A fusion protein of β2 microglobulin, HLA heavy chain polypeptide, and CD47-SIRPα inhibitor.

Abstract

The present invention relates to a fusion protein, the fusion protein comprising: a β2m (beta-2 microglobulin) polypeptide, wherein the β2m polypeptide is linked to an HLA heavy chain polypeptide containing an HLAα3 domain by a β2m linker peptide; optionally, an Fc polypeptide linked to an HLA heavy chain polypeptide via a peptide linker, optionally linked to a polypeptide inhibitor of the interaction between cellular CD47 and cellular signal regulatory protein alpha (SIRPα) by a short peptide linker; the any Fc polypeptide. Fusion proteins and their dimers are particularly useful as cancer treatments.

Description

[Technical Field]

[0001] This invention relates to a fusion protein and its use in patients diagnosed with cancer. The fusion protein described in this invention comprises β2m, a soluble HLA (human leukocyte antigen) heavy chain polypeptide, and an inhibitor of the interaction between CD47 and signal regulatory protein alpha (SIRPα).

[0002] This application claims priority to European Patent Application (EP) No. 23155693.7, filed on 8 February 2023, and European Patent Application (EP) No. 23213177.1, filed on 29 November 2023, which are incorporated herein by reference. [Background technology]

[0003] The tumor microenvironment contains a large number of macrophages, sometimes accounting for up to 50% of the tumor mass. Importantly, macrophages perform immune surveillance. Therefore, both innate immunity and tumor-associated macrophages are gradually gaining attention as targets for innate immunotherapy. However, there is a complex relationship between tumors and macrophages, and in some cases, macrophages can actually promote tumor growth or metastasis.

[0004] The dual role of macrophages has been shown to stem from their high degree of heterogeneity and plasticity. Macrophages can be broadly classified into two broad groups: type 1 or classically activated (M1) and type 2 or alternatively activated (M2). In vitro, M1 cells are characterized by a pro-inflammatory phenotype, exhibiting antimicrobial activity and leading to tumor suppression, while M2 macrophages can promote tissue repair, matrix remodeling, and angiogenesis-supporting tumorigenesis. Different macrophage phenotypes result in different phagocytic activities. In the context of cancer, macrophages can rapidly detect membrane molecules on tumor cells and engulf tumor cells through phagocytosis, a multi-step cellular process involving target cell recognition, cell engulfment, and lysosomal digestion, regulated by receptor-ligand interactions between target cells and phagocytic cells. Multiple antiphagocytic signals present in cancer cells have been identified, including LILRB1 and LILRB2. The leukocyte Ig-like receptor subfamily B (LILRB) is a group of type I transmembrane glycoproteins that possess an extracellular Ig-like domain that binds to a ligand, and an intracellular ITIM that can recruit tyrosine phosphatases SHP-1 and SHP-2, and inositol phosphatase SHIP.

[0005] Macrophages express both LILRB1 and LILRB2 receptors, and activation of these receptors downregulates macrophage activity in the tumor microenvironment. LILRB1 blockade in immune cells has been demonstrated to be effective against solid and liquid tumors using in vitro models. For example, LILRB1 signaling can inhibit monocyte activation and macrophage phagocytosis (Colonna M. et al. 1997 J.Exp.Med. 186(1):1809). LILRB2 blockade reprograms TAMs into a pro-inflammatory phenotype, suppresses T regulatory cell infiltration, and enhances the effectiveness of immune checkpoint inhibitors. Furthermore, LILRB2 blockade inhibits receptor-mediated activation of SHP-1 / SHP-2 and enhances the pro-inflammatory response (Alsina-Beauchamp D. et al. 2018, J.Clin.Invest. 128(12):5647).

[0006] ImmunOs Therapeutics is developing human leukocyte antigen (HLA) heavy chain-based molecules that bind to the LILRB1, LILRB2, and KIR3DL1 receptors (see, for example, International Publication (WO) 2017153438 (A1)). [Prior art documents] [Patent Documents]

[0007] [Patent Document 1] International Publication (WO) No. 2017153438 (A1) [Non-patent literature]

[0008] [Non-Patent Document 1] Colonna M.et al.1997 J.Exp.Med.186(1):1809 [Non-Patent Document 2] Alsina-Beauchamp D.et al.2018, J.Clin.Invest.128(12):5647 [Overview of the project] [Problems that the invention aims to solve]

[0009] Based on the cutting-edge technologies described above, the object of the present invention is to provide means and methods for enhancing the antitumor effect of molecules based on therapeutic HLA heavy chains. [Means for solving the problem]

[0010] This objective is achieved by the subject matter of the independent claims herein, with further advantageous embodiments described herein by the dependent claims, examples, figures and general description.

[0011] Summary of the Invention The interaction of inhibitory CD47 inhibits phagocytosis by binding to SIRPα. Targeting either CD47 or SIRPα with checkpoint inhibitors may improve immune outcomes. HLA class I heavy chain fusion proteins that block LILRB1 / 2 immune checkpoint inhibitors have been previously developed by the inventors and share the ability to modulate phagocytic checkpoints. The inventors planned an experiment to evaluate whether the pro-inflammatory effect of HLA fusion proteins on macrophage phagocytosis of tumor cells is enhanced by co-administering a drug that inhibits CD47 binding to SIRPα within the same molecule (Figure 1). The results show that the candidate HLA fusion protein iosH2 enhances the phagocytic activity of human primary macrophages against solid and liquid tumors, both as a monotherapy and, in particular, in combination with anti-CD47 and anti-SIRPα antibodies.

[0012] A first aspect of the present invention relates to a fusion protein, wherein the fusion protein is: a. A β2m (beta-2 microglobulin) polypeptide comprising sequence number 001, or a sequence having at least 85% identity with sequence number 001; Here, the β2m (β2 microglobulin) polypeptide is linked to the following HLA heavy chain polypeptide by a β2m linker peptide: b. An HLA heavy chain polypeptide comprising the HLA α3 domain; An Fc polypeptide optionally connected to the HLA heavy chain polypeptide via a peptide linker and optionally connected to the following peptide inhibitor by an optional short peptide linker: c. A polypeptide inhibitor of the interaction between cell CD47 and cell signal regulatory protein alpha (SIRPα), relates to a fusion protein comprising.

[0013] Another aspect of the present invention relates to a protein dimer comprising one or more β2m (beta2 microglobulin) polypeptides, one or more HLA heavy chain polypeptides, and one or more polypeptide inhibitors of the interaction between CD47 and SIRPα.

[0014] A further aspect of the present invention relates to a nucleic acid encoding the polypeptide of the present invention.

[0015] In certain embodiments, the fusion protein according to the present invention comprises a β2m polypeptide, an HLA heavy chain polypeptide (the entire extracellular domain of the HLA heavy chain, or a part thereof, particularly the α3 domain) bound to an immunoglobulin crystallizable fragment (Ig Fc) polypeptide. In a more specific embodiment of the fusion protein according to the present invention, the HLA heavy chain polypeptide is HLA-B * , * , * , * , * , * 57, HLA-C * 08, HLA-A * 25, HLA-B * 58, HLA-B * 27, HLA-A * 30, HLA-B * 53, or HLA-C * Selected from 12. In other more specific embodiments of the fusion protein according to the present invention, the HLA heavy chain polypeptide is HLA-A * 45, HLA-A * 24, HLA-B * 44, HLA-B * 78, HLA-B * 81, HLA-C *Selected from 06. In other specific embodiments, the HLA heavy chain polypeptide is a variant that is sequence-similar to the HLA heavy chain defined above by at least (≧)95% and has similar biological activity. The fusion protein contains a beta-2 microglobulin (B2m) polypeptide.

[0016] The fusion protein according to the present invention includes, as part of its polypeptide chain, an inhibitor of the interaction between CD47 and SIRPα.

[0017] Another aspect relates to the use of the fusion protein according to the present invention in the treatment of cancer, particularly solid tumors (malignant neoplastic diseases).

[0018] Terms and Definitions For the purposes of interpreting this Specified, the following definitions apply, wherever a term is used in the singular form, it also includes the plural form, and vice versa. In the event of any conflict between the following definitions and any document incorporated herein by reference, the definitions set forth herein shall prevail.

[0019] The terms “comprising,” “having,” “containing,” “including,” and other similar forms, as used herein, and their grammatical equivalents, are intended to be semantically equivalent and open-ended, meaning that the one or more items following any one of these words do not mean to exhaustively list the one or more items concerned, or to limit themselves to only the one or more items listed. For example, an item “comprising” components A, B, and C may consist of components A, B, and C (i.e., contain only components A, B, and C), or it may contain one or more other components in addition to components A, B, and C. Thus, “comprising” and its similar forms, and their grammatical equivalents, are intended and understood to include disclosures of embodiments of “consisting essentially of” or “consisting of.”

[0020] Where a range of values ​​is provided, unless the context clearly indicates otherwise, each intermediate value between the upper and lower limits of that range, up to one-tenth of the lower limit unit, and any other stated or intermediate values ​​within that range are understood to be included in this disclosure, subject to the specific exclusions of the stated range. Where the stated range includes one or both of the limit values, the range excluding one or both of those limit values ​​is also included in the disclosure.

[0021] In this specification, a “about” reference to a value or parameter includes (and describes) the variation directed toward the value or parameter itself. For example, a statement referring to “about X” also includes the statement “X”.

[0022] As used herein, including in the attached claims, the singular forms "a," "or," and "the" include plural references unless the context makes it clear otherwise.

[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art (e.g., cell culture, molecular genetics, nucleic acid chemistry, hybridization techniques, and biochemistry). Standard methods are used for molecular, genetic, and biochemical methods (see generally below: Sambrook et al., Molecular Cloning: A Laboratory Manual, Part IV (2012), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY and Ausubel et al., Short Protocols in Molecular Biology (2002), Part V, John Wiley & Sons, Inc.) and chemical methods.

[0024] In the context of this specification, the term "polypeptide" refers to a molecule consisting of 50 or more amino acids that form a linear chain linked by peptide bonds. The amino acid sequence of a polypeptide may also represent the amino acid sequence of an entire protein or a fragment thereof (as found physiologically). The terms "polypeptide" and "protein" are used interchangeably herein and include proteins and fragments thereof. Polypeptides are disclosed herein as amino acid residue sequences.

[0025] The sequence of amino acid residues is written from the amino terminus to the carboxyl terminus. Uppercase letters indicating the sequence position refer to the single-letter code for the L-amino acid (Stryer, Biochemistry, Vol. 3, p. 21). Lowercase letters indicating the position of the amino acid sequence represent the corresponding D- or (2R)-amino acid. The sequence is written from left to right, from the amino terminus to the carboxyl terminus. Following standard nomenclature, the sequence of amino acid residues is represented by either a three-letter or one-letter code, as follows: Alanine (Ala, A), arginine (Arg, R), asparagine (Asn, N), aspartic acid (Asp, D), cysteine ​​(Cys, C), glutamine (Gln, Q), glutamic acid (Glu, E), glycine (Gly, G), histidine (His, H), isoleucine (Ile, I), leucine (Leu, L), lysine (Lys, K), methionine (Met, M), phenylalanine (Phe, F), proline (Pro, P), serine (Ser, S), threonine (Thr, T), tryptophan (Trp, W), tyrosine (Tyr, Y), and valine (Val, V).

[0026] The terms “gene expression” or “expression,” or “gene product,” may refer to either or both the process of generating nucleic acids (RNA) or peptides or polypeptides, and their products, also known as transcription and translation, respectively, or to any intermediate process that modulates the processing of genetic information to produce polypeptide products. The term “gene expression” may also apply to the transcription and processing of RNA gene products, such as regulatory RNA or structural (e.g., ribosomal) RNA. When the expressed polynucleotides originate from genomic DNA, expression may include the splicing of mRNA in eukaryotic cells. Expression can be assessed at both the transcription and translation levels, i.e., at the mRNA and / or protein product levels.

[0027] In the context of this specification, the terms “sequence identity,” “sequence similarity,” and “percentage of sequence identity” refer to a single quantitative parameter representing the result of a sequence comparison determined by comparing two polypeptide sequences to be aligned position by position. Methods for aligning sequences for comparison are well known in the art. Sequence alignment for comparison can be performed by Smith and Waterman’s local homology algorithm, Adv.Appl.Math.2:482 (1981), Needleman and Wunsch’s global alignment algorithm, J.Mol.Biol.48:443 (1970), Pearson and Lipman’s similarity search method, Proc.Nat.Acad.Sci.85:2444 (1988), or computerized implementations of these algorithms, including but not limited to CLUSTAL, GAP, BESTFIT, BLAST, FASTA, and TFASTA. Software for performing BLAST analysis is publicly available, for example, through the National Center for Biotechnology Information (http: / / blast.ncbi.nlm.nih.gov / ).

[0028] As an example of amino acid sequence comparison, there is the BLASTP algorithm using default settings: Expect threshold: 10; Word size: 3; Max matches in a query range: 0; Matrix: BLOSUM62; Gap Costs: Existence 11, Extension 1; Compositional adjustments: Conditional compositional score matrix adjustment. Unless otherwise specified, the sequence identity values ​​provided herein refer to values ​​obtained using the BLAST series program with the above-defined default parameters of the protein (Altschul et al., J.Mol.Biol.215:403-410 (1990)).

[0029] A reference to identical sequences without specifying a percentage value includes the meaning of 100% identical sequences (i.e., the same sequence).

[0030] As used herein, the term “pharmaceutical composition” refers to an HLA fusion protein as described herein, accompanied by at least one pharmaceutically acceptable carrier. In certain embodiments, the pharmaceutical composition according to the present invention is provided in a form suitable for parenteral administration, particularly by injection.

[0031] As used herein, the term “pharmaceutically acceptable carrier” includes, as is known to those skilled in the art, any solvent, dispersion medium, coating, surfactant, antioxidant, preservative (e.g., antimicrobial, antifungal), isotonic, absorption retarder, salt, preservative, drug, drug stabilizer, binder, excipient, disintegrant, lubricant, sweetener, flavoring, coloring, etc., and combinations thereof (see, for example, Remington: The Science and Practice of Pharmacy, ISBN 0857110624).

[0032] As used herein, the terms “to treat” or “to cure” any disease or disorder (e.g., cancer) mean, in one embodiment, mitigating the disease or disorder (e.g., delaying, preventing, or reducing the onset of at least one of the disease or its clinical symptoms, e.g., slowing or reducing tumor growth). In another embodiment, “to treat” or “to cure” means mitigating or improving at least one physical parameter, including one that may not be identifiable to the patient. In yet another embodiment, “to treat” or “to cure” means modulating the disease or disorder in any way, either physically (e.g., stabilizing identifiable symptoms), physiologically (e.g., stabilizing physical parameters), or both. Methods for evaluating the treatment and / or prevention of diseases are generally known in the art unless specifically described below herein.

[0033] The terms “cancer” and “malignant neoplasm” are used synonymously herein. Certain alternative forms of any aspect and embodiment disclosed herein are directed toward the use of the combination of the present invention in the treatment of solid tumors. Other alternative forms of any aspect and embodiment disclosed herein are directed toward the use of the combination of the present invention in the treatment of myeloid leukemia or granulocytic leukemia, particularly AML, lymphocytic leukemia, lymphocytic leukemia, or lymphoblastic leukemia and lymphoma, polycythemia vera or erythropenia.

[0034] In the context of this specification, the terms “peptide linker” or “amino acid linker” mean a variable-length polypeptide used to link two polypeptides to produce a single-chain polypeptide. Exemplary embodiments of linkers useful for carrying out the invention as defined herein are oligopeptide chains consisting of 1, 2, 3, 4, 5, 10, 20, 30, 40, or 50 amino acids. A non-limiting example of an amino acid linker is polypeptide GGGGSGGGGS (SEQ ID NO: 003), which can link an HLA heavy chain polypeptide with a stabilizing peptide, for example, linking HLA-B57 and IgG4 Fc polypeptide in an HLA fusion protein.

[0035] In the context of this specification, the terms “human leukocyte antigen (HLA) heavy chain” and “HLA heavy chain” refer to proteins encoded by MHC class I histocompatibility antigen genes, particularly the classical MHC class 1a heavy chain. In humans, the HLA heavy chain may be monomeric, or may form part of a dimeric structure comprising a heavy chain having three extracellular domains (α1, α2, and α3) non-covalently bonded to a β2m light chain, or optionally, may form a trimer structure in which a small peptide is associated with a peptide bond cleft. A full-length HLA heavy chain polypeptide comprises an extracellular domain containing the α1, α2, and α3 domains, a transmembrane domain, and an intracellular domain. The “HLA heavy chain polypeptide” as described in claim 1 refers to a polypeptide containing at least α3 of the HLA heavy chain.

[0036] Table 1 shows a list of naturally occurring HLA heavy chains that are considered to be embodiments of the terminology according to this aspect of the present invention.

[0037] [Table 1]

[0038] In the context of this specification, the term “variant” refers to an HLA heavy chain polypeptide sequence having at least one amino acid substitution that differs from the naturally occurring polypeptide sequence. For example, a variant HLA heavy chain polypeptide has one or more amino acid substitutions introduced so that it differs from the original naturally occurring HLA heavy chain polypeptide sequence from which it originates. Variant HLA heavy chain polypeptides are characterized by sequence similarity of at least (≧)95%, particularly 98% or more, compared to the alignment of the extracellular domain of the naturally occurring HLA heavy chain from which it originates. In addition, since the modified amino acids do not inhibit the ability of the variant HLA heavy chain to interact with its ligand, this variant has similar biological activity to the original sequence from which it originates.

[0039] The "biological activity" of variant HLA heavy chain peptides can be evaluated in their incorporation into HLA fusion proteins according to the present invention, particularly by measuring their binding ability to the ligand LILRB2. The "similar biological activity" of a variant is defined as the binding ability of the variant HLA heavy chain polypeptide used in the HLA fusion protein to LILRB2 being at least 65%, particularly 85%, or even 95%, compared to the equivalent non-variant sequence, as measured by enzyme-linked immunosorbent assay (ELISA). This can be evaluated by calculating the EC50, which is the concentration of the fusion protein that gives a semi-maximal response, and in this case, half of the maximum binding to the biotinylated LILRB2 molecule. For example, for the variant HLA-B57 heavy chain-IgG4 fusion protein evaluated in the example (Figure 3), the equivalent non-variant wild-type HLA-B57-based structure has an EC50 of approximately 21 nM (nanomoles / L) for LILRB2 binding. Therefore, the EC50 threshold for a suitable variant with approximately 65% ​​biological function, as measured by ELISA, is approximately 32 nM, for 85% it is approximately 29 nM, and for 95% it is approximately 22 nM.

[0040] To determine the EC50 of LILRB2 binding according to the present invention, a streptavidin-coated high-binding capacity 96-well plate is coated with 50 μl of c-terminally biotinylated LILRB2 (e.g., available from BPS Bioscience #100335) with a final concentration of 5 μg / ml in PBS buffer. PBS and IgG isotypes can be used as negative controls. Serial dilutions of the HLA fusion protein are applied as a titration series (e.g., eight concentration points): 10, 2.5, 1, 0.25, 0.1, 0.025, 0.01, 0.0025 μg / ml), preferably 50 μl applied in double measurements. Next, a labeled antibody capable of detecting the fusion protein (for example, if the fusion protein contains IgG Fc, APC-conjugated goat anti-human IgG antibody (Jackson Immuno Research #109-135-098, 1:100 dilution in 50 μl of TBS)) can be applied to detect HLA fusion protein binding. Finally, 50 μl of TBS is added to each well, and fluorescence excitation and emission are measured at appropriate wavelengths (e.g., 650 nm and 660 nm, respectively). A log (agonist) model based on three parameters is one suitable method for determining the EC50 of the HLA fusion protein binding to LILRB2.

[0041] In this specification, the term “extracellular domain” applied to HLA heavy chains or variants of HLA heavy chains refers to the extracellular portion of an HLA heavy chain protein (or a variant protein in which amino acid substitutions have been introduced into a naturally occurring HLA heavy chain protein sequence). The extracellular portion of a naturally occurring HLA class 1a polypeptide includes the alpha(α)1, α2, and α3 domains, which contribute to varying degrees to receptor-ligand interactions that mediate the immunomodulatory effects of HLA fusion proteins in pharmaceutical compositions for use according to the present invention. The “extracellular domain” excludes the transmembrane domain and the intracellular domain.

[0042] The term "fusion protein" refers to a fusion protein monomer containing a single β2 microglobulin attached to an HLA polypeptide, where these portions are bound to a single stabilized immunoglobulin (Ig) Fc domain having SIRPα at its C-terminus. Alternatively, the term "fusion protein" may also refer to a dimer formed by the association of a first such fusion protein monomer and a second such fusion protein monomer, particularly linked via a stabilized Ig Fc domain.

[0043] In the context of this specification, the terms “β2-microglobulin (β2m, B2m, B2M),” “B2m polypeptide,” or “β2m polypeptide” refer to the beta (β) chain of an MHC class I heterodimer, also known as the HLA light chain. The term “β2-microglobulin” first encompasses untreated β2-microglobulin containing a secretory signal, e.g., the sequence of Uniprot P61769, or the sequence of Sequence ID No. 001, and the post-secretionary form of the protein in which the secretory signal portion of the protein has been removed, particularly by cleavage during the secretory process.

[0044] In the context of this specification, the terms “secretion signal,” “secretion signal peptide,” or “signal sequence” refer to the N-terminal leader sequence that initiates the open reading frame (ORF) of a polypeptide, typically about 6 to 30 amino acids long. Rarely, the secretion signal may be located at the C-terminus of the polypeptide. “Secretion signals” may also be called targeting signals, localization signals, transit peptides, leader sequences, or leader peptides. “Secretion signals,” which enable the efficient secretion of polypeptides from cells, are well known in the art and can be included in the ORF of recombinant proteins to facilitate the efflux of polypeptides into the supernatant in cell-based polypeptide production systems, allowing for the purification of polypeptides from the cell supernatant. Once the mRNA encoding the secretion signal is translated, it is recognized by a cytoplasmic protein that mediates the translocation of the mRNA-ribosome complex to a channel protein within the endoplasmic reticulum (ER). The newly synthesized polypeptide containing the secretion signal peptide translocates to the ER lumen via the channel protein and enters the cell’s secretory pathway. “Signal sequences” used particularly in accordance with this invention are those that are cleaved from the final polypeptide product after translation.

[0045] Binding; binding agent, ligand, antibody: In the context of this invention, the term "specific binding" refers to the properties of a ligand, which binds to its target with a certain affinity and target specificity. Such ligand affinity is indicated by the ligand's dissociation constant. A specifically reactive ligand has a dissociation constant of 10 when bound to its target. -7 Although the dissociation constant is less than mol / L, interactions between molecules with nearly the same chemical composition but different three-dimensional structures exhibit a dissociation constant at least three orders of magnitude higher.

[0046] In the context of this specification, "dissociation constant (K DThe term "equilibrium constant" is used in the sense known in the fields of chemistry and physics: it refers to the tendency of a complex consisting of [mainly two] different components to reversibly dissociate into its (two) constituent components. This complex can be, for example, an antibody-antigen complex AbAg, consisting of an antibody Ab and an antigen Ag. D This is expressed as a molar concentration [mol / l] and corresponds to the concentration of [Ab], which occupies half of the binding sites of [Ag]. In other words, the concentration of unbound [Ab] is equal to the concentration of the [AbAg] complex. The dissociation constant can be calculated using the following formula:

number

[0047] In the context of this specification, "off-rate (dissociation rate) (K off ;[1 / sec]) and "On-rate (meeting rate) (K on ;[L / (sec * The term "mol" is used in the sense known in the fields of chemistry and physics: these refer to the dissociation of an antibody with its target antigen (K off ) or meeting (K on This refers to the rate constant used to measure (K). off and K on This can be determined experimentally using methods well established in the art. The K of the antibody off and K on Surface plasmon resonance is employed as a method for measuring this. This is the principle behind biosensor systems such as the Biacore® or ProteOn® systems. These also use the following equation to determine the dissociation constant K D It can be used to find:

number

[0048] On Rate K on The natural upper limit is 109 L / sec * (It is a mole.)

[0049] In the context of this specification, the term “antibody” refers to all antibodies, including but not limited to immunoglobulins G (IgG), A (IgA), D (IgD), E (IgE), or M (IgM), any antigen-binding fragment or single chain thereof, and constructs associated with or derived therefrom. A whole antibody is a glycoprotein comprising at least two heavy chains (H) and two light chains (L) linked together by disulfide bonds. Each heavy chain has a heavy chain variable region (V H ) and heavy chain constant region (C H ) is composed of. The heavy chain constant region of IgG is C H 1. C H 2, C H It consists of three domains. Each light chain has a light chain variable region (V in this specification). L (abbreviated as) and the light chain steady region (C L It is composed of ). The light chain constant region is a single domain called C L It is composed of the following: The variable regions of the heavy and light chains contain binding domains that interact with the antigen. The constant region of the antibody contains various cells of the immune system (e.g., effector cells) and the first component of the classical complement system, and can mediate the binding of immunoglobulins to host tissues or factors. Similarly, this term encompasses so-called nanobodies or single-domain antibodies, antibody fragments consisting of a single monomeric variable antibody domain.

[0050] In the context of this specification, the term "antibody-like molecule" refers to a molecule or target with high affinity / Kd ≤ 10E. -8This refers to molecules that can specifically bind at a mol / l level. Antibody-like molecules bind to their targets in a similar manner to the specific binding of antibodies. The term "antibody-like molecule" encompasses repeat proteins, such as engineered ankyrin repeat proteins (Molecular Partners, Zurich), and modified antibody-mimetic proteins exhibiting high specificity and high affinity target protein binding (see U.S. Patent Publication Nos. 2012142611, 2016250341, 2016075767, and 2015368302, all of which are incorporated herein by reference). The term "antibody-like molecule" further encompasses, but is not limited to, polypeptides derived from armadillo repeat proteins, polypeptides derived from leucine-rich repeat proteins, and polypeptides derived from tetratricopeptide repeat proteins. The term "antibody-like molecule" further includes protein A domain, fibronectin domain FN3, consensus fibronectin domain, lipocalin (see Skerra, Biochim, Biophys, Acta 2000, 1482(1-2):337-50), polypeptides derived from zinc finger proteins (see Kwan et al. Structure 2003, 11(7):803-813), Src homology domain 2 (SH2) or Src homology domain 3 (SH3), PDZ domain, gamma crystallin, ubiquitin, cysteine ​​notch polypeptide or nottin, cystatin, Sac7d, triple helix coiled-coil (also known as the alpha form), Knitz domain or Knitz-type protease inhibitors, and specifically binding polypeptides derived from sugar-binding module 32-2.

[0051] The term "protein A domain-derived polypeptide" refers to a derivative of protein A that can specifically bind to the Fc and Fab regions of immunoglobulins.

[0052] The term "armadillo repeat protein" refers to a polypeptide containing at least one armadillo repeat, which is characterized by a pair of alpha helices that form a hairpin structure.

[0053] In the context of this specification, the term “humanized camel antibody” refers to an antibody consisting solely of a heavy chain or a variable domain (VHH domain) of a heavy chain, whose amino acid sequence has been modified to increase its similarity to antibodies naturally produced in humans, resulting in reduced immunogenicity when administered to humans. General strategies for humanizing camel antibodies are described below: Vincke et al., “General strategy to humanize a camelid single-domain antibody and identification of a universal humanized nanobody scaffold,” J Biol Chem. January 30, 2009; 284(5):3273-3284, and U.S. Patent Application Publication No. 2011165621(A1).

[0054] In this specification, the term "Fc polypeptide" refers to an immunoglobulin crystallizable fragment (Fc) region. H 2 domains and C H This refers to a portion of an antibody or immunoglobulin (Ig) consisting of three domains. "Fc polypeptide" refers, where relevant, to monomers contained within a single HLA fusion protein, or immunoglobulins covalently linked by disulfide bonds. H 2 and C HThe invention encompasses a dimer containing two HLA fusion proteins, each containing three domains. In the context of the dimeric protein according to the present invention, a disulfide bond can link two fusion protein molecules, each containing an IgFc domain. The presence of IgFc in the fusion protein improves solubility, stability, affinity, and half-life, and from a technical standpoint facilitates cost-effective production and purification (protein A or G purification) in mammalian systems. Examples of Fc polypeptides particularly used according to the present invention include SEQ ID NOs: 009 and SEQ ID NOs: 010.

[0055] In the context of this specification, the term “inhibitor of interaction between CD47 and SIRPα” encompasses any agent capable of disrupting the inhibitory signaling cascade between intracellularly presented CD47 and SIRPα, thereby limiting phagocytic cell activation, when forming part of the fusion protein according to the present invention. This includes antibody fragments and antibody-like molecules capable of specifically binding to either CD47 or SIRPα.

[0056] The term "signal regulatory protein alpha" (SIRPα, Uniprot P78324) refers to the immunoglobulin-like surface receptor for CD47 (Uniprot Q08722). Ligation of CD47 and SIRPα mediates the negative regulation of phagocytosis. [Modes for carrying out the invention]

[0057] Detailed description of the invention A first aspect of the present invention relates to a fusion protein comprising the following: a. A β2m (beta-2 microglobulin) polypeptide comprising, or consisting of, sequence number 001, or a sequence having at least 85% identity with sequence number 001; A β2m linker peptide that connects a β2m polypeptide and an HLA heavy chain polypeptide, wherein the β2m polypeptide is linked to the HLA heavy chain polypeptide via the β2m linker peptide; b. HLA heavy chain polypeptides containing the HLAα3 domain; The Fc polypeptide is then optionally connected to an HLA heavy chain polypeptide via a peptide linker, particularly a peptide linker having a length of 5 to 25 aa, and this optional Fc polypeptide is linked to a polypeptide inhibitor of the interaction between cellular CD47 and cellular signal regulatory protein alpha (SIRPα) by a short optional peptide linker; in other words, this optional Fc polypeptide can act as a bridge between the β2m-HLA domain and the SIRPα-CD47 inhibitory domain, and c. Polypeptide inhibitors of the interaction between cell CD47 and cell signaling regulatory protein alpha (SIRPα).

[0058] In a particular embodiment, a single fusion protein would contain all of the above functionalities as part of the same polypeptide sequence. In that particular embodiment, all of the above functionalities, including the Fc polypeptide, would be arranged in the order described, i.e., from the N-terminus to the C-terminus of the fusion protein, in the order of β2m, HLA, Fc, and inhibitor.

[0059] In other embodiments, all of the above functional entities containing the Fc polypeptide are part of two protein chains, one chain being physically linked to β2m-HLA and the other polypeptide chain to a SIRPα-CD47 inhibitory domain, and the two chains may be linked by a crystallizable (Fc) fragment composed of, for example, two different polypeptide chains to form a single fusion construct.

[0060] β2m polypeptide In certain embodiments, the β2m polypeptide contains or consists of SEQ ID NO: 001. In certain embodiments, the β2m polypeptide contains or consists of a sequence that is at least 85% identical to SEQ ID NO: 001. In certain embodiments, the β2m polypeptide contains or consists of a sequence that is at least 90% identical to SEQ ID NO: 001. In certain embodiments, the β2m polypeptide contains or consists of a sequence that is at least 95% identical to SEQ ID NO: 001. In certain embodiments, the β2m polypeptide contains or consists of a sequence that is at least 98% identical to SEQ ID NO: 001.

[0061] Any sequence used in the context of the present invention that is very similar to the sequences expressly disclosed herein (e.g., 85% or more identity) is understood to be required to perform its biological function; in the case of the β2m polypeptide, this would be stable and synergistically binding to the receptor it is intended to interact with, in conjunction with the HLA polypeptides described herein (see Determination of the Biological Function of HLA above).

[0062] In certain embodiments, the β2m polypeptide consists of a single copy of sequence number 001 of the β2m sequence, or a sequence having at least (≧)85% identity with sequence number 001, particularly 90% or more, 93% or more, 96% or more, or 98% or more.

[0063] In other specific embodiments, the β2m polypeptide consists of two copies of sequence number 001 of the β2m sequence, or a sequence having at least (≧)85% identity with respect to sequence number 001, particularly 90% or more, 93% or more, 96% or more, or 98% or more, and the two copies of the β2m sequence are linked by an internal (β2m)2 peptide linker.

[0064] β2m linker that links β2m polypeptide to HLA polypeptide In certain embodiments of any various construct designs contemplated herein, the β2m linker peptide is characterized by a length of 10aa to 30aa. In certain embodiments, the β2m linker peptide is a sequence consisting essentially of G residues and S residues, particularly (G n S) m Characterized by, where n is 2, 3, 4, and 5, and m is 2 to 10. In certain embodiments, the β2m linker peptide is characterized by Sequence ID No. 002.

[0065] In certain embodiments, a β2m linker peptide having a length of 1 to 30 amino acids, particularly 10 to 20 amino acids, connects the β2m polypeptide to the HLA heavy chain polypeptide.

[0066] HLA heavy chain polypeptide allele The fusion protein comprises a variant HLA heavy chain extracellular domain polypeptide. The variant HLA heavy chain polypeptide is characterized by at least (≧)95% sequence similarity (at the protein level) to the non-variant extracellular domain of the HLA heavy chain and has similar biological activity (as defined under the term "similar biological activity"). In certain embodiments, the variant HLA heavy chain is similar by 98% or more to the naturally occurring HLA heavy chain extracellular domain from which it is derived. In further specific embodiments, such a variant HLA heavy chain extracellular domain polypeptide is B * 57, B * 58, B * 27, B * 44, B * 81, C * 08, and C * A variant of an HLA polypeptide selected from 12. (For example, HLA-B57, B57, B * 57 may be used; this type of specification is compatible.

[0067] The fusion protein according to the present invention contains a beta-2-microglobulin (β2m) polypeptide linked to the HLA polypeptide portion of the fusion protein by a peptide linker.

[0068] The inventors analyzed available HLA polypeptides in terms of parameters including excellent stability in manufacturing, shelf life, in vivo performance, and binding to LILRB1 and / or LILRB2. The following HLA allele polypeptides were particularly advantageous in their analysis:

[0069] In certain embodiments, the HLA polypeptide is HLA-A * It is a polypeptide with 11 polypeptides.

[0070] In certain embodiments, the HLA polypeptide is HLA-A * It is a 24 polypeptide.

[0071] In certain embodiments, the HLA polypeptide is HLA-A * It is a polypeptide with 25 parts.

[0072] In certain embodiments, the HLA polypeptide is HLA-A * It is a 45 polypeptide.

[0073] In certain embodiments, the HLA polypeptide is HLA-B * It is a polypeptide with 27 polypeptides.

[0074] In certain embodiments, the HLA polypeptide is HLA-B * It has 44 polypeptides.

[0075] In certain embodiments, the HLA polypeptide is HLA-B * It is a polypeptide with 57 parts.

[0076] In certain embodiments, the HLA polypeptide is HLA-B * It is a 58 polypeptide.

[0077] In certain embodiments, the HLA polypeptide is HLA-B * It is a 78 polypeptide.

[0078] In certain embodiments, the HLA polypeptide is HLA-B * It is polypeptide 81.

[0079] In certain embodiments, the HLA polypeptide is HLA-C * It is polypeptide 06.

[0080] In certain embodiments, the HLA polypeptide is HLA-C * It is polypeptide 08.

[0081] In certain embodiments, the HLA polypeptide is HLA-C * It has 12 polypeptides.

[0082] The HLA polypeptide portion of the HLA fusion protein included in the combination drug for use according to the present invention does not necessarily contain specific domains of the HLA heavy chain protein that are not required for cognate-ligand interaction. In certain embodiments, the intracellular domain and transmembrane domain are absent from this HLA heavy chain polypeptide.

[0083] Structural data suggests that HLA heavy chains interact with ligands such as killer immunoglobulin-like receptors (KIRs) and leukocyte immunoglobulin-like receptors (LILRs) via regions distant from the transmembrane domain. Amino acids close to the membrane generally do not interact with receptors. Furthermore, the inventors hypothesize that the high concentration of hydrophobic amino acids within the 5C-terminal amino acid of the extracellular domain of naturally occurring HLA heavy chain sequences is likely to introduce undesirable properties, such as a tendency for protein aggregation, into recombinant proteins, potentially affecting production, purification, stability, and toxicity in downstream production processes.

[0084] In certain embodiments, the extracellular domain polypeptide of the HLA heavy chain of the fusion protein includes a core structure of the extracellular portion of the HLA heavy chain protein sequence, which contains alpha-1, alpha-2, and alpha-3 domains, so as to confer to the HLA fusion protein the ability of this portion to interact with surface molecules on the target cell.

[0085] In certain embodiments, the HLA heavy chain polypeptide portion of the fusion protein has a polypeptide sequence of the extracellular domain of a naturally occurring HLA heavy chain listed in Table 1. In certain embodiments, the HLA heavy chain extracellular domain is that of a naturally occurring HLA heavy chain that possesses immunomodulatory qualities such as specific binding to regulatory KIR3DL1, LILRA, and LILRB1 / 2 cell surface proteins that alter the function of innate and adaptive immune cells.

[0086] In certain embodiments, the HLA heavy chain portion of the fusion protein is derived from the extracellular domain of HLA-B58. In certain embodiments, the fusion protein comprises a soluble HLA heavy chain polypeptide which is an HLA-B58 polypeptide having the sequence designated as SEQ ID NO: 024.

[0087] In certain embodiments, the HLA heavy chain portion of the fusion protein is HLA-A * It is derived from 30 extracellular domains.

[0088] In certain embodiments, the HLA heavy chain portion of the fusion protein is HLA-B * It originates from, or is essentially derived from, 27 extracellular domains.

[0089] In certain embodiments, the HLA heavy chain portion of the fusion protein is HLA-B * It originates from, or is essentially derived from, 44 extracellular domains.

[0090] In certain embodiments, the HLA heavy chain portion of the fusion protein is HLA-B * It originates from, or is essentially derived from, 81 extracellular domains.

[0091] In certain embodiments, the HLA heavy chain portion of the fusion protein is HLA-C * It is derived from, or essentially derived from, the extracellular domain of 08. In certain embodiments, the HLA heavy chain polypeptide comprises the sequence designated as SEQ ID NO: 023.

[0092] In further embodiments, the HLA heavy chain portion of the fusion protein is HLA-C * It originates from, or is essentially derived from, 12 extracellular domains.

[0093] In certain embodiments, the HLA heavy chain polypeptide is a variant of the extracellular domain of a naturally occurring HLA polypeptide sequence, where one or two amino acid substitutions are carried out so that they are characterized by E at position 46 and R at position 97 (numbering assigned from the G, S, H motifs indicating the start of the extracellular domain, see, for example, SEQ ID NO: 007 for SEQ ID NO: 006). That is, an amino acid other than E is substituted for E at position 46, and / or an amino acid other than R is substituted for R at position 97. The numbering of these amino acids refers to assigning sequential integers to the G, S, H motifs that initiate the extracellular domain of secreted HLA polypeptides lacking secretory signals, starting with numbers 1, 2, 3. We have found that these substitutions correlate with higher yields and better solubility of fusion proteins based on the resulting variants compared to wild-type sequences.

[0094] In certain embodiments, the HLA heavy chain polypeptide is a variant HLA-A containing the M97R mutation, which is better expressed and more soluble, corresponding to the HLA present in SEQ ID NO: 056. * It is a polypeptide of 24. In a more specific embodiment, the fusion protein comprises or consists of two polypeptides of SEQ ID NO: 056.

[0095] HLA heavy chain allele B * 57 In certain embodiments, the HLA polypeptide is B *57 polypeptides. The examples described herein are B * This molecule was obtained using a 57-heavy chain polypeptide, and its particular usefulness has been demonstrated by the inventors.

[0096] In a specific embodiment of the fusion protein according to the present invention, the HLA-B57 polypeptide contained in the HLA fusion protein is a naturally occurring HLA-B * The 57 heavy chain protein contains alpha-1, alpha-2, and alpha-3 domains, which are essential for receptor-ligand interactions that mediate the immunomodulatory effects of the HLA fusion protein according to the present invention.

[0097] In certain embodiments, the fusion protein comprises a variant HLA-B57 heavy chain extracellular domain polypeptide characterized by at least one or two amino acid substitutions that differ from the polypeptide sequence, which is a naturally occurring MHC class 1a heavy chain extracellular domain polypeptide. The HLA-B57 heavy chain gene family currently encompasses 221 known variants, and HLA-B * 57:01~HLA-B * It has a unique nucleic acid sequence numbered 57:141 and encodes dozens of unique protein sequences. These protein sequences are known and can be found, for example, in the MGT / HLA Allele Query Form provided by the European Bioinformatics Institute Immuno Polymorphism Database (Robinson J. et al. 2013 Nucleic Acids Res. 41:D1234, https: / / www.ebi.ac.uk / ipd / imgt / hla / allele.html) using the search term "B * You can search by entering "57".

[0098] In more specific embodiments, the HLA heavy chain polypeptide is a variant of the naturally occurring extracellular domain of the HLA-B57 heavy chain polypeptide sequence, characterized by one or two amino acid substitutions, namely E at position 46 and R at position 97 (numbers assigned from the G, S, H motifs indicating the start of the extracellular domain). That is, an amino acid other than E is substituted with E at position 46, and / or an amino acid other than R is substituted with R at position 97. The numbering of these amino acids is determined by sequentially assigning integers 1, 2, and 3, starting from the G, S, H motifs that initiate the extracellular domain of the secreted HLA-B57 portion lacking secretory signaling. The inventors have found that these substitutions correlate with a higher yield of fusion proteins based on the resulting variant compared to the wild-type sequence.

[0099] In a particular embodiment, the β2m polypeptide contained in the fusion protein has a sequence designated as sequence number 001.

[0100] In more specific embodiments, the HLA heavy chain polypeptide portion of the fusion protein comprises a variant HLA-B57 sequence designated as SEQ ID NO: 007. In even more specific embodiments, the HLA heavy chain polypeptide essentially consists of the sequence designated as SEQ ID NO: 007.

[0101] In a more specific embodiment of the fusion protein according to the present invention, the variant HLA-B57 polypeptide protein is the alpha-1, alpha-2, and alpha-3 domains of the naturally occurring HLA heavy chain protein, but without the C-terminal isoleucine-valing dipeptide, preceded by a threonine-valine-proline residue in the extracellular domain, which is located within the HLA-B57 region preceding the transmembrane domain, annotated or referred to as the "connection peptide."

[0102] In a more specific embodiment, HLA-B57 comprises one or more mutations selected from the following positions in the sequence of B57 of sequence number 006: - 7th, 46th, 97th, 143rd, 194th, 197th (according to UniProt numbering) (Sequence ID 006)

[0103] According to our current knowledge, positions 7 and 143 particularly influence peptide bonding in HLA. To ensure that open conformer polypeptides are stable and active without peptide bonding, we intend to alter these positions to avoid nonspecific peptide bonding.

[0104] These positions are structurally well conserved in all classical HLA (MHC-1 class) for interacting with peptides. However, because peptides change, these positions are not completely conserved at the aa position level.

[0105] According to our homology modeling and simulations, positions 23, 25, 206, and 211 are part of a hydrophobic patch, which affects solubility and stability. Replacing these amino acids with more hydrophilic or charged amino acids would affect the overall behavior of the protein variant.

[0106] [ka]

[0107] Sequence ID 006 of the above sequence shows all residues mentioned in the above and below descriptions in bold: 7(Y), 23(I), 25(V), 45(M), 46(A), 62(G), 66(N), 74(Y), 79(R), 94(I), 97(V), 116(S), 131(S), 143(I), 147(W), 163(L), ​​194(I), 197(H), 206(L), 211(A).

[0108] Based on our homology alignment, positions 45, 46, 62, 66, 74, 79, 94, 97, 116, 131, 147, and 163 were determined to influence either binding to LILRB1 and LILRB2 via structural changes or allosteric binding. These positions are based on the protein structure and are neither directly involved in the interaction between HLA and LILRB1 or LILRB2 nor are they part of the binding epitope. These positions do not appear to be conserved, are variable, and are likely drivers of the high stability of other HLA subtypes, thus potentially affecting stability and solubility.

[0109] Positions 194 and 197 of HLA are characterized by T and F of the human HLA-G protein, respectively. In order to mimic the high affinity of human HLA-G to LILRB1 / LILRB2 in the context of fusion proteins, the I194T and / or H197F substitutions are included in SEQ ID NO: 006 or SEQ ID NO: 007 (A46E / V97R). * Polypeptide 57 was designed to enhance binding to the LILRB receptor.

[0110] HLA polypeptide having an α1-α2-α3 domain structure In certain embodiments, the β2m polypeptide consists of a single copy of the β2m sequence SEQ ID NO: 001, or a sequence having at least (≧)85% identity with SEQ ID NO: 001, particularly 90% or more, 93% or more, 96% or more, or 98% or more, and the HLA heavy chain polypeptide consists of a complete HLA heavy chain from the N-terminus to the C-terminus, comprising an α1 main, an α2 domain, and an α3 domain. The "β2m linker" peptide links the β2m polypeptide and the HLA heavy chain polypeptide.

[0111] In that particular embodiment, the HLA heavy chain polypeptide is HLA-B, which consists of α1, α2, and α3 domains. * It is a 57-chain polypeptide.

[0112] In more specific embodiments, the β2m polypeptide and the HLA heavy chain polypeptide together are characterized by the sequence of SEQ ID NO: 017, or a sequence having at least (≧)85% identity, particularly 90% or more, 93% or more, 96% or more, or 98% or more, to SEQ ID NO: 017, and a sequence having at least 75% of the biological function of SEQ ID NO: 017 (see the section "Biological Activity of Variant HLA Heavy Chain Peptides" above for the determination of biological function; this applies to all references made herein to biological activity in the context of HLA or β2m-HLA polypeptides).

[0113] [ka]

[0114] HLA polypeptide having an α3 domain structure In certain embodiments, the β2m polypeptide consists of a single copy of the β2m sequence SEQ ID NO: 001, or a sequence having at least (≧)85% identity to SEQ ID NO: 001, particularly 90% or more, 93% or more, 96% or more, or 98% or more, and the HLA heavy chain polypeptide consists of one, two, or three HLAα3 domains from the N-terminus to the C-terminus. The "β2m linker" peptide links the β2m polypeptide and the HLA heavy chain polypeptide.

[0115] In that particular embodiment, the HLA heavy chain polypeptide consists of one, two, or three HLA-B * HLA-B consisting of the 57α3 domain * It is a 57-chain polypeptide.

[0116] In a particular embodiment, the first α3 domain and the second α3 domain are linked by an "(α3)2 peptide linker." This (α3)2 peptide linker may be a glycine-serine linker consisting of 1 to 30 amino acids, represented by an n-mer of (GGGGS) (Sequence ID: 003), where n is 1, 2, 3, 4, or 5. In a particular embodiment, the two α3 domains are linked by a monomer of this sequence.

[0117] In more specific embodiments, the β2m polypeptide and the HLA heavy chain polypeptide together are characterized by the sequence of SEQ ID NO: 019, or a sequence having at least (≧)85%, particularly 90%, 93%, 96%, or 98% identity to SEQ ID NO: 019, and having at least 75% of the biological function of SEQ ID NO: 019.

[0118] [ka]

[0119] HLA polypeptides having one or more copies of β2m In certain embodiments, the β2m polypeptide consists of two copies of the β2m sequence linked by an internal (β2m)2 peptide linker. In those particular embodiments, the HLA heavy chain polypeptide consists of two HLAα3 domains.

[0120] A non-limiting specific example of an internal (β2m)2 peptide linker is a glycine-serine linker composed of 1 to 30 amino acids. In a particular embodiment, the internal (β2m)2 peptide linker is represented by formula (GGGGS)n (SEQ ID NO: 003), where n is 1, 2, 3, 4, or 5.

[0121] In a particular embodiment, the first α3 domain and the second α3 domain are linked by an (α3)2 peptide linker. This (α3)2 peptide linker may be a glycine-serine linker composed of 1 to 30 amino acids, represented by the n-mer of (GGGGS) (SEQ ID NO: 003), where n is 1, 2, 3, 4, or 5. In a particular embodiment, the two α3 domains are linked by a monomer of this sequence.

[0122] In a more specific embodiment, the HLA heavy chain polypeptide consists of two HLA-B * It consists of 57α3 domains.

[0123] In a more specific embodiment, the β2m polypeptide and the HLA heavy chain polypeptide together are characterized by the sequence of SEQ ID NO: 021, or a sequence having at least (≧)85% identity, particularly 90% or more, 93% or more, 96% or more, or 98% or more identity with SEQ ID NO: 021, and having at least 75% of the biological function of SEQ ID NO: 021.

[0124] [ka]

[0125] Inhibitors of the interaction between CD47 and SIRPα (SIRPα-CD47 inhibitory domain) Inhibitors of the CD47-SIRPα interaction reported in preclinical and clinical trials encompass a variety of agents capable of disrupting the inhibitory signaling cascade between CD47 and SIRPα that limits phagocytic cell activation (Zhang W. et al. 2020, Front.Immunol. 11(19) doi:10.3389 / fimmu.2020.00018). These include antibodies, antibody fragments, antibody-like molecules, or innate ligand receptors capable of binding to either CD47 or SIRPα and inhibiting ligand-to-ligand interaction, or recombinant proteins incorporating the extracellular ligand-binding domain of either CD47 or SIRPα. Any of these, either in whole or as an active portion, can be attached to a fusion molecule to serve as an inhibitor of the CD47-SIRPα interaction as used herein.

[0126] Inhibitors of the CD47-SIRPα interaction known in this technology include ALX148 (ALX Oncology), TTI-662 or TTI-621 (Trillium Therapeutics), Hu5F9-G4 (Stanford University 47), TI-061 (Arch Oncology), SRF231 (Surface Oncology), SHR-1603 (Hengrui), NI-1701, NI-1801 (Novimmune TG Therapeutics), IBI188 (Innovent Biologics), CC-90002 (Celgene Inibrx), AO-176 (Arch Oncology), lemzoparlimap / TJC4 (I-MAB Biopharma), SY102 (Salyuan), SL-172154 (Shattuck Labs), and PSTx-23 (Paradigm Shift). Examples include IMM2504, IMM2502, IMM03 (ImmuneOnco Biopharma), and IMC-002 (ImmuneOncia Therapeutics).

[0127] In a particular embodiment, the inhibitor of the interaction between CD47 and SIRPα is 10 -7 It binds to the extracellular domain of CD47 or SIRPα with a dissociation constant (Kd) of M or less (high affinity). Furthermore, in certain embodiments, the affinity of this interaction is at least 10 -8 M, or even 10 more -9 It is M.

[0128] In some specific embodiments, the inhibitor of the interaction between CD47 and SIRPα comprises or consists of an antibody fragment or an antibody-like molecule. In other specific embodiments, the inhibitor of the interaction between CD47 and SIRPα is essentially an antibody fragment or an antibody-like molecule.

[0129] In certain embodiments, the inhibitor of the interaction between CD47 and SIRPα comprises an antibody fragment. In certain embodiments, the inhibitor of the interaction between CD47 and SIRPα is an antibody fragment characterized by its specific binding to CD47. In more specific embodiments, the inhibitor of the interaction between CD47 and SIRPα is an antibody fragment characterized by its specific binding to CD47, as disclosed in WO2020 / 068752(A1), in particular CC-90002 (Celgene). In certain embodiments, the inhibitor of the interaction between CD47 and SIRPα is an antibody characterized by its specific binding to SIRPα.

[0130] The antibody fragment may be the Fab domain or variable fragment (Fv) domain of the antibody, or it may be a single-chain antibody fragment which is a fusion protein in which the variable regions of the light and heavy chains of the antibody are linked by a peptide linker. The antibody may also be a single-domain antibody consisting of a heavy chain or a variable domain isolated from the light chain. Furthermore, the antibody may be a heavy-chain antibody consisting only of a heavy chain, such as antibodies found in camelids. The antibody-like molecule may also be a repeat protein such as the designed ankyrin repeat protein (Molecular Partners, Zurich).

[0131] All fragments must retain an antigen-binding moiety characterized by specifically binding to either CD47 or SIRPα in order to function according to the present invention. In other specific embodiments, the inhibitor is an antibody fragment which comprises an antigen-binding moiety characterized by specifically binding to CD47. In other specific embodiments, the inhibitor is an antibody fragment comprising an antigen-binding moiety characterized by specifically binding to SIRPα. In other specific embodiments of the present invention, the inhibitor of the interaction between CD47 and SIRPα comprises or is essentially the ligand-binding extracellular domain of CD47 or SIRPα, or a variant of the ligand-binding extracellular domain of CD47 or SIRPα. In specific embodiments, the inhibitor portion of the fusion protein is essentially the extracellular domain of the SIRPα polypeptide. In specific embodiments, the inhibitor of the interaction between CD47 and SIRPα is selected from the CD47-binding moiety of ALX148, TTI-622, or TTI-621. ALX148 is a fusion of a modified SIRPα D1 domain and an inactive human IgG1 Fc, binding to CD47 with a Kd of 1E-9M (Kauder S. et al. 2018, PloS One 13(8):e0201832). TTI-662 and TTI-621 are extracellular binding domains of human SIRPα fused to IgG4 or IgG1 Fc, respectively (Trillium Therapeutics).

[0132] Other anti-SIRPα antibodies useful as SIRPα inhibitors include, for example, KWAR23 (Ring et al., 2017, Proc Natl Acad Sci 114(49):E10578-E10585), CC-95251 (Celgene), and Bl765063 (also known as OSE-172; OSE Immunotherapeutics). Further examples of anti-SIRPα antibodies are described in WO2019 / 023347, WO2013 / 056352, and WO2018 / 190719.

[0133] Other such antibodies include, but are not limited to, GS-0189 (Gilead, formerly FSI-189), ES-004, and ADU1805 (Voets et al., J Immunother Cancer. 2019 Dec 4;7(1):340).

[0134] In certain embodiments, the SIRPα inhibitor is a soluble CD47 polypeptide, for example, as described in U.S. Patent No. 2010 / 0239579. In certain embodiments, the soluble CD47 polypeptide comprises an extracellular domain of CD47 containing a signal peptide (SEQ ID NO: 2 of WO2016 / 118754), the extracellular portion of CD47 being typically 142 amino acids long and having the amino acid sequence described in SEQ ID NO: 3 of WO2016 / 118754. The soluble CD47 polypeptides described herein also include CD47 extracellular domain variants comprising at least 65%–75%, 75%–80%, 80%–85%, 85%–90%, or 95%–99% (or any percentage of identity between 65% and 100% not specifically enumerated) of the amino acid sequence, which retain the ability to bind to SIRPα without stimulating SIRPα signaling.

[0135] In certain embodiments, the fusion protein initially comprises a β2m polypeptide, an HLA B57 class I polypeptide, an IgGFc polypeptide, and an antigen-binding fragment of the anti-CD47 antibody Hu5F9-G4.

[0136] In certain embodiments, the fusion protein initially comprises a β2m polypeptide, an HLA B57 class I polypeptide, an IgGFc polypeptide, and an antigen-binding fragment of the anti-CD47 antibody TI-061.

[0137] In certain embodiments, the fusion protein initially comprises a β2m polypeptide, an HLA B57 class I polypeptide, an IgGFc polypeptide, and an antigen-binding fragment of the anti-CD47 antibody IBI188.

[0138] In certain embodiments, the fusion protein initially comprises a β2m polypeptide, an HLA B57 class I polypeptide, an IgGFc polypeptide, and an antigen-binding fragment of the anti-CD47 antibody CC-90002 (Inhibrx).

[0139] In certain embodiments, the fusion protein initially comprises a β2m polypeptide, an HLA B57 class I polypeptide, an IgGFc polypeptide, and an antigen-binding fragment of the anti-CD47 antibody AO-176 (Arch Oncology).

[0140] In certain embodiments, the fusion protein initially comprises a β2m polypeptide, an HLA B57 class I polypeptide, an IgGFc polypeptide, and an antigen-binding fragment of the anti-CD47 antibody remzoparimab / TJC4 (I-MAB Biopharma).

[0141] In certain embodiments, the fusion protein initially comprises a β2m polypeptide, an HLA B57 class I polypeptide, an IgGFc polypeptide, and an antigen-binding fragment of the bispecific anti-CD47 antibody SL-172154 (Shattuck Labs).

[0142] In certain embodiments, the fusion protein initially comprises a β2m polypeptide, an HLA B57 class I polypeptide, an IgGFc polypeptide, and an antigen-binding fragment of the anti-CD47 antibody IMC-002 (ImmuneOncia Therapeutics).

[0143] In some embodiments, the inhibitors of the interaction between CD47 and SIRPα used in the present invention can sterically block the interaction between CD47 or SIRPα and its binding partner upon binding. In other embodiments, this interaction is a non-agonist interaction.

[0144] In certain embodiments, the polypeptide inhibitor of the interaction between cellular CD47 and cellular SIRPα is the SIRPα extracellular domain polypeptide.

[0145] In a particular embodiment, the polypeptide inhibitor of the interaction between cell CD47 and cell SIRPα is characterized by SEQ ID NO: 011. In a particular embodiment, the polypeptide inhibitor of the interaction between cell CD47 and cell SIRPα is characterized by SEQ ID NO: 012. In a particular embodiment, the polypeptide inhibitor of the interaction between cell CD47 and cell SIRPα is characterized by SEQ ID NO: 013. In a particular embodiment, the polypeptide inhibitor of the interaction between cell CD47 and cell SIRPα is characterized by SEQ ID NO: 014. In a particular embodiment, the polypeptide inhibitor of the interaction between cell CD47 and cell SIRPα is characterized by SEQ ID NO: 015. In a particular embodiment, the polypeptide inhibitor of the interaction between cell CD47 and cell SIRPα is characterized by SEQ ID NO: 016.

[0146] The inventors envision any known SIRPα variant, particularly one that includes a wild-type sequence modified at position 53 (K53R) to enhance affinity for CD47 (see SEQ ID NO: 014). In certain embodiments, the fusion protein comprises, from N-terminus to C-terminus, a β2m polypeptide, a peptide linker, an HLA heavy chain polypeptide, an IgG fragment crystallizable (Fc) region of an immunoglobulin heavy chain fragment polypeptide, another peptide linker, and SEQ ID NO: 014 of the SIRPα variant. This arrangement is favorable for signal transduction by both the HLA polypeptide and SIRPa components, while simultaneously providing a dimer compound with desirable high yield.

[0147] In certain embodiments, the fusion protein consists of an HLA heavy chain polypeptide, an IgG fragment crystallizable (Fc) region of an immunoglobulin heavy chain fragment polypeptide, and SEQ ID NO: 014 of the SIRPα variant, from the N-terminus to the C-terminus.

[0148] Fusion construct containing any (optional) crystallizable (Fc) fragments The fusion protein according to the present invention comprises a β2m polypeptide, an HLA heavy chain polypeptide (the extracellular domain of the HLA heavy chain, or a portion thereof), and a CD-47 SIRPα interaction inhibitor. When used as a pharmaceutical, the heavy chain has the advantage of being stabilized by incorporating a protein moiety that enhances the manufacturability and stability of the drug and improves its plasma half-life. In a particularly advantageous embodiment, the inventors have found that these advantages can be obtained by conjugating the heavy chain and the inhibitor to an immunoglobulin crystallizable fragment (Ig Fc) polypeptide.

[0149] Further advantages of the presence of a stabilizing portion of the fusion protein include its ability to confer stability between expression and purification, increased yield and solubility by reducing the degradation and oligomerization of the fusion protein, and increased viability of cells expressing the fusion protein.

[0150] In certain embodiments, the stabilizing polypeptide of the fusion protein is a human IgFc polypeptide. Furthermore, the IgFc moiety can extend the in vivo half-life of the molecule by binding to the recycling receptor.

[0151] The present invention will be further described in terms of such fusion proteins having an IgG Fc domain, but those skilled in the art will understand that other forms of stabilization may be available to provide similar advantages.

[0152] In certain embodiments, the stabilizing polypeptide is isotype IgG Fc. The IgG Fc stabilizing peptide domain offers further advantages in the purification of HLA fusion proteins by enabling absorption onto a surface coated with protein A or G. We believe that bovine serum albumin could be a possible alternative that offers similar advantages when conjugated to a functional HLA polypeptide instead of Ig Fc. Our previous research has established that albumin molecules, such as bovine serum albumin, can also function as stabilizing polypeptides. It is also known that PEGylation can extend the half-life of circulating proteins, and therefore this is another feasible stabilizing peptide.

[0153] In certain embodiments, the fusion protein comprises, from the N-terminus to the C-terminus, an HLA heavy chain polypeptide, a crystallizable (Fc) region which is an IgG fragment of an immunoglobulin heavy chain fragment polypeptide, and a polypeptide inhibitor of the interaction between cellular CD47 and cellular SIRPα.

[0154] Here again, as described above, Fc may be any immunoglobulin subtype, with particular advantages provided by IgG Fc. Within embodiments where Fc is IgG-Fc, any subclass or isotype can be selected. Known mutations in IgG1 and IgG4, such as mutations that silence ADCC activity, e.g., LALAPG mutations in IgG1, are conceivable.

[0155] The inventors obtained favorable results by using IgG (particularly IgG4 crystallizable fragment (Fc) (see SEQ ID NO: 009)) inserted between the HLA polypeptide and the inhibitor polypeptide. While not wishing to be constrained by theory, the inventors hypothesize that IgG-Fc may play various roles in contributing to the efficacy of the fusion protein, including the following: A. Enhanced expression B. Role as an ideal linker, or role in effectively widening the synaptic distance between tumor cells and immune cells. C. Fc effector function (here, CD16 binding and ADCC (antibody-dependent cell-mediated cytotoxicity) are considered undesirable). Fc may facilitate binding and cross-linking with other FcgRs (Fc gamma receptors), and further facilitate the binding of FcRn (neonatal Fc receptor) for half-life extension via FcRn recycling.

[0156] IgG4-derived Fc is a desirable isotype for therapeutic fusion proteins due to its low cytotoxicity. In more specific embodiments, the HLA fusion protein contains a modified IgG4 S228P.dk molecule having the sequence of SEQ ID NO: 009. This is characterized by a mutation in the hinge region of IgG4, where proline (P) at position 228 of the original IgG4 antibody is replaced with serine (S), and dK represents a deletion of lysine (K), the last amino acid on the original IgG4 sequence. These modifications stabilize the IgG4 format and reduce heterogenicity. Both modifications are well-established and commonly used in a variety of Fc constructs.

[0157] Fc polypeptides can also be selected from a variety of Fc sequences known to those skilled in the art, which have been engineered to exhibit reduced or absent Fc-type effector function in order to suppress antibody-dependent cell-mediated cytotoxicity (ADCC) or antibody-dependent cell-mediated phagocytosis (ADCP). These include the LALA (IgG1 L234A / L235A) variant (Lund et al., J Immunol. 1991; 147: 2657-2662; WO 1988007089), LALAPG (WO 2012130831), and the LULAG variant (Wilkinson et al., PLOS ONE / https: / / doi.org / 10.1371 / journal.pone.0260954; PCT / GB2021 / 051233); see also Schlothauer et al., Protein Engineering, Design and Selection, Volume 29, Issue 10, October 2016, Pages 457-466. Other Fc variants envisioned in the art include the G236R, L328R variants (U.S. Patent Application Publication (US) 20060235208); the L234A, G237A variants (US2009060906); the L234A, L235E variants (WO2018044948), the D265A, P329A variants (WO2016207858); the L234F, L235E, P331S variants (WO2009100309); and the variants disclosed in WO2013165690, WO2011066501 and WO1999058572, all of which are incorporated herein by reference in their entirety. The location of Fc substitutions is determined by referring to the EU index, similar to Kabat (Sequences of Proteins of Immunological Interest, Part 5, U.S. Public Health Service, National Institutes of Health (NIH), Bethesda, Maryland, USA).

[0158] Furthermore, it is conceivable that IgG-Fc may not be used. For example, simply directly linking the β2m-HLA domain to the SIRPα-CD47 inhibitory domain by other means such as a polypeptide linker or HSA (human serum albumin) would yield the same (A) enhanced expression, (B) linking, and a portion of (C) with FcRn binding.

[0159] In more specific embodiments, the HLA heavy chain polypeptide and Ig Fc polypeptide of the fusion protein are linked by a peptide linker. In specific embodiments, the peptide linker has an amino acid length between 5 and 20. In even more specific embodiments, this binding peptide linker has the sequence of SEQ ID NO: 004.

[0160] In certain embodiments, the encoded fusion protein further includes a secretory signal upstream of the HLA sequence. The signal sequence is cleaved during the secretory and expression processes. These sequences are typically not part of the biologically active construct. These sequences are modified and optimized during the progression of the process and the generation of the master cell bank.

[0161] In more specific embodiments, the fusion protein essentially consists of the sequence specified by SEQ ID NO: 018, 020, or 022 (containing a β2m polypeptide, a variant extracellular domain of HLA-B57 fused to IgG4 Fc) and SIRPα.

[0162] In certain embodiments, the fusion protein comprises an HLA polypeptide linked to an IgG polypeptide as part of a single polypeptide chain by a peptide linker, which is a short sequence of amino acids having a length of 5, 10, 15, or 20 residues. In certain embodiments, the peptide linker is a non-immunogenic sequence, such as a serine and glycine-rich sequence like SEQ ID NOs: 005, 058, 059, or 060.

[0163] In more specific embodiments, the peptide linker is about 15 amino acids long, which has been shown to be related to favorable amounts of yield and efficacy. In more specific embodiments, the peptide linker has the sequence of SEQ ID NO: 059.

[0164] In more specific embodiments, the fusion protein consists essentially of a sequence designated as SEQ ID NO: 049, 053, 054, or 055 (including a β2m polypeptide, an extracellular domain variant of HLA-B57 fused to IgG4 Fc, and a variant SIRPα). In other embodiments, the fusion protein consists essentially of a sequence designated as SEQ ID NO: 056 (including a β2m polypeptide, an extracellular domain variant of HLA-A24 fused to IgG4 Fc, and a variant SIRPα). * 57 extracellular domain variant, and a variant SIRPα). * 24 extracellular domain variant, and a variant SIRPα).

[0165] In certain embodiments, the fusion protein is in a dimeric format. The dimer includes a first monomer and a second monomer, and each monomer independently includes a β2m polypeptide, an HLA heavy chain extracellular domain polypeptide fused to an Ig Fc portion (the Ig Fc portions may associate via disulfide bonds), and a SIRPα portion.

[0166] The Fc fragment may form a homodimer, particularly by interaction in the hinge region.

[0167] Another aspect of the invention relates to a fusion protein as described above, or a dimeric protein (dimer) comprising two fusion proteins. In certain embodiments, the fusion protein includes the Fc region of human immunoglobulin G (IgG).

[0168] In more specific embodiments thereof, the Fc region is characterized by SEQ ID NO: 009 or 010, or a sequence that is at least 95% identical to SEQ ID NO: 009 or 010.

[0169] Any Fc sequence known in the art to be useful for protein stabilization may be employed, as may variants of that Fc sequence that maintain the stability of the protein scaffold, up to an identity difference of 85% or more relative to the sequence number (Fc sequence).

[0170] In certain embodiments, both polypeptide chains are characterized by the same amino acid sequence.

[0171] In certain embodiments, the two fusion protein chains are characterized by different amino acid sequences.

[0172] In one particular embodiment of this aspect of the present invention, the dimeric protein comprises two polypeptide chains comprising or consisting of the amino acid sequence SEQ ID NO: 049.

[0173] In one particular embodiment of this aspect of the present invention, the dimeric protein comprises two polypeptide chains comprising or consisting of the amino acid sequence SEQ ID NO: 053.

[0174] In one particular embodiment of this aspect of the present invention, the dimeric protein comprises two polypeptide chains comprising or consisting of the amino acid sequence SEQ ID NO: 054.

[0175] In one particular embodiment of this aspect of the present invention, the dimeric protein comprises two polypeptide chains comprising or consisting of the amino acid sequence SEQ ID NO: 055.

[0176] In one particular embodiment of this aspect of the present invention, the dimeric protein comprises two polypeptide chains comprising or consisting of the amino acid sequence SEQ ID NO: 056.

[0177] nucleic acid sequence encoding the fusion protein according to the present invention Another aspect of the present invention relates to nucleic acid sequences encoding fusion proteins defined according to the above aspects of the present invention, and any embodiment thereof.

[0178] Furthermore, a pharmaceutical composition is provided comprising a fusion protein according to any one of the above embodiments and examples of the present invention, and a pharmaceutically acceptable excipient.

[0179] Similarly, the use of fusion proteins in pharmaceuticals is also included in the present invention, and in particular, its use in the treatment or prevention of cancer.

[0180] Dosage and administration of HLA fusion protein pharmaceutical compositions Further aspects of the present invention relate to pharmaceutical compositions, in particular to use for the treatment of forms of cancer or malignant neoplastic diseases, the compositions comprising fusion proteins as defined herein.

[0181] The pharmaceutical compositions according to the present invention, in particular pharmaceutical compositions for use in the treatment of cancer, comprise a fusion protein disclosed herein and are typically formulated into pharmaceutical dosage forms to provide a readily controllable drug dosage form and a clear and easy-to-handle product for the patient. The pharmaceutical composition further comprises a pharmaceutically acceptable carrier. In further embodiments, the composition comprises at least two pharmaceutically acceptable carriers.

[0182] The administration regimen of the pharmaceutical composition of the present invention varies depending on known factors such as the pharmacodynamic properties of the specific drug and its mode and route of administration; the recipient's species, age, sex, health status, condition, and weight; the nature and severity of symptoms; the type of concurrent treatment; the frequency of treatment; the route of administration, the patient's renal and hepatic function, and the desired effect. In certain embodiments, the pharmaceutical composition of the present invention may be administered once daily, or the total daily dose may be divided into two, three, or four doses per day. In certain embodiments, the concomitant drug is administered every one, two, or three weeks.

[0183] Many procedures and methods for preparing pharmaceutical compositions are known in the art; see, for example, L. Lachman et al., Theory and Practice of Industrial Pharmacy, Vol. 4, 2013 (ISBN 8123922892).

[0184] Medical and manufacturing methods In certain embodiments, the fusion proteins according to the present invention are provided for use in the treatment of various forms of cancer. Preclinical studies of other forms of similar LILRB2-targeted anti-cancer therapies have demonstrated efficacy in renal and ovarian cancer. The safety and tolerability of the LILRB2-targeted antibody MK-4830 are currently being investigated in clinical trials in combination with chemotherapy for a wide range of solid organ cancers (mesothelioma, triple-negative breast cancer, ovarian cancer, lung cancer, glioblastoma, pancreatic cancer, and gastric cancer) (ClinicalTrials.gov Identifier: NCT03564691). Preclinical studies using similar LILBR2-targeted anti-cancer therapies containing HLA-A, B, or C heavy chain polypeptides have demonstrated efficacy in lymphoma, leukemia, myeloma, lung cancer, colon cancer, pancreatic cancer, and melanoma (see WO2017153438(A1);WO2016124661(A1);WO2018029284(A1);WO20230147(A1)). Each of these indications may be treatable with the LILBR2-binding fusion protein according to the present invention.

[0185] In certain embodiments, the fusion protein according to the present invention is provided for use in treating hematological cancers. The inventors believe that the characteristic T cell exhaustion induced by the pharmaceutical composition according to the present invention, as well as the accessibility of circulating hematological cancer cells to T cells and macrophage activation, means that this fusion protein is likely to be effective. In particular, in such embodiments, the fusion protein is provided for use in patients diagnosed with multiple myeloma, such as modeled by the RPMI-8226 cell line (Figure 4). In other specific embodiments, the fusion protein is provided for use in patients diagnosed with solid tumors or metastatic solid tumors. In other specific embodiments, the fusion protein is provided for use in patients diagnosed with a form of breast cancer. In more specific embodiments, the cancer is estrogen receptor positive. In other specific embodiments, the cancer is progesterone receptor positive. In other specific embodiments, the cancer is human epidermal growth factor receptor 2 positive.

[0186] In certain embodiments, the fusion protein is provided for use in treating patients diagnosed with a form of lung cancer. In even more specific embodiments, the fusion protein is provided for use in treating non-small cell lung cancer (NSCLC) in diagnosed patients (see Figure 14).

[0187] Similarly, the scope of the present invention includes methods for treating cancer patients, comprising administering the fusion protein according to the present invention to a patient. In a further different aspect, the present invention further encompasses the use of the fusion protein as detailed above for use in the production of pharmaceuticals for the treatment or prevention of cancer.

[0188] The dosage form may be parenteral administration, such as subcutaneous injection, intravenous injection, intrahepatic injection, or intramuscular injection. Optionally, pharmaceutically acceptable carriers and / or excipients may be present.

[0189] The present invention can be further illustrated by the following embodiments and figures, from which further embodiments and advantages can be derived. These embodiments are for illustrative purposes only and do not limit the scope of the present invention. [Brief explanation of the drawing]

[0190] [Figure 1] Figure 1(A) shows how the interaction of SIRPα and LILRB2 on macrophages associates with their native ligands on tumor cells and leads to immunosuppressive signals. (B) outlines an experimental design to investigate whether blocking LILRB1 / 2 with the HLA fusion protein candidate iosH2, thereby releasing the brakes on macrophage-mediated cancer cell killing, can be combined with blocking the SIRPα / CD47 axis using antibodies or native ligands to enhance tumor-killing activity. [Figure 2AB] Figure 2 shows the superior cell viability and expression characteristics of the HLA-B57(A46E / V97R)IgG4 fusion protein. Fusion protein expression from clones transfected with HLA B57.β2m(DGC8-T39, DGC8-T64, and DGC8-73) and HLA-B57(A46E / V97R).β2m(DGC8-T54, DGC8-T75, and DGC8-91) is shown based on cell viability (A) and expressed protein titer (B). [Figure 2B] The table summarizes the yields at different transfection ratios tested. [Figure 2CD] (C) Equivalent RNA profiles of HLA-B57(A46E / V97R)IgG4 fusion protein and (D) β2m expressed from a vector within CHO cell clones. [Figure 3]Figure 3 shows quantitative estimations of the binding affinity of LILRB2 to non-β2m-associated HLA-B57 or HLA-B57(A46E / V97R) fusion protein and HLA-B57(A46E / V97R).β2m, as measured by ELISA. The EC50 of β2m-free HLA-B57 was 21 nM, the EC50 of HLA-B57(A46E / V97R) was 8.3 nM, and the EC50 of HLA-B57(A46E / V97R).β2m was 5.72 nM, indicating that amino acid substitution does not reduce the binding of the HLA-B57 heavy chain to LILRB2. [Figure 4A] Figure 4 shows the increased phagocytic activity of HLA-B57(A46E / V97R).β2m(iosH2) combined with anti-SIRPα antibody or anti-CD47 antibody in (A) liquid cancer using RPMI8226 multiple myeloma cells and (B) solid cancer using BT474 breast cancer cells. [Figure 4B] Figure 4 shows the increased phagocytosis of (A) liquid tumors using RPMI8226 multiple myeloma cells and (B) solid tumors using BT474 breast cancer cells, using HLA-B57(A46E / V97R).β2m(iosH2) in combination with anti-SIRPα antibody or anti-CD47 antibody. Cancer cell lines derived from solid and liquid tumors were co-cultured with human primary macrophages, and phagocytosis was measured for 36 hours in the IncuCyte live-cell imaging system. The experiment was repeated using at least two biologically independent samples. IgG1 10ug / ml, IgG4 20ug / ml, iosH2 20ug / ml, anti-CD47 antibody 10ug / ml, MK4830 10ug / ml, anti-SIRPα antibody 10ug / ml. Error bars represent SEM of n=2 biological repeats, each including two technical repeats. Statistical analysis was performed using two-way repeated measures ANOVA, followed by Dunnett's post-hoc analysis, with p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. [Figure 5]Figure 5 is a schematic diagram of a structure according to the present invention. A) A dimer of a fusion protein comprising a β2m-HLA containing a1, a2, and a3 domains, IgG Fc, and a SIRPα extracellular domain (see claims 2, 3; SEQ ID NOs: 17, 18); B) A dimer of a fusion protein comprising a β2m-HLA composed of two a3 domains, IgG Fc, and a SIRPα extracellular domain (see claims 4, 5; SEQ ID NOs: 19, 20); C) A dimer of a fusion protein comprising two β2m sequences and an HLA polypeptide composed of two a3 domains, IgG Fc, and a SIRPα extracellular domain (all parts are linked by a short peptide linker). [Figure 6] Figure 6 shows a series of HLA fragment constructs. The fragment constructs were designed as dimers by ligating a repeating number of HLAB57(α3) domains to a c-terminal fused human IgG4-Fc(S228P). [Figure 7] Figure 7 shows a series of SIRPα bispecific constructs and controls. The SIRPα bispecific constructs were designed as dimers by ligating a B2m-binding HLA domain to a c-terminal fused human IgG. The control molecules were designed as dimers by fusing the V1 domain of human SIRPα to either the N-terminus or C-terminus of human IgG. [Figure 8] Figure 8 shows the binding of the compound to human CD47 in Jurkat T cells. Binding was analyzed by flow cytometry. EC50 values ​​(nM) were generated using nonlinear regression fitting of data to the three-parameter response to the agonist using Prism (Graphpad), and the best-fitted values ​​were shown. [Figure 9] Figure 9 shows the binding of compounds to the LILRB1 and LILRB2 receptors expressed in transduced CHO cells. Binding was analyzed by flow cytometry. EC50 values ​​(nM) were generated using nonlinear regression fitting of data to the three-parameter response to the agonist using Prism (GraphPad), and the best-fitted values ​​were shown. [Figure 10]Figure 10 shows that affinity optimization of the SIRPα arm reduced erythrocyte binding compared to maglorimab and TTI-622 (top). Furthermore, the erythrocyte agglutination ability of the compounds was tested compared to competing molecules (bottom). The longer the linker connecting HLA and SIRPa, the less erythrocyte agglutination was observed. In addition, modifications to the HLA arm improved the erythrocyte agglutination profiles of IOS-2057, IOS-2058, IOS-2059, and IOS-2060. [Figure 11] Figure 11 shows the inhibition rate (%) of CD47 / SIRPa axis signaling during compound treatment using a reporter cell-based system. The inhibition rate was calculated by fitting the data to a three-parameter response to the inhibitor using nonlinear regression, and the best-fit value is shown. [Figure 12] Figure 12 shows the phagocytic activity of macrophages in a myeloma cell line (RPMI-8226) using HLA / SIRPα bispecificity. The results indicate that a longer linker between Fc and SIRPα leads to a superior phagocytic response. Different symbols indicate different biological donors, and each point indicates a technical replica. [Figure 13] Figure 13 shows that macrophage phagocytosis of myeloma cells (RPMI-8226) is enhanced using IOS-1155 (B2M-HLA_B57(A46E / V97R / I194T)-IgG4(S228P)Fc), IOS-1161 (B2M-HLA_B57DM(A46E / V97R / H197F)-IgG4(S228P)Fc), and IOS-1164 (B2m-HLA-A24(M97R)-IgG4-Fc) compared with IgG4 isotype and IOS-1131 (B2m-HLA-B57(A46E / V97R)). [Examples]

[0191] Example 1: Generation of HLA fusion protein In previous research by the present inventors, the HLA fusion protein (iosH1) was designed by ligating the heavy chain extracellular domain of the HLA-B57:01:01 polypeptide (SEQ ID NO: 006) to the IgG4 Fc polypeptide. To increase the yield of this HLA fusion protein, amino acids associated with low yield in the recombinant cell expression environment were identified in the amino acid sequence of the natural HLA-B57 extracellular domain. This was modified by substituting the alanine (A) residue at position 46 with glutamine (E) and the valine (V) residue at position 97 with arginine (R), providing a variant HLA-B57 polypeptide (SEQ ID NO: 007). This was then fused to the IgG4 polypeptide via a linking peptide (SEQ ID NO: 004) to provide a variant HLA-B57 fusion protein. Recombinant HLA-B57 (A46E / V97R) The cDNA encoding the fusion protein, and a natural HLA-B57-derived fusion protein control lacking two mutations, were cloned downstream of the nucleic acid sequence encoding the secretion signal into a commercially available expression vector (Probiogen). HLA-B57-Fc and HLA-B57 (A46E / V97R) A vector construct expressing -Fc was co-transfected into Chinese hamster ovary (CHO) cells by microporation (MP) using the NEON Transfection Kit (Life Technologies #MPK10096) along with a plasmid containing nucleic acid encoding the β2m protein (SEQ ID NO: 001). CHO-DG44 starter cells were transfected with varying ratios of HLA fusion protein to β2m plasmid (4:1, 2:1, 1:1, 1:2). Selected clone pools were cultured in standard shaker flasks in 125 mL of PBG-CD-C4 medium supplemented with puromycin and methotrexate at a specified cell seeding density of 4E5 vc / mL. After selective pressure adjusted with antibiotics, individual clone pools were selected for analysis. Viability and viable cell density were measured using the Vi-CELL XR system and trypan blue cell exclusion. Titer quantification was performed using an Octet RED instrument (ForteBio, Pall Division) equipped with a protein A biosensor, and measured at different time points (number of days).

[0192] Analysis of clones expressing HLA-B57 or variant fusion proteins showed that the survival rate and titer of wild-type HLA-B57.β2m cells were significantly lower than those of HLA-B57 (A46E / V97R) .β2m (Figures 2A and B). Equimolar amounts of RNA of native or modified HLA-B57 (A46E / V97R) against β2m were confirmed in the selected clone cells (Figures 2C and D), suggesting that the amino acid modification increased the expression.

[0193] HLA-B57 (A46E / V97R) .β2m complex was isolated from the filtered CHO cell supernatant by affinity column purification. Protein purification under acidic conditions and removal of β2m were performed as a two-step purification protocol. As the first step, Protein G Sepharose [(4 Fast Flow) Sigma, #GE-17-0618-01)] beads were used to capture HLA-B57 (A46E / V97R) associated with β2m from the supernatant. After incubating overnight at 4°C in a rocker, the recovered beads were washed with PBS, and then the HLA-B57 (A46E / V97R) fusion protein in which each HLA polypeptide was still associated with β2m was eluted using standard IgG-Elution Buffer (pH 2.8) (Pierce™ IgG Elution Buffer, Thermo Fischer #21004). As the second step of purification based on size exclusion chromatography, HLA-B57 (A46E / V97R) was separated from β2m under acidic conditions to obtain an HLA fusion protein not associated with β2m. For separation, a Superdex10 / 300 gel filtration column pre-equilibrated with sodium citrate (100 mM, pH 3.0) was used. A protein at a concentration of 2.0 mg / ml was injected at 0.5 ml, and the peak of the target HLA-B57 (A46E / V97R) protein eluted at 12.7 ml, and the peak of β2m eluted at 22.0 ml.

[0194] Example 2: Quantification of the interaction between LILRB2 and HLA fusion proteins that are not associated with β2m or are associated with β2m. The inventors of the present invention have identified HLA-B57 (A46E / V97R) Detailed analysis of the immunological properties most relevant to tumor immunity associated with the format revealed that they all lacked association with peptide epitopes associated with the peptide binding groove of the HLA class I domain of the fusion protein. (Examples: wild-type HLA-B57 fusion protein (dimer) lacking β2m, HLA-B57 lacking β2m) (A46E / V97R) Fusion protein (dimer) and HLA-B57 still bound to β2m (A46E / V97R) The affinity of the interaction between the fusion protein dimer and LILRB2 was measured using enzyme-linked immunosorbent assay (ELISA). 50 μl of c-terminal biotinylated antigen molecules (LILRB2, BPS Bioscience #100335), immobilized at a final concentration of 5 μg / ml in PBS buffer, were coated onto a flat-bottomed Pierce® streptavidin-coated 96-well plate (Pierce #15500). PBS and IgG isotypes were used as negative controls. HLA-B57 or HLA-B57 lacking β2m was used. (A46E / V97R) , and HLA-B57 (A46E / V97R) β2m serial dilutions (8 concentration points: 10, 2.5, 1, 0.25, 0.1, 0.025, 0.01, 0.0025 μg / ml) were applied in duplicate (50 μl). For detection, APC-conjugated goat anti-human IgG antibody (Jackson Immuno Research #109-135-098) diluted 1:100 with TBS (50 μl) was used. Finally, 50 μl of TBS was added to each well, and fluorescence scanning was performed at the excitation wavelength of APC (650 nm) and emission wavelength of 660 nm. Graphpad Prism v9.1.2 and a 3-parameter-based log(agonist)-pair response model were used to determine the EC50 of the interaction. Binding of variant HLA-B57 fusion protein to LILRB2 is not associated with HLA-B57 β2m. (A46E / V97R)Improvement was observed in both cases, and especially when associated with β2m, suggesting that this variant possesses high immunomodulatory capacity (Figure 3).

[0195] Example 3: Increased tumor cell killing power in vitro Next, HLA-B57 associates with β2m. (A46E,V97R) The ability of IgG4 Fc fusion protein (referred to as iosH2) to induce phagocytosis of tumor cells by human primary macrophages was evaluated against liquid cancer cells (RPMI8226 multiple myeloma, DSMZ, catalog number ACC402) and solid cancer cells (BT474 ductal carcinoma, DSMZ, catalog number ACC64) in comparison with PBS, IgG1 (Biolegend; catalog number 403502) and IgG4 (Biolegend; catalog number 403702) isotype controls, and a benchmark anti-LILRB2 antibody (MK4830) (in-house produced based on the sequence of U.S. Patent Application Publication No. 2018 / 0298096 (A1)). Furthermore, to investigate whether the effects of iosH2 can be enhanced by complexly inhibiting two pathways that regulate phagocytic cell activation (Figure 1), we evaluated concurrent treatment using either an anti-CD47 antibody (LubioScience; catalog number BE0019-1) or an anti-SIRPα antibody (in-house manufactured based on the sequence of WO2020068752(A1)).

[0196] The indicated liquefied and solid tumor-derived cancer cell lines were co-cultured with human primary macrophages, and the phagocytic activity of tumor cells was monitored for 36 hours using an IncuCyte incubator and live-cell imaging system (Vitaris AG, MCO-230AICUVH-PE) according to the manufacturer's instructions. Cancer cells were stained with CellTrace® CFSE (ThermoFisher) according to the manufacturer's instructions, and then 1000 cells / well were seeded together with 1000–5000 primary T cells in a flat-bottomed 96-well plate (Greiner). The culture medium contained 250 nM Cytotox Red (Sartorius). Live-cell imaging was performed using the IncuCyte S3 Live-Cell Analysis System (Sartorius). Four non-adjacent images per well were analyzed using IncuCyte software v2020C. CFSE signals were separated by green objects, and all objects were counted as cancer cells. Dead cells were identified by the Cytotox signal, which is delimited by red objects. Cancer cell death was detected by the colocalization of green and red objects.

[0197] The results in Figure 4 are for HLA-B57. (A46E,V97R)The study showed that β2m(iosH2) significantly and rapidly increased macrophage phagocytosis as early as 3 hours after co-culture. Although the rate and dynamics of phagocytosis differed among cell lines, the effect of iosH2 was independent of cancer cell type. IosH2 stimulation was also superior to the effect induced by the benchmark anti-LILRB2 antibody MK4830, which induced delayed phagocytosis in breast cancer cells but had no effect on myeloma cells. Anti-CD47 antibodies or anti-SIRPα antibodies did not affect the phagocytosis rate in myeloma cells when used alone; however, the combination of iosH2 and anti-CD47 antibody increased the phagocytosis rate in myeloma cells, which was not observed with the combination of anti-CD47 antibody and MK4830. In the breast cancer cell line BT474, anti-CD47 antibody also enhanced IosH2-induced phagocytosis, but had little effect when used alone. Combination therapy with anti-SIRPα antibody and iosH2 showed the highest phagocytosis rate compared to all monotherapy and combination therapies tested in breast cancer, but it was ineffective in treating myeloma cell lines.

[0198] Example 4: Dual-specific construct [Table 2]

[0199] 1) Note: In some cases, production yields were higher than the Octet-based quantitative titer, showing recovery rates exceeding 100% after two purifications (affinity + SEC). Titer estimation using Octet-based quantitative methods appears to significantly underestimate the actual expressed protein. 2) The expression and production yields of IOS-2024 and IOS-2025 are the average values ​​from two expression cycles and three production cycles. 3) IOS-2027 and IOS-2028 did not proceed to the purification process. 4) Quality control (QC) using SEC-HPLC has not been performed for IOS-2035, IOS-2037, and IOS-3039.

[0200] Manufacturing and Characterization of Bispecific Constructs The inventors designed a bispecific construct with the following concept: B2M (SEQ ID NO: 001) is linked via a general-purpose linker (SEQ ID NO: 002) to either the optimized α1-α2-α3 domain of HLA B57 (e.g., SEQ ID NO: 007) or a repeat of only the HLA B57α3 domain, where the HLA B57α1-α2-α3 domain or HLA B57α3 domain repeat is linked via a general-purpose linker (SEQ ID NO: 004) to IgG4-Fc(S228P)dk (SEQ ID NO: 009), where the IgG4-Fc(S228P)dk is linked via a general-purpose linker (SEQ ID NO: 005) to various versions of Sirpα (e.g., SEQ ID NO: 011). Various designs were intended to: i) increase the stability and overall yield of bispecific constructs (e.g., IOS-2027 and IOS-2028); ii) reduce complexity by using only the α3HLA domain (e.g., IOS-2035 and IOS-2037); or iii) alter the interaction between the Sirpα domain and its natural binding partner CD47 (e.g., IOS-2024, IOS-2025, IOS-2026, IOS-2039, IOS-2041).

[0201] cDNA encoding recombinant bispecific fusion proteins was cloned into a commercially available expression vector (Evitria) downstream of a nucleic acid sequence encoding a secretory signal. Vector constructs expressing different bispecific designs were transfected into Chinese hamster ovary (CHO) cells by lipofection at a defined cell density and using a standardized protocol to allow comparison of production parameters based on design differences. Transient recombinant protein expression was performed in defined medium in a shaking flask. Viability and viable cell density were measured using the Vi-CELL XR system and trypan blue cell exclusion method. Titer quantification was performed at different time points (days) using an Octet RED instrument (ForteBio, Pall Division) equipped with a protein A biosensor.

[0202] Structures IOS-2027 and IOS-2028 showed very low titers, and no further processing was performed.

[0203] Next, if the yield was sufficient for further steps, the bispecific construct protein was isolated from the CHO cell supernatant filtered by affinity column purification and further purified using size exclusion chromatography. More specifically, in the first step, to capture the bispecific construct protein from the supernatant, Protein A Prisma [(HiTrap Fibro Prisma) Cytiva, #17549856] was used, the column was washed with PBS, and elution was performed using standard IgG-Elution Buffer (pH 2.8) (Pierce® IgG Elution Buffer, Thermo Fischer #21004). In the second step, purification was performed by size exclusion chromatography using a Superdex 16 / 600 (Cytiva; #28989335) gel filtration column, pre-equilibrium was performed with running buffer (PBS and 0.05 polysorbate 20), and peaks indicating the monomer fraction of the bispecific protein were collected and pooled. Protein concentrations after two-step purification were measured using UV absorption at a wavelength of 280 nm and a standard method with nanodrops. HLPC-SEC was performed using a TSKgel G3000 SW XL with the following running buffer: 50 mM sodium phosphate, 150 mM sodium chloride, 50 mM L-arginine, pH 7.4. Monomer peak content was calculated from the chromatogram.

[0204] The constructs showed high recovery rates after purification, indicating that titer estimation using Octet-based quantitative methods significantly underestimated the actual expressed protein content. As evident from the expression titers of IOS-2027 and IOS2028, further mutations in the HLA heavy chain domain using one of several combinations did not result in increased stabilization of the desired protein. By reducing the complexity of the fusion proteins expressed in IOS-2035 and IOS2037, high titers and yields were maintained compared to more complex proteins such as IOS-2024, which holds great promise for further development. Construct IOS-2041, with fewer mutations in the SIRPα domain, maintained high titers and expression. This was intended to maintain good biophysical properties while reducing the potential risk of immunogenicity.

[0205] Example 5: Functionality and manufacturing quality of bispecific fusion proteins Construction, expression, and purification of recombinant proteins Next, a series of SIRPα bispecific constructs and controls were prepared to investigate the functional and manufacturable qualities related to alternative sequences and arrangements of SIRPα or HLA features of the bispecific protein. Recombinant proteins in Tables 3 and 4, and Figures 6 and 7, were constructed as dimers with human IgG Fc polypeptide positioned at either the N-terminus or C-terminus. Control molecules were designed as dimers by fusing the V1 domain of human SIRPα to either the N-terminus or C-terminus of human IgG Fc polypeptide. Proteins were transiently expressed in CHO or HEK cell lines as shown. After affinity purification, proteins were purified from the supernatant using size exclusion chromatography (SEC) (Cytiva, HiLoad 16 / 600 Superdex 200pg column). Affinity purification was performed using a HiTrap PrismA protein A 0.4 mL chromatography column (Cytiva). Subsequently, the purified protein was concentrated using an Amicon Ultra (50kDa) concentrator (Millipore), and the stability and purity of the protein were measured using SEC-HPLC and SDS-PAGE.

[0206] Quantification of binding affinity by BLI and ELISA Comparison of constructs IOS-1225 and IOS-1228 showed that the N-terminal configuration was associated with the loss of CD47 receptor binding activity due to the conserved SIRPα variant (Table 3). In the context of bispecific fusion proteins, we observed that the reduced affinity of the N-terminal SIRPα wild-type sequence to CD47 in IOS-2024 could be overcome in IOS-2025 by introducing a single amino acid substitution K53R to restore binding to CD47 (Table 4). The length of the peptide linker connecting SIRPα and the fusion protein did not affect receptor binding. However, the long peptide linker of IOS-2046 was associated with lower expression titer, while the 10-15 amino acid peptide linkers of IOS-2044 and 2045 were associated with better manufacturing quality (Table 4).

[0207] The human receptor molecules LILRB1, LILRB2, and KIR3DL1 were constructed by ligation with C-terminal mouse IgG2aFc possessing a soluble extracellular domain and a distal Avi tag (GLNDIFEAQKIEWHE), and transiently expressed in CHO cells. The expressed LILRB1, LILRB2, and KIR3DL1 were further purified by affinity purification using HiTrap PrismA protein A followed by SEC. The c-terminal Avi tag in the purified LILRB1, LILRB2, and KIR3DL1 was biotinylated using BirA ligase (Avidity). The c-terminal biotinylated human CD47 receptor molecules were obtained from AcroBiosystems.

[0208] To measure binding affinity, Octet-based bio-layer interferometry (BLI) was used. For the measurement of equilibrium binding constants, the assay was set up so that c-terminally biotinylated receptor proteins (LILRB1, LILRB2, KIR3DL1, and CD47) were immobilized on a streptavidin (SAXS) biosensor, while mFc-bound LILRB1-mFc, LILRB1-mFc, LILRB2-mFc, and KIR3DL1-mFc proteins were immobilized on an anti-mouse IgG Fc Capture AMC biosensor (Sartorius), and the interaction between each analyte molecule was tested. Each biosensor was incubated while increasing the concentration of the analyte, and sensorograms were recorded. Data analysis, double-reference subtraction, and quantification of kinetic parameters and binding affinity were performed using Data Analysis HT 12.0.2.59 software. 50Enzyme-linked immunosorbent assay (ELISA) was used for quantification. Biotinylated LILRB1, LILRB2, KIR3DL1, and CD47 were coated in PBS at 5 μg / ml in a 96-well streptavidin plate (Pierce). Analyte proteins at different concentration series were added to the wells in protein buffer (PBS, 0.01% Tween-20). HRP-conjugated goat anti-human IgG antibody (Jackson ImmunoResearch) was used as the secondary antibody. Optical density (OD) values ​​were recorded at 450 nm using Infinite M Nano (Tecan). The data were analyzed, and EC was calculated using Graphpad Prism software (9.4.1) by nonlinear fitting of XY data to dose / response. 50 The result was calculated.

[0209] [Table 3] JPEG2026514293000010.jpg74146

[0210] [Table 4]

[0211] Cell binding assay Next, the binding of various representative constructs to human CD47 was measured using a direct binding assay. CD47+ human Jurkat T cells were incubated with different constructs in stepwise amounts ranging from 0.08 to 1185 nM, and binding was analyzed by flow cytometry using a polyclonal anti-IgG antibody (Jackson ImmunoResearch). Briefly, Jurkat cells were washed once with PBS buffer containing 2% FBS, and 50,000 cells per well were incubated with the compound at 37°C for 2 hours. The cells were washed again, a secondary antibody was added, and the cells were incubated at 4°C for 20 minutes. EC50 values ​​were prepared using Prism (GraphPad) with a formula for the response to the agonist (3 parameters). CD47 knockout (KO) cells were used as a negative control. The moderate affinity binding to CD47 achieved by the SIRPα polypeptide is desirable in the fusion protein of the present invention, as it enhances the safety of the compound by blocking CD47-based signaling, inducing phagocytosis, and targeting the CD47-rich tumor environment while reducing interactions with erythrocytes. The presence of HLA in the IOS-2041 fusion protein did not inhibit the binding of C-terminal SIRPα to CD47, as it showed similar binding to the construct lacking the HLA moiety (Figure 8).

[0212] Conversely, when comparing the binding of compounds to the LILRB1 and LILRB2 receptors, the presence of the SIRPα moiety did not inhibit the binding of the bispecific fusion protein compared to the control HLA IgG4 fusion protein (Figure 9). CHO LILRB1 cells or CHO LILRB2 cells (LILRB1 and LILRB2 lentiviral particles obtained from G&P Biosciences) were incubated with different constructs in stepwise amounts ranging from 0.08 to 1185 nM, and binding was analyzed by flow cytometry using polyclonal anti-IgG antibody (Jackson ImmunoResearch). Briefly, CHO cells were scraped from plates, washed once with PBS buffer containing 2% FBS, and incubated with the compound at 37°C for 2 hours at a rate of 50,000 cells per well. The cells were washed again, a secondary antibody was added, and incubated at 4°C for 20 minutes. EC50 values ​​were prepared using Prism (GraphPad) with the formula for the response to the agonist (3 parameters). As a negative control, CHO cells transduced with an empty control vector were used.

[0213] Erythrocyte binding and hemagglutination assays Undesirable binding to red blood cells is associated with the pathological characteristics of existing SIRPα-targeting agents such as maglorimab. To evaluate the potential of HLA-SIRPa bispecific constructs to bind to red blood cells (RBCs), heparinized whole blood samples were multiplexed to isolate RBCs and obtain a 10% (v / v) RBC solution in PBS. For the binding of the compounds to red blood cells, the compounds were incubated with red blood cells at concentrations ranging from 0.03 nM to 2000 nM for 30 minutes at 4°C and detected by flow cytometry using a secondary antibody. We showed that the binding of different HLA-SIRPa bispecific constructs was lower than that of maglorimab antibody or another splicing variant type of SIRPα present in TTI622 (Figure 10, top).

[0214] For the hemagglutination assay, isolation of RBCs was performed by transferring 2 ml of human heparinized whole blood into a 15 ml tube, supplementing with 10% phosphate buffered saline (PBS), and centrifuging at 200 x g for 10 minutes at room temperature. After aspirating the supernatant until 4 ml remained in the tube, the pellet was supplemented, mixed with fresh PBS by pipetting, and centrifuged for the second time. This process was repeated a total of 3 times to completely remove platelets. After the third centrifugation, the supernatant was aspirated, and the RBC pellet was resuspended in 20 ml of PBS in a final volume in a 50 ml tube to obtain a 10% red blood cell (RBC) solution (which is stable up to 1 week at 4 °C). Immediately before performing the hemagglutination assay, a new 2 - 5% red blood cell solution was prepared by diluting with PBS. A dilution series of the construct in the desired concentration range of 0.98 - 1000 nM in PBS was made. After seeding the red blood cells in a round - bottom 96 - well plate, the step - wise adjusted construct was added to the plate in the same volume and incubated overnight at 37 °C, 5% CO2. Photos were taken after overnight incubation.

[0215] As a result, compared with magrolimab, it was shown that the hemagglutination ability of the HLA - SIRPa bispecific construct was decreased (Figure 10, bottom). In addition to this, IOS - 2057, IOS - 2058, IOS - 2059, and IOS - 2060 showed a decrease in hemagglutination compared to IOS - 2041, IOS - 2044, IOS - 2045, IOS - 2046, and their respective SIRPa arms.

[0216] Cell reporter CD47 / SIRPα potency assay The SIRPα bispecific construct was tested for inhibition or blockade of the interaction between CD47 and SIRPα on cells expressing the receptor (PathHunter® Jurkat SIRPα Signaling Bioassay Eurofins DiscoveRx). Briefly, ligand cells presenting CD47 were incubated with SIRPα signaling cells and compounds in the range of 0.015 - 1184.6 nM at 37 °C for 24 hours. The detection reagent was added to the cell mixture and the chemiluminescence signal was measured. The background value was subtracted from all samples, and the sample without compound was taken as the maximum value. The inhibition rate was calculated as follows: [(maximum value - sample) / maximum value] × 100%. The IC50 value (nM) was generated by non-linearly fitting the data to a three-parameter response to the inhibitor using Prism (GraphPad). A representative bispecific fusion protein containing the optimized N-terminal SIRPα polypeptide potently inhibited the CD47 / SIRPα signaling axis in this assay (Figure 11, Table 5).

[0217]

Table 5

[0218] Phagocytosis of Macrophages Primary monocyte-derived macrophages were isolated and plated at 2.5x10 4Cells were plated in growth medium (containing 20 ng / ml M-CSF) at a cell / well concentration and incubated at 37°C for 5-7 days. To generate M1-like macrophages, IFN-g (200 ng / ml) was added to the wells, and the cells were incubated at 37°C for a further 48 hours. On the day of co-culture with cancer cells, LPS (40 ng / ml) was added over 1 hour. On the day of the experiment, macrophages were treated with the indicated compounds; cancer cells were labeled with CellTrace CFSE (ThermoFisher, C34554) and pHrodo iFL Red STP Ester (Invitrogen, P36010) according to the manufacturer's instructions and added to the wells in an E:T ratio of 1:2. Co-cultures were monitored using an S3 or SX5 IncuCyte live cell imaging system, and four non-adjacent 10x objective lens fields per well were imaged every two hours and analyzed using IncuCyte software v2020C / 2022a. Phagocytic events were quantified based on segmentation of red signals originating from pHrodo within red objects. Green objects were defined using green signals generated by cancer cells stained with CFSE. Background red signals originating from cancer cells, where present, were quantified by identifying red and green co-localized objects. The final relative phagocytic index was expressed as the integrated red intensity per well (red calibration unit × μm) normalized to the confluence per well. 2 This represents the ratio of wells to confluence. Each phagocytic reaction was performed using technically double or triple replication.

[0219] The effects of variant HLA sequences and peptide linker length on fusion protein function were investigated using in vitro phagocytosis assays. Figure 12 shows macrophage phagocytosis of myeloma cell line (RPMI-8226) with HLA / SIRPα bispecificity, demonstrating that a longer linker between Fc and SIRPα leads to a better phagocytic response, as observed between IOS-2041 and IOS-2045. Macrophage phagocytosis of myeloma cells was enhanced using IOS-1155, IOS-1161, and IOS-1164 compared to IOS-1131, which only has an IgG4 isotype or A46E / V97R substitution, indicating that optimization of the HLA polypeptide sequence can also enhance the function of the fusion protein (Figure 13).

[0220] ex vivo tumor biopsy Surgically resected samples from non-small cell lung cancer (NSCLC) patients were obtained from a tissue biobank. This material consisted of a heterogeneous mixture of tumor cell clusters and single cells, both of which are part of the tumor microenvironment (TME). Heterogeneous complexes of tumor clusters and immune components isolated from the resected material were seeded onto hydrogel (Crown Bioscience's proprietary technology) using 384-well plates. The test compound, control, and reference were added to the culture in n=8 repeats. The 3D cultures were maintained for 6-8 days. The supernatant was collected in each repeat and stored at -20°C. Quantitative image analysis of tumor volume reduction and immune cell count in the 3D cultures was performed using 3D High Content Imaging (HCI) (Crown Bioscience's proprietary technology).

[0221] Ex vivo tumor biopsy assays using five NSCLC donors showed that the reduction in total tumor-like size (%) was donor-specific in both the bispecific format and the HLA+SIRPa(K53R) combination, with a response observed in three out of five donors compared to the culture medium and / or IgG4 control.

[0222] [Table 6]

[0223] All scientific publications and patent documents cited herein are incorporated herein by reference.

[0224] JPEG2026514293000015.jpg193134JPEG2026514293000016.jpg190134JPEG2026514293000017.j pg179134JPEG2026514293000018.jpg180134JPEG2026514293000019.jpg166127JPEG20265142930 00020.jpg198133JPEG2026514293000021.jpg204134JPEG2026514293000022.jpg172134JPEG202 6514293000023.jpg178135JPEG2026514293000024.jpg202135JPEG2026514293000025.jpg122135

Claims

1. It is a fusion protein: a. β2m (beta-2 microglobulin) polypeptide; β2m linker peptide; b. HLA heavy chain polypeptides containing the HLAα3 domain; The Fc polypeptide is optionally connected to the HLA heavy chain polypeptide via a peptide linker, particularly a peptide linker having a length of 5 to 25 aa, and optionally connected to the following polypeptide inhibitors by a short peptide linker: c. Polypeptide inhibitors of the interaction between CD47 and signal regulatory protein alpha (SIRPα) The fusion protein, which includes the above.

2. The β2m polypeptide consists of a single copy of the β2m sequence, and the HLA heavy chain polypeptide consists of an HLA heavy chain composed of an α1 domain, an α2 domain, and an α3 domain. In particular, HLA heavy chain polypeptides are composed of α1, α2, and α3 domains, HLA-B ＊ A fusion protein according to claim 1, comprising a 57-stranded chain.

3. The fusion protein according to claim 2, wherein the β2m polypeptide and the HLA heavy chain polypeptide are collectively characterized by the sequence of SEQ ID NO: 017, or a sequence having at least (≧) 85% identity with SEQ ID NO: 017 and having at least 75% of the biological function of SEQ ID NO:

017.

4. The β2m polypeptide consists of a single copy of the β2m sequence, and the HLA heavy chain polypeptide consists of one, two, or three HLAα3 domains. In particular, HLA heavy chain polypeptides contain one, two, or three HLA-B ＊ A fusion protein according to claim 1, comprising a 57α3 domain.

5. The fusion protein according to claim 4, wherein the β2m polypeptide and the HLA heavy chain polypeptide together are characterized by the sequence of SEQ ID NO: 019, or a sequence that has at least (≧) 85% identity, particularly 90% or more, 93% or more, 96% or more, or 98% or more identity with SEQ ID NO: 019, and has at least 75% of the biological function of SEQ ID NO:

019.

6. β2m polypeptide is internal (β2m) ２ The HLA heavy chain polypeptide consists of two copies of a β2m sequence linked by a peptide linker, and the HLA heavy chain polypeptide consists of two HLAα3 domains. In particular, HLA heavy chain polypeptides contain two HLA-B ＊ A fusion protein according to claim 1, comprising a 57α3 domain.

7. The aforementioned interior (β2m) ２ Peptide linkers are glycine-serine linkers composed of 1 to 30 amino acids. Especially the interior (β2m) ２ The fusion protein according to claim 6, wherein the peptide linker is represented by formula (GGGGS)n (Sequence ID: 003), where n is 1, 2, 3, 4, or 5.

8. The fusion protein according to claim 6 or 7, wherein the β2m polypeptide and the HLA heavy chain polypeptide together are characterized by the sequence of SEQ ID NO: 021, or a sequence having at least (≧) 85% identity, particularly 90% or more, 93% or more, 96% or more, or 98% or more identity with SEQ ID NO: 021, and having at least 75% of the biological function of SEQ ID NO:

021.

9. The first and second α3 domains are (α3) ２ It is connected by a peptide linker, and the (α3) ２ The fusion protein according to any one of claims 4 to 8, wherein the peptide linker is a glycine-serine linker composed of 1 to 30 amino acids, and is represented as an n-mer of (GGGGS) (Sequence ID: 005), where n is 1, 2, 3, 4, or 5.

10. A β2m linker peptide having a length of 1 to 30 amino acids, particularly 10 to 20 amino acids, is used to link a β2m polypeptide to an HLA heavy chain polypeptide, as described in any one of claims 1 to 9.

11. The HLA heavy chain polypeptide is B ＊ 57 polypeptide, B ＊ 58 polypeptide, B ＊ 27 polypeptide, B ＊ 44 polypeptide, B ＊ 81 polypeptide, C ＊ 08 polypeptide, and C ＊ The fusion protein according to any one of claims 2 to 11, selected from the group consisting of 12 polypeptides.

12. HLA heavy chain polypeptides are HLA-A ＊ 11 polypeptides, A ＊ 24 polypeptides, A ＊ 25 polypeptides, A ＊ 45 polypeptides, B ＊ 78 polypeptides, B ＊ 81 polypeptide, C ＊ 06 Polypeptides and C ＊ A fusion protein according to any one of claims 1, 2, 4, 6, 7, 9, or 10, selected from the group consisting of 12 polypeptides.

13. HLA heavy chain polypeptides are B ＊ A fusion protein according to any one of claims 1 to 11, wherein the polypeptide is 57.

14. HLA heavy chain polypeptide is B of SEQ ID NO: 006 ＊ For an array of 57, - 7th place, 46th place, 97th place, 143rd place, 194th place, 197th place A fusion protein according to any one of claims 8 to 12, comprising one or more mutations selected from.

15. HLA heavy chain polypeptide is B of SEQ ID NO: 006 ＊ A fusion protein according to any one of claims 8 to 12, comprising the A46E and V97R mutations in sequence 57.

16. HLA heavy chain polypeptide is B of SEQ ID NO: 006 ＊ A fusion protein according to any one of claims 8 to 12, comprising the A46E, V97R, and I194T mutations in sequence 57.

17. HLA heavy chain polypeptide is B of SEQ ID NO: 006 ＊ A fusion protein according to any one of claims 8 to 12, comprising the A46E, V97R, and H197F mutations in sequence 57.

18. HLA heavy chain polypeptide is B of SEQ ID NO: 006 ＊ A fusion protein according to any one of claims 8 to 12, comprising the mutations A46E, V97R, I194T, and H197F in sequence 57.

19. The fusion protein according to any one of claims 1 to 4, 7, or 10 to 18, wherein the HLA heavy chain polypeptide comprises or consists of a polypeptide selected from SEQ ID NOs: 007, 061, 62, 63, and 64.

20. The fusion protein according to any one of claims 1 to 19, wherein the polypeptide inhibitor of the interaction between CD47 and SIRPα is a SIRPα extracellular domain polypeptide.

21. A polypeptide inhibitor of the interaction between CD47 and SIRPα is characterized by a sequence selected from the group of SEQ ID NOs: 011, 012, 013, 014, 015, 016, particularly SEQ ID NO: 014, according to any one of claims 1 to 20.

22. The fusion protein according to claim 15 or 16, comprising an Fc polypeptide and a SIRPα linker peptide having a length of 10 to 20 amino acids, particularly a SIRPα linker peptide having a length of 10 to 15 amino acids that connects the Fc polypeptide to a polypeptide inhibitor of the interaction between CD47 and SIRPα.

23. The fusion protein according to any one of claims 1 to 22, comprising, from the N-terminus to the C-terminus, a β2m polypeptide, a β2m linker peptide, an HLA heavy chain polypeptide, an IgG fragment crystallizable (Fc) region of an immunoglobulin heavy chain fragment polypeptide, a SIRPα linker peptide, and a polypeptide inhibitor of the interaction between CD47 and SIRPα.

24. The fusion protein according to any one of claims 1 to 23, comprising, from the N-terminus to the C-terminus, a β2m polypeptide, an HLA heavy chain polypeptide, an IgG fragment crystallizable (Fc) region of an immunoglobulin heavy chain fragment polypeptide, and a polypeptide inhibitor of the interaction between CD47 and SIRPα.

25. A dimer protein comprising a fusion protein as defined in any one of claims 1 to 24, wherein the fusion protein comprises the Fc region of human immunoglobulin G (IgG).

26. The dimeric protein according to claim 25, wherein the Fc region is characterized by sequence number 009 or 010, or a sequence that is at least 95% identical to sequence number 009 or 010.

27. The dimeric protein according to either claim 25 or 26, wherein both polypeptide chains are characterized by the same amino acid sequence.

28. The dimeric protein according to claim 25 or 26, wherein the two polypeptide chains are characterized by different amino acid sequences.

29. A nucleic acid encoding a fusion protein or dimer protein as defined in any one of claims 1 to 28.

30. A pharmaceutical composition comprising a fusion protein according to any one of claims 1 to 24, or a dimer protein according to any one of claims 25 to 28, and a pharmaceutically acceptable excipient.

31. A fusion protein according to any one of claims 1 to 24, or a dimer protein according to any one of claims 25 to 28, for use in pharmaceuticals.

32. A pharmaceutical composition according to claim 30, a fusion protein according to any one of claims 1 to 24, or a dimer protein according to any one of claims 25 to 28, for use in the treatment or prevention of cancer.

Images