BISPECIFIC ANTIBODIES FOR EPCAMM-CD3epsilon

Abstract

The present invention relates to a humanized EPCAM CD3 epsilon chain (CD3e) bispecific antibody. The invention provides EpCAM-CD3 antibodies of different structures, and a method for producing antibodies in cells based on mRNA of an in vitro transcription method. The EpCAM-CD3 human Fc bispecific antibody prepared by adopting an mRNA-lipid nanoparticle (LNP) technology shows high activity both in vitro and in vivo.

Description

TECHNICAL FIELD

[0001] The present invention relates to EPCAM-CD3 epsilon chain (CD3e) bispecific antibodies. The present invention also relates to a method of killing EPCAM positive tumor cells by T cells by administering to a patient an EPCAM-CD3e bispecific antibody. BACKGROUND

[0002] Immunotherapy is emerging as a promising approach to treat cancer. T cells or T lymphocytes, as the "armed forces" of the immune system, continuously search for foreign antigens and recognize abnormal cells (such as cancer cells or infected cells) from normal cells. The use of bispecific antibodies to combine T cells and tumor-associated antigens to direct cytotoxic T cells to cancer cells and kill them is the most common approach to designing bispecific antibodies. Bispecific antibodies can be infused into patients through different routes. The advantage of bispecific antibodies compared to chemotherapy or antibody therapy is that they can specifically target antigen-positive cancer cells and simultaneously activate T cells.

[0003] Redirecting the activity of T cells against tumor cells by bispecific antibodies, independent of their T cell receptor (TCR) specificity, is a powerful approach to treat cancer. The concept is based on the recognition of cell surface tumor antigens and simultaneous binding to the CD3 epsilon chain (CD3e or CD3) in the T cell receptor (TCR) complex on the surface of T cells. This can trigger T cell activation, including the release of cytotoxic molecules, cytokines and chemokines, and induction of T cell proliferation.

[0004] EPCAM

[0005] EPCAM (epithelial cell adhesion molecule) is encoded by the EPCAM gene. EPCAM is a cell surface glycoprotein of approximately 40 kDa that is highly expressed in epithelial cancers, while its expression level in normal epithelial tissues is low (1), (2), (3). EPCAM has multiple names, including TROP1, CD326, HEA125, EGP40, KSA and ESA (1). EPCAM regulates cell-cell adhesion and tissue plasticity, and controls Cell proliferation and differentiation (1), (4).

[0006] There is great potential for developing anti-cancer therapies and immunotherapies targeting EpCAM using monoclonal antibodies or bispecific antibodies.

[0007] Human EPCAM is a polypeptide of 314 amino acids, comprising an extracellular domain (N-terminal, amino acids 24-265), a single transmembrane domain (amino acids 266-288, underlined) and a short cytoplasmic domain (amino acids 289-314, C-terminal) (1). EPCAM sequence is available at Uniprot database (www.uniprot.org / uniprot / P16422). BRIEF DESCRIPTION OF DRAWINGS

[0008] Figure 1A EPCAM-CROSS-FAB-CD3 Knob in hole (KIH) Cross FAB structure antibody is shown. Figure 1B EPCAM scfv-CD3 ScFv (BITE) structure antibody is shown. Figure 1C EPCAM scFv-CD3 ScFv-Fc domain structure antibody is shown.

[0009] Figure 2 Construction scheme of DNA vector template for RNA (lower) in vitro transcription (upper) is shown. 5'UTR, 5' untranslated region; 3'UTR, 3' untranslated region; poly A tail for increased stability.

[0010] Figure 3 EPCAM-CD3 antibody in combination with T cells killing EPCAM positive cells is shown. RTCA assay was performed with Lovo cells (antibody is Figure 1A structure).

[0011] Figure 4 PBM005 induces T cells to secrete IFN-g in a dose dependent manner against EPCAM positive Lovo cells. PBM005 does not induce T cells to secrete IFN-g against EPCAM negative HL-60 cells is shown.

[0012] Figure 5A EpCAM ScFv-CD3 ScFv antibody (PBM0065) in combination with T cells secreting IFN-g in a dose dependent manner against EPCAM positive Lovo cells but not against HL-60 cells is shown (antibody is Figure 1B structure). Figure 5B EPCAM-CD3 antibody (antibody is Figure 1B structure) induces IFN-g secretion with an EC50 value of 5.3 ng / ml is shown.

[0013] Figure 6EpCAM ScFv-CD3 ScFv-hinge-Fc antibody (PBM0070) in combination with T cells secreted IFN-g in a dose dependent manner against EPCAM positive Lovo cells, but not against HL-60 cells (antibody is Figure 1C Structure).

[0014] Figure 7 shows the EC50 value of a representative IFN-g ELISA assay of 6.1 ng / ml (antibody is Figure 1C Structure).

[0015] Figures 8A-8C EPCAM-CD3 antibody in combination with T cells killed EPCAM positive Lovo cells, but not EPCAM negative Colo 741 or Colo 320 cells. RTCA assay was performed with T cells only, antibody only, or T cells in combination with different dilutions of antibody supernatant (antibody is Figure 1C Structure).

[0016] Figure 9 EpCAM-CD3 antibody in combination with T cells secreted IFN-g against EPCAM positive Lovo cells, but not against EPCAM negative Colo 741 and Colo 329 cells (antibody is Figure 1C Structure).

[0017] Figure 10 Intratumoral injection of EPCAM-CD3-hinge-Fc-LNP in combination with T cells significantly inhibited the growth of mouse xenograft tumors. OVCAR-5 cells were transplanted by subcutaneous injection, and RNA-LNP was administered (i) intratumorally (i.t.) at a dose of 20 pL (4 pg) per tumor on days 1, 8, and 15, or (ii) intracellularly (i.c.) at a dose of 20 pL (4 pg) on day 0, and intratumorally (i.t.) at a dose of 2-fold per tumor (40 pL) on day 16. i.t. indicates intratumoral administration; i.c + i.t. indicates intracellular and intratumoral administration.

[0018] Figures 11A-11BIntratumoral delivery of EpCAM-CD3 hFc mRNA-LNPs to OVCAR-5 tumors, in combination with a single intravenous injection of T cells, significantly inhibited the growth of OVCAR-5 xenograft tumors. A. Dosing regimen of intratumoral mRNA-LNP injection and a single intravenous injection of T cells in OVCAR-5 xenograft tumor model. B. EpCAM-CD3 hFc mRNA-LNPs in combination with a single T cell injection significantly inhibited OVCAR-5 tumor growth (antibody is Figure 1C STRUCTURE). DETAILED DESCRIPTION

[0019] DEFINITIONS

[0020] As used herein, "affinity" refers to the strength of the binding of a single molecule to its ligand. Affinity is typically measured and reported as the equilibrium dissociation constant (KD), which is used to assess and compare the strength of the interaction between two molecules.

[0021] As used herein, "bispecific antibody" is an artificial protein that can bind to two different types of antigens or different epitopes of the same antigen at the same time.

[0022] As used herein, "CD3 epsilon (CD3e)" is a polypeptide encoded by the CD3E gene located on human chromosome 11. The CD3 epsilon polypeptide, together with CD3 gamma, delta, and zeta, and T cell receptor alpha / beta and gamma / delta heterodimers, make up the T cell receptor-CD3 complex. This complex plays an important role in coupling antigen recognition to multiple intracellular signaling pathways. The CD3 epsilon polypeptide plays a critical role in T cell development. In this application, CD3 epsilon, CD3e, and CD3 are used interchangeably.

[0023] As used herein, "domain" refers to a region of a polypeptide that folds into a specific structure independently of other regions.

[0024] As used herein, "single chain variable fragment (scFv)" refers to a single chain polypeptide derived from an antibody that retains the ability to bind to an antigen. One example of an scFv includes an antibody polypeptide prepared by linking the Fv regions of immunoglobulin heavy (H) and light (L) chain fragments via a spacer sequence by recombinant DNA techniques. Various methods of preparing scFv are well known to those skilled in the art.

[0025] As used herein, "tumor antigen" refers to a biological molecule that is antigenic, whose expression leads to cancer.

[0026] The present invention relates to a bispecific antibody that can specifically bind to human EPCAM and human CD3e. The EPCAM-CD3e bispecific antibody targets the EPCAM tumor antigen, which is highly overexpressed in various cancers such as ovarian cancer, seminoma and colon cancer. The EPCAM-CD3 bispecific antibody of the present invention has high cytotoxic activity on EPCAM positive colon cancer cell lines, but no cytotoxic activity on EPCAM negative cell lines. The bispecific antibody can activate T cells and redirect T cells to EPCAM positive cancer cells.

[0027] The three bispecific antibody structures of the present invention are shown in Figures 1A-1C Figure 1A A "knob-in-hole" heterodimeric cross-Fab antibody is shown, which has one arm binding to the human CD3e chain expressed on T cells and two arms binding to the human EPCAM expressed on EPCAM positive cancer cells. Figure 1B A BITE antibody is shown, which has one arm binding to the human CD3e chain and one arm binding to the human EPCAM antigen. Figure 1C An EPCAM ScFv-CD3 ScFv antibody is shown, which has a human Fc to improve stability.

[0028] Figures 1A-1C The structure of the bispecific humanized EPCAM and CD3 antibody is shown. Figure 1A A bivalent EPCAM-cross-FAB-CD3 knob-in-hole cross-Fab format antibody is shown. Its knob-in-hole structure as well as the Fc P329G mutation and the leucine to alanine (L234A, L235A or LA-LA) mutation are indicated in Figure 1A The amino acid numbering in CH3 starts with human IgGl according to [6]. Figure 1B A DNA construct encoding the single polypeptide of the humanized EPCAM-CD3 BITE antibody is shown. Figure 1C A DNA construct encoding the single polypeptide of the humanized EPCAM-CD3-hFc antibody is shown. Figure 1A The antibody shown has two EPCAM binding groups and one CD3 binding group. Figure 1B and 1C The antibodies shown each have one EPCAM binding group and one CD3 binding group and are either in BITE format (B) or dimer with human Fc (C), respectively.

[0029] Figure 1A The bispecific antibody structure of

[0030] The present invention relates to a bispecific antibody having Figure 1A ​The bispecific antigen-binding molecule of the structure shown (EpCAM-CD3 cross Mab socket-socket structure). In one aspect, the EPCAM antibody is a humanized antibody, the bispecific antibody comprises: (a) first and second two antigen-binding groups, each group is a humanized Fab molecule capable of specifically binding to human EPCAM, and each group comprises a heavy chain variable region (EPCAM VH), such as having the amino acid sequence of SEQ ID NO: 10, and a light chain variable region (EPCAM VL), such as having the amino acid sequence of SEQ ID NO: 4; (b) a third antigen-binding group, which is a Fab molecule capable of specifically binding to human CD3 epsilon, the third antigen-binding group comprises a heavy chain variable region (CD3 VH), such as having the amino acid sequence of SEQ ID NO: 11, and a light chain variable region (CD3 VL), such as having the amino acid sequence of SEQ ID NO: 7, and in addition, the third antigen-binding group is a cross Fab molecule, in which the constant regions of the Fab light chain and the Fab heavy chain are exchanged; and (c) a human IgG Fc domain, which comprises a first subunit and a second subunit capable of stable binding; wherein the Fab heavy chain of the third antigen-binding group is (i) fused at the N-terminus to the C-terminus of the Fab heavy chain of the first antigen-binding group (CH1), and (ii) its C-terminus is fused to the N-terminus of the first subunit of the Fc socket domain, and in addition, the C-terminus of the Fab heavy chain of the second antigen-binding group (CH1) is fused to the N-terminus of the second subunit of the Fc socket domain.

[0031] The bispecific antibody of the present application involves the use of a cross FAB technology, which reduces the undesirable pairing errors by cross-linking the constant and variable regions and exchanging the CH1 and CL regions in the CD3e Fab molecule.

[0032] In one embodiment, the bispecific antibody of the present application comprises: (1) a humanized EPCAM light chain, (2) a CD3e cross FAB, CD3 VL-CH1; (3) a humanized EPCAM VH-CH1-CD3e cross FAB (VH-CL)-Fc (socket), and (4) a humanized EPCAM VH-CH1-Fc (socket). Figure 1A )

[0033] As shown in Figure 1A , the human IgG Fc comprises a hinge region, CH2 and CH3.

[0034] In one embodiment, the VH of the humanized EPCAM antibody has the amino acid sequence of SEQ ID NO: 10, and the VL has the amino acid sequence of SEQ ID NO: 4.

[0035] In one embodiment, the Fc domain comprises a modification that promotes binding of a first subunit to a second subunit. In one embodiment, in the CH3 region of the first subunit of the Fc domain, one amino acid residue is replaced with an amino acid residue having a larger side chain volume, thereby forming a protuberance in the CH3 region of the first subunit that can fit into a cavity in the CH3 region of the second subunit. At the same time, in the CH3 region of the second subunit of the Fc domain, one amino acid residue is replaced with an amino acid residue having a smaller side chain volume, thereby forming a cavity in the CH3 region of the second subunit that can accommodate the protuberance in the CH3 region of the first subunit.

[0036] In one embodiment, the Fc domain has reduced binding affinity to an Fc receptor and / or reduced effector function compared to a native IgG Fc domain.

[0037] In one embodiment, the Fc domain comprises one or more amino acid substitutions that reduce binding to an Fc receptor and / or effector function. In one embodiment, the one or more amino acid substitutions in the Fc domain are selected from the group of L234, L235, and P329 (Kabat numbering). In one embodiment, the amino acid substitutions are L234A, L235A, and P329G.

[0038] In one embodiment, the Fc region P329G, and L234A and L235A mutations are used to prevent Fc-dependent immune responses.

[0039] In one embodiment, only the L234A and L235A mutations are included to prevent Fc-dependent immune responses.

[0040] In one specific embodiment, the Fc domain is modified by so-called "knobs-into-holes" modifications, which involve a "knob" modification in one of the two subunits of the Fc domain and a "hole" modification in the other subunit of the Fc domain. The "knobs-into-holes" technology is described, for example, in U.S. Patent No. 5,731,168. Generally, this approach involves introducing a protuberance ("knob") at the interface of a first polypeptide and a corresponding recess ("hole") at the interface of a second polypeptide, such that the protuberance can be positioned in the recess to promote heterodimer formation and inhibit homodimer formation. The protuberance is constructed by replacing a small amino acid side chain at the interface of the first polypeptide with a larger side chain (such as tyrosine or tryptophan). At the interface of the second polypeptide, a compensatory recess of the same or similar size as the protuberance is created by replacing a large amino acid side chain with a smaller side chain (such as alanine or threonine).

[0041] In one embodiment, the "knob" is formed by mutating an Fc at S354C and T366W, and the corresponding "hole" is formed by mutating a corresponding Fc at Y349C, T366S, L368A, and Y407V.

[0042] In one embodiment, the bispecific antibody molecule comprises two binding groups that bind to EPCAM and one binding group that binds to CD3e, the molecule comprises the amino acid sequences of SEQ ID NOs: 5, 8, 12, and 14 in a molar ratio of 2:1:1:1.

[0043] Figure 1B Bispecific antibody structure of the present invention

[0044] The present invention also relates to a bispecific antigen binding molecule (EpCAM ScFv-CD3 ScFv (BITE)) having the structure shown below. Figure 1B

[0045] Figure 1B The structure of a recombinant humanized bispecific EPCAM ScFv-CD3e ScFv antibody consisting of one DNA construct is shown. The BITE structure comprises one binding group that binds to EPCAM and one binding group that binds to CD3e and carries a 6xHis tag at the end for purification.

[0046] The present invention relates to a bispecific antigen binding molecule comprising an EPCAM VH, for example having the amino acid sequence of SEQ ID NO: 10; a first linker; an EpCAM VL, for example having the amino acid sequence of SEQ ID NO: 4; a second linker; a CD3 VH, for example having the amino acid sequence of SEQ ID NO: 11; a third linker; a CD3 VL, for example having the amino acid sequence of SEQ ID NO: 7; wherein the first, second, and third linkers can be the same or different.

[0047] Figure 1C Bispecific antibody structure of the present invention

[0048] The present invention also relates to a bispecific antigen binding molecule (EpCAM ScFv-CD3 ScFv- human Fc domain) having the structure shown below. Figure 1C

[0049] Figure 1C ​​A bispecific antibody structure is shown, which is composed of a monovalent humanized EPCAM ScFv and a monovalent CD3e ScFv fused to a human Fc domain; the structure is composed of one DNA construct. The structure comprises an EPCAM scFv and a CD3e scFv connected by a linker, followed by a Fc domain (hinge linker-CH2-CH3) fused to the CD3 ScFv for stability. The bispecific antibody (BITE-human Fc format) comprises a binding group that binds to EPCAM, and a binding group that binds to CD3e. The antibody can have a dimeric conformation.

[0050] The present application provides a bispecific antigen binding molecule comprising a humanized EPCAM VH, a first linker, a humanized EPCAM VL, a second linker, a CD3 VH, a third linker, a CD3 VL, and a Fc domain.

[0051] In one embodiment, the EPCAM VH has the amino acid sequence of SEQ ID NO: 10, the EPCAM VL has the amino acid sequence of SEQ ID NO: 4, the CD3 VH has the amino acid sequence of SEQ ID NO: 11, and the CD3 VL has the amino acid sequence of SEQ ID NO: 7.

[0052] In one embodiment, the Fc domain comprises a hinge linker and the CH2-CH3 region of human IgGl, with one or two amino acid substitutions.

[0053] In one embodiment, the Fc domain comprises one or more amino acid substitutions selected from the group consisting of L234A, L235A, and P329G (EU numbering).

[0054] In one embodiment, the first, second, and third linkers have the amino acid sequence (GGGGS)n, where n = 1-5. In a preferred embodiment, the linker sequence has the amino acid sequence (GGGGS)3.

[0055] In one embodiment, the bispecific antigen binding molecule has the amino acid sequence of SEQ ID NO: 19.

[0056] Methods of making bispecific antibodies

[0057] Nucleic acids encoding antibodies or antigen-binding fragments thereof can be inserted into a vector and expressed in mammalian 293S or CHO cells using serum-free media. Antibodies containing human-derived Fc fragments can be purified by protein A or protein G columns and used for related research. Antibodies with His tags can be purified by one-step affinity chromatography, i.e., immobilized metal ion affinity chromatography (IMAC), which is commercialized in different forms, with Ni-NTA matrix being the most commonly used.

[0058] The present application provides an isolated DNA sequence comprising: (a) a promoter coding sequence, (b) a 5'-UTR (untranslated region) sequence, (c) a sequence encoding an antibody or antigen-binding fragment thereof, (d) a 3'-UTR sequence, and (e) a Poly A tail sequence. In one embodiment, the antibody or antigen-binding fragment thereof is a bispecific antigen-binding molecule of the present application.

[0059] In the DNA sequence, the promoter is a T7 promoter, a T7AG promoter, or a SP6 promoter. The Poly A tail sequence is 20 to 170 nucleotides in length. The Poly A tail sequence can optionally comprise one or more linker sequences between the poly A segments. If the Poly A tail is longer than 60 nucleotides, it typically comprises a linker sequence comprising non-adenine nucleotides. The linker sequence is 5-30 or 5-25 nucleotides in length, for example, 10 nucleotides or 20 nucleotides. In one example, the Poly A tail is 110 nucleotides in length, consisting of a contiguous sequence of 30 adenine residues, followed by a 10-nucleotide linker sequence, followed by another 70 adenine residues. In another example, the Poly A tail is 90 nucleotides in length, consisting of a contiguous sequence of 40 adenine residues, followed by a 30-nucleotide linker sequence, and followed by another 30 adenine residues. In another example, the Poly A tail is 150-160 nucleotides in length, comprising two linker sequences.

[0060] DNA expression is finely regulated at the post-transcriptional level. Untranslated regions (UTRs) are not translated into amino acids, however, the UTRs of mRNA can play a role in translation, degradation, and localization by controlling their stem-loop structures, upstream start codons and open reading frames, internal ribosome entry sites, and various cis-acting elements bound by RNA-binding proteins. UTRs play an important role in the post-transcriptional regulation of DNA expression, including regulating mRNA export from the nucleus, translation efficiency, subcellular localization, and stability.

[0061] The 5'-UTR typically contains 10-1000 nucleotides, or 20-500 nucleotides, or 30-200 nucleotides, or 30-100 nucleotides. For example, the 5'-UTR is 50 nucleotides. The 3'-UTR typically contains 10-3000 nucleotides, for example 50-500 nucleotides, or 100-300 nucleotides. Preferred 5'-UTRs and 3'-UTRs are the UTRs of beta-globin, or the UTRs of the Pfizer COVID-19 vaccine.

[0062] The beta-globin gene is shown below:

[0063] https: / / www.ncbi.nlm.nih.gov / nucleotide / V00497.1?report=genbank&log$=nuc lalign&blast_rank=5&RID=TDDZ1K98016

[0064] In one embodiment, the 5' untranslated region is derived from human alpha-globin RNA and has an optimized Kozak sequence. The 3' untranslated region contains two sequence elements derived from the amino-terminal split enhancer (AES) mRNA and the mitochondrially encoded 12S ribosomal RNA to confer RNA stability and high total protein expression levels.

[0065] Any suitable vector, such as the pSP64 Poly(A) vector (Promega) or the pGEM3Z vector (Promega), can be used as a cloning vector for the above DNA sequences.

[0066] For example, to construct pEM3Z-beta-globin UTR-UTR-poly A tail, the 3'-UTR of the beta-globin molecule can be amplified from human bone marrow with restriction enzyme sites flanking the sequence. For example, a single fragment (pEM3Z-l beta-globin-UTR-A

[120] ) or two consecutive fragments (pEM3Z-2 beta-globin-UTR-A

[120] ) can be inserted before the poly(A) tail.

[0067] The present application provides a method of producing an antibody or antigen-binding fragment thereof in a cell and in an animal. The method comprises the steps of obtaining a DNA sequence as described above, transcribing the DNA sequence into mRNA in vitro using an RNA polymerase, introducing the mRNA into a cell or mouse xenograft tumor by electroporation or transfection, and translating in the cell to produce the antibody or antigen-binding fragment thereof. In one embodiment, the antibody or antigen-binding fragment thereof is a bispecific antigen-binding molecule of the present application.

[0068] Figure 2A schematic of a DNA vector template (top) for in vitro transcription of RNA (bottom) is shown. 5'UTR, 5' untranslated region; 3'UTR, 3' untranslated region; poly A tail for increased stability.

[0069] The present application also relates to a pharmaceutical composition comprising the bispecific antigen binding molecule and a pharmaceutically acceptable carrier.

[0070] The inventors have demonstrated that in vitro transcription of DNA templates can produce EpCAM-CD3 antibodies when the obtained RNA is transfected into 293 cells.

[0071] Antibodies can also be produced by adenoviral or other viral transfection of cells, enabling in vivo manufacturing, reducing production costs, and generating stable cell lines for antibody production.

[0072] In one embodiment, antibody DNA is inserted into a DNA template vector containing a T7 or SP6 promoter to facilitate transcription of the DNA into antibody RNA by an RNA polymerase through in vitro transcription. The RNA is mixed with a lipid component to make lipid nanoparticles (LNPs) in which the RNA is encapsulated. Subsequently, the LNP-encapsulated mRNA is introduced into T cells, primary NK cells, NK-92 cells, dendritic cells (DCs), macrophages, or other types of immune cells or mammalian cells by electroporation or transfection, and the mRNA is translated in the cells to produce antibody proteins that can be used to kill tumor cells.

[0073] Generally, mRNA has a short and degradable nature after delivery into the body. The present application provides a method of producing an antibody or antigen binding molecule by encapsulating mRNA of the antibody or antigen binding molecule in a lipid nanoparticle and delivering it into a cell. By encapsulating the mRNA in a lipid nanoparticle, the stability of the mRNA can be improved. The method comprises the following steps: (i) obtaining a DNA sequence comprising: (a) a promoter sequence, (b) a 5'-UTR (untranslated region) sequence, (c) a sequence encoding an antigen binding molecule, (d) a 3'-UTR coding sequence, and (e) a poly-A tail sequence; (ii) transcribing the DNA sequence into mRNA in vitro using an RNA polymerase, (iii) encapsulating the mRNA in a lipid nanoparticle (LNP), (iv) transfecting the mRNA-encapsulated LNP into a cell, and (v) translating the mRNA in the cell to produce the antigen binding molecule.

[0074] In one embodiment, the lipid nanoparticle comprises 8-[(2-hydroxyethyl)[6-oxo-6- (undecyloxy)hexyl]amino]octanoic acid, 1-octyl nonyl ester (SM-102), distearoylphosphatidylcholine (DSPC), cholesterol, and 1,2-dimyristoyl-rac-glycero-3- methoxypolyethyleneglycol-2000 (DMG-PEG2000).[LNP-102(ii)]

[0075] In one embodiment, the lipid nanoparticle comprises 8-[(2-hydroxyethyl)[6-oxo-6- (undecyloxy)hexyl]amino]octanoic acid, 1-octyl nonyl ester (SM-102), distearoylphosphatidylcholine (DSPC), cholesterol, and 1,2-dimyristoyl-rac-glycero-3- methoxypolyethyleneglycol-2000 (DMG-PEG2000).[LNP-102(ii)]

[0076] In one embodiment, the lipid nanoparticle comprises 2-hexyldecanoic acid, 1,1'-[[(4- hydroxybutyl)imino]di-6,1-hexanediyl] ester (ALC-0315), DSPC, cholesterol, and a-[2- (tetradecylamino)-2-oxoethyl]-w-methoxy-poly(oxy-1,2-ethanediyl) (ALC-0159).[LNP-315]

[0077] Encapsulation of mRNA into LNP nanoparticles protects mRNA from degradation and improves its stability; mRNA is subsequently released from the LNPs into cells in the body to produce proteins. Methods for the preparation of mRNA-lipid nanoparticles are described in detail in Schoenmaker (International J. Pharmaceutics, 601: 120856, 2021), which is hereby incorporated by reference in its entirety, particularly the section on LNPs.

[0078] Methods of treating cancer

[0079] The present invention relates to a method of treating cancer, the method comprising injecting a bispecific EPCAM-CD3e antigen binding molecule or antibody into a subject having cancer by intratumoral injection, wherein the cancer is selected from the group of cancers consisting of colon cancer, lung cancer, pancreatic cancer, gastric cancer, testicular cancer, teratocarcinoma, seminoma, ovarian cancer, and cervical cancer, and the cancer is EPCAM positive.

[0080] In one embodiment, the method comprises the steps of: obtaining a mRNA sequence comprising a coding sequence encoding a bispecific EPCAM-CD3e antigen binding molecule; mixing the mRNA with a lipid nanoparticle to form a mRNA-lipid nanoparticle complex; and injecting the mRNA-lipid nanoparticle complex into a cancer cell.

[0081] In one embodiment, the sequence of the bispecific EPCAM-CD3e antigen binding molecule is as described in the present application.

[0082] In one embodiment, the method further comprises intravenous injection of T cells.

[0083] The present application demonstrates the effectiveness of bispecific antibodies against the EPCAM antigen that is overexpressed in colon cancer tumors. The present application demonstrates that EPCAM-CD3e antibodies can bind to CD3e antigen and EPCAM antigen. When the antibody is co-administered with T cells, it can specifically reduce the survival rate of EPCAM-positive colon cancer cells, but does not affect EPCAM-negative cancer cells. When EPCAM-CD3e antibodies are combined with T cells, they can induce a significant level of interferon-gamma (IFN-gamma) secretion when co-cultured with EPCAM-positive colon cancer cells, but not when co-cultured with EPCAM-negative target cancer cells. The present application demonstrates that the use of EPCAM-CD3e antibodies in combination with T cells can significantly inhibit the growth of Lovo (EPCAM-positive colon cancer cells) tumors in vitro.

[0084] In one embodiment, the RNA is encapsulated into lipid nanoparticles (LNPs) and transfected into normal cells, cancer cells, and tumors in vivo. Nanoparticle drug delivery systems exhibit numerous advantages, including high bioavailability, solubility, stability, ability to cross the blood-brain barrier, and low toxicity and minimal side effects.

[0085] The present inventors demonstrate that the use of EPCAM-CD3 antibodies in combination with T cells can significantly kill all EPCAM-positive cancer cells, but not EPCAM-negative colon cancer cells. This indicates that EPCAM-CD3 antibodies have high specificity.

[0086] The present inventors demonstrate the high effectiveness of the bispecific antibodies of the three different designs shown in Figures 1A-1C

[0087] The present inventors demonstrate that the transfection of EPCAM-CD3-hFc RNA-LNP into cancer cell lines can produce the bispecific antibodies.

[0088] ​Furthermore, the RNA template containing the EpCAM bispecific antibody can comprise a human Fc fragment, wherein silent mutations (such as L234A, L235A, P329G) can be introduced to reduce Fc-dependent immune responses (ADCC).

[0089] The RNA template containing the EpCAM bispecific antibody can also be prepared as a version without the Fc fragment to achieve a shorter sequence and shorter antibody half-life.

[0090] The inventors have shown that co-injection of EPCAM-CD3-hFc antibody RNA-LNP with T cells into OVCAR-5 tumors significantly inhibited the growth of xenograft tumors.

[0091] This application shows that the antibodies produced in vivo by injection of RNA-LNP can provide a high degree of anti-tumor effect in vivo.

[0092] In the following example, three different designs of humanized EpCAM-CD3 antibodies were manufactured and tested for targeting anti-colorectal tumors. These antibodies showed high efficiency and specificity. Furthermore, this study demonstrates a method of delivering bispecific antibodies using mRNA-lipid nanoparticle (LNP) technology. By delivering EpCAM-CD3 humanized Fc (hFc) mRNA-LNPs to mouse tumors and intravenously injecting T cells, the in vivo growth of OVCAR-5 xenograft tumors was significantly blocked.

[0093] There are several advantages of producing bispecific antibodies in vivo by mRNA-LNP injection compared to RNA or antibody protein. mRNA-LNP complexes are more stable than regular mRNA, the cost of producing antibodies by mRNA-LNP is lower than antibody protein manufacturing, and there is potential for further optimization in future studies. By intravenously injecting T cells in combination with intratumoral delivery of EpCAM-CD3 hFc mRNA-LNPs, the growth of xenograft tumors can be blocked. This method can be used to target solid tumors.

[0094] This disclosure describes a safe method of delivering bispecific antibodies using mRNA-LNP technology for local intratumoral administration. This invention combines cell therapy, using T cells to kill tumors. This method can similarly utilize specific immune cell receptors to recruit natural killer cells (NK cells) or gd T cells to the tumor. Furthermore, different types of immune cell stimulators can be used to improve the efficacy of this therapy, such as checkpoint inhibitors, chemokines, cytokines, growth factors, and tumor microenvironment modulators. The lysed tumor releases tumor neoantigens (natural vaccines), which can be recognized by antigen-presenting cells or dendritic cells in the presence of immunomodulators, and promote the activity of memory T cells and B cells.

[0095] The following examples further illustrate the application. These examples are intended merely to be illustrative of the application and are not to be construed as limiting.

[0096] Example

[0097] Example 1. Materials and methods

[0098] Cells and media

[0099] HEK293FT cells were obtained from Alstem (Richmond, CA) and cultured in Dulbecco's Modified Eagle's Medium (DMEM) with 10% fetal bovine serum (FBS) and 1% penicillin / streptomycin. Human peripheral blood mononuclear cells (PBMCs) were isolated from whole blood obtained from the Stanford Hospital Blood Center (Stanford, CA) using Ficoll-Paque solution (GE Healthcare) following an Institutional Review Board (IRB) approved protocol. Colon cancer cell lines: EPCAM negative: Colo-741 and Colo-320, and EPCAM positive: Lovo cells were used for this study. Cells were cultured in a humidified environment with 5% carbon dioxide (CO2) (5).

[0100] Antibodies

[0101] Allophycocyanin-(APC) labeled anti-human CD326 (EpCAM) antibody was obtained from Biolegend. PE conjugated anti-His tag antibody was obtained from Biolegend. APC labeled anti-human CD3 antibody was obtained from Biolegend. Anti-human IgG was obtained from Jackson ImmunoResearch. PE and 7-AAD cell viability stain was obtained from Biolegend.

[0102] EPCAM-CD3 antibody constructs

[0103] The four constructs in Example 2 were designed according to the Cross-Fab design method described in the literature (6). These constructs contain the P329G mutation and substitution of leucines 234, 235 to alanines, referred to as LA-LA, to reduce Fc immune activity. In addition, Fc silent mutations and knob-in-hole mutations were also engineered as described in the literature (6). We also expressed two constructs shown in Figure 1B and Figure 1C . Figure 1A All constructs in Example 2 were cloned into the pYD11 vector between the Nhe I and Nsi I sites.

[0104] Design and cloning of bispecific antibody DNA constructs

[0105] AUA1 antibodies were produced by conventional methods, i.e. by immunizing BALB / c mice with the colon adenocarcinoma cell line Lovo. Mouse AUA1 EPCAM antibodies were characterized. Humanization of mouse AUA1 antibodies (7) was performed according to the methods described in (8, 9). Humanized antibodies were first tested for their functional activity in the form of chimeric antigen receptors (CAR) according to the methods described in (9). Humanization of the variable heavy (VH) and light chain fragments (VL) of the AUA1 antibody was performed according to the CrossMab knob-in-hole engineering design described in (10) to construct bivalent antibody constructs. The human Fc (IgGl) contains P329G and L234A, L235A (LALA) mutations to silence the Fc region to prevent binding to gamma receptors and activation of innate immune cells.

[0106] Another design was in the form of EpCAM ScFv-CD3 ScFv-His tag (BITE).

[0107] A third design was in the form of EpCAM ScFv-CD3 ScFv-human Fc construct.

[0108] All constructs were cloned into the pYD11 vector and confirmed by sequencing. For RNA expression based DNA template design, EpCAM ScFv-CD3 ScFv-human Fc sequences were cloned into a DNA vector containing a T7AG promoter, 50U TR, 30U TR and a 150 poly A tail as described in (11). DNA template sequences for in vitro transcription were verified by sequencing.

[0109] Expression of EPCAM-CD3 antibodies

[0110] For constructs of Figure 1A , four antibody constructs were mixed with NanoFect transfection reagent in a mass ratio of 2 (EPCAM VL-CL): 1:1:1 (pg / mL) for transfection of 293S cells. For constructs 1B and 1C, antibody constructs were mixed with NanoFect transfection reagent in a mass ratio for transfection of 293S cells. Cells were cultured in Freestyle F17 media containing 8 mM L-glutamine (or GlutaMAX) and 0.1% Pluoronic F-68 in a flask for one week at 37 °C in a shaker incubator. Supernatants or Fc fragment containing antibodies purified by protein A column ( Figure 1A , Figure 1C ) or His tag containing antibodies purified by Ni-NTA column ( Figure 1B ) were analyzed on SDS gels and tested by flow cytometry and functional assays.

[0111] Bispecific antibody DNA transfection of HEK-293 cells

[0112] HEK293S cells were transfected with DNA constructs using bispecific antibodies and ALSTEM NanoFect transfection reagent. Cells were grown in suspension flasks in Freestyle F17 media containing 8 mM glutamine and 0.1% Pluronic F68 surfactant, and incubated at 37°C and 5% CO2. Antibody-containing supernatants were collected 3-7 days post-transfection. Human Fc fragment-containing antibody supernatants were purified using protein A or protein G columns, and EpCAM-CD3-His tagged antibody supernatants were purified using Ni-NTA columns. Purified antibodies were tested for molecular weight by SDS gel, and subsequently used for functional analysis.

[0113] PBMC

[0114] PBMC were activated using CD3 / CD28 Dynabeads (Invitrogen) at a concentration of 1 x 10 6 Cells were suspended in AIM V-AlbuMAX media (Thermo Fisher) containing 10% fetal bovine serum (FBS) and 300 U / mL interleukin-2 (IL-2) (Thermo Fisher) at a concentration of 1 x 10

[0115] Flow cytometry (FACS) analysis

[0116] FACS analysis was performed using allophycocyanin (APC)-labeled anti-CD3 antibody (eBioscience, San Diego, CA) and a FACSCalibur flow cytometer (BD Biosciences). For experiments using colon cancer cell lines for FACS to detect EPCAM levels, bispecific EPCAM-CD3 antibodies, or AUA1 murine monoclonal antibody from the University of Oxford, or Biolegend murine monoclonal antibody (Biolegend Ant-hu-EpCAM-APC, Cat No: 324207) were used, and analysis was performed on a FACSCalibur, as described in the literature (5).

[0117] Real-time cytotoxicity (RTCA) analysis

[0118] Adherent colon cancer target cells (10,000 cells per well) were seeded onto 96-well E-plates (Acea Biosciences, San Diego, CA) and incubated overnight for analysis using the impedance-based real-time cell analysis (RTCA) iCELLigence system (Acea Biosciences). After 20-24 hours, the medium was replaced with AIM V-AlbuMAX medium containing 10% fetal bovine serum (FBS) containing 1 x 10 5 Effector cells T cells, or bispecific antibody-containing T cells, or antibody only, with three replicates per group. Cells were then monitored for over 40 hours using the RTCA system, and impedance (proportional to cell index) was plotted against time. Cytotoxicity was calculated as follows: (impedance value of target cells without effector cells - impedance value of target cells with effector cells) x 100 / impedance value of target cells without effector cells.

[0119] ELISA detection of IFN-g cytokine secretion

[0120] Target cells were incubated with effector cells or reagents in U-bottom 96-well plates using AIM V-AlbuMAX medium containing 10% FBS, with three replicates per group. After 16 hours, supernatants were removed and centrifuged to remove residual cells. In some experiments, supernatants after RTCA detection were subjected to IFN-g cytokine detection using ELISA. Supernatants were transferred to new 96-well plates and subjected to ELISA detection using a kit from Thermo Fisher according to the manufacturer’s protocol to analyze human cytokines. EC50 values were calculated using GraphPad Prism software.

[0121] Example 2. Sequences of humanized EPCAM-CD3E bispecific antibody (Fig. 1A)

[0122] Figure 1A The structure of a humanized EPCAM-CD3 bivalent antibody is shown, which is composed of 4 DNA constructs. The structure contains a CD3 CROSS-Fab, designated PBM005.

[0123] Figure 1A The EPCAM-CD3e bispecific antibody shown is composed of 4 constructs:

[0124] 1. Humanized EPCAM light chain (VL-CL): EPCAM VL (PCT / US20 / 14999) and codon-optimized (shown below)

[0125] 2. CD3 CROSSFAB (VL-CH1)

[0126] 3. Humanized EPCAM VH-CH1-CD3 CROSSFAB (VH-CL)-Fc (knob)

[0127] P329G LA-LA (EPCAM VH from PCT / US20 / 14999 and codon optimized as follows)

[0128] 4. Humanized EPCAM VH-CH1-Fc (hole) P329G LA-LA (EPCAM VH from PCT / US20 / 14999 and codon optimized as follows)

[0129] The P329G mutation eliminates the interaction between Fcy receptors and Clq, preventing the process of target cell clearance by antibody-dependent cell-mediated cytotoxicity (ADCC), antibody-dependent cellular phagocytosis (ADCP), or complement-dependent cytotoxicity (CDC). The P329G mutation eliminates Fcy-mediated immune receptor function by delivering a silent Fc region (7) to the cell when it is delivered. The other two mutations, LA-LA mutations replace Leu 234 and Leu 235 with alanine (A), which completely blocks the interaction of FcyR with Clq, eliminating Fc-mediated ADC, ADCC, and other immunogenicity (6).

[0130] All sequences were codon optimized, synthesized as GBlocks, and inserted into the Nhe I and Nsi I sites in the pYD11 vector. To avoid light chain domain pairing errors, the CrossFAB technology was used, in which the CD3 VH was linked to the CL, and the CD3 VL was linked to the CH1. This patent also used the “knob-in-hole” mutation strategy proposed by Crick in 1952, which used T366W and S354C mutations to create a knob and (Y349C, T366S, L368A, and Y407V) to create a hole, which can fix the two Fc chains together. All sequences start with a signal peptide (underlined):

[0131] Nucleotide sequence of the signal peptide

[0132]

[0133] Amino acid sequence of the signal peptide

[0134] METDTLLLWVLLLWVPGSTG AAS (SEQ ID NO: 2)

[0135] Construct #1. EPCAM light chain: LC-EPCAM

[0136] The DNA artificial sequence for humanized EPCAM ((EPCAM VL (bold) - CL (italic) of LC (light chain) is shown below. The EPCAM VL nucleotide sequence shown in WO2020 / 154627 was codon optimized and inserted with the constant region CL into the Nhe I site (GCTAGC site is annotated in italic and underlined) and Nsi I site (atgcat site is annotated in italic and underlined, pYD11 vector). The sequence starts with a signal peptide (signal peptide is annotated with underlining and AAS amino acids are added to the cloning site): METDTLLLWVLLLWVPGSTG AAS (SEQ ID NO: 2).

[0137] A stop codon was added in front of the human Fc sequence to express a light chain without Fc in the vector. The signal peptide is annotated in bold italic and underlined; the VL is annotated in bold; the CL is annotated in italic.

[0138]

[0139]

[0140] Amino acid sequence

[0141] Signal peptide (annotated with underlining) + AAS: METDTLLLWVLLLWVPGSTG AAS (SEQ ID NO: 2)

[0142] Construct #2. CD3 CROSSFAB (VL-CH1 )

[0143] CD3 VL is shown in bold, CH1 is shown in italic, the nucleotide sequence is codon optimized. Nhe I and Nsi I sites are shown in italic. A stop codon TAA was added in front of the Fc to terminate the sequence.

[0144] Nucleotide sequence: signal peptide is shown in underlined and italic bold, then AAS is shown in italic regular font; VL is shown in bold, CH1 is shown in regular font italic.

[0145]

[0146] Amino acid sequence (without signal peptide)

[0147]

[0148] Construct #3.

[0149]

[0150] Signal peptide is in bold, italic and underlined, followed by 3 amino acids - AAS (according to cloning site); cloning site Nhe I and Nsi I Underlined and larger font

[0151] Bold; CH1 underlined; 2xG4S linker; Bold italic; CL italic underlined; IgG Fc sequence contains LA-LA mutations (L234 and L235 changed to A) which are shown in bold underlined, and P329G mutation (P changed to G) which is bold underlined.

[0152] Knob mutation sites in Fc domain are S354C and T366W, shown in bold, larger font and italic.

[0153] AS amino acids can be included after EPCAM VH, before start of CH1.

[0154] Nucleotide sequence:

[0155]

[0156]

[0157] Amino acid sequence of EPCAM VH, SEQ ID NO: 10

[0158]

[0159] Amino acid sequence of Construct #3 (does not include signal peptide):

[0160]

[0161]

[0162] Construct #4 EPCAM VH-CH1-Fc (Oval) P329GLA-LA

[0163] Construct #4 uses the same P329G and LALA mutations as Construct #3, which are shown in bold. The socket mutations are Y349C, T366S, L368A and Y407V, which are shown in bold, larger font and italic. Cloning site Nhe I GCTAGC and Nsi I ATGCAT Underlined

[0164] Signal peptide is underlined, bold and italic, followed by 9 nucleotides AAS (cloning site) encoding 3 amino acids, regular font, italic; EPCAM-VH - bold, CH1 is underlined, followed by Fc, including P329GLA-LA mutation and hole mutation.

[0165] Nucleotide sequence, PBM0065:

[0166]

[0167] Amino acid sequence

[0168]

[0169]

[0170] Example 3. EPCAM-CD3 antibody sequences (Fig. 1. B constructs)

[0171] Figure 1B The structure of the humanized EPCAM ScFv-Linker-CD3 ScFv-His tag (BITE) is shown, which consists of one DNA construct (PBM0065). The signal peptide is underlined

[0172] Nucleotide sequence, PBM0065:

[0173]

[0174] Amino acid sequence

[0175] Signal peptide is underlined

[0176] EPCAM ScFv (VH-G4Sx3-VL): VH is bold, VL is bold and underlined

[0177] CD3 ScFv (VH-G4Sx3-VL): VH is italic, VL is italic and underlined, followed by 6xHis tag

[0178]

[0179]

[0180] Example 4. EPCAM-CD3 antibody sequences (Fig. 1. c constructs)

[0181] Figure 1C The structure of the humanized EPCAM ScFv-Linker G4Sx3-CD3 ScFv-human Fc is shown, which consists of one DNA construct (PBM0070).

[0182] DNA (PMC1367) was cloned into the same site as Example 2, inserted into the pYD11 vector.

[0183] Nucleotide sequence: signal peptide is bold and underlined, humanized EpCAM scFv-G4Sx3 CD3 scFv (BITE SEQ PBM0065, without 6xHis tag); human Fc part is underlined:

[0184]

[0185]

[0186] Amino acid sequence:

[0187] Signal peptide, indicated in bold underlined;

[0188]

[0189] The amino acid sequence of the EPCAM scFv-G4Sx3-CD3 ScFv (BITE, bold) - human Fc domain (underlined) is shown below:

[0190]

[0191] The amino acid sequence of CH2-CH3 of human IgGl Fc is shown below.

[0192] APELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK (SEQ ID NO: 20)

[0193] Example 5. DNA template sequences for transcription of EPCAM-CD3 antibody RNA of EPCAM-CD3 antibody (Figure. 1C construct)

[0194] Nucleotide sequence PMC1367 (used for construction Figure 1CThe template for the construct) was inserted into DNA that served as a template for in vitro RNA transcription to obtain PMC1396. Its structure is as follows: SP6 promoter: 5'-ATTTAGGTGACACTATAG-3' (underlined, SEQ ID NO: 21); 5' UTR (italic), EpCAM-CD3-human Fc sequence (bold), beginning with the ATG start codon (underlined) and ending with the stop codon TGA (underlined); 3' UTR (italic); 110 amino acid poly A tail with a linker in between.

[0195] Prior to the in vitro transcription reaction, the sequence was linearized using BglII enzyme AGATCT , underlined at the end of the poly A tail or the Asc I site GGCGCGCC (italic bold), located 6 bases 3' of the poly A tail).

[0196] During or after the in vitro transcription reaction, a 5' cap was added to improve the stability of the RNA. The RNA was detected on an agarose gel and then used to transfect 293 cells to produce the EPCAM-CD3-hFc antibody.

[0197]

[0198]

[0199] The present application also provides a DNA template PMC1549 with the T7 (AG) promoter underlined, where EpCAM-CD3-hFc is shown in bold. The structure of the PMC1549 EPCAM-CD3-hFc containing the T7 Ag promoter is shown below.

[0200]

[0201]

[0202]

[0203] The present application also prepared an EpCAM-CD3 antibody with the same nucleotide sequence as SEQ ID NO: 22, but with the human Fc region of the antibody containing L234A, L235A and P329G mutations, designated as PMC1970 (SEQ ID NO: 24). The sequence of the Fc mutations of the EPCAM-CD3 antibody is shown in bold. The T7 AG promoter is underlined.

[0204]

[0205]

[0206]

[0207] The translated antibody sequences are shown below, with the mutation sites (L234A, L235A and P329G) underlined and bolded.

[0208]

[0209]

[0210] The CH2-CH3 sequence of human IgGl Fc containing L234A, L235A and P329G substitutions is shown below.

[0211]

[0212] Example 6. Expression of EPCAM-CD3 antibody (Fig. 1A)

[0213] 293S cells were cultured in serum-free Freestyle F17 Expression media containing 8 mM L-glutamine (or GlutaMAX); 0.1% Pluoronic F-68. NanoFect transfection reagent was used at a ratio of 3:1 (3 microliters reagent per 1 microgram DNA) for transfection. Supernatant was collected 3-7 days post-transfection.

[0214] Antibody protein supernatant was run on SDS gel under reducing and non-reducing conditions (addition of beta-mercaptoethanol to lysis buffer) Figure 2 ). The gel under reducing conditions showed 4 bands. Protein was also purified through a protein A or G column. Purification used Millipore Sigma Protein A beads and Thermo IgG elution buffer (Cat. No. 21004). The purified EPCAM antibody showed a band of approximately 190 kDa under non-reducing conditions in SDS gel.

[0215] Example 7. FACS detection of binding of CD3 and EPCAM antigens Figure 1A Structure

[0216] FACS experiments using bispecific EPCAM-CD3 antibody Figure 1A showed that the antibody could bind EPCAM on EPCAM positive cells and CD3 on T cells Figure 3

[0217] ​​Bispecific antibodies were tested on EPCAM-positive and EPCAM-negative cell lines. CD3 binding was performed using CD3-positive T cells. The bispecific antibodies bound positively to both EPCAM and CD3 antigens. By FACS, the EPCAM-CD3 antibody bound to the EPCAM-positive Lovo cell line but not to the EPCAM-negative HL-60 cell line. Furthermore, the EPCAL-CD3 antibody showed positive staining on CD3-positive T cells.

[0218] Example 8. Cytotoxic activity of EPCAM-CD3 antibody with T cells against EPCAM positive target cell lines ( Figure 1A structure)

[0219] Antibody supernatant was used together with T cells for RTCA detection. The combined action of EPCAM-CD3 antibody and T cells can kill EPCAM-positive cells. Figure 3 Individual T cells also failed to kill target cells. Figure 3 This indicates that bispecific antibodies have high specificity when used in combination with T cells. The results confirm the mechanism by which bispecific antibodies guide T cells to cancer cells through binding to the CD3 antigen on T cells. Figure 3 It also showed that the bispecific antibody had dose-dependent activity.

[0220] Example 9. Bispecific humanized EPCAM-CD3 cross antibody (CrossFAB) SCFV binding to T cell specific secretion of interferon-γ Interferon-gamma Figure 1A Structure)

[0221] This bivalent, bispecific humanized EPCAM CrossFAB and CD3 ScFv antibody (such as...) Figure 1A As shown, PBM005 can induce T cells to secrete IFN-γ in EPCAM-positive Lovo cells. However, it does not induce IFN-γ secretion in EPCAM-negative HL-60 cells. Figure 4 The results showed that the antibody had high activity and specificity.

[0222] Example 10. EPCAM SCFV-CD3 SCFV antibody in combination with T cell specific killing of EPCAM positive cells and secretion of interferon-γ IFN-γ Figure 1B structure)

[0223] The EpCAM Scfv-CD3 ScFv-His tag (PBM0065) was purified using a Ni-NTA purification column and subjected to SDS gel electrophoresis. The results showed a band with a molecular weight of approximately 55 kDa.

[0224] By flow cytometry (FACS), EpCAM Scfv-CD3 ScFv (PBM0065) was shown to bind to Lovo cells but not to HL-60 cells.

[0225] When EpCAM Scfv-CD3 ScFv (PBM0065) acts in conjunction with T cells, it can target IFN-γ secreted at significant levels in EPCAM-positive Lovo cells, but has no such effect on EPCAM-negative HL-60 cells. Figure 5A The EC50 value of IFN-γ secretion was 5.28 ng / ml. Figure 5B ).

[0226] Example 11. EPCAM SCFV-CD3 SCFV-human Fc antibody (PBM0070) in combination with T cell specific killing of EPCAM positive cells and secretion of interferon-γ Figure 1C ( Figure 1C structure)

[0227] like Figure 6 The bispecific monovalent humanized EPCAM ScFv-CD3 ScFv-human Fc antibody with the structure shown exhibits a single band on an SDS gel under non-reducing conditions with a molecular weight of 162 kDa, while under reducing conditions, the molecular weight is 81 kDa corresponding to monomeric and dimeric proteins.

[0228] The humanized EPCAM ScFv-CD3 ScFv antibody binds to EPCAM in EPCAM-positive Lovo cells but not to EPCAM-negative HL-60 cells. The antibody also showed binding to CD3 by flow cytometry.

[0229] When EPCAM-CD3 antibody works in conjunction with T cells, it can target EPCAM-positive cells to secrete significant levels of IFN-γ, but has no effect on EPCAM-negative HL-60 cells. Example 12. EPCAM SCFV-CD3 SCFV-human Fc (PMC1396) produced by RNA transcription shows good activity in vitro When EPCAM-CD3 antibody interacts with T cells, the EC50 value of INF-γ secretion is 6.1 ng / ml (Figure 7).

[0230] Figure 1C Figure 1C ( Figure 9 structure)

[0231] EPCAM ScFV-CD3 ScFv-human Fc (PBM0070) antibody structure Example 13. LNP preparation materials The RNA was used as a template for RNA transcription. RNA was generated using an in vitro transcription kit. This RNA was transfected into 293 cells, and the supernatant was collected at 24 and 48 hours post-transfection for SDS-PAGE. An 81 kDa band was observed under reducing conditions at 24 and 48 hours post-transfection with PCM1396EPCAM Scfv-CD3 Scfv-hFc RNA. These results indicate that the bispecific antibody was successfully prepared via in vitro RNA transcription.

[0232] EPCAM-CD3-hFc protein in the supernatant collected 48 hours after RNA transfection could bind to EPCAM-positive Lovo cells and CD3-positive T cells, but not to EPCAM-negative HL-60 cells, by flow cytometry (FACS) analysis. We also detected two known EPCAM-negative cell lines (COLO320 and COLO741) by FACS, further confirming the specificity of the antibody.

[0233] EPCAM-CD3 antibody killed EPCAM-positive Lovo cells but not EPCAM-negative COLO741 or COLO320 cells by RTC (Figure 8). The antibody secreted IFN-γ in a dose-dependent manner when co-acting with T cells, but only against Lovo cells, not against COLO741 and COLO320 cells Example 14. In vitro transcription and mRNA encapsulation into LNP ).

[0234] Figure 1C

[0235] SM-102: 8-[(2-hydroxyethyl)[6-oxo-6-(undecyloxy)hexyl]amino]octanoic acid, 1- octylnonyl ester; CAS number: 2089251-47-6

[0236] ALC-0315: 2-hexyldecanoic acid, 1,1'-[[(4-hydroxybutyl)imino]di-6,1-hexanediyl] ester CAS number: 2036272-55-4.

[0237] ALC-0159: a-[2-(Tetracosylamino)-2-oxoethyl]-w-methoxy-poly(oxy-1,2- ethanediyl) CAS: 1849616-42-7

[0238] DMG-PEG2000: 1,2-dimyristoyl-rac-glycerol-3-methoxypolyethylene glycol-2000

[0239] DSPC: distearoylphosphatidylcholine

[0240] Cholesterol

[0241] Nanoparticles can be prepared using SM-102 or ALC315 ionizable lipids.

[0242] Example 15. EPCAM-CD3-HFC antibody expression in cancer cell lines transfected with EPCAM-CD3-LNP ( Figure 1C Structure)

[0243] mRNA was obtained by in vitro transcription from DNA template with HiScribe T7 mRNA kit with CleanCap reagent AG (NEB #E2080). Briefly, DNA template, 0.5x T7 CleanCap reagent AG reaction buffer, 5 mM ATP, CTP, pseudo UTP and GTP were added to 4 mM CleanCap AG and T7 polymerase mix, and reacted at 37 °C for 2 hours. Subsequently, DNAse I treatment was performed at 37 °C for 15 minutes. mRNA was purified using Monarch RNA Cleanup Kit (T2050) according to the manufacturer's protocol.

[0244] To generate mRNA-LNP complexes, an aqueous solution of mRNA (100 mM sodium acetate, pH 4.0) was mixed with a lipid mixture comprising SM-102 (Cayman), DSPC (Avanti), cholesterol (Sigma), and DMG-PEG2000 (Cayman) (molar percentages of 50:10:38.5:1.5, respectively) in an ethanol phase, using a PreciGenome Flex S system (San Jose, CA, USA) at a flow rate ratio of 3:1 (aqueous phase: organic phase). mRNA-LNPs were purified and concentrated by Ultra-15 centrifugal filter unit (30-100 kDa). The particle size, zeta potential, and polydispersity index (PDI) of mRNA-LNPs were detected using an Anton Paar Litesizer 500 system, and the encapsulation efficiency was detected using a Quant-it™ RiboGreen RNA assay kit. The average particle size of mRNA-LNPs was 104 nm, the PDI was 0.14, and the encapsulation efficiency was 91.8%.

[0245] Example 16. In vivo effectiveness of EPCAM-CD3-HFC-LNP with T cells in mice in vivo ( Figure 1C structure) Figure 10 Figure 10

[0246] The present application transfected 1 pg of EPCAM-CD3 mRNA-LNP (Example 14) into 0.5-1 x 10 6 cells of different ovarian cancer SKOV-3, OVCAR-3, and colorectal cancer OVCAR-5 cell lines. It was found that EpCAM-CD3 was secreted from the OVCAR-5 cancer cell line. The supernatant was combined with Lovo cells positive for EpCAM, but not with cells negative for EpCAM. The results showed that the antibody was produced inside the cancer cell line.

[0247] Figure 11A ( Figure 11BStructure)

[0248] To verify whether EPCAM-CD3 antibodies can be produced in vivo by EPCAM-CD3 RNA-LNP, the present application utilized EpCAM-CD3-hFc RNA-LNP (Example 14) and injected it into the OVCAR-5 xenograft tumor model in mice, either alone or in combination with T cells. We injected 4 x 10 6 colorectal cancer OVCAR-5 cell line subcutaneously into NSG mice, followed by injection of EpCAM-CD3-hFc-RNA-LNP into the tumor on days 1, 8, and 15. Human T cells were injected into the mice by intravenous injection two days after EpCAM-CD3-hFc-RNA-LNP injection (days 3, 10, 17). Injection of EPCAM-CD3 RNA-LNP into the tumor in combination with T cells significantly inhibited the growth of OVCAR-5 xenograft tumors compared to injection of T cells alone or EpCAM-CD3-hFc RNA-LNP alone ​ ).

[0249] In another experiment, ​ It was shown that using the OVCAR-5 xenograft mouse model, 3 intratumoral injections of EpCAM-CD3 RNA-LNP and only 1 injection of T cells significantly inhibited tumor growth. EpCAM-CD3 hFc mRNA-LNPs were injected intratumorally on days 6, 13, 20, and 25, while T cells were injected intravenously as a single injection on day 8 ​ ). GFP mRNA-LNPs were used as a negative control. EpCAM-CD3 mRNA-LNPs significantly inhibited the growth of OVCAR-5 xenograft tumors, while GFP mRNA-LNPs did not have this effect ​ ). Tumor images were evaluated at the end of treatment. The tumor volume and weight were significantly reduced in the EpCAM-CD3 hFc mRNA-LNP combined with T cells treatment group compared to the PBS and GFP mRNA-LNP treatment groups. In addition, serum was collected from the treated mice, and EpCAM-CD3 hFc mRNA-LNP treatment did not lead to an increase in hematotoxicity markers (AST, ALT, amylase, and LDH) compared to the GFP mRNA-LNP treatment group, indicating that EpCAM-CD3 mRNA-LNPs were not toxic. Therefore, delivery of EpCAM-CD3 human Fc mRNA-LNPs to the OVCAR-5 tumor xenograft model by intratumoral injection in combination with intravenous injection of T cells significantly inhibited the growth of xenograft tumors.

[0250] This result shows that bispecific antibodies can be generated in vivo, which significantly reduces the cost of antibody production. Bispecific EPCAM-CD3 antibodies produced by RNA-LNP delivery in combination with T cells suppressed the growth of mouse xenograft tumors.

[0251] References

[0252] 1. U. Schnell, V. Cirulli and B. N. Giepmans: EpCAM: structure and function in health and disease. Biochim Biophys Acta, 1828(8), 1989-2001 (2013) doi: 10.1016 / j.bbamem.2013.04.018

[0253] 2. J. Ni, P. J. Cozzi, W. Duan, S. Shigdar, P. H. Graham, K. H. John and Y. Li: Role of the EpCAM (CD326) in prostate cancer metastasis and progression. Cancer Metastasis Rev, 31(3-4), 779-91 (2012) doi: 10.1007 / s10555-012-9389-1

[0254] 3. L. Huang, Y. Yang, F. Yang, S. Liu, Z. Zhu, Z. Lei and J. Guo: Functions of EpCAM in physiological processes and diseases (Review). Int J Mol Med, 42(4), 1771-1785 (2018) doi: 10.3892 / ijmm.2018.3764

[0255] 4. E. Bremer and W. Helfrich: EpCAM-targeted induction of apoptosis. Front Biosci, 13, 5042-9 (2008) doi: 10.2741 / 3062

[0256] 5. R. Berahovich, H. Zhou, S. Xu, Y. Wei, J. Guan, J. Guan, H. Harto, S. Fu, K. Yang, S. Zhu, L. Li, L. Wu and V. Golubovskaya: CAR-T Cells Based on Novel BCMA Monoclonal Antibody Block Multiple Myeloma Cell Growth. Cancers (Basel), 10(9) (2018) doi:10.3390 / cancers10090323

[0257] 6. M. Bacac, T. Fauti, J. Sam, S. Colombetti, T. Weinzierl, D. Ouaret, W. Bodmer, S. Lehmann, T. Hofer, R. J. Hosse, E. Moessner, O. Ast, P. Brϋnker, S. Grau-Richards, T. Schaller, A. Seidl, C. Gerdes, M. Perro, V. Nicolini, N. Steinhoff, S. Dudal, S. Neumann, T. von Hirschheydt, C. Jaeger, J. Saro, V. Karanikas, C. Klein and P. Umana: A Novel Carcinoembryonic Antigen T-Cell Bispecific Antibody (CEA TCB) for the Treatment of Solid Tumors. Clin Cancer Res, 22(13), 3286-97 (2016) doi:10.1158 / 1078-0432.CCR-15-1696

[0258] 7. Wong, N. A.; Warren, B. F.; Piris, J.; Maynard, N.; Marshall, R.; Bodmer, W. F. EpCAM and gpA33 are markers of Barrett’s metaplasia.

[0259] J. Clin. Pathol. 2006, 59, 260-263

[0260] 8. Almagro, J. C.; Fransson, J. Humanization of antibodies. Front. Biosci. 2008, 13, 1619-1633.

[0261] 9. Golubovskaya, V.; Berahovich, R.; Zhou, H.; Xu, S.; Harto, H.; Li, L.; Chao, C. C; Mao, M. M.; Wu, L. CD47-CAR-T Cells Effectively Kill Target Cancer Cells and Block Pancreatic Tumor Growth. Cancers 2017, 9, 139.

[0262] 10. Bacac, M.; Fauti, T.; Sam, J.; Colombetti, S.; Weinzierl, T.; Ouaret, D.; Bodmer, W.; Lehmann, S.; Hofer, T.; Hosse, R. J.; et al. A Novel Carcinoembryonic Antigen T-Cell Bispecific Antibody (CEA TCB) for the Treatment of Solid Tumors. Clin. Cancer Res. 2016, 22, 3286-3297.

[0263] 11. Reinhard, K.; Rengstl, B.; Oehm, P.; Michel, K.; Billmeier, A.; Hayduk, N.; Klein, O.; Kuna, K.; Ouchan, Y.; Woll, S.; et al. An RNA vaccine drives expansion and efficacy of claudin-CAR-T cells against solid tumors. Science 2020, 367, 446-453

Claims

1. A bispecific antigen binding molecule comprising a humanized EPCAM VH, a first linker, a humanized EPCAM VL, a second linker, a CD3 VH, a third linker, a CD3 VL, and a human Fc domain.

2. The bispecific antigen binding molecule of claim 2, wherein the EPCAM VH has the amino acid sequence of SEQ ID NO: 10, the EPCAM VL has the amino acid sequence of SEQ ID NO: 4, the CD3 VH has the amino acid sequence of SEQ ID NO: 11, and the CD3 VL has the amino acid sequence of SEQ ID NO:

7.

3. The bispecific antigen binding molecule according to claim 1 or 2, wherein the human Fc domain comprises a hinge linker, and the CH2-CH3 region of human IgGl, optionally with one or two amino acid substitutions.

4. The bispecific antigen binding molecule according to claim 3, wherein the Fc domain has one or two amino acid substitutions selected from the group of L234A, L235A, and P329G (EU numbering).

5. The bispecific antigen binding molecule according to any one of claims 1 to 3, wherein each of the first, second, and third linkers has the amino acid sequence (GGGGS)n, and n = 1-5.

6. The bispecific antigen binding molecule according to claim 1, having the amino acid sequence of SEQ ID NO:

19.

7. The bispecific antigen binding molecule according to claim 1, wherein the human Fc domain comprises the amino acid sequence of SEQ ID NO: 20 or 26.

8. A bispecific antigen binding molecule comprising an EPCAM VH having the amino acid sequence of SEQ ID NO: 10, a first linker, an EPCAM VL having the amino acid sequence of SEQ ID NO: 4, a second linker, a CD3 VH having the amino acid sequence of SEQ ID NO: 11, a third linker, a CD3 VL having the amino acid sequence of SEQ ID NO: 7, wherein the first, second, and third linkers can be the same or different.

9. The bispecific antigen binding molecule according to claim 6, wherein each of the linkers has the amino acid sequence (GGGGS)3.

10. An isolated DNA sequence comprising: (a) a promoter coding sequence, (b) a 5'-UTR (untranslated region) coding sequence, (c) a coding sequence encoding a bispecific antigen binding molecule according to any one of claims 1 to 9, (d) a 3'-UTR coding sequence, and (e) a Poly A tail sequence.

11. A method of producing a bispecific antigen binding molecule in a cell, comprising the steps of: obtaining the DNA sequence of claim 10, transcribing the DNA sequence into mRNA in vitro using an RNA polymerase, mixing the mRNA with lipid nanoparticles (LNP) to form an mRNA-LNP complex, transfecting the mRNA-encapsulated LNP into a cell, and culturing the cell. ​ ​ ​ ​ The mRNA is translated in a cell to produce the bispecific antigen binding molecule.

12. The method of claim 11, wherein the cell is a cancer cell.

13. A method of treating cancer, comprising the steps of: obtaining the DNA sequence of claim 10, transcribing the DNA sequence into mRNA in vitro using an RNA polymerase, mixing the mRNA with a lipid nanoparticle (LNP) to form an mRNA-LNP complex, and injecting the mRNA-lipid nanoparticle complex into a tumor.

14. The method of claim 13, further comprising intravenous injection of T cells.

15. A bispecific antigen binding molecule comprising: (a) a first and a second antigen binding moiety, each moiety being a humanized Fab molecule capable of specific binding to human EpCAM, and each moiety comprising a heavy chain variable region having the amino acid sequence of SEQ ID NO: 10 (EPCAM VH) and a light chain variable region having the amino acid sequence of SEQ ID NO: 4 (EPCAM VL); (b) a third antigen binding moiety being a Fab molecule capable of specific binding to human CD3 epsilon (CD3), the third antigen binding moiety comprising a heavy chain variable region having the amino acid sequence of SEQ ID NO: 11 (CD3 VH) and a light chain variable region having the amino acid sequence of SEQ ID NO: 7 (CD3 VL), wherein the third antigen binding moiety is a crossover Fab molecule with exchange of the constant regions of the Fab light chain and the Fab heavy chain; and (c) a human Fc domain comprising a first subunit and a second subunit capable of stable association; wherein, the Fab heavy chain of the third antigen binding moiety (i) is fused at the N-terminus to the C-terminus of the Fab heavy chain of the first antigen binding moiety (CHI), and (ii) is fused at its C-terminus to the N-terminus of the first subunit of the Fc knob domain, and wherein the C-terminus of the Fab heavy chain of the second antigen binding moiety (CHI) is fused to the N-terminus of the second subunit of the Fc hole domain.

16. The bispecific antigen binding molecule of claim 15, wherein the human Fc domain comprises one or more amino acid substitutions that promote binding of the first subunit to the second subunit of the Fc domain.

17. A bispecific antigen binding molecule comprising two binding moieties that bind to EPCAM and one binding moiety that binds to CD3e, the molecule comprising the amino acid sequences of SEQ ID NOs: 5, 8, 12, and 14 in a molar ratio of 2: 1 : 1 : 1.

Images