Single-chain fab proteins and uses thereof

Abstract

The present invention relates to the field of protein engineering and therapeutics, in particular, improved protein formats and uses thereof.

Description

10024-W001-SECSINGLE-CHAIN FAB PROTEINS AND USES THEREOFCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Application No.63 / 699,367, filed September 26, 2024.DESCRIPTION OF THE TEXT FILE SUBMITTED ELECTRONICALLY

[0002] The present application contains a Sequence Listing, which has been submitted electronically in XML format and is hereby incorporated by reference in its entirety. The computer readable format copy of the Sequence Listing, which was created on August 29, 2025, is named 10024-W001-SEC_ST26.xml and is 18,645 bytes in size.FIELD OF THE INVENTION

[0003] The present invention relates to the field of protein engineering and therapeutics, in particular, improved protein formats and uses thereof.BACKGROUND OF THE INVENTION

[0004] Multi-specific antibodies have shown great promise as the next generation of therapeutic drugs against cancers, infections, and immunological disorders. As of December 2023, the U.S. Food and Drug Administration approved 11 bispecifics in therapeutic areas that cover Cardiovascular / hemostasis (Emicizumab), Ophthalmology (Faricimab), and 9 bispecifics approved for Cancer (Lim et. al, Clinical Pharmacology & Therapeutics. 2024;116:315-327; Yoneyama et. al, Expert Review of Clinical Pharmacology. 2023;16:775-790; and Sharma et. al, Expert Opinion on Biological Therapy. 2024;24:263-268). Every year there are increasing number of bispecifics at various stages of clinical trials with well over 100 bispecifics (Huang et. al, Journal of Cancer Research and Clinical Oncology. 2020). These multi-specifics range in formats from single chain Fv (scFv), such as BiTE® molecules, to antibody fragments linked to IgG chains. Examples include DVD-IgG and Fab-IgG (Brinkmann and Kontermann, MAbs. 2017;9: 182-212).

[0005] There have been advances in optimization and characterization of scFvs as therapeutics, however, not much effort has been made to advance single-chain Fabs (scFabs).10024-W001-SECHere, we report on the biophysical and biochemical properties of scFab-scFc constructs with varying linker compositions and scFab geometries, and the characterization of a single chain Fab - single chain Fc protein.SUMMARY OF THE INVENTION

[0006] The present disclosure provides a single-chain protein comprising at least one single-chain Fab (scFab) connected to a single chain Fc (scFc). In an embodiment, the singlechain protein comprises an scFab connected by an scFab N-terminus or an scFab C-terminus to an N-terminus of the scFc. In an embodiment, the single-chain protein comprises an scFab connected by an scFab N-terminus or an scFab C-terminus to a C-terminus of the scFc. In an embodiment, the scFab is connected by the scFab N-terminus to the scFc. In an embodiment, the scFab is connected by the scFab C-terminus to the scFc. In an embodiment, the single-chain protein comprises (i) a first scFab connected by an scFab N-terminus or an scFab C-terminus to an N-terminus of the scFc; and (ii) a second scFab connected by an scFab N-terminus or an scFab C-terminus to a C-terminus of the scFc. In an embodiment, the first and second scFab each bind to a same target. In an embodiment, the first and second scFab each bind to a different target.

[0007] The present disclosure provides a single-chain protein comprising at least three scFabs, wherein each scFab is connected to at least one other scFab. In an embodiment, the at least three scFabs each bind a same target. In an embodiment, the at least three scFabs each bind a different target. In an embodiment, at least one scFab binds a target, and at least one other scFab binds a different target. In an embodiment, two scFabs bind two different epitopes of the same target.

[0008] In an embodiment, the scFab has an orientation in the following order, from N- terminus to C-terminus, VL, either CK or CX, VH, and CHI. In an embodiment, the scFab has an orientation in the following order, from N-terminus to C-terminus, VH, CHI, VL, and either CK or CX. In an embodiment, the scFab comprises CK. In an embodiment, the scFab comprises CX. In an embodiment, the scFab comprises a linker, and the linker is (G4S)6, (G4S)7, (G4S)8, (G4E)7, (G4Q)7, (G3SE)7, (G4SG4Q)3(G4S), or (G4Q)3(G4S)4. In an embodiment, the scFab linker is (G4S)6. In an embodiment, the scFab linker is (G4S)7. In an embodiment, the scFab linker is (G4S)8. In an embodiment, the scFab linker is (G4E)7. In an embodiment, the scFab linker is (G4Q)7. In an embodiment, the scFab linker is (G3SE)7. In an embodiment, the scFab10024-W001-SEC linker is (G4SG4Q)3(G4S). In an embodiment, the scFab linker is (G4Q)3(G4S)4. In an embodiment, the scFab linker is (G4S)6, (G4S)7, (G4S)8, (G4E)7, (G4Q)7, (G3SE)7, (G4SG4Q)3(G4S), or (G4Q)3(G4S)4. In an embodiment, the scFab linker is (G4S)6, (G4S)7, (G4S)8, (G4E)7, (G4Q)7, (G3SE)7, (G4SG4Q)3(G4S), or (G4Q)3(G4S)4.

[0009] In an embodiment, the scFab CHI, CK and / or CX domains are IgG, IgM, IgA, IgD, or IgE. In an embodiment, the domains are IgG. In an embodiment, the domains are IgGl .

[0010] In an embodiment, the scFab contains a cysteine clamp between CHI and either CK or CX.

[0011] In an embodiment, the scFc comprises a linker, and the linker is (G4S)6, (G4E)6, (G4Q)6, (G4Q)3(G4S)3, or (G4QG4S)3, G3SE)6. In an embodiment, the scFc linker is (G4S)6. In an embodiment, the scFc linker is (G4S)6. In an embodiment, the scFc linker is (G4E)6. In an embodiment, the scFc linker is (G4Q)6. In an embodiment, the scFc linker is (G4Q)3(G4S)3. In an embodiment, the scFc linker is (G4QG4S)3. In an embodiment, the scFc linker is G3SE)6.

[0012] In an embodiment, the scFc comprises a linker, and the linker is an FcRn binding peptide shown in CONI 7. In an embodiment, the scFab is connected to the scFc by a G4S linker.

[0013] The present disclosure provides a method of treating a patient comprising administering to the patient an effective amount of a single-chain protein of the present invention. In an embodiment, the patient has cancer. In an embodiment, the cancer is a solid tumor. In an embodiment, the cancer is non-small cell lung cancer, gastric cancer, head and neck squamous cell carcinoma, hepatocellular carcinoma, triple-negative breast cancer, colorectal cancer, pancreatic cancer, or metastatic castrate-resistant prostate cancer. In an embodiment, the patient has an auto-immune disease. In an embodiment, the patient has a cardiometabolic disease.

[0014] The present disclosure provides a single-chain protein of the present invention for use in treating a disease. In an embodiment, the disease is cancer. In an embodiment, the disease is cancer. In an embodiment, the cancer is a solid tumor. In an embodiment, the cancer is non-small cell lung cancer, gastric cancer, head and neck squamous cell carcinoma, hepatocellular carcinoma, triple-negative breast cancer, colorectal cancer, pancreatic cancer, or metastatic castrate-resistant prostate cancer. In an embodiment, the disease is an auto-immune disease. In an embodiment, the disease is a cardiometabolic disease.10024-W001-SEC

[0015] The present disclosure provides use of a single-chain protein of the present invention in the manufacture of a medicament for treating a disease. In an embodiment, the disease is cancer. In an embodiment, the cancer is a solid tumor. In an embodiment, the cancer is non-small cell lung cancer, gastric cancer, head and neck squamous cell carcinoma, hepatocellular carcinoma, triple-negative breast cancer, colorectal cancer, pancreatic cancer, or metastatic castrate-resistant prostate cancer. In an embodiment, the disease is an auto-immune disease. In an embodiment, the disease is a cardiometabolic disease.BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 depicts on-cell binding of each construct to Human TAA1 transiently expressed on 293T cells is reported as fold increase in TAA1 Geomean over an irrelevant protein (TAA3) Geomean. All constructs, except for Con20, were tested at lOOnM. Con20 was tested at 43.5nM.

[0017] Figure 2 depicts dose response curve for TAA1 activity measured as phosphor- ERK1 / 2 activation. Each construct and antibody were titrated 3-fold starting from 33 nM. Monovalent scFab-scFc, Con7 and Con8, constructs showed activity comparable to media control. Bi-valent scFab-scFc and parental IgG demonstrated activation of phosphor-ERK through TAA1 dimerization.

[0018] Figure 3 depicts Octet® sensorgram demonstrating binding of the bispecific scFab-scFc (Con23) to two target proteins, TAA2 and TAA1. Streptavidin biosensors were used to load a Goat Anti -Human Fc IgG, followed by loading of Con23. Sensorgram illustrates baseline step after loading of Con23, addition to solution of either TAA2, TAA1, or combination of both TAA2 and TAA1 at 100 nM. Both TAA2 and TAA1 are monomeric.Sequential binding of TAA2 and TAA1, or visa-versa, demonstrates Con23 binds both target proteins. Response signal from TAA2 and TAA1 in the same solution is approximate to the signal after sequential binding of both.

[0019] Figure 4 depicts Octet® sensorgrams demonstrating binding of the TsAb (Con25) to each of target, TAA3 (A), TAA2 (B) and TAA1 (C). Streptavidin biosensors were used to capture Biotinylated TAA3 and TAA2, while AR2G biosensors were used to directly immobilize TAA1 onto the surface. Affinities were determined using a 1 : 1 binding model. Con25 bound to all three targets, with TAA3 and TAA1 demonstrating heterogeneous binding profiles. Approximate affinities could be determined for each target.10024-W001-SEC

[0020] Figure 5 depicts schematic representation and linker description of constructs used in this study. Generally, constructs (Con) 1-21 form the scFab by connecting the heavy chain at the N-termini to the light chain at the C-termini, except Coni 1 which switches the orientation of the heavy and light chains. The scFc is connected to the scFab at the C-termini of the light chain through a GGGGS linker and is composed of two identical units of IgGl hinge- CH2-CH3 domains. Conl-12 maintain the same linker of the scFc and differ in the linker of the scFab. While Coni 3- 19 maintain the same linker of the scFab and differ in the linker of the scFc. Con20 and 21 are composed of the same linkers as Con8 but contain different TAA1 binding paratopes.

[0021] Figure 6 depicts schematic representation and linker description of differing scFab geometries, valencies, and BsAb and TsAb constructs used in this study. Con22 is similar to Con8, with the scFab and scFc modules flipped in orientation. Con23 is a monovalent BsAb to two different targets. Con24 is a bivalent version of Con8. Con25 is a monovalent TsAb to three different targets.

[0022] Figure 7 depicts representative chromatograph for Con4. The main peak at 13.59 min represents the monomer construct, while the shorter retention time peak labeled at 11.34 min represents constructs with higher molecular weight.

[0023] Figures 8A-8C depicts representative MS analysis for Con4. Total ion chromatograph illustrated two peaks (Fig. 8 A); the peak at 9.5 min did not give any mass, while the 3.1 min peak gave the MS spectrum in Fig. 8B. (Fig. 8C) Deconvolution resulted in a major peak corresponding to the expected mass (103,769.7 Da) under non-reducing conditions for Con4.

[0024] Figures 9A-D depict MS analysis for Coni under non-reducing conditions. Total ion chromatograph illustrated four peaks (Fig. 9A); the 3.0 min peak gave the MS spectrum in Fig. 9B. (Fig. 9C) Deconvolution resulted in a major peak corresponding to the expected mass (105,016.5 Da) for Coni. (Fig. 9D) A minor peak demonstrated a fraction contained a species lacking the first 18 amino acids at the N-term.

[0025] Figures 10A-C depict MS analysis for Coni under reducing conditions. Total ion chromatograph illustrated two peaks (Fig. 10 A); the 3.6 min peak gave the MS spectrum in Fig. 10B. (Fig. 10C) Deconvolution resulted in a major peak corresponding to the expected mass (105,042.7 Da) for Coni. The presence of a second peak demonstrated a fraction contained a species clipped at F253-V254.10024-W001-SEC

[0026] Figures 11 A-C depicts MS analysis for ConlO under reducing conditions. Total ion chromatograph illustrated one peak (Fig. 11 A); the 3.4 min peak gave the MS spectrum in Fig.1 IB. (Fig. 11C) Deconvolution resulted a major peak corresponding to the expected mass (104,111.2 Da) for ConlO. The presence of a second peak demonstrated a fraction contained a species clipped at S257-S258.

[0027] Figures 12A and 12B depict: (Fig. 12A) Schematic of the initial trispecific format designed to investigate whether bi-epitopic TAA2 scFabs could demonstrate TAA2- dependent TAA1 agonist activity. (Fig. 12B) Table listing the combinations of TAA2 scFabs used for each trispecific construct, with corresponding epitope bins indicated. The legend shows how bins correspond to each TAA2 domain. Most combinations involved TAA2 scFabs binding to different epitope bins within a TAA2 domain (first 20) and across TAA2 domains (last three). ScFabs were oriented with the heavy chain at the N’-terminus, linked to the light chain through a (G4S)8 linker. Each scFab and scFc were connected by a single G4S linker, and the two Fc halves were joined by a (G4S)6 linker to form a scFc.

[0028] Figure 13 depicts expression and purification of all 23 trispecific constructs. DNA encoding each construct was transfected into 40 mL of 293 T cells and expressed for five days. Proteins were purified using Protein A magnetic beads, yielding 0.2-19 mg / L.

[0029] Figures 14A and B depict: (Fig. 14A) TAA1 activity of all 23 trispecific constructs at 100 nM, tested in triplicate. Trispecific constructs showing >2.5-fold TAA1 activity over negative control were considered false positives. The plot on the right shows titration of an active TAA1 antibody (40E5) and an inactive antibody (41C2). (Fig. 14B) TAA2-dependent TAA1 activity (phospho-ERK readout) of the 15 trispecific constructs nonactive on TAA1 alone, tested in triplicate at 100 nM. Titration curves for positive (40E5), negative (41C2), and ligand (TAAL) controls are included.

[0030] Figure 15 is a shematic of three trispecific formats used to investigate TAA2- dependent TAA1 agonist activity. The TAA1 scFab was placed either between the two TAA2 scFabs or at the C’ -terminal position. Constructs incorporated unique linker sequences for simplified cloning and sequence verification:

[0031] - G1 scFab with scFab-display linker

[0032] - G2 scFab with (G4E)2-FcRn peptide-(G4E)2 linker

[0033] - R scFab with (G3 SE)6 linker10024-W001-SEC

[0034] Figure 16 depicts SDS-PAGE under non-reducing conditions of purified trispecific constructs. Fl molecules were expressed and purified in duplicate. Gels were imaged on a Gel Doc EZ Imager (Bio-Rad) and processed with Image Lab software (Version 6.1.0).

[0035] Figure 17 depicts mass spectra of trispecific constructs confirming expected molecular weights. All non-reduced samples match expected masses with many additional contaminant masses. The -17209 Da, -49291 Da, and -75800 Da masses found in multiple samples do match potential clipped forms of the expected sequence and suggest they are contaminant masses.

[0036] Figure 18 depicts titration curves of TAA1 activity for Fl and F2 trispecific constructs. No trispecific constructs displayed activity above background (threshold defined as 2.5-fold over baseline). All constructs were subsequently tested in the TAA2 / TAA1 agonist assay. Measurements were performed in quadruplicate. Positive control: 40E5; negative control: 41C2.

[0037] Figure 19 depicts titration curves of TAA2 / TAA1 activity for Fl and F2 trispecific constructs. Ranking of activity: Fl-lmer > Fl-2mer > Fl-3mer > F2-lmer > F2- 2mer. Measurements were performed in quadruplicate. TAAL and 40E5 served as positive controls; 41C2 as negative control.

[0038] Figure 20 is a schematic of a two-step cloning process for Format 0 trispecific constructs.DETAILED DESCRIPTION

[0039] The linking of the heavy and light chain domains into a single binding module is an attractive property of the scFv format. However, most scFvs possess lower stabilities than Fabs, are prone to aggregation, are relatively unstable over longer periods of time (Worn and Pliickthun, Journal of molecular biology. 2001;305:989-1010), and substantially decreased serum half-life (Ahamadi-Fesharaki et. al, Oncolytics. 2019;14:38-56). Additionally, reformatting of the scFvs to a more stable Fab or IgG scaffold can result in a reduction in affinity for the target antigen (Bird RE, Walker BW. Single chain antibody variable regions. Trends in biotechnology. 1991;9: 132-7). Conversely, Fab fragments are stable under long term storage conditions (Kramer et. al, Biosensors and Bioelectronics. 2002;17:305-13), however, the heavy and light chains are separate domains.

[0040] Advantages of scFabs include greater stability over scFvs, as well as less prone to aggregation and longer longer serum half life (Worn and Pliickthun 2001) (Ahamadi-10024-W001-SECFesharaki et al. 2019). Reformatting from Fab to scFv can result in a reduction in affinity for the target antigen (Bird and Walker 1991). When developing multispecifics, engineering strategies are required to pair the correct heavy and light chains. By using scFabs, additional purification steps can be avoided to isolate the correctly paired species (Husain and Ellerman 2018) and prevention of higher immunogenicity and development of anti-drug antibodies (AD As) in patients (Nie et al. 2020; Staton et al. 2019). The consequence for multi-specifics is that engineering strategies are required to pair the correct heavy and light chains and additional purification steps are required to isolate the correctly paired species. Moreover, engineering strategies could greatly affect immunogenicity and patients might develop anti-drug antibodies (AD As) (Husain B, Ellerman D. Expanding the Boundaries of Biotherapeutics with Bispecific Antibodies. BioDrugs. 2018;32:441-64). One possible approach is to combine Fab fragments into one single chain construct.

[0041] scFab modules have been described in phage display platform (Hust et. al, BMC Biotechnology. 2007;7:14; and Koerber et. al, Journal of molecular biology. 2015;427:576-86) and as single chain IgGs (Schirrmann et. al, MAbs. 2010;2:73-6; and Lang et. al, BioRxiV, 2019:2019.12.25.888586; and Lee et. al, Molecular immunology. 1999;36:61-71). Description on linker composition and geometry of heavy and light variable-constant domains has been limited.

[0042] When making bispecifics comprising of multiple Fab domains, they may not retain correct HC-LC pairing. Multiple species can often be expressed making purification of the correct HC-LC difficult. A scFab at any position of a bispecific can be interchanged with another while retaining proper HC / LC pairing. We demonstrate a bivalent scFab-scFc construct maintains affinity and functional activity comparable to a standard bivalent IgG. Swapping one scFab for an alternate target binding scFab, we can generate a bispecific which retains binding to both targets. Utilizing scFab and scFc as modules allows for the interchange of different target binding arms and as a single molecule which simplifies multi-specific design. We also demonstrated a variety of linkers that can be used to pair Fab or Fc fragments. ScFab-scFc constructs have similar expression yields, thermal stabilities, and aggregation propensities to one another and similar affinities as canonical IgG molecules. Furthermore, a bivalent scFab- scFc exhibited comparable functional activity to its IgG counterpart, EC50 288 pM to EC50 161 pM respectively, and when made into a bi-specific, can engage both targets simultaneously. This platform can not only serve to recapitulate the binding and functional characteristics of an IgG, but also serve as a platform for multi-specifics.10024-W001-SEC

[0043] Advantages of using scFabs include; 1) retaining HC / LC pairing, (this becomes a greater advantage when constructing a multispecifics (>2 binding domains) and avoids the use of different Charge Pair Mutations to maintain HC / LC pairing.), 2) simpler purification of the desired bispecific (less possible side products), and 3) maintaining binding geometry of HC and LC (when switching format to scFv, geometry of VH / VL could change and binding could be lost; this retention in binding is observed). Another consequence of having all domains on one DNA plasmid is that you do not have to adjust the ratios of plasmids containing different HC and LC when expressing bi / multi-specifics.

[0044] A “single-chain protein” refers to a single polypeptide chain.

[0045] A “trispecific construct” refers to a single-chain protein of the present disclosure that binds to three different places on at least one target. A trispecific construct may, for example, bind one target at two different epitopes, and another target. A trispecific construct may, for example, bind three different targets. A trispecific construct may, for example, comprise three scFabs. In an embodiment, a first scFab is connected to a second scFab, and the second scFab is connected to a third scFab. In an embodiment, the scFabs are connected by linkers. A trispecific construct may, for example, comprise three scFabs and an scFc. In an embodiment, a first scFab is connected to a second scFab, and the second scFab is connected to an scFc, and the scFc is connected to an scFab. In an embodiment, two scFabs each bind a first target at two different epitopes, and the third scFab binds a second target. In an embodiment, the first scFab and second scFab each bind a first target, and the third scFab binds a second target. In an embodiment, the first scFab and third scFab each bind a first target, and the second scFab binds a second target. In an embodiment, the second scFab and third scFab each bind a first target, and the second scFab binds a second target. A trispecific construct may be constructed in an orientation as described herein and / or shown in the figures (e.g. Figures 12A and 15). A trispecific construct may be constructed with linkers as described herein and / or shown in the figures (e.g. Figures 12A and 15).

[0046] Two scFabs may bind a target (e.g. protein) at two different epitopes. The two different epitopes may overlap. This may also be referred to as bi-epitopic. In an embodiment, two scFabs can bind the target at the same time at two different epitopes. If two scFabs can bind a target at the same time as each other, they bind two different epitopes. An epitope is the specific part of a target (such as a protein, polysaccharide, or other molecule) that is recognized and bound by an scFab. An epitope is the binding site on the target to which the scFab binds.10024-W001-SEC

[0047] A linker is a synthetic sequence of amino acid residues. Encompassed herein are linkers composed of G4S (GGGGS; SEQ ID NO: 1), G4E (GGGGE; SEQ ID NO: 2), G4Q (GGGGQ; SEQ ID NO: 3), or combinations of these (example: G4QG4S (GGGGQGGGGS; SEQ ID NO: 4)) and can be of varying repeats (e g. (G4S)3 is GGGGS GGGGS GGGGS; SEQ ID NO: 5). The linker may be, for example, a (G4S)6 (SEQ ID NO: 6), (G4S)7 (SEQ ID NO: 7), or (G4S)8 linker (SEQ ID NO: 8). A G4S linker is a linker made of amino acids GGGGS (SEQ ID NO: 9), from N-terminus to C-terminus, and may be repeated multiple times. A (G4S)4 linker, for example, means a linker comprising the following amino acids, from N- terminus to C-terminus: GGGGSGGGGSGGGGSGGGGS (SEQ ID NO: 10). Alternatively, the linker may be, for example, a (G4Q)6 (SEQ ID NO: 11), (G4Q)7 (SEQ ID NO: 12), or (G4Q)8 linker (SEQ ID NO: 13). A G4Q linker is a linker made of amino acids GGGGQ (SEQ ID NO: 14), from N-terminus to C-terminus, and may be repeated multiple times. A (G4Q)4 linker, for example, means a linker comprising the following amino acids, from N-terminus to C-terminus: GGGGQGGGGQGGGGQGGGGQ (SEQ ID NO: 15). For an scFab, the length of the linker should be sufficient to cover the distance between the C-terminus of one domain (for example the heavy chain) to the N-terminus of the corresponding domain (the light chain for example).

[0048] A “single-chain variable fragment” (“scFv”) is a fusion protein in which a VL and a VH region are joined via a linker to form a continuous protein chain wherein the linker is long enough to allow the protein chain to fold back on itself and form a monovalent antigen binding site (see, e.g., Bird et al., Science 242:423-26 (1988) and Huston et al., 1988, Proc. Natl. Acad. Sci. USA 85:5879-83 (1988)). The scFv can be arranged VH-linker-VL, or VL- linker-VH, for example. An anti-target scFv is an scFv that binds an antigen, such as a tumor antigen.

[0049] A “single-chain antigen-binding fragment” (“scFab”) is a fusion protein in which a VH and CHI are joined via a linker to a VL and CK or CA to form a continuous protein chain wherein the linker is long enough to allow the protein chain to fold back on itself and form a monovalent antigen binding site independent of the orientation. An scFab comprises a first heavy chain variable region (scFab VH), a CHI domain, a first light chain variable region (scFab VL), and a CK or CA domain.

[0050] A single-chain fragment crystallizable (“scFc”) is a fusion protein in which a CH2 and CH3 (Fcl) are joined via a linker to another CH2 and CH3 (Fc2) to form a continuous protein chain wherein the linker is long enough to allow the protein chain to fold back on itself.10024-W001-SECIn some embodiments, the scFc comprises cysteine clamps. An scFc may also comprise an Ig- Fc hinge region, or part of an Ig-Fc hinge region. The hinge is amino terminal to the CH2 domain, and the scFc may have the following orientation: (Fcl : hinge, CH2, CH3), linker, (Fc2: hinge, CH2, CH3). It is envisaged that the hinge region promotes dimerization. Such Fc polypeptide molecules can be obtained by papain digestion of an immunoglobulin region (resulting in a dimer of two Fc polypeptide), for example and not limitation. In an embodiment, the polypeptide sequence of an Fc monomer is substantially similar to an Fc polypeptide sequence of: an IgGi Fc region, an IgG? Fc region, an IgGs Fc region, an IgG4 Fc region, an IgM Fc region, an IgA Fc region, an IgD Fc region and an IgE Fc region. (See, e.g., Padlan, Molecular Immunology, 31(3), 169-217 (1993)).

[0051] The scFab, scFv, and / or scFc may also have a cysteine clamp. A "cysteine clamp" involves the introduction of a cysteine into a polypeptide domain at a specific location, typically through replacing an existing amino acid at the specific location, so that when in proximity with another polypeptide domain, also having a cysteine introduced at a specific location, a disulfide bond (a “cysteine clamp”) may be formed between the two domains. In certain embodiments, an scFc comprises at least one cysteine clamp that results in a disulfide bond across both CH2 domains. In a further specific embodiment, an scFc comprises at least two cysteine clamps that results in a disulfide bond across both CH2 domains. In other embodiments, a binding construct’s VH and VL domains may comprise the cysteine clamp(s) to result in disulfide bond formation between the VH and VL domains. These cysteine clamps will stabilize the VH and VL domains in an antigen-binding configuration.

[0052] A cysteine clamp may be naturally occurring or it may be a result of a molecule engineered to contain cysteines. For example, a scFab may have a natural cysteine clamp between the heavy and light chain constant domains. An scFab may also have a natural cysteine clamp between the heavy and light chain constant domains and an engineered cysteine clamp between cysteines at residue 44 of the heavy chain variable region and residue 100 of the light chain variable region. In addition, an anti-target scFv may also contain a cysteine clamp between cysteines at residue 44 of the heavy chain variable region and residue 100 of the light chain variable region. An scFc may contain hinge cysteine clamps, natural CH2 / CH3 cysteine clamps, and / or an engineered CH2 cysteine clamp (intrachain).

[0053] The VH and VL contain CDRs, which are interspersed with regions that are more conserved, termed framework regions (“FR”). Each variable region is composed of 3 CDRs and 4 FRs, arranged from amino-terminus to carboxy -terminus in the following order:10024-W001-SECFR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. The 3 CDRs of the VL are referred to as “LCDR1, LCDR2, and LCDR3,” and the 3 CDRs of the VH are referred to as “HCDR1, HCDR2, and HCDR3.” The CDRs contain most of the residues which form specific interactions with the antigen. That is, the CDRs contain most of the residues that are in contact with the antigen's residues. Assignment of amino acids to CDR domains within the VL and HL regions can be based on well-known naming conventions such as the Kabat numbering convention (Kabat, et al., Ann. NY Acad. Sci. 190:382-93 (1971); Kabat et al., Sequences of Proteins of Immunological Interest, Fifth Edition, U.S. Department of Health and Human Services, NIH Publication No. 91-3242 (1991)). It is understand that other numbering conventions may also be used, such as Chothia (Chothia et al., “Canonical structures for the hypervariable regions of immunoglobulins”, Journal of Molecular Biology, 196, 901-917 (1987); Al-Lazikani et al., “Standard conformations for the canonical structures of immunoglobulins”, Journal of Molecular Biology, 273, 927-948 (1997)), and / or North (North et al., “A New Clustering of Antibody CDR Loop Conformations”, Journal of Molecular Biology, 406, 228-256 (2011)).

[0054] In another embodiment, the present invention provides vectors comprising a nucleic acid encoding a polypeptide of the invention or a portion thereof. Examples of vectors include, but are not limited to, plasmids, viral vectors, non-episomal mammalian vectors and expression vectors, for example, recombinant expression vectors.

[0055] The recombinant expression vectors of the invention can comprise a nucleic acid of the invention in a form suitable for expression of the nucleic acid in a host cell. The recombinant expression vectors include one or more regulatory sequences, selected on the basis of the host cells to be used for expression, which is operably linked to the nucleic acid sequence to be expressed. Regulatory sequences include those that direct constitutive expression of a nucleotide sequence in many types of host cells (e.g., SV40 early gene enhancer, Rous sarcoma virus promoter and cytomegalovirus promoter), those that direct expression of the nucleotide sequence only in certain host cells (e.g., tissue-specific regulatory sequences, see Voss et al., 1986, Trends Biochem. Sci. 11 :287, Maniatis et al., 1987, Science 236:1237, incorporated by reference herein in their entireties), and those that direct inducible expression of a nucleotide sequence in response to particular treatment or condition (e.g., the metallothionin promoter in mammalian cells and the tet-responsive and / or streptomycin responsive promoter in both prokaryotic and eukaryotic systems. It will be appreciated by those skilled in the art that the design of the expression vector can depend on such factors as the choice of the host cell to be transformed, the level of expression of protein desired, etc. The expression vectors of the10024-W001-SEC invention can be introduced into host cells to thereby produce proteins or peptides, including fusion proteins or peptides, encoded by nucleic acids as described herein.

[0056] In another embodiment, the present invention provides host cells into which a recombinant expression vector of the invention has been introduced. A host cell can be any prokaryotic cell or eukaryotic cell. Prokaryotic host cells include gram negative or gram positive organisms, for example E. coli or bacilli. Higher eukaryotic cells include insect cells, yeast cells, and established cell lines of mammalian origin. Examples of suitable mammalian host cell lines include Chinese hamster ovary (CHO) cells or their derivatives such as Veggie CHO and related cell lines which grow in serum -free media (see Rasmussen et al., 1998, Cytotechnology 28:31) or CHO strain DXB-11, which is deficient in DHFR (see Urlaub et al., 1980, Proc. Natl. Acad. Sci. USA 77:4216-20). Additional CHO cell lines include CHO-K1 (ATCC#CCL-61), EM9 (ATCC# CRL-1861), and UV20 (ATCC# CRL-1862). Additional host cells include the COS-7 line of monkey kidney cells (ATCC CRL 1651) (see Gluzman et al., 1981, Cell 23: 175), L cells, C127 cells, 3T3 cells (ATCC CCL 163), AM-l / D cells (described in U.S. Patent No. 6,210,924), HeLa cells, BHK (ATCC CRL 10) cell lines, the CV1 / EBNA cell line derived from the African green monkey kidney cell line CV1 (ATCC CCL 70) (see McMahan et al., 1991, EMBO J. 10:2821), human embryonic kidney cells such as 293, 293 EBNA or MSR 293, human epidermal A431 cells, human Colo205 cells, other transformed primate cell lines, normal diploid cells, cell strains derived from in vitro culture of primary tissue, primary explants, HL-60, U937, HaK or Jurkat cells. Appropriate cloning and expression vectors for use with bacterial, fungal, yeast, and mammalian cellular hosts are described by Pouwels et al. (Cloning Vectors: A Laboratory Manual, Elsevier, New York, 1985).

[0057] Typically, expression vectors used in any of the host cells will contain sequences for plasmid maintenance and for cloning and expression of exogenous nucleotide sequences. Such sequences, collectively referred to as “flanking sequences” in certain embodiments will typically include one or more of the following nucleotide sequences: a promoter, one or more enhancer sequences, an origin of replication, a transcriptional termination sequence, a complete intron sequence containing a donor and acceptor splice site, a sequence encoding a leader sequence for polypeptide secretion, a ribosome binding site, a polyadenylation sequence, a polylinker region for inserting the nucleic acid encoding the polypeptide to be expressed, and a selectable marker element. The leader sequence may comprise an amino acid sequence given by SEQ ID NO: 16 (MDMRVPAQLL10024-W001-SECGLLLLWLRGA RC) which is encoded by SEQ ID NO: 17 (atggacatga gagtgcctgc acagctgctg ggcctgctgc tgctgtggct gagaggcgcc agatgc). The leader sequence may comprise an amino acid sequence given by SEQ ID NO: 18 (MAWALLLLTL LTQGTGSWA) which is encoded by SEQ ID NO: 19 (atggcctggg ctctgctgct cctcaccctc ctcactcagg gcacagggtc ctgggcc). The leader polynucleotide sequence may comprise a polynucleotide sequence given by SEQ ID NO: 20 (ATGGACATGAGAGTGCCTGCACAGCTGCTGGGCCTGCTGCTGCTGTGGCTGAG AGGCGCCAGATG).

[0058] Vector DNA can be introduced into prokaryotic or eukaryotic cells via conventional transformation or transfection techniques. For stable transfection of mammalian cells, it is known that, depending upon the expression vector and transfection technique used, only a small fraction of cells may integrate the foreign DNA into their genome. In order to identify and select these integrants, a gene that encodes a selectable marker (e.g., for resistance to antibiotics) is generally introduced into the host cells along with the gene of interest. Additional selectable markers include those which confer resistance to drugs, such as G418, hygromycin and methotrexate. Cells stably transfected with the introduced nucleic acid can be identified by drug selection (e.g., cells that have incorporated the selectable marker gene will survive, while the other cells die), among other methods.

[0059] A polynucleotide encoding an amino acid sequence of an scFab-containing protein of the present invention can be any length as appropriate for the desired use or function, and can comprise one or more additional sequences, for example, regulatory sequences, and / or be part of a larger nucleic acid, for example, a vector. The skilled artisan will appreciate that, due to the degeneracy of the genetic code, each of the polypeptide sequences disclosed herein is encoded by a large number of other nucleic acid sequences. Mutations can also be introduced into a nucleic acid without significantly altering the biological activity of a polypeptide that it encodes. For example, one can make nucleotide substitutions leading to amino acid substitutions at non-essential amino acid residues.

[0060] Transformed cells can be cultured under conditions that promote expression of the polypeptide, and the polypeptide recovered by conventional protein purification procedures. Polypeptides contemplated for use herein include substantially homogeneous recombinant mammalian polypeptides substantially free of contaminating endogenous materials. Cells containing the nucleic acid encoding the scFab-containing proteins of the present invention also include hybridomas.10024-W001-SEC

[0061] In some embodiments, a vector comprising a nucleic acid molecule as described herein is provided. In some embodiments, the invention comprises a host cell comprising a nucleic acid molecule as described herein. In some embodiments, a nucleic acid molecule encoding a scFab-containing protein as described herein is provided. In some embodiments, a pharmaceutical composition comprising at least one scFab-containing protein described herein is provided.

[0062] Glutaminyl and asparaginyl residues are frequently deamidated to the corresponding glutamyl and aspartyl residues, respectively. Alternatively, these residues are deamidated under mildly acidic conditions. Either form of these residues falls within the scope of this invention.

[0063] Other modifications include hydroxylation of proline and lysine, phosphorylation of hydroxyl groups of seryl or threonyl residues, methylation of the a-amino groups of lysine, arginine, and histidine side chains (T. E. Creighton, Proteins: Structure and Molecular Properties, W. H. Freeman & Co., San Francisco, 1983, pp. 79-86), acetylation of the N-terminal amine, and amidation of any C-terminal carboxyl group.

[0064] Another type of covalent modification of the scFab-containing proteins included within the scope of this invention comprises altering the glycosylation pattern of the protein. As is known in the art, glycosylation patterns can depend on both the sequence of the protein (e.g., the presence or absence of particular glycosylation amino acid residues, discussed below), or the host cell or organism in which the protein is produced. Particular expression systems are discussed below.

[0065] Glycosylation of polypeptides is typically either N-linked or O-linked. N-linked refers to the attachment of the carbohydrate moiety to the side chain of an asparagine residue. The tri-peptide sequences asparagine-X-serine and asparagine-X-threonine, where X is any amino acid except proline, are the recognition sequences for enzymatic attachment of the carbohydrate moiety to the asparagine side chain. Thus, the presence of either of these tri-peptide sequences in a polypeptide creates a potential glycosylation site. O-linked glycosylation refers to the attachment of one of the sugars N-acetylgalactosamine, galactose, or xylose, to a hydroxyamino acid, most commonly serine or threonine, although 5-hydroxyproline or 5- hydroxylysine may also be used.

[0066] As used interchangeably herein, "treatment" and / or "treating" and / or "treat" are intended to refer to all processes wherein there may be a slowing, interrupting, arresting, controlling, stopping, or reversing of the progression of the disorders described herein, but does10024-W001-SEC not necessarily indicate a total elimination of all disorder symptoms. Treatment includes administration of an scFab-containing protein of the present invention for treatment of a disease or condition in a human that would benefit from activity of an scFab-containing protein of the present invention, and includes: (a) inhibiting further progression of the disease; and (b) relieving the disease, i.e., causing regression of the disease or disorder or alleviating symptoms or complications thereof.

[0067] In the context of oncology, for example, the present disclosure contemplates an scFab-containing protein of the present invention to reduce the size of a patient’s tumor. In an embodiment, the tumor is reduced by 10%. In an embodiment, the tumor is reduced by 20%. In an embodiment, the tumor is reduced by 30%. In an embodiment, the tumor is reduced by 40%. In an embodiment, the tumor is reduced by 50%. In an embodiment, the tumor is reduced by 60%. In an embodiment, the tumor is reduced by 70%. In an embodiment, the tumor is reduced by 80%. In an embodiment, the tumor is reduced by 90%. In an embodiment, the tumor is no longer present.

[0068] The size of a patient's tumor (and / or metastatic lesions) can be determined by methods known in the art. Such methods include computer tomography (CT), MRI, and / or bone scans.

[0069] As used herein, an “effective amount” means the amount of a single-chain protein of the present invention or pharmaceutical composition comprising such single-chain protein that will elicit the biological or medical response of or desired therapeutic effect on a tissue, system, animal, mammal, or human that is being sought by the researcher, medical doctor, or other clinician. An effective amount of the single-chain protein may vary according to factors such as the disease state, age, sex, and weight of the individual, and the ability of the antibody to elicit a desired response in the individual. An effective amount is also one in which any toxic or detrimental effect of the antibody is outweighed by the therapeutically beneficial effects. Such benefit includes, for example, improving signs or symptoms of cancer.

[0070] A single-chain protein, or a pharmaceutical composition containing such a single-chain protein, can be administered by any feasible method. Protein therapeutics will ordinarily be administered by a parenteral route, for example by injection, since oral administration, in the absence of some special formulation or circumstance, would lead to hydrolysis of the protein in the acid environment of the stomach. Subcutaneous, intramuscular, intravenous, intraarterial, intralesional, or peritoneal bolus injection are possible routes of10024-W001-SEC administration. A single-chain protein can also be administered via infusion, for example intravenous or subcutaneous infusion.

[0071] Single-chain proteins can be administered in the form of a composition comprising one or more additional components such as a physiologically acceptable carrier, excipient or diluent. Optionally, the composition additionally comprises one or more physiologically active agents. In various particular embodiments, the composition comprises one, two, three, four, five, or six physiologically active agents in addition to one or more single-chain proteins.EXAMPLESMaterials and Methods

[0072] Cloning and recombinant antibody expression

[0073] Twenty-four constructs were engineered, CONI through CON24 (Fig. 5).

[0074] Single chain Fab and Fc modules were split into two DNA fragments with unique golden gate BsmBI restriction sites and synthesized at either Integrated DNA or Twist Bioscience. A Gly-Gly-Gly-Gly-Ser linker was used to connect each scFab / scFc module and linkers connecting heavy and light chains or two CH2-CH3 chains are described in Figure 5. The CHI, hinge, CH2, and CH3 domains were based on human IgGl subclass sequence. Cloning was performed into an in-house vector for expression. Transfection into human embryonic kidney (HEK) 293T cells (CRL-3216,™ ATCC) was performed using 293Fectin according to manufacturer’s protocols, modified to include 1 pL of 293Fectin per 1 pg of DNA in total precomplex volume of 100 pL. Supernatants were harvested 5 days after transfection.

[0075] Purification of constructs

[0076] Two step purification was performed on all constructs; i) first, Protein A magnetic beads (Genscript, #L00672), followed by ii) SEC FPLC (AKTA pure, Cytiva). Supernatants from 293T transfections were incubated with Protein A magnetic beads on a rotisserie (~10 rpm) overnight at ambient temperature. Using a custom build magnetic bead separator, Protein A beads were washed four times with PBS, eluted using 1.0 mL of 0.1M Glycine (pH 3.0) (Sigma, # G8898), and neutralized with 100 pL of 1.0M TRIS (pH8.0) (Fisher, # BP 152). Finally, the eluted constructs were filtered using Spin-X centrifuge tubes (Costar, #8160). One milliliter of filtered eluent was injected onto a SuperdexTM 200 Increase 10 / 300 GL column (Cytiva, 28990944) and run at 0.5 mL / min using an isocratic gradient of10024-W001-SECPBS (Cytiva, # SH30256.02). SDS-PAGE was run as a preliminary check to determine fractions containing construct with correct molecular weights. Fractions were pooled and concentrated to 500 pL using Amicon Ultra-0.5mL (Milipore, UFC501024). The concentrations of purified antibodies were determined by NanoDrop TM 2000 (ThermoFisher) using extinction coefficient for each scFab-scFc format. Percent monomer of each construct was determined from SEC chromatographs.

[0077] Melting temperature determination using Differential Scanning Fluorimetry (DSF)

[0078] Melting temperatures of purified constructs were measured using the Thermal Cycler & CFX Real-Time System (Bio-Rad). Each construct (at concentration 0.5 - 1 mg / mL) and RNase A Reference Standard (Worthington-BioChem, # LS005650) (at 1 mg / mL) were prepared and 25 pL transferred into a 96-well PCR plate (Bio-Rad # HSP9601). 5 pL of a lOOx SYPRO Orange protein gel stain (Invitrogen, # S6651) dye was added into each sample well and mixed. The plate was covered with a micro-seal film (Bio-Rad, # MSB 1001) and centrifuged at 1,300 rpm for 5 min. The plate was placed into the autosampler and the heating step was set from 20 °C to 95 °C in 1 min per 1°C. All chromatograms and melting temperature (Tm) values are generated automatically.

[0079] Mass spectrophotometry (MS) analysis

[0080] Purified constructs were analyzed under reduced and non-reduced conditions on an Agilant 1200 HPLC system connected to an Agilent 6200 TOF MS mass spectrometer. Constructs under reduced conditions were treated with TCEP and all constructs under both non-reducing and reducing conditions were prepared with 0.1% formic acid. Constructs were subjecting it to reverse phase chromatography on a 2.1 mm x 75 mm C4 column (Mac-mod, Chadds Ford, PA) at a flow rate of 0.75 mL / min with a gradient of 5-95% B (Solvent A: 0.05% TFA in water, Solvent B: 0.1% TFA in acetonitrile) over 12 min. Data was analyzed either using MaxEnt deconvolution algorithm through Masshunter or using Parsimonious charge deconvolution algorithm through Protein Metrics -PMI Intact Mass software.

[0081] SDS polyacrylamide gel electrophoresis (PAGE)

[0082] SDS-PAGE was performed with a Mini-PROTEAN system (Bio-Rad, Munich, Germany). The samples were run on a 4-20% Mini-Protean TGX stain free gel (Bio-Rad, #4568096) under two conditions: 1) Under reducing conditions, 2 pg of each sample was reduced using 10X NuPAGE sample reducing agent (Fisher, #NP004) in 4X Laemmli (BioRad, #1610747). 2) Under non-reducing conditions, 2 pg of each sample was added to 4X10024-W001-SECLaemmli (Bio-Rad, #1610747). Samples were incubated at 100 °C for 10 min and run in lx Tris / Glycine / SDS buffer (Bio-Rad, # 1610772) for approximately one hour at 150V.

[0083] Affinity assay using BLI

[0084] The Octet® HTX instrument was used to determine affinities of all constructs. Prior to the binding measurements, the sensor tips were pre-hydrated for 10 min in Octet® running buffer containing 10 mM Tris (Fisher, #BP152-1), 150 mM NaCl (Fisher, #S271-10), 1 mM CaC12 (Fisher, #BP510-500), 0.1 mg / mL BSA (BioShop, #ALB001.1), and 0.1% Triton-X 100 (Calbiochem, #9410-OP), pH7.4. Streptavidin (SA) biosensors (ForteBio, #18-5021) were used to load biotinylated anti -human Fc antibody (Invitrogen, #A18827) followed by one baseline step of 60 s in Octet® running buffer. Constructs were captured by the anti -human Fc antibody prior to being submerged in wells containing different concentrations of antigen for 5 min, followed by a 10 min dissociation. Antibodies and antigen concentrations were all prepared in Octet® running buffer. SA biosensors were used once without regeneration. For data evaluation, ForteBio Data Analysis vl 1.0 software was used. The kinetic rate constants, association rate constant (ka, M-ls-1), dissociation rate constant (kd, s-1), and the equilibrium rate constant (KD, M) were determined by using a 1 : 1 Langmuir model.

[0085] On-cell binding assay

[0086] Vectors containing either Human TAA1 (Target Antigen 1) or and irrelevant protein, Human TAA3 (Target Antigen 3), were transfected HEK293T cells as described previously. Constructs were prepared at double the assay concentration and added to the same volume as 50,000 transfected cells per well in FACS buffer (PBS, 2% FBS). Test constructs with cells were incubated for 1 hr at 4 °C. Cells were washed three times, followed by addition of secondary solution (5 pg / mL Goat a Human AlexaFluor 647 (Jackson Immunoresearch, # 109-605-098) and 2.5 pg / mL 7-Aminoactinmycin D (7AAD) (Sigma- Aldrich, # A9400) in FACS buffer and incubated for 15 minutes at 4 °C. Cells were washed three times, resuspended in FACS buffer and FACS was performed using iQue 2 (Sartorius).

[0087] TAA1 agonist assay

[0088] A stable expressing Human TAA2 (Target Antigen 2) and tetTAAl CHO cell line was incubated overnight (ON) at 37 °C, in 5 % CO2 in culture media (Fl 2 media with GlutaMax (Gibco, # 11765054), 10 % FBS (Hyclone, SH30396.03) , lx Antibiotic- Antimycotic (Gibco, 15240-062), lOug / mL Blastincin (Gibco, Al 1139-03), 0.5 mg / mL G418 (Corning, 30-234-C1), 0.4mg / mL Zeocin (Gibco, R25001) ) supplemented with Doxycycline (Fisher Scientificc, # J67043) for TAA1 induction. The following day, the culture media was10024-W001-SEC replaced with serum free assay media (F12 (Gibco, # 11765054) + 0.5% BSA (Sigma, A7979)) and incubated for 3 hrs at 37 °C in 5 % CO2. Cells were detached using a cell scraper, washed, and re-suspended in serum free assay media to a final concentration of 1.875x106 cells / mL. 8 pL of cell suspension (15,000 cells) was added to 384-well low volume plates (PerkinElmer, #ProxiPlate-384 plus) followed by 4 pL of a 3x concentrated sample solution (resulting in a three-fold dilution of each construct). Cells with each construct were incubated for 12-30 minutes at 37 °C in 5 % CO2 followed by the addition of 4 pL of lysis buffer (from the Cisbio kit # 64AERPEH). The plate was sealed and placed on a shaker for 30 minutes at room temperature. The CisBio Advanced phospho-ERK 1 / 2 assay kit (Cisbio, # 64AERPEH) was followed according to manufacturer's protocol. Fluorescence at 665 and 620 nm was read on Envision plate reader (PerkinElmer).Results

[0089] To gain an understanding of different linker compositions on the biophysical and biochemical properties of Fab and Fc modules, we generated monovalent scFab-scFc constructs. These scFab-scFc constructs included scFv linkers (G4S, G4Q, G4E), alternating repeats of G4S and G4Q, linker applied in previously reported scFab constructs (G3SE) and scFv display (GGSSRSSSSGGGGSGGGGS) [Barbas CF, Burton DR, Scott JK, Silverman GJ. Phage Display: A Laboratory Manual: Cold Spring Harbor Laboratory Press; 2001], and a linker based on a FcRn peptide sequence to enhance serum half-life (QRFVTGHFGGLYPANG) [Datta-Mannan A, Boyles J, Huang L, Jin ZY, Peariso A, Murphy AT, et al. Engineered FcRn Binding Fusion Peptides Significantly Enhance the Half-Life of a Fab Domain in Cynomolgus Monkeys. Biotechnology journal. 2019; 14:el 800007], As our proof of concept target to demonstrate biochemical and biophysical properties, we used anti- TAA1 antibody sequences identified from a hybridoma screen in monovalent scFab-scFc constructs. All scFab-scFc constructs, except for Construct 20 (Con20), resulted in similar expression yields (Table 1). Specifically, expression yields as determined after Protein A purification ranged from 10 to 66 mg / L scale averaging 31 mg / L; Con20 yielded ~1 mg / L, thus, only affinity determination was possible. Size exclusion chromatography (SEC) profiles indicated >55 % of the species for each construct were monomeric with the remaining dimeric or trimeric. For Con22, whereby the scFc is at the N-terminus and the scFab at the C-terminus, the amount of monomeric species fell to ~40 % suggesting that positioning the scFc module at the N-terminus may lead to greater aggregation (Table 1).10024-W001-SEC

[0090] Furthermore, analysis by Liquid Chromatography Mass Spectroscopy (LCMS) revealed most constructs were expressed with the correct mass, with constructs Coni, Con 10, Conl2, and Conl3, generating significant clipped species (Table 1). Coni and ConlO each showed a major contribution from the intact construct, however, demonstrated a mass contribution from a minor species that had been clipped at the scFab linker. A minor species of Coni showed clipping at residues L18-R19 which is in the FR1 and F253-V254 which is the FcRn peptide of the scFab. Considering the scFc with FcRn peptide did not show clipping suggests having multiple FcRn repeating peptides as part of the scFab linker could result in clipping. A minor species of ConlO also showed clipping at residues S257-S258 which is the scFab linker, indicating this linker is prone to clipping. Clipping in Conl2 and Conl3 did not occur at the linker position and instead at the C-terminal position of the Fc which is a common area to be clipped. Conl2 demonstrated clipping at residues S978-L979 removing the peptide sequence, LSLSPGK, from the C-terminus and Conl3 showed clipping at residues P973-G974 removing ‘GK’ from the C-terminus (Table 1).Table 1. Biophysical and biochemical properties of all constructs generated. Brief format descriptions indicated as were al expression yield post Protein A purification, intact mass analysis determined by LCMS, monomer peaks by SEC, thermal stabilities (Tm) by DSF, and affinities (KD) to TAA1 determined by BLI, Expression yields were normalized to concentrations at a liter scale.10024-W001-SEC_CMS, SEC, and DSF analysis of Con20 was not performed due to lack of sample.

[0091] Although Con 12 and Con 13 failed to give the correct mass, the minor loss of 7 or 2 amino acid residues from the C-terminus should not impact function, whereas the presence of clipped species at the scFab linker would decrease the active fraction in solution. Melting temperatures of each construct was determined to be between 72 - 76 °C. Con23 gave two melting temperatures at 69 and 78 °C, suggesting two different targeting scFabs may impact thermal stability of each scFab. Con24, consisting of two identical scFabs demonstrated only one melting temperature at 74 °C.

[0092] To identify the impact each construct has on binding to the target protein TAA1, both on-cell binding and soluble affinities were determined. Binding was measured to TAA1 transiently transfected on HEK293 cells and 23 out of 24 constructs showed significant geomean shifts, >10-fold, over an irrelevant protein, TAA3 (Figure. 1). Con20 bound to TAA1 at ~2.7 fold over irrelevant due to lower test concentration (Con20 was tested at 43.5 nM, as opposed to lOOnM for all constructs). Affinities to soluble Human TAA1 were determined to monomeric TAA1 protein using Biolayer Interferometry (BLI). All constructs with the scFab found at the N-terminus demonstrated affinities comparable to the parental IgG molecule (40E5), <2 fold. Two scFab modules based on IgG sequences demonstrated affinities <3 fold. For Con22, the second module refers to the scFab linked to the C-term part of the scFc, i.e. going from 5' end to 3' end of the molecule, the scFab is the second domain. For Con23, the molecule can be divided into 3 domains with the linker separating each one. By positioning the scFab module as the second module from the N-terminus, Con22, the affinity decreased ~4 fold, compared Con8, a construct with equivalent linker composition and length. When the10024-W001-SEC scFab module was positioned as the third module, Con23, affinity decreased ~5 fold, compared to Con8. The decrease in affinity resulting from moving the scFab module further from the N- terminus could be due to steric hinderance from the short linker (G4S) connecting two single chain modules.

[0093] We have demonstrated a variety of linear linkers can be used to connect Fab and Fc fragments together. Generally, there is no significant difference between constructs in terms of expression level, percent monomer, stability, binding to on-cell expressing TAA1, or affinity to soluble TAA1 protein. The exceptions include: i) Con20 which has a variable sequence from Con8 and Con21, ii) the scFab linker sequence of Coni and Con 10 may be more prone to clipping than other sequences. We did not observe any significant differences in biophysical or biochemical properties between constructs with linker lengths in the scFab module between 30 - 40 amino acids (Con6-8), nor in switching of heavy and light chains (Con8 and Coni 1). The minor impact of linkers on the scFab module could be a result of maintaining the native disulfide between the heavy and light chains, thus preserving geometry of the variable domains.

[0094] We confirmed that monomeric scFab modules can demonstrate on-cell binding and affinity to Human TAA1. We aimed to determine if bi-valent scFab format can recapitulate IgG function to agonize the TAA1 pathway. TAA1 activation typically occurs through a homodimer ligand, TAAL, first binding two TAA2 coreceptors, followed by recruitment of two TAA1 molecules. One of the signaling pathways the TAA1 dimer can activate is the RAS / extracellular signal-regulated kinase (ERK) [Takahashi M. The GDNF / RET signaling pathway and human diseases. Cytokine & growth factor reviews. 2001;12:361-73], We used activation of this pathway as readout for functional activity of an anti-TAAl antibody and a bivalent scFab construct. We observed the EC50 for the canonical IgG and bivalent scFab (Con24) within two-fold from each other (IgG EC50 of 161 pM and Con24 EC50 of 288 pM) (Fig. 2). Interestingly the height of activation appeared depressed for Con24 compared to the parental IgG. The decreased affinity, KD = 4.73 nM for Con22 compared to KD = 1.21 nM for Con8 (~4-fold), of the scFab module at the C-termini and / or steric hinderance from the short linker connecting the scFab and scFc modules might be contributing factors. Monovalent constructs, Con7 and Con8, demonstrated baseline activity for all concentrations tested. Furthermore, to demonstrate the feasibility of using single chain modules as bi-specifics, we incorporated an anti-TAA2 scFab at the N-terminus and tethered a scFc - anti-TAAl scFab (Con23). Both scFab modules engaged monomeric TAA2 and TAA1 proteins simultaneously (Fig. 3). Octet® illustrated Con23 sequential binding of TAA2 and TAA1, or visa-versa. The10024-W001-SEC response signal from TAA2 and TAA1 in the same solution is approximate to the signal after sequential binding of both proteins.Table 2, Affinities to Human TAA1 of parental IgG antibodies used as scFab modules determined by Octet® . 40E5 is the parental sequence to Coni -19, 22-24, 41B11 is the parental sequence to Con20, and 44G4 the sequence to Con21,

[0095] Here, we showed that scFab and scFc modules are amenable to various linkers without impacting binding or affinity. Variable heavy and light chain orientation and linker lengths greater than 30 amino acids for the scFab module and 25 amino acids for the scFc module had no negative impacts. The lack of codon complexity and length of G4S, G4E, or G4Q repeats did present challenges and various cloning strategies were employed. Utilizing linkers of greater complexity and target specificity gives additional function to the scFab / scFc modules, i.e. addition of an FcRn binding peptide in the current study could promote greater serum half-life. Treating scFab and scFc as modules allowed for assembly into bivalent and monovalent bispecific like molecules with compared biophysical and biochemical properties to IgGs. Given these results, future work will focus on demonstrating the scFab-scFc format as a therapeutic agent. scFab Trispecific Constructs

[0096] ScFab domains as building blocks to trispecific engineering

[0097] A variety of linkers were established to connect heavy and light chains to form a scFab. These linkers demonstrated stability, and the scFab s retained binding to their cognate antigen. An effort was undertaken to design trispecific constructs using scFabs to mimic a ligand (TAAL) that binds to TAA1 and TAA2, followed by activation of TAA1 cellular signaling through TAA2-dependent recruitment. The canonical mechanism of action for TAAL involves binding to two molecules of the TAA2 receptor, leading to a conformational change that recruits an additional two molecules of the TAA1 receptor. Two TAA1 molecules are required to initiate intracellular signaling.10024-W001-SEC

[0098] To identify a trispecific mimetic of TAAL, clustering of TAA2 was proposed using bi-epitopic scFabs combined with a monovalent TAA1 scFab. The first trispecific effort focused on identifying TAA2 bi-epitopic scFabs paired with a monovalent TAA1 scFab. The TAA1 scFab was inactive in monovalent form but active in bivalent form (clone 40E5 from Figure 2). TAA2 scFabs were selected based on classification into distinct epitope bins of the TAA2 ECD.

[0099] The initial format (F0), arranged from N’- to C’-terminal sequence, contained a TAA2 scFab to one epitope (TAA2-arml) linked to a TAA2 scFab to a second epitope (TAA2- arm2), connected to a scFc, followed by a TAA1 scFab (Figure 12A). An epitope binning experiment identified at least two epitope bins within each TAA2 domain, with mAbs grouped accordingly. Representatives from each bin were selected to generate 23 trispecific constructs combining TAA2 scFabs with a TAA1 scFab. Most combinations included TAA2 scFabs binding to different epitope bins within a TAA2 domain (20 of 23), while three included scFabs across domains (Figure 12B).

[0100] All trispecific constructs were expressed and purified, yielding between 0.2 and 19 mg / L after Protein A purification (Figure 13). Activity evaluation against TAA1 identified false positives by testing each molecule in a transiently transfected CHO cell line containing TAA1, ERK-1, and Luciferase (Luc) dependent on ERK-1 activation. Testing at a single concentration (100 nM) revealed that eight trispecific constructs demonstrated greater than 2.5- fold TAA1 activity compared to negative control (Figure 14A). The other 15 trispecific constructs were non-active in TAA1 alone and tested in a TAA2 / TAA1 functional assay. These displayed TAA2-dependent TAA1 activities ranging from baseline to 3.4 times the 665nm / 620nm ratio (Figure 14B). Two molecules, F0-GGR5 and F0-GGR18, showed the highest activities (3.3x and 3.4x), corresponding to 48% and 49% of ligand activity. This initial screen identified active bi-epitopic TAA2 scFab pairs, enabling optimization of scFab geometry.

[0101] To further enhance TAA2-dependent TAA1 activity, the position of the TAA1 scFab and the length between scFabs were evaluated. Using TAA2 and TAA1 scFabs from F0- GGR5, the following designs were generated

[0102] i) Format 1 (Fl) contained TAA2 scFab (epitope 1) - TAA2 scFab (epitope 2) - TAA1 scFab.

[0103] ii) Format 2 (F2) contained TAA2 scFab (epitope 1) - TAA1 scFab - TAA2 scFab (epitope 2).10024-W001-SEC

[0104] iii) A control format incorporated an irrelevant TAA3 scFab: TAA2 scFab (epitope 2) - TAA1 scFab (Figures 4 and 6).Linker lengths between scFabs ranged from one to three G4S units (Figure 15). Fl, F2, and control constructs were expressed, purified, and mass confirmed (Figures 16-17). Testing revealed no measurable TAA1 activity for any molecule (Figure 18). However, in the TAA2- dependent TAA1 assay, Format 1 trispecific constructs displayed higher activity than Format 2. Within Format 1, shorter linkers yielded stronger activities (Fl-lmer > Fl-2mer > Fl-3mer) (Figure 19). Although activity continued to increase at concentrations above 100 nM, Fl-lmer exhibited the highest effect, reaching 83% of TAAL activity (Figure 19). Collectively, results indicated that the most potent agonist mimetic of TAAL required paratopes in close proximity, with bi-epitopic TAA2 paratopes positioned adjacent to each other.Materials and Methods

[0105] Cloning and recombinant antibody expression

[0106] Format 0 trispecific constructs (scFab #1 - scFab #2 - scFc - scFab #3) were cloned through a multistep process. In the first step, the two halves of the scFc were combined with scFab #3 using unique BsmBI golden gate restriction sites. The resulting scFc-scFab #3 fragment was inserted into a vector using Nhel and Pmel restriction sites. The 5' end of the first half of the scFc contained Sall and a unique Bsal site for cloning downstream of Nhel. In the second step, scFab #1 and scFab #2 were combined using unique Bsal restriction sites and inserted into the vector from step one using Sall and Bsal restriction sites. A schematic of the cloning process is provided in Figure 20.

[0107] Step 1 includes cloning three fragments (representing half of the trispecific) into a mammalian expression vector. Features of each fragment from 5’ - to 3 ’-ends include: i) Nhel restriction site, G4 linker with unique Bsal site for Step 2, CH2-CH3, (G4S)6 linker with a BsmBI restriction site (first half of scFc), ii) (G4S)6 linker with a BsmBI restriction site, CH2- CH3, G4 linker with second BsmBI restriction site (second half of scFc), iii) G4 linker with BsmBI restriction site, VH-CH1, (G4S)8 linker, VL-CL, and Pmel restriction site (scFab #3). Fragment i will ligate to the vector at the 5’-end through Nhel restriction sites. BsmBI sites at the 3 ’-end of fragment i and 5 ’-end of fragment ii are complementary and will ligate to each other. Similarly, the BsmBI sites at the C’-terminal of fragment ii and 5’-end of fragment iii are complementary and will ligate to each other. Fragment iii will ligate to the vector through blunt ends generated by Pmel and EcoRV digestion. Step 2 will ligate two scFab fragments to the10024-W001-SEC scFc-scFab vector generated in step 1. Features of each fragment from 5’-end to 3’-end include: iv) Sall restriction site, leader peptide, VH-CH1, (G4S)8 linker, VL-CL, and G4S linker with a Bsal restriction site (scFab #1), and v) G4S linker, leader peptide, VH-CH1, (G4S)8 linker, VL-CL, and G4S linker with a Bsal restriction site (scFab #2). The vector from Step 1 was opened using Sall and Bsal, with the 5 ’-end of fragment iv ligated to the vector through Sall restriction sites. Bsal sites at the 3’-end of fragment iv and 5’-end of fragment v are complementary and will ligate to each other. The Bsal site at the 3 ’-end of fragment v will ligate to the vector to complete the trispecific sequence.

[0108] Formats 1, 2, and control trispecific constructs were cloned using a single digestion / ligation step. Synthesized fragments for each scFab included unique linker sequences incorporating Bsal sites and sequencing primers. DNA fragments were synthesized by Integrated DNA Technologies. Cloning was carried out into an in-house vector for expression. Transfection into human embryonic kidney (HEK) 293T cells (CRL-3216™, ATCC) was performed with 293Fectin according to the manufacturer’s protocol, modified to use 1 pL of 293Fectin per 1 pg of DNA in a precomplex volume of 100 pL. Supernatants were harvested five days after transfection.

[0109] Purification of constructs

[0110] All trispecific constructs were purified in a single step using either Protein A (Genscript, #L00672) or NTA (Genscript, #L00295) magnetic beads. Supernatants from 293T transfections were incubated with beads overnight at ambient temperature. Protein A beads were washed with PBS, eluted with 1.0 mL of 0.1M Glycine (pH 3.0; Sigma, #G8898), and neutralized with 100 pL of 1.0M TRIS (pH 8.0; Fisher, #BP152). Alternatively, constructs were eluted with 1.0 mL of 500 mM Imidazole in HBS (30 mM HEPES, 150 mM NaCl, pH 7.6). Eluted proteins were filtered using Spin-X centrifuge tubes (Costar, #8160), buffer exchanged into PBS (Cytiva, #SH30256.02), and concentrated using Amicon Ultra-0.5 mL devices (Millipore, #UFC501024). Concentrations of purified antibodies were determined with a NanoDrop™ 2000 spectrophotometer (ThermoFisher) using extinction coefficients calculated for each format.

[0111] Mass spectrometry (MS) analysis

[0112] Purified Format 1 and 2 trispecific constructs (as represented in Figures 12 and 15) were prepared for mass spectrometry using two deglycosylation methods: (1) treatment with 1 pL PNGaseF and 1 pL of each sialidase and O-glycanase (Agilent, #GK80110) in 50 mM Tris buffer (pH 7.8) overnight at 37 °C, or (2) Rapid PNGaseF (New England Biolabs,10024-W001-SEC#P0710 reduced, #P7011 non-reduced) following manufacturer’s protocol. Proteins were analyzed under reduced and non-reduced conditions using an Agilent 1200 HPLC system coupled to an Agilent 6200 TOF MS.

[0113] Reduction for non-deglycosylated proteins was performed with 8 M Guanidine hydrochloride and 10 mM TCEP or DTT. For Rapid PNGaseF -treated proteins, no denaturing or reducing agents were added. Protein samples were acidified to 1% formic acid before reverse phase chromatography on a 2.1 mm * 75 mm C4 column (Mac-mod, Chadds Ford, PA) at 0.75 mL / min, with a 5-95% gradient of solvent B (0.1% formic acid in acetonitrile; solvent A: 0.1% formic acid in water) over 12 minutes. Data were analyzed using MaxEnt deconvolution (MassHunter) or Parsimonious charge deconvolution (Protein Metrics, PMI Intact Mass).

[0114] SDS-polyacrylamide gel electrophoresis (PAGE)

[0115] SDS-PAGE was conducted with a Mini-PROTEAN system (Bio-Rad, Munich, Germany) using 4-20% Mini-Protean TGX stain-free gels (Bio-Rad, #4568096) under reducing and non-reducing conditions. Samples (2 pg) were prepared with 4X Laemmli buffer (Bio-Rad, #1610747). For reducing conditions, 10X NuPAGE sample reducing agent (Fisher, #NP004) was added. Samples were incubated at 100 °C for 10 minutes and electrophoresed in IX Tris / Glycine / SDS buffer (Bio-Rad, #1610772) for ~1 hour at 150 V (Figure 16).

[0116] TAA1 agonist assay

[0117] CHO cells stably expressing human TAA2 and inducible TAA1 were cultured overnight at 37 °C in 5% CO2 in F12 media with GlutaMax (Gibco, #11765054), 10% FBS (Hyclone, SH30396.03), IX Antibiotic- Antimycotic (Gibco, #15240-062), 10 pg / mL Blasticidin (Gibco, #A11139-03), 0.5 mg / mL G418 (Corning, #30-234-0), and 0.4 mg / mL Zeocin (Gibco, #R25001), supplemented with doxycycline (Fisher, #167043) for TAA1 induction. The following day, cells were incubated in serum-free assay media (F12 + 0.5% BSA; Sigma, #A7979) for 3 hours at 37 °C in 5% CO2.

[0118] Cells were detached, washed, and resuspended to 1.875 x 106cells / mL. Fifteen thousand cells (8 pL) were seeded per well of a 384-well low-volume plate (PerkinElmer, #ProxiPlate-384 plus), followed by 4 pL of 3X concentrated construct. After 12-30 minutes incubation at 37 °C in 5% CO2, 4 pL of lysis buffer (Cisbio kit #64AERPEH) was added. Plates were sealed, shaken for 30 minutes, and processed with the Cisbio phospho-ERK 1 / 2 assay kit according to manufacturer’s instructions. Fluorescence at 665 and 620 nm was measured on an Envision plate reader (PerkinElmer).10024-W001-SEC

[0119] Independent TAA1 agonist assay

[0120] CH0-AM-1 / D cells were cultured overnight at 37 °C in 5% CO2 in DMEM supplemented with 10% FBS (Hyclone, SH30396.03), 2 mM L-glutamine, 1 mM sodium pyruvate, IX NEAA, IX HT, and 10 mM HEPES at three million cells per flask. Cells were transfected the next day with vectors encoding TAA1, ERK-1, and luciferase (Luc) using Lipofectamine 2000 (Invitrogen, #11668-027) and incubated overnight.The following day, cells were incubated in serum-free assay medium (DMEM + supplements as above) for three hours. Cells were detached, washed, replated at 25,000 cells / well in 96-well half-area plates (Greiner, #675083), and treated with test samples or controls. After overnight incubation, Bio-Gio reagent (Promega, #G7940) was added, shaken for 5 minutes in the dark, and luminescence was measured with an Envision plate reader (PerkinElmer).10024-W001-SECSEQUENCES10024-W001-SECREFERENCES1. Ahamadi-Fesharaki R. et. al., “Single-Chain Variable Fragment-Based Bispecific Antibodies: Hitting Two Targets with One Sophisticated Arrow”, Oncolytics , vol. 14 (38-56): 2012. Bird R.E. and Walker B.W., “Single chain antibody variable regions”, Trends Biotechnol., vol. 9 (132-137): 19913. Brinkmann U. and Kontermann R.E., “The making of bispecific antibodies”, mAbs, vol. 9 (182-212): 20174. Huang S. et. al., “Bispecific antibodies targeting dual tumor-associated antigens in cancer therapy”, J. Cancer Res. Clin. Oncol., vol. 146 (3111-3122): 20205. Husain B. and Ellerman D., “Expanding the Boundaries of Biotherapeutics with Bispecific Antibodies”, BioDrugs, vol. 32 (441-464): 20186. Hust M. et. al., “Single chain Fab (scFab) fragment”, BMC Biotech., col. 7 (14): 20077. Koerber J.T. et. al., “An improved single-chain Fab platform for efficient display and recombinant expression”, J. Mol. Bio., vol. 427 (576-586): 20158. Kramer K. et. al., “A generic strategy for subcloning antibody variable regions from the scFv phage display vector pCANTAB 5 E into pASK85 permits the economical production of Fab fragments and leads to improved recombinant immunoglobulin stability”, Biosens. Bioelectron., vol. 17 (305-313): 20029. Schirrmann T. et. al., “Oligomeric forms of single chain immunoglobulin (scIgG)”, mAbs, vol. 2 (73-76): 201010. Lang S. et. al., “LegoBody: facile generation of bispecific and multi-specific antibodies”, BioRxiV, 2019.12.25.88858611. Lee H.S. et. al., “Generation and characterization of a novel single-gene-encoded single-chain immunoglobulin molecule with antigen binding activity and effector functions”, Mol. Immunol., vol. 36 (61-71): 199912. Lim K.S. et. al., “Clinical Pharmacology Strategies for BispecificAntibody Development: Learnings from FDA-Approved Bispecific Antibodies in Oncology”, Clin. Pharmacol. Ther., vol. 116 (315-327): 202413. Sharma A. et. Al., “Global experience of faricimab in clinical settings - a review”, Expert Opin. Biol. Ther., vol. 24 (263-268): 202410024-W001-SEC14. Yoneyama K. et. Al., “Clinical pharmacology of emicizumab for the treatment of hemophilia A”, Expert Rev. Clin. Pharmacol., vol. 16 (775-790): 2023

Claims

10024-W001-SECCLAIMSWhat is claimed:

1. A single-chain protein comprising at least one single-chain Fab (scFab) connected to a single chain Fc (scFc).

2. The single-chain protein of Claim 1, comprising an scFab connected by an scFab N-terminus or an scFab C-terminus to an N-terminus of the scFc.

3. The single-chain protein of Claim 1, comprising an scFab connected by an scFab N-terminus or an scFab C-terminus to a C-terminus of the scFc.

4. The single-chain protein of any one of Claims 1-3, wherein the scFab is connected by the scFab N-terminus to the scFc.

5. The single-chain protein of any one of Claims 1-3, wherein the scFab is connected by the scFab C-terminus to the scFc.

6. The single-chain protein of Claim 1, comprising (i) a first scFab connected by an scFab N-terminus or an scFab C-terminus to an N-terminus of the scFc; and (ii) a second scFab connected by an scFab N-terminus or an scFab C-terminus to a C-terminus of the scFc.

7. The single-chain protein of Claim 6, wherein the first and second scFab each bind to a same target.

8. The single-chain protein of Claim 6, wherein the first and second scFab each bind to a different target.

9. A single-chain protein comprising at least three scFabs, wherein each scFab is connected to at least one other scFab.

10. The single-chain protein of any one of Claims 1-9, wherein the scFab has an orientation in the following order, from N-terminus to C-terminus, VL, either CK or C , VH, and CHI.

11. The single-chain protein of any one of Claims 1-9, wherein the scFab has an orientation in the following order, from N-terminus to C-terminus, VH, CHI, VL, and either CK or CX.

12. The single-chain protein of any one of Claims 1-11, wherein the scFab comprises a linker, and the linker is (G4S)6, (G4S)7, (G4S)8, (G4E)7, (G4Q)7, (G3SE)7, (G4SG4Q)3(G4S), or (G4Q)3(G4S)4.

13. The single-chain protein of any one of Claims 1-12, wherein the scFc comprises a linker, and the linker is (G4S)6, (G4E)6, (G4Q)6, (G4Q)3(G4S)3, or (G4QG4S)3, G3SE)6.10024-W001-SEC14. The single-chain protein of any one of Claims 1-12, wherein the scFc comprises a linker, and the linker is an FcRn binding peptide shown in C0N17.

15. The single-chain protein of any one of Claims 1-14, wherein the scFab is connected to the scFc by a G4S linker.

16. A single-chain protein comprising at least a first single-chain Fab (scFab) connected to a second scFab.

17. The single-chain protein of Claim 16, further comprising a third scFab.

18. The single-chain protein of Claim 16 or Claim 17, comprising an scFc.

19. The single-chain protein of any one of Claims 16-18, wherein: a) the first and second scFab each bind a first target at two different epitopes, and the third scFab binds a second target; b) the first and third scFab each bind a first target at two different epitopes, and the second scFab binds a second target; c) the second and third scFab each bind a first target at two different epitopes, and the second scFab binds a second target; or d) the first scFab binds a first target, the second scFab binds a second target, and the third scFab binds a third target.

Images