Bispecific antibodies that bind to the protease-like domain of the human transferrin receptor HTFR1.

Abstract

This disclosure relates to a binding protein comprising an immunoglobulin heavy chain variable region (VH) and an immunoglobulin light chain variable region (VL), and comprising M1, which is a first transferrin receptor 1 binding portion having selective binding ability to an epitope located in the protease-like domain of hTfR1, and M2, which is a second portion containing an antibody Fc domain. In the binding protein, M1 and M2 are bound to each other by at least one peptide linker.

Description

[Technical Field]

[0001] field This disclosure relates to a binding protein comprising an immunoglobulin heavy chain variable region (VH) and an immunoglobulin light chain variable region (VL), and comprising M1, which is a first transferrin receptor 1 binding portion having selective binding ability to an epitope located in the protease-like domain of hTfR1, and M2, which is a second portion containing an antibody Fc domain. In the binding protein, M1 and M2 are bound to each other by at least one peptide linker. [Background technology]

[0002] background Therapeutic modalities for brain and neurological diseases are limited because the brain's blood vessels are impermeable to most substrates transported into the bloodstream (Freskgard and Urich (2017), Neuropharmacology 120:38-55; Stanimirovic et al. (2018), BioDrugs 32:547-559). The brain's microvessels (capillaries), collectively known as the blood-brain barrier (BBB), are unique compared to the blood vessels surrounding the body. BBB endothelial cells (ECs), tightly juxtaposed with nerve cells such as astrocytes, pericytes, and neurons, give rise to phenotypic characteristics that contribute to the observed impermeability. The tight binding between ECs in the BBB restricts paracellular transport, while the absence of passive vesicles and fenestra restricts nonspecific intercellular transport. These elements work together to limit molecular flux from blood to the brain to molecules that are generally smaller than 500 Da and lipophilic. Therefore, unless a drug with the desired pharmacological properties happens to be small enough and lipophilic to cross the blood-brain barrier (BBB), it will not pass through the large transport surface area of ​​the bloodstream (600 km of capillaries in the human brain, up to 20 m). 2Another promising possibility—the use of a surface area (larger than 50%) as a delivery vehicle—becomes virtually impossible to realize. Due to these limitations, it is estimated that over 98% of all small molecule drugs and nearly 100% of the emerging class of protein and gene therapies will not cross the BBB.

[0003] International Publication WO91 / 03259 proposes a principle for the transport of neuropharmacological drugs across the blood-brain barrier (BBB), which involves conjugating the drug to an antibody that is reactive with the transferrin receptor. According to this disclosure, the binding of the conjugate to the transferrin receptor results in active transport across the BBB. Subsequent studies have further developed this concept, as exemplified by WO2012 / 075037, WO2014 / 033074, WO2018 / 011353, and WO2022 / 258841, demonstrating different formats for achieving the transport of biopharmaceuticals that can cross the BBB using the transferrin receptor.

[0004] Human transferrin receptors exist in two forms. Transferrin receptor 1 (TfR1) is the target of the binding protein described herein. TfR1 is an iron transport protein that maintains intracellular iron concentration by specifically binding the iron transport proteins transferrin (Tf) and ferritin (Ft) into the cell through clathrin-coated vesicle-mediated endocytosis, thereby promoting their internal translocation. TfR1 is expressed in numerous cells and organelles, but at varying levels. Importantly, TfR1 is expressed at a higher degree in BBB endothelial cells than in other endothelial cells, indicating that the receptor is a target for neuropharmacological transport. Structurally, TfR1 is a dimeric transmembrane glycoprotein containing an amino acid sequence (SEQ ID NO: 66) with a large external domain (residues 89-760), an intramembrane domain (residues 62-88), and a cytoplasmic domain (residues 1-61). The external domain has three distinct domains in sequence, separated from the cell surface by the Stark domain (residues 89-120). These three parts of the external domain are the helical domain (residues 606-760), the protease-like domain (residues 121-183, 384-605), and the apical domain (residues 184-383) (Lawrence et al. (1999), Science 286:779-782).

[0005] When using TfR1 to transport molecules containing TfR1-binding agents that cross the blood-brain barrier in this manner, binding to TfR1 occurs first in circulation, i.e., before the molecule is transported to brain components. While circulating, the molecule-TfR1 complex is exposed to blood components and peripheral system cells, and the molecule localizes and is presented on TfR1-expressing cells. The complex presented on the cell surface is then exposed to various endogenous blood components or other factors present in the local environment surrounding the TfR1-expressing cell. The potential interaction between this presented TfR1 / binding agent complex and the environment can induce or stimulate multiple pathways with various forms of action. If this occurs, it can cause undesirable peripheral side effects before the molecule is transported to the brain. This is especially important if the molecule containing the TfR1-binding agent contains antibodies or fragments thereof that are known to mediate immune responses.

[0006] The drawbacks of targeting TfR1 with antibodies have been demonstrated, such as acute clinical symptoms and a decrease in circulating reticulocytes. Previously, this was addressed by removing the antibody's Fc effector function. This has resulted in improvements in acute clinical symptoms and partial rescue of reticulocytes. Both antibody-dependent cell-mediated cytotoxicity (ADCC) and complement-dependent cell-mediated cytotoxicity (CDC) have been shown to be involved in these processes. Introducing mutations into the Fc portion of antibodies to reduce or eliminate ADCC and / or CDC has been suggested as a promising mitigation strategy in the development of TfR1-based therapies designed to cross the blood-brain barrier (Couch et al. (2013), Sci Transl Med 5:183ra57). However, this strategy modifies the antibody structure, removing its effector function when binding to the target. Therefore, Fc mutations may reduce or even interfere with the therapeutic function of the therapeutic antibody. For example, the mechanism of amyloid-beta removal from the brain is hypothesized to be facilitated by the Fc effector function of antibodies, which includes the roles of microglia and expressed Fc gamma receptors, as well as immune cells and phagocytes resident in the central nervous system (Condello et al. (2015), Nat Commun 6:6176). Therefore, in certain applications of antibodies, such as the treatment of brain diseases, Fc mediation is necessary to achieve the desired clinical effect.

[0007] Administration of therapeutic monoclonal antibodies is often accompanied by severe first infusion reactions (FIRs). These have been shown to be caused by the effector function of the antibody, which may complicate the clinical applications of therapeutic antibodies, especially when TfR1 is one of their targets (Weber et al. (2018), Pharm Res 35(9):169). Another finding is that when an antibody is bound to TfR1 via an scFab TfR1 binding molecule linked to the C-terminus of the antibody's Fc portion, the antibody's Fab arm may be able to block, to some extent, the ADCC and / or CDC pathways mediated by FcγR bound to the antibody's Fc portion (Weber et al. (2018), Cell Rep 22:149-162). Weber et al. showed that this effect is due to the antibody binding in the opposite direction compared to a normal antibody.

[0008] There remains a need for biopharmaceuticals for the treatment of brain diseases, such as those that can be delivered across the blood-brain barrier and exhibit beneficial characteristics in reducing the risk of side effects, such as those mediated by the Fc effector function of antibodies. [Overview of the Initiative]

[0009] Disclosure of the invention The purpose of this disclosure is to address this need by providing a binding protein that can cross the blood-brain barrier.

[0010] Another objective of this disclosure is to provide a binding protein that has therapeutic function on the brain while mitigating or eliminating the risk of Fc-mediated adverse effects such as ADCC, ADCP, or CDC.

[0011] Another objective of this disclosure is to improve existing binding proteins based on a dual or multispecific format for therapeutic purposes of neurological diseases.

[0012] Another objective of this disclosure is to provide a binding protein with novel binding orientation in its interaction with hTfR1.

[0013] In various aspects of this specification, one or more of these objectives, and / or any other objectives that are apparent to those skilled in the art from the present disclosure, are satisfied.

[0014] Thus, in a first aspect, the present disclosure provides a binding protein comprising: - A first portion M1, which is a human transferrin receptor 1 (hTfR1) binding portion comprising an immunoglobulin heavy chain variable region (VH) and an immunoglobulin light chain variable region (VL), wherein the VH and VL regions form a VH / VL pair comprising an antigen binding surface, and the antigen binding surface has the ability to selectively bind to an epitope located in the protease-like domain of hTfR1 defined by amino acid residues 121-183 and 384-605 of SEQ ID NO: 66; - A second portion M2 comprising an antibody Fc domain where M1 and M2 are linked to each other by at least one peptide linker between M1 and M2, and the linker is arranged such that when administered to a human and when M1 binds to hTfR1 present on the cell, M2 elicits a reduced Fc-mediated response.

[0015] In an alternative first aspect, the present disclosure provides a binding protein comprising: - A first portion M1, which is a human transferrin receptor 1 (hTfR1) binding portion comprising an immunoglobulin heavy chain variable region (VH) and an immunoglobulin light chain variable region (VL), wherein the VH and VL regions form a VH / VL pair comprising an antigen binding surface, and the antigen binding surface has the ability to selectively bind to an epitope located in the protease-like domain of hTfR1 defined by amino acid residues 121-183 and 384-605 of SEQ ID NO: 66; - A second portion M2 comprising an antibody Fc domain The present invention provides a binding protein in which M1 and M2 are linked to each other by at least one peptide linker between M1 and M2, the linker being positioned such that when M1 binds to hTfR1 present on the cell, M2 is oriented toward the cell surface. [Brief explanation of the drawing]

[0016] Brief explanation of the drawing [Figure 1] Figure 1 shows the X-ray structure of the complex between the ectodomain of human transferrin receptor 1 (hTfR1) and the VH / VL pair of the hTfR1 binder h26D3, which here represents M1, as determined as described in Example 5. As shown, h26D3 or other binders with the same epitope specificity bind to the lateral surfaces of hTfR1 on the protease-like domain, placing the binder at a unique position on the hTfR1 structure. Since hTfR1 is a homodimer, each hTfR1 can bind one h26D3 on each side. When used as part of the binding protein described herein, the VH / VL pair provides four different possible binding sites as shown in the structural image: (VL-C) C-terminal amino acid residue of the light chain variable region; (VH-N) N-terminal amino acid residue of the heavy chain variable region; (VH-C) C-terminal amino acid residue of the heavy chain variable region; and (VL-N) N-terminal amino acid residue of the light chain variable region. [Figure 2]Figure 2 shows two different design options for creating an hTfR1-binding scFv for use here as M1. Depending on the order in which the VH and VL variable domains are bound in the scFv, different sites of the scFv can be used as binding sites for a second portion, M2, for example, in the form of an antibody or other Fc-containing protein. (A) shows the "VH-first" configuration of the hTfR1-binding scFv, showing that the scFv linker between the variable regions is located below the scFv with respect to the orientation of hTfR1 when fixed to the cell membrane. In this configuration, the binding of the M2 portion can be one or both of the binding sites VL-C and VH-N. (B) shows the "VL-first" configuration of the hTfR1-binding scFv, showing that the scFv linker between the variable regions is located above the scFv with respect to the orientation of hTfR1 when fixed to the cell membrane. In this configuration, the bond in the M2 portion can be one or both of the bond points VH-C and VL-N. [Figure 3] Figure 3 is a schematic diagram of hTfR1 immobilized on the cell surface, showing how the Stark domain of hTfR1 provides space ("hidden space") beneath the interaction site between portion M1 and hTfR1. The figure also shows the binding site of M2 in one embodiment of this disclosure, where M1 is a scFv in a "VL-first" orientation. In this preferred embodiment, M2 appropriately binds to M1 via one or both of VH-C and VL-N. [Figure 4]Figure 4 is a schematic diagram of two embodiments of the binding protein in this disclosure. In the embodiment designated “Gen 2A”, M1 is provided as an scFv fused between two antibody light chains bound to two identical heavy chains to form an intact antibody as M2. This M2 antibody binds to two binding sites on the M1 scFv via two linkers at the C-terminus of the first light chain and the N-terminus of the second light chain of the M2 antibody. This is a symmetric IgG structure constructed from two different polypeptide chains, with each binding protein containing one hTfR1 binder (M1 portion). Further details of the Gen 2A format are disclosed in WO2022 / 258841. In the embodiment designated “Gen 2D”, M1 is provided as an scFv fused on the C-terminal amino acid residue of a single heavy chain HC(knob) containing the knob portion of a knob-into-hole asymmetric IgG construct. The light chain (LC) and heavy chain (HC) combine with the corresponding knob-into-hole knob portion, which is not scFv, to form the complete IgG structure as the M2 portion of this disclosure. The M1 portion is then attached to the C-terminus of the Fc portion in the form of scFv using one of the available binding sites within the VH / VL pair of M1. [Figure 5] Figure 5 shows the results of IgG antibody binding screening from immunization of unpurified fusion cell supernatants by biolayer interferometry (BLI) against human (hTfR1), cynomolgus monkey (cTfR1), and mouse (mTfR1) cells, as described in Example 1. [Figure 6] Figure 6 shows the results of BLI binding analysis for the Fab fragments of mouse antibodies 24B4, 26D3, and 37D10, as well as the Fab fragment of the control antibody 8D3, as described in Example 2. [Figure 7]Figure 7 shows the mapping of antibody-binding epitopes to the protease-like domain of hTfR1 by selective antibody binding on ELISA plates coated with human, mouse, or one of three different chimeric human / mouse TfR1 receptors, as described in Example 2. Antibodies 24B4, 26D3, and 37D10 bind to hTfR1 (A) but not to mTfR1 (B). Furthermore, 24B4, 26D3, and 37D10 also bind to the h / m protease-like domain chimeric receptor (D) but not to any of the plates coated with the other chimeric receptors (C and E). [Figure 8] Figure 8 illustrates the epitope binning assay described in Example 2, which consists of the following four main steps: Step 1 - Immobilization of bio-TfR1 onto the sensor chip; Step 2 - Washing of unbound material; Step 3 - Binding of competing binders to TfR1; Step 4 - Association of binders to the TfR1:binder complex formed in Step 3. The data from Step 4 determines whether the two binders under investigation compete for binding to hTfR1. [Figure 9A] Figure 9 shows the results of the epitope binning assay described in Example 2, illustrating the degree of competition among antibodies that simultaneously bind to hTfR1. (A) Antibody 26D3 binds to any of the antibodies shown, along with a pre-formed hTfR1 complex. The binding response of all antibodies was normalized by the binding response measured against free hTfR1 (without competing antibodies). [Figure 9B] (B) Binding of antibody 24B4 to any of the antibodies shown, which are pre-formed hTfR1 complexes. The binding response of all antibodies was normalized to the binding response measured against free hTfR1 (no competing antibodies). [Figure 9C] (C) Binding of control antibody 15G11-1 to any of the antibodies shown, which were pre-formed hTfR1 complexes. The binding response of all antibodies was normalized to the binding response measured against free hTfR1 (no competing antibodies). [Figure 10]Figure 10 shows the binding of the indicated binder to hTfR1 on the cell surface, as studied as described in Example 2. The Y-axis in both figures shows the mean fluorescence intensity when cells were stained with (A) the entire antibody and (B) the Fab fragment of the indicated binder. No background staining is detected with the negative isotype control IgG (A) or the unrelated Fab fragment Lys128 (B). [Figure 11] Figure 11 shows the results of the competitive analysis of ferritin and transferrin against the indicated binder, as described in Example 3. The figure shows (A) the MFI of the indicated binder binding to TfR1 expressed on the surface of THP-1 cells, (B) the MFI of ferritin on the cell surface when exposed to the indicated binder and the positive control antibody MA-712 which competes with ferritin, and (C) the MFI of transferrin on the cell surface when exposed to the indicated binder. [Figure 12] Figure 12 is a collection of sensorgrams showing the results of SPR analysis of the original 26D3 and humanized 26D3 as described in the Fab format of Example 4 (h26D3) when bound to hTfR1 and cTfR1, as shown. [Figure 13] Figure 13 shows the results of BLI and ELISA binding studies performed on mouse and humanized versions of 26D3 in scFv format, as described in Example 4. (A) Sensorgram obtained by BLI measurement of binding of the indicated construct to hTfR1. (B) Binding response from ELISA measurement of binding of the indicated construct to coated TfR1. [Figure 14] Figure 14 is a depiction of the X-ray structure of the 26D3-Fab and hTfR1 complex determined as described in Example 5. The names of the chains used in the coordinate file are shown. (A) A precise structure showing the overall folding of the three independent complexes in the asymmetric unit. (B) An example of the electron density (2m|Fo|-D|Fc|) formed at the 1σ level. Protein chains are depicted in cartoon notation, and sugar moieties are shown in stick notation. [Figure 15]Figure 15 shows the ribbon representation of the h26D3-Fab human TfR1 complex as determined by X-ray crystallography, as described in Example 5. h26D3-Fab is shown in dark gray, and hTfR1 in white. The binding surface (epitope / paratope) is enclosed. [Figure 16] Figure 16 shows the surface area representation of hTfR1 with the binding sites of the shown natural ligands ferritin and transferrin, as well as the epitope of the binder 26D3 of this disclosure. Different binding sites and epitopes are indicated by circles around their respective specific sites. [Figure 17A] Figure 17 illustrates the studies on the generation and characterization of the hTfR1-KI mouse model described in Example 6. (A) Schematic diagram of the transgenic hTfR1-KI mouse construct. The extracellular domain of human TFRC was inserted into the mouse Tfrc gene by homologous recombination. [Figure 17B] (B) Quantitative reverse transcription PCR (RT-qPCR) analysis of mouse Tfrc and human TFRC gene expression in the brain (N=3 / genotype). hTfR1-KI mice (gray circles) expressed both human TFRC and mouse Tfrc throughout the brain homogenate, while wild littermates expressed only mouse Tfrc (white). [Figure 17C] (C) Western blot analysis of hTfR1, total TfR1, and hTfR1-KI in the brain. hTfR1-KI animals at 6-8 months (N=5) and 15 months (N=4) express comparable levels of hTfR1 protein. Total TfR1 levels are comparable between hTfR1-KI transgenic animals and wild littermates (N=3). [Figure 18]Figure 18 shows the results of in vivo brain and plasma exposure analyses of various indicated hTfR1 binding molecules in hTfR1-KI transgenic mice, as described in Example 7. (A) Brain exposure 24 hours after intravenous administration of the indicated hTfR1 binding agent. (B) Plasma exposure 24 hours after intravenous administration of the indicated hTfR1 binding agent. (C) Brain:plasma ratio 24 hours after intravenous administration of the indicated hTfR1 binding agent. Negative controls are denoted as "158" and positive controls as "15G11-1". Error bars indicate mean ± SD (n=4 for each construct tested). [Figure 19] Figure 19 shows the results of in vivo brain exposure analysis of various shown hTfR1-binding molecules in hTfR1-KI mice by immunohistochemistry, as described in Example 8. Cortical capillary staining was observed for several binding molecules, including h26D3. The reference hTfR1 binder "15G11-1" and the non-TfR1 binder "Rec158" were used as positive and negative controls, respectively. [Figure 20] Figure 20 shows the BLI sensorgrams for the h26D3 alanine variant described in Example 9. Each variant exhibits a different kinetic profile, suggesting the possibility of creating variants with different affinities to human TfR1 that have specific mutations in the heavy chain or light chain CDR region. [Figure 21] Figure 21 shows a representative SPR sensorgram of the interaction between hTfR1 and cTfR1 and the indicated alanine mutant h26D3, as measured as described in Example 9. [Figure 22] Figure 22 shows the results of indirect ELISA analysis of the binding of hTfR1 and cTfR1 to the alanine variant of h26D3, as measured as described in Example 9. [Figure 23] Figure 23 shows an SPR sensorgram of the interaction between shown alanine variants of h26D3, which was studied as an scFv component in the bispecific protein format described in Example 9. [Figure 24]Figure 24 shows two different Gen 2A constructs designed and produced as described in Example 10. [Figure 25-1] Figure 25 shows the purification results of different Gen 2A constructs prepared and purified as described in Example 11. The monomer content of bispecific binding proteins was high (generally >98%), and they were produced at low mg / l levels. Purification was analyzed using Coomassie blue SDS-PAGE staining. [Figure 25-2] Figure 25 shows the purification results of different Gen 2A constructs prepared and purified as described in Example 11. The monomer content of bispecific binding proteins was high (generally >98%), and they were produced at low mg / l levels. Purification was analyzed using Coomassie blue SDS-PAGE staining. [Figure 26] Figure 26 shows sensorgrams obtained from SPR binding analysis of 14 Gen 2A-binding protein constructs and controls described in Example 12. One sensorgram is shown for each indicated variant, and all constructs are shown to be functional and to bind to hTfR1. [Figure 27] Figure 27 shows cell binding data for the Gen 2A constructs shown, measured as described in Example 13. All tested constructs bound to hTfR1-expressing cells in a similar manner. [Figure 28-1] Figure 28 is a collection of figures showing the CDC measurement results in Ramos cells for the indicated test constructs #1-7 ("VH-first" configuration) as described in Example 14. [Figure 28-2] Figure 28 is a collection of figures showing the CDC measurement results in Ramos cells for the indicated test constructs #1-7 ("VH-first" configuration) as described in Example 14. [Figure 29-1] Figure 29 is a collection of figures showing the CDC measurement results in Ramos cells for the indicated test constructs #8-14 ("VL-first" configuration) as described in Example 14. [Figure 29-2] Figure 29 is a collection of figures showing the CDC measurement results in Ramos cells for the indicated test constructs #8-14 ("VL-first" configuration) as described in Example 14. [Figure 30] Figure 30 is a collection of figures showing the results of the in vivo pharmacokinetic studies described in Example 15. Plasma and brain exposures are shown from terminal samples taken from n=3 mice at 4, 24, 72, 168, and 240 hours after intravenous administration of the indicated test constructs. Data are presented as mean ± SD. [Figure 31] Figure 31 shows the plasma concentration-time profiles collected sequentially from n=3 mice for each test construct after intravenous administration, as described in Example 15. The data are presented as mean ± SD. [Figure 32] Figure 32 shows 40x z-stack images of the cerebral cortex of hIgG-stained hTfR-KI mice, described 24 hours after administration of Example 16. Three replicates (n=3) are shown for each group (LC1, HC6, and LC5). The perfusion score (0-3) for each brain is indicated by a white square in the lower left corner. Blood vessels are indicated by arrows. [Figure 33] Figure 33 shows representative 63x z-stack images of each group of hIgG staining versus collagen IV, as described in Example 16. The perfusion score (0-3) for each brain is indicated by a white square in the lower left corner. [Figure 34] Figure 34 is a pair of figures showing the results of the in vivo study described in Example 17. The concentrations of the indicated test construct in plasma (A) and brain (B) at 24 hours at three indicated doses: 11 nmol / kg (circle), 40 nmol / kg (triangle), and 60 nmol / kg (inverted triangle). The levels of 2A2#2-8D3 are based on a standard prepared from 2A3#2-WT. [Figure 35A]Figure 35 shows the results of the cytokine response analysis described in Example 17, presenting the individual and median plasma levels of the indicated cytokines (A: TNF and KC / GRO) for the indicated test constructs before and 2 hours after administration of three indicated doses: 11 nmol / kg (diamond), 40 nmol / kg (triangle), and 60 nmol / kg (inverted triangle). [Figure 35B] Figure 35 shows the results of the cytokine response analysis described in Example 17, presenting the individual and median plasma levels of the indicated cytokines (B: IL-6 and IL-10) for the indicated test constructs before and 2 hours after administration of three indicated doses: 11 nmol / kg (diamond), 40 nmol / kg (triangle), and 60 nmol / kg (inverted triangle). [Figure 36] Figure 36 shows the production and purification results of the Gen 2D construct after gel analysis of the purified construct by in-gel protein detection staining using Coomassie blue, as described in Example 18. The lanes are as follows: (1) Marker; (2) mAb158-2D-h26D3-HC6, unreduced; (3) mAb158-2D-h26D3-HC6, reduced; (4) mAb158-2D-h26D3, unreduced; (5) mAb158-2D-h26D3, reduced. [Figure 37] Figure 37 shows sensorgrams from SPR binding analysis of the Gen 2D binding protein constructs and controls described in Example 19. Sensorgrams of the binding of each shown variant to both hTfR1 and cTfR1 are shown, and all constructs are shown to be functional and binding to both hTfR1 and cTfR1. The Fab fragment of h26D3 is included as a positive control. [Figure 38]Figure 38 shows the results of the ADCC assay using rituximab as a positive control, as described in Example 20. (A) Verification of the assay setting using rituximab, which induces strong folding in the presence of target cells (labeled "Rituximab" in the figure). In the absence of target cells, there is no ADCC induction by rituximab (labeled "Rituximab (w / o Target Cells)" in the figure). Similarly, the mAb158 antibody, i.e., an antibody without binding to the hTfR1 binding moiety M1, was tested in the presence or absence of target cells. (B) Analysis of the Gen 2D format, which does not show ADCC induction, with h26D3 scFv in a "VL-first" configuration as M1 and mAb158 as M2. In the same assay, rituximab showed strong induction, but the antibody mAb158 alone showed no indication of ADCC activity. [Figure 39] Figure 39 shows the cell binding data of the demonstrated Gen 2D construct in K562 cells, measured as described in Example 20. [Figure 40] Figure 40 shows the results of the in vivo pharmacokinetic study described in Example 21. A: Mean (±SD) terminal plasma (black) and brain (white) concentration-time profiles (n=3-5 for each sampling time point) for mAb000-Gen2D-h26D3-HC6 (diamond) and mAb000 (square) after a single intravenous administration of 40 nmol / kg to hTfR-KI mice. B: Mean (±SD) plasma concentration-time profiles (n=5) from continuous sampling for mAb000-Gen2D-h26D3-HC6 (diamond) and mAb000 (square) after a single intravenous administration of 40 nmol / kg to hTfR-KI mice. [Figure 41A] Figure 41 shows 40x Z-stack images of the frontal cortex of 5XFAD x hTfR-KI mice 72 hours after administration of mAb000-Gen2D-h26D3-HC6(A) as described in Example 22. Representative images of total amyloid-beta and hIgG from two replicates (n=2) per group. [Figure 41B]Figure 41 shows 40x Z-stack images of the frontal cortex of 5XFAD x hTfR-KI mice 72 hours after administration of mAb000(B) as described in Example 22. Representative images of total amyloid-beta and hIgG from two replicates (n=2) per group. [Figure 42] Figure 42 shows the competition among the indicated test constructs for hTfR1 binding by the antibody M-A712 on the cell surface, as measured by flow cytometry as described in Example 23. [Figure 43] Figure 43 shows the competition by human ferritin for hTfR1 binding by the antibody M-A712 on the cell surface, as measured by flow cytometry as described in Example 23. [Modes for carrying out the invention]

[0017] In one embodiment, M2 is selected from the group consisting of antibodies and Fc fusion proteins. In a particular embodiment, M2 is an antibody. In another particular embodiment, M2 is an Fc fusion protein.

[0018] In one embodiment, the Fc domain of M2 has the ability to induce an Fc-mediated response, such as an Fc-mediated cytotoxic response, when administered to humans. The binding protein of the first embodiment is designed in such a way that it reduces this Fc-mediated response. While we do not wish to be constrained by theory, it is thought that the binding of M1 to the epitope on the protease-like domain of hTfR1, upon binding to hTfR1, orients the entire binding protein, including M1 and M2, such that when fixed to the cell surface, the Fc domain is detached from the environment and oriented toward the space created beneath and / or to the sides of the TfR1 protein. In this way, the interaction between the Fc domain and the Fc receptor necessary for the Fc-mediated response to occur is prevented or reduced. This effect is thought to occur only when the binding protein is bound to hTfR1. As a result, once the binding protein crosses the BBB and enters the environment, if the Fc domain of the binding protein has the intended function and desired effect, this Fc-mediated effect becomes functional as needed.

[0019] hTfR1 binding part M1 As described above, in a first embodiment, the disclosure provides a binding protein in which portion M1 is a human transferrin receptor 1 (hTfR1) binding portion and which has the ability to selectively bind to an epitope located in the protease-like domain of hTfR1 as defined by amino acid residues 121-183 and 384-605 of SEQ ID NO: 66. Although we do not wish to be bound by theory, the binding of hTfR1 to an epitope or binding site within the protease-like domain is considered to offer advantages in that it avoids the drawbacks associated with known binders for hTfR1, particularly those known binders that have affinity for epitopes or binding sites located in the apical domain of TfR1.

[0020] The binding of M1 to hTfR1 in the protease-like domain epitope is shown in Figure 1, which is a surface density model of the structure between the VH / VL pairs of the exemplary M1 module described herein, determined by X-ray crystallography as shown in Example 5. The pairing of the VH and VL regions and the binding orientation of the M1 portion on hTfR1, as shown in the M1 portion bound to the left side of the hTfR1 homodimer, provide four peptide chain ends that allow the binding of a second portion M2 via a linker between M1 and M2.

[0021] In certain embodiments, the epitope or binding site of the hTfR1 binding moiety includes amino acid residues 150, 151, 154, 158, 159, 161, 163, and 385 of SEQ ID NO: 66. In other embodiments, the epitope or binding site of the hTfR1 binding protein of the present disclosure consists of amino acid residues 150, 151, 154, 158, 159, 161, 163, and 385 of SEQ ID NO: 66. In alternative, certain embodiments, the epitope or binding site of the hTfR1 binding protein includes or consists of at least one, at least two, at least three, at least four, at least five, at least six, at least seven, or all eight amino acid residues 150, 151, 154, 158, 159, 161, 163, and 385 of SEQ ID NO: 66. As shown in the following example, referring to Figure 16, for instance, this embodiment relating to the epitopes of the binding proteins identified and disclosed herein ensures that the binding is non-interfering to the natural hTfR1 ligands transferrin and ferritin.

[0022] As is known to those skilled in the art, an epitope (or antigenic determinant) is a group of amino acids or other chemical groups exposed on the surface of a molecule or, in many cases, a protein, here hTfR1, that can generate an antigenic reaction and bind to an antibody. An epitope is a localized region on the surface of an antigen that is recognized by the immune system, particularly by an antibody. A conformational epitope consists of adjacent amino acid residues located on the surface structure of an antigen protein. Conformational epitopes bind to their complementary paratopes in B cell receptors and / or antibodies. In one embodiment of this disclosure, an epitope bound by a binding molecule is a conformational epitope.

[0023] In one embodiment, the bond to hTfR1 by the M1 bond is monovalent.

[0024] As described above, the first portion M1 contains a VH / VL pair having an antigen-binding surface. For clarity, the term "VH / VL" used in relation to the VH / VL pair does not restrict the configuration to a specific order of the VH and VL regions in the polypeptide chain, but is used solely to indicate that both VH and VL regions are present and can be associated in pairs with the formation of an Ig domain having an antigen-binding site. Therefore, the term VH / VL pair includes, for example, constructs in which the VL region precedes the VH region in a single-chain Fv, constructs in which the VH region precedes the VL region in a single-chain Fv, and constructs in which the VH and VL regions are associated with each other non-covalently. In certain embodiments of the binding protein, the VH / VL pair in M1 is arranged such that the VL region precedes the VH region in a single-chain Fv construct.

[0025] The VH / VL pair contained in M1 includes an antigen-binding surface. In one embodiment, the antigen-binding surface is composed of three complementarity-determining regions (CDRs) from each of the VH and VL regions. In one embodiment, the CDRs include the following amino acid sequence: VHCDR1:X1X2NMX3 (Sequence ID 1) Here, X1 is selected from D and A, X2 is selected from Y and A, and X3 is selected from D and A. VHCDR2:X4INPX5X6X7TTSX8NEKFKG(Sequence ID 2) Here, X4 is selected from D and A, X5 is selected from D and A, X6 is selected from Y and A, X7 is selected from D and A, and X8 is selected from Y and A. VHCDR3:GGX9SGSSX10X11HPMX12X13 (Sequence ID 3) Here, X9 is selected from Y and A, X10 is selected from Y and A, X11 is selected from Y and A, X12 is selected from D and A, and X13 is selected from Y and A. VLCDR1:KSSQSLLX14STNQKNX15LA (Sequence ID 4) Here, X14 is selected from Y and A, and X15 is selected from Y and A. VLCDR2:X16ASTRES(Sequence ID 5) Here, X16 is selected from W and A. VLCDR3:QQX17FIX18PRT (Sequence ID 6) Here, X17 is selected from Y and A, and X18 is selected from Y and A.

[0026] In one embodiment, the amino acid sequence of VHCDR1 is selected from the group consisting of SEQ ID NOs: 7 and 13-15.

[0027] In one embodiment, the amino acid sequence of VHCDR2 is selected from the group consisting of SEQ ID NOs: 8 and 16-20.

[0028] In one embodiment, the amino acid sequence of VHCDR3 is selected from the group consisting of SEQ ID NOs. 9 and 21-25.

[0029] In one embodiment, the amino acid sequence of VLCDR1 is selected from the group consisting of SEQ ID NOs: 10, 26, and 27.

[0030] In one embodiment, the amino acid sequence of VLCDR2 is selected from the group consisting of SEQ ID NOs: 11 and 28.

[0031] In one embodiment, the amino acid sequence of VLCDR3 is selected from the group consisting of SEQ ID NOs: 12, 29, and 30.

[0032] In some embodiments, the CDR sequences can be freely combined from the options listed above. Such embodiments include, but are not limited to, the combinations exemplified in Example 9 for a typical M1 portion h26D3 alanine substitution mutant.

[0033] In a particular embodiment of the binding protein disclosed herein, the CDR sequence of the antigen-binding surface of binding moiety M1 is as follows: VHCDR1:DYNMD (Sequence ID 7) VHCDR2:DINPDYDTTSYNEKFKG(Sequence ID 8) VHCDR3:GGYSGSSYYHPMDY (Sequence ID 9) VLCDR1:KSSQSLLYSTNQKNYLA (Sequence ID 10) VLCDR2:WASTRES (Sequence ID 11) VLCDR3:QQYFIYPRT(Sequence ID 12) That is the case.

[0034] In another specific embodiment of the binding protein disclosed herein, the CDR sequence of the antigen-binding surface of binding moiety M1 is as follows: VHCDR1:DYNMD (Sequence ID 7) VHCDR2:DINPDADTTSYNEKFKG (Sequence ID 18) VHCDR3:GGYSGSSYYHPMDY (Sequence ID 9) VLCDR1:KSSQSLLYSTNQKNYLA (Sequence ID 10) VLCDR2:WASTRES (Sequence ID 11) VLCDR3:QQYFIYPRT(Sequence ID 12) That is the case.

[0035] In one embodiment, the CDR sequence of the antigen-binding interface contained in the binding protein of this disclosure is defined using the Kabat method well known to those skilled in the art of antibody technology (see, for example, Kabat (1991), Sequences of Proteins of Immunological Interest, 5th edition, NIH Publication no. 91-3242).

[0036] In one embodiment, the VH region in the VH / VL pair of M1 is as follows: (i) The group consisting of sequence numbers 31 to 44, for example, the group consisting of sequence numbers 31 and 37; and, (ii) A sequence is provided which has at least 80%, at least 90%, at least 92%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% identity with the sequence defined in (i), and the sequence of the CDR region is 100% identical to the sequence defined in (i). It contains or consists of an amino acid sequence selected from the following.

[0037] In one embodiment, the VH region in the VH / VL pair of M1 is as follows: (i) The group consisting of sequence numbers 45 to 51; and, (ii) A sequence is provided which has at least 80%, at least 90%, at least 92%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% identity with the sequence defined in (i), and the sequence of the CDR region is 100% identical to the sequence defined in (i). It contains or consists of an amino acid sequence selected from the following.

[0038] In certain embodiments, the VH region and the VL region are both as defined immediately above, namely, VH includes or consists of sequences selected from sequence numbers 31-44 and sequences having at least 80% sequence identity thereto, and VL includes or consists of sequences selected from sequence numbers 45-51 and sequences having at least 80% sequence identity thereto.

[0039] In one embodiment, the VH region includes sequence number 31, and the VL region includes sequences selected from sequence numbers 45 to 51.

[0040] In one embodiment, the VH region includes an array selected from sequence numbers 31 to 44, and the VL region includes sequence number 45.

[0041] In one embodiment, the VH region includes sequence number 31, and the VL region includes sequence number 45.

[0042] In one embodiment, the VH region includes sequence number 37, and the VL region includes sequence number 45.

[0043] In certain embodiments, the VH and VL sequences of the binding molecule are selected from any one of the listed sequences and sequences having at least 80%, at least 90%, at least 92%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, or at least 100% identity thereto.

[0044] Combination of M1 and M2 In one key embodiment of the binding protein in this disclosure, the M1 portion contains scFv. In other words, the VH / VL pair of M1 forms part of scFv, and the VH and VL regions are linked to each other by a peptide scFv linker. In such an embodiment, the scFv linker binds to the C-terminal amino acid residue of the VH region and the N-terminal amino acid residue of the VL region, or to the C-terminal amino acid residue of the VL region and the N-terminal amino acid residue of the VH region. In the first configuration, the VH region precedes the VL region in the polypeptide chain constituting scFv, while in the second configuration, the VL region precedes the VH region. The two different configurations are denoted as "VH-first" and "VL-first," and are illustrated in Figures 2A and 2B, respectively, by exemplary structures of the M1 binding portion h26D3 complexed with hTfR1. As shown in Figure 2A, in the "VH-first" configuration, the scFv linker between VH and VL is located below portion M1, near the cell membrane surface. This allows M2 to be bound to M1 using the N-terminal amino acid residue of VH (VH-N) or the C-terminal amino acid residue of VL (VL-C), or both. As shown in Figure 2B, in the "VL-first" configuration, the scFv linker between VH and VL is located above partial M1 and further away from the cell membrane surface. This allows M2 to be bound to M1 using the C-terminal amino acid residue of VH (VH-C) or the N-terminal amino acid residue of VL (VL-N), or both. When the M1 binding portion is used as scFv in the "VL-first" configuration, or in any setting where the VH-C and / or VL-N binding sites are actually available, the Fc domain containing the second partial M2 is oriented adjacent to the cell membrane in the space directly beneath the bulk of the hTfR1 homodimer through binding to one or both of these binding sites (schematically shown in Figure 3, again referring to the X-ray structure of the hTfR1 and h26D3 complex). While we don't want to be constrained by theory, this suggests that when hTfR1 is bound, the antibody may reduce the likelihood of harmful interactions with other circulating components.

[0045] The design and selection of appropriate peptide linkers to be used within and between domains and portions of fusion proteins, antibody constructs, and other such designed polypeptides are within the capabilities of those skilled in the art. In some embodiments in which M1 comprises or consists of scFv, the scFv linker is a flexible peptide linker comprising 5 to 40 amino acid residues, e.g., 10 to 30 amino acid residues, e.g., 15 to 25 amino acids, e.g., approximately 15 amino acid residues, e.g., 15 amino acid residues, e.g., sequence (G4S)3 (SEQ ID NO: 88).

[0046] The same or similar design considerations also apply to the linker used to bind the hTfR1 binding portion M1 to the second portion M2. In one embodiment, the at least one peptide linker between M1 and M2 binds, on the M1 side, to the C-terminal amino acid residue of the VH region of M1, or to the N-terminal amino acid residue of the VL region of M1.

[0047] In one such embodiment, the at least one linker between M1 and M2 binds to the C-terminal residue of the CH3 region of the Fc domain on the M2 side, and to the N-terminal amino acid residue of the VL region of M1 on the M1 side. One example of this embodiment has a full-length antibody as M2, which is hereby denoted as "Gen 2D". The "Gen 2D" design is shown in the right-hand panel of Figure 4. As shown in the figure, in this design, the M1 portion bound to hTfR1 is fused to the C-terminus of one antibody heavy chain of M2, and pairs with the other heavy chain without an hTfR1 binder. Along with two copies of the M2 antibody light chain, the complete construct is formed with one hTfR1 binder bound to a normal Y-shaped antibody structure. In the example where scFv with a "VL-first" setting was tested as M1, i.e., when the portion M1 to which hTfR1 is bound is an scFv fused to the antibody via the VL-N site, no detectable antibody-dependent cell-mediated cytotoxicity (ADCC) was observed in the assay shown in Example 18 (Figure 32B).

[0048] In other embodiments of this kind, M2 comprises an antibody having two antibody light chains, with M1 and M2 linked to each other via two peptide linkers, the first linker bound to the C-terminal amino acid residue of the first light chain of M2 on the M2 side and to the N-terminal amino acid residue of the VL region of M1 on the M1 side, and the second linker bound to the N-terminal amino acid residue of the second light chain of M2 on the M2 side and to the C-terminal amino acid residue of the VH region of M1 on the M1 side. One example of this embodiment has a full-length antibody as M2, which is hereby denoted as "Gen 2A". The design of "Gen 2A" is shown in the left panel of Figure 4 and is described in detail in WO2022 / 258841 (incorporated by reference). By ligating a single-chain hTfR1-bound portion M1 to the C-terminus of one light chain of M2 and the N-terminus of the other light chain of M2, a symmetrical construct consisting of only two different polypeptide chains is created. In an example where an scFv with a "VL-first" setting was tested as M1 in this "Gen 2A" format, i.e., an scFv fused to an antibody via both the VL-N and VH-C sites, no detectable complement-dependent cell-mediated cytotoxicity (CDC) was observed in the assay shown in Example 14 (Figure 29).

[0049] As described above, the design and selection of appropriate peptide linkers used within and between domains and portions of fusion proteins, antibody constructs, and other such designed polypeptides are within the capabilities of those skilled in the art. In one embodiment of the binding protein of this disclosure, M1 and M2 are linked by at least one flexible peptide linker. In one embodiment, the at least one flexible peptide linker comprises glycine, serine, alanine, and / or threonine residues. In a more specific embodiment, the linker is (G n S m ) p and (S n G m ) pThere is a general formula selected from, where independently, n=1 to 7, m=0 to 7, n+m≦8, and p=1 to 10. In some embodiments, at least one linker is 10 to 50 amino acid residues long, e.g., 10 to 30 amino acid residues long, e.g., 15 to 25 amino acid residues long, or 10 to 20 amino acid residues long. If M1 and M2 are linked via two or more linkers, all of the disclosed linker designs apply individually to each linker, which exists independently of the other linkers. Thus, for example, if there are two linkers, their lengths may be the same or different, and they may have the same or different amino acid sequences.

[0050] Second part M2 With respect to the second portion M2 in the binding protein of this disclosure, it is selected from an antibody and an Fc fusion protein. Through the presence of its Fc domain, M2 is capable of eliminating Fc-mediated responses such as Fc-mediated cytotoxic responses. In one embodiment, such a cytotoxic response is selected from the group consisting of antibody-dependent cell-mediated cytotoxicity (ADCC), antibody-dependent cell-mediated phagocytosis (ADCP), complement-dependent cell-mediated cytotoxicity (CDC), and combinations thereof. In one embodiment, the response is selected from ADCC, CDC, and combinations thereof. In a particular embodiment, the response is ADCC. In another particular embodiment, the response is CDC.

[0051] In one embodiment of the binding protein of this disclosure, M2 is an antibody capable of selectively binding to a target present in the mammalian brain. In some embodiments, the target is selected from the group consisting of amyloid-β peptide or its derivatives or fragments, alpha-synuclein or its derivatives or fragments, TAR DNA-binding protein 43 (TDP-43) or its derivatives or fragments, trigger receptor (TREM2) expressed in bone marrow cells 2, beta-secretase 1 (BACE1), superoxide dismutase (SOD), huntingtin, transthyretin, P-secretase 1, epidermal growth factor, epidermal growth factor receptor 2, Tau, phosphorylated Tau or its fragments, apolipoprotein E4, CD20, prion protein, leucine-rich repeat kinase 2, parkin, presenilin 2, gamma-secretase, cell death receptor 6, amyloid-β precursor protein, p75 neurotrophin receptor, neuregulin, and caspase 6. In a further specific embodiment, the target is selected from the group consisting of amyloid-β peptide or its derivatives or fragments, alpha-synuclein or its derivatives or fragments, TAR DNA-binding protein 43 (TDP-43) or its derivatives or fragments, trigger receptor (TREM2) expressed in bone marrow cells 2, Tau, phosphorylated Tau or its fragments, and apolipoprotein E4. In a further specific embodiment, the target is selected from the group consisting of amyloid-β peptide or its derivatives or fragments, alpha-synuclein or its derivatives or fragments, TAR DNA-binding protein 43 (TDP-43) or its derivatives or fragments.

[0052] In one embodiment of the binding protein of the present disclosure, M2 is an antibody capable of selectively binding to a target present in the mammalian brain, wherein the antibody is selected from the group of anti-Aβ antibodies, such as lecanemab, gantenerumab, aducanumab, donanemab, PBD-C06, and KHK6640.

[0053] In another embodiment of the binding protein of this disclosure, M2 is an antibody capable of selectively binding to a target present in the mammalian brain, wherein the antibody is selected from the group of anti-alpha-synuclein antibodies, such as pracinezumab, UCB7853, Lu AF82422, TAK-341, and BAN0805.

[0054] Affinity to the target As used herein, the terms “specific binding to X,” “selective binding to X,” and “affinity to X” refer to properties of the binding protein, such as the antibody or its antigen-binding fragment, or the properties of a bispecific or multispecific construct incorporating such an antibody or its antigen-binding fragment, where X is the target (e.g., an antigen or epitope such as TfR1 bound by a VH / VL pair at the M1 portion of the binding protein described above), and which can be tested by methods such as ELISA, surface plasmon resonance (SPR), or biolayer interferometry (BLI). Those skilled in the art are aware of these methods and others.

[0055] For example, the binding affinity to a target, antigen, or epitope X can be tested in an experiment in which the binding protein to be tested is captured on an ELISA plate coated with a molecule containing X or epitope X, a biotinylated detection antibody is added, followed by the addition of streptavidin-conjugated horseradish peroxidase (HRP). Alternatively, the detection antibody can be directly conjugated with HRP. A tetramethylbenzidine (TMB) substrate is added, and the absorbance at 450 nm is measured using an ELISA multiwell plate reader. Those skilled in the art will then interpret the results obtained from such an experiment to establish at least a qualitative measurement of the binding affinity of the binding protein to X. If quantitative measurement is desired, ELISA can also be used, for example, to determine the EC50 value (half of the maximum effective concentration) for the interaction. The response of the binding protein to a dilution series of X can be measured using ELISA as described above. Those skilled in the art can then interpret the results obtained from such experiments and calculate the value of EC50 from the results, for example, using GraphPad Prism v.9 and nonlinear regression.

[0056] As used here, the term "EC50" refers to half of the maximum effective concentration of the binding protein, which induces an intermediate response between baseline and maximum after a particular exposure time.

[0057] Alternatively, inhibitory ELISA is used to obtain a quantitative measurement of the interaction by determining the IC50 (median inhibitory concentration). In inhibitory ELISA measures the concentration of target X in a liquid sample by detecting interference in the expected signal output. As a rule, a known target or epitope-related substrate is used to coat a multi-well plate. In parallel, a binding protein with putative affinity to X is added and incubated with a solution containing various concentrations of the target. Following the usual blocking and washing steps, a sample containing the mixture of the binding protein and target is added to the wells. A labeled detection antibody with affinity to the binding protein is then applied to detection using the relevant substrate (e.g., TMB). As a rule, if the target concentration in the liquid sample is high, a significant decrease in signal output will be observed. In contrast, if the target is very low in the liquid sample, the expected decrease in signal output will be very small. Those skilled in the art will recognize that the signal output also depends on the affinity of the binding protein to the target.

[0058] As used here, the term "IC50" refers to the half-inhibitory inhibitory concentration of a binding protein, which induces an intermediate response between baseline and maximum inhibition after a specific exposure time. Here, a lower IC50 value indicates that a lower concentration of the target is required to prevent the detection antibody from binding to the known target covering the plate, compared to a higher IC50 value. Therefore, a lower IC50 value typically corresponds to higher affinity.

[0059] The binding affinity of a binding protein is also tested by surface plasmon resonance (SPR). For example, the affinity is tested in an experiment where the target or epitope X is immobilized on the sensor chip of the device and a sample containing the binding protein to be tested is passed over the chip. Alternatively, the binding protein to be tested can be immobilized on the sensor of the device and tested by passing a sample containing X over the chip. A person skilled in the art will then interpret the results obtained from such an experiment to establish at least a qualitative measurement of the binding affinity of the binding protein for X. If a quantitative measurement is desired, for example, SPR is also used to determine the D K value of the interaction. The binding values can be defined, for example, using a Biacore (Cytiva) or ProteOn XPR 36 (Bio-Rad) device. The target or epitope is appropriately immobilized on the sensor chip of the device, and the binding protein sample for determining the affinity is prepared by serial dilution and injected. The K D value can be calculated from the results using, for example, the 1:1 Langmuir binding model of Biacore Insight Evaluation Software 2.0 or other suitable software typically provided by the device manufacturer.

[0060] Binding affinity can also be measured by biolayer interferometry (BLI), a label-free technique for measuring biomolecule interactions in an interactome. This is an optical analytical technique that analyzes the interference pattern of white light reflected from two surfaces: a layer of immobilized protein on a biosensor chip and an internal reference layer. The binding between a ligand (target or epitope X) immobilized on the surface of the biosensor chip and an analyte in solution (e.g., a binding protein with a putative affinity for X) causes an increase in the optical thickness in the biosensor chip, resulting in a wavelength shift Δλ, which is a direct measurement of the change in the thickness of the biological layer. The interaction is measured in real time and provides precise and accurate capabilities for observing binding specificity, binding and dissociation rates, or concentrations.

[0061] Those skilled in the art are aware of the above and other methods for measuring the affinity of a protein bound to a target or epitope X qualitatively, quantitatively, or both.

[0062] Pharmaceutical composition In a second aspect, the disclosure provides a pharmaceutical composition comprising the binding protein shown herein and at least one pharmaceutically acceptable excipient or carrier.

[0063] The technology for formulating polypeptides such as antibodies and their derivatives for human therapeutic use is well known in this field, and is outlined, for example, in Wang et al. (2007), J Pharm Sci, 96:1-26, the entire content of which is incorporated here.

[0064] pharmaceutically acceptable excipients that can be used in the formulation of compositions include, but are not limited to, ion exchangers, alumina, aluminum stearate, lecithin, serum proteins such as human serum albumin, buffer substrates such as phosphates, glycine, sorbic acid, potassium sorbate, partial glyceride mixtures of saturated vegetable fatty acids, water, salts or electrolytes such as protamine sulfate, disodium hydrogen phosphate, potassium hydrogen phosphate, sodium chloride, zinc salts, colloidal silica, magnesium trisilicate, polyvinylpyrrolidone, cellulosic substrates (e.g., sodium carboxymethylcellulose), polyethylene glycol, polyacrylates, waxes, polyethylene-polyoxypropylene block polymers, polyethylene glycol, and wool fats.

[0065] In certain embodiments, the pharmaceutical composition is formulated for administration to a subject via any suitable route of administration, including but not limited to intramuscular, intravenous, intradermal, intraperitoneal injection, subcutaneous, epidural, nasal, oral, rectal, topical, inhalation, buccal (e.g., sublingual), and transdermal administration. In preferred embodiments, the composition is formulated for intravenous or subcutaneous administration.

[0066] Methods of prevention, treatment, diagnosis, prognosis, and detection The binding proteins described herein are useful as therapeutic, prophylactic, diagnostic, and / or prognostic agents.

[0067] Accordingly, in a further aspect of this disclosure, a binding protein as described in the first aspect or a pharmaceutical composition as described in the second aspect is provided for use as a pharmaceutical agent.

[0068] In yet another aspect of this disclosure, a binding protein as described in the first aspect or a pharmaceutical composition as described in the second aspect is provided for use as a diagnostic agent.

[0069] In yet another aspect of this disclosure, a binding protein as described in the first aspect or a pharmaceutical composition as described in the second aspect is provided for use as a prognostic agent.

[0070] Furthermore, methods are provided for administering the binding proteins disclosed herein to subjects who require them, typically human subjects, for the prevention, treatment, or diagnosis of disease, or for evaluating the prognosis of disease.

[0071] Also provided is the use of the disclosed binding proteins for the manufacture of compositions (e.g., pharmaceuticals) for use in the prevention, treatment, diagnosis, and / or prognosis of any one of the listed diseases.

[0072] Therefore, in one embodiment, a binding protein, or a pharmaceutical composition containing the same, is useful in the treatment, prevention, diagnosis, and / or prognosis of neurodegenerative disorders, such as Alzheimer's disease and other disorders associated with Aβ protein aggregation, traumatic brain injury (TBI), Lewy body dementia (LBD), Down syndrome (DS), amyotrophic lateral sclerosis (ALS), frontodontic dementia, tauopathy, systemic amyloidosis, atherosclerosis, Parkinson's disease (PD), Parkinson's disease dementia (PDD), Lewy body degeneration of Alzheimer's disease, multiple system atrophy, psychosis, schizophrenia, Creutzfeldt-Jakob disease, Huntington's disease, and familial amyloid neurological disorders.

[0073] In more specific embodiments, the disease is selected from Alzheimer's disease and other disorders associated with Aβ protein aggregation, Lewy body dementia (LBD), Down syndrome (DS), amyotrophic lateral sclerosis (ALS), frontal dementia, tauopathy, Parkinson's disease (PD), Parkinson's disease dementia (PDD), and Lewy body degeneration of Alzheimer's disease.

[0074] In more specific embodiments, the disease is selected from Alzheimer's disease and other Aβ protein aggregation-related disorders, Lewy body dementia (LBD), amyotrophic lateral sclerosis (ALS), and Parkinson's disease (PD), particularly Alzheimer's disease.

[0075] In alternative embodiments, the binding protein, or a pharmaceutical composition containing the same, may be useful in the treatment, prevention, diagnosis, and / or prognosis of other disorders, such as brain cancer, multiple sclerosis, and lysosomal storage disorders.

[0076] In another embodiment, a method for treating, preventing, diagnosing, and / or prognosing a disorder is provided, the method comprising administering to the mammal an amount such as a therapeutically effective dose of a binding protein, or a pharmaceutical composition containing the same.

[0077] Built-in by reference Various publications are cited in this application, and each of them is incorporated herein by reference in its entirety. [Examples]

[0078] Examples While the present invention has been described with reference to various exemplary aspects and embodiments, it will be understood by those skilled in the art that various modifications may be made and their elements replaced with corresponding ones without departing from the scope of the invention. Furthermore, many modifications may be made to adapt the teachings of the invention to specific situations or molecules without departing from the essential scope of the invention. Thus, the invention is not limited to any particular embodiment, but is intended to include all embodiments that fall within the scope of the appended claims.

[0079] The present invention will be further illustrated by the following non-limiting embodiments, which are provided for illustrative purposes only and are not intended to limit the invention in any way. Those skilled in the art will readily recognize various non-essential parameters that can be changed or modified to obtain essentially the same results. With regard to the figures used (e.g., quantities, temperatures, etc.), efforts have been made to ensure accuracy, but some experimental errors and deviations may exist. Unless otherwise indicated, the implementation of the present invention employs conventional methods of protein chemistry, biochemistry, recombinant DNA technology, and pharmacology in the art. Such techniques are fully described in existing literature. Furthermore, it will be apparent to those skilled in the art that the protein engineering techniques applied herein are also applicable to other constructs described herein and which the inventors assume to be within the scope of this disclosure.

[0080] Example 1 Identification of human TfR1 binders by immunization and screening. Immunization and Hybridoma Screening To identify monoclonal antibodies that bind to human transferrin receptor 1 (hTfR1), four 6-10 week old Balb / c or C57BL / 6 mice were immunized by subcutaneous injection of an immunogen with an adjuvant. The hTfR1 immunogen consisted of the ectodomain of the human TfR1 protein, which was N-terminally fused to the tetanus toxin-derived T cell epitope P2 (Kovacs-Nolan and Mine (2006), Biochim Biophys Acta 1760:1884-1893) via a GSS linker, and an N-terminal 10× histidine tag (His 10 Designed to include -P2-hTfR1 (sequence number 52). Following gene construction, recombinant His 10 -P2-hTfR1 protein is Expi293 TM The TfR1 immunogen was prepared by transient transduction in Hek293 cells using an expression system (Gibco), purified using a nickel column (HisTrap FF, catalog number 17-5255-01, GE Healthcare), buffered in PBS, and concentrated to 1 mg / ml. The expressed TfR1 immunogen was aliquoted and stored at -80°C until use. Quil-A adjuvant (vac-quil, InvivoGen) was used in all immunizations except for the final supplemental immunization injection which did not contain the adjuvant. Before use, Quil-A was resuspended in ddH2O at a concentration of 1 mg / ml, filtered, sterilized, and aliquoted into 0.1 ml portions and stored at -80°C. Quil-A was administered at a dose of 10 μg / mouse.

[0081] Animals are recombinantly produced immunogen His 10 -P2-hTfR1 was immunized monthly with Quil-A in combination and co-administered. Three weeks after each immunization, blood samples were collected, and the plasma was analyzed for the presence of antibodies reactive to recombinant human TfR1 and mouse TfR1. The titer was considered sufficiently high when the ELISA response at a 1 / 100,000 dilution exceeded the mean value of the blank (i.e., background) plus the standard deviation of three blanks. The four mice used in this study were each immunized between 4 and 6 times.

[0082] Three days prior to fusion, the mice were given a final intraperitoneal booster immunization injection without adjuvant. The mice were anesthetized with isoflurane before sacrifice. Intact spleens were harvested by dissection after opening the peritoneum. Briefly, a single-cell suspension of immunized mouse spleens was prepared and mixed with Sp2 / 0 cells in a 3:1 ratio. The cells were fused using PEG, and the cells were then fused into ClonaCells. TM-HY Medium D (STEMCELL Technologies) was added to a bottle. Then, 60-70 μl per well was dispensed into a 96-well plate. After 6-7 days, 150 μl of HAT medium was added to each well of the semi-solid 96-well plate. The following day, 120 μl of supernatant was discarded from each well, and 100 μl of fresh HAT medium was added. The next day, 100 μl of supernatant was taken from each well, transferred to a storage plate, and tested for the presence of antibodies against mouse TfR1 using indirect ELISA on a nickel-coated plate as described in the protocol below. Repeated screening of hybridoma plates was performed by adding 120 μl of HAT medium on day 12, and then transferring 25 μl of supernatant to an ELISA plate on day 3 to screen for mouse TfR1 reactivity (both screenings are referred to as "primary screening"). Clones that tested positive for mouse TfR1 with an OD > 0.2 were transferred to 24-well plates and cultured for at least 3 days. Secondary screening was then performed on the reactivity of mouse, human, and cynomolgus monkey TfR1 in solution using biolayer interferometry (referred to as "secondary screening"). Secondary screening showed binding to both hTfR1 and cynomolgus monkey TfR1, while only very weak binding or no binding at all was detected for mTfR1. The supernatant from the 24-well plates was screened for binding to His-tagged hTfR1 and lack of binding to His-tagged amyloid-beta precursor protein (APP; negative control) using both directly coated TfR1 plates and nickel-coated plates, as described below. Binding to cynomolgus monkey TfR1 (cTfR1) was also analyzed using the directly coated TfR1 method. Notably, the ELISA response (OD450 value) was significantly lower for mTfR1 compared to hTfR1 and cTfR1, and all positive clones showed weaker binding to mTfR1 compared to binding to hTfR1 and cTfR1.

[0083] The selected clones were diluted using a limiting dilution assay (LDA) to obtain a single clone. Reactivity to mouse and human TfR1 was re-examined by LDA and ELISA in post-growth monoclonal cultures.

[0084] Indirect ELISA screening ELISA assays were performed according to standard ELISA protocols to screen for immunization-induced reactivity to target antigens in plasma samples, or to identify hybridoma clones that produce antibodies reactive to the TfR1 target protein. In short, 96-well half-area plates (Corning) were immunized with 1 μg / ml His 10 -mTfR1 (SEQ ID NO: 53) or His 10 Coated with -hTfR1 (sequence number 54). 10 -mTfR1 and His 10 -hTfR1 is the above His 10 Recombinant production and purification were performed using the method with the -P2-hTfR1 immunogen. Plates were blocked with 150 μl of protein-free blocking solution (Pierce) per well for 1 hour at room temperature with shaking (600-900 rpm). Plates were then treated with 0.1% TWEEN®-20 and Kathon TMThe plates were washed four times with PBS containing [the specified substance]. Serially diluted plasma samples, starting at a 1 / 450 dilution, or 1 / 2 dilution hybridoma supernatant were added to the plates (50 μl / well; dilution buffer: PBS containing 0.1% BSA and 0.05% TWEEN®-20), incubated at room temperature for 2 hours, and then washed four times. 50 μl of detection antibody (HRP conjugate anti-mouse IgG, Southern Biotech, catalog no. 1030-05, diluted 1 / 5000 with dilution buffer) was added per well, and the plates were incubated at room temperature for 1 hour. After another wash (as described above), 50 μl of TMB substrate (K-Blue® Aqueous, Noegen) was added per well, and the reaction was stopped after 10–15 minutes with 50 μl of 0.5 M H2SO4 per well. The optical density at 450 nm was read using a plate reader (Tecan). As diluted above, the endpoint titer was defined as the mean value of the blank well (background) plus the standard deviation of three blank wells.

[0085] Primary screening of hybridomas that produce antibodies reactive to target proteins was performed using nickel-coated ELISA plates. In short, 96-well nickel-coated plates (PIERCE) supplied with BSA preblocking were treated with 3 μg / ml (100 μl) of His 10 -mTfR1 was incubated overnight at 4°C without shaking. The plates were treated with 0.1% TWEEN®-20 and Kathon TMThe plates were washed four times with PBS containing [component name missing]. A 1 / 4 diluted hybridoma supernatant was added to the plate (dilution buffer: PBS containing 0.1% BSA and 0.05% TWEEN®-20), incubated at room temperature for 2 hours, and then washed four times. 100 μl of detection antibody (HRP conjugate anti-mouse IgG, Southern Biotech, catalog no. 1030-05, diluted 1 / 5000 with dilution buffer) was added per well, and the plate was incubated at room temperature for 1 hour. After washing again (as described above), 100 μl of K-Blue® Aqueous substrate (Neogen) was added per well, and the reaction was stopped after 10-15 minutes with 100 μl of 0.5 M H2SO4 per well. Optical density at 450 nm was read using an ELISA plate reader (Tecan).

[0086] Table 1 shows examples of clones that are considered positive for binding to mouse TfR1 and human TfR1. These clones were confirmed by ELISA to bind to both His-tagged hTfR and cTfR, and to lack binding to His-tagged APP (negative control). The selected clones were further characterized by various assays.

[0087] [Table 1]

[0088] Biolayer interferometry measurement The selected clones were examined using biolayer interferometry (BLI) with an Octet instrument (Octet Red384, ForteBio). The setup used included IgG capture from each clone on individual sensor chips to enable detection of antibodies binding to the target in solution. In addition to providing binding measurements, BLI measurements offer further details about the overall binding characteristics, including on-rate and off-rate projections.

[0089] Figure 5 shows the results of BLI measurements by direct binding measurement in unpurified hybridoma supernatant for three selected clones provided as an example. Briefly, mouse IgG antibody clones in hybridoma supernatant diluted 1:1 with running buffer (PBS, 0.02% TWEEN®-20 and 0.01% BSA) were captured with an anti-mouse capture biosensor (anti-mouse capture, AMC, Molecular devices, catalog no. 18-5580). Next, to establish a baseline signal, the sensor with immobilized IgGs was briefly washed for 10 seconds before incubation in running buffer. Binding to target antigens was measured by incubating the sensor for 120 seconds in wells of an assay plate containing the following concentrations of each target antigen: 500 nM mTfR1, 250 nM hTfR1, and 250 nM cTfR1. All proteins were diluted in running buffer. Target dissociation was measured by incubating the biosensor in running buffer for 90 seconds. All tested clones, namely 24B4, 26D3, and 37D10, bound to both human and cynomolgus monkey TfR1, but very weakly to mouse TfR1. Overall, most clones showed greater cross-reactivity to human and cynomolgus monkey TfR1 than to mouse TfR1.

[0090] Sequence determination of selected clones The target clones were cryopreserved and sequenced by whole transcriptome shotgun sequencing. Among the sequenced hybridoma clones were those designated 26D3, 24B4, and 37D10. The amino acid sequences of the heavy chain variable (VH) and light chain variable (VL) regions of each antibody were obtained, and the complementarity-determining regions (CDRs) of these antibodies were identified using the Kabat definition. The amino acid sequence of the CDR of the selected mouse antibody 26D3 is shown in Table 2 below.

[0091] [Table 2]

[0092] Example 2 In vitro binding and epitope screening of TfR1 in humans and cynomolgus monkeys Further detailed binding analysis using BLI was performed with purified selected antibodies. Binding of Fab fragments derived from mouse antibodies 26D3, 24B4, and 37D10 to human TfR1 and cynomolgus monkey TfR1 was investigated. For example, the BLI instrument Octet Red384 was used to measure binding between immobilized TfR1 in solution and the Fab fragment under test. Antibody binding to TfR1 was measured using TfR1 complexed with human transferrin ligand (Tf). The Tf / TfR1 complex was formed on a streptavidin biosensor by a complex formation step in which either hTfR1 or cTfR1 was captured on the sensor, after first loading a sensor with biotinylated human holotransferrin. The final complex density on the sensor was comparable for both hTfR1 and cTfR1. Antibody binding to TfR1 was measured between a 120-second binding phase and a 300-second dissociation phase. Figure 6 shows sensorgrams of 15 nM 24B4-Fab, 26D3-Fab, and 37D10-Fab, as well as Fab derived from the known TfR1-binding antibody 8D3 (Boado et al. (2009), Biotechnol Bioeng 102:1251-1258). The data showed similar binding profiles to human and cTfR1 for both 24B4-Fab and 26D3-Fab, and cross-reactivity to both species was also detected in 37D10-Fab, but no significant binding of 8D3-Fab to human or cynomolgus monkey TfR1 was detected. Importantly, this experiment showed that 24B4-Fab, 26D3-Fab, and 37D10-Fab all bind to TfR1 when the natural ligand transferrin is complexed with TfR1.

[0093] Next, ELISA experiments showed that antibodies 26D3, 24B4, and 37D10 bound to the protease-like domain of TfR1. In the ELISA experiments, human, mouse, or three different chimeric TfR1 receptors were used to coat the ELISA plates (Figure 7). The ELISA protocol was slightly modified from the indirect ELISA described in Example 1, as follows: In short, the ELISA plates were coated with the following His-tagged antigen at 1 μg / ml: the ectodomain of human TfR1 (His 10 -hTfR1; Sequence ID 55), ectodomain of mouse TfR1 (His 10Chimeric TfR1 consisting of -mTfR1 (SEQ ID NO: 56), a human apical domain transplanted into the mouse TfR1 ectodomain (h / m apical domain chimera, mhHD_TfR1; SEQ ID NO: 57), a human helical domain transplanted into the mouse TfR1 ectodomain (h / m helical domain chimera, mhHD_TfR1; SEQ ID NO: 58), or a human protease-like domain transplanted into the mouse TfR1 ectodomain (h / m protease-like domain chimera, mhPLD_TfR1; SEQ ID NO: 59). The coated plates were then blocked. Dilution series of mouse IgG for the antibodies to be analyzed were prepared in PBS and incubated on ELISA plates. Unbound antibodies were then washed before incubating the wells with HRP-conjugated secondary anti-mouse IgG for 1 hour. The plates were then washed again before adding the HRP substrate TMB for expression and detection of antibodies bound to the wells. TMB expression was stopped by adding 0.5 M H2SO4 to the wells, and the ELISA response was measured as OD at 450 nm on an ELISA plate reader. As shown in Figure 7, 26D3, 24B4, and 37D10 bound only to hTfR1 (A) and not to mTfR1 (B). In constructs in which the human apical domain was transplanted onto the remainder of the mTfR1 ectodomain, 26D3, 24B4, and 37D10 did not bind (C). The control antibody 15G11-1 (Yu et al. (2014), Sci Transl Med 6:261ra154), known to bind to the human apical domain, showed expected binding to the h / m apical domain chimera (C). Furthermore, 26D3, 24B4, and 37D10 bind to the h / m protease-like domain chimera (D), but do not bind to any plates coated with other chimeric receptors (C and E). In addition, the control antibody 8D3, which has an epitope in the apical domain of mTfR1, binds to all plates coated with TfR1 antigens containing this domain, i.e., mTfR1 (B), the h / m protease-like domain chimera (D), and the h / m helical domain chimera (E).In summary, this experiment demonstrated that the epitopes, or epitopes 26D3, 24B4, and 37D10, are primarily located within the protease-like domain of hTfR1, which is in contrast to the control antibodies 15G11-1 and 8D3.

[0094] Further BLI experiments performed for the purpose of epitope binning (binding competition) showed that binding by both 26D3 and 24B4 targeted the same or overlapping regions of hTfR1 with epitopes located outside the apical domain (Figure 8). Epitope binning experiments by BLI were performed using an Octet Red384 instrument (ForteBio) by first (Step 1) immobilizing biotinylated hTfR1 onto streptavidin biosensors (High precision biosensors, ForteBio). Next (Step 2), a washing step was performed. Then (Step 3), the sensors loaded with hTfR1 were incubated with either a buffer (non-competitive reference) or 200 nM of the respective antibody (Ab) to form hTfR1:Ab complexes on the sensors. Finally (Step 4), sensors with free hTfR1 (reference) or their respective pre-formed hTfR1:Ab complexes were incubated with 200 nM of each antibody, and binding to hTfR1 in the complex with the competing antibody was measured. Figure 8 shows representative BLI sensorgrams obtained during the main assay steps shown. The signal in Step 4 represents the degree of competition between the two analyzed antibodies. If the antibodies compete for binding to the same or overlapping epitopes, the signal in the sensorgram from Step 4 will not increase. Conversely, if the two tested antibodies bind to separate and different epitopes, the signal from Step 4 will increase.

[0095] The results of competitive screening of antibody binding to the hTfR1 epitope by epitope binning as described above are shown in Figure 9. Antibodies 26D3 (dark gray bars) and 24B4 (light gray bars) were shown to bind to overlapping epitopes different from the hTfR1 apical domain epitope of the control antibody 15G11-1 (black bars). In Figure 9A, the binding response of 26D3 is reduced by more than 70% when hTfR1 is complexed with 24B4. As expected, the binding of 26D3 to the pre-formed hTfR1:26D3 complex is almost completely inhibited, indicating self-blocking. Similarly, in Figure 9B, the binding response of 24B4 is reduced by 70% when hTfR1 is complexed with 26D3, indicating almost complete self-inhibition. Both 24B4 and 26D3 retain a complete binding response to hTfR1 when hTfR1 is conjugated with control antibody 15G11-1, which has a binding epitope within the apical domain of hTfR1 (Figures 9A and 9B, black bars). As shown in Figure 9C, control antibody 15G11-1 exhibits a similar binding response to the apical domain of hTfR1 regardless of whether it is tested against hTfR1 without a competing antibody or when the receptor is conjugated with 24B4 or 26D3. In Figure 9, all responses are normalized by the maximum binding response of each antibody to free hTfR1.

[0096] Furthermore, we studied antibody binding to endogenous hTfR1 on brain endothelial cells. Binding to endogenous hTfR1 on the cell surface was monitored using flow cytometry and human hCMEC / D3 cells known to express significant levels of hTfR1 on their surface (Weksler et al. (2013), Fluids Barriers CNS 10:16). Positively stained cells were plotted, and the mean fluorescence intensity (MFI) is shown in Figure 10. In both Figure 10A (IgG1 antibody) and 10B (Fab fragment), cells were positively stained with hTfR1 possessing 24B4 and 26D3, showing a comparable degree (MFI) to the positive control antibody 15G11-1, which has high hTfR1 affinity, and a higher degree than the low-affinity control antibody 15G11-2 (Yu et al. (2014), see above). No background staining was detected with the negative isotype control (Figure 10A) or the unrelated Fab fragment Lys128 (Figure 10B). These data indicate that both 24B4 and 26D3 bind to hTfR1 expressed on the cell surface.

[0097] Example 3 Competition for hTfR1 binding with ferritin and transferrin The unique binding of the binders described herein to hTfR1, which bind to the protease-like domain of hTfR1 and were identified as described in Example 1, was evaluated for competition with the natural TfR1 ligands ferritin (Ft) and transferrin (Tf). To test ferritin competition with antibodies, the human mononuclear cell line THP-1 (Sigma / ECACC) was used. Binding of the scFv-Fc format (see Example 4 below) and the control antibody (M-A712) to hTfR1 on the THP-1 cell surface was confirmed, as shown in Figure 11A. To evaluate competition between ferritin and the binders of this disclosure, cells were incubated with human liver-derived ferritin (BioRad, 4420-4804) and serially diluted test binders for 1 hour at 4°C. After incubation, ferritin bound to hTfR1 on the cell surface was captured using a primary sheep antibody against human liver ferritin (BioRad, AHP2179G) and analyzed using flow cytometry. The results are shown in Figure 11B. The scFv-Fc of 26D3 does not compete with ferritin on the cell surface, while the control antibody anti-CD71, clone M-A712 (Maier et al. (2016), Mol Ther Nucleic Acids 5:e321), known to bind to the same epitope as Ft on hTfR1, clearly competes with Ft binding. Furthermore, the effect of Ft binding on the identified 26D3 hTfR1 binding agents is much smaller, indicating that 26D3 possesses an epitope on hTfR1 different from the Ft binding site (Figure 11B).

[0098] For transferrin competition, K562 lymphoblastic cells (Sigma / ECACC) were used. The cells were incubated for 1 hour at 4°C with serially diluted test conjugates, along with human holotransferrin (Thermo Fisher; T13342) conjugated with Alexa Fluor 488. Transferrin bound to hTfR1 on the cell surface was captured using flow cytometry, and the mean fluorescence intensity was plotted. Figure 11C shows no competition between the 26D3 conjugate and transferrin. When unlabeled (unconjugated) Tf was used as a positive control for competition, the signal for labeled (AF488) Tf binding decreased in a concentration-dependent manner. This experiment demonstrates that conjugates directed towards the protease-like domain of TfR1 do not directly compete with the same epitope, such as transferrin.

[0099] Overall, this example demonstrates that the binding of 26D3 to hTfR1 does not negatively affect the ability of the two endogenous ligands, ferrin and transferrin, to bind to their receptors.

[0100] Example 4 Humanization of hTfR1 binder 26D3 The Fab sequences of the mouse antibody 26D3, identified and characterized as described in Examples 1-3, were analyzed, and an in silico model of the 26D3 Fab 3D structure was created using Bioluminate Software (Schrodinger). This mouse Fab model was used as input for humanization. In this process, the CDRs in the VH and VL regions of 26D3 (see Table 2; SEQ ID NOs. 10-15) were transplanted in silico into various human variable domains, and some residues were mutated to return to the mouse framework at some positions. Three mutants with the fewest reverse mutations and otherwise desirable characteristics were generated and extracted from the software. One such humanized mutant was selected for expression and designated as h26D3. h26D3 has the VH region sequence defined by SEQ ID NO. 31 and the VL region sequence defined by SEQ ID NO. 45. Both the humanized version of h26D3 and the original mouse sequence 26D3 were expressed as His-tagged Fabs by transient transduction in Chinese hamster ovary cells (ExpiCHO; Thermo Fisher Scientific) according to the manufacturer's instructions. The recovered supernatant was purified by molecular sieve chromatography using HiTrap IMAC Sepharose FF (Cytiva) followed by HiLoad Superdex 200pg 26 / 600 (Cytiva). The following buffers were used: Ni-NTA washing buffer: 20mM Tris pH8.0, 10mM imidazole and 200mM NaCl; Ni-NTA elution buffer: 20mM Tris pH8.0, 200mM NaCl and 500mM imidazole; Molecular sieve buffer (SEC): 1xdPBS (Thermo Fisher).

[0101] The binding of purified Fabs to human and cynomolgus monkey TfR1 was evaluated using surface plasmon resonance (SPR) in a Biacore 8K instrument (Cytiva), and the results are shown in Figure 12. 1 μg / ml human TfR1 (cleaved hTfR1 of SEQ ID NO: 89) or cynomolgus monkey TfR1 (cleaved cTfR1 of SEQ ID NO: 90) was immobilized on a Cm5 sensor chip (Cytiva, #BR100399) using an amine coupling kit type 2 (Cytiva, #BR100633) according to the manufacturer's instructions. h26D3 and 26D3 Fabs were injected onto the chip using a five-step 2x dilution series starting at 25 nM. Interactions were measured using single-cycle kinetics with a flow rate of 30 μl / ml, a contact time of 120 seconds, followed by a dissociation time of 600 seconds. Surface regeneration between cycles was performed by injecting 3 M MgCl2. Binding data were fitted to a 1:1 interaction model. Fabs were diluted with HBS-EP+ (Cytiva, #BR100669). Experiments were performed at 25°C. The data confirmed that the humanized mutant of 26D3, i.e., h26D3, retains binding ability to human and cynomolgus monkey TfR1 (Figure 12). The kinetic parameters obtained from the experiment are shown in Table 3 below.

[0102] [Table 3]

[0103] Both mouse 26D3 and the humanized mutant h26D3 were converted to the scFv format and maintained target binding as scFv (Figure 13). Mouse and humanized 26D3 were reformatted to scFv (SEQ ID NOs. 60 and 61, respectively) and produced as monovalent Fc-fused scFv antibody fragments using the knob-into-hole (KiH) technique. In this format, one scFv fragment fuses to only half of the Fc knob (SEQ ID NOs. 62), while the hole half of the Fc knob (SEQ ID NOs. 63) remains unfused. The resulting antibody format is a one-arm scFv-Fc. The 26D3 scFv fused to half of the Fc knob has the complete amino acid sequence, SEQ ID NOs. 64, while the h26D3 scFv fused to half of the Fc knob has the complete amino acid sequence, SEQ ID NOs. 65. The binding profiles of mouse and humanized 26D3 in this scFv format are similar, confirming binding activity in the scFv format. The binding response is consistent with that of these antibodies in the Fab format. This was confirmed by multiple methods, including kinetics experiments using BLI (results shown in Figure 13A) and ELISA (results shown in Figure 13B). The binding kinetics of mouse and humanized 26D3-scFv-Fc were first measured by BLI with biotinylated hTfR1 immobilized on a streptavidin biosensor (ForteBio). The sensor was then washed with a buffer (Kinetics buffer, ForteBio), and the binding of 26D3-scFv-Fc (mouse) and h26D3-scFv-Fc (humanized) at a concentration of 25 nM, followed by the 500-second dissociation phase, was measured. In the ELISA experiment, hTfR1 was used to coat the plates for the standard binding ELISA experiment, using the indirect ELISA protocol described in Example 1.

[0104] Example 5 Crystallization and structural determination of h26D3-Fab in complex with hTfR1 This example demonstrates the crystallization of the complex between h26D3-Fab and hTfR1 and the determination of the binding interface. The ectodomain of human TfR1 (SEQ ID NO: 55) was expressed by transient transduction in human embryonic kidney cells (Expi297; Thermo Fisher Scientific) according to the manufacturer's instructions. The recovered supernatant was purified by molecular sieve chromatography using HiTrap IMAC Sepharose FF (Cytiva) followed by HiLoad Superdex 200pg 26 / 600 (Cytiva). The buffers used and the purification of humanized Fab were as described in Example 4.

[0105] Complex formation between humanized h26D3-Fab and hTfR1 was performed by mixing the two components in a 1:1 molar ratio in 1xdPBS and incubating at room temperature for 1 hour. The complex was then purified by molecular sieve chromatography using HiLoad Superdex 200pg 26 / 600 (Cytiva), as described in Example 4.

[0106] Crystallization was performed using a storage solution of hTfR1-h26D3 prepared to 15 mg / ml in PBS, which was then diluted to 4 mg / ml in PBS with 4 mM β-mercaptoethanol. 100+10nl drops were established using additive screening in a reservoir: 0.1 M potassium / sodium phosphate, pH 6.5, 10% PEG 3000, 0.05% dichloromethane, and 2 mM β-mercaptoethanol. The crystals were flash-frozen in a reservoir solution supplemented with 8% glycerol and 16% PEG 400.

[0107] X-ray data acquisition and refinement were performed as follows. Data was acquired at 3.87 Å at the Diamond Light Source beamline I04. The beamline was equipped with a DECTRIS Eiger2 XE 16M detector. The dataset was integrated using XDS (Kabsch (2010), Acta Crystallogr D Biol Crystallogr 66:125-132) with STARANISO anisotropic scaling (Tickle et al. (2018), Global Phasing Ltd), and c * 3.87 Å along the direction of the reciprocal lattice, and a * / b * The plane was diffracted to 4.82 Å. Three complexes were found as asymmetric units. The structure was refined using Buster refinement software, and the model was constructed using Coot. Data acquisition, refinement parameters, and statistics are shown in Table 4 below.

[0108] [Table 4]

[0109] Finally, the overall folding shown by the refined structure of the complex is shown in Figure 14. As shown in Figure 14A, three independent complexes existed as asymmetric units. The names of the chains used in the coordinate file are shown. Figure 14B shows an example of electron density formed at the interface between human TfR1 and the heavy / light chains of h26D3-Fab. Protein chains are shown in cartoon representation, and sugar portions are shown in stick representation. The binding interface interaction between h26D3 and human TfR1 was extracted from the X-ray structure and is shown below to provide information about the precise binding of h26D3 to human hTfR1.

[0110] The binding interface between human TfR1 and h26D3-Fab is shown in Figures 14 and 15, and the interaction was observed between the amino acid residues shown in Table 5.

[0111] [Table 5]

[0112] Figure 5 shows key residues from both sides involved in the epitope / paratope interface, as determined from the crystal structure. Additional residues in the vicinity may also be important for the binding between h26D3 and human TfR1. Furthermore, as shown in Example 9 below, several locations outside the observed binding interaction indicate a significant involvement of h26D3 in the binding of human TfR1.

[0113] Table 6 below lists the amino acids of human TfR1 involved in interactions with h26D3, Ft, and Tf, respectively. In particular, the amino acids involved in the binding of h26D3 do not form any part of the binding interface of the endogenous ligand. This indicates that the binders of this disclosure, such as those exemplified by h26D3, bind to human TfR1 outside the binding sites used by Ft and Tf.

[0114] [Table 6]

[0115] Different epitopes on the hTfR1 structure (pdb:1SUV) are further shown in Figure 16. As shown in Figure 16, the Ft binding site is located in the apical domain of hTfR1, the Tf binding site is mainly located in the helical domain of hTfR1, and the h26D3 epitope is located in the protease-like domain of hTfR1. The structure demonstrates that different ligands and binders utilize different, specific surface areas on the hTfR1 structure. hTfR1 is a homodimer with two identical chains, and the epitopes are shown on only one of these chains.

[0116] Example 6 Preparation and characterization of hTfR1 knock-in mice Human TfR1 knock-in (hTfR1-KI; TFR1C-KI) mice were generated by homologous recombination (experimental procedures were performed at Cyagen US). A cDNA vector containing the TFR1C (NCBI reference sequence: NM_001128148.3) ectodomain and mouse Tfrc transmembrane and intracellular domains was introduced into C57BL / 6N ES cell Tfrc cells by pronuclear microinjection. The coding region of Tfrc exon 2 and a partial intron 2 were replaced with the TFR1C chimeric cassette (Figure 17A). Correct insertion of hTfR1 cDNA was verified by Southern blotting and PCR. Transgene expression in hTfR1-KI mice was confirmed by qRT-PCR (Figure 17B) and Western blotting (Figure 17C) in brain tissue, showing endogenous expression levels. hTfR1-KI mice were maintained in a C57BL / 6N background, and only heterozygous hTfR1-KI mice were used in the experiments.

[0117] Example 7 Brain uptake of hTfR1-binding constructs in vivo To evaluate hTfR1-mediated brain uptake in vivo, monovalent Fc-scFv constructs (see Example 4) were produced for four different binding proteins. The known hTfR1 conjugate 15G11-1 was used as a control (Yu et al. (2014), see above). This hTfR1 conjugate has been shown to be effective in vivo and to serve as a positive reference control for brain uptake. Furthermore, a construct containing a non-hTfR1 scFv conjugate based on the anti-amyloid-β antibody mAb158 was designed and incorporated as a negative control in the form of an Fc fusion construct (Fc-scFv158, also abbreviated as "158" here and in the figures). Different Fc-scFv constructs were intravenously (iv) injected at an equimolar dose of 30 nmol / kg (approximately 2.3 mg / kg) into hTfR1 knock-in (hTfR1-KI) mice (n=4 per construct) prepared as described in Example 6. Plasma and brain exposure were evaluated 24 hours after administration.

[0118] Animals were anesthetized with isoflurane, and terminal blood samples were collected from the orbital plexus into BD Microtainer K2EDTA tubes. The samples were inverted and centrifuged at 2400xg for 10 minutes at 4°C. Plasma was extracted, transferred to Eppendorf tubes, and frozen at -80°C. Immediately following blood sampling, the animal's abdomen was dissected, and a cannula (21G) was inserted into the left ventricle of the heart. A small incision was made in the right ventricle, and transcardiac perfusion was performed with a minimum of 50 ml of cold PBS. Following perfusion, the brain was extracted, and the olfactory bulb was removed. The brain was separated into left and right hemispheres, and after removing the cerebellum from the left hemisphere, the left hemisphere was weighed, flash-frozen with dry ice, and stored at -80°C until further preparation and analysis of the injected construct concentration using a Meso Scale Discovery (MSD) based assay. The right hemisphere was placed in 4% formaldehyde and stored at 4°C for 24 hours, then rinsed with cold PBS and transferred to a cold 30% sucrose solution prepared in PBS, and stored at 4°C for further immunohistochemical (IHC) treatment (see Example 8).

[0119] To measure brain concentration, the frozen left hemisphere was thawed on ice and homogenized in TBS by automated bead homogenization. Before centrifugation at 16000xg, Triton® was added to the homogenate at a final Triton® concentration of 0.5%, and the supernatant was then collected.

[0120] Brain and plasma concentrations of anti-hTfR1 Fc-scFv were determined using a custom-constructed MSD plate for detecting human Fc. Standard 96-well MSD plates (MSD, #L15XA-3) were coated with 0.5 μg / ml goat anti-human IgG, Fcγ fragment-specific antibody (Jackson Immuno Research Europe Ltd, #109-005-098) diluted in 1x PBS (Medicago AB, #09-9400-100). After incubation overnight at 4°C, the plates were washed four times with 1x PBS-TWEEN® (Fisher Scientific, #09-9410-100) and blocked with 150 μl of PBS-TWEEN® (MSD, #R93BA-4) containing 1% blocker A per well. The samples and corresponding standards, ranging from 400 pM to 0.1 pM in a 1:4 dilution step, were incubated for 2 hours at 900 rpm at room temperature. This included a 1-hour incubation step with 0.5 μg / ml diluted mouse anti-human IgG (Mabtech, 3850-1-1000, MT145), followed by a further 1-hour incubation of the plates at room temperature and 900 rpm, during which they were incubated for 1 hour with 0.5 μg / ml diluted SULFO-TAG conjugate anti-mouse antibody (MSD, R32AC-1). Before reading the plates with an MSD SECTOR Imager, 150 μl of MSD read buffer (MSD, #R92TC) was added per well. Four washes with 1x PBS-TWEEN® were performed between each incubation step. All antibodies except the coated antibody and the samples were diluted with PBS-TWEEN® containing 1% blocker A and added in 50 μl volumes per well. The concentrations of analytes in the samples were evaluated using the MSD Discovery Workbench software with the 4PL curve fitting algorithm and a curve weighting of 1 / Y2 to the standard curve. Statistical analysis was performed using GraphPad Prism (v.9.0.0) with one-way ANOVA using the Turkey post hoc test.

[0121] The results are shown in Figure 18. As shown in Figure 18, substantially higher brain concentrations were observed in the two test constructs and the positive control 15G11-1 compared to the negative control (158) 24 hours after administration. As shown in Figure 18B, plasma concentrations in the two test constructs and the positive control 15G11-1 at 24 hours were lower compared to 158, indicating that hTfR1 involvement leads to faster plasma clearance. The brain-to-plasma concentration ratio is shown in Figure 18C. The two test constructs and the positive control 15G11-1 showed significantly enhanced brain exposure compared to plasma compared to the negative control. In summary, the data support hTfR1-mediated BBB transport in this experiment for the novel hTfR1 conjugates tested.

[0122] Example 8 Immunohistochemical data in brain exposure The in vivo binding of hTfR1 by the Fc-scFv construct was further investigated using qualitative immunohistochemistry (IHC) analysis. Briefly, 20 μm thick coronal brain sections were obtained from PBS-perfused mouse cerebral hemispheres as described in Example 7 using a cryostat (Microm NX50 CryoStar, Epredia). Sections were collected on Superfrost plus slides (Menzel-Glaser, #J1800AMNZ) and air-dried before IHC. Brain sections were washed with PBS (pH 7.4) for 15 minutes and incubated in blocking buffer (PBS containing 5% BSA and 0.25% Triton®-X) for 2 hours at room temperature. To visualize the intravenously administered construct, brain sections were incubated with secondary goat anti-human IgG (heavy and light chain specific) conjugated to Alexa Fluor 488 (Invitrogen, #A11013) for 120 minutes at room temperature, followed by washing with PBS for 3 x 15 minutes. Slides were mounted on Fluoromount-G (Invitrogen, #00-4958-02) for image analysis. Confocal images from the cerebral cortex were captured using a Leica Stellaris 5 confocal system equipped with an HC PL APO 40x / 1.25 GLYC motCORR CS2 objective (Leica, #11506423).

[0123] A clear IHC immunofluorescence signal was observed in the brain capillaries of the positive reference module 15G11-1, while minimal IHC signaling was detected in brain sections from mice injected with the negative control 158 (Figure 19). IHC signaling in brain capillaries was observed in two test constructs, h26D3 and 37D10, of which h26D3 showed the strongest immunofluorescence signal, comparable to that of the positive control 15G11-1. In summary, the MSD (Example 7) and IHC (This Example) analyses demonstrated that the hTfR1 conjugate of this disclosure in scFv format resulted in increased brain exposure in hTfR1-KI mice.

[0124] Example 9 Generation of affinity mutants and determination of affinity Multiple mutants of the parent antibody h26D3 were created by substituting one tyrosine, tryptophan, and aspartic acid residue in the CDR with an alanine residue. The resulting mutant VH regions are denoted as HC1 to HC13, and their amino acid sequences are provided in sequence lists SEQ ID NOs. 32 to 44, respectively. The mutant CDR sequences contained within these mutant VL regions are listed in SEQ ID NOs. 13 to 25, respectively. The resulting mutant VL regions are denoted as LC1 to LC6, and their amino acid sequences are provided in sequence lists SEQ ID NOs. 46 to 51, respectively. The mutant CDR sequences contained within these mutant VL regions are listed in SEQ ID NOs. 26 to 30, respectively. Table 7 below provides a summary of the specific mutations in each alanine mutant.

[0125] [Table 7]

[0126] The prepared alanine mutants were expressed as single-mutant His-tagged Fabs in Chinese hamster ovary cells (ExpiCHO; Thermo Fisher Scientific) via transient transduction, according to the manufacturer's instructions. The culture medium secreted by the Fabs was used to evaluate their binding to hTfR1 using BLI (Octet Red 384, ForteBio). The expressed Fabs were loaded from the cell supernatant onto an anti-Fab biosensor for 240 seconds. The binding of the ectodomain of hTfR1 (SEQ ID NO: 55), diluted to 3.75 μg / ml in 1x kinetics buffer (ForteBio), to the loaded sensor was measured for 300 seconds, followed by dissociation for 300 seconds. All mutants were confirmed to bind to hTfR1, but affected it to varying degrees (Figure 20).

[0127] Mutants that showed an effect on binding to hTfR1 during screening were selected for further characterization. Furthermore, dual mutants were created by combining heavy and light chains containing alanine substitutions. Table 8 below provides a summary of the specific mutations of each selected alanine mutant.

[0128] [Table 8]

[0129] The selected mutants were expressed as His-tagged Fabs in Chinese hamster ovary cells (ExpiCHO; Thermo Fisher Scientific) via transient transfusion, according to the manufacturer's instructions. The Fabs were then expressed as HisPur according to the manufacturer's instructions. TM Small-scale purification was performed using Ni-NTA Magnetic Beads (Thermo Scientific), followed by buffer exchange to DPBS pH 7.4. Selected mutants were also purified on a large scale by application to a HisTrap Excel column (Cytiva) and washed with 20 mM Tris, 200 mM NaCl, and 5 mM imidazole. Proteins were eluted with 20 mM Tris, 200 mM NaCl, and 500 mM imidazole, followed by buffer exchange to DPBS pH 7.4 using a HiPrep 26 / 10 Desalting column (Cytiva). Proteins were concentrated using an Amicon Ultra centrifugal concentrator (30MWCO; Millipore). Selected mutants were further purified by molecular sieve chromatography (SEC; HiLoad 26 / 600 Superdex 200; Cytiva) in DPBS pH 7.4. Protein analytical characterization was performed using UV protein quantification, SDS-PAGE, and HPLC-SEC.

[0130] The binding of purified Fabs to human and cynomolgus monkey TfR1 was evaluated using either SPR (Figure 21) or indirect ELISA (Figure 22). For SPR, a Biacore 8K instrument (Cytiva) was used. 1 μg / ml of hTfR1 (SEQ ID NO: 89) or cTfR1 (SEQ ID NO: 90) was immobilized onto a Cm5 sensor chip (Cytiva, #BR100399) using an amine coupling kit type 2 (Cytiva, #BR100633) according to the manufacturer's instructions. Fabs were injected onto the chip using a four-step 2x dilution series starting at 100 nM. Interactions were measured using single-cycle kinetics with a flow rate of 30 μl / ml, a contact time of 120 seconds, followed by a dissociation time of 1000 seconds. Surface regeneration between cycles was performed by injecting 3M MgCl2. Binding data were fitted to a 1:1 interaction model. Fabs were diluted with HBS-EP+ (Cytiva, #BR100669). The experiment was conducted at 25°C. The results are shown in Figure 21, and the calculated K D The values ​​are shown in Table 9 below.

[0131] [Table 9]

[0132] In the indirect ELISA, half-area 96-well plates (Corning, #3690) were coated overnight at 4°C with PBS containing 1 μg / ml recombinant hTfR1 ectodomain (SEQ ID NO: 74). The coated plates were blocked by shaking at room temperature for 1 hour using Pierce protein-free blocking solution (Thermo Fisher Scientific, #37572) and washed four times with PBS containing 0.1% TWEEN®-20. Dilution series (1:3) of various expression constructs were incubated at room temperature for 1 hour in incubation buffer (PBS containing 1% BSA and 0.1% TWEEN®-20). After the four washing steps, the bound test constructs were detected by adding anti-human IgG F(ab′)2-HRP antibody (Jackson Immuno Research, #109-036-003) diluted 1:5000 in incubation buffer (1 hour, room temperature). After four washing steps, K-Blue® Aqueous TMB substrate (Neogen, #331177) was added to the wells at room temperature for 15 minutes before stopping the reaction with a 1:1 dilution with 0.5 M H2SO4. Optical density at 450 nm was recorded (Spark, Tecan), and background signals were subtracted before analysis. The results are shown in Figure 22.

[0133] Based on Biacore and ELISA measurements, several mutants were identified as having broad affinities to human TfR1. Many mutants were shown to retain cross-reactivity to cynomolgus monkey TfR1.

[0134] Finally, the selected mutants were reformatted to scFv and used in the bispecific binding molecule format disclosed in WO2022 / 258841. The bispecific binding molecules, including the scFv module constructed from h26D3 and the selected alanine mutant, were expressed in ExpiCHO cells as described above. The filtered supernatant was added to a MabSelect SuRe column (Cytiva) and subsequently washed with DPBS pH 7.4. The expressed binding molecules were eluted by adding 0.7% HAc pH 2.5, and the sample was immediately neutralized to pH 7.5. The purified samples were further purified by molecular sieve chromatography (SEC; HiLoad 26 / 600 Superdex 200; Cytiva) in DPBS pH 7.4. The purified constructs were concentrated using centrifugal concentrators Amicon Ultra (39MWCO, Milipore). Each purified expression construct was characterized using SDS-PAGE, molecular sieve chromatography (Superdex 200 Increase 3.2 / 300; Cytiva), and UV protein quantification. Binding to hTfR1 was evaluated using the above SPR with adjusted concentration intervals depending on the mutant. As shown in Figure 23 and Table 10 below, the range of affinity to the hTfR1 target was shown for different test mutants.

[0135] [Table 10]

[0136] Example 10 Design of Gen 2A bispecificity binding proteins with "VH-first" or "VL-first" scFv modules Fourteen different bispecificity-binding protein constructs were designed using the "Gen 2A" format, first described in WO2022 / 258841, with the above hTfR1 conjugate h26D3 HC6 in scFv format in two different configurations: "VH-first," represented by SEQ ID NO: 67, and "VL-first," represented by SEQ ID NO: 68 (Figure 4, left panel). In the language of this disclosure, "scBM" in WO2022 / 258841 is hereby referred to as the first sub-part M1. Seven constructs were designed using the "VH-first" configuration (#1-7, each with single-chain components shown in SEQ ID NOs: 69-75), while the other seven constructs were designed using the "VL-first" configuration (#8-14, each with single-chain components shown in SEQ ID NOs: 76-82). The antibody heavy chain used in all of these constructs is identical and is shown in SEQ ID NO: 83.

[0137] Figure 24 provides a schematic overview of the different test constructs. Constructs #1-7 (VH-first) and #8-14 (VL-first) were produced as a series of different linker length combinations, as shown in Table 11, to investigate the effects of linker length and VH vs. VL-first configuration in the h26D3 HC6 scFv binding agent combinations on binding to hTfR1 and antibody positioning when hTfR1 is expressed on the cell surface.

[0138] [Table 11]

[0139] Example 11 Production and purification of the designed Gen 2A construct The 14 constructs designed as described in Example 10 were functionally expressed by transient transduction in Chinese hamster ovary cells (ExpiCHO; Thermo Fisher Scientific) according to the manufacturer's instructions. The filtered cell culture supernatant was added to a MabSelect SuRe column (Cytiva) and subsequently washed with DPBS pH 7.4. The expressed binding proteins were eluted by adding 0.7% HAc pH 2.5, and the sample was then neutralized to pH 7.5. The purified samples were further purified by molecular sieve chromatography (SEC; HiLoad 26 / 600 Superdex 200; Cytiva) in DPBS pH 7.4. Each purified expression construct was characterized using SDS-PAGE, molecular sieve chromatography (Superdex 200 Increase 3.2 / 300; Cytiva), and UV protein concentration quantification.

[0140] The purification results are shown in Table 12. A representative SDS-PAGE analysis of the purified construct is shown in Figure 25. The non-reduced gel showed one band of approximately 175 kDa. The reduced gel showed two band profiles expected in Gen 2A format, consisting of a single-chain component containing an antibody heavy chain of approximately 50 kDa and two light chains bound to an hTfR1-conjugated scFv with an approximate molecular weight of 75 kDa. As shown in Table 12, the monomer content of the bispecific binding protein was high (generally >98%), and they were produced at low mg / l levels.

[0141] [Table 12]

[0142] Example 12 Gen 2A binding analysis to hTfR1 using SPR The binding of the bispecific binding protein expressed and purified as described in Example 11 to hTfR1 was evaluated using SPR (Biacore 8K, Cytiva). 2 μg / ml of hTfR1 was immobilized on a Cm5 sensor chip (Cytiva, #BR100399) using an amine coupling kit type 2 (Cytiva, #BR100633) according to the manufacturer's instructions. The bispecific binding protein was injected onto the chip using a four-step 2x dilution series starting at 200 nM. The interaction was measured using single-cycle kinetics with a flow rate of 30 μl / ml, a contact time of 120 seconds, followed by a dissociation time of 600 seconds. Surface regeneration between cycles was performed by injecting 3 M MgCl2. Binding data were fitted to a 1:1 interaction model. The bispecific binding protein was diluted with HBS-EP+ (Cytiva, #BR100669). Experiments were performed at 25°C. The data in Figure 26 show that all designed and produced Gen 2A constructs bind to hTfR1. All constructs exhibit similar on- and off-rates compared to the control Fab construct with the hTfR1 conjugate h26D3 HC6. This indicates that all expression constructs are functional and that neither linker length nor "VH-first" / "VL-first" configuration directly affects the binding of the construct to hTfR1.

[0143] Example 13 Binding to hTfR1 expressed on the cell surface Binding to hTfR1 on cells was measured on the immortalized human B lymphocyte cell line Ramos (Sigma, catalog number: 85030802). This cell line is known to express high levels of hTfR1 on its cell surface. Fcγ receptors were blocked for 30 minutes at 4°C using an Fc receptor blocker (Innovex biosciences, #NB309-4X-40), and the cells were then washed with PBS. Cells were seeded in a 96-well V-bottom plate (#249570, Thermo Scientific Nunc), serially diluted Gen 2A construct was added, and the plate was incubated overnight at 4°C. Cells were washed with PBS containing 1% BSA and then fixed at room temperature for 15 minutes using freshly prepared 4% formaldehyde (Thermo Scientific Pierce, #28906) diluted in PBS. Cells were washed with PBS containing 1% BSA and then stained. Bispecific binding proteins bound to hTfR1 on the cell surface were detected via their shared IgG heavy chain components using fluorescently labeled secondary goat F(ab′)2 anti-human IgG(γ)-Alexa fluor 488 (Invitrogen Life Technologies, #H10120). Staining was performed for 30 minutes at 4°C. After incubation with the detection reagent, cells were washed with PBS containing 1% BSA. Cells were finally resuspended in 200 μl of PBS containing 1% BSA and acquired using the BD FACSLyric flow cytometry system (BD Biosciences). Samples were analyzed using flowJo software (BD Biosciences). Measured central fluorescence intensity (MFI) was plotted against binding protein concentration and is shown in Figure 27. The results indicate that all Gen 2A constructs bind similarly to hTfR1 expressed on the cell surface, regardless of linker length and "VH-first" / "VL-first" configuration.

[0144] Example 14 Complement-dependent cell-mediated cytotoxicity (CDC) analysis Complement activity is initiated by the binding of C1q to the Fc moiety, for example, an antibody, which further induces the binding of other complement factors, ultimately leading to cell death. To evaluate whether a given bispecificity test construct results in any increase in CDC activity by enabling C1q binding to Fc, Ramos cells (Sigma, catalog number: 85030802) were used as target cells for CDC analysis. To measure cell death, Ramos cells were labeled with the cell viability dye calcein-AM (Sigma, 17783). These labeled cells were then treated with serially diluted bispecificity test constructs in the presence of pooled human complement serum (Innovative Research Inc, #39337) for 4 hours at 37°C, 5% CO2. As a negative control, cells were treated with similar concentrations of bispecificity constructs in the presence of C1q-depleted human serum (Sigma, #234401). As a positive control, the monoclonal antibody rituximab (MabThera; Roche) was also tested under both conditions (pooled complement serum and C1q-depleted serum).

[0145] Treated cells were acquired using a BD BDLyric flow cytometer (BD Biosciences). Samples were analyzed using flowJo software (BD Biosciences). The frequency of cell death was determined on gated cells quenched with calcein AM and plotted against the concentrations of the tested bispecific construct or control. Results are shown in Figure 28 for constructs #1-7 ("VH-first") and in Figure 29 for constructs #8-14 ("VL-first"). It was clearly observed that bispecific binding proteins with the "VH-first" configuration mediated CDC activity and caused cell death (Figure 28). On the other hand, bispecific binding proteins with the "VL-first" configuration did not mediate any CDC activity (Figure 29), leading to the conclusion that hTfR1 binding mediated by the "VL-first" configuration inhibits C1q binding to the antibody's Fc domain.

[0146] Example 15 Brain and plasma exposure of Gen2A constructs with hTfR1 affinity mutations To evaluate the time-course brain and plasma exposure of the bispecific binding protein constructs described herein, additional constructs were prepared based on the Gen 2A format, similar to Example 10. This example examines the antibody mAb158 in hIgG1 format with the Fc mutation K322A, both with and without three affinity variants of the hTfR1 binding module h26D3 (see Example 9). The constructs tested and their amino acid sequences are shown in Table 13.

[0147] [Table 13]

[0148] Different affinity mutants and the comparative mAb158 hIgG1 were intravenously (iv) injected at equimolar doses of 40 nmol / kg (approximately 6-7 mg / kg) into hTfR1 knock-in (hTfR1-KI) mice prepared as described in Example 6 (n=15 per test item). Plasma and brain exposure were evaluated at five consecutive end times of 4, 24, 72, 168, and 240 hours, with n=3 mice per time point and test compound. Serum for evaluation of continuous plasma concentration versus time profiles was collected at 0.25, 4, 24, 48, 72, 120, 168, and 240 hours post-administration, from a group of animals sacrificed at 240 hours (n=3). Serum from living animals was collected from the saphenous vein into Sarstedt Microvette CB300 K2E tubes.

[0149] At each endpoint, animals were deeply anesthetized with isoflurane, and terminal serum samples were collected from the orbital plexus into BD Microtainer K2EDTA tubes. Samples were inverted and centrifuged at 2400xg for 10 minutes at 4°C. Plasma was extracted and transferred to Eppendorf tubes and frozen at -80°C. Immediately following serum sampling, the animal's abdomen was dissected, and a cannula (21G) was inserted into the left ventricle of the heart. A small incision was made in the right ventricle, and transcardiac perfusion with cold PBS was performed. Following perfusion, the brain was extracted and the olfactory bulb was removed. The brain was separated into left and right hemispheres, and after removing the cerebellum from the left hemisphere, the left hemisphere was weighed, flash-frozen with dry ice, and stored at -80°C until further preparation and analysis of the concentrations of the injected test constructs using a Meso Scale Discovery (MSD) based assay. The right hemisphere was placed in 4% formaldehyde and stored at 4°C for 24 hours, then rinsed with cold PBS and transferred to a cold 30% sucrose solution prepared in PBS, and stored at 4°C for further immunohistochemical (IHC) treatment (see Example 16 below).

[0150] To measure brain concentration, the frozen left hemisphere was thawed on ice and homogenized in Tris-buffered saline (TBS) containing the cOmplete protease inhibitor and the phosSTOP phosphatase inhibitor (#11836145001 and #04906837001, Roche) using automated bead homogenization with MP Biomedical's FastPrep-24 5G system and Lysing Matrix D for 5 seconds at 6 m / s. Triton® X-100 (#X100, Merck) was added to the homogenate, resulting in a final Triton® X-100 concentration of 0.5% and a weight-to-volume ratio of 1:10. The homogenate was vortexed for 10 seconds, centrifuged at 16000xg for 1 hour at 4°C, and the supernatant was collected and used for brain antibody exposure measurement.

[0151] Brain and plasma concentrations of 2A3#2-LC1-K322A, 2A3#2-HC6-K322A, 2A3#2-LC5-K322A, and mAb158 hIgG1-K322A were determined using a custom MSD assay for detecting human Fc. 96-well MSD plates (#L15XA-3) were coated overnight at 4°C with 25 ng per well of goat anti-human IgG, Fcγ fragment-specific antibody (#109-005-098, Jackson Immuno Research Europe Ltd) diluted in 1x PBS (#09-9400-100, Medicago AB). The coating was removed, and the wells were blocked with PBS-0.5% TWEEN® 20 (PBS-T) (#09-9410-100, Medicago AB) containing 1% blocker A (#R93BA-4, MSD). Following four washes with 1x PBS-T, samples and test construct standards diluted with PBS-T containing 1% blocker A were added to the plates and incubated at room temperature (RT) for 2 hours at 900 rpm. Detection of bound antibodies was performed by a 1-hour incubation at 900 rpm at room temperature with a secondary antibody (mouse anti-human IgG #3850-1-1000, MT145, Mabtech), followed by a 1-hour incubation at 900 rpm at room temperature with a SULFO-TAG-conjugated anti-mouse detection antibody (R32AC-1, MSD). Both the secondary and detection antibodies were diluted to 25 ng per well with PBS-T containing 1% blocker A. Four washes with 1x PBS-T were performed between all incubation steps. Following the final wash, 2X Read Buffer T (#R92TC, MSD) was added before plate reading with an MSD SECTOR Imager. The concentration of the test construct in the sample was evaluated using the MSD Discovery Workbench software with the 4PL curve fitting algorithm and a curve weighting of 1 / Y2 to the corresponding test construct standard curve.

[0152] The results are shown in Figures 30 and 31. As shown from the terminal sample in Figure 30, higher maximal concentrations in the brain and higher brain exposure over time were observed with the test construct containing the hTfR1 binding module compared with mAb158 hIgG1-K322A alone. As shown in Figure 31, plasma exposure of the test construct containing the hTfR1 binding module was lower than that of mAb158 hIgG1-K322A, indicating hTfR1 binding and clearance of the test construct from plasma to hTfR1-expressing tissues. In summary, the data support the conclusion that the test construct undergoes hTfR1-mediated BBB transport and that the affinity of the hTfR1-binding h26D3 variant influences both brain and plasma exposure profiles.

[0153] Example 16 Immunohistochemistry of the Gen 2A construct with hTfR1 affinity mutants The in vivo binding of hTfR1 using the test constructs of Example 15 (2A3#2-LC1-K322A, 2A3#2-HC6-K322A, 2A3#2-LC5-K322A, and mAb158 hIgG1-K322A) was further studied using qualitative immunohistochemistry (IHC) analysis. Briefly, the right hemisphere of the animals completed in Example 15 was embedded in sucrose in an OCT compound (LAMB / OCT, Thermo Fisher Scientific) and rapidly frozen in dry ice. The embedded right hemisphere was dissected, and sagittal 20 μm slides were collected on Superfrost cryoslides (J1800AMNZ, Thermo Fisher Scientific) and air-dried before IHC. The brain sections were pre-treated at room temperature for 1 hour with MOM mouse IgG blocking reagent (MKB-2213-1, Vector Laboratories). The primary antibody was diluted in 1x PBS containing 0.1% Triton® X-100 and incubated overnight at 4°C. The secondary antibody was diluted in 1x PBS and incubated at room temperature for 1.5 hours. Blood vessels were visualized with anti-collagen IV (1:100) (2150-1470, Biorad) and Alexa488 anti-rabbit IgG H+L (1:500) (A21206, Invitrogen). Constructs administered intravenously were visualized with Alexa647 anti-human IgG H+L (1:500) (A21206, Invitrogen). All incubations were performed in a PBS humidified chamber. After incubation, slides were washed in a cuvette with 1x PBS (usually 5x5 minutes). Sections were mounted on Fluoromount-G (00-4958-02, Invitrogen, USA), and images were captured using a Leica Stellaris 5 confocal system equipped with HC PL APO 40x / 1.25 GLYC motCORR CS2 (Figure 32) and HC PL APO 63x / 1.40 OIL CS2 (Figure 33) objectives (Leica, #11506423).

[0154] Brains 24 hours post-administration showed clear IHC immunofluorescence signals in the capillaries of the LC1 and HC6 constructs (Figure 32). The LC5 variant and control (mAb158 hIgG1-K322A) did not show detectable immunofluorescence signals in the capillaries (Figure 32). Since some variation was observed among brains of the same group, a subjective score of 0 to 3 for macroscopic perfusion success was considered when interpreting the images. The score was established based on a visual inspection of the brain after extraction, with a score of 0 corresponding to a white brain with no visible sign of blood in any vessels, a score of 1 corresponding to a brain that was slightly pink and barely showed a visible sign of blood in some vessels, a score of 2 corresponding to a brain that was pink and showed a visible sign of blood in the vessels, and a score of 3 corresponding to a red brain showing a clear visible sign of remaining blood covering most of the brain or larger vessels. Perfusion scores were in good agreement with imaging, and brains that showed immunofluorescence signals in the capillaries after injection of 2A3#2-LC5-K322A and mAb158 hIgG1-K322A were among those indicating insufficient perfusion. Staining with collagen IV was performed to visualize all capillaries on the slide and to compare with the immunofluorescence signals of the administered constructs (Figure 33). The data show active uptake into brain capillaries of constructs containing hTfR1 binding mutants when administered intravenously.

[0155] Example 17 Investigation of injection-related immune responses In vivo injection of Fc-containing biomolecules may induce an immune response, such as the presentation of acute clinical symptoms as demonstrated by Couch et al. (2013), Sci Transl Med 5(183):183ra57, 1-12. To evaluate the test constructs of this disclosure in relation to such responses, the test and control constructs shown in Table 14 were designed and expressed.

[0156] [Table 14]

[0157] Constructs were created in which a high-affinity apical TfR1 conjugate from mouse (2A2#2-8D3) or human (2A2#2-15G11) was bound to IgG1 with complete effector function in the Fc domain. 2A2#2-8D3-K322A was prepared as a complementary and resilient comparison to 2A2#2-8D3. The "VL-first" Gen 2A mutants 2A3#10-WT and 2A3#14-WT, and the "VH-first" mutants 2A3#3-WT and 2A3#2-WT, all contained the h26D3 TfR1 binding module of this disclosure in either the "VL-first" or "VH-first" scFv format, and were prepared to investigate whether infusion reactions could be mitigated by epitope binding and binding module orientation. All test constructs were administered as a single intravenous (iv) injection at doses of 2.5, 11, 40, or 60 nmol / kg to hTfR1 knock-in (hTfR1-KI) mice expressing both mouse and human TfR1 (Example 6) (n=1-3 per administration and test construct). Separate animals were used for each administration and each test construct. The first cohort of mice in each test construct was given a dose of 11 nmol / kg.

[0158] Progression to higher or lower doses depended on the presence or absence of an infusion reaction observed at the previous dose level, and the severity and duration of observed symptoms were thoroughly evaluated in accordance with Swedish and EU animal welfare laws, ethical approvals, and guidelines. Observational symptoms were evaluated as mild, moderate, or severe for each individual mouse. Symptoms ranged from asymptomatic to hunched over, isolation and inactivity, severe post-infusion drowsiness, slight motor dysfunction, and increased heart rate and respiration approximately 15–25 minutes after administration. Mild to moderate symptoms resolved completely over time, while severe or prolonged symptoms resulted in immediate animal death. Observational results of the first infusion reaction (FIR), reported as none, mild, moderate, or severe, are shown in Table 15.

[0159] [Table 15]

[0160] hTfR1-KI mice express both mouse and human TfR1 and are therefore capable of cross-reacting with both mouse-specific 8D3 and human-specific 15G11 TfR1 conjugates, as well as with the h26D3 WT mutant. Mild to moderate FIR was observed up to a dose of 60 nmol / kg with 2A2#2-8D3. Consistent with in vitro data, the response was abolished by introducing the K322A mutation into the Fc domain of the cargo antibody, as performed with 2A2#2-8D3-K322A, preventing the CDC response and thus eliminating complement activation-dependent FIR. Administration of the human apical conjugate 2A2#2-15G11 already resulted in severe FIR at a dose of 11 nmol / kg. No symptoms of FIR were observed after administration of 2A3#10-WT, 2A3#14-WT, or 2A3#3-WT at doses up to 60 nmol / kg, but after infusion of 2A3#2-WT at 60 nmol / kg, symptoms were absent or possibly mild. This supports the hypothesis of a hidden space where the antibody ("M2" portion) is located beneath the hTfR1-binding scFv domain ("M1"), thereby being closer to the plasma membrane. The results indicate that both the binding epitope and binding molecule orientation of scFv on hTfR1 cause observational reduction of FIR. In contrast, FIR was observed with both apical TfR1 conjugates 8D3 and 15G11-1 in the hTfR1-KI mouse model, regardless of whether they bind to mouse or human TfR1.

[0161] All mice were subjected to pre-infusion blood sampling, and mice that showed a mild to moderate infusion response were also subjected to 2-hour blood sampling. Both living blood samples were collected from the saphenous vein into Sarstedt Microvette CB300 K2E tubes for further plasma processing and subsequent cytokine analysis. Brain and plasma exposure of the test constructs was examined at the end of 24 hours.

[0162] Independent of end-of-procedure and subsequent sample analysis, animals were deeply anesthetized with isoflurane, and terminal blood samples were collected from the orbital plexus into BD Microtainer K2EDTA tubes. The samples were inverted and centrifuged at 2400xg for 10 minutes at 4°C. Plasma was extracted, transferred to Eppendorf tubes, and frozen at -80°C. Immediately following serum sampling, the animal's abdomen was dissected, and a cannula (21G) was inserted into the left ventricle of the heart. A small incision was made in the right ventricle, and transcardiac perfusion with cold PBS was performed. Following perfusion, the brain was extracted, and the olfactory bulb was removed. The brain was separated into left and right hemispheres, and after removing the cerebellum from the left hemisphere, the left hemisphere was weighed, flash-frozen with dry ice, and stored at -80°C until further preparation and analysis of the concentrations of the injected test constructs using a Meso Scale Discovery (MSD) based assay.

[0163] To measure brain concentration, the frozen left hemisphere was thawed on ice and homogenized in Tris-buffered saline (TBS) containing the cOmplete protease inhibitor and the phosSTOP phosphatase inhibitor (#11836145001 and #04906837001, Roche) using automated bead homogenization with MP Biomedical's FastPrep-24 5G system and Lysing Matrix D for 5 seconds at 6 m / s. Triton® X-100 (#X100, Merck) was added to the homogenate, resulting in a final Triton® X-100 concentration of 0.5% and a weight-to-volume ratio of 1:10. The homogenate was vortexed for 10 seconds, centrifuged at 16000xg for 1 hour at 4°C, and the supernatant was collected and used for brain antibody exposure measurement.

[0164] Brain and plasma concentrations of test constructs 2A3#2-WT, 2A3#3-WT, 2A3#10-WT, 2A3#14-WT, and 2A2#2-8D3 were determined using a custom MSD assay for detecting human Fc. 96-well MSD plates (#L15XA-3) were coated overnight at 4°C with 25 ng per well of goat anti-human IgG, Fcγ fragment-specific antibody (#109-005-098, Jackson Immuno Research Europe Ltd) diluted in 1x PBS (#09-9400-100, Medicago AB). The coating was removed, and the wells were blocked with PBS-0.5% TWEEN® 20 (PBS-T) (#09-9410-100, Medicago AB) containing 1% blocker A (#R93BA-4, MSD). Following four washes with 1x PBS-T, samples and test construct standards diluted with PBS-T containing 1% blocker A were added to the plates and incubated at room temperature (RT) for 2 hours at 900 rpm. Detection of bound antibodies was performed by a 1-hour incubation at 900 rpm at room temperature with a secondary antibody (mouse anti-human IgG #3850-1-1000, MT145, Mabtech), followed by a 1-hour incubation at 900 rpm at room temperature with a SULFO-TAG-conjugated anti-mouse detection antibody (R32AC-1, MSD). Both the secondary and detection antibodies were diluted to 25 ng per well with PBS-T containing 1% blocker A, and four washes with 1x PBS-T were performed between all incubation steps. Following the final wash, 2X Read Buffer T (#R92TC, MSD) was added before plate reading with an MSD SECTOR Imager. The concentrations of the test constructs in the samples were evaluated using the MSD Discovery Workbench software with the 4PL curve fitting algorithm and a curve weighting of 1 / Y2 to the corresponding test construct standard curve. For construct 2A2#2-8D3, 2A3#2-WT was used as the standard material.

[0165] Brain and plasma concentrations 24 hours after administration of the presented test construct are shown in Figure 34. As shown in Figure 34, dose increases are reflected in increased plasma and brain exposure, demonstrating the successful administration of the test item and its binding to TfR1 in vivo.

[0166] Plasma concentrations of 10 different cytokines (IFNγ, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12p70, KC / GRO, and TNF) were determined using a pre-made panel V-PLEX Plus Proinflammatory Panel 1 Mouse Kit (K15048G, Meso Scale Discovery (MSD)) according to the manufacturer's instructions. Briefly, plates were incubated with standards, control samples, and 10-fold diluted plasma samples for 2 hours, followed by the addition of a mixture of all 10 SULFO-TAG detection antibodies for 2 hours. All incubations were performed at 900 rpm and room temperature. Four washes with PBS-0.05% TWEEN® 20 (PBS-T) (#09-9410-100, Medicago AB) were performed before and after each incubation step. After the final wash, 2x Read Buffer T (#R92TC, MSD) was added before plate reading with the MSD SECTOR Imager. Analyte concentrations in the samples were evaluated using the MSD Discovery Workbench software with the 4PL curve fitting algorithm and a curve weighting of 1 / Y2 to the corresponding test construct standard curve.

[0167] Results for a subset of relevant cytokines tested are shown in Figure 35. The tested constructs induced different cytokine responses, with constructs 2A3#3-WT and 2A3#2-WT ("VH first") responding more strongly than constructs 2A3#10-WT and 2A3#14-WT ("VL first"). Generally reinforcing the hidden space hypothesis, the results demonstrate that positioning the antibody moiety ("M2") beneath the h26D3 conjugate moiety ("M1") and closer to the plasma membrane induces lower levels of cytokines and chemokines. Conjugating to TfR1 on the apical domain and showing FIR based on the above observational data (Table 15), construct 2A2#2-8D3 generated high levels of cytokines / chemokines, particularly KC / GRO and IL-10.

[0168] Example 18 Production and refinement of designed Gen 2D constructs Two bispecific constructs designed as knob-into-hole antibody variants were expressed, each containing an h26D3 scFv as a portion M1 ligated to the C-terminal amino acid residue of the knob heavy chain of the M2 antibody (see the right panel of Figure 4). In the first variant, "mAb158-Gen2D-h26D3 VH-first" (SEQ ID NO: 84), the N-terminal amino acid residue of the VH region of the M1 scFv was ligated to Fc. In the second variant, "mAb158-Gen2D-h26D3 VL-first" (SEQ ID NO: 85), the N-terminal amino acid residue of the VL region of the M1 scFv was ligated to Fc. The hole heavy chain used in both constructs is shown in SEQ ID NO: 86, and the light chains present in two copies in the "M2" antibody portion are shown in SEQ ID NO: 87.

[0169] The designed "Gen 2D" construct was expressed by transient transduction in Chinese hamster ovary cells (ExpiCHO; Thermo Fisher Scientific) according to the manufacturer's instructions. The filtered cell culture supernatant was added to a MabSelect SuRe column (Cytiva) and subsequently washed with DPBS pH 7.4. The expressed binding protein was eluted by adding 0.7% HAc pH 2.5, and the sample was then neutralized to pH 7.5. The purified sample was further purified by molecular sieve chromatography (SEC; HiLoad 26 / 600 Superdex 200; Cytiva) in DPBS pH 7.4, or by anion exchange using, for example, a HiTrap Q HP column (Cytiva), 20 mM Trizma as binding buffer, and sodium chloride for elution. Each purified expression construct was characterized using SDS-PAGE, molecular sieve chromatography (Superdex 200 Increase 3.2 / 300; Cytiva), and UV protein concentration quantification. Examples of purity for different constructs are shown in Figure 36 from representative SDS-PAGE analysis. The non-reduced gel showed a single band of approximately 175 kDa, while the reduced gel showed three bands as expected from the Gen 2D format: a "knob" heavy chain fused to the h26D3 scFv showing approximately 75 kDa, an identical light chain of approximately 25 kDa, and a "hole" heavy chain without a fused scFv of approximately 50 kDa.

[0170] Example 19 Analysis of Gen 2D bispecificity-binding proteins to hTfR1 using SPR The binding of purified bispecific binding proteins to hTfR1 and cTfR1 was evaluated as shown for the Gen 2A construct in Example 12. The data in Figure 37 show that all designed and produced Gen 2D constructs bind to hTfR1. All constructs exhibit similar on- and off-rates to hTfR1 (SEQ ID NO: 89) and cTfR1 (SEQ ID NO: 90) compared to the control Fab fragment of the hTfR1 binding agent h26D3. The results indicate that all produced constructs are functional and that the "VH-first" / "VL-first" configuration does not affect the binding of the constructs to hTfR1.

[0171] Example 20 ADCC measurement To investigate the effector function of the Gen 2D construct, an antibody-dependent cell-mediated cytotoxicity (ADCC) assay was used. Jutkat effector cells (Promega; #G7018) were used to evaluate ADCC activity. These cells stably express the FcγRIIIa receptor, a V158 (high affinity) mutant, and an NFAT response element that drives the expression of firefly luciferase, as measures of ADCC activity. When the Fc moiety binds to FcγR, an activation signal is triggered in the effector cells, causing the death of target cells coated with antibodies on their surface. Ramos cells (Sigma, catalog number: 85030802) expressing high levels of hTfR1 on their cell surface were used as target cells. Effector and target cells were used in an effector:target ratio of 6:1, both with and without dilution series of the test construct. The controls used were the antibody alone (i.e., without hTfR1-bound scFv) as a negative control, and rituximab as a positive control. Target cells with the test construct were seeded in a 96-well assay plate (Corning, #3917), mixed with effector cells, and incubated for 6 hours at 37°C and 5% CO2. Luciferase activity is induced when Fc-containing proteins form a bridge between the target and effector cells (through the interaction between Fc and FcγR). After 6 hours of incubation, Bio-Glo luciferase reagent was added, and the luciferase signal was quantified using a SPARK plate reader (Tecan).

[0172] First, the antibody rituximab was used as a positive control to verify ADCC activity and folding induction (Figure 38A). Rituximab is known as a potent inducer of ADCC, and this was confirmed in the assay setting. ADCC was not induced when target cells were not present. Importantly, the M2 partial antibody used in the Gen 2D assay construct (mAb158) did not exhibit ADCC activity when tested alone without M1, and it has also been shown that mAb158 has no ADCC activity when target cells are not present (Figure 38A). Next, the Gen 2D construct "mAb158-Gen2D-h26D3 VL first," which was expressed and analyzed as described in Examples 18-19, was examined. Importantly, despite this construct strongly binding to target cells via the hTfR1 binder, no ADCC activity was detected (Figure 38B).

[0173] Next, K562 cells (Sigma / ECACC) were used in cell binding experiments. The Fcγ receptor was blocked for 30 minutes at 4°C using an Fc receptor blocker (Innovex biosciences, #NB309-4X-40), and then the cells were washed with PBS. The cells were seeded into a 96-well V-bottom plate (#249570, Thermo Scientific Nunc), and a dilution series of the Gen 2D construct to be tested, "mAb158-Gen2D-h26D3 VL Fast," was added. The plate was incubated overnight at 4°C. The cells were washed with PBS containing 1% BSA, and then fixed for 10 minutes at room temperature using freshly prepared 4% formaldehyde (Thermo Scientific™ Pierce™, #28906) diluted in PBS. Cells were washed with PBS containing 1% BSA and then stained with fluorescently labeled secondary goat F(ab′)2 anti-human IgG(γ)-Alexa fluor 488 (Invitrogen Life Technologies, #H10120). Staining was performed for 30 minutes at 4°C. After incubation with the detection reagent, cells were washed with PBS and 1% BSA. Finally, cells were resuspended in 200 μl of PBS containing 1% BSA and acquired using the BD FACSLyric flow cytometry system (BD Biosciences). Samples were analyzed using flowJo software (BD Biosciences). The measured central fluorescence intensity (MFI) was plotted against the binding protein concentration and is shown in Figure 39. The results indicate that the tested Gen 2D construct binds to hTfR1 on the cell surface.

[0174] In summary, ADCC and cell binding experiments show that in the Gen 2D construct with h26D3 scFv in a "VL-first" configuration, the Fc moiety present in the "M2" binding protein, despite binding to the cell surface via hTfR1, fails to bind to the Fcγ receptor, thus causing ADCC.

[0175] Example 21 Plasma and brain exposure in the Gen 2D construct in vivo To further validate the h26D3 HC6 binding module as the scFv M1 portion of the Gen 2D format, along with antibody-bound amyloid-beta, brain and plasma exposure of such test constructs was examined over time in hTfR1 knock-in (hTfR1-KI) mice (Example 6). In vivo target binding to brain amyloid-beta pathology was examined 72 hours after administration in cross-reciprocated 5xFAD x hTfR-KI mice. 5xFAD x hTfR-KI mice were generated by mating 5xFAD male mice with hTfR1-KI female mice in the C57BL / 6J background (Northwestern University). The 5xFAD mouse model is an Alzheimer's disease (AD) model using human APP and PSEN1 transgene-expressing mice that have a total of five AD-related mutations, including the Swedish (K670N / M671L), Florida (I716V), and London (V717I) mutations in APP, and the M146L and L286V mutations in PSEN1.

[0176] In the construct tested in this example, the amyloid-beta binding antibody mAb158 used in the previous example was replaced with another amyloid-beta binding antibody, denoted here as mAb000. The construct was prepared on the same format as "mAb158-Gen2D-h26D3 VH first" in Example 18 and was denoted as mAb000-Gen2D-h26D3-HC6. To investigate exposure to the test construct and control antibody in the brain and plasma, mAb000-Gen2D-h26D3-HC6 and mAb000 were prepared and intravenously (iv) injected into hTfR1-KI mice at equimolar amounts of 40 nmol / kg (approximately 6-7 mg / kg) (n=11 per test construct). Terminal plasma and brain exposure were evaluated at three consecutive end times of 24, 72, and 336 hours, with n=3-5 per time point and test construct. Blood samples for evaluating continuous plasma concentration-time profiles were collected from animals sacrificed at 336 hours (n=5), and at subsequent time points of 0.25, 4, 24, 48, 72, 168, 240, and 336 hours after administration of the test construct. Live blood samples were collected from the saphenous vein into Sarstedt Microvette CB300 K2E tubes.

[0177] To investigate target binding of cerebral amyloid-beta, mAb000 and mAb000-Gen2D-h26D3-HC6 were intravenously (iv) injected into 5xFAD x hTfR1-KI mice at equimolar doses of 40 nmol / kg (approximately equivalent to 6-7 mg / kg) (n=2-3 per test item). The animals were then sacrificed 72 hours after administration.

[0178] At each end point, independent of subsequent sample analysis, animals were deeply anesthetized with isoflurane, and terminal blood samples were collected from the orbital plexus into BD Microtainer K2EDTA tubes. The samples were inverted and centrifuged at 2400xg for 10 minutes at 4°C. Plasma was extracted, transferred to Eppendorf tubes, and frozen at -80°C. Immediately following blood sampling, the animal's abdomen was dissected, and a cannula (21G) was inserted into the left ventricle of the heart. A small incision was made in the right ventricle, and transcardiac perfusion with cold PBS was performed. Following perfusion, the brain was extracted, and the olfactory bulb was removed. The brain was separated into left and right hemispheres, and after removing the cerebellum from the left hemisphere, the left hemisphere was weighed, flash-frozen with dry ice, and stored at -80°C until further preparation and analysis of the concentrations of the injected test constructs using a Meso Scale Discovery (MSD) based assay. The right hemisphere was placed in 4% formaldehyde and stored at 4°C for 24 hours, then rinsed with cold PBS and transferred to a cold 30% sucrose solution prepared in PBS, and stored at 4°C for further immunohistochemical (IHC) treatment (see Example 22 below).

[0179] To measure brain concentration, the frozen left hemisphere was thawed on ice and homogenized in Tris-buffered saline (TBS) containing the cOmplete protease inhibitor and the phosSTOP phosphatase inhibitor (#11836145001 and #04906837001, Roche) using automated bead homogenization with MP Biomedical's FastPrep-24 5G system and Lysing Matrix D for 5 seconds at 6 m / s. Triton® X-100 (#X100, Merck) was added to the homogenate, resulting in a final Triton® X-100 concentration of 0.5% and a weight-to-volume ratio of 1:10. The homogenate was vortexed for 10 seconds, centrifuged at 16000xg for 1 hour at 4°C, and the supernatant was collected and used for brain antibody exposure measurement.

[0180] Brain and plasma concentrations of mAb000 and mAb000-Gen2D-h26D3-HC6 were determined using a custom MSD assay for detecting human Fc. 96-well MSD plates (#L15XA-3) were coated overnight at 4°C with 25 ng per well of goat anti-human IgG, Fcγ fragment-specific antibody (#109-005-098, Jackson Immuno Research Europe Ltd) diluted in 1x PBS (#09-9400-100, Medicago AB). The coating was removed, and the wells were blocked with PBS-0.5% TWEEN® 20 (PBS-T) (#09-9410-100, Medicago AB) containing 1% blocker A (#R93BA-4, MSD). Following four washes with 1x PBS-T, samples and test construct standards diluted with PBS-T containing 1% blocker A were added to plates and incubated at room temperature (RT) for 2 hours at 900 rpm. Detection of conjugated antibodies was performed by a 1-hour incubation at 900 rpm at room temperature with a secondary antibody (mouse anti-human IgG #3850-1-1000, MT145, Mabtech), followed by a 1-hour incubation at 900 rpm at room temperature with a SULFO-TAG-conjugated anti-mouse detection antibody (R32AC-1, MSD). Both the secondary and detection antibodies were diluted to 25 ng per well with PBS-T containing 1% blocker A, and four washes with 1x PBS-T were performed between each incubation step. Following the final wash, 2X Read Buffer T (#R92TC, MSD) was added before plate reading with an MSD SECTOR Imager. The concentration of the test construct in the sample was evaluated using the MSD Discovery Workbench software with the 4PL curve fitting algorithm and a curve weighting of 1 / Y2 to the corresponding test construct standard curve.

[0181] The results are shown in Figure 40. As shown in Figure 40A, higher maximum concentration and higher brain exposure in the brain were observed during the test period of mAb000-Gen2D-h26D3-HC6, as shown by the area under the curve, compared to mAb000. As shown in Figure 40B, the plasma exposure of mAb000-Gen2D-h26D3-HC6 was lower compared to mAb000, indicating hTfR1 binding and clearance of the test construct from plasma to hTfR1-expressing tissues. Overall, the data support the conclusion that the test construct mAb000-Gen2D-h26D3-HC6 undergoes hTfR1-mediated BBB transport.

[0182] Example 22 Immunohistochemistry of Gen 2D constructs in vivo The in vivo binding to both hTfR1 and amyloid β by the test construct mAb000-Gen2D-h26D3-HC6 and the control antibody mAb000 (see Example 21) was further investigated using qualitative immunohistochemistry (IHC) analysis. Briefly, the right hemisphere in sucrose of the animals terminated in Example 21 was embedded in O.C.T compound (LAMB / OCT, Thermo Fisher Scientific) and snap-frozen in dry ice. The embedded right hemisphere was sectioned, and 20-μm sagittal slides were taken onto Superfrost cryoslides (J1800AMNZ, Thermo Fisher Scientific) and air-dried before IHC. Brain sections were pretreated with 4% PFA (HL96753. *1000*, HistoLab, Sweden) for 20 minutes, followed by washing with dH2O for 5 minutes and incubating with 70% FA for 5 minutes for antigen retrieval. After washing with 1XPBS for 2 x 10 minutes, the slides were blocked with M.O.M. mouse IgG blocking reagent (MKB-2213-1, Vector Laboratories) for 1 hour at room temperature. The primary antibody was diluted with 1XPBS 0.1% Triton® X-100 and incubated overnight at 4°C, and the secondary antibody was diluted with 1XPBS and incubated for 1.5 hours at room temperature.

[0183] Amyloid-beta was visualized using mouse antibodies 6E10 (1 μg / ml) (803002, Biolegend), 4G8 (1 μg / ml) (800702, Biolegend), and Alexa555 anti-mouse IgG H+L (1:500) (A21424, Invitrogen). The tested compounds were visualized using Alexa647 anti-human IgG H+L (1:500) (A21206, Invitrogen). All incubations were performed in a PBS humidified chamber. After incubation, slides were washed in a cuvette with 1xPBS (usually 5x5 minutes). The sections were mounted on a Fluoromount-G (00-4958-02, Invitrogen, USA), and images were captured using a Leica Stellaris 5 confocal system equipped with HC PL APO 40x / 1.25 GLYC motCORR CS2 objectives (Leica, #11506423).

[0184] The resulting images are shown in Figure 41. Co-localization of hIgG1 and amyloid-β antibodies was observed, and extensive amyloid-β target binding of mAb000-Gen2D-h26D3-HC6 was shown in the brain of a 7-month-old 5XFAD / hTfR-KI animal 72 hours after administration (Figure 41A). Core plaques were predominantly positive with mAb000-Gen2D-h26D3-HC6, but low amyloid-β plaque co-localization was observed with the mAb000 antibody lacking the hTfR1 binding module (Figure 41B).

[0185] Example 23 Competition of Gen 2A constructs in hTfR1 binding with antibody M-A712 In experiments complementing the research described in Example 3, a different method was used to study competition among test constructs for hTfR1 binding. The test constructs examined were both the "VH-first" and "VL-first" configurations of "mAb158-Gen2D-h26D3-HC6" from Example 18 and "mAb000-Gen2D-h26D3-HC6" from Example 21, which are here referred to as "mAb000-Gen2D-h26D3-HC6 VH-first" and "mAb000-Gen2D-h26D3-HC6 VL-first," respectively.

[0186] The method of this embodiment uses the anti-CD71 (anti-hTfR1) antibody M-A712 as a marker for a specific epitope on the apical domain of hTfR1. The M-A712 antibody has been reported to bind to hTfR1 residues 208-212 in the apical domain (Radoshitzky et al. (2008), PNAS 105(7):2664-2669; Maier et al. (2016), Molecular Therapy Nucleic Acids 5:e321). This site overlaps with the binding site shown to human ferritin on human TfR1 (Montemiglio et al. (2019), Nat Commun. 10(1):1121). The binding of the binders described in this disclosure to hTfR1, i.e., to the protease-like domain of hTfR1, as identified as described in Example 1, was evaluated in competition with the M-A712 antibody. Furthermore, competition of M-A712 with recombinant human ferritin heavy chain 1 (FTH1) was also investigated.

[0187] In the competitive experiment, K562 lymphoblastic cells (Sigma) were used. To evaluate competition between M-A712 and the disclosed conjugate and to confirm binding of the labeled M-A712 antibody, cells were first incubated with human Fc block (BD Pharmingen, 564220) for 30 minutes at 4°C to block nonspecific Fc receptor-mediated antibody binding. Cells were then incubated with serially diluted test constructs and PE-conjugated M-A712 antibody (monoclonal, BD Pharmingen, 555537) for 1 hour at 4°C. After incubation, cells were washed three times with staining buffer (1X DPBS containing 1% BSA and 0.1% sodium azide). M-A712 antibodies that appeared to be bound to hTfR1 on the cell surface were analyzed using flow cytometry, and the mean fluorescence intensity (MFI(PE)) was plotted. Figure 42 shows no direct competition between h26D3 conjugates in different constructs and M-A712. When unlabeled (unconjugated) M-A712 antibody was used as a positive control for competition, binding of labeled (PE)M-A712 signal decreased in a concentration-dependent manner. The experiment demonstrates that direct conjugates to the protease-like domain of TfR1 do not directly compete with the same epitope, such as M-A712 antibody.

[0188] Similar experiments were performed with human ferritin heavy chain 1 (FTH1; Sino Biologicals #13217-HNAE, lot number LC15NO0415) labeled with Alexa647. As shown in Figure 43, competition was observed at higher concentrations compared to the positive control (M-A712). This data clearly demonstrates that the hTfR1 conjugates described herein, which bind to the protease-like domain of hTfR1, do not compete with the indicated epitopes on the apical domain corresponding to the binding sites used by the ferritin protein.

[0189] List of embodiments by category [1] A binding protein, - A first portion M1 is a human transferrin receptor 1 (hTfR1) binding region comprising an immunoglobulin heavy chain variable region (VH) and an immunoglobulin light chain variable region (VL), wherein the VH and VL regions form a VH / VL pair comprising an antigen-binding surface, and the antigen-binding surface provides the binding selectivity of the binding protein to an epitope located in the protease-like domain of hTfR1 as defined by amino acid residues 121-183 and 384-605 of SEQ ID NO: 66, - A second portion M2 containing an antibody Fc domain, selected from the group consisting of, for example, an antibody and an Fc fusion protein. The binding protein comprises M1 and M2, where M1 and M2 are bound to each other by at least one peptide linker between M1 and M2, the linker being positioned to elicit a reduced Fc-mediated response from M2 when administered to a human and when M1 binds to an hTfR1 site present on a cell.

[0190] [2] The binding protein according to item 1, wherein the epitope of M1 located in the protease-like domain of hTfR1 contains or consists of amino acid residues 150, 151, 154, 158, 159, 161, 163, and 385 of SEQ ID NO: 66.

[0191] [3] The antigen-binding surface of M1 is composed of three complementarity-determining regions (CDRs) derived from the VH region and three CDRs derived from the VL region, and the CDRs are as follows: VHCDR1:X1X2NMX3 (Sequence ID 1) Here, X1 is selected from D and A, X2 is selected from Y and A, and X3 is selected from D and A. VHCDR2:X4INPX5X6X7TTSX8NEKFKG(Sequence ID 2) Here, X4 is selected from D and A, X5 is selected from D and A, X6 is selected from Y and A, X7 is selected from D and A, and X8 is selected from Y and A. VHCDR3:GGX9SGSSX10X11HPMX12X13 (Sequence ID 3) Here, X9 is selected from Y and A, X10 is selected from Y and A, X11 is selected from Y and A, X12 is selected from D and A, and X13 is selected from Y and A. VLCDR1: KSSQSLLX14STNQKNX15LA (SEQ ID NO: 4) Here, X14 is selected from Y and A, and X15 is selected from Y and A. VLCDR2: X16ASTRES (SEQ ID NO: 5) Here, X16 is selected from W and A. VLCDR3: QQX17FIX18PRT (SEQ ID NO: 6) Here, X17 is selected from Y and A, and X18 is selected from Y and A. The binding protein according to any one of Items 1 to 2, comprising:

[0192] [4] The binding protein according to Item 3, wherein the amino acid sequence of the VHCDR1 is selected from the group consisting of SEQ ID NO: 7 and SEQ ID NOs: 13 to 15.

[0193] [5] The binding protein according to any one of Items 3 to 4, wherein the amino acid sequence of the VHCDR2 is selected from the group consisting of SEQ ID NO: 8 and SEQ ID NOs: 16 to 20.

[0194]

[0195] [6] The binding protein according to any one of Items 3 to 5, wherein the amino acid sequence of the VHCDR3 is selected from the group consisting of SEQ ID NO: 9 and SEQ ID NOs: 21 to 25.

[0196] [7] The binding protein according to any one of Items 3 to 6, wherein the amino acid sequence of the VLCDR1 is selected from the group consisting of SEQ ID NO: 10, SEQ ID NOs: 26, and SEQ ID NO: 27.

[0197] [8] The binding protein according to any one of Items 3 to 7, wherein the amino acid sequence of the VLCDR2 is selected from the group consisting of SEQ ID NO: 11 and SEQ ID NO: 28.​​​​​​​​​​​​​[9] The amino acid sequence of VLCDR3 is selected from the group consisting of SEQ ID NOs: 12, 29, and 30. The binding protein described in any one of items 3 to 8.

[0198]

[10] The amino acid sequences of the six CDRs are as follows: VHCDR1:DYNMD (Sequence ID 7) VHCDR2:DINPDYDTTSYNEKFKG(Sequence ID 8) VHCDR3:GGYSGSSYYHPMDY (Sequence ID 9) VLCDR1:KSSQSLLYSTNQKNYLA (Sequence ID 10) VLCDR2:WASTRES (Sequence ID 11) VLCDR3:QQYFIYPRT(Sequence ID 12) The binding protein described in any one of items 3 to 9.

[0199]

[11] The amino acid sequences of the six CDRs are as follows: VHCDR1:DYNMD (Sequence ID 7) VHCDR2:DINPDADTTSYNEKFKG (Sequence ID 18) VHCDR3:GGYSGSSYYHPMDY (Sequence ID 9) VLCDR1:KSSQSLLYSTNQKNYLA (Sequence ID 10) VLCDR2:WASTRES (Sequence ID 11) VLCDR3:QQYFIYPRT(Sequence ID 12) The binding protein described in any one of items 3 to 9.

[0200]

[12] The VH area is as follows: (i) The group consisting of sequence numbers 31 to 44, for example, the group consisting of sequence numbers 31 and 37; and, (ii) A sequence is provided which has at least 80%, at least 90%, at least 92%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% identity with the sequence defined in (i), and the sequence of the CDR region is 100% identical to the sequence defined in (i). A binding protein according to any one of items 1 to 11, comprising or consisting of an amino acid sequence selected from the above.

[0201]

[13] The VL region is as follows: (i) The group consisting of sequence numbers 45 to 51; and, (ii) A sequence is provided which has at least 80%, at least 90%, at least 92%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% identity with the sequence defined in (i), and the sequence of the CDR region is 100% identical to the sequence defined in (i). A binding protein according to any one of items 1 to 12, comprising or consisting of an amino acid sequence selected from the above.

[0202]

[14] The binding protein according to any one of items 12 to 13, wherein the VH region is as defined in item 12 and the VL region is as defined in item 13.

[0203]

[15] The binding protein according to item 14, wherein the VH region includes sequence number 31 and the VL region includes a sequence selected from sequence numbers 45 to 51.

[0204]

[16] The binding protein according to item 14, wherein the VH region comprises a sequence selected from sequence numbers 31 to 44, and the VL region comprises sequence number 45.

[0205]

[17] The binding protein according to any one of items 15 to 16, wherein the VH region includes SEQ ID NO: 31 and the VL region includes SEQ ID NO: 45.

[0206]

[18] The binding protein according to any one of items 15 to 16, wherein the VH region includes SEQ ID NO: 37 and the VL region includes SEQ ID NO: 45.

[0207]

[19] The binding protein according to any one of items 1 to 18, wherein the VH / VL pair of the first portion M1 forms part of the scFv, and the VH and VL regions are linked by a peptide scFv linker.

[0208]

[20] The binding protein according to item 19, wherein the scFv linker is bound to the N-terminal amino acid residue of the VH region and the C-terminal amino acid residue of the VL region.

[0209]

[21] The scFv linker is a flexible peptide linker comprising 5 to 0 amino acid residues, e.g., 10 to 30 amino acid residues, e.g., 15 to 25 amino acid residues, e.g., about 15 amino acid residues, e.g., 15 amino acid residues, or e.g., sequence (G4S)3 (SEQ ID NO: 88), or comprising, or comprising, the binding protein according to any one of items 19 to 20.

[0210]

[22] The binding protein according to any one of items 1 to 21, wherein the Fc-mediated response elicited by the second portion M2 is selected from the group consisting of Fc-mediated cytotoxic responses, such as antibody-dependent cell-mediated cytotoxicity (ADCC), antibody-dependent cell phagocytosis (ADCP), complement-dependent cell-mediated cytotoxicity (CDC), and combinations thereof.

[0211]

[23] The binding protein described in item 22, wherein the Fc-mediated cytotoxic response is selected from ADCC, CDC, and combinations thereof.

[0212]

[24] The binding protein described in item 23, wherein the Fc-mediated cytotoxic response is ADCC.

[0213]

[25] The binding protein described in item 23, wherein the Fc-mediated cytotoxic response is CDC.

[0214]

[26] The binding protein according to any one of items 1 to 25, wherein at least one peptide linker between M1 and M2 is bound to the C-terminal amino acid residue of the VH region of M1 or the N-terminal amino acid residue of the VL region of M1 on the M1 side.

[0215]

[27] The binding protein according to item 26, wherein the peptide linker between M1 and M2 is bound to the C-terminal residue of the CH3 region of the Fc domain on the M2 side and to the N-terminal amino acid residue of the VL region of M1 on the M1 side.

[0216]

[28] M2 contains an antibody with two antibody light chains, and M1 and M2 are linked to each other via two peptide linkers. The first linker is bound to the C-terminal amino acid residue of the first light chain of M2 on the M2 side, and to the N-terminal amino acid residue of the VL region of M1 on the M1 side, The second linker binds to the N-terminal amino acid residue of the second light chain of M2 on the M2 side, and to the C-terminal amino acid residue of the VH region of M1 on the M1 side. A binding protein listed in any one of items 1 through 25.

[0217]

[29] The binding protein according to any one of items 1 to 28, wherein at least one peptide linker between M1 and M2 is a flexible linker.

[0218]

[30] The binding protein according to item 29, wherein the flexible linker comprises glycine, serine, alanine, and / or threonine residues.

[0219]

[31] The linker is (G n S m ) p and (S n G m ) pA binding protein described in item 30, having a general formula selected from the following, where independently n=1 to 7, m=0 to 7, n+m≦8 and p=1 to 10.

[0220]

[32] The binding protein according to any one of items 1 to 31, wherein at least one linker is 10 to 50 amino acid residues long, for example 10 to 30 amino acid residues long, for example 15 to 25 amino acid residues long, or 10 to 20 amino acid residues long.

[0221]

[33] A binding protein according to any one of items 29 to 32, wherein M1 and M2 are linked to each other via two peptide linkers, and both linkers are defined in any one of items 29 to 32.

[0222]

[34] The binding protein described in item 33, wherein both linkers are of the same length.

[0223]

[35] The binding protein described in item 33, wherein both linkers are of different lengths.

[0224]

[36] M2 is an antibody that selectively binds to a target present in the mammalian brain, and is a binding protein as described in any one of items 1 to 35.

[0225]

[37] The binding protein described in item 36, wherein the target is selected from the group consisting of amyloid-β peptide or its derivatives or fragments, alpha-synuclein or its derivatives or fragments, TAR DNA binding protein 43 (TDP-43) or its derivatives or fragments, trigger receptor (TREM2) expressed in bone marrow cells 2, beta-secretase 1 (BACE1), superoxide dismutase (SOD), huntingtin, transthyretin, P-secretase 1, epidermal growth factor, epidermal growth factor receptor 2, Tau, phosphorylated Tau or its fragments, apolipoprotein E4, CD20, prion protein, leucine-rich repeat kinase 2, parkin, presenilin 2, gamma-secretase, cell death receptor 6, amyloid-β precursor protein, p75 neurotrophin receptor, neuregulin, and caspase 6.

[0226]

[38] The binding protein described in item 37, wherein the target is selected from the group consisting of amyloid-β peptide or its derivatives or fragments, alpha-synuclein or its derivatives or fragments, TAR DNA binding protein 43 (TDP-43) or its derivatives or fragments, trigger receptor (TREM2) expressed in bone marrow cells 2, Tau, phosphorylated Tau or fragments thereof, and apolipoprotein E4.

[0227]

[39] The binding protein according to item 38, wherein the target is selected from the group consisting of amyloid-β peptide or its derivatives or fragments, alpha-synuclein or its derivatives or fragments, and TAR DNA binding protein 43 (TDP-43) or its derivatives or fragments.

[0228]

[40] The binding protein according to any one of items 36 to 39, wherein the antibody having selective binding ability to a target present in the mammalian brain is an anti-Aβ antibody selected from the group consisting of, for example, lecanemab, gantenerumab, aducanumab, donanemab, PBD-C06, and KHK6640.

[0229]

[41] The binding protein according to any one of items 36 to 39, wherein the antibody having selective binding ability to a target present in the mammalian brain is an anti-alpha-synuclein antibody selected from the group consisting of, for example, pracinezumab, UCB7853, Lu AF82422, TAK-341, and BAN0805.

[0230]

[42] A pharmaceutical composition comprising a binding protein described in any one of items 1 to 41 and a pharmaceutically acceptable carrier or excipient.

[0231]

[43] A binding protein as described in any one of items 1 to 41 or a composition as described in item 42, for use in therapeutic purposes such as a treatment or prophylactic treatment.

[0232]

[44] A binding protein according to any one of items 1 to 41 or a composition according to item 42, for use in in vivo diagnosis or in vivo prognosis.

[0233]

[45] The treatment, prevention, in vivo diagnosis or in vivo prognosis relating to neurodegenerative disorders, for example, Alzheimer's disease and other diseases associated with Aβ protein aggregation, traumatic brain injury (TBI), Lewy body dementia (LBD), Down syndrome (DS), amyotrophic lateral sclerosis (ALS), frontodontic dementia, tauopathy, systemic amyloidosis, atherosclerosis, Parkinson's disease (PD), Parkinson's disease dementia (PDD), Lewy body degeneration of Alzheimer's disease, multiple system atrophy, psychosis, schizophrenia, Creutzfeldt-Jakob disease, Huntington's disease and familial amyloid neurological disorders, the binding protein or composition for use as described in any one of items 43 to 44.

[0234]

[46] The treatment, prevention, in vivo diagnosis or in vivo prognosis relating to neurodegenerative disorders, and being selected from Alzheimer's disease and other diseases associated with Aβ protein aggregation, Lewy body dementia (LBD), Down syndrome (DS), amyotrophic lateral sclerosis (ALS), frontal dementia, tauopathy, Parkinson's disease (PD), Parkinson's disease dementia (PDD), and Lewy body degeneration of Alzheimer's disease, the binding protein or composition for use as described in item 45.

[0235]

[47] The treatment, prevention, in vivo diagnosis or in vivo prognosis relating to neurodegenerative disorders, selected from Alzheimer's disease and other diseases associated with Aβ protein aggregation, Lewy body dementia (LBD), amyotrophic lateral sclerosis (ALS), and Parkinson's disease (PD), is the binding protein or composition for use as described in item 46.

[0236]

[48] ​​A binding protein or composition for use as described in item 47, wherein the treatment, prevention, in vivo diagnosis or in vivo prognosis relates to Alzheimer's disease.

[0237]

[49] A binding protein or composition for use according to any one of items 43 to 44, wherein the treatment, prevention, in vivo diagnosis, or in vivo prognosis relates to a disease selected from brain cancer, multiple sclerosis, and lysosomal storage disorders.

[0238]

[50] A method for treating or prophylactic treatment of a mammal having a disease or being at risk of developing a disease, the method comprising administering to the mammal a therapeutically effective amount of a binding protein described in any one of items 1 to 41 or a composition described in item 42.

[0239]

[51] The method according to item 50, wherein the disease is a neurodegenerative disorder, for example, a neurodegenerative disorder described in any one of items 45 to 48.

[0240]

[52] The method according to item 51, wherein the disease is the disease described in item 49.

Claims

1. It is a binding protein, - A first portion M1 is a human transferrin receptor 1 (hTfR1) binding region comprising an immunoglobulin heavy chain variable region (VH) and an immunoglobulin light chain variable region (VL), wherein the VH and VL regions form a VH / VL pair including an antigen-binding surface, and the antigen-binding surface provides the binding selectivity of the binding protein to an epitope located in the protease-like domain of hTfR1 as defined by amino acid residues 121-183 and 384-605 of SEQ ID NO: 66, - A second portion M2 containing an antibody Fc domain, selected from the group consisting of, for example, an antibody and an Fc fusion protein. A binding protein comprising, where M1 and M2 are linked to each other by at least one peptide linker between M1 and M2, wherein the linker is positioned to elicit a reduced Fc-mediated response in M2 when administered to a human and when M1 binds to an hTfR1 site on a cell.

2. The binding protein according to claim 1, wherein the M1 epitope located in the protease-like domain of hTfR1 comprises or consists of amino acid residues 150, 151, 154, 158, 159, 161, 163, and 385 of SEQ ID NO:

66.

3. The antigen-binding surface of M1 is composed of three complementarity-determining regions (CDRs) derived from the VH region and three CDRs derived from the VL region, and the CDRs are as follows: VHCDR1:X1X2NMX3 (Sequence ID 1) Here, X1 is selected from D and A, X2 is selected from Y and A, and X3 is selected from D and A. VHCDR2:X4INPX5X6X7TTSX8NEKFKG (Sequence ID 2) Here, X4 is selected from D and A, X5 is selected from D and A, X6 is selected from Y and A, X7 is selected from D and A, and X8 is selected from Y and A. VHCDR3:GGX9SGSSX10X11HPMX12X13 (Sequence ID 3) Here, X9 is selected from Y and A, X10 is selected from Y and A, X11 is selected from Y and A, X12 is selected from D and A, and X13 is selected from Y and A. VLCDR1:KSSQSLLX14STNQKNX15LA (Sequence ID 4) Here, X14 is selected from Y and A, and X15 is selected from Y and A. VLCDR2:X16ASTRES (Sequence ID 5) Here, X16 is selected from W and A. VLCDR3:QQX17FIX18PRT (Sequence ID 6) Here, X17 is selected from Y and A, and X18 is selected from Y and A. A binding protein according to any one of claims 1 to 2, comprising:

4. The amino acid sequences of the six CDRs mentioned above are as follows: VHCDR1:DYNMD (SEQ ID NO: 7) VHCDR2:DINPDDYDTTSYNEKFKG (Sequence ID 8) VHCDR3:GGYSGSSYYHPMDY (Sequence ID 9) VLCDR1:KSSQSLLYSTNQKNYLA (Sequence ID 10) VLCDR2:WASTRES (Sequence ID 11) VLCDR3:QQYFIYPRT (Sequence ID 12) The binding protein according to claim 3.

5. The amino acid sequences of the six CDRs mentioned above are as follows: VHCDR1:DYNMD (SEQ ID NO: 7) VHCDR2:DINPDADTTSYNEKFKG (Sequence ID 18) VHCDR3:GGYSGSSYYHPMDY (Sequence ID 9) VLCDR1:KSSQSLLYSTNQKNYLA (Sequence ID 10) VLCDR2:WASTRES (Sequence ID 11) VLCDR3:QQYFIYPRT (Sequence ID 12) The binding protein according to claim 3.

6. The binding protein according to claim 4, wherein the VH region includes SEQ ID NO: 31 and the VL region includes SEQ ID NO:

45.

7. The binding protein according to claim 5, wherein the VH region includes SEQ ID NO: 37 and the VL region includes SEQ ID NO:

45.

8. The binding protein according to any one of claims 1 to 7, wherein the VH / VL pair of the first portion M1 forms a part of the scFv, and the VH and VL regions are linked by a peptide scFv linker.

9. The binding protein according to claim 8, wherein the scFv linker is bound to the N-terminal amino acid residue of the VH region and the C-terminal amino acid residue of the VL region.

10. The binding protein according to any one of claims 1 to 9, wherein the Fc-mediated response elicited by the second portion M2 is an Fc-mediated cytotoxic response, selected from the group consisting of, for example, antibody-dependent cell-mediated cytotoxicity (ADCC), antibody-dependent cell-mediated phagocytosis (ADCP), complement-dependent cell-mediated cytotoxicity (CDC), and combinations thereof.

11. The binding protein according to any one of claims 1 to 10, wherein at least one peptide linker between M1 and M2 is bound to the C-terminal amino acid residue of the VH region of M1 or the N-terminal amino acid residue of the VL region of M1 on the M1 side.

12. The binding protein according to claim 11, wherein the peptide linker between M1 and M2 is bound to the C-terminal residue of the CH3 region of the Fc domain on the M2 side and to the N-terminal amino acid residue of the VL region of M1 on the M1 side.

13. M2 contains an antibody with two antibody light chains, and M1 and M2 are linked to each other via two peptide linkers. The first linker is bound to the C-terminal amino acid residue of the first light chain of M2 on the M2 side, and to the N-terminal amino acid residue of the VL region of M1 on the M1 side, The second linker is bound to the N-terminal amino acid residue of the second light chain of M2 on the M2 side, and to the C-terminal amino acid residue of the VH region of M1. The binding protein according to any one of claims 1 to 11.

14. The binding protein according to any one of claims 1 to 13, wherein M2 is an antibody capable of selectively binding to a target present in the mammalian brain.

15. The binding protein according to claim 14, wherein the target is selected from the group consisting of amyloid-β peptide or its derivatives or fragments, alpha-synuclein or its derivatives or fragments, TAR DNA-binding protein 43 (TDP-43) or its derivatives or fragments, trigger receptor (TREM2) expressed in bone marrow cells 2, beta-secretase 1 (BACE1), superoxide dismutase (SOD), huntingtin, transthyretin, P-secretase 1, epidermal growth factor, epidermal growth factor receptor 2, Tau, phosphorylated Tau or fragments thereof, apolipoprotein E4, CD20, prion protein, leucine-rich repeat kinase 2, parkin, presenilin 2, gamma-secretase, cell death receptor 6, amyloid-β precursor protein, p75 neurotrophin receptor, neuregulin, and caspase 6.

16. A pharmaceutical composition comprising a binding protein according to any one of claims 1 to 15 and a pharmaceutically acceptable carrier or excipient.

17. A binding protein according to any one of claims 1 to 15 or a composition according to claim 16, for use in therapeutic applications such as therapeutic or preventive treatment.

18. The binding protein or composition for use according to claim 17, wherein the treatment or prevention relates to a neurodegenerative disorder, for example, a disease selected from Alzheimer's disease and other diseases associated with Aβ protein aggregation, traumatic brain injury (TBI), Lewy body dementia (LBD), Down syndrome (DS), amyotrophic lateral sclerosis (ALS), frontodontic dementia, tauopathy, systemic amyloidosis, atherosclerosis, Parkinson's disease (PD), Parkinson's disease dementia (PDD), Lewy body degeneration of Alzheimer's disease, multiple system atrophy, psychosis, schizophrenia, Creutzfeldt-Jakob disease, Huntington's disease, and familial amyloid neurological disease.

Images