand of a urokinase receptor (e.g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain) and one or more drugs for treating cancer, such as, but not limited to, chemotherapy, immunotherapy, biologic or targeted therapy agents. Alternatively, such agents can be contained in a separate composition from the composition comprising a ligand of a urokinase receptor and co-administered concurrently, before, or after the composition comprising the ligand of the urokinase receptor.

E. Bioconjugation of Ligands of the Urokinase Receptor

Bioconjugates may comprise a ligand of the urokinase receptor (e.g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain) conjugated to one or more diagnostic or therapeutic agents, or a combination thereof. A ligand of the urokinase receptor may be attached to diagnostic and/or therapeutic agents in a variety of manners. For example, an agent may be attached at the N-terminus, C -terminus, at both the N-terminus and C -terminus, and/or internally, for example, at a residue in the peptide linker between the GFD and SMB domains. Diagnostic and/or therapeutic agents may be connected directly to the ligand of the urokinase receptor or indirectly through an intervening linker or chelating agent (e.g., for metal labeling such as with a radionuclide or paramagnetic metal ion).

Conjugation of ligands can be performed using methods well-known in the art. For a discussion of bioconjugation techniques, see, e.g., Chemistry of Bioconjugates: Synthesis, Characterization, and Biomedical Applications (R. Narain ed., Wiley, 2014), G.T. Flermanson Bioconjugate Techniques (Academic Press, 3 ^rd edition, 2013), and Bioconjugalion Protocols: Strategies and Methods (Methods in Molecular Biology, S.S. Mark ed., Humana Press, 2 ^nd edition, 201 1), van Vught et al. (2014) Comput Struct Biotechnol J. 9 :e201402001; Gao et al. (2016) Curr Cancer Drug Targets. May 12 [Epub ahead of print]; Massa et al. (2016) Expert Opin Drug Deliv 13: 1-15; Yeh et al. (2015) PLoS One 10(7):e0129681; Freise et al. (2015) Mol Immunol. 67(2 Pt A): l42-152; herein incorporated by reference in their entireties.

For example, imaging and/or therapeutic agents can be conjugated to the side chain e-amine of lysine residues or the free thiol of cysteine residues. In particular, reactions of cysteine thiols with maleimides are commonly used for bioconjugation of proteins. Maleimide-functionalized imaging probes and compounds to facilitate bioconjugation for various imaging modalities are commercially available from a number of companies (e.g., ThermoFisher Scientific (Waltham, MA), GE Healthcare Life Sciences (Pittsburgh, PA), SigmaAldrich (St. Louis, MO), and CF1EMATECH (Dijon, France)), including, for example, maleimide lipid derivatives and maleimide albumin derivatives, which can be incorporated into a microbubble shell for ultrasound imaging, maleimide fluorescent dye derivatives for fluorescence imaging, maleimide chelating agent derivatives for binding metals such as radionuclides and paramagnetic cations, and maleimide gold nanoparticle derivatives, which can be used in a variety of ways including as detection agents for electron microscopy and surface enhanced Raman spectroscopy, enhancement agents for radiotherapy, phototherm al agents for surface plasmon resonance, and delivery agents for attached drugs or other therapeutic agents.

In certain embodi m ents, the l igand of the urokin ase receptor is engineered to include an N-terminal or C-terminal cysteine residue providing a free thiol group to facilitate conjugation to a reagent comprising a functional group that is reactive with thiols. In other embodiments, a cysteine is incorporated into the linker peptide to allow attachment of reagents to the linker region between the VH and VL domains.

Alternatively, additional cysteine residues may be introduced into the single-chain antibody, for example, by site-directed mutagenesis to allow attachment at other sites. A site of attachment away from the urokinase receptor binding sites of the GFD and SMB domains should be chosen to avoid interfering with targeting of the bioconjugate to the urokinase receptor. In one embodiment, a diagnostic or therapeutic agent is conjugated to a ligand of the urokinase receptor comprising a sequence selected from the group consisting of SEQ ID NOS: 12, 13, and 17-20.

An alternative bioconjugation method uses click chemistry. Click chemistry reactions include the Huisgen 1,3 -dipolar cycloaddition copper catalyzed reaction (Tornoe et al, 2002, 1 Organic Chem 67:3057-64), cycloaddition reactions such as Diels- Alder reactions, nucleophilic substitution reactions (especially to small strained rings like epoxy and aziridine compounds), reactions involving form ation of urea compounds, and reactions involving carbon-carbon double bonds, such as alkynes in thiol -yne

reactions. See, e.g., Kolb et al., 2004, Angew Chem Int Ed 40:3004-31; Evans, 2007, Aust J Chem 60:384-95; Millward et al. (2013) Integr Biol (Camb) 5(1 ): 87-95), Lai 1 ana et al. (2012) Pharm Res 29(1): 1-34, Gregoritza et al. (2015) Eur J Pharm Biopharm. 97(Pt B):438-453, Musumeci et al. (2015) Curr Med Chem. 22(17):2022-2050, McKay et al. (2014) Chem Biol21(9): 1075-1101, Ulrich et al. (2014) Chemistry 20(1):34-41, Pasini (2013) Molecules 18(8):9512-9530, and Wangler et al. (2010) Curr Med Chem.

17(11): 1092-1116; herein incorporated by reference in their entireties.

F. Urokinase Receptor-Targeted Imaging Agents

Ligands of the urokinase receptor can be conjugated to diagnostic agents (e.g., probes or detection agents) suitable for various imaging modalities, including, but not limited to, ultrasound imaging (UI), positron emission tomography (PET), single photon emission computed tomography (SPECT), magnetic resonance imaging (MRI), computed tomography (CT), optical imaging (OI), photoacoustic imaging (PI), or fluorescence imaging. Conjugation of a diagnostic agent comprising a detectable moiety or label to a ligand of the urokinase receptor localizes the diagnostic agent to tumors or cancerous cells expressing the urokinase receptor. Useful diagnostic agents that can be used in the practice of the invention include, but are not limited to, contrast agents, photoactive agents, radioisotopes, nonradioactive isotopes, dyes, fluorescent compounds or proteins, chemiluminescent compounds, bioluminescent proteins, enzymes, and enhancing agents (e.g., paramagnetic ions). In certain embodiments, the ligand of the urokinase receptor is conjugated to a contrast agent. Exemplary contrast agents include ultrasound contrast agents (e.g.

SonoVue microbubbles comprising sulphur hexafluoride, Optison microbubbles comprising an albumin shell and octafluoropropane gas core, Levovist microbubbles comprising a lipi d/galactose shell and an air core, Perflexane lipid microspheres comprising perfluorocarbon microbubbles, and Perflutren lipid microspheres comprising octafl uoropropane encapsulated in an outer lipid shell), magnetic resonance imaging (MRI) contrast agents (e.g., gadodi mide, gadobenic acid, gadopentetic acid, gadoteridol, gadofosveset, gadoversetamide, gadoxetic acid), and radiocontrast agents, such as for computed tomography (CT), radiography, or fluoroscopy (e.g., diatrizoic acid, metrizoic acid, iodamide, iotalamic acid, ioxitalamic acid, ioglicic acid, acetrizoic acid, iocarmic acid, methiodal, diodone, metrizamide, iohexol, ioxaglic acid, iopamidol, iopromide, iotrolan, ioversol, iopentol, iodixanol, iomeprol, iobitridol, ioxilan, iodoxamic acid, iotroxic acid, ioglycamic acid, adipiodone, iobenzamic acid, iopanoic acid, iocetamic acid, sodium iopodate, tyropanoic acid, and calcium iopodate).

In other embodiments, the diagnostic agent is a radioactive metal, paramagnetic ion, or other diagnostic cation. In this case, the ligand of the urokinase receptor can be conjugated to a chelating group for binding cations. Exemplary chelating agents include ethylenedi aminetetraaceti c acid (EDTA), di ethyl enetriaminepentaacetic acid (DTP A), l,4,7, 10-tetraazacyclododecane-l,4,7, 10-tetraacetic acid (DOT A), 1,4,7- triazacyclononane-N,N’,N”-triacetic acid (NOTA), NET A, p-bromoacetami do-benzyl - tetraethylaminetetraacetic acid (TETA), porphyrins, polyamines, crown ethers, bis- thiosemicarbazones, polyoximes, and like groups known to be useful for this purpose. Chelates are coupled to the single-chain antibodies using standard chemistries, which then can be used to bind diagnostic isotopes such as ¹²⁵ I, ¹³¹ I, ¹²³ I, ¹²⁴ 1, ⁶² Cu, ^b4 Cu, ¹⁸ F, ^m In, ⁶⁷ Ga, ⁶⁸ Ga, ^99m Tc, ²²³ Ra, ⁿ C, ¹³ N, ^!5 0, and ⁷⁶ Br for radioimaging. The same chelates, when complexed with non-radioactive metals, such as manganese, iron and gadolinium are useful for MRI, when used along with the ligands of the urokinase receptor. Diagnostic agents comprising ¹⁸ F or ⁿ C can be used in PET imaging. For example, a ligand of the urokinase receptor can be isotopically labeled with ^1S F or ^U C or conjugated to ¹⁸ F or ^u C-labeled compounds for use in PET imaging.

In certain embodiments, the ligand of the urokinase receptor is conjugated to a fluorescent label. Exemplar _} fluorescent labels include fluorescein derivatives, rhodamine derivatives, coumarin derivatives, cyanine derivatives, acridine derivatives, squaraine derivatives, naphthalene derivatives, oxadiazol derivatives, anthracene derivatives, pyrene derivatives, oxazine derivatives, arylmethine derivatives, and tetrapyrrole derivatives. Alternatively, the fluorescent label may comprise a fluorescent protein, such as, but not limited to, green fluorescent protein (GFP), enhanced green fluorescent protein (EGFP), yellow fluorescent protein (YFP), enhanced yellow fluorescent protein (EYFP), blue fluorescent protein (BFP), red fluorescent protein (RFP), TagRFP, Dronpa, Padron, mApple, mCherry, rsCherry, and rsCherryRev.

In other embodiments, the ligand of the urokinase receptor is conjugated to a bioluminescent label such as a bioluminescent protein. Exemplary bioluminescent proteins, include, but are not limited to, firefly luciferase, bacterial luciferase, Renilla luciferase, and aequorin.

Fluorescence or bioluminescence images may be recorded by any suitable method. For example, a CCD image sensor (e.g., intensified CCD (ICCD) or electron multiplying CCD (EMCCD)), a CMOS image sensor, or a digital camera may be used to capture images. The image may be a still photo or a video in any format (e.g., bitmap, Graphics Interchange Format, JPEG file interchange format, TIFF, or mpeg).

Alternatively, images may be captured by an analog camera and converted into an electronic form. In addition, luminescence can be detected by a luminometer, and fluorescence can be detected by a fluorimeter, a fluorescence microscope, a fluorescence microplate reader, a fluorometric imaging plate reader, fluorescence-activated cell sorting, a fiber-optic fluorescence imaging system, or a medical fluorescence imaging device (e.g., a handheld fluorescence microscope, a laparoscope, an endoscope, or a microendoscope). Various medical imaging systems have been developed for open surgery as well as for laparoscopic, thoracoscopic, and robot-assisted surgery and can be used in the practice of the invention. Conventional laparoscopes and endoscopes can be equipped with a photodetector (e.g., camera or CCD detector) to provide guidance during medical procedures. Fiber-optic imaging systems can also be used, which include portable handheld microscopes, flexible endoscopes, and microendoscopes. An illumination source can be added to such devices to allow fluorescence imaging. For fluorescence imaging, the excitation light source and photodetector can be integrated into a single medical imaging device or the excitation light source and/or photodetector may reside apart, in which case, imaging is performed with remote delivery of excitation light. Miniaturized imaging systems can be used that allow imaging inside small cavities and constricted spaces. In addition, miniaturized imaging devices (e.g., microendoscopes) may be implanted within a subject for long-term imaging studies. An imaging system that can simultaneously detect fluorescence or bioluminescence at multiple wavelengths can be used for detection of multiple fluorescent and/or bioluminescent agents that emit light at different wavelengths. In addition, a camera may be used to take both photographic images of a subject and to detect fluorescence and bioluminescence, so that photographic images and fluorescent or bioluminescent images can be superimposed to allow regions of fluorescence or bioluminescence to be mapped to the subject’s anatomy for identification of the source of light emissions. For a review of medical imaging devices and methods of using them in image-guided surgery and other medical procedures, see, e.g., Gray et al. (2012) Biomed. Opt. Express. 3(8): 1880-1890; Flusberg et al. (2005) Nat. Methods 2(12):941-950; Choyke et al. (2012) IEEE J. Sel. Top.

Quantum. Electron. 18(3): 1140-1146; Gray et al. (2012 ) Proc. SPIE February 3: 8207; Vahrmeijer et al. (2013) Nat. Rev. Clin. Oncol. 10:507-518; Braks et al. (2013) Methods Mol. Biol. 923 :353-368; Yong et al. (2011) Diabetes 60(5): 1383-1392; Wilson et al. (2008) J. Vis. Exp. May 2(14) pii: 740; Engel sm an et al. (2009) J. Biomed. Mater. Res B Appl Biomater. 88(1): 123-129; Franke-Fayard et al. (2006) Nat. Protoc. l(l):476-485; Rehemtulla et al. (2000) Neoplasia. 2(6):491-495; and Close et al. (2011) Sensors 11 : 180-206; herein incorporated by reference in their entireties. Administration of Urokinase receptor-Targeted Imaging Agents

Preferably, a detectably effective amount of a urokinase receptor-targeted imaging agent (e.g., a ligand of the urokinase receptor conjugated to a diagnostic agent) is administered to a subject; that is, an amount that is sufficient to yield an acceptable image using the imaging equipment that is available for clinical use. A detectably effective am ount of the urokinase receptor-targeted imaging agent may be admini stered in more than one injection if needed. The detectably effective amount of the urokinase receptor- targeted imaging agent needed for an individual may vary according to factors such as the degree of binding of the imaging agent to cancerous tissue, the age, sex, and weight of the individual, and the particular medical imaging method used. Optimization of such factors is within the level of skill in the art.

Imaging with urokinase receptor-targeted imaging agents can be used in assessing efficacy of therapeutic drugs in treating cancer. For example, images can be acquired after treatment with an anti -cancer therapy to determine if the individual is responding to treatment. In a subject with cancer, imaging with a urokinase receptor-targeted imaging agent can be used to evaluate whether a tumor is shrinking or growing. Further, the extent of cancerous disease (stage of cancer progression) can be determined to aid in determining prognosis and evaluating optimal strategies for treatment (e.g., surgery, radiation, or chemotherapy).

Additionally, urokinase receptor-targeted imaging agents can be used in image- guided surgery. Cells or tissue of interest can be contacted with a urokinase receptor- targeted imaging agent, such that the urokinase receptor-targeted imaging agent binds to urokinase receptors present on the surface of cells or tissue (e.g., urokinase receptors overexpressed on tumors or cancerous cells). Imaging of tissues labeled with the urokinase receptor-targeted imaging agent in this way can be used, for example, for detection of pathology, tumor margin delineation, evaluation of the completeness of resection, and evaluation of the efficacy of treatment. G. Tumor-Targeted Therapeutic Agents

The ligands of the invention localize specifically to uPAR-expressing tumors where they are internalized by cancerous cells (see Example 2). Thus, ligands can also be used to target therapeutic agents to the location of uPAR-expressing tumors or cancerous cells to directly treat cancer in a subject. A ligand of the invention (e.g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain) can be conjugated to one or more therapeutic agents, such as, but not limited to, drugs, toxins, radioisotopes, immunomodulators, angiogenesis inhibitors, therapeutic enzymes, and cytotoxic or pro- apoptotic agents for treatment of cancer.

For example, a ligand can be conjugated to one or more chemotherapeutic agents such as, but not limited to, abitrexate, adriamycin, adrucil, amsacrine, asparaginase, anthracyclines, azacitidine, azathioprine, bicnu, blenoxane, busulfan, bleomycin, camptosar, camptothecins, carboplatin, carmustine, cerubidine, chlorambucil, cisplatin, cladribine, cosmegen, cytarabine, cytosar, cyclophosphamide, cytoxan, dactinomycin, docetaxel, doxorubicin, daunorubicin, ellence, el spar, epirubicin, etoposide, fludarabine, fluorouracil, fludara, gemcitabine, gemzar, hycamtin, hydroxyurea, hydrea, idamycin, idarubicin, ifosfamide, ifex, irinotecan, lanvis, leukeran, leustatin, matulane,

mechlorethamine, mercaptopurine, methotrexate, mitomycin, mitoxantrone, mithramycin, mutamycin, myleran, mylosar, navelbine, nipent, novantrone, oncovin, oxaliplatin, paclitaxel, paraplatin, pentostatin, platinol, plicamycin, procarbazine, purinethol, ralitrexed, taxotere, taxol, teniposide, thioguanine, tomudex, topotecan, valrubicin, velban, vepesid, vinblastine, vindesine, vincristine, vinorelbine, VP- 16, and vumon.

Alternatively or additionally, a ligand can be conjugated to, one or more tyrosine- kinase inhibitors, such as Imatinib mesylate (Gleevec, also known as STI-571), Gefitinib (Iressa, also known as ZD 1839), Erlotinib (marketed as Tarceva), Sorafenib (Nexavar), Sunitinib (Sutent), Dasatinib (Sprycel), Lapatinib (Tykerb), Nilotinib (Tasigna), and Bortezomib (Velcade); Janus kinase inhibitors, such as tofacitinib; ALK inhibitors, such as crizotinib; Bel -2 inhibitors, such as obatoclax and gossypol; PARP inhibitors, such as Iniparib and Olaparib; PI3K inhibitors, such as perifosine; VEGF Receptor 2 inhibitors, such as Apatinib; AN-152 (AEZS-108) doxorubicin linked to [D -Ly s(6)] - LHRF1 ; Braf inhibitors, such as vemurafenib, dabrafenib, and LGX818; MEK inhibitors, such as trametinib; CDK inhibitors, such as PD-0332991 and LEE011; Hsp90 inhibitors, such as salinomycin; and/or small molecule drug conjugates, such as Vintafolide;

serine/threonine kinase inhibitors, such as Temsirolimus (Torisel), Everolimus (Afmitor), Vemurafenib (Zelboraf), Trametinib (Mekinist), and Dabrafenib (Tafinlar).

In another example, a ligand is conjugated to a hormonal blocking therapeutic agent for treatment of a cancer depending on estrogen for growth (e.g., cancer expressing estrogen receptors (ER+ cancer)). For example, the anti-B7-H3 antibody can be conjugated to a drug that blocks ER receptors (e.g. tamoxifen) or a drug that blocks the production of estrogen, such as an aromatase inhibitor (e.g. anastrozole, or letrozole).

In another example, the ligand is conjugated to a toxin. The toxin can be of animal, plant or microbial origin. Exemplary toxins include Pseudomonas exotoxin, ricin, abrin, Staphylococcal enterotoxin-A, pokeweed antiviral protein, gelonin, diphtheria toxin, and Pseudomonas endotoxin.

In a further example, the ligand is conjugated to an immunomodulator, such as a cytokine, a lymphokine, a monoline, a stem cell growth factor, a lymphotoxin (LT), a hematopoietic factor, a colony stimulating factor (CSF), an interferon (IFN), parathyroid hormone, thyroxine, insulin, proinsulin, relaxin, prorelaxin, follicle stimulating hormone (FSH), thyroid stimulating hormone (TSH), luteinizing hormone (LH), hepatic growth factor, prostaglandin, fibroblast growth factor, prolactin, placental lactogen, OB protein, a transforming growth factor (TGF), such as TGF-oc or TGF-b, insulin-like growth factor (IGF), erythropoietin, thrombopoietin, a tumor necrosis factor (TNF) such as TNF-ct or TNF-b, a mullerian-inhibiting substance, mouse gonadotropin-associated peptide, inhibin, activin, vascular endothelial growth factor, integrin, granulocyte-colony stimulating factor (G-CSF), granulocyte macrophage-colony stimulating factor (GM- CSF), an interferon such as interferon-a, interferon-b, or interferon-v, SI factor, an interleukin (IL) such as IL-l, IL-lcc, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, IL-10, IL-l l, IL-l 2, IL-13, IL-l 4, IL-15, IL-l 6, IL-l 7, IL-l 8 IL-21 or IL-25, LIF, kit-ligand, FLT-3, angiostatin, thrombospondin, endostatin, and LT. In another embodiment, the ligand is conjugated to a radioactive isotope.

Particularly useful therapeutic radionuclides include, but are not limited to ^m In, ¹⁷ Lu,

In certain embodiments, the therapeutic radionuclide has a decay energy in the range of 20 to 6,000 keV (e.g., 60 to 200 keV for an Auger emitter, 100-2,500 keV for a beta emitter, and 4,000-6,000 keV for an alpha emitter). In one embodiment, the radionuclide is an Auger-emitter (e.g., Co-58, Ga-67, Br-80m, Tc-99m, Rh-103m, Pt- 109, In-111, Sb- 1 19, 1-125, Ho-161 , Os-189m and Ir-192). In another embodiment, the radionuclide is an alpha-emitter (e.g., Dy-152, At-21 1, Bi-212, Ra-223, Rn-219, Po-215,

Bi-21 1, Ac-225, Fr-221, At-217, Bi-213 and Fm-255).

Additional therapeutic radioisotopes include ⁴¹ C, ^l3 N, ¹⁵ 0, ⁷ ¾r, ¹⁹⁸ Au, ²²⁴ Ac, ¹²⁶ I,

2 ²⁵ AC, ^/b Br, ¹⁶⁹ Yb, and the like.

Ligands may also be conjugated to a boron addend-loaded carrier for thermal neutron activation therapy. For example, boron addends such as carboranes, can be attached to B7-H3-targeting agents. Carboranes can be prepared with carboxyl functions on pendant side chains, as is well-known in the art. Attachment of carboranes to a carrier, such as aminodextran, can be achieved by activation of the carboxyl groups of the carboranes and condensation with amines on the carrier. The intermediate conjugate is then conjugated to the B7-H3-targeting agent. After administration of the B7-H3- targeting agent conjugate, a boron addend is activated by thermal neutron irradiation and converted to radioactive atoms which decay by alpha-emission to produce highly toxic, short-range effects.

H. Administration

The methods of the invention can be used for treating a subj ect for cancer. For example, ligands of the urokinase receptor (or bioconjugates thereof) can be used for treating a subject for a tumor, cancer, or metastasis that is progressing, worsening, stabilized or in remission as well as precancerous lesions.

At least one therapeutically effective cycle of treatment with a ligand of the urokinase receptor (e.g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain or a bioconjugate thereof) will be admini stered to a subject for treatment of cancer. By "therapeutically effective dose or amount" of a ligand of the urokinase receptor is intended an amount that when administered brings about a positive therapeutic response with respect to treatment of an individual for cancer. Of particular interest is an amount of a ligand of the urokinase receptor that provides an anti-tumor effect, as defined herein. By "positive therapeutic response" is intended the individual undergoing the treatment according to the invention exhibits an improvement in one or more symptoms of the cancer for which the individual is undergoing therapy.

Thus, for example, a "positive therapeutic response" would be an improvement in the disease in association with the therapy, and/or an improvement in one or more symptoms of the disease in association with the therapy. Therefore, for example, a positive therapeutic response would refer to one or more of the following improvements in the disease: (1) reduction in tumor size; (2) reduction in the number of cancer cells;

(3) inhibition (i.e., slowing to some extent, preferably halting) of tumor growth; (4) inhibition (i.e., slowing to some extent, preferably halting) of cancer cell infiltration into peripheral organs; (5) inhibition (i.e., slowing to some extent, preferably halting) of tumor metastasi s; and (6) some extent of relief from one or more symptoms associated with the cancer. Such therapeutic responses may be further characterized as to degree of improvement. Thus, for example, an improvement may be characterized as a complete response. By "complete response" is documentation of the disappearance of all symptoms and signs of all measurable or evaluable disease confirmed by physical examination, laboratory, ultrasound, nuclear, radiographic studies (i.e., CT (computer tomography), and/or MRI (magnetic resonance imaging)), and other non-invasive procedures repeated for all initial abnormalities or sites positive at the time of entry into the study. Alternatively, an improvement in the disease may be categorized as being a partial response. By "partial response" is intended a reduction of greater than 50% in the sum of the products of the perpendicular diameters of all measurable lesions when compared with pretreatment measurements.

In certain embodiments, multiple therapeutically effective doses of compositions comprising a ligand of the urokinase receptor (e g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain or a bioconjugate thereof) and/or one or more other therapeutic agents, such as other drugs for treating cancer, or other medications will be administered according to a daily dosing regimen, or intermittently. For example, a therapeutically effective dose can be administered, one day a week, two days a week, three days a week, four days a week, or five days a week, and so forth. By "intermittent" administration is intended the therapeutically effective dose can be administered, for example, every other day, every two days, every three days, and so forth. For example, in some embodiments, a ligand will be administered twice-weekly or thrice-weekly for an extended period of time, such as for 1, 2, 3, 4, 5, 6, 7, 8...10...15...24 weeks, and so forth. By "twice-weekly" or "two times per week" is intended that two therapeutically effective doses of the agent in question is administered to the subject within a 7-day period, beginning on day 1 of the first week of administration, with a minimum of 72 hours, between doses and a maximum of 96 hours between doses. By "thrice weekly" or "three times per week" is intended that three therapeutically effective doses are administered to the subject within a 7-day period, allowing for a minimum of 48 hours between doses and a maximum of 72 hours between doses. For purposes of the present invention, this type of dosing is referred to as "intermittent" therapy. In accordance with the methods of the present invention, a subject can receive intermittent therapy (i.e., twice-weekly or thrice-weekly administration of a therapeutically effective dose) for one or more weekly cycles until the desired therapeutic response is achieved. The agents can be administered by any acceptable route of administration as noted herein below.

The compositions of the present invention are typically, although not necessarily, administered orally, via injection (subcutaneously, intravenously, or intramuscularly), by infusion, or locally. Additional modes of administration are also contemplated, such as intra-arterial, intraperitoneal, pulmonary, nasal, topical, transderm al, intralesion, intrapleural, intraparenchymatous, rectal, transdermal, transmucosal, intrathecal, pericardial, intra-arterial, intraocular, and so forth. When administering the ligand of the urokinase receptor by injection, the administration ay be by continuous infusion or by single or multiple boluses.

The preparati ons according to the invention are also suitable for local treatment.

In a particular embodiment, a composition of the invention is used for localized delivery of a ligand of the urokinase receptor for the treatment of cancer. For exampl e, compositions may be administered directly into a tumor or cancerous cells.

Administration may be by perfusion through a regional catheter or direct intralesional injection.

The pharmaceutical preparation can be in the form of a liquid solution or suspension immediately prior to administration, but may also take another form such as a syrup, cream, ointment, tablet, capsule, powder, gel, matrix, suppository, or the like. The pharmaceutical compositions comprising a ligand of the urokinase receptor and other agents may be administered using the same or different routes of administration in accordance with any medically acceptable method known in the art.

In another embodiment, the pharmaceutical compositions comprising a ligand of the urokinase receptor and/or other agents are administered prophylactically, e.g., to prevent cancer progression or metastasis in tissue. Such prophylactic uses will be of particular value for subjects with a potentially precancerous or premalignant condition (e.g., precancerous lesions, dysplasia or benign neoplasia), or who have a genetic predisposition to developing cancer.

In another embodiment of the invention, the pharmaceutical compositions comprising a ligand of the urokinase receptor and/or other agents are in a sustained- release formulation, or a formulation that is administered using a sustained-release device. Such devices are well known in the art, and include, for example, transdermal patches, and miniature implantable pumps that can provide for drug delivery over time in a continuous, steady-state fashion at a variety of doses to achieve a sustained-release effect with a non-sustained-release pharmaceutical composition. The invention also provides a method for administering a conjugate comprising a ligand of the urokinase receptor (e.g., conjugated to a diagnostic or therapeutic agent) as provided herein to a patient suffering from cancer. The method comprises administering, via any of the herein described modes, a therapeutically effective amount of the conjugate or drug delivery system, preferably provided as part of a pharmaceutical composition.

The method of administering may be used to treat any cancer that is responsive to treatment with a ligand of the urokinase receptor. More specifically, the compositions herein are effective in treating cancer.

Those of ordinary skill in the art will appreciate which conditions a ligand of the urokinase receptor can effectively treat. The actual dose to be administered will vary depending upon the age, weight, and general condition of the subject as well as the severity of the condition being treated, the judgment of the health care professional, and conjugate being administered. Therapeutically effective amounts can be determined by those skilled in the art, and will be adjusted to the particular requirements of each particular case.

Generally, a therapeutically effective amount will range from about 0.50 mg to 5 grams of a ligand of the urokinase receptor daily, more preferably from about 5 mg to 2 grams daily, even more preferably from about 7 mg to 1.5 grams daily. Preferably, such doses are in the range of 10-600 mg four times a day (QID), 200-500 mg QID, 25 - 600 mg three times a day (TID), 25-50 mg TID, 50-100 mg TID, 50-200 mg TID, 300-600 mg TID, 200-400 mg TID, 200-600 mg TID, 100 to 700 mg twice daily (BID), 100-600 mg BID, 200-500 mg BID, or 200-300 mg BID. The amount of compound administered will depend on the potency of the specific ligand of the urokinase receptor and the magnitude or effect desired and the route of administration.

A purified ligand of the urokinase receptor (again, preferably provided as part of a pharmaceutical preparation) can be administered alone or in combination with one or more other therapeutic agents, such as chemotherapy, immunotherapy, biologic or targeted therapy agents, or other medications used to treat a particular condition or disease according to a variety of dosing schedules depending on the judgment of the clinician, needs of the patient, and so forth. The specific dosing schedule will be known by those of ordinary skill in the art or can be determined experimentally using routine methods Exemplary dosing schedules include, without limitation, administration five times a day, four times a day, three times a day, twice daily, once daily, three times weekly, twice weekly, once weekly, twice monthly, once monthly, and any combination thereof Preferred compositions are those requiring dosing no more than once a day.

A ligand of the urokinase receptor can be administered prior to, concurrent with, or subsequent to other agents. If provided at the same time as other agents, the ligand of the urokinase receptor can be provided in the same or in a different composition. Thus, the ligand of the urokinase receptor and other agents can be presented to the individual by way of concurrent therapy. By "concurrent therapy" is intended administration to a subject such that the therapeutic effect of the combination of the substances is caused in the subject undergoing therapy. For example, concurrent therapy may be achieved by administering a dose of a pharmaceutical composition comprising a ligand of the urokinase receptor and a dose of a pharmaceutical composition comprising at least one other agent, such as another drug for treating cancer, which in combination comprise a therapeutically effective dose, according to a particular dosing regimen. Similarly, the ligand of the urokinase receptor and one or more other therapeutic agents can be administered in at least one therapeutic dose. Administration of the separate

pharmaceutical compositions can be performed simultaneously or at different times (i.e., sequentially, in either order, on the same day, or on different days), as long as the therapeutic effect of the combination of these substances is caused in the subject undergoing therapy.

Where a subject undergoing therapy in accordance with the previously mentioned dosing regimens exhibits a partial response or a relapse following a prolonged period of remission, subsequent courses of concurrent therapy may be needed to achieve complete remission of the disease. Thus, subsequent to a period of time off from a first treatment period, a subject may receive one or more additional treatment periods with the ligand of the urokinase receptor. Such a period of time off between treatment periods is referred to herein as a time period of discontinuance. It is recognized that the length of the time peri od of discontinuance is dependent upon the degree of tumor response (i.e., complete versus partial) achieved with any prior treatment periods of concurrent therapy with these therapeutic agents

Additionally, treatment with a ligand of the urokinase receptor may be combined with any other medical treatment for cancer, such as, but not limited to, surgery, radiation therapy, chemotherapy, hormonal therapy, immunotherapy, or molecularly targeted or biologic therapy. Any combination of these other medical treatment methods with a ligand of the urokinase receptor may be used to effectively treat cancer in a subject.

For example, treatment with a ligand of the urokinase receptor may be combined with chemotherapy with one or more chemotherapeutic agents such as, but not limited to, abitrexate, adriamycin, adrucil, amsacrine, asparaginase, anthracyclines, azacitidine, azathioprine, bicnu, blenoxane, busulfan, bleomycin, camptosar, camptothecins, carboplatin, carmustine, cerubidine, chlorambucil, cisplatin, cladribine, cosmegen, cytarabine, cytosar, cyclophosphamide, cytoxan, dactinomycin, docetaxel, doxorubicin, daunorubicin, ellence, el spar, epirubicin, etoposide, fludarabine, fluorouracil, fludara, gemcitabine, gemzar, hycamtin, hydroxyurea, hydrea, idamycin, idarubicin, ifosfamide, ifex, irinotecan, lanvis, leukeran, leustatin, matulane, mechlorethamine, mercaptopurine, methotrexate, mitomycin, mitoxantrone, mithramycin, mutamycin, myleran, mylosar, navelbine, nipent, novantrone, oncovin, oxaliplatin, paclitaxel, paraplatin, pentostatin, platinol, plicamycin, procarbazine, purinethol, ralitrexed, taxotere, taxol, teniposide, thioguanine, tomudex, topotecan, valrubicin, velban, vepesid, vinblastine, vindesine, vincri stine, vinorelbine, VP- 16, and vumon.

In another example, treatment with a ligand of the urokinase receptor may be combined with targeted therapy with one or more small molecule inhibitors or monoclonal antibodies such as, but not limited to, tyrosine-kinase inhibitors, such as Imatinib mesylate (Gleevec, also known as STI-571), Gefitinib (Iressa, also known as ZD1839), Erlotinib (marketed as Tarceva), Sorafenib (Nexavar), Sunitinib (Sutent), Dasatinib (Sprycel), Lapatinib (Tykerb), Nilotinib (Tasigna), and Bortezomib (Velcade); Janus kinase inhibitors, such as tofacitinib; ALK inhibitors, such as crizotinib; Bcl-2 inhibitors, such as obatoclax and gossypol; PARP inhibitors, such as Iniparib and Olaparib; PI3K inhibitors, such as perifosine; VEGF receptor 2 inhibitors, such as Apatinib; AN-152 (AEZS-108) doxorubicin linked to [D-Lys(6)]-LHRH; Braf inhibitors, such as vemurafenib, dabrafenib, and LGX818; MEK inhibitors, such as trametinib; CDK inhibitors, such as PD-0332991 and LEE011 ; Hsp90 inhibitors, such as salinomycin; small molecule drag conjugates, such as Vintafolide; serine/threonine kinase inhibitors, such as Temsirolimus (Torisel), Everolimus (Afmitor), Vemurafenib (Zelboraf),

Trametinib (Mekinist), and Dabrafenib (Tafmlar); and monoclonal antibodies, such as Rituximab (marketed as MabThera or Rituxan), Trastuzumab (Herceptin), Alemtuzumab, Cetuximab (marketed as Erbitux), Panitumumab, Bevacizumab (marketed as Avastin), and Ipilimumab (Yervoy).

In a further example, treatment with a ligand of the urokinase receptor may be combined with immunotherapy, including, but not limited to, using any of the following: a cancer vaccine (e.g., E75 HER2-derived peptide vaccine, nelipepimut-S (NeuVax), Sipuleucel-T), antibody therapy (e.g., Trastuzumab, Ado-trastuzumab emtansine, Alemtuzumab, Ipilimumab, Ofatumumab, Nivolumab, Pembrolizumab, or Rituximab), cytokine therapy (e.g., interferons, including type I (IFNa and PTNGb), type II (LFNy) and type III (ITNl) and interleukins, including interleukin-2 (IL-2)), adjuvant

im unochemotherapy (e.g., polysaccharide-K), adoptive T-cell therapy, and immune checkpoint blockade therapy. I. Kits

The invention also provides kits comprising one or more containers holding compositions comprising at least one ligand of a urokinase receptor (e.g., a fusion protein comprising a urokinase GFD linked to a vitronectin SMB domain), or a bioconjugate thereof (e.g. conjugated to a therapeutic and/or diagnostic agent), or a nucleic acid encoding such a ligand and optionally one or more other drugs for treating cancer.

Compositions can be in liquid form or can be lyophilized, as can individual peptides, polypeptides, or nucleic acids. Suitable containers for the compositions include, for example, bottles, vials, syringes, and test tubes. Containers can be formed from a variety of materials, including glass or plastic. A container 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).

The kit can further comprise a second container comprising a pharmaceutically- acceptable buffer, such as phosphate-buffered saline, Ringer's solution, or dextrose solution. It can also contain other materials useful to the end-user, including other pharmaceutically acceptable formulating solutions such as buffers, diluents, filters, needles, and syringes or other delivery devices. The delivery device may be pre-filled with the compositions.

The kit can also comprise a package insert containing written instructions for methods of treating cancer. The package insert can be an unapproved draft package insert or can be a package insert approved by the Food and Drug Administration (FDA) or other regulatory body.

In certain embodiments, the kit comprises a ligand comprising a fusion protein comprising an amino acid sequence selected from the group consisting of SEQ ID NO: 12, SEQ ID NO: 13, and SEQ ID NOS: 17-20, or a sequence displaying at least about 80-100% sequence identity thereto, including any percent identity within this range, such as 81 , 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% sequence identity thereto, wherein the ligand is capable of binding to and/or inhibiting the urokinase receptor.

III. Experimental

Below are examples of specific embodiments for carrying out the present invention. The examples are offered for illustrative purposes only, and are not intended to limit the scope of the present invention in any way.

Efforts have been made to ensure accuracy with respect to numbers used (e.g., amounts, temperatures, etc.), but some experimental error and deviation should, of course, be allowed for. Example 1

Engineering High-Affinity Ligands of The Urokinase Receptor

INTRODUCTION

Over the last 30 years the data implicating uPAR in cancer metastasis has continued to grow, with over 5,000 research articles and reviews published on the topic to date. In addition to uncovering uPAR’s role in metastasis, a tremendous effort has been made to develop molecules that block uPAR activity as a means of combating the spread of cancer in patients. Although numerous molecules targeting uPAR have shown promising efficacy in animal models of cancer, surprisingly, none have been further developed and made available in the clinic for treating patients. One molecule that inhibits uPA activity did complete two separate phase II clinical trials in 2008 and 2012 and was well tolerated ^{1 2} , suggesting toxicity is not impeding progress. However, further developm ent of this molecule has not been published, likely due to the underwhelming efficacy of only extending median progression free survival from 49 to 100 days in patients with epithelial ovarian cancer ¹ , and having no positive effect on patients with recurrent epithelial ovarian cancer, fallopian tube cancer, or primary peritoneal carcinoma ² . Nevertheless, the significant, although low, efficacy demonstrated by this first-in-class molecule and the absence of toxic side effects confirms uPA and uPAR as viable clinical targets for cancer treatment. Our hypothesis is the lack of efficacy delivered by previous molecules is due to their inadequate inhibition of uPAR activity. The goal of this thesis is to apply lessons learned from these previous attempts toward engineering a next-generation uPAR antagonist capable of more completely inhibiting uPAR activity in aggressive and metastatic cancers.

Previously developed uPAR antagonists can be categorized by the specific uPAR functions they inhibit. For example, there are inhibitors of uPAR’s interaction with its soluble ligand urokinase, and with an assortment of integrins, FPRL1, vitronectin, and P AI- 1 , and there are direct inhibitors of urokinase catalytic activity. A close evaluation of these antagonists reveals key shared characteristics between them, 1) they each effectively reduce cancer establishment, growth, and/or metastasis in vivo, even though 2) they bind uPAR with similar or weaker affinity compared to the competing wild-type uPAR ligands, and 3) each antagonist alone only partially inhibits uPAR activity.

Targeting the uPAR receptor with an antagonist engineered to: a) have orders-of- magnitude higher binding affinity than uPAR’s native ligands, and thus more effectively outcompete native ligand binding in vivo, and b) simultaneously block multiple cancer- associated uPAR interactions is therefore an attractive strategy for developing a next- generation uPAR antagonist. Such an antagonist would overcome the limitations of previously developed molecules, and potentially have a more dramatic effect on cancer metastasis in vivo.

Towards this goal, we sought to engineer a bispecific protein that simultaneously binds to two distinct functional uPAR epitopes with an affinity superior to the native ligands. Possible strategies for engineering said protein include 1) developing and fusing together two antibodies that each bind a separate functional uPAR epitope, such that both simultaneously bind uPAR, 2) developing two antibodies that each bind a separate functional uPAR epitope, and then creating a "knob-in-hole" heterodimer ^{3 6} composed of one arm from each antibody that simultaneously bind uPAR, 3) engineering non-antibody scaffold proteins de novo to bind two distinct uPAR epitopes, and then chemically or genetically fusing them such that each simultaneously bind uPAR, or 4) Re-engineering two natural uPAR ligands that already simultaneously bind uPAR to function as antagonists, and fusing them together such that they simultaneously bind the receptor. There are two key challenges inherent in the first three strategies mentioned; the first challenge is developing a high-throughput assay capable of separating off-target uPAR binders from binders of the desired epitope, and the second is engineering de novo ligand with binding affinities superior to the wild-type ligands. The latter challenge is particularly difficult in the case of interrupting the uPA-uPAR interaction, which has an affinity of approximately 230 pM'. The fourth strategy mentioned - re-engineering uPAR’s native ligands as antagonists, avoids both of these challenges. Specifically, starting with the native ligands preserves the mechanism of binding, ensuring specific inhibition of only the target uPAR epitopes. Additionally, each ligand starts with a binding affinity equal to the wild-type interaction; any further improvement in affinity from engineering efforts would result in ligands already more capable of outcompeting the natural ligands for binding. Therefore, we focused our efforts on re-engineering uPAR’s natural ligands to form a bispecific ligand.

Our first step toward engineering a ligand-based bi specific uPAR antagonist was to select two ligands that simultaneously bind uPAR, and when re-engineered, are capable of blocking multiple cancer-associated uPAR interaction s. Domains of uPA and the ECM component vitronectin meet these criteria. uPA contains a growth factor domain, a kringle domain, and a protease domain. The growth factor domain (GFD, amino acids 1 -48) binds uPAR with an affinity equal to that of full-length uPA (KD =

0.23 nM) ⁷ , and has no other known binding partner. Additionally, GFD lacks both the protease domain and kringle domain that are required for focalizing ECM degradation at the cell surface, and for uPAR interactions with anb3 integrin ⁸ ’ ⁹ , gp 130 (ref ¹⁰ ) and PAI- 1 (ref ^u ) that play roles in cancer metastasis (FIG. 1 A, FIG. IB, i). Furthermore, GFD by itself already has a baseline level of anti -cancer activity; treating mice with GFD fused to the Fc domain of an antibody (GFD-Fc) has been shown to reduce tumor growth in vivo ¹² ’ ¹³ . However, the GFD-uPAR complex also binds vitronectin ¹⁴ , which induces cell migration. Additionally, GFD is predicted to be the only urokinase domain required to induce EGFR activation via formation of "the proliferasome" ¹⁵ , and downstream signaling for cell proliferation, migration, and angiogenesis. Therefore, GFD alone is a sub optimal uPAR antagonist; it inhibits a subset of uPAR functions, but activates others that facilitate cancer metastasis. The somatomedin B domain (SMB, amino acids 1 -44) of vitronectin binds the urokinase-uP AR complex with a cooperative affinity of 300 nM, or binds uPAR alone with a K _d of ~ 2 pM ⁷ . SMB blocks uPAR from binding vitronectin ⁷ and FPR1/2 ⁷ , and is predicted to block EGFR activation by binding the same domain required for uRAII/aίbi integrin/EGFR complex formation ^{16 17} (FIG. IB, ii). Thus, a bispecific ligand comprising GFD linked to SMB is expected to simultaneously block multiple uPAR interactions that coordinate cancer metastasis (FIG. I B, iii). Furthermore, fusing the wild-type GFD and SMB proteins is expected to create a bispecific ligand with substantially higher affinity compared to native uPA and vitronectin due to the avidity effect of GFD and SMB simultaneously binding uPAR We therefore chose to develop a bispecific uPAR ligand comprised of GFD genetically fused to SMB

In addition to engineering the human uPAR (huPAR) ligand, we also followed the same strategy to engineer a mouse uPAR (muPAR) ligand in parallel, which will be required to demonstrate the full potential of a uPAR-targeted therapy in mouse model s of cancer. This is because human uPA/GFD does not bind mouse uPAR, and mouse uPA/GFD does not bind human uPAR ¹⁸ . Therefore, when testing huPAR ligands in a xenograft model, the cancer cell-surface huPAR will not be stimulated by endogenous mouse uPA, and thus antagonizing the receptor is expected to produce a lower, less physiologically accurate response. Furthermore, the huPAR ligand will not inhibit mouse uPAR present on the vasculature that facilitates tumor-associated angiogenesis and tumor growth. Inhibiting muPAR in a syngeneic mouse model avoids these issues. In this case, the murine cancer cell-surface muPAR would be stimulated by endogenous mouse uPA, which when inhibited is expected to result in a more substantial and accurate response. In addition, the muPAR on endothelial cells facilitating tumor-associated angiogenesis will also be inhibited. Targeting muPAR activity in a syngeneic mouse model is thus expected to more accurately represent the full potential of a uPAR antagonist.

We used yeast surface display technology and directed evolution to engineer the bi specific uPAR ligands. First, the individual domains and various GFD-SMB fusion designs were displayed on the surface of yeast and their binding affinities for uPAR were measured. As predicted, fusing GFD and SMB resulted in an 8.3 -fold and 8.7-fold improvement in binding affinity for the mouse and human proteins, respectively, compared to indivi dual GFD. This improvement is due to the avidity effect of the domains simultaneously binding uPAR. Subsequently, libraries of GFD and SMB mutants were generated and screened by FACS to identify amino acid mutations that confer enhanced binding affinity for uPAR This yielded mutants with additional 6-to- 235 -fold improvements in binding affinity. To our knowledge, these are the highest affinity uPAR ligands reported to date, and represent significant progress towards developing a more effective uPAR-targeted therapy. 3.3 RESULTS

3.3.1 ENGINEERING BISPECIFIC UPAR LIGANDS.

To design a ligand-based bispecific uPAR ligand it was critical to engineer the peptide linker connecting the termini of GFD and SMB such that both ligands can simultaneously bind uPAR, while avoiding excess conformational entropy in the linker that would otherwise hinder the overall affinity of the fusion protein. We optimized two parameters to achieve this goal: 1) linker length, and 2) linker amino acid composition.

To first approximate the minimum and maximum boundaries of the linker length to explore, we used data from the crystal structure of the solved uPA-uPAR-SMB complex ¹⁹ (PDB 3BT1) to model various linker lengths connecting the COOH-terminus of GFD to the NH2 -terminus of SMB. Specifically, we wrote amino acid coordinates into the PDB text file bridging the GFD and SMB domains, and then loaded the edited text file into UCSF chimera software ²⁰ to assign possible structures to the added linker sequence. We started by modeling two different linker lengths comprising repeating GGGGS (GlyrSer) sequences, a [Gly ₄ Ser] ₄ and [Gly ₄ Ser]s linker, predicted to have total lengths of 40 and 50 angstroms, respectively. Linkers composed of Gly ₄ Ser repeats are commonly used due to their favorable solubility and inherent flexibility ¹⁸ , allowing the fused proteins to sample a vast landscape of orientations, including, ideally, the orientation required for both to simultaneously engage the target.

Fusing GFD and SMB with a [Gly ₄ Ser] ₄ linker appears to allow both ligands to simultaneously bind uP AR (FIG. 2), with backbone dihedral angles in the linker predicted to be energetically allowed by Ramachandran plots (not shown). However, the linker has a limited set of conformations that bridge GFD and SMB, as one of the three possibilities shown clashes sterically with the receptor. This suggests a longer linker may be more optimal. Indeed, the [Gly ₄ Ser]s linker appears long enough to orient around all angles of the receptor without causing steric clash (FIG. 3), also with backbone dihedral angles predicted to be favorable by Ramachandran plots (not shown). We therefore generated and tested three different GFD -SMB fusion proteins containing a [Gly ₄ Ser] ₃ , [Gly ₄ Ser] ₄ , or [Gly ₄ Ser]s linker. The [Gly ₄ Ser] ₃ linker was included as a negative control, as modeling predicts it is too short to allow GFD and SMB to simultaneously bind uPAR. To test the different linker designs, the mouse GFD-SMB proteins were first displayed on the surface of yeast as N-terminal fusions to the aga2p yeast coat protein, and their binding affinities for soluble muPAR were measured. In addition, mouse GFD (mGFD) and SMB (mSMB) were also displayed and their binding affinities were measured for comparison. This is a rapid method for generating proteins and measuring their affinities for a soluble target, as it does not require the laborious soluble expression and purification of each individual protein. The three mGFD-mSMB designs with varying linker lengths all expressed well on the surface of yeast and showed significant binding to soluble recombinant muPAR (FIG. 4). The mGFD-mSMB proteins with 3x, 4x, and 5x Gly4Ser linkers had Kd values of 2 34 nM, 0.83 nM, and 0.55 nM, respectively, indicating the fusion protein with the 5x(Gly ₄ Ser) linker is the most optimal of the three as predicted by our modeling results. When compared head-to-head with mGFD, mGFD-mSMB bound muPAR with 8.3 -fold higher affinity (K _d = 0.22 nM, FIG. 5), indicating the GFD and SMB domains are simultaneously binding muPAR. To further confirm the enhanced affinity observed for mGFD-mSMB compared to mGFD alone is due to the added SMB domain specifically and simultaneously binding muPAR, we made two amino acid mutations in the SMB domain of mGFD-mSMB (Y27A and Y28A) that are known to abolish SMB-uPAR binding ^{21 23} . As expected, the mutant mGFD- mSMB ^{Y2 /A} _’ ^Y28A protein bound muPAR with the same affinity as mGFD (FIG. 6). The affinity of mSMB alone is too low (~ 2 mM) ⁷ to detect binding to muPAR using this assay, and therefore those data are not included here. These results indicate the bi specific mGFD-mSMB protein binds with superior affinity compared to the wild-type mGFD protein, and simultaneously engages both the uPA (GFD) and vitronectin (SMB) binding sites on muPAR.

Given the [Gly ₄ Ser] ₅ linker was optimal in the mGFD-mSMB design and the high structural similarity between the hGFD-huPAR and mGFD-muPAR complexes (FIG. 7) we also used this linker to generate the bispecific human uPAR ligand comprising human GFD (hGFD) linked to SMB (hSMB), hereafter called "hGFD- hSMB". As before, this protein was displayed on the surface of yeast and its binding affinity for soluble huPAR was measured and compared with the individual hGFD and hSMB domains. All three proteins expressed well on the surface of yeast and showed significant binding so soluble recombinant huPAR (FIG. 8), except hSMB which again has too low of an affinity (~2 mM) ⁷ to detect binding to huPAR using this assay (data not shown). Similar to the hGFD- hSMB protein, hGFD-hSMB bound huPAR with an 8.7-fold higher affinity (K _d = 0.12 nM) compared to hGFD alone (K _d = 1.04 nM). To again confirm the enhanced affinity observed for hGFD-hSMB compared to hGFD is due to the added hSMB domain specifically and simultaneously binding huPAR, we made the same two amino acid mutations in the SMB domain of hGFD-hSMB (Y27A and Y28A) that abolish its affinity for uPAR ^{21 23} . As expected, the mutant hGFD-hSMB ^Y27A,Y28A protein bound uPAR with the same affinity as hGFD (FIG. 9).

Without further engineering, the hGFD-hSMB and hGFD-hSMB proteins already block multiple functions of uPAR with superior affinities compared to the wild-type ligands. However, that is assuming both epitopes of uPAR are available for mGFD- mSMB to bind, which will not always be the case in vivo. For example, if cancer cell- surface muPAR is already bound to vitronectin, the mGFD-mSMB protein affinity will be effectively reduced to that of wild-type mouse uPA. Likewise, if muPAR is already bound to mouse uPA, the mGFD-mSMB affinity will be effectively reduced to that of wild-type vitronectin. In these cases, mGFD-mSMB will be less capable of outcompeting uPAR’ s natural ligands for binding. To solve this problem, we next used directed evolution to identify amino acid mutations in the GFD and SMB domains that enhance their individual affinities for uPAR, thus making them more effective competitors. We also evolved the [Gly4Ser]5 linkers in parallel to search for amino acid mutations that reduce excess linker entropy and favor simultaneous binding of each ligand.

3.3.2 RECOMBINANT SOLUBLE UPAR EXPRESSION AND

PURIFICATION

Directed evolution and characterization of the uPAR ligands was expected to require significant amounts of recombinant uPAR protein. Therefore, we established methods for producing and purifying these proteins. The cDNA sequences coding for soluble human and mouse uPAR were cloned into the mammalian expression vector pCEP4, and the proteins were expressed in suspension Freestyle 293 -F cells using transient transfection. The recombinant proteins lack amino acid G305 (huPAR) or G298 (muPAR) required for lipidation, and are thus released from the cell as soluble proteins after being exported to the extracellular space. Two versions of huPAR and muPAR were expressed, one with an NH2-terminal FLAG tag as a handle for detection in vitro by a fluorescent anti -FLAG antibody, and one without. All proteins were expressed with C- terminal 6x -histidine tags as handles for purification. The 6x-histidine tagged huPAR and muPAR (henceforth simply called "huPAR" and "muPAR") and the FLAG-tagged muPAR (henceforth called muPAR-FLAG") were expressed and purified by nickel-NTA affinity chromatography followed by size exclusion chromatography (FIG. 10A). The final products were > 95-99% pure (FIG. 10B), and the yields were 1.8 mg/L, 12 mg/L, and 1.1 mg/L for huPAR, muPAR, muPAR-FLAG, respectively. Unfortunately, we were unable to separate the desired FLAG-tagged huPAR protein from expression byproducts using size-exclusion chromatography. Therefore, we used the 6x -histidine-tagged huPAR to evolve high-affinity hGFD and hSMB ligands. To our knowledge, this is the first documented method of producing these proteins using FIEK 293 cells as an expression host.

3.3.3 MOUSE GFD LIBRARY 1 GENERATION, SORT PROGRESSION, AND VARIANT CHARACTERIZATION

Amino acids 19 - 31 in the mouse GFD protein are primarily responsible for forming the interaction between muPA and muPAR, with additional potential interactions between amino acids 39-42. Exhaustively testing every possibl e combination of amino acids in these regions using saturation mutagenesis would require a library of 1.3 ^c IQ ²² variants, which is far larger than the ~10 ⁸ variant limit that can be practically screened using FACS. Therefore, we instead used error-prone PCR to make random mutations across the entire 46 amino acid sequence. In addition to exploring mutations at the binding interface, this approach also has the potential to identify mutations distal from the binding interface that have a positive allosteric effect on binding affinity, and/or that enhance the stability and expression of the protein ²⁴ . Five separate DNA librari es were generated using varying concentrations of dNTP analogs that induce A-to-C, G-to-T, A- to-G, and G-to-A mutations. The DNA libraries were sequenced to identify which contained the desired frequency of mutation (FIG. 11). Two libraries were pooled that had amino acid mutations ranging from 0-to-3 and 0-to-5 per gene, with averages of 1 and 2 mutations per gene. The pooled DNA library was transformed into yeast resulting in a protein library of approximately 5.9 x 10 ' variants. Preliminary analysis indicated ~ 20% of the library retained significant binding affinity to soluble muPAR (data not shown), indicating the level of mutation produced the desired stratification of binding affinities (i.e. within 20 - 80% of variants retaining binding).

The mGFD library was sorted using equilibrium binding conditions to separate variants with improved binding affinities from those with weaker affinities. For the first sort, the library was incubated with 1 nM muPAR-FLAG and approximately 50% of the binding population (10% of the library) was collected (FIG. 12). The purpose of the first sort round is primarily to remove the non-binding and low-binding variants and reduce the library size to a level that can be sorted at higher coverage (e.g., lOx) in the next sort. For the second sort, the library was again incubated with 1 nM muPAR-FLAG, but only 1% of the binding population was collected to increase the rate of enrichment of high- affinity variants. Finally, for sort 3 and 4 the concentration of muPAR-FLAG was reduced to 200 pM to increase the selective pressure for improved variants, and only the top 0.5-to-l% of binders were collected in each round. There was a noticeable increase in binding from sort 3 to 4 (FIG. 12), indicating the sorting strategy successfully isolated superior variants in the library.

DNA from the variants remaining after the 4 ^th sort was isolated and sequenced to identify consensus mutations that may confer enhanced affinity for muPAR. Of the 20 sequences obtained, 6 unique variants had an E41K mutation (FIG. 13 A), suggesting this mutation may be responsible for the increase in muPAR binding observed between sorts 3 and 4. Indeed, when analyzed in bulk, the library after sort 4 had a 2.5-fold

improvement in apparent K _d compared to wild-type mGFD (FIG. 13B). Our goal was to improve the affinity of mGFD by a minimum of 10-fold, so we submitted the remaining vari ants to another round of directed evolution. 3.3,4 MOUSE GFD LIBRARY 2 GENERATION, SORT PROGRESSION,

AND VARIANT CHARACTERIZATION

A similar strategy was used to generate the second mGFD library. DNA isolated after the 4 ^th sort of the first library was mutagenized again using error-prone PCR. Five separate DNA libraries were generated using a range of dNTP analog concentrations and sequenced to determine which had the desired mutation frequency (FIG. 14). DNA from three of the libraries with averages of 3.4, 2.5, and 1.5 amino acid mutations per gene were pooled and transformed into yeast resulting in a protein library of approximately 2 x 10' variants. Preliminary analysis indicated ~ 20% of the library retained binding to soluble muPAR (data not shown).

The library was sorted first using equilibrium binding conditions to reduce the library to an amenable size, and then using a kinetic off-rate screen to isolate variants with slower off-rates and thus improved binding affinities for muPAR. For the first sort, the library was incubated with 1 nM muPAR-FLAG and approximately 50% of the binding population (10% of the l ibrary) was collected (FIG. 15). The concentration of muPAR-FLAG was reduced to 500 pM in the next sort to increase the selective pressure for binding, and only the top 1% of the binding population was collected to further enrich for higher-affinity variants.

After the second sort, the remaining library was estimated to contain around 4.5 x 10 ⁵ unique variants, which is an amenable size to sort using a kinetic off-rate screen. This type of screen is used when the average affinity of the library approaches the 1 ^x 10 ¹⁰ - to - 1 x 10 ¹¹ M range, whereafter issues with ligand depletion and the incubation times required for binding to go to completion in an equilibrium binding screen become impractical to manage. Using a kinetic off-rate screen avoids these challenges, and allows the application of a more stringent selection pressure for variants with lower off-rates, and thus improved binding affinities (K _d = k _o n/k _on ). In this strategy, the library is completely saturated with muPAR-FLAG such that all surface-displayed variants are bound, and then the unbound muPAR-FLAG is thoroughly washed away and replaced with a molar excess of a "competitor" protein that renders dissociation of muP AR-FLAG from displayed variants irreversible. In this case, a 1, 000-fold molar excess of muPAR without a FLAG tag was used as the competitor. When muPAR-FLAG dissociates from a displayed variant, its place is rapidly taken by the competing muPAR protein.

Subsequently, the library is sorted for yeast cells expressing variants that remain bound to FLAG-muPAR by using an anti-FLAG fluorescent antibody.

For the first off-rate sort (3 ^rd library sort), the library was incubated with 50 nM muPAR-FLAG to saturate the displayed variants. The excess unbound muPAR-FLAG was then removed and replaced with 2 mM muPAR and the library was incubated for an additional 2 hours (here called "2 hour off). Subsequently, the library was sorted for cells that had the highest percentage of displayed variants still bound to muPAR-FLAG. The incubation time with 2 mM muPAR was successively increased from 2 hours in the 3rd sort to 3 hours in the 4 ^th and 5 ^th sorts, and finally to 12 hours in the 6 ^th sort to increase the selective pressure for variants with improved off-rates. In each sort, 0.5 - to - 1% of the top binding variants were collected. There was a noticeable increase in the amount of muPAR bound to the library between sort 4 and 5, and sort 5 and 6 (FIG. 15), indicating the kinetic screen successfully enriched for superior variants in the library.

35 clones and 10 clones were isolated from the library after the 5 ^th and 6 ^th sorts, respectively, and their DNA were sequenced to identify consensus amino acid mutations responsible for the improved binding properties. There was a strong consensus for 4 amino acid mutations (K37G, Q39R, E41R, and F142Y) after the 5 ^th sort, with variants containing all 4 making up 46% of the 35 cl ones sequenced (FIG. 16 A). After the 6th sort the library converged on a single variant, with 100% of the 10 clones isolated having identical sequences (FIG. 16B). Surprisingly, the remaining variant was missing consensus mutations identified after the previous sort, and had a new mutation, R30W, that was yet to be detected in previous sequences. Judging by the sort plots, it appears this new R30W mutation is responsible for the substantially superior binding population that emerged in the 6 ^th sort. We therefore suspected adding the R30W mutation to the dominant variant remaining after sort 5 would generate a superior variant. Indeed, the variant with all 5 mutations (henceforth referred to as *mGFD, FIG. 45) had

approximately 2-fold higher affinity for muPAR compared to the final variant isolated after sort 6 (data not shown). When compared head-to-head with wild-type mGFD, the *mGFD protein bound muPAR with 235 -fold higher affinity (K _d = 9 pM, FIG. 17A). As would be expected, 4 out of the 5 mutations in *mSMB are at the interface of the *mGFD-muPAR complex (FIG. 17B), suggesting they may be contributing additional binding interactions with the receptor. To our knowledge, this is the highest affinity uPAR-binding protein reported to date, and represents a noteworthy protein engineering achievement and a significant step towards developing a next generation uPAR ligand.

3.3.5 HUMAN GFD LIBRARY 1 GENERATION, SORT PROGRESSION,

AND VARIANT CHARACTERIZATION

Although there is no cross-reactivity between human urokinase and mouse uPAR, or mouse urokinase and human uPAR, an alignment of the uPA-uPAR complexes from both species reveals high structural similarity between the two (FIG. 7). Therefore, hGFD was engineered in parallel with mGFD using the same strategy Error-prone PCR was used to generate 5 separate DNA libraries of hGFD variants, and the libraries were sequenced to identify those with the desirable level of mutation (FIG 18). Two libraries that had an average of 1 -2 mutations per gene were pooled and transformed into yeast resulting in a protein library of - 3 ^c 10 ' variants. Preliminary analysis indicated approximately 30% of the library retained significant binding to soluble huPAR (data not shown).

The hGFD library was sorted using equilibrium binding conditions to isolate variants with improved affinities for huPAR. For the first sort, the library was incubated with 1 nM huPAR and approximately 40% of the binding population (-10% of the library) was collected (FIG. 19). There was an interesting characteristic of this library not observed in others - the binding population was separated into two distinct groups with different levels of expression. It’s still not clear what caused this, but we decided to continue sorting the library' nonetheless. Two gates were drawn to collect variants from both binding populations. The same conditions were used for the second sort, and approximately 5% of the top binding populations were collected. The concentration of huPAR was reduced to 500 pM and 400 pM in the 3 ^rd and 4 ^th sorts, respectively, and only the top 1% of binders were collected to increase the selective pressure and further enrich for higher-affinity variants. By the third sort the two distinct binding populations converged and it was only possible to gate cells from the left-most population.

DNA from the variants remaining after the 4 ^th sort was isolated and sequenced to identify consensus mutations that may confer enhanced affinity for huPAR. Of the 20 sequences obtained, 6 unique variants had a D12G mutation, and three had a K46R mutation (FIG. 20A). To parse the benefits of each mutation, variants with all combinations of the two mutations were compared with wild-type hGFD for their huPAR binding affinity (FIG. 20B). The variant containing both mutations had the highest improvement in affinity (3.3 -fold, ¾ = 130 pM), indicating the sorting strategy successfully enriched for improved mGFD variants. However, as with mGFD, our goal was to improve the affinity of hGFD by a minimum of 10-fold. We therefore submitted the variants isolated after the 4 ^th sort to a subsequent round of directed evolution. 3.3.6 HUMAN GFD LIBRARY 2 GENERATION, SORT PROGRESSION,

AND VARIANT CHARACTERIZATION

A new strategy was explored to generate the second hGFD library. DNA isolated after the 4 ^th sort of the first library was "shuffled" using the PCR-based staggered extension process (StEP) ²⁵ . Rather than adding additional mutation to the library, this process combines, or shuffles, mutations from different variants to generate a library of new variants. As a result, favorable mutations present in the enriched pool of binders are combined to create variants with potentially further enhanced binding affinities. To increase the probability of identifying improved binders, we generated two separate libraries, one using the StEP method and the other using the standard error-prone PCR method. The DNA libraries were transformed into yeast resulting in two protein libraries with approximately 2 5 x lO ⁷ variants each. Approximately 50% of the library generated by error-prone PCR retained binding to soluble huPAR, and interestingly, over 95% of the library generated by the StEP method retained binding to huPAR. The higher percentage of binders in the library generated by StEP is likely due to the mutations already being selected for in the previous round; shuffling neutral or beneficial mutations is not expected to dramatically reduce binding affinity.

Both hGFD libraries were first sorted separately using an equilibrium binding screen to reduce them to a size amenable to sorting using an off-rate screen. Each library was incubated with 1 nM huPAR and approximately 5 - 15% of the top binders were collected and combined. For sorts 2-4 the concentration of huPAR incubated with the library was reduced to 0.5 nM, 0.25 nM, and 0.15 nM respectively, and the top 0.5 - 1% of binders were collected. In the following sorts the remaining variants were subjected to an off-rate screen to further increase the selective pressure for binding. After saturating the displayed variants with huPAR, the remaining unbound uPAR was removed and replaced with a 100-fold molar excess of soluble hGFD fused to the Fc domain of an antibody (hGFD-Fc). The hGFD-Fc competitor effectively renders dissociation of huPAR irreversible, as unbound uPAR is rapidly bound by the excess hGFD-Fc. The library was then incubated with hGFD-Fc for an additional 24 hours and sorted for cells displaying the highest percentage of variants that remained bound to huPAR. In each sort, 0.5 - to - 1% of the top binding variants were collected. There was a noticeable increase in the amount of huPAR bound to the library between sort 5 and 6 (FIG. 21), indicating the kinetic screen successfully enriched for superior variants in the library.

10 clones and 40 clones were isolated from the library after the 4 ^th and 6 ^th sorts, respectively, and their DNA were sequenced to reveal amino acid mutations driving their enhanced affinity for huPAR (FIG. 22). Consensus mutations already emerged after the 4 ^th sort with only minor changes after two additional sorts. Six mutations (L4W, Q6R, K36G, Q40R, 1141R, and K46R) were present in over 60% of the 40 clones sequenced after the 6 ^th sort. Interestingly, the library didn’t converge on a single dominant variant after the 6 ^th sort, with the most frequent variant only making up 22% of the sequences.

We therefore screened seven unique variants of interest (FIG 23 A) to identify which had the slowest off-rate, and thus the highest affinity for muPAR. Of the seven tested, one variant (L2.S6.15) stood out from the others as having a slower kinetic dissociation rate (FIG. 23B). When compared directly with wild-type hGFD, the L2.S6.15 mutant (henceforth referred to as *hGFD, FIG. 45) had a 30-fold higher affinity for huPAR (K _d == 28 pM, FIG. 24A). Four out of seven mutations in *hGFD are at the interface of the *hGFD-muPAR complex (FIG. 24B), and may thus be contributing additional binding interactions with the receptor.

3.3.7 MOUSE LINKER-SMB LIBRARY GENERATION, SORT

PROGRESSION, AND VARIANT CHARACTERIZATION

Our next goals were to 1) enhance the affinity of mSMB for muPAR, and 2) search for mutations in the [Gly+Serjs linker joining mGFD and mSMB that further facilitate simultaneous binding of both ligands to muPAR, resulting in an overall increase in binding affinity of the mGFD-mSMB protein. We developed a strategy to accomplish both goals in parallel. Error-prone PCR was used to make random mutations across only the linker and mSMB domain of the mGFD-mSMB gene. As described previously, 5 DNA libraries were generated and sequenced (FIG. 25), and two were pooled that had the desired frequency of amino acid mutations ranging from of 3 to 4.8 per gene. The pooled DNA library was transformed into yeast resulting in a protein library of approximately 7 x 10 ⁷ variants, of which approximately 35% retained binding to soluble muPAR (data not shown).

The linker-mSMB library was first sorted using an equilibrium binding screen, followed by a kinetic off-rate screen to isolate high-affinity variants. In the first two sorts, the library was incubated with 500 pM and 100 pM muPAR-FLAG, and the top 10% and 1% of binders were collected, respectively (FIG. 26). From the 3 ^rd sort onward, the variants were sorted based on their off-rates as previously described for the mGFD library- in Section 3.3.4. The library was incubated with 50 nM muPAR-FLAG to saturate all displayed variants. Then, unbound muPAR-FLAG was removed and replaced with a molar excess of muPAR without a FLAG tag, and the library was incubated for an additional 2, 5, 12, and 16 hours for sorts 3-6, respectively, to increase the selective pressure for variants with slower dissociation rates. In each sort, 0.5 - to - 1% of the top binding variants were collected. There was a noticeable increase in the amount of muPAR remaining bound between sort 3 and 4, and sort 5 and 6 (FIG. 26), indicating the kinetic screen successfully enriched for superior variants in the library. 10 clones were isolated after the 6 ^th sort and their DNA were sequenced to identify consensus mutations behind the observed improvement in binding affinity. All 10 sequences were identical (FIG. 27), indicating the library had converged on a single superior variant. The variant contains a total of seven mutations, three in the linker region, and four in the SMB domain. When compared with the parent mGFD-mSMB fusion protein, the variant (henceforth referred to as mGFD-*mSMB, FIG. 45) bound muPAR with a 6-fold higher affinity (K _d = 125 pM, (FIG. 28), indicating the engineering strategy successfully evolved improved muPAR binding. Unfortunately, however, none of the mutations in the SMB domain are at the binding interface of the mSMB-muPAR interaction (FIG. 29), suggesting the affinity for the individual SMB-muPAR interaction was not improved.

The enhanced affinity is thus likely a result of the mutations improving the linker’s proficiency to coordinate simultaneous binding of the GFD and SMB domains to muPAR. Whereas mutations that add additional binding interactions reduce the off-rate of the interaction, these mutations are effectively increasing the on-rate of the second ligand after the first has bound muPAR.

3.3.8 HUMAN LINKER-SMB LIBRARY GEN ERATION, SORT

PROGRESSION, AND VARIANT CHARACTERIZATION

A slightly different strategy was used to improve the linker-SMB region of the hGFD-hSMB fusion protein. Given the high affinity interaction between hGFD-hSMB and huPAR (K _d = 115 pM), we postulated yeast display may not offer a large margin for improvement, as it becomes challenging to isolate variants with improved affinity when the general K _d of the library approaches the femtomolar range. To avoid this potential problem, we searched for mutations in the GFD region that reduced the overall affinity of the fusion protein. Three mutants were tested, each with a single amino acid change in the main binding loop of hGFD: F25A, W30A, and I28A. Of the three tested, the hGFD ^F25A -hSMB mutant had the desired reduction in affinity (K _d = 15 nM), while still maintaining a characteristic binding interaction with huPAR (FIG. 30). We therefore engineered the linker-hSMB domain in the context of the hGFD ^F25A -hSMB mutant. Error-prone PCR was used to generate ten separate DNA libraries with varying levels of mutation. The libraries were sequenced (FIG. 31), and seven were pooled to form an overall library with an average of 3-4 amino acid mutations per gene. The pooled DNA library was transformed into yeast resulting in a protein library of approximately 5 x 10 ⁶ variants, of which only approximately 5% retained binding to soluble huPAR (data not shown). Both the library size and the percent of variants retaining huPAR binding were about 10-fold lower than desired. However, due to limited time we decided to sort the library regardless.

Given the higher starting affinity of this library (K _d - 15 nM), equilibrium binding screens were sufficient for creating selective pressure in all sorts. The library was first incubated with 250 nM huPAR and a generous gate was drawn to collect the small fraction of variants retaining binding to huPAR (FIG. 32). In sorts 2 through 5, the huPAR concentration was reduced sequentially from 5 nM to 0.025 nM to increase the selective pressure for improved variants. The top 0.17 - 0.34 % of binders were collected after each sort. There was a noticeable increase in huPAR binding between sorts 3 and 4 even though the huPAR concentration was reduced by 10-fold (FIG. 32), indicating the sorting successfully enriched for superior variants.

10 clones were isolated after the 5 ^th sort and their DNA were sequenced to identify consensus mutations responsible for improving binding. Only two unique variants were present, with one making up 80% of the 10 sequences (FIG. 33A). Both variants had approximately the same affinity for huPAR (0.44 nM and 0.37 nM, FIG. 33B). When compared to the parent hGFD ^F25A -h SMB protein, the slightly superior variant (henceforth referred to as hGFD ^F2'A -*hSMB, FIG. 45) had a 10-fold improvement in binding affinity (FIG. 34), indicating the engineering strategy successfully evolved improved huPAR binding. The hGFD ^F25A -*hSMB protein contains 5 amino acid mutations total, 1 in the linker region and 4 in the SMB domain. Just as with the mGFD- *mSMB variant, none of the mutations in the hSMB domain are at the binding interface of the h SMB -huPAR interaction (FIG. 35). This suggests the amino acid mutations in hGFD-*hSMB are instead increasing the capability of both ligands to simultaneously engage huPAR. 3.4 DISCUSSION

Since the discovery of uPAR’s role in cancer growth and metastasis, considerable progress has been made toward developing uPAR-targeted therapeutics for clinical use. The most noteworthy was the significant in vivo efficacy demonstrated by a first-in-class molecule targeting uPA in patients with epithelial ovarian cancer ¹ Although a promi sing proof of concept, the underwhelming efficacy of this molecule is suspected to have impeded further development toward the clinic, as is also the case for other previously developed uPAR ligands. The lack of efficacy delivered by these molecules can be attributed to their inability to completely and effectively antagonize uPAR, as they only block a subset of uPAR interactions with affinities equal to or weaker than the native binding interactions they target.

Our hypothesis is a ligand capable of blocking multiple uPAR interactions with superior affinity may deliver more significant efficacy in vivo. Towards this goal, we have generated a bispecific uPAR ligand that simultaneously blocks both the urokinase and vitronectin binding sites with orders-of-magnitude higher affinity compared to the native uPAR ligands. This was accomplished by genetically fusing the GFD domain of uPA to the SMB domain of vitronectin using a 25-amino acid [Gly ₄ Ser]5 linker, which resulted in an 8.3 and 8.7-fold improvement in apparent binding affinity for the mouse and human uPAR ligands, respectively, due to the avidity effects of multivalent binding. Yeast display was then used to enhance the individual affinities of GFD and SMB, and improve the linkers abili ty to coordinate simultaneous binding of both ligands to uPAR. The affinities of the human and mouse GFD proteins were improved by 30-fold, and 235- fold, respectively. The linker and SMB regions were then engineered as a single unit, and additional 10-fold and 6-fold improvements in affinity were achieved for the bispecific human and mouse uPAR ligands, respectively. Since the mutations driving the enhanced affinity for the linker-SMB regions are not at the binding interface of the SMB-uPAR interaction, we suspect these mutations are improving the overall affinity of the fusion proteins by increasing the ability of both ligands to simultaneously bind uPAR, or more specifically, by increasing the on-rate of the second ligand after the first ligand has bound. The bispecific proteins comprising the improved human and mouse GFD and linker-SMB domains (henceforth referred to as hGS and mGS, respectively, Table 1 and FIG 36) are thus expected to have over 1, 000-fold higher affinity for uPAR relative to the wild-type ligands, representing the highest affinity uPAR ligand developed to date that also simultaneously block multiple functions of uPAR involved in cancer metastasis.

We suspected amino acid mutations improving the mouse uPAR ligand affinities would not be transferable to the human ligands due to the absence of species cross reactivity between hGFD and muPAR, and mGFD and huPAR Therefore, the human and mouse ligands were both engineered separately. Indeed, unique mutations were identified in each ligand, suggesting this strategy was appropriate. Interestingly, however, 2 of the 5 mutations in the engineered *mGFD protein were also present in the *hGFD protein, (K37G and E41R in *mGFD corresponding to K36G and Q40R in *hGFD protein). The numbers designating amino acid position are offset by 1 between the mouse and human proteins due to mGFD having an additional amino acid at the Fb-termimus. These mutations are located outside of the growth factor-like binding loop (amino acids 19 - 31) that is primarily responsible for the high-affinity interaction with uPAR. Therefore, although the interaction between the primary binding loop of uPA and uPAR has diverged significantly between the human and mouse proteins (eliminating cross reactivity between them) it appears the region in human and mouse uPAR that interacts with amino acids 37 and 41 in mGFD, and 36 and 40 in hGFD, has not. It is still possible other mutations identified in *mGFD may benefit *hGFD, and vice versa. Of particular interest i s the R30W mutation found in *mGFD that appears to dramatically improve binding affinity. Although we did not add this mutation to *hGFD, it may be worth investigating if future improvement of *hGFD is desired.

In summary, we used insights from previous structural analysis of the uPA-uPAR- SMB binding interaction to develop an engineering strategy focused on producing an improved uPAR ligand. Using yeast display as a platform for engineering, we generated bispecific ligands that simultaneously engage two functional uPAR epitopes with orders- of-magnitude higher affinity than the wild-type uPAR ligands. 3.6 MATERIALS AND METHODS

3.6.1 MEDIA AND REAGENTS

SD-CAA media was composed of 20 g/L dextrose, 6.7 g/L yeast nitrogenous based without amino acids, 5.4 g/L Na ₂ HP0 ₄ , 8.6 g/L NatLPQ^ILO, and 5 g/L Bacto casamino acid, pH 4.5. SG-CAA media was identical except for galactose replacing the dextrose, and was titrated to pH 6.0. YPD media contained 20 g/L dextrose, 20 g/L peptone, and 10 g/L yeast extract. SD-CAA plates contained 182 g/L sorbitol and 15 g/L agar in addition to the media components. PBS A was lx phosphate buffered saline with 1 g/L bovine serum albumin added and sterile filtered using a 0.22-micron filter

(Millipore). Chicken anti-c-myc (A21281) was purchased from Invitrogen, Goat anti chi cken-PE (ab72482) was from Abeam, goat anti-chicken-488 (A11039) and rabbit anti- 6histidine-488 (NC0585792) were from Fisher Scientific, and anti-FLAG-PE (PJ315) was from PROzyme. The yeast-display pCTcon2 vector and EBY100 yeast cells were a gift from Professor Dane Wittrup’s research group at MIT.

3.6.2 SOLUBLE RECOMBINANT PROTEIN EXPRESSION AND

PURIFICATION

Human and murine uPAR were expressed in Freestyle 29 -F cells (Thermo Fisher Scientific) suspension cells, and purified by nickel -NT A metal affinity chromatography followed by size exclusion chromatography. cDNA encoding amino acids 1-303 for huPAR and 1-296 for muPAR were cloned into the mammalian expression vector pCEP4 (Thermo Fisher Scientific) between the Hindlll and Xhol (huPAR) and Hindlll and BamHI (muPAR) restriction sites. The human signal peptide was used in place of the murine signal peptide for the muPAR gene. Each gene was immediately preceded by a Kozak sequence (GCCACC), and followed by a C -terminal 6x-histidine tag. A separate version of the muPAR protein was expressed that contained a GGGS linker (SEQ ID NO: 16) and FLAG tag (GGGGSDYKDDDDK, SEQ ID NO: 10) between the signal peptide and L-terminus of the receptor. Freestyle 293F cells were transiently transfected with the expression vectors following the manufacturers protocols. Briefly, cells were grown in suspension culture at 37°C in a humidified atmosphere with 5% CO ₂ in Freestyle 293 -F culture medium (Thermo Fisher Scientific) to density of

approximately 1-3 million cells/mL, after which they were passaged in fresh medium to 0.3 million cells/mL. Cells were passaged a minimum of 5 times prior to transfection.

The day before transfection, cells were diluted to 0.5 million/mL in fresh medium, cultured overnight to a density of 1 million cells/mL, and then transfected with a DNA/polyethyleneimine mixture. For example, for a 0.5 L culture, 0.5 mg plasmid DNA and 1 01L of 1 g/L linear polyethyl eneimine (MW 25 kDa, Polysciences), pH 7.0 were separately diluted in 10 mL OptiPro SFM (Thermo Fisher Scientific) and incubated for 15 minutes. The DNA and PEI dilutions were then gently mixed and incubated for an additional 15 minutes, and finally added dropwise to the cell culture while mixing.

Cultures were incubated for 5-6 days, after which the cells were removed by

centrifugation and the desired protein in the supernatant was purified by nickel -NT A (Quigen) affinity chromatography followed by size exclusion chromatography (Superdex 75, 10/300 GL, GE Healthcare) following the manufacturers protocols. The eluted protein was kept in storage buffer (lx PBS, pH 7.4, 0.1% BSA, and 10% glycerol) at concentrations of 10 mM (huPAR and muP AR-F LAG) or 100 mM (muPAR) at -80°C.

3.6.3 LIBRARY CREATION

The genes encoding human and murine GFD, SMB, and the GFD-SMB fusion proteins were generated by PCR-diiven overlap extension and used as templates for error-prone PCR. Mutations were introduced by PCR amplification of the genes in the presence of the dNTP analogs 8-oxo-dGTP and dPTP (TriLink Biotech) using low- fidelity Taq polymerase (New England Biolabs). Five separate PCR reactions were used with varying analog concentrations and PCR cycles to generate a range of mutation frequency: ten cycles with (400 pM, 200 pM, 100 pM), and twenty cycles with (80 pM, 40 pM). The DNA libraries were sequenced and those with the desired level of mutation (-2-5%) were combined and used to generate yeast-displayed protein libraries. The combined DNA libraries were PCR amplified absent dNTP analogs using primers that generate 50-base pair overlapping regions upstream and downstream of the Nhel and BamHI cloning sites in the yeast display pCTcon2 vector. The pCTcon2 vector was linearized by double digestion with Nhel and BamHI-HF (New England Biolabs), and the PCR amplification and double digestion products were purified by agarose gel electrophoreses on a 1% gel, extracted using a Gene Jet kit (Fisher Scientific), ethanol precipitated, and resuspended in sterile nuclease-free water. Purified mutant DNA "insert" and linear plasmid were mixed at a 4: 1 weight ratio (12 pg insert: 4 pg plasmid) and electroporated into EBY100 yeast following the protocol by Benetuil L. ei a/. ²⁶ Four separate electroporations were performed for per library, and each were recovered in 8 mL of a 1 : 1 mixture of 1 M sorbitol :YPD shaking in an incubator at 225 RPM and 30°C for 1 hour. The library was then propagated in selective SD-CAA medium to dilute out untransfected cells, and expression of surface-displayed protein variants was induced by switching to SG-CAA medium. Library sizes were estimated by dilution plating on selective SD-CAA plates.

3.6.4 LIBRARY SCREENING

Yeast cells displaying variants with a range of affinities for uPAR were stratified by their equilibrium dissociation constants (K _d ) or their kinetic off-rate constants (koff) and then sorted by FACS to isolate variants with superior affinities. For equilibrium binding sorts, yeast libraries were incubated at room temperature gently mixing with the indicated concentrations of uPAR in PBS A buffer for the necessary time required to reach 95-99% equilibrium that was calculated using the formula t = (£ _on [L]o + / _f ) ¹ , where t is the equilibrium time constant; k _oa and k _{0t F} are the on and off rates of binding, respectively, and [L]o is the initial concentration of ligand in solution ²⁷ At 3t and 4.6t the reactions have reached approximately 95% and 99% equilibrium, respectively. The binding reaction volumes were adjusted to prevent [L]o from changing significantly throughout the reaction (i.e. to prevent ligand depletion). After reaching > 95% equilibrium, cells were pelleted by centrifugation, excess supernatant was removed, and a chicken anti-cMyc antibody was added at a dilution of 1 :250 and mixed with the cells for 1 hour at room temperature. The concentration of uPAR remained relatively unchanged during this step. Yeast cells were then washed with cold PBS A and kept on ice for the remainder of the procedure. Cells were pelleted and resuspended with secondary antibodies (anti -chi cken-PE and anti-6histidine-488 for reactions with huPAR, and anti chicken-488 and anti-FLAG-PE for reactions with muPAR) at a 1 : 100 dilution in PBS A and mixed for 30 minutes at 4°C, washed with cold PBS A, pelleted, and immediately sorted by FACS.

For kinetic off-rate sorts, cells were incubated with the indicated concentrations of uPAR required to saturate high-affmity variants for 3 hours at room temperature, after which cells were washed 3 times to remove unbound uPAR, resuspended in PBS A containing a molar excess of either muPAR or hGFD-Fc for the mouse or human uPAR ligand libraries, respectively, and incubated for the indicated times at room temperature. During the last hour, a 1 :250 dilution of chicken anti-c-myc was added to the reaction. Cells were then pelleted, washed, and labeled with secondary antibodies as described for the equilibrium binding sorts.

The labeled yeast libraries were sorted using a BD Aria FACS instrument (Stanford FACS Core Facility) and FACSDiva software (BD biosciences). For the first sort, the number of cells sorted was between l-10x the estimated library diversity, depending on the library size and available instrument time. After the first sort, the number of cells sorted was at least lOx the estimated library diversity to ensure adequate sampling of the library diversity. Collected cells were propagated in SD-CAA and induced in SG-CAA as previously described, and subjected to additional rounds of sorting. Following sorts 4, 5, and/or 6 of each library, plasmid DNA was extracted from the library using a Zymoprep kit (Zymo Research Corp.), transformed into DHlOb electrocompetent cells, isolated using a plasmid miniprep kit (Qiagen), and sequenced by MC Lab.

3.6.5 EQUILIBRIUM BINDING TITRATIONS OF YEAST-DISPLAYED

PROTEINS

Clonal cultures of yeast cells expressing a single protein variant were grown and induced following the same procedure used for the protein libraries. Approximately 50,000 cells were incubated in PBSA with varying concentrations of uPAR ranging from 500 fM to 200 nM for the required time to reach > 95% binding equilibrium (see section 3.6.4) in volumes sufficient to avoid >10% ligand depletion Immunofluorescent labeling was performed as described for library sorting, and cell fluorescence was analyzed using a C6 Accuri flow cytometer (BD Biosciences). Data were analyzed using Flow Jo software (Tree star Inc.) or BD Accuri C6 software (BD Biosciences) Binding data from triplicate experiments were fit to a four-parameter variable slope sigmoidal curve using Prism 7 software (GraphPad) to calculate equilibrium dissociation constants.

Table 1 Assembled sequences for the engineered *mGFD-*mSMB (mGS) and *hGFD-*hSMB (hGS) bispecific uPAR ligands

3.9 REFERENCES

1. Ghamande, S. A. et al. A phase 2, randomized, double-blind, placebo- controlled trial of clinical activity and safety of subcutaneous Aό in women with asymptomatic CA125 progression after first-line chemotherapy of epithelial ovarian cancer. Gynecol. Oncol. Il l, 89-94 (2008).

2. Gold, M. A. et al. A phase II study of a urokinase-derived peptide (A6) in the treatment of persistent or recurrent epithelial ovarian, fal lopian tube, or primary peritoneal carcinoma: a Gynecologic Oncology Group study. Gynecol. Oncol. 125, 635-9 (2012).

3. Byrne, H., Conroy, P. J., Whisstock, J. C. & O’Kennedy, R. J. A tale of two specificities: bi specific antibodies for therapeutic and diagnostic applications Trends Biotechnol. 31, 621-632 (2013). 4. May, C., Sapra, P. & Gerber, H.-P. Advances in bi specific biotherapeutics for the treatment of cancer. Biochem. Pharmacol. 84, 1 105-1112 (2012).

5. Klein, C. et al. Progress in overcoming the chain association issue in bispecific heterodimeric IgG antibodies. MAbs 4, 653-663 (2012).

6. Kontermann, R. Dual targeting strategies with bispecific antibodies. MAbs 4, 182-197 (2012).

7. Gardsvoll, H & Ploug, M. Mapping of the vitronectin-binding site on the urokinase receptor: involvement of a coherent receptor interface consisting of residues from both domain I and the flanking interdomain linker region. J. Biol. Chem. 282, 13561-72 (2007).

8. Taaii, T. et al Direct interaction of the kringle domain of urokinase-type plasminogen activator (uPA) and integrin alpha v beta 3 induces signal transduction and enhances plasminogen activation. Thromb. Haemost. 95, 524- 34 (2006).

9. Kwak, S .-H. et al. The kringle domain of urokinase-type plasminogen activator potentiates LPS-induced neutrophil activation through interaction with (alpha}V(beta)3 integrins. Leukoc. Biol. 78, 937-45 (2005).

10. Liang, O. D . et al. Binding of urokinase plasminogen activator to gpl30 via a putative urokinase-binding consensus sequence. Biol. Chem. 384, 229-36 (2003).

1 1. Degryse, B., Sier, C. F., Resnati, M , Conese, M. & Blasi, F. PAI-1 inhibits urokinase-induced chemotaxis by internalizing the urokinase receptor. FEBSLett. 505, 249-54 (2001).

12. Bu, X et al. Species-specific urokinase receptor ligands reduce glioma growth and increase survival primarily by an anti angiogenesis mechanism. Lab. Invest. 84, 667-78 (2004).

13. Min, H. Y. et al. Urokinase receptor antagonists inhibit angiogenesis and primary tumor growth in syngeneic mice. Cancer Res. 56, 2428-33 (1996). 14. Gardsvoll, H et al. Conformational regulation of urokinase receptor function: impact of receptor occupancy and epitope-mapped monoclonal antibodies on lameliipodia induction. J. Biol. Chem. 286, 33544-56 (201 1).

15. Eden, G., Archinti, M., Furlan, F , Murphy, R. & Degryse, B. The urokinase receptor interactome. Curr. Pharm. Des. 17, 1874-89 (2011).

16. Liu, D., Aguirre Ghiso, J., Estrada, Y. & Ossowski, L. EGFR is a transducer of the urokinase receptor initiated signal that is required for in vivo growth of a human carcinoma. Cancer Cell 1, 445-57 (2002).

17. Sidenius, N. & Blasi, F. Domain 1 of the urokinase receptor (uPAR) is required for uPAR-mediated cell binding to vitronectin. FEES Lett. 470, 40-6 (2000).

18. Quax, P. H. et al. Binding of human urokinase-type plasminogen activator to its receptor: residues involved in species specificity and binding. Arterioscler . Thromb. Vase. Biol. 18, 693-701 (1998).

19. Huai, Q et al. Crystal structures of two human vitronectin, urokinase and urokinase receptor complexes. Nat. Struct. Mol. Biol. 15, 422-3 (2008).

20. Pettersen, E. F. et al UCSF Chimera— a visualization system for exploratory research and analysis. J. Comput. Chem. 25, 1605-12 (2004).

21. Deng, G., Royle, G., Wang, S., Crain, K. & Loskutoff, D. J Structural and functional analysis of the plasminogen activator inhibitor- 1 binding motif in the somatomedin B domain of vitronectin. J. Biol. Chem. 271, 12716-23 (1996).

22. Deng, G., Curriden, S. A., Wang, S., Rosenberg, S. & Loskutoff, D. J. Is plasminogen activator inhibitor- 1 the molecular switch that governs urokinase receptor-mediated cell adhesion and release? J. Cell Biol. 134, 1563-71 (1996).

23. Okumura, Y. et al. Kinetic Analysis of the Interaction between Vitronectin and the Urokinase Receptor. J. Biol. Chem. 277, 9395-9404 (2002)

24. Traxlmayr, M. W. & Obinger, C. Directed evolution of proteins for increased stability and expression using yeast display. Arch. Biochem. Biophys. 526, 174- 80 (2012). 25. Zhao, H. & Zha, W. In vitro‘sexual’ evolution through the PCR-based staggered extension process (StEP). Nat. Protoc. 1, 1865-1871 (2006).

26. Benatuil, L., Perez, J. M., Bel , J. & FIsieh, C.-M. An improved yeast transformation method for the generation of very large human antibody libraries. Protein Eng. Des. Sel. 23, 155-159 (2010).

27. Van Deventer, J. A. & Wittrup, K. D Yeast surface display for antibody isolation: library construction, library screening, and affinity maturation. Methods Mol. Biol. 1131, 151-81 (2014).

28. Lin, L., Gardsvoll, H., Huai, Q., Huang, M. & Ploug, M. Structure-based engineering of species selectivity in the interaction between urokinase and its receptor: implication for preclinical cancer therapy. J. Biol. Chem. 285, 10982- 92 (2010).

Example 2

Evaluating the Urokinase Receptor Ligands in Vitro and in Vivo in Animal Models

of Cancer

4.2 INTRODUCTION

The urokinase receptor has been a target for therapeutic intervention of cancer metastasis since its discovery in 1988 ¹ . To date, countless molecules have been developed to antagonize the cancer-promoting functions of the receptor, including small molecules, linear peptides, cyclic peptides, proteins, and antibodies, though these have failed to reach the clinic because of their insufficient efficacy in vivo, owing to their inability to effectively antagonize the urokinase receptor. In Example 1, we outlined a strategy for creating an improved uPAR ligand that has the potential to overcome these shortfalls. Using yeast display, we generated bispecific ligands of murine and human uPAR comprising the GFD domain of urokinase and the SMB domain of vitronectin genetically fused by a 25 -amino acid [Gly- Serjs linker (mGFD-mSMB and hGFD-hSMB, respectively). The bi specific ligands simultaneously engage the urokinase and vitronectin binding sites of uPAR with an apparent affinity approximately 8 -fold higher than wild- type urokinase due to the avidity effects of multivalent binding. Subsequently, the affinities of the individual GFD and linker-SMB domains were enhanced using directed evolution; mGFD, and the linker-mSMB domains were improved 235 -fold and 6-fold, respectively, and the hGFD and linker-hSMB regions were improved 30-fold and 10- fold, respectively. Combining the evolved domains is expected to generate bispecific ligands with over 1,000-fold higher affinity for uPAR relative to the native ligands. The high affinity and multi -functionality of the engineered proteins thus have the potential to overcome the limitations of previously developed ligands.

Here we describe further work with the engineered uPAR ligands, which were expressed, purified, and characterized in vitro and in vivo for their ability to localize to tumors and inhibit growth and experimental metastasis. Given their small size (~12 kDa), these proteins are expected to have a serum half-life of only 10-20 minutes in vivo, leaving a limited window of therapeutic efficacy. Therefore, we expressed the affinity matured *mGFD-*mSMB and *hGFD-*hSMB proteins as fusions to the Fc domain of an antibody (henceforth referred to as hGS-Fc and mGS-Fc, respectively) to increase their molecular weight above the renal clearance cutoff of 60 kDa, and as a result extend serum half-life to approximately 24 hours in vivo. The human and murine GFD domains were also expressed as Fc fusion proteins (hGFD-Fc and mGFD-Fc, respectively) to include as positive controls and for comparison in these studies. We show the hGS-Fc and mGS-Fc proteins bind to multiple uPAR-expressing cancer cell lines and inhibit the uPA-uPAR binding interaction. Additionally, the hGS-Fc protein demonstrates increased tumor targeting and localization in vivo compared to the hGFD-Fc control, and is also shown to internalize into cancer cells in vitro via a uPAR-dependent mechanism, thus offering exciting potential for delivery of cytotoxic agents specifically to cancer cells. 4.3 RESULTS

4.3.1 REFORMATTING THE UPAR LIGANDS INTO FC FUSIONS

For the bispecific uPAR ligands to be effective in vivo, their concentration in the blood must be maintained at or above the level required to sufficiently inhibit uPAR function. This is problematic given the 12 kDa proteins will rapidly filter out of the blood through the kidneys with a circulating half-life of only 10-20 minutes, necessitating frequent and/or high doses to be effective. Furthermore, the proteins must be delivered intravenously or subcutaneously due to their instability in the denaturing and proteolytic environment of the gastrointestinal tract. Therefore, to avoid expensive and impractical dosing requirements, we reformatted the bi specific ligands as fusions to the fragment crystal lizable (Fc) region of the mouse IgG2a antibody. The resulting cysteine-linked dimers have molecular weights of 77 kDa for the hGS-Fc and mGS-Fc proteins, and 64 kDa for hGFD-Fc and mGFD-Fc, which are above the 60 kDa renal clearance cutoff and thus have extended serum half-lives of approximately 24 hours in vivo.

The Fc-fusion proteins were recombinantly expressed by transient transfection of suspension Freestyle 293 cells and purified by protein A affinity chromatography, followed by size exclusion chromatography to remove undesired oligomers of the Fc domain (FIG. 37). The purified products were subsequently analyzed by polyacrylamide gel electrophoresis; they were of high purity and ran at the expected homodimer mol ecular weights of approximately 64 kDa and 77 kDa for the GFD-Fc and GS-Fc proteins, respectively (FIG. 38). The final protein yields from cell culture were 11.5 mg/L, 7.5 mg/1,, 7 mg/L and 4 mg/L for hGFD-Fc, hGS-Fc, mGFD-Fc, and mGS-Fc, respectively.

4.3.2 INHIBITION OF UPA-UPAR BINDING IN VITRO BY THE HGS-FC

AND MGS-FC

We next verified the engineered hGS-Fc and mGS-Fc proteins are capable of inhibiting the uPA-uPAR binding interaction. Since uPA-uPAR binding is coordinated by the growth factor domain (GFD) of uPA, we displayed GFD on the surface of yeast cells as a fusion to the aga2p coat protein and measured its binding to soluble uPAR in the presence of the hGS-Fc and mGS-Fc proteins. The wild-type hGFD-Fc and mGFD-Fc proteins were also included as a positive control. All Fc fusion proteins successfully inhibited tire GFD-uPAR interaction in a typical dose-dependent manner (FIG. 39). The assay used required a minimum of 2 nM uPAR to obtain a signal with sufficient dynamic range to measure inhibition of uPAR binding. Since 2 nM uPAR is orders of magnitude above the K _d of the engineered proteins, the binding reaction between hGS-Fc and huPAR, and mGS-Fc and muPAR were concentration-driven, rather than K _d -driven, and thus accurate values for the half-maximal inhibitory concentrations (ICso) could not be determined. Regardless, these results qualitatively indicate the engineered bispecific uPAR ligands hGS-Fc and mGS-Fc inhibit the uPA-uPAR binding interaction.

4.3.3 BINDING OF MGS-FC AND HGS-FC TO UPAR EXPRESSED ON

THE SURFACE OF VARIOUS CANCER CELL LINES

In preparation for testing the ability of the bispecific uPAR ligands to localize to tumors and inhibit growth and metastasis in vivo, we first confirmed the proteins bind to uPAR expressed on the surface of multiple types of cancer cells. Adherent cancer cells were grown in tissue culture flasks, harvested by non-enzymatic dissociation to maintain the integrity of the surface proteins, and incubated with 100 nM of the hGS-Fc and mGS- Fc proteins. Binding was detected by flow cytometry following immunostaining with a fluorophore-labeled anti-Fc antibody. The hGS-Fc protein bound to uPAR on the surface of all cell lines tested, including the human breast (MDA-MB-231), brain (U-87-MG), and ovarian (SKOV-3) cancer cell lines (data not shown). Additionally, the hGS-Fc protein also bound to huPAR on the surface of human umbilical vein endothelial cells (HUVEC), which is expected given the pivotal role uPAR plays in angiogenesis ^2,3 . The mGS-Fc protein bound to two of the three murine cancer cell lines tested: murine breast cancer (4T1) and colon adenocarcinoma (MC-38) cell lines, but not the murine skin melanoma (B16-F10) cell line (data not shown). These data are consistent with the finding that uPAR is overexpressed in numerous types of cancer ⁴ , and validate use of the uP AR-positive lines to assess the efficacy of mGS-Fc and hGS-Fc in vivo. 4.3,4 LOCALIZATION OF HGS-FC AND HGFD-FC TO HUMAN

BREAST CANCER TUMOR XENOGRAFTS IN VIVO

We first verified the engineered uPAR ligands localize specifically to uPAR- expressing tumors in vivo. Aggressive human breast (MDA-MB-231) tumors that express high levels of uPAR were implanted into the left shoulders of athymic nude mice.

Additionally, human embryonic kidney (HEK 293T) tumors that do not express uPAR were implanted into the right shoulder as a negative control for uPAR-specific tumor localization. After tumors reached ~10 mm in diameter, mice were injected with 1.5 nmol of either Alex a 680 fluorophore-labeled hGFD-Fc (hGFD-Fc-680) or hGS-Fc (hGS-Fc- 680), and localization of the labeled proteins was tracked via whole-body fluorescence imaging using an IVIS Lumina III in vivo imaging system. Compared to the wild-type hGFD-Fc-680 control, the hGS-Fc-680 protein demonstrated enhanced tumor localization that persisted for over 3 weeks after injection (FIG. 40). To quantify uPAR-specific tumor localization, the fluorescence signal from the uPAR-positive breast cancer tumor was divided by the signal from the uPAR-negative tumor and plotted over the course of the experiment (FIG. 41). The hGS-Fc-680 protein exhibited substantially higher uPAR- positive tumor specificity. These results were consistent when the experiment was repeated in two additional pairs of mice (FIG. 42). The engineered hGS-Fc fusion protein thus demonstrates enhanced tumor targeting and localization in vivo compared to the wild-type hGFD-Fc control. Given GFD-Fc has been shown to effectively inhibit cancer growth and metastasis in vivo ^{5 7} , these results suggest the engineered hGS-Fc protein may demonstrate more significant efficacy in vivo.

In addition to testing uPAR-specific tumor localization of the Fc-fusion proteins in vivo, we also tested localization of the monomeric hGFD and hGS proteins. Given their small size of 6 and 12 kDa, respectively, the injected proteins that remained unbound to cell-surface uPAR were expected to rapidly clear from the blood though the kidneys with a circulating serum half-life of only 10-to-20 minutes. Although longer serum half-lives are typically desired for cell -surface receptor ligands to be effective, the rapid clearance of these molecules may be advantageous for other applicati ons such as cancer diagnostics or cancer-specific cytotoxic drug delivery, where clearance of unbound protein increases contrast for tumor detection, or reduces side effects from off- target toxicity. As with the Fc-fusion proteins, 1.5 nmol of Alexa 680-1 abeled hGFD (hGFD-680) and hGS (hGS-680) were injected into mice with uPAR-positive MDA-MB- 231 and uPAR-negative HEK 293T tumors in their left and right shoulders, respectively. As expected, both proteins rapidly cleared from circulation through the kidneys (FIG.

43). However, to our surprise, neither hGFD-680 nor hGS-680 showed detectable uPAR- specific tumor localization. The 680-1 abeled proteins showed potent inhibition of uPA- uPAR binding in vitro (data not shown), indicating their activity was not compromised prior to injection. These results suggest the increased circulating serum half-life afforded by fusion to the Fc domain of an antibody is required for detectable levels of uPAR- specific tumor accumulation lndeed, the hGFD-Fc-680 and hGS-Fc-680 proteins only show uPAR-specific tumor accumulation higher than background signal between four and five days after injection (FIG. 40).

4.3.4 UPAR-DEPENDENT INTERNALIZATION OF HGFD-FC AND HGS-

FC INTO CANCER CELLS

Given the prolonged tumor accumulation of hGS-Fc-680 observed in vivo , we suspected the protein may be internalizing into cancer cells following uPAR binding. The hGS-Fc protein was actually designed to inhibit active uPAR internalization; after formation of the uPA-uPAR complex, pl asminogen activator inhibitor 1 (P AI- 1 ) binds to the protease domain of uPA, and PAI-1 is then bound by the low-density lipoprotein receptor-related protein (LRP1) ⁸ , which drives internalization of the entire protein assembly ^9,10 , degradation of PAI-1 -uP A in the lysosome ⁹ , and recycling of uPAR back to the cell surface ¹¹ . The absence of the uPA protease domain in the hGS-Fc protein should therefore inhibit this process of internalization, which has been shown to contribute to cancer migration and metastasis ¹² . However, the hGS-Fc-uPAR complex may still be internalized passively given uPAR’s localization at focal adhesion sites ¹³ , where various adhesion receptors such as integrins are constantly internalized and recycled to the cell surface ¹⁴ . Intemalization of the hGFD-Fc and hGS-Fc proteins was confirmed in vitro. Human breast cancer (MDA-MB-231) cells were incubated with Alexa 488-labeled hGS- Fc (hGS-Fc -488) and hGFD-Fc (hGFD-Fc-488) for 15 hours, harvested by non- enzymatic cell dissociation, and incubated with an anti -Alexa 488 antibody that quenches extracellular Alexa 488 fluorescence (FIG. 44 A). The cells were then analyzed by flow cytometry to detect the presence of intracellular hGFD-Fc-488 or hGS-Fc-488. Both proteins internalized into the cancer cells (FIG. 44B). Interestingly, the fluorescent signal for internalization of 0.2 nM and 2 nM hGFD-Fc-488 was 3.79 ± 0.17 and 19.74 ± 0.27 a.u., respectively, whereas the signal for internalization of hGS-Fc-488 was significantly higher at both concentrations (Mean ± SEM = 10 84 ± 0.35, p value < 0.0001, and 41.88 ± 1.11 a.u., p value = 0.0016, respectively). Given both proteins were conjugated with equal amounts of fluorophore (approximately 0.5 Alexa 488 fluorophores per protein), this suggests the hGS-Fc protein binds uPAR and/or internalizes more effectively than the wild-type hGFD-Fc protein. Internalization was dependent on incubation time (FIG. 44C), and on uPAR binding (FIG. 44B), as a 100-fold molar excess of unlabeled hGFD-

Fc or hGS-Fc protein blocked internalization of hGFD-Fc-488 and hGS-Fc-488, respectively. The results from this assay and the uPAR-specificity of hGS-Fc in vivo suggest the engineered protein may also be an effective vehicle for intracellular delivery of cytotoxic agents specifically to cancer cells in vivo, which may further increase the efficacy of the hGS-Fc protein for inhibiting cancer growth and metastasis.

Table 2 Amino acid sequences of the engineered uPAR ligands fused to the Fc domain of the mouse IgG2a antibody or human IgGl antibody (bold). The human and murine GFD sequences are preceded by the native 20-amino acid signal peptide for human and murine uPA, respectively (underlined).

4.8 REFERENCES

1. Nielsen, L. S. el al. A 55,000-60,000 Mr receptor protein for urokinase- type plasminogen activator. Identification in human tumor cell lines and partial purification. J. Biol. Chem. 263, 2358-63 (1988).

2. Montuori, N. & Ragno, P in Chemical immunology and allergy 99, 1 OS- 122 (2013).

3. Herkenne, S. et al. The interaction of uPAR with VEGFR2 promotes VEGF-induced angiogenesis. Sci. Signal. 8, ral l7-ral l7 (2015).

4. de Bock, C. E & Wang, Y. Clinical significance of urokinase-type plasminogen activator receptor (uPAR) expression in cancer. Med. Res. Rev. 24, 13-39 (2004).

5. Hu, X.-W. et al. Inhibition of tumor growth and metastasis by ATF-Fc, an engineered antibody targeting urokinase receptor. Cancer Biol. Ther. 7, 651-9

(2008).

6. Ignar, D. M. et al. Inhibition of establishment of primary and

micrometastatic tumors by a urokinase plasminogen activator receptor antagonist. Clin. Exp. Metastasis 16, 9-20 (1998).

7. Zhou, H. et al. Synergistic inhibitory' effects of an engineered antibody like molecule ATF-Fc and trastuzumab on tumor growth and invasion in a human breast cancer xenograft mouse model. Oncol. Lett. (2017).

doi: 10.3892/ol.2017.6896 8. Nykjaer, A. et al. Purified alpha 2-macroglobulin receptor/LDL receptor- related protein binds urokinase plasminogen activator inhibitor type-l complex Evidence that the alpha 2 -macroglobulin receptor mediates cellular degradation of urokinase receptor-bound complexes. J. Biol. Chem. 267, 14543-6 (1992).

9. Cubeilis, M. V, Wun, T. C. & Blasi, F. Receptor-mediated internalization and degradation of urokinase is caused by its specific inhibitor PAI-1. EMBO ,/. 9, 1079-85 (1990).

10. Olson, D. et al. Internalization of the urokinase-plasminogen activator inhibitor type-1 complex is mediated by the urokinase receptor. J. Biol. Chem.

267, 9129- 33 (1992).

11. Nykjar, A. et al. Recycling of the urokinase receptor upon internalization of the uPA:serpin complexes. EMBO J. 16, 2610-2620 (1997).

12. Chazaud, B. et al. Promigratory Effect of Plasminogen Activator

Inhibitor-! on Invasive Breast Cancer Cell Populations. Am. J. Pathol. 160, 237- 246 (2002).

13. Salasznyk, R. M. et al. The uPA receptor and the somatomedin B region of vitronectin direct the localization of uPA to focal adhesions in microvessel endothelial cells. Matrix Biol. 26, 359-370 (2007).

14. Hiilsbusch, N., Solis, G. P., Katanaev, V. L. & Stuermer, C. A. O. Reggie- l/Flotillin-2 regulates integrin trafficking and focal adhesion turnover via Rabl la. Eur. J. Cell Biol 94, 531-545 (2015).

While the preferred embodiments of the invention have been illustrated and described, it will be appreciated that various changes can be made therein without departing from the spirit and scope of the invention.