Expression of antigen-binding proteins in the nervous system

Abstract

A method for delivering therapeutic proteins to the central nervous system for antibody-based therapy is provided. [Solution] A method for expressing a bivalent binding member in a cell of the nervous system, comprising the step of introducing into the cell an expression cassette encoding a polypeptide comprising an antibody heavy chain variable domain (VH), an antibody light chain variable domain (VL), and an IgG Fc region, wherein the VH and the VL form an antigen-binding site that specifically binds to a target protein, and two molecules of the polypeptide, when expressed in the cell, form a disulfide-linked homodimeric bivalent binding member specific to the target protein.

Description

[Technical Field]

[0001] Sequence Listing This application contains a Sequence Listing that has been submitted electronically in ASCII format and is incorporated by reference herein in its entirety. This ASCII copy, created on April 30, 2020, is named 022548_WO060_SL.txt and is 7,612 bytes in size. [Background technology]

[0002] Alzheimer's disease (AD) is characterized by progressive neurodegeneration resulting in memory impairment and cognitive decline. Its pathological hallmarks include the accumulation of extracellular amyloid plaques and intraneuronal tau fibrils. Therapies targeting amyloid beta (Aβ) have been under active investigation for many years due to its genetic and pathological involvement in AD (Non-Patent Document 1). Elevated levels of amyloid precursor protein (APP) and Aβ are associated with AD pathogenesis, but Aβ peptides exist in various conformations and fibrillar states, and it is unclear which species should be targeted for therapeutic benefit (Non-Patent Document 2).

[0003] Despite this uncertainty, passive immunotherapy against various forms of Aβ has been extensively tested in clinical settings, but such approaches are hampered by additional problems. First, the blood-brain barrier (BBB) ​​limits the transport of large biomolecules, necessitating peripheral injection of high doses to achieve therapeutically relevant levels in the brain. At high doses, some anti-Aβ antibodies in clinical trials have caused adverse reactions, such as amyloid-associated imaging abnormalities (ARIA), which are thought to result from antibody accumulation at vascular amyloid sites and induce local inflammation via Fc-dependent effector functions (Non-Patent Document 3). Second, the need to maintain levels above the minimum therapeutic dose necessitates long-term passive immunotherapy, which requires patient cooperation and compliance, as well as significant and substantial economic burden.

[0004] Gene transfer into the central nervous system (CNS) allows for the production of therapeutic proteins in neuronal cells, thereby bypassing the BBB. AAV-mediated expression of either whole immunoglobulin (IgG) or single-chain variable fragments (scFv) has been attempted in the CNS, but both of these approaches have inherent limitations (Non-Patent Document 4; Non-Patent Document 5; Non-Patent Document 6; Non-Patent Document 7; Non-Patent Document 8; Non-Patent Document 9; Non-Patent Document 10). Expression of IgG heavy and light chains in the CNS has only been achieved using a self-cleaving F2A sequence, which generates both chains from a single promoter cassette. The F2A peptide remains attached to either the heavy or light chain and is potentially immunogenic (Non-Patent Document 11). On the other hand, gene-based delivery of scFv proteins often results in a substantial loss of affinity due to reduced valency. Furthermore, removal of the Fc region reduces FcRn binding, shortens peripheral half-life, and reduces the efflux of antigen (Ag)-binding scFv from the brain by reverse transcytosis (Non-Patent Document 12; Non-Patent Document 13; Non-Patent Document 14; Non-Patent Document 15). Thus, antibody therapy for CNS diseases, such as Alzheimer's disease, is promising but is limited by the challenge of delivering therapeutic proteins into affected brains. Thus, there is a need for improved access to the central nervous system for antibody-based therapies. [Prior art documents] [Non-patent literature]

[0005] [Non-Patent Document 1] Tcw and Goate, Cold Spring Harb Perspect Med.(2017)7(6):pii a024539 [Non-patent document 2] Benilova et al., Nat Neurosci. (2012) 15:349~57 [Non-patent document 3] Mo et al., Ann Clin Transl Neu. (2017)4:931~42 [Non-patent document 4] Sudol et al., Mol Ther. (2009) 17:2031~40 [Non-Patent Document 5] Ryan et al., Mol Ther. (2010) 18:1471-81 [Non-patent document 6] Levites et al., J Neurosci. (2006) 26:11923~28 [Non-Patent Document 7] Levites et al., J Neurosci. (2015)35:6265~76 [Non-patent document 8] Kou et al., JAD.(2011)27:23-38 [Non-Patent Document 9] Fukuchi et al., Neurobio Dis. (2006) 23:502-11 [Non-Patent Document 10] Liu et al., J Neurosci. (2016)36:12425~35 [Non-Patent Document 11] Saunders et al., J Vir. (2015) 89:8334–45 [Non-Patent Document 12] Deane et al., J Neurosci. (2005) 25:11495-503 [Non-Patent Document 13] Boado et al., Bioconjug Chem. (2007) 18:447~55 [Non-Patent Document 14] Zhang et al., J Neuroimm. (2001) 114:168-72 [Non-Patent Document 15] Schlachetzki et al., J Neurochem. (2002) 81:203~6 Summary of the Invention [Means for solving the problem]

[0006] The present disclosure provides a method for expressing a bivalent binding member in cells of the nervous system, comprising: H ), antibody light chain variable domain (V L) and an expression cassette encoding a polypeptide comprising an IgG Fc region into a cell, H and V L forms an antigen binding site that specifically binds to a target protein, and two molecules of the polypeptide, when expressed in a cell, form a disulfide-linked homodimeric bivalent binding member specific for the target protein.

[0007] In some embodiments, the cells of the nervous system are neurons, glial cells, ependymal cells, or brain epithelial cells. In further embodiments, the glial cells are selected from oligodendrocytes, astrocytes, pericytes, Schwann cells, and microglial cells. In some embodiments, the cells are human cells, e.g., cells in the brain of a human patient.

[0008] In some embodiments, the target protein is a protein expressed in the brain and may be amyloid beta peptide (Aβ), tau, SOD-1, TDP-43, ApoE, or alpha-synuclein.

[0009] In some embodiments, the polypeptide comprises, from N-terminus to C-terminus: (i) V H , a peptide linker and V L ; or V L , a peptide linker and V H and (ii) an IgG Fc region. In a further embodiment, the peptide linker comprises the sequence GGGGS (SEQ ID NO: 3), e.g., the peptide linker has the sequence [G4S]3 (SEQ ID NO: 2). do.

[0010] In some embodiments, a bivalent binding member of the disclosure binds to the neonatal Fc receptor (FcRn) but does not bind to the Fc gamma receptor due to one or more mutations in the IgG Fc region.

[0011] In some embodiments, the method comprises administering a viral vector comprising an expression cassette. The viral vector can be a recombinant viral vector. In further embodiments, the recombinant virus is introduced into the patient's brain by intracranial injection, intrathecal injection, or intracisternal injection. The recombinant virus can be, for example, a recombinant adeno-associated virus (rAAV), for example, serotype 1 or 2 rAAV.

[0012] In some embodiments, expression of the polypeptide is under the transcriptional control of a constitutively active promoter or an inducible promoter.

[0013] The method may be used to treat patients with neurodegenerative diseases, such as Alzheimer's disease, cerebral amyloid angiopathy, synucleinopathies, tauopathies, or amyotrophic lateral sclerosis (ALS).

[0014] In another aspect, provided herein is a method of treating a neurodegenerative disease comprising administering to a patient in need thereof a therapeutically effective amount of a composition comprising a viral vector disclosed herein that expresses a bivalent binding member of the disclosure.

[0015] In another aspect, the disclosure provides a bivalent binding member for use in treating a patient in need thereof, and the use of a bivalent binding member for the manufacture of a medicament for treating a patient in need thereof, wherein the patient has, for example, a neurodegenerative disease, such as Alzheimer's disease, cerebral amyloid angiopathy, a synucleinopathy, a tauopathy or ALS.

[0016] Other features, objects, and advantages of the present invention will become apparent from the following detailed description. It should be understood, however, that this detailed description, while indicating embodiments and aspects of the present invention, is given by way of illustration only, and not by way of limitation. Various changes and modifications within the scope of the present invention will become apparent to those skilled in the art from this detailed description. [Brief explanation of the drawings]

[0017] [Figure 1A] Figure 1 shows the construction and characterization of AAV-IgG vectors. Vector design for efficient expression of heavy and light chains is shown. Genome size is indicated. [Figure 1B] Figure 1 shows construction and characterization of AAV-IgG vectors. The left panel shows durable expression and secretion of AAV-αAβIgG from the brain compared to huIgG measured from PBS-injected control mice. Graphed points represent mean + / - SEM, n=8 mice per group. The right panel shows the kinetics of AAV-mediated expression of AAV-αAβIgG versus traditional peripherally administered αAβIgG in the brain. Graphs show mean + / - SEM. **p<0.01, one-way ANOVA 7 weeks post-injection, n=5 mice per time point. [Figure 1C] Construction and characterization of AAV-IgG vectors. Color micrographs of neurons expressing the huIgG transgene throughout the hippocampus (CA2 shown in detail), with some nearby GFAP+ astrocytes also expressing huIgG. Cc = corpus callosum. Green: human IgG (huIgG). Red: glial fibrillary acidic protein (GFAP). Blue: DAPI. [Figure 2A] Figure 1 shows antigen binding by AAV-αAβIgG in a mouse model of Alzheimer's disease. The study design for intracranial (AAV-αAβIgG or AAV-IgG control) and peripheral doses (αAβIgG) is shown. [Figure 2B] Figure 1 shows antigen binding by AAV-αAβIgG in a mouse model of Alzheimer's disease. Expression of AAV-αAβIgG or AAV-IgG control is shown throughout the hippocampus and overlying cortex. The image in the right panel shows IgG binding to plaques in the frontal cortex. Scale bar = 10 μm. Blue: DAPI. Green: huIgG. Red: 4G8+GFAP. [Figure 3A]Figure 1. Assessment of AAV-αAβIgG neuronal expression and neurotoxicity. The left panel shows peptides detected from huIgG heavy and light chains in hemibrain lysates from SCID mice injected with AAV-αAβIgG compared to PBS-injected animals (sham), or in sham brain homogenates spiked with huIgG at levels equivalent to those in the AAV-αAβIgG group. The right panel shows quantification of functional huIgG compared to total huIgG expressed centrally or peripherally in SCID mice. Data are presented as mean + / - SEM. **p<0.01, unpaired Student's t-test. [Figure 3B] Figure 1. Assessment of AAV-αAβ IgG neuronal expression and neurotoxicity. The left panel shows H&E staining of the hippocampus of a C57BL / 6 mouse brain after 16 weeks of intrahippocampal AAV-αAβ msIgG expression compared to a PBS control. The inset shows details, and the arrows point to representative hyaline inclusions. Scale bar = 100 μm. Results are summarized in the table on the right panel as the number of animals scored with or without this pathology. [Figure 3C] Figure 1 shows assessment of AAV-αAβ IgG neuronal expression and neurotoxicity. Evidence of neuroinflammation by immunohistochemical (IHC) glial fibrillary acidic protein (GFAP) analysis is shown compared to PBS. The left panel shows quantitative IHC for GFAP+ area. In the right panel, each circle represents one mouse. Bars indicate group mean + / - SEM of GFAP+ area normalized to PBS. ***p<0.001, unpaired Student's t-test, n=8 mice per group. [Figure 4A] Figure 1 shows the construction and characterization of AAV-scFv-IgG vectors. The left panel shows a schematic of the scFv-IgG design. The center panel shows reducing or non-reducing SDS-PAGE analysis of purified scFv-IgG, demonstrating protein purity and proper disulfide-dependent dimer formation. The table in the right panel compares the antigen-binding affinity (M) of scFv-IgG versus IgG formats. [Figure 4B]

[0023] Figure 1 shows construction and characterization of AAV-scFv-IgG vectors. The left panel shows serum expression of AAV-scFv-IgG as measured by antigen enzyme-linked immunosorbent assay (ELISA) one month after peripheral IV injection of AAV into C57BL / 6 mice. The right panel shows brain expression of AAV-scFv-IgG. ***p<0.001, unpaired Student's t-test, n=5 mice per group for intracranial injection and 2 mice per group for IV injection. [Figure 4C]

[0033] Figure 4. Construction and characterization of AAV-scFv-IgG vector. The left panel shows hippocampal targeting of the vector and its transduction within the hippocampal formation after IHC on sagittal sections of mouse brain taken from the same animal as the right panel of Figure 4B. The right panel shows ELISA-based quantification of scFv-IgG in various dissected brain regions after bilateral hippocampal injection of AAV-scFv-IgG. Hipp = hippocampus. Ctx = overlying cortical region. Str = striatum. [Figure 5A] Figure 1. Expression, diffusion, and plaque association of anti-Aβ scFv-IgG. Whole scans of the hippocampus and overlying cortex of an adult mouse one month after injection with anti-Aβ AAV-scFv-IgG are shown. Sections were immunostained for Aβ plaques (4G8, red) and 6xHis (SEQ ID NO: 9) (green). Scale bar = 300 μm. Cc = corpus callosum. Images in the right panel show individual plaque ROIs (numbered in A) proximal (1) to distal (6) from the injection site. Abundant plaque formation was observed throughout the cortex (left panel), and staining with anti-His antibody colocalized with the plaques (right panel). Regions of interest (ROIs) are 150 μm in diameter. Red: 4G8. Green: anti-HIS antibody. Blue: DAPI. [Figure 5B] Figure 1 shows the expression, diffusion, and plaque binding of anti-Aβ scFv-IgG. The left panel shows a schematic of the study design. The image in the right panel shows the hippocampus of a coronal section from an AAV-injected mouse. IHC revealed labeling throughout the hippocampus on the injected side (red arrow), which also had additional transduction in the contralateral hippocampus. Empty AAV-injected brains did not show any anti-His labeling. Scale bar = 1 mm. [Figure 5C] Figure 1 shows the expression, diffusion, and plaque binding of anti-Aβ scFv-IgG. Quantification of plaque deposition in the cortex and hippocampus of each animal in each group is shown. n = 10–13 animals per group, 3 sections per animal. ***p < 0.001, one-way ANOVA with multiple comparisons. Error bars represent the standard error of the mean (SEM). DETAILED DESCRIPTION OF THE INVENTION

[0018] The present disclosure provides a method for expressing bivalent binding members in cells of the nervous system without the side effects of current expression methods. Nervous system cells do not naturally express antibodies. Previous studies have shown that expression of intact antibodies in the brain results in neurotoxicity. Compared to previous methods of expressing wild-type IgG in brain cells, the disclosed expression method results in unexpectedly high yields (e.g., two-fold or higher) and low toxicity (e.g., as indicated by the absence of detectable accumulation of intraneuronal hyaline protein at the injection site). Without being bound by theory, the inventors believe that cells in the nervous system lack the ability to efficiently express and assemble natural antibodies, and that unpaired antibody chains form inclusion bodies that are toxic to cells. However, the present expression method overcomes this problem by reducing the number of polypeptide chains from two to one. Furthermore, the present expression method is advantageous over previous methods of expressing scFvs in the brain because it enables the expression of binding molecules with higher binding activity and better pharmacokinetic profiles (e.g., half-lives).

[0019] Nervous system cells The present disclosure provides methods for expressing (e.g., secreting) bivalent molecules specific for target proteins expressed in the nervous system, e.g., the central nervous system, including the brain and spinal cord, in cells of the nervous system. Cells of the nervous system for expressing binding members of the present disclosure can be any cell type in the nervous system, e.g., any cell type in the brain. For example, the present methods may express binding members in neuronal cells (e.g., interneurons, motor neurons, sensory neurons, brain neurons, dopaminergic neurons, cholinergic neurons, glutamatergic neurons, GABAergic neurons, or serotonergic neurons); glial cells (e.g., oligodendrocytes, astrocytes, pericytes, Schwann cells, or microglial cells); ependymal cells, or brain epithelial cells. In some embodiments, such cells are human cells. The cells may also be located in any target region of the human brain, e.g., the hippocampus, cortex, basal ganglia, midbrain, or hindbrain.

[0020] Bivalent binding members The present disclosure provides a bivalent binding member that is expressed in cells of the nervous system and binds to a target antigen expressed in the nervous system, for example, the brain.The target antigen can be, for example, a protein that mediates a neurological disease, for example, a neurodegenerative disease.The antigen of interest includes, but is not limited to, amyloid beta peptide (Aβ), tau, SOD-1, TDP-43, ApoE, and α-synuclein.

[0021] A bivalent binding member is a homodimer of polypeptide chains, wherein the polypeptide chain comprises an antigen-binding domain and a constant region of an antibody (e.g., the hinge region, CH2 domain, and CH3 domain of an IgG, e.g., human IgG). Thus, the homodimer comprises two antigen-binding sites and an Fc domain of an antibody.

[0022] In some embodiments, the antigen-binding domain of the polypeptide chain is a single-chain Fv (scFv). The scFv domain is a domain. The scFv domain comprises an antibody heavy chain variable region (VH) and an antibody light chain variable region (VL), where the VH and VL are optionally separated by a peptide linker and interact to form an antigen-binding site. Methods for obtaining scFv polypeptides against a target antigen are well known in the art. For example, one method can be to screen a phage display library to obtain a combination of VH and VL that binds to the antigen with high affinity, or one method can obtain VH and VL sequences from an existing antibody that specifically binds to the antigen.

[0023] An antigen-binding domain, e.g., an scFv domain, can be fused to an antibody constant region, with or without a peptide linker (such as one exemplified herein, comprising a 9-Gly repeat linker (SEQ ID NO: 7)), where the antibody Fc domain is formed by the constant regions of the two polypeptide chains through one or more disulfide bonds. As used herein, the term "Fc region" or "Fc domain" refers to the portion of a native immunoglobulin formed by the dimeric association of one or more constant domains of the immunoglobulin.

[0024] In some embodiments, each polypeptide sequence of the Fc domain may comprise a portion of a single immunoglobulin (Ig) heavy chain, beginning in the hinge region just upstream of the papain cleavage site and ending at the C-terminus of the Ig heavy chain. The Fc domain may comprise the immunoglobulin hinge region, CH2, and CH3. Depending on the Ig isotype from which the Fc domain is derived, the Fc domain may comprise an additional constant domain (e.g., the CH4 domain of IgE or IgM). The Fc domain may comprise a mutation, compared to the wild-type sequence, that, for example, enhances the stability (e.g., half-life) of the fusion dimeric protein and / or alters the effector function of the fusion dimeric protein. The mutation may be the addition, deletion, or substitution of one or more amino acids.

[0025] In some embodiments, the Fc domain is derived from IgG, e.g., human IgG, and can be any IgG subtype, e.g., human IgG1, IgG2, IgG3, or IgG4 subtype. In such cases, the scFv-Fc of the present disclosure is also referred to as scFv-IgG. The Fc domain can include the entire hinge region or only a portion thereof of an IgG, e.g., an IgG1, IgG2, IgG3, or IgG4 hinge region. In some embodiments, the Fc domain is derived from human IgG1 and includes L234A and L235A mutations ("LALA") (EU numbering), such that the Fc domain does not bind to high-affinity Fc gamma (γ) receptor(s) and has reduced ADCC / CDC effector function. Other Fc mutations that can be introduced into human IgG1 include, but are not limited to, N297Q, N297A, N297G, C220S / C226S / C229S / P238S, C226S / C229S / E233P / L234V / L235A, and L234F / L235E / P331S (EU numbering). See, e.g., Wang et al., Protein Cell. (2018) 9(1):63-73; Strohl, Curr Opin Biotechnol. (2009) 20(6):685-91; Johnson et al., NatMed. (2009) 15(8):901-6. In some embodiments, the binding member has a hinge region derived from human IgG4, wherein the hinge region comprises a S228P mutation (EU numbering), which reduces dissociation of the two polypeptide chains of the binding member. In certain embodiments, the Fc domain is derived from human IgG4 and comprises S228P and L235E mutations (EU numbering; corresponding to S241P and L248E in Kabat numbering), which reduce Fcγ half molecule exchange and effector function, respectively (Reddy et al., J Imm. (2000) 164:1925-33). Loss or reduction of ADCC / CDC effector function allows the binding member to bind to target antigens without causing cytotoxicity or inducing unwanted inflammation in the nervous system. In further embodiments, the modified Fc domain binds to the neonatal Fc receptor FcRn. Retaining FcRn-binding ability allows the antigen-bound binding member to be removed from the nervous system, for example, the brain, by FcRn-mediated reverse transcytosis.

[0026] In some embodiments, the VH and VL domains of an scFv-Fc binding member and / or the scFv and Fc domains of a binding member are linked via a peptide linker. Suitable peptide linkers are well known in the art. See, for example, Bird et al., Science (1988) 242:423-26; and Huston et al., PNAS. (1988) 85:5879-83. The peptide linker may be rich in glycine and / or serine. Examples of peptide linkers are G, GG, G3S (SEQ ID NO: 1), G4S (SEQ ID NO: 3), and [G4S]n (n = 1, 2, 3, or 4; SEQ ID NO: 4). In some embodiments, a 9-Gly repeat linker (SEQ ID NO: 7) is used to link the scFv to the IgG portion in the scFv-IgG format of the present disclosure.

[0027] In certain embodiments, the scFv-IgG of the present disclosure is designed to link the variable domains via a peptide linker using a [G4S]3-type peptide linker (SEQ ID NO: 2). [G4S]3-type linkers (SEQ ID NO: 2) have been widely used to link variable domains in scFv structures (Huston, supra). As used herein, a [G4S]3-type linker (SEQ ID NO: 2) refers to [G4S]3 (SEQ ID NO: 2) or a functional variant thereof (e.g., a peptide linker having up to four amino acid modifications (e.g., insertions, deletions, and / or substitutions) of [G4S]3 (SEQ ID NO: 2)). By way of example, a functional variant of [G4S]3 (SEQ ID NO: 2) can be the amino acid sequence SGGGSGGGGSGGGGS (SEQ ID NO: 5) or the amino acid sequence GGGGSGGGGXGGGGYGGGGS (X=S, A, or N, and Y=A or N; SEQ ID NO: 6).

[0028] In some embodiments, the amino acid sequence of the linker may be modified. Modifications may include deletions or insertions that change the linker length (e.g., to adjust flexibility), or amino acid substitutions, including, for example, Gly to Ser, or vice versa.

[0029] An scFv-Fc polypeptide against Aβ is shown below, merely to illustrate one format of the scFv-Fc polypeptide: The sequence below includes, from N- to C-terminus, a signal peptide (italics), a VL, a [G4S]3 linker (SEQ ID NO: 2) (underlined), a VH, a G9 (SEQ ID NO: 7) (boxed), an IgG1 hinge and Fc domain, and a short linker (SEQ ID NO: 9) (bold) connecting to a 6xHis tag.

[0030] [ka]

[0031] Expression of binding members in the nervous system An expression construct containing a binding member expression cassette can be introduced into cells of the nervous system by well-known methods. For example, viral vectors can be used for in vivo or ex vivo delivery. In some embodiments, the expression vector remains stable episomal in the cell. In other embodiments, the expression vector is integrated into the genome of the cell. The expression vector can include expression control sequences, such as promoters, enhancers, transcription signal sequences, and transcription termination sequences, that enable expression of the binding member coding sequence in cells of the nervous system. Suitable promoters include, but are not limited to, the retroviral RSV LTR promoter (optionally with the RSV enhancer), the CMV promoter (optionally with the CMV enhancer), the CMV immediate early promoter, the SV40 promoter, the dihydrofolate reductase (DHFR) promoter, the β-actin promoter, the phosphoglycerate kinase (PGK) promoter, the EFlα promoter, the MoMLV LTR, the CK6 promoter, the transthyretin promoter (TTR), the TK promoter, the tetracycline-responsive promoter (TRE), the HBV promoter, the hAAT promoter, the LSP promoter, the chimeric liver-specific promoter (LSP), the E2F promoter, the telomerase (hTERT) promoter, and the CMV enhancer / chicken β-actin / rabbit β-globin promoter (CAG promoter; Niwa et al., Gene (1991) 108(2):193-9). In some embodiments, the promoter comprises the CMV enhancer linked to the human β-glucuronidase promoter or the chicken β-actin (CBA) promoter. The promoter may be a constitutive, inducible or repressible promoter.

[0032] Any method of introducing a nucleotide sequence into a cell may be utilized, including, but not limited to, electroporation, calcium phosphate precipitation, microinjection, cationic or anionic liposomes, liposomes combined with nuclear localization signals, naturally occurring liposomes (e.g., exosomes), or viral transduction.

[0033] Viral transduction may be used for in vivo delivery of the binding member expression cassette. Various viral vectors known in the art may be modified by those skilled in the art for use in the present disclosure, including, for example, recombinant adeno-associated viruses (rAAV), recombinant adenoviruses, recombinant retroviruses, recombinant poxviruses, and recombinant lentiviruses. In some embodiments, the viral vector used in the present invention is an rAAV vector. AAV vectors are particularly suitable for CNS gene delivery because they infect both dividing and non-dividing cells, exist as stable episomal structures for long-term expression, and have very low immunogenicity (Hadaczek et al., Mol Ther. (2010) 18:1458-61; Zaiss et al., Gene Ther. (2008) 15:808-16). Any suitable AAV serotype may be used. For example, AAV serotypes 1, 2, or 9 may be used. AAV may be modified to reduce the immunogenicity of its capsid protein in humans. In some embodiments, AAV1 serotype is used because it has shown excellent spread in the parenchyma and, although neuronal transduction predominates (like most AAV vectors), this serotype also transduces astrocytes, which may be particularly suitable for high-level protein expression and secretion.

[0034] The viral vectors described herein can be produced using methods known in the art. Any suitable permissive or packaging cell can be used to produce viral particles. For example, mammalian or insect cells can be used as packaging cell lines.

[0035] The expression construct, e.g., a recombinant AAV virus, can be injected intracranially, intrathecally, or It may be introduced into the patient's brain by intracisternal injection.

[0036] Applicable The expression methods of the present disclosure can be used to deliver therapeutic binding members to a patient's nervous system. The binding members are then expressed and secreted from transfected / transduced cells in the nervous system, exerting their therapeutic activity locally in the nervous system, e.g., the brain. These methods are useful for treating neurodegenerative diseases, such as Alzheimer's disease (e.g., Aβ and ApoE), cerebral amyloid angiopathy, synucleinopathies (e.g., α-synuclein), tauopathies (e.g., tau), or ALS (e.g., SOD-1 and TDP-43 (Pozzi et al., JCI (2019) doi:10.1172 / JCI123931)), Parkinson's disease (e.g., α-synuclein), dementia (e.g., tau (Sigurdsson, J Alzheimers Dis. (2018) 66(2):855-6)), Lewy body syndrome (e.g., α-synuclein (Games et al., J Neurosci. (2014) 34(28):9441-54), Huntington's disease (e.g., huntingtin (WO 2016016278)), and multiple system atrophy (e.g., P25α and α-synuclein (Games, supra)). In certain embodiments, the neurodegenerative disease is Alzheimer's disease. Binding members expressed locally in the nervous system target pathogenic antigens and remove them from the nervous system, e.g., the brain.

[0037] Thus, the present disclosure provides a method of treating a neurological disease (e.g., a neurodegenerative disease) in a subject, e.g., a human patient, in need thereof, comprising introducing into the nervous system of the subject a therapeutically effective amount (e.g., an amount that allows expression of the binding member sufficient to produce a desired therapeutic effect) of a viral vector (e.g., rAAV) comprising a coding sequence for a binding member for a target antigen that operably binds to a transcriptional control element(s) active in cells of the nervous system.

[0038] Pharmaceutical Composition In some embodiments, the present disclosure provides a pharmaceutical composition comprising a viral vector, such as a recombinant rAAV, containing an scFv-Fc binding member expression cassette in its recombinant genome.The pharmaceutical composition may further comprise a pharmaceutically acceptable carrier, such as water, saline (e.g., phosphate-buffered saline), dextrose, glycerin, sucrose, lactose, gelatin, dextran, albumin, or pectin.In addition, the composition may contain auxiliary substances, such as wetting and emulsifying agents, pH buffers, stabilizers, or other reagents that enhance the effectiveness of the pharmaceutical composition.The pharmaceutical composition may also comprise a delivery vehicle, such as liposomes, nanocapsules, microparticles, microspheres, lipid particles, and vesicles.

[0039] Delivery of rAAV to a subject can be achieved, for example, by intravenous administration. In certain instances, it may be desirable to locally deliver rAAV to brain tissue, spinal cord, cerebrospinal fluid (CSF), neurons, glial cells, meninges, astrocytes, oligodendrocytes, interstitial spaces, etc. In some cases, recombinant AAV can be directly delivered to the CNS by injection into the ventricular region, striatum and neuromuscular junction, or cerebellar lobes. AAV can be delivered by needle, catheter, or related device using neurosurgical techniques known in the art, e.g., by stereotactic injection (see, e.g., Stein et al., J Vir. (1999) 73:3424-9; Davidson et al., PNAS. (2000) 97:3428-32; Davidson et al., Nat Genet. (1993) 3:219-23; and Alisky and Davidson, Hum. Gene Ther. (2000) 11:2315-29).

[0040] The routes of administration are intracerebral, intrathecal, intracranial, intracerebral, intraventricular, intrathecal, intracisternal, intravenous. In some embodiments, the viral vector is administered intravenously, intranasally, or intraocularly, for example, by intrathecal and / or intracerebral injection, or by intracisternal injection, and then spreads throughout the CNS tissue. In other embodiments, the viral vector crosses the blood-brain barrier and achieves widespread distribution throughout the CNS tissue of a subject after intravenous administration. In some aspects, the viral vector has a unique CNS tissue targeting ability (e.g., CNS tissue tropism), which allows stable and non-toxic gene transfer to be achieved with high efficiency.

[0041] For example, pharmaceutical compositions can be administered to patients by intraventricular administration, for example, to the ventricular region of the patient's forebrain, for example, the right ventricle, the left ventricle, the third ventricle or the fourth ventricle.Pharmaceutical compositions can be administered to patients by intracerebral administration, for example, by injecting the composition into or near the cerebrum, medulla, pons, cerebellum, intracranial cavity, meninges, dura mater, arachnoid mater or pia mater.Intracerebral administration can, in some cases, include administering the drug to the cerebrospinal fluid (CSF) in the subarachnoid space surrounding the brain.

[0042] In some cases, intracerebral administration involves injection using stereotaxy. Stereotaxy is well known in the art and typically involves the use of a computer and a three-dimensional scanning device, which are used together to guide injection into a specific brain region, such as the ventricular region. A microinjection pump (e.g., from World Precision Instruments) can also be used. In some cases, a microinjection pump is used to deliver a composition containing a viral vector. In some cases, the infusion rate of the composition ranges from 1 μl / min to 100 μl / min. As will be understood by those skilled in the art, the infusion rate will depend on various factors, including, for example, the subject's species, the subject's age, the subject's weight / size, the AAV serotype, the required dose, and the brain region to be targeted. Therefore, other infusion rates can be determined to be appropriate by those skilled in the art in specific situations.

[0043] Unless otherwise defined herein, scientific and technical terms used in connection with this disclosure shall have the meanings commonly understood by those of ordinary skill in the art. Exemplary methods and materials are described below; however, methods and materials similar or equivalent to those described herein can also be used in the practice or testing of this disclosure. In case of conflict, the present specification, including definitions, will control. Generally, the nomenclature used in connection with and techniques of cell and tissue culture, molecular biology, immunology, microbiology, genetics, analytical chemistry, organic synthetic chemistry, medicinal chemistry, and protein and nucleic acid chemistry and hybridization described herein are those well known and commonly used in the art. Enzymatic reactions and purification techniques are performed according to manufacturer's specifications, as commonly accomplished in the art or as described herein. Furthermore, unless otherwise required by context, singular terms shall include pluralities, and plural terms shall include the singular. Throughout this specification and embodiments, the words "have" and "comprise" or variations thereof, such as "has," "having," "comprises," or "comprising," are understood to mean the inclusion of a specified integer or group of integers, but not the exclusion of any other integer or group of integers. "About" can be understood as being within 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.1%, 0.05%, or 0.01% of the specified value. Unless otherwise clear from the context, all numerical values ​​provided herein are modified by the term "about." It is understood that aspects and variations of the invention described herein include those "consisting of" and / or "consisting essentially of" aspects and variations. All publications and other references mentioned herein are incorporated by reference in their entirety. Although numerous documents are cited herein, this citation does not constitute an admission that any of such documents form part of the common general knowledge in the art. In case of conflict, the present specification, including definitions, will control. In addition, the materials, methods, and examples are not intended to be limiting. These are illustrative examples and are not intended to be limiting.

[0044] In order that this invention may be better understood, the following examples are set forth. Such examples are for illustrative purposes only and are not to be construed in any way as limiting the scope of the invention. [Example]

[0045] In the following examples, we demonstrate that a single-chain antibody (Ab), termed silent scFv-IgG, fused to an Fc domain that maintains FcRn binding but lacks Fc gamma receptor (FcγR) binding, can be expressed and released into the CNS after AAV-mediated gene transfer. Incorporating Fc into the scFv-IgG design restores the bivalency of canonical IgG, conferring high avidity to multimeric targets such as aggregated amyloid, and provides the molecule with the ability to modulate Fc-dependent signaling as needed. Preservation of Fc binding to FcRn at the blood-brain barrier may be enhanced by allowing antibody-antigen clearance from the brain via FcRn-mediated efflux upon reduction of amyloid pathology previously observed with scFv alone. Although expression of canonical IgG in the brain resulted in signs of neurotoxicity, this engineered antibody (Ab) was efficiently secreted from neurons and retained target specificity. Brain steady-state levels exceeded peak levels achieved by intravenous injection of Ab. In a transgenic ThyAPPmut mouse model of progressive amyloid plaque accumulation, AAV expression of this scFv-IgG reduced cortical and hippocampal plaque burden compared with controls. These findings suggest that CNS gene delivery of a silent anti-Aβ scFv-IgG is well-tolerated, durably expressed, and functional in a relevant disease model, demonstrating the potential of this modality for the treatment of Alzheimer's disease and other neurological disorders.

[0046] The materials and methods used in the studies described in the Examples below are described below.

[0047] Study design This study was initiated to design an anti-Aβ IgG for AAV-mediated delivery to the brain for the treatment of Alzheimer's disease. Such an IgG construct was designed and first After two to four in vitro studies to confirm appropriate expression, association, and antigen-binding activity, in vivo studies were performed. The sample size for C57BL / 6 or SCID animal studies was determined based on the variability observed in previous studies using stereotactic AAV delivery to express transgenes in vivo and was defined for each experiment. Studies testing in vivo expression were performed two to three times. The sample size for ThyAPPmut mice for quantifying amyloid plaque burden was determined to account for the expected interanimal variability in plaque formation. Based on previous studies using this strain, efficacy studies were performed once per group with an n of ≥ 10. Animals were randomly assigned to each group for all studies. ROI identification for automated image analysis was performed by researchers blinded to the experimental conditions. All animal studies were performed in accordance with appropriate guidelines.

[0048] AAV-IgG design The variable regions were derived from either the original 13C3 mouse anti-Aβ antibody (for AAV-αAβmsIgG) or humanized sequences (for AAV-αAβIgG) (Schupf et al., PNAS (2008) 105:14052-7), as described in patent applications WO 2009 / 065054 and WO 2010 / 130946, respectively. huIgG expression vectors were generated by inserting the coding sequences for a human IgG4 heavy chain (Reddy et al., J Imm. (2000) 164:1925-33) and a kappa light chain containing two amino acid substitutions described to reduce half-molecule (S241P) and effector function (L248E) into a dual promoter cassette (without requiring the 2A peptide cleavage sequence shown in Figure 1A). Experiments requiring a mouse IgG1 framework were performed. In this study, the original 13C3 antibody (Vandenberghe et al., Sci Rep. (2016) 6:20958) was used to reduce effector function by adding the N297A mutation to the heavy chain. The AAV control IgG vector encoded the huIgG4PE isotype control antibody, which targets a non-mammalian antigen.

[0049] scFv-IgG design The design of the scFv-IgG is shown (Figure 4A; SEQ ID NO: 8). Briefly, the variable light and variable heavy chain regions of the 13C3 anti-amyloid beta parent antibody were linked by three repeats of a flexible G4S linker (SEQ ID NO: 2) to form a VL-VHscFv. The scFv sequence was followed by an additional nine-repeat glycine linker (SEQ ID NO: 7) containing the native mouse IgG1 hinge and CH2 and CH3 domains, thereby comprising the Fc region of the scFv-IgG (Balazs et al., Nature (2011) 481:81-4). Similar to AAV-αAβmsIgG, asparagine 297 of the Fc was mutated to alanine (N297A) to attenuate effector function (Chao et al., Immunol Invest. (2009) 38:76-92; Jefferis et al., Immunol Rev. (1998) 163:59-76). A C-terminal 6xHis epitope tag (SEQ ID NO: 9) was included to facilitate both in vitro purification and in vivo detection in mice. Expression of scFv-IgG was driven by a hCMV / hEF1a promoter expression cassette with a Tbgh polyA.

[0050] immune tolerance To induce immune tolerance, mice were injected IP with 7.5 mg / kg of GK1.5 anti-CD4 monoclonal antibody (Bioxcell) on days 0, 2, and 10. To confirm CD4 T cell depletion, blood was collected by retro-orbital sampling into heparin-coated tubes on day 12. CD4+ T lymphocytes were quantified using CD45-FITC (clone 104 BD Pharmigen™), CD3e-AlexaFluor647 (clone 17A2, eBioscience), and CD4-PE (RM4-4 clone, BioLegend) antibodies using FACS analysis on a BD Fortessa using standard protocols. GK1.5-treated animals had a reduction in CD4+ lymphocytes, as evidenced by a ratio of CD4+ lymphocytes / total CD3+ lymphocytes of 0.04 + / - 0.008 (mean + / - SEM) in treated mice compared with 0.47 + / - 0.003 in untreated mice.

[0051] Cell culture, protein expression and purification Expi293™ cells (LifeTech) were passaged in Expi293T™ serum-free medium (LifeTech) and used for protein expression. Expression plasmids were transfected into Expi293™ cells by lipid transfection (Fectopro, Polyplus), and cell culture medium containing secreted proteins was harvested 4 days later. After sterile filtration, 6xHis (SEQ ID NO: 9)-tagged proteins were purified by immobilized metal affinity chromatography (IMAC). Briefly, proteins were batch-adsorbed onto cobalt resin (Thermo Scientific™) overnight at 4°C, washed with 10 column volumes of phosphate-buffered saline, and then eluted with 500 mM imidazole. Proteins were dialyzed overnight into HEPES-buffered saline, concentrated (Centricon®), and frozen at -80°C until use.

[0052] ELISA 96-well Immulon™ IIHB (Thermo) plates were filled with 1 μg / mL Aβ for antigen ELISA. 1-42(Bachem H-1368) or 1 μg / mL mouse anti-huIgG polyclonal Ab (Jackson 209-005-088), and incubated overnight at 25°C in carbonate buffer. The wells were washed five times with TBS-0.5% tween (TBST) and blocked for 1 hour in TBSTB (TBST + 1.5% BSA). A standard curve using purified proteins was run in parallel with serum or brain homogenate to allow quantification of bound scFv-IgG or huIgG. Samples were incubated for 2.5 hours, washed three times with TBST, and then incubated with an HRP-conjugated secondary antibody for 1 hour. After five TBST washes, the wells were incubated with TMB substrate for 5 minutes and then quenched with 0.5 M H2SO4. Plate-bound signal was quantified by absorbance at 450 nm (Spectramax M5). All samples were run in triplicate.

[0053] LC-MS / MS LC / MS / MS experiments were performed on a Q Exactive™ mass spectrometer (Thermo Scientific™) equipped with a NanoAcQuity LC system (Waters). IgG from tissue homogenates was specifically enriched and isolated using CaptureSelect™ HuIgG affinity resin (Thermo Fisher). The enriched IgG was digested by DTT reduction and alkylation followed by overnight incubation at 37°C with trypsin / Lys-C (1:100 w / w). The digestion was terminated by the addition of 1% formic acid (FA). The resulting tryptic peptide mixture was loaded onto a microcapillary column (id 75 μm, HSST3 15 cm, 1.8 μm, Waters) and resolved. Data were collected in PRM mode at a resolution of 70,000 (m / z 200) and an AGC target of 5 × 10. 6and a maximum injection time of 500 ms. A planned inclusion list was created based on the profiling data of a control IgG. The PRM method utilized target ion isolation with a 2 Da isolation window and fragmentation with a normalized collision energy (NCE) of 25. MS / MS scans were acquired with a starting mass range of 100 m / z and acquired as profile spectrum data types. Precursor and fragment ions were quantified using Skyline (MacCoss Lab Software).

[0054] Surface plasmon resonance 1 mg / mL Aβ 1-42 Peptide (Bachem H-1368) was incubated overnight in 10 mM HCl at 37°C with shaking at 600 rpm. The resulting fibril solution was directly immobilized via amine coupling onto a CM5 sensor chip (GE Healthcare). Antibody or scFv-IgG solutions prepared at 50, 30, 20, 10, and 5 nM in PBS-+P buffer (GE Healthcare) were injected at a relatively high flow rate (50 μL / min) to limit avidity effects. Data were processed using Biacore™ T200 evaluation software, and after double referencing by subtraction of a blank surface and buffer-only injections, the data were globally fitted to a 1:1 binding model.

[0055] Preparation of AAV ITR plasmids and adeno-associated virus vectors The IgG or scFv-IgG expression cassette was subcloned into an AAV2-ITR-containing plasmid, which, as needed, retained the A1AT stuffer DNA to maintain the AAV genome size for proper packaging. In the case of the dual-promoter IgG ITR plasmid, the stuffer DNA was not included because the cassette was already the maximum size allowed for efficient packaging. The AAV empty vector consisted of the CBA promoter, Tbgh polyA, and the A1AT stuffer DNA. AAV2 / 1 virus was generated by transient transfection. Briefly, HEK293 cells were transfected with a 1:1:1 ratio of three plasmids (containing ITR, AAVrep / cap, and Ad helper) using PEI (polyethyleneimine). The Ad helper plasmid (pHelper) was obtained from Stratagene / Agilent Technologies (Santa Clara, CA). Purification was previously performed. This was performed using column chromatography as previously described (Burnham et al., Hum Gene Ther Methods (2015) 26:228-42). Virus titer was tested using qPCR against poly(A) sequences, and AAV was stored at 80°C in 180 mM sodium chloride, 10 mM sodium phosphate (5 mM monobasic + 5 mM dibasic), 0.001% F68, pH 7.3 until use.

[0056] animal The animals used were 2-month-old C57BL / 6 males obtained from The Jackson Laboratory (Bar Harbor, USA) unless otherwise stated. 2-month-old adult SCID mice (B6.CB17-Prkdc) were obtained from The Jackson Laboratory. scid / SzJ). ThyAPPmut transgenic mice backcrossed to C57BL / 6 have been described by Blanchard et al., Exp Neurol. (2003) 184:247-63. The surgical group was housed singly to allow adequate recovery from brain surgery. Mice were maintained on a 12-hour light / dark cycle and had free access to food and water. Animals were randomized into various groups, and analyses were performed by an operator blinded to the treatment groups.

[0057] Stereotactic injection Surgery was performed according to procedures approved by the Animal Care and Use Committee. Mice were deeply anesthetized by intraperitoneal injection of a mixture (volume 10 ml / kg): ketamine (100 mg / kg; Imalgene; Merial, France) and xylazine (10 mg / kg; Rompun; Bayer, France). Before placing the animals in a stereotaxic frame (Kopf Instruments, USA), the scalp was shaved and disinfected with Vetidine (Vetoquinol, France). The local anesthetic bupivacaine (2 mg / kg, volume 5 ml / kg; Aguettant, France) was injected subcutaneously onto the cranial skin, and Emla (Lidocaine, Astrazeneca) was applied intra-ear. During surgery, eyes were protected from light with vitamin A Dulcis, and body temperature was maintained constant at 37°C with a heating blanket.

[0058] Samples were injected at a rate of 0.5 microliters per minute. The needle was left in place for no more than 2 minutes to prevent backflow of the sample from the needle tract and then slowly withdrawn from the brain. Unilateral hippocampal injections were performed in ThyAPPmut mice, and bilateral injections were performed in all other mice. The coordinates for hippocampal injection were AP-2.0, DV-2.0, and ML+ / -1.5. After surgery, mice were kept warm and given a subcutaneous injection of carprofen (5 mg / kg, volume 5 ml / kg, Rimadyl®, Zoetis) and continuously observed until recovery. At the end of the study, mice were euthanized by anesthesia overdose with Euthasol® (USA) or ketamine / xylazine (France). After overdose, mice were kept warm until perfused with ice-cold PBS.

[0059] Immunohistochemistry After perfusion with cold PBS, brain tissue was fixed in 10% neutral buffered formalin (NBF). Formalin-fixed tissue was embedded in paraffin and then sectioned in 5 μm sagittal or coronal planes. All tissues were stained using a Leica BOND RX automated stainer. For immunofluorescence staining, heat-mediated antigen retrieval was performed using epitope retrieval solution 1 (ER1; citrate buffer, pH 6.0) for 10 minutes. Tissues were then blocked / permeabilized with goat serum + 0.25% Triton X-100, incubated with primary antibody for 1 hour at room temperature, washed with TBST, and then incubated with secondary antibody for 30 minutes. Nuclei were detected using Spectral DAPI (Life). Plaque quantification was performed using biotin-conjugated 4G8 antibody (4G8 clone, BioLegend, Inc. 800701) using the Vectastain® ABC (PK-7100) kit according to the manufacturer's instructions without antigen retrieval and formic acid extraction. The tissues were immunostained.

[0060] antibody 6xHis (SEQ ID NO: 9) (Abcam Ab9108, 1:1000 IHC, Invitrogen™ R931-25, 1:1000 Western, ELISA), GFAP (Ebiosciences 41-9892-82, 1:200 or Abcam Ab4674, 1:500 IHC), 4G8 (BioLegend 800701, 1:500 IHC). Life Technologies secondary antibodies: Cy3 goat anti-mouse, Alexa Fluor® 647 goat anti-rabbit, Alexa Fluor® 488 goat anti-chicken; all 1:500. Amyloid DAB: 4G8-biotin (BioLegend 800705 1:250).

[0061] Image analysis Immunohistochemistry slides were scanned at 20x magnification using a Scanscope® XT brightfield image scanner (Aperio, Vista, CA) or AxioScanZ1 (Carl Zeiss Microscopy GmbH, Germany). Whole-slide images (WSIs) of GFAP IHC were viewed and analyzed using HALO™ image analysis software (Indica Labs, Corrales, NM, USA). For each WSI, hippocampal area was manually annotated and analyzed for GFAP immunopositive area using HALO's automated area quantification algorithm. For each sample, for selected ROIs, the GFAP-positive area was divided by the total tissue area to obtain the percentage of immunopositive area. For plaque analysis, 5 μm coronal brain sections were taken from 6-month-old ThyAPPmut mice at three different levels, 50 μm apart. Cortical and hippocampal ROIs were manually annotated. Amyloid plaque burden was analyzed using ZEN2 software (Carl Zeiss Microscopy GmbH, Corrales, NM, USA). DAB+ tissue area was quantified using a custom image analysis algorithm developed using Zeiss Microscopy GmbH, Germany. Data were plotted using GraphPad Prism version 6 (GraphPad Software, LaJolla, CA, USA).

[0062] statistics Statistical analysis was performed using Graphpad Prism (v6 and v7) using one-way analysis of variance with multiple comparisons (Dunnett) for experiments with more than two groups. An unpaired Student's t-test was used for comparisons of two groups. *p<0.05, **p<0.01, ***p<0.001. Sample sizes varied and are indicated for each experiment.

[0063] Example 1: Construction and characterization of AAV-IgG vectors targeting β-amyloid To generate gene-based expression of the antibody, we used a dual promoter expression cassette to express a humanized version of the 13C3 antibody, which binds to fibrillar and fibrillar Aβ but has no affinity for the monomeric form, as described in Schupf, supra. The IgG4 heavy chain contained S228P and L248E mutations that reduce Fcγ effector function and half-molecule exchange (Yang et al., Curr Opin Biotechnol. (2014) 30:225-9; Reddy et al., J Imm. (2000) 164:1925-33).

[0064] The heavy and light chains were expressed from different promoters, and the entire cassette was designed to be compatible with the AAV genome packaging restrictions (Figure 1A). The dual promoter design used here avoids potential immunogenicity or expression trends induced by other designs that use a single promoter but do not require the use of an F2A cleavage sequence or internal ribosome entry site for bicistronic expression (Saunders, supra). (Mizuguchi et al., Mol Ther. (2000) 1:376-82). Because the AAV1 serotype has demonstrated excellent parenchymal spread and predominantly transduces neurons (like most AAV vectors), this serotype also transduces astrocytes, making it particularly suitable for high-level protein expression and secretion. Therefore, we packaged this cassette within an AAV1 capsid and injected it directly into the brain (AAV-αAβIgG). To test AAV-αAβIgG expression, we used C57BL / 6-SCID (SCID) mice to prevent anti-huIgG immune responses that could interfere with transgene expression. Antibodies are actively transported out of the brain by reverse transcytosis. Therefore, we monitored brain expression of AAV-αAβIgG by biweekly serum collection. Serum was drawn at 2-week intervals for 16 weeks after bilateral injection of AAV-αAβ IgG into the hippocampus of SCID mice (2E10GC per side). 1-42 A fiber-binding immunoassay was used to measure the expression levels of functional antibodies after bilateral hippocampal injection of 2E10GC AAV-αAβIgG.

[0065] The vector demonstrated stable expression for up to 16 weeks (Figure 1B, left). To gain insight into how AAV-mediated antibody expression in the brain compares to levels observed with standard passive immunotherapy approaches, we measured huIgG levels in the hippocampus of SCID mice at various time points in parallel with another group subjected to a single intravenous (IV) bolus injection of 20 mg / kg αAβIgG. SCID mice received a single IV injection of 2E10GC AAV-αAβIg bilaterally into the hippocampus or a single IV injection of 20 mg / kg purified IgG. After tissue harvest at the indicated times, brains were exposed to IgG over a time course. The ipsilateral hippocampus was homogenized, and huIgG was quantified by antigen ELISA. The AAV-αAβIgG vector sustained hippocampal expression of approximately 300 ng / g over the time course, as measured by antigen ELISA (Figure 1B, right). IgG levels in the hippocampus 24 hours after IV injection approached 200 ng / g, but as IgG was cleared from the brain (in a strain with a known serum half-life), this level declined and by 7 weeks was 11-fold lower compared to AAV-αAβ IgG.

[0066] Figure 1C shows that neuronal and glial expression of AAV-IgG was detectable in the hippocampus. Specifically, expression in both neurons and astrocytes was confirmed by IHC against the huIgG expression product; neurons were easily identifiable by morphology in the CA2 region of the hippocampus, and astrocytic expression was demonstrated by colocalization with GFAP (Figure 1C).

[0067] These data demonstrate that the AAV-αAβIgG vector can maintain steady-state levels of antibody in the brain significantly higher than those achievable with traditional passive immunotherapy protocols.

[0068] Example 2: Antigen binding by AAV-αAβIgG in a mouse model of Alzheimer's disease We then expressed AAV-αAβIgG in an amyloid plaque mouse model expressing a mutant amyloid precursor protein (ThyAPPmut) to assess the extent of brain transduction and determine whether the antibody is secreted into the extracellular space and binds to plaques. This model demonstrates progressive amyloid plaque accumulation in the cortex from approximately 2-3 months of age (Blanchard et al., Exp Neurol. (2003) 184:247-63). To prevent anti-huIgG antibody responses, animals were immunotolerized with CD4-depleting antibodies before and after vector administration (Figure 2A). Briefly, to easily detect IgG in mice, we injected AAV expressing AAV-αAβIgG or isotype control IgG (AAV-IgG control) into the hippocampus of 2-month-old male ThyAPPmut mice. ThyAPPmut mice were CD4 T cell depleted between days 2 and 10. Mice were immunotolerized with AAV-αAβ IgG or isotype control vector AAV-IgG control were injected bilaterally into the hippocampus on days 4-5 (2E10GC per injection). Another group received weekly IP injections of 10 mg / kg purified αAβ huIgG as a positive control for plaque-binding activity throughout the study. After 8 weeks, 5 μm sagittal brain sections were harvested and immunostained. This αAβ IgG dose and IP delivery paradigm has previously been shown to result in plaque binding in ThyAPPmut animals in vivo (Pradier et al., Alzheimer's & Dementia (2013) 9(4):808-809).

[0069] Two months after injection, at the age when these animals exhibit plaque deposition in the frontal cortex, sagittal sections of the brain were subjected to IHC. Specifically, huIgG IHC staining revealed expression throughout the hippocampus surrounding the needle tract and the overlying cortex. Enlargement of the region of interest (ROI) (500 μm wide) reveals detailed huIgG expression in neurons and the hippocampal neuropil. In contrast, the IP-injected αAβ IgG group, which stained exclusively in amyloid plaques, did not show any expression in cell bodies (Figure 2B, left). Fluorescent IHC of huIgG, Aβ plaques, and GFAP demonstrated colocalization of huIgG with cortical plaques in both the AAV-αAβ IgG and IV αAβ IgG groups, but not in the AAV-IgG control group. In particular, AAV-αAβIgG and peripherally delivered αAβIgG exhibited clear binding to 4G8+ amyloid deposits, whereas the AAV-IgG control exhibited no detectable binding (FIG. 2B, right).

[0070] These data indicate that AAV-αAβIgG was secreted into the extracellular space and was able to bind to Aβ plaques in brain regions distal to the injection site.

[0071] Example 3: Evaluation of AAV-αAβIgG neuronal expression and neurotoxicity Neurons are highly specialized to secrete factors related to neurotransmission, rather than large macromolecules such as IgG. It is unclear whether efficient processing and secretion of IgG can occur in such cells. To determine whether there is inappropriate processing of neurally expressed IgG, we performed mass spectrometry to measure the overall levels of heavy and light chains in the brain of SCID mice 1 month after AAV-αAβIgG expression. Expression of AAV-αAβIgG in the hippocampus, similar to saline-injected brain lysates spiked with purified αAβIgG, was at the expected level of heavy chains, but the cognate light chain was unexpectedly low compared to spiked controls (Figure 3A). This finding suggests that AAV-αAβIgG expression in brain cells results in insufficient light chain production, resulting in an imbalance in the ratio of heavy to light chains.

[0072] We also used ELISA to quantify the percentage of total IgG (heavy and light chains) versus this population capable of binding antigen (Ag). Specifically, the levels of functional Aβ antibodies in brain extracts from AAV-αAβIgG-expressing SCID mice were quantified by antigen ELISA and compared in parallel with pan-huIgG ELISA. We observed that approximately 20% of total IgG expressed from the brain (2E10 total GC injected into the hippocampus) was functional, whereas AAV-αAβIgG expressed from peripheral tissues by IV vector injection (1E12 total GC injected IV) had no imbalance in total IgG / functional IgG. Specifically, the level of antigen-bound huIgG accounted for only 21% of total huIgG when expressed from the brain, but this discrepancy was not detected in serum 1 month after peripheral vector expression (Figure 3A, right).

[0073] We then investigated whether there was evidence of neurotoxicity as a result of IgG expression. In our initial characterization of the AAV-αAβIgG vector, we used a huIgG version of this antibody, which has more direct translation to humans, allowing clear detection in mice. However, we were unable to determine any neurotoxicity that may be associated with brain IgG expression. To test the toxicity or neuroinflammation of αAβ IgG without the confounding variables of cross-species huIgG exposure, we used an AAV vector called AAV-αAβmsIgG, which expresses the original murine version of αAβ IgG (Schupf, supra; Pradier, supra; Vandenberghe et al., Sci Rep. (2016) 6:20958). This vector was injected into the hippocampus of C57BL / 6 mice, and brain tissue was processed for histological examination one month later. Histopathological analysis revealed a high incidence of hyaline / eosinophilic cytoplasmic deposits in hippocampal neurons, suggestive of high glycoprotein expression (Figure 3B). Neuronal inclusion of eosinophilic hyaline-like material, suggestive of glycoprotein accumulation, was observed only in brains injected with the antibody-expressing vector. Such structures were also observed in the hippocampus of mice injected with the AAV-IgG control (6 / 12 mice), indicating that this toxicity was not specific to αAβIgG expression.No such hyaline deposits were observed in the hippocampus of mice injected with the AAV1 empty vector or PBS alone (Figure 3B).

[0074] We also observed evidence of neuroinflammation by immunohistochemical GFAP analysis compared to PBS. In this experiment, C57BL / 6 mice were injected with either PBS or AAV-αAβmsIgG (2E10GC into the hippocampus), and 5 μm sagittal brain sections were collected 16 weeks later (Figure 3C). Furthermore, AAV1 empty vector did not induce significant gliosis compared to PBS (1.11 + / - 0.12, 5 mice, mean + / - SEM normalized to PBS GFAP+ area), suggesting that neuroinflammation was due to IgG expression.

[0075] These data indicate that although brain cells can express and secrete IgG, only a subset of approximately 20% of this IgG is functional and capable of binding antigen, and that this expression induces detectable neuroinflammation throughout the transduced area.

[0076] Example 4: Construction and characterization of AAV-scFv-IgG vectors While IgG delivered by our vector secreted and bound to amyloid plaques in vivo, we hypothesized that a selective Ig format could minimize AAV-IgG-induced mispairing and neurotoxicity. Based on the same mouse αAβ antibody (Schupf, supra), we synthesized an engineered single-chain Fv with the variable region of an IgG light chain fused by its COOH-terminus (C-terminus) to the heavy chain variable region linked to the hinge, CH2, and CH3 domains of mouse IgG1 (Figure 4A, scFv-IgG). To minimize the pro-inflammatory effects of the Fc region, we mutated the mouse IgG1 Fc domain to eliminate glycosylation at asparagine 297 (N297A), thereby preventing binding to all FcγRs (Johnson, supra; Chao, supra). Specifically, the scFv-IgG was designed to have the variable regions of mouse anti-Aβ IgG linked via three repeats of a flexible GGGGS (SEQ ID NO: 3) linker sequence. The scFv was linked to mouse IgG1 via a 9-Gly repeat linker (SEQ ID NO: 7). The scFv-IgG was conjugated to N297A Fc. A 6xHis tag (SEQ ID NO: 9) was added to the C-terminus. The scFv-IgG was expressed in Expi293 cells and purified by immobilized metal affinity chromatography (IMAC) using the C-terminal histidine (His) tag sequence.

[0077] SDS-PAGE analysis confirmed that the protein efficiently assembled into disulfide-linked dimers (Figure 4A). Surface plasmon resonance (SPR) analysis of this scFv-IgG demonstrated that it produced fibrillar Aβ fragments comparable to those of the parent antibody. 1-42 The affinity (M) was measured by SPR using scFv-IgG or IgG at various molar concentrations on immobilized Aβ. 1-42 The parental IgG exhibited a slightly lower binding affinity of 5.2 × 10 by the scFv-IgG. -10 compared to 1.3 × 10 -10 The apparent dissociation constant (KD ) (Fig. 4A, Table).

[0078] This expression cassette was inserted into an AAV1 vector to determine whether the modified IgG could be synthesized in vivo. Because peripheral tissues have been well validated for the expression and secretion of IgG molecules, IV injection of AAV was used as a positive control for the activity of the virus by the inventors (Saunders, supra; Shimada et al., PloS ONE (2013) 8:e57606; Hicks et al., Sci Transl Med. (2012) 4:140ra187; Chen et al., Sci Rep. (2017) 7:46301; Balazs et al., Nature (2011) 481:81-4; Balazs et al., Nat Biotech. (2013) 31:647-52; Balazs et al., Nat Med. (2014) 20:296-300). One month after IV injection of AAV-scFv-IgG (total 1E12GC), serum levels reached 63 μg / mL, demonstrating robust AAV vector activity in peripheral tissues ( Fig. 4B , left).

[0079] To assess vector brain expression, we quantified scFv-IgG levels in extracts from one sagittal brain hemisection, referred to as the hemibrain, 1 month after hippocampal injection of AAV whole 2E10GC into C57BL6 mice. Expression levels averaged approximately 600 ng / g (Figure 4B, right). Notably, this concentration was more than threefold higher than that observed 24 h after IV injection of 20 mg / kg IgG and 2.5-fold higher than that observed with AAV-αAβIgG (Figure 1B). Histological analysis revealed that despite having higher levels of expression in the brain than the AAV-αAβIgG vector, AAV-scFv-IgG transduction did not result in any detectable intraneuronal hyaline protein accumulation in the injected hippocampus (0 / 5 mice), suggesting that scFv-IgG was processed more efficiently by neurons than IgG.

[0080] To clarify the brain distribution of scFv-IgG-transduced cells, DAB-6xHis IHC (disclosed as "6xHis" SEQ ID NO: 9) was performed on sagittal sections 1 month after hippocampal injection using an antibody against the His tag. The AAV-scFv-IgG vector transduced the entire hippocampus, with low transduction density in the hippocampus surrounding the needle tract and in the cortical region overlying the subiculum (Figure 4C). Brains transduced with the negative control empty AAV (AAV control) vector showed no detectable anti-His immunostaining (Figure 4C). It should be noted that only intracellular expression was detected by anti-His IHC in C57BL6 mice, as any secreted extracellular scFv-IgG may have been washed away due to a lack of available antigen.

[0081] Both intracellular and extracellular scFv-IgG expression was assessed biochemically in ipsilateral brain regions both proximal and distal to the injection site. One month after AAV injection, brain regions from three mice were dissected, and expressed protein was quantified by antigen ELISA in each brain region. Background signals were subtracted using PBS-injected brain homogenates. Specifically, the hippocampus, overlying cortex, and striatum were dissected, homogenized, and scFv-IgG was quantified by antigen ELISA (Figure 4C, right). A concentration gradient was observed, with the highest levels observed at the injection site (hippocampus) and progressively lower levels observed in more distal brain regions (Figure 4C, right). Despite having lower levels compared to the injection site, the concentration of scFv-IgG in striatal tissue remained near 200 ng / g, a brain steady-state level not typically achieved by passive IgG infusion.

[0082] Example 5: Antigen binding by scFv-IgG in a mouse model of β-amyloidosis We determined whether AAV-delivered scFv-IgG could be secreted into the extracellular space and bind to antigens in vivo. AAV-scFv-IgG vectors were injected into the hippocampus of 5-month-old female ThyAPPmut mice (Blanchard, supra), an age at which they had already developed plaques throughout the neocortex. One month after unilateral injection of 1 μL of 1E10GC total acetaminophen, 5 μm sagittal sections of the brain were processed for IHC and stained for His-tag reactivity and Aβ plaques. The image on the right shows individual plaque ROIs (numbered in A) proximal (1) to distal (6) from the injection site. The image was overlaid with 6xHis (SEQ ID NO: 9) immunostaining (green) and DAPI (blue) (Figure 5A). As expected, abundant plaque formation was observed throughout the cortex (Figure 5A, left), and anti-His antibody staining colocalized with the plaques (Figure 5A, right). Notably, there was a gradual but widespread decrease in the intensity of 6xHis (SEQ ID NO: 9) labeling for plaques more distal to the hippocampus and occipital cortex regions where AAV-scFv-IgG was expressed, indicating a clear concentration gradient of plaque-bound scFv-IgG, with plaques distal to the hippocampus showing progressively lower levels of bound scFv-IgG than plaques closer to the injection site. These data indicate that anti-Aβ scFv-IgG is expressed and secreted from hippocampal cells, enabling it to bind to plaques distal to the injection site.

[0083] These data provided evidence that viral vector-delivered scFv-IgG binds to physiologically relevant targets in vivo. We then determined whether long-term expression in this mouse model of amyloidosis could reduce plaque formation. The study design is outlined below. Four groups of 2-month-old ThyAPPmut male mice (approximately the age at which plaques begin to form) were unilaterally injected into the hippocampus with either AAV-scFv-IgG, an AAV control vector, Aβ IgG, or a control isotype IgG. These groups were compared with animals treated with passive immunotherapy using weekly IP injections of 10 mg / kg mouse anti-Aβ antibody (Aβ IgG) or an isotype control antibody (Figure 5B, left). Brains were harvested 16 weeks (4 months) after treatment, and coronal sections were immunostained for amyloid plaques or 6xHis (SEQ ID NO: 9) and analyzed for transgene expression. AAV-scFv-IgG was expressed throughout the injected hippocampus, and there was also clear transport of the vector to the contralateral subiculum, as evidenced by αHis staining of cell bodies (Figure 5B, right). Aβ plaque burden in the cortex and hippocampus was quantified by anti-His IHC on coronal brain sections. ROIs from both hemispheres were quantified combined, and plaque burden was expressed as a percentage of tissue ROI area as DAB-positive staining. Despite differences in plaque burden between control groups compared with their respective controls, a single injection of AAV-scFv-IgG resulted in a reduction in plaques in the hippocampus of the same magnitude as the αAβ IgG criterion (Figure 5C). Plaque reduction was also significantly reduced in the cortex (Figure 5C), consistent with evidence that scFv-IgG diffuses from the expression site and binds to distal plaques.

[0084] These results demonstrated that a single injection of AAV-scFv-IgG in an amyloid mouse model resulted in durably expressed and secreted AAV from the injection site and binding to plaques throughout the brain. A typical passive immunotherapy regimen of 10 mg / kg anti-Aβ IgG weekly for 16 weeks resulted in a significant reduction in amyloid plaque formation in ThyAPPmut animals. In contrast, a single intracranial injection of AAV-scFv-IgG resulted in comparable efficacy 4 months after expression.

[0085] In summary, in the above studies, scFv-IgG derived antibodies specific for fibrillar and fibrillar Aβ species reduced amyloid plaque burden in vivo. Our scFv-IgG expressed well in vitro, allowing for purification and subsequent analysis of antigen-binding affinity by SPR. Compared to the IgG form, scFv-IgG bound to the antigen to a similar extent. AAV1 was selected as the serotype for this efficacy because this capsid facilitates vector dissemination in the CNS after parenchymal injection. This serotype primarily infects neuronal cells but also transduces some non-neuronal cell types, expanding the potential repertoire of cells available for transgene expression. Using a relatively high dose (1E10GC in the hippocampus), steady-state levels at the injection site were significantly higher than the serotypes we used for comparison. These results were 3-4 times higher than the levels maximally achievable by the passive IgG criterion chosen for this study (antibody levels in the brain 24 hours after IV injection of 20 mg / kg purified IgG). A peripheral IV dose of approximately 60 mg / kg was required to reach the levels achieved by the AAV-scFv-IgG vector.

[0086] After a single injection, expression in the hippocampus persisted for at least 4 months, and protein concentrations exceeded passive immunotherapy criteria, even in brain regions several millimeters distal to the injection site. Because 6xHis (SEQ ID NO: 9)-positive cells were not found far beyond the injection site or needle tract (data not shown), it is unlikely that transduced cells migrated from the injection site to secrete protein in regions distal to the hippocampus. Long-term expression of this vector in ThyAPPmut mice resulted in a reduction of plaques in both the cortex (52% reduction) and hippocampus (87% reduction). This is a more efficient reduction than that observed in other studies utilizing scFv, where plaque reduction ranged from 0 to 60% (Levites, 2006, supra; Levites, 2015, supra; Kou, supra; Fukuchi, supra; and Wang et al., Brain, Behavior, and Immunity (2010) 24:1281-93). The observed magnitude of plaque reduction in animals treated with a single intracranial injection of AAV-scFv-IgG was similar to that in animals treated with weekly IV injections of 10 mg / kg anti-Aβ antibody for 4 months, highlighting the utility of gene delivery for long-term treatment paradigms.

[0087] [Table 1]

Claims

1. The use of recombinant viruses containing expression constructs for the manufacture of pharmaceuticals for treating human patients who require them, The expression construct comprises an expression cassette encoding a polypeptide containing an antibody heavy chain variable domain (VH), an antibody light chain variable domain (VL), and an IgGFc region. VH and VL form antigen-binding sites that specifically bind to target proteins in the nervous system. The use wherein two polypeptide molecules, when expressed in cells of the nervous system of a human patient, form a disulfide bond homodimer divalent bond member specific to the target protein.

2. The use according to claim 1, wherein the target protein is selected from amyloid-beta peptide (Aβ), tau, SOD-1, TDP-43, ApoE, or α-synuclein.

3. The use according to claim 1 or 2, wherein the cells are human cells.

4. The use according to claim 3, wherein the cells of the nervous system are neurons; optionally, glial cells selected from oligodendrocytes, astrocytes, Schwann cells, pericytes and microglial cells; ependymal cells, and brain epithelial cells.

5. The use according to any one of claims 1 to 4, wherein the target protein is a protein expressed in the brain.

6. The use according to any one of claims 1 to 5, wherein the bivalent binding member binds to the neonatal Fc receptor (FcRn), but does not bind to the Fc gamma receptor due to one or more mutations in the IgG Fc region selected from C220S, C226S, C229S, N297Q, N297A, N297G, E233P, L234V, L234F, L234A, L235A, L235E, P331S, and P238S (EU numbering).

7. The polypeptide moves from the N-terminus to the C-terminus. (i) VH, peptide linker and VL; or VL, peptide linker and VH, and (ii) IgG Fc region The use according to any one of claims 1 to 6, including the use described in any one of claims 1 to 6.

8. The use according to claim 7, wherein the peptide linker comprises the sequence of GGGGS (SEQ ID NO: 3).

9. The use according to any one of claims 1 to 8, wherein the recombinant virus is introduced into the patient's brain by intracranial injection, subarachnoid injection, or cisterna magna injection.

10. The use according to any one of claims 1 to 9, wherein the recombinant virus is recombinant adeno-associated virus (AAV).

11. The use according to claim 10, wherein the recombinant AAV is a recombinant AAV of serotype 1 or 2.

12. The use according to any one of claims 1 to 11, wherein the expression of the polypeptide is under the transcriptional control of a constitutively active promoter or an inducible promoter.

13. The use according to any one of claims 1 to 12, wherein the patient has a neurodegenerative disease.

14. The use according to claim 13, wherein the patient has Alzheimer's disease, cerebral amyloid angiopathy, synuclein disease, tauopathy, or amyotrophic lateral sclerosis.

15. The use according to any one of claims 1 to 14, wherein the target protein is amyloid-beta peptide (Aβ).

Images