Enzyme capable of degrading amyloid-β, and pharmaceutical composition using same

Abstract

According to the present invention, it is found that a snake venom metaloproteinase SVMP (snake venom metalloproteinase) from a snake belonging to the family Viperidae, the family Elapidae or the like decomposes Aβ and cuts the Aβ at an α-position to suppress the formation of Aβ fibrils. When a protein containing a metaloproteinase domain of SVMP is administered or a gene for the protein is introduced and is expressed in the brain, the formation of Aβ fibrils can be suppressed and consequently Alzheimer's disease can be treated.

Description

Enzyme that degrades amyloid beta and pharmaceutical composition using the same

[0001] The present invention relates to a pharmaceutical composition that contributes to the treatment of Alzheimer's disease.

[0002] Alzheimer's disease is a progressive brain disease that gradually impairs memory and thinking ability. Alzheimer's disease is a neurodegenerative disease characterized by the accumulation of amyloid beta (Aβ) plaques in the brain, known as senile plaques (amyloid plaques), and neurofibrillary tangles (Non-Patent Document 1). Although the cause of Alzheimer's disease is unclear, it has been revealed that multiple factors, including genetic factors, environment, and lifestyle, are involved. Alzheimer's disease accounts for approximately 70% of dementia in the elderly, and it is estimated that there are 55 million patients worldwide. Furthermore, as the elderly population increases, the number of Alzheimer's disease patients is also increasing, posing a major global problem.

[0003] It is said that amyloid deposition in the brain begins more than 20 years before clinical symptoms of Alzheimer's disease appear. Therefore, early detection of Alzheimer's disease is difficult. While progress has been made in the development of diagnostic methods, such as biomarkers for Alzheimer's disease, the lack of diagnostic accuracy makes it difficult to analyze the etiology and develop a fundamental treatment. Clinical trials of over 400 candidate drugs for Alzheimer's disease have been conducted to date, but most have failed. Currently, cholinesterase inhibitors such as dopenesil and galantamine and NMDA receptor antagonists are approved as treatments, but their effectiveness is temporary and limited. Antibody drugs such as lecanemab and aducanumab, which remove accumulated Aβ, have been developed and have been approved or approved in various countries. However, antibody drugs also suffer from individual differences in efficacy, side effects, and high treatment costs, making the development of new drugs awaited.

[0004] Development of therapeutic drugs for Alzheimer's disease has progressed from drugs such as cholinesterase inhibitors that improve symptoms and delay progression to drugs that suppress the progression of brain lesions and fundamentally treat Alzheimer's disease. While senile plaques and neurofibrillary tangles are considered to be pathological changes in the brain in Alzheimer's disease, the "Aβ hypothesis," which posits that the accumulation of Aβ, which constitutes senile plaques, is the underlying cause of the disease, and drug development is progressing. Patent Document 1 discloses a therapeutic drug for Alzheimer's disease that regulates Aβ levels by selectively regulating γ-secretase activity. Patent Document 2 discloses the use of ADAMTS (a disintegrin and metalloproteinase with thrombospondin motif) polypeptides in the treatment of many diseases, including Alzheimer's disease as one of the target diseases. However, no data demonstrating therapeutic effects for any of the diseases has been disclosed, making it impossible to evaluate efficacy.

[0005] Furthermore, numerous proteases that degrade Aβ have been identified, and their therapeutic applications are also being investigated. It has been reported that intracerebral Aβ clearance is reduced in sporadic patients, who account for approximately 95% of all Alzheimer's disease patients. Therefore, if proteases that degrade Aβ or proteases that suppress Aβ production can be supplemented or their activity enhanced, Alzheimer's disease may be treatable. Proteases that degrade or suppress Aβ production include metalloproteases such as neprilysin (NEP), ADAM (ADAM-9, 10, 17), and matrix metalloproteases (MMP-2, -3, -6, -9, -14), as well as serine proteases such as kallikrein-related peptidase 7 (KLK7) and plasmin, BACE1, BACE2, etc. (Non-Patent Documents 2-4). NEP has been reported to have effects such as reducing brain Aβ levels and amyloid burden in Alzheimer's disease model mice, but no therapeutic drug has yet been developed.

[0006] JP-T-2012-508770A JP-A-2001-327297A

[0007] Selkoe, D.J. &Hardy, J. EMBO Mol. Med. 2016, 8, 595-608.Saido, T. &Leissring, M.A. Cold Spring Harb. Perspect. Med. 2012, 2, a006379.Kikuchi,K. et al., J. Cell. Biochem. 2017, 118, 4183-4190.Kidana,K. et al., EMBO Mol. Med. 2018, 10, e8184.Oliveira,A.L. et al., Nat. Rev. Chem., 2022, 6, 451-469.Muttenthaler,M. et al., Nat. Rev. Drug Discov., 2021, 20, 309-325.Yan,P. et al., J. Biol. Chem. 2006, 281, 24566-24574.Iwatsubo, T. etal., Neuron, 1994, 13, 45-53.Szaruga, M. etal., J. Exp. Med., 2015, 212, 2003-2013.De Strooper, D.et al., Neurol., 2010, 6, 99-107.Panza, F. et al.,Drugs Aging, 2009, 26, 537-555.Takeda, S. Toxins2016, 8, 155Shibata, H. et al., Sci. Rep., 2018, 8,11300.Ogawa,T. et al., Toxins, 2019, 11, 581.Wijeyewickrema,L.C. et al., Toxicon, 2007, 50, 1064-1072.Kurosumi,et al., Bioorg. Med. Chem. Lett. 2010, 20, 5282-5285.

[0008] As described above, although efforts have been made to develop pharmaceutical compositions that can fundamentally treat Alzheimer's disease by regulating Aβ levels, inhibiting its accumulation, and inhibiting the formation of senile plaques, no effective therapeutic drug has yet been developed. As described below, an objective of the present invention is to develop a pharmaceutical composition that can fundamentally treat Alzheimer's disease.

[0009] The present invention relates to the following amyloid fibril formation inhibitors and pharmaceutical compositions utilizing the same. (1) An amyloid fibril formation inhibitor containing, as an active ingredient, a region containing at least the metalloprotease domain of snake venom metalloprotease SVMP (snake venom metalloproteinase). The inventors previously demonstrated through genome analysis that Habu venom contains a metalloprotease similar to the ADAM family, a type of protease that degrades APP. Purification of Habu venom and analysis of protease activity revealed that Habu venom metalloprotease cleaves Aβ at the α site. Snake venom metalloproteases are highly conserved among venomous snakes throughout evolution. Snake venom metalloproteases highly homologous to the ADAM family act as amyloid fibril formation inhibitors.

[0010] (2) The amyloid fibril formation inhibitor according to (1), wherein the snake is a snake of the Viperidae family or the Elapidae family. Snake venoms of the Viperidae family, to which the Habu snake shown in the following Examples belongs, and the Elapidae family are known to contain proteases homologous to ADAM. While endogenous proteases in humans have a mechanism for inhibiting them, exogenous Viperidae or Elapidae snake venom proteases are not thought to be inhibited by the endogenous inhibitory mechanism, and therefore are thought to be highly therapeutically effective in the human body as an amyloid fibril formation inhibitor.

[0011] (3) The amyloid fibril formation inhibitor according to (2), wherein the snake of the Viperidae family is a Habu, and the SVMP is at least one of HV1, flavorase, VMP-III_like, jerdonitin-like precursor, HR1a, H2pro, HR2a, flavoridin precursor, HR1b, elegantin precursor, and NaMP. As shown below, metalloprotease contained in Habu venom inhibits the formation of amyloid fibrils by cleaving Aβ at the α site. Since amyloid fibril formation is thought to correlate with the progression of Alzheimer's disease, Habu venom, which has been shown to be effective in the examples, is expected to be a therapeutic agent for inhibiting the progression of Alzheimer's disease.

[0012] (4) The amyloid fibril formation inhibitor according to (3), wherein the SVMP is a flavoridin precursor. Detailed analysis has revealed that, among snake venom metalloprotease SVMPs, the flavoridin precursor exhibits strong cleavage activity. The flavoridin precursor is highly effective as an amyloid fibril formation inhibitor.

[0013] (5) A pharmaceutical composition for treating Alzheimer's disease, comprising the amyloid fibril formation inhibitor according to any one of (1) to (4) as an active ingredient. As shown below, an amyloid fibril formation inhibitor containing the metalloprotease domain of SVMP degrades Aβ. Therefore, it is expected that a protein preparation containing this as an active ingredient will function effectively in the human body.

[0014] (6) A pharmaceutical composition comprising a gene therapy product for treating Alzheimer's disease, wherein the gene therapy product is a nucleic acid sequence encoding a region containing at least the metalloprotease domain of snake venom SVMP. Since Alzheimer's disease progresses over a long period of time, progression can be suppressed by introducing at least the metalloprotease domain of SVMP into the brain, allowing it to be continuously expressed, and cleaving Aβ. Gene transfer can be performed by any method that can introduce the metalloprotease domain and ensure long-term gene expression.

[0015] (7) The pharmaceutical composition according to (6), wherein the gene therapy product comprises a viral vector. Introduction using a viral vector is preferably used because the method is well established.

[0016] (8) The pharmaceutical composition according to (7), wherein the viral vector is an adeno-associated virus (AAV). Adeno-associated viruses are non-pathogenic and can be efficiently introduced into non-dividing cells, making them suitable for persistently expressing snake venom metalloproteases in the brain.

[0017] (9) The pharmaceutical composition according to any one of (6) to (8), wherein the SVMP is at least one of the peptides represented by SEQ ID NOs: 1 to 9, 16, and 17. For the treatment of Alzheimer's disease, it is preferable to use a snake venom metalloprotease represented by SEQ ID NOs: 1 to 9, 16, and 17.

[0018] (10) The pharmaceutical composition according to (9), wherein the SVMP is a peptide represented by SEQ ID NO: 1. Analysis has revealed that the flavoridin precursor has strong cleavage activity, and therefore, a gene therapy product that expresses a region containing at least the metalloprotease domain of the flavoridin precursor represented by SEQ ID NO: 1 can be preferably used as a gene therapy product.

[0019] Schematic diagram showing the amyloidogenic and non-amyloidogenic pathways of human amyloid precursor protein (APP). Schematic diagram showing the domain structure of ADAM (a disintegrin and metalloproteinase) family proteins. Chromatogram of gel filtration purification of crude Habu snake venom. Purification process of gel filtration G2 fraction by metal ion affinity chromatography (left), and SDS-PAGE of crude snake venom (CV), gel filtration G2 fraction (GF), and affinity-purified fraction (AF) (right). Diagram showing protease activity of each fraction: CV, GF, and AF. Diagram showing the results of analysis of the cytotoxicity of SVMPs. Diagram showing the results of analysis of the cleavage activity of SVMPs using APP Swedish mutant cells. 1. A diagram showing the results of analyzing the cleavage activity of SVMPs using synthetic peptides. 2. A diagram showing the results of analyzing the effect of SVMPs on Aβ fibril degradation. 3. A diagram showing the results of analyzing the effect of SVMPs on Aβ fibril formation. 4. The affinity-purified SVMPs were further purified by strong cation exchange chromatography. 5. A diagram showing the elution pattern. 6. SDS-PAGE analysis of the eluted fraction. 7. A diagram showing the results of analyzing the Aβ-degrading activity of purified SVMP. 8. A diagram showing the results of quantitative evaluation of the Aβ-degrading activity of purified SVMP. 9. A diagram showing the results of analyzing the degradation activity of SVMP against substance P. 10. A diagram showing the results of analyzing the degradation activity of SVMP against Met-enkephalin. A diagram showing the homology of the metalloprotease domains of SVMPs from the Viperidae snake (svMP01_HV1 to svMP11_NaMP: SEQ ID NOs: 18 to 28) and four species of Elapidae snakes (Cb_svMP to Nk_svMP: SEQ ID NOs: 29 to 32).

[0020] ADAMs are known as proteases that suppress the production of Aβ, which causes Alzheimer's disease. The present inventors hypothesized that snake venom metalloproteinases (SVMPs), which are major components of the venom acquired by Habu snakes during the evolutionary process, may suppress the production of Aβ because they are derived from a common ancestor with the ADAMs family proteins present in humans. They analyzed the action of SVMPs and completed the present invention.

[0021] The "venoms" of living organisms contain a variety of physiologically active substances and have been used for medical purposes. Snake venom also contains toxic substances such as cytotoxins, neurotoxins, and hematotoxins, and their potential for medical applications has long been recognized (Non-Patent Documents 5, 6). In fact, small peptides and proteases have been developed as pharmaceuticals, such as the antihypertensive drug captopril, the antiplatelet drugs tirofiban and eptifibatide, the antithrombotic peripheral circulation improver batroxobin, the hemostatic agent hemocoagulase, and the analgesic α-cobrotoxin. As mentioned above, ADAMs, which are derived from a common ancestor with SVMP, act as proteases that suppress Aβ production. Therefore, we analyzed SVMP as a potential therapeutic agent for Alzheimer's disease by targeting Aβ.

[0022] Attempts have already been made to use proteases that cleave Aβ as therapeutic agents for Alzheimer's disease, but when human-derived proteases are used, there are concerns that substrates other than Aβ may be present, which may cause side effects, and that endogenous protease inhibitors may exist. In fact, it is known that MMPs have endogenous inhibitors, TIMPs (tissue inhibitors of MMPs), and it has also been reported that TIMP-1 expression increases in the cerebrospinal fluid (CSF) of Alzheimer's disease patients (Non-Patent Document 7). Therefore, proteases that do not have endogenous inhibitors may be able to more effectively degrade Aβ and treat Alzheimer's disease.

[0023] Here, analysis was performed using the venom of Habu snakes (Protobothrops flavoviridis), but the venom of snakes from the Viperidae family is also known as hemorrhagic toxins, contains powerful digestive enzymes, and has similar toxicity. Therefore, SVMPs from the Viperidae family, other than the Habu snakes that exhibited Aβ-degrading activity below, can also be used to degrade Aβ. In addition to the Habu snake, there are many other species of Viperidae snakes, such as the Japanese pit viper and the bushmaster snake. Furthermore, snakes from the Elapidae family also possess metalloproteases homologous to Habu snake SVMPs. Therefore, it is highly likely that SVMPs from Elapidae snakes can also be used for treatment.

[0024] When SVMP is used for therapeutic purposes, the protein itself may be used, or a gene may be introduced and expressed in the brain. When an SVMP gene is introduced and expressed in the brain, any known method for gene introduction may be used, and it can be introduced using lipid nanoparticles, viral vectors, etc. As the viral vector, known vectors such as baculovirus vectors, herpesvirus vectors, lentivirus vectors, retrovirus vectors, adenovirus vectors, and adeno-associated virus (AAV) vectors can be used, but it is preferable to use an AAV vector because it can be introduced into neurons and has high gene introduction efficiency.

[0025] Pharmaceutical compositions for gene therapy contain a therapeutically effective amount of a gene therapy product and a known pharmaceutically acceptable carrier. The term "carrier" refers to a diluent, adjuvant, excipient, or vehicle, and can be, for example, a sterile liquid, such as water or oil, including oils of petroleum, animal, vegetable, or synthetic origin, such as peanut oil, soybean oil, mineral oil, and sesame oil. Water is a preferred carrier when the pharmaceutical composition is administered intravenously. Saline solutions and aqueous dextrose and glycerol solutions can also be used as liquid carriers, particularly for injectable solutions. Suitable pharmaceutical excipients include starch, glucose, lactose, sucrose, sodium stearate, glycerol monostearate, talc, sodium chloride, nonfat powdered milk, glycerol, propylene glycol, water, ethanol, and the like. Pharmaceutical compositions may also contain minor amounts of wetting or emulsifying agents, or pH buffering agents, if desired. These compositions can take the form of solutions, suspensions, emulsions, sustained-release formulations, and the like. The pharmaceutical composition is formulated in accordance with routine procedure as a pharmaceutical composition adapted for intravenous administration to human beings. Typically, compositions for intravenous administration are solutions in sterile isotonic aqueous buffer.

[0026] Any AAV vector may be used, but a neurotropic vector is preferred. Alternatively, the full-length SVMP may be used, or only the metalloprotease domain with proteolytic activity may be utilized. Furthermore, when using an SVMP gene for therapy, the SVMP gene or a gene 100% homologous to its metalloprotease domain may be used, or a modified gene with proteolytic activity may be used. Even in this case, it is preferable to maintain 80% or more, preferably 90% or more, more preferably 95% or more, and even more preferably 98% or more homology with the metalloprotease domain of SVMP. Genome analysis has revealed that hubs contain 11 types of proteases. One or more of these may be used for gene therapy. The SVMP gene contains a signal peptide and a prodomain that are removed in activated proteases. A gene expressing an activated protease without these sequences may be incorporated into an AAV vector, or the full-length gene may be incorporated. Even when the full-length gene is integrated, it is cleaved by endogenous proteases and functions as an active protease.

[0027] When a protein is used, administration by injection, infusion, or the like is preferred. Specifically, it can be administered intravenously, subcutaneously, or into bone marrow fluid. When the protein itself is used for treatment, the SVMP or a region containing the metalloprotease domain of the SVMP can be dissolved in a pharmaceutically acceptable carrier such as an aqueous carrier. As the aqueous carrier, a known pharmaceutically and physiologically acceptable solution, specifically the above-mentioned injectable solution, can be used as the liquid carrier.

[0028] As mentioned above, Alzheimer's disease is a neurodegenerative disease characterized by the accumulation of Aβ in the brain. Cleavage of the extracellular domain of amyloid precursor protein (APP) involves enzymes (BACE-1, β-secretase) that cleave at the β-site of APP, and proteases such as ADAM that cleave at the α-site. The susceptibility of Aβ to aggregation varies depending on the cleavage site, suggesting that this is involved in the pathology of Alzheimer's disease (Figure 1A). Cleavage of APP by BACE-1 and γ-secretase produces Aβ peptides of 38 to 43 amino acids, Aβ38, Aβ40, Aβ42, and Aβ43. Aβ42 and Aβ43 are thought to be involved in the formation of amyloid plaques. On the other hand, cleavage of APP by ADAM and γ-secretase produces the non-amyloidogenic p3 peptide (Non-Patent Documents 8 and 9). It is believed that treatment could be achieved by activating ADAM proteases through signal transduction pathways involved in the control of ADAM activity, thereby increasing non-amyloidogenic cleavage and preventing the production of neurotoxic Aβ42 and Aβ43, but this has not yet been applied clinically (Non-Patent Documents 10 and 11).

[0029] ADAM family proteins are transmembrane proteases with an active center in the extracellular region and are involved in protein degradation on the cell surface. As shown in Figure 1B, a typical ADAM protein has a metalloprotease (M) domain, a disintegrin (D) domain, and a cysteine-rich (C) domain, and in the membrane-bound form, it has an EGF-like (E) domain, a transmembrane region, and an intracellular domain. Note that Figure 1B shows the active form and does not show the signal peptide and prodomain that are cleaved and removed.

[0030] There are two similar protein families of ADAMs: ADAMTSs and SVMPs. In addition to the domain structure of ADAMs, ADAMTSs have a spacer (S) domain, a type 1 thrombospondin repeat (T) sequence, and various other domain structures (X). SVMPs also have a metalloprotease (M) domain, a disintegrin (D) domain, and a cysteine-rich (C) domain.

[0031] SVMPs are the major components of Viperidae snake venoms, and similar proteases are known to be present in Elapidae snakes, several mole vipers, and Colubridae snakes (Non-Patent Document 12). Proteome analysis has revealed that SVMPs account for more than 30% of the total protein content in the venoms of many Viperidae snakes, including the Habu snake. SVMPs are classified into three classes based on their domain structure: P-I, which contains only an M domain; P-II, which contains a D domain; and P-III, which contains a C domain (Figure 1B). Genomic analysis has revealed that 11 SVMPs are present in the Habu snake (Non-Patent Documents 13, 14), with the highest homology to ADAMs 7 and 28, and ADAM DEC-1 (decycin-1), which contains only a metalloprotease domain.

[0032] [Purification of SVMP] Crude Habu venom was first purified using a Sephadex G-100 column (2.8 (diameter) × 75 cm). ２ , 50 mM CH ３ COONH ３ Elution with 1000kDa solution at 0.32 mL / min resulted in three fractions (Figure 2A). Based on molecular weight, the three fractions were separated into a fraction greater than 50 kDa (G1), a fraction between 20 and 50 kDa (G2), and a fraction smaller than 20 kDa (G3). The G2 fraction, which contained a large amount of protein, was purified by metal ion affinity chromatography following the method for purifying the venom of the white-lipped viper (Pittschutz et al., 2004). Specifically, the lyophilized G2 fraction was dissolved in 1 mL of PBS, and the insoluble fraction (ppt) was removed by centrifugation. The solution was then mixed with 1 mL of His60 Ni Superflow Resin (Clontech) and incubated overnight at 4°C. The flow-through (FT), the fraction eluted with a wash solution (10 mM Tris / HCl, pH 8.0, 150 mM NaCl) (Wash), and the fraction eluted stepwise by adding 10 to 30 mM imidazole to the wash solution (10 mM, 20 mM, 30 mM, 2 mL each) (Elute) were analyzed by SDS-PAGE. Among the 1 mL eluted fractions, the third to fifth fractions (E3-E5, lanes 8-10) contained a large amount of SVMP (Fig. 2B, left).

[0033] Crude snake venom (CV), gel filtration G2 fraction (GF), and affinity-purified fraction (AF) were analyzed by non-reducing SDS-PAGE (Figure 2B, right). Based on molecular weight, it was determined that they contained a large amount of SVMPs P-I and P-III, from which the signal peptide and prodomain had been removed. From 20.3 mg of crude snake venom, 11.6 mg was recovered by gel filtration, and 1.7 mg by affinity purification.

[0034] To evaluate the proteolytic activity of SVMPs, a casein degradation assay was performed (Figure 2C). 30 mg of casein was added to 1 mL of 1 M Tris-HCl (pH 8.0) and mixed by end-over-end mixing at 4°C overnight. This solution was heated at 40°C for 10 minutes to completely dissolve the casein. The casein solution was diluted and each snake venom-containing fraction was added to prepare the assay solution (total volume: 250 μL, 3 mg / mL casein, 10 μg / mL enzyme solution, 100 mM Tris-HCl (pH 8.0)). The reaction was initiated by adding 25 μL of enzyme solution, i.e., crude venom fraction (CV), G2 fraction (GF), or AF fraction (100 μg / mL protein). The reaction was incubated at 37°C for 2.5 hours and stopped with 250 μL of 10% TCA. After incubation on ice for 20 minutes, the supernatant was centrifuged at 20,600 × g for 10 minutes. The absorbance (A280) of the supernatant was measured. An increase in absorbance of 1.0 per minute was defined as 1 unit of caseinolytic activity, and the caseinolytic protease activity of the CV, GF, and AF fractions was measured. The activity of each fraction was significantly reduced in the presence of EDTA, which inhibits the metal ions present in the active center of SVMPs by chelating them. This indicated that the protease activity in the fractions was mediated by SVMPs. The specific activity of the AF fraction was significantly increased compared with the GF and CV fractions. The specific activity of the AF fraction was approximately 1.5-fold higher than that of the GF fraction and approximately 2-fold higher than that of the crude venom (CV). The recovery of SVMP activity in the AF fraction was 22.8% of that of the GF fraction and 16.7% of that of the crude venom. Data shown in Figure 2C are means ± standard deviation; n = 3. In the following data, ****: p<0.0001, ***: <0.001, **: <0.01, *: <0.05, n.s.: >0.05.

[0035] Mass spectrometry (LC-MS / MS) analysis of the AF fraction identified nine of the 11 SVMPs present in the Habu snake genome (Non-Patent Documents 13, 14) (Table 1). The purified sample contained SVMPs classified as P-I, P-II, and P-III. Two SVMPs, NaMP and the precursor to elegantin, were not identified by LC-MS / MS, despite the presence of their genes. These two SVMPs were also not detected in proteomic analysis of crude Habu snake venom. In the following experiments, the AF fraction of nine SVMPs identified in snake venom, purified by gel filtration and affinity chromatography, was used as a cocktail (SVMPs).

[0036]

[0037] The molecular weight and base length in Table 1 are based on sequence information in the database. SVMP has a signal peptide and propeptide at the amino terminus (1-90), which are cleaved during maturation, but the values ​​of molecular weight, base length, and sequence coverage in the table, as well as the amino acid sequence shown in the sequence listing, include these regions. Protein ID: svMP08_flavoridin represented by SEQ ID NO: 1 represents a flavoridin precursor having cleavage activity, Protein ID: svMP06_H2pro represented by SEQ ID NO: 2 represents H2pro, Protein ID: svMP09_HR1b represents HR1b, Protein ID: svMP07_HR2a represented by SEQ ID NO: 4 represents HR2a, Protein ID: svMP05_HR1a represented by SEQ ID NO: 5 represents HR1a, Protein ID: svMP02_flavorase represented by SEQ ID NO: 6 represents flavorase, Protein ID: svMP03_VMP-III_like represented by SEQ ID NO: 7 represents VMP-III_like, and Protein ID: svMP04_VMP-III_like represented by SEQ ID NO: 8 represents ID: svMP01_HV1 indicates HV1, and Protein ID: svMP04_jerdonitin-like indicates a jerdonitin-like precursor represented by SEQ ID NO: 9. Both amino acid sequences represented by SEQ ID NOs are proteins containing a protease domain.

[0038] [Effect of SVMPs on Aβ Production] Aβ production assay of SVMPs was performed using cultured cells. HEK293 cells (human embryonic kidney cell line, obtained from ATCC) were transfected with the Swedish mutant APP (a mutant with K670N and M671L mutations in APP, see Non-Patent Document 16) and stably expressed in these cells (hereinafter referred to as NLNTK cells). SVMPs were added to the culture medium of NLNTK cells, and cell viability was measured. Cell viability was evaluated using Cell Counting Kit-8 (Dojindo Laboratories). Cell viability was comparable to that of the control when 2, 4, or 8 μg / mL SVMPs were added to the medium. However, viability decreased to approximately 80% at 16 μg / mL SVMPs and to less than 40% at 32 μg / mL SVMPs (Figure 3A). IC ５０ The concentration of SVMPs was 24.8 μg / mL (95% confidence interval: 20.6-29.9 μg / mL). Observation by phase contrast microscope revealed many floating cells at 16 μg / mL and 32 μg / mL SVMPs concentrations, but almost no floating cells were observed at 2-8 μg / mL SVMPs concentrations. Based on these results, subsequent experiments were performed at an SVMP concentration of 5 μg / mL.

[0039] NLNTK cells were cultured for 48 hours in medium supplemented with 5 μg / mL SVMPs, and secreted Aβ and intracellular APP fragments (C99 or C83) were analyzed by immunoblotting (Figure 3B, top). Aβ and APP fragments were detected with a monoclonal antibody (82E1) against the N-terminus of Aβ, and tubulin was detected with an anti-tubulin monoclonal antibody (DM1A) as a control (both obtained from Immuno-Biological Laboratories, Inc.). Visualization was performed using the ECL system, and quantification was performed using a lumino-image analyzer LAS-4000 (Fujifilm). SVMPs reduced Aβ levels by 10% compared with the control (no SVMPs) (Figure 3B, bottom left). On the other hand, APP P-terminal fragments (C99 and C83) were unaffected by SVMPs (Figure 3B, bottom center and right). These results suggest that SVMPs degrade secreted Aβ peptides but not APP.

[0040] [SVMP Cleavage Site] SVMPs directly act on and cleave Aβ, so their cleavage activity was analyzed in vitro using synthetic Aβ peptides (Peptide Institute, Inc.). 10 nM Aβ40 or Aβ42 synthetic peptides, 60 μg / mL SVMPs, and 100 mM Tris-HCl (pH 8.0) were incubated at 37°C for 2.5 hours in the presence or absence of 5 mM EDTA, and Aβ levels were analyzed by immunoblotting (Figure 4, left). SVMPs significantly reduced Aβ40 and Aβ42 levels compared with the control (no SVMP, Figure 4, left lanes 1-3) (Figure 4, left lanes 7-9, Figure 4, right). Furthermore, the proteolytic activity of SVMPs was significantly reduced by the addition of EDTA (Figure 4, left lanes 10-12, Figure 4, right). These results indicate that SVMP can degrade Aβ40 and Aβ42 in vitro.

[0041] LC-MS / MS was performed to identify the cleavage site of Aβ by SVMP. Peptides were reacted with SVMPs under the following conditions, and the cleavage site was identified by LC-MS / MS. 0.1 mg / mL Aβ40 or Aβ42 synthetic peptide was mixed with 5.6 μg / mL SVMPs in 100 mM Tris-HCl (pH 8.0), and the mixture was incubated at 37°C for 24 hours. After the reaction, the mixture was analyzed by LC-MS / MS.

[0042] The amino-terminal fragments of the identified Aβ peptides are shown in Table 2 (Aβ40) and Table 3 (Aβ42). LC-MS / MS analysis revealed that the cleavage sites of Aβ40 / 42 were Lys16-Leu17, Val18-Phe19, and Ala30-Ile31. The Lys16-Leu17 cleavage site of Aβ corresponds to Lys687-Leu688 of APP, which is the α-cleavage site at which APP is cleaved by ADAM9, 10, and 17. In other words, SVMP cleaves Aβ at the same cleavage site as ADAM. LC-MS / MS analysis revealed that 99.5% of Aβ40 peptides and 80.6% of Aβ42 peptides were cleaved at the α-site, indicating that the α-site is the major cleavage site of SVMP.

[0043] The full-length sequence of Aβ40 is as follows: Aβ40: DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVV (SEQ ID NO: 14)

[0044] The full-length sequence of Aβ42 is as follows: Aβ42: DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVIA (SEQ ID NO: 15)

[0045] [Effect of SVMPs on Aβ fibrils] The effect of SVMPs on Aβ fibrils was analyzed. Synthesized Aβ40 and Aβ42 were allowed to form Aβ fibrils in Tris-buffered saline (100 mM Tris-HCl (pH 8.0), 50 mM NaCl). SVMPs were then added, and the mixture was incubated for 24 hours, after which analysis was performed to determine whether scission occurred. Aβ fibril scission was detected by measuring the fluorescence intensity of the dye thioflavin T (ThT), which specifically binds to amyloid fibrils and emits fluorescence. To analyze the ability to degrade Aβ fibrils, SVMPs were added before or after fibril formation, and the mixture was incubated for 24 hours, after which a ThT assay was performed using a control solution without SVMPs.

[0046] First, we analyzed whether SVMPs degrade Aβ40 or Aβ42 fibrils. 5 μL of a mixture (100 mM Tris-HCl (pH 8.0), 50 mM NaCl) of SVMPs and 0.1 mg / mL of generated Aβ40 or Aβ42 fibrils was dispensed into a 96-well plate immediately after mixing and 24 hours later. This mixture was mixed with 200 μL of ThT solution (5 μM ThT, 100 mM glycine (pH 7.5)). The mixture was incubated in the dark at room temperature for 5 minutes, and fluorescence was measured. ThT fluorescence was analyzed using a Gemini XPS microplate spectrofluorometer (Molecular Devices) in endpoint mode with an excitation wavelength of 440 nm and an emission wavelength of 480 nm. Data analysis was performed using SoftMax Pro Software v5.4.1 (Molecular Devices). A ThT solution containing no Aβ was used as a control, and the fluorescence intensity was calculated by subtracting the control value. The fluorescence intensity of ThT from both Aβ40 and Aβ42 fibrils was unaffected by SVMP, indicating that SVMP does not degrade Aβ40 or Aβ42 after amyloid fibril formation (Figure 5A).

[0047] Next, we analyzed whether SVMP inhibits Aβ amyloid formation (Figure 5B). When SVMPs were added to the Aβ fibrillation solution and incubated for 24 hours, ThT fluorescence increased approximately 5.4-fold in the control, but the increase in ThT fluorescence was only 2.3-fold with the addition of SVMPs, indicating significant suppression of fibrillation. On the other hand, in the presence of EDTA, ThT fluorescence intensity increased 3.3-fold, indicating a decrease in the inhibition by SVMPs. Therefore, it is believed that SVMP inhibits amyloid fibril formation by degrading Aβ40.

[0048] As described above, SVMP cleaves Aβ at the α site. Aβ cleaved at the α site is cleaved by γ-secretase into p3 peptide and the APP intracellular domain (AICD), and is incapable of forming Aβ fibrils (see Figure 1A). Although SVMP does not cleave formed Aβ fibrils, it may be possible to suppress the progression of Alzheimer's disease by inhibiting the formation of Aβ fibrils. As described above, in addition to the SVMPs represented by SEQ ID NOS: 1 to 9, SVMPs represented by SEQ ID NOS: 26 (elegantin precursor) and 27 (NaMP) have been identified by genome analysis. Because these SVMPs also contain a metalloprotease domain, they can be introduced by viral vectors such as AAV and used for treatment. Note that the elegantin represented by SEQ ID NOS: 26 is an elegantin precursor with cleavage activity.

[0049] The SVMP used in the above examples was a cocktail containing the enzymes represented by SEQ ID NOS: 1-9. Flavoridin precursor (SEQ ID NOS: 1) and H2pro (SEQ ID NOS: 2) are proteases abundant in crude snake venom, but their Aβ degradation activity is unknown. Therefore, further purification and analysis were performed. As described above, the SVMP cocktail containing nine SVMPs purified by gel filtration and metal ion affinity chromatography was purified using strong cation exchange chromatography and a ResourceS column (1 mL resin). The concentrated SVMP fraction was applied at 1.0 mL / min to bind to the column, washed with 20 mM MES / HEPES (pH 7.2), and eluted with increasing NaCl concentrations. Six peak fractions were identified (Figure 6). The flow-through fraction (FT), which did not bind to the resin, the fraction eluted by washing with a wash solution (20 mM MES / HEPES, pH 7.2, 0 M NaCl) (Wash), and the fraction eluted by adding 0 to 1 M NaCl to the wash solution (0 M to 0.5 M, 1 mL gradient elution, followed by a 1 M step elution) (Elute) were analyzed by SDS-PAGE (Figure 7). Seven fractions, Nos. 2 to 7, including the FT fraction (No. 1), were obtained.

[0050] To evaluate the Aβ-degrading activity of purified SVMPs, we analyzed their cleavage activity in vitro using synthetic Aβ peptides as described above. After incubation at 37°C for 2.5 hours, Aβ levels were analyzed by immunoblotting. SVMP fractions 3, 4, 5, and 6 significantly reduced Aβ40 levels compared with the control (no SVMP, Fig. 8A, upper left, lanes 1–3) (Fig. 8A, upper right, lanes 8–9, lower left, lanes 1–3, lower left, lanes 7–9, lower right, lanes 1–3).

[0051] Furthermore, the proteolytic activity of these fractions was significantly reduced by the addition of EDTA (Figure 8B). Therefore, it was confirmed that the Aβ-degrading activity was due to SVMP. The degrading activity, compared with the activity after EDTA (100%), ranged from 43.9% to 78.0%, indicating that all of the SVMPs in the fractions were capable of degrading Aβ40 (Table 4).

[0052] The SVMP fractions were analyzed by mass spectrometry (LC-MS / MS) (Table 4). Fraction No. 6, which had the highest Aβ40 cleavage activity, was identified as a flavoridin precursor among the 11 types of SVMP present in the hub genome. It is believed that the flavoridin precursor acts as a protease with high Aβ40 cleavage activity. Since all fractions No. 2 to No. 6 had Aβ-degrading activity, it goes without saying that other SVMPs could also be potential therapeutic candidates. In particular, H2 protease (H2pro) has no hemorrhagic activity, is non-toxic, and is therefore expected to be a potential therapeutic candidate.

[0053]

[0054] Cleavage of neuropeptides when expressed in the brain may result in side effects. Therefore, the No. 6 fraction, which had high Aβ degradation activity, was used to analyze its neuropeptide degradation activity. Synthetic substance P (SEQ ID NO: 37) and Met-enkephalin (SEQ ID NO: 38) (both from Peptide Institute, Inc.) were used to perform an in vitro cleavage reaction, and LC-MS / MS was performed to identify the cleavage site by SVMP. The peptide was reacted with SVMPs under the following conditions, and the cleavage site was identified by LC-MS / MS. 0.1 mg / mL substance P or Met-enkephalin synthetic peptide was mixed with 10 μg / mL SVMP No. 6 fraction (flavoridin precursor) in 100 mM Tris-HCl (pH 8.0), and the mixture was incubated at 37°C for 24 hours. After the reaction, analysis was performed by LC-MS / MS. Although no cleavage activity was observed for Met-enkephalin, it was revealed that there is one cleavage site for substance P (Fig. 9A, B). The effect of cleavage activity on substance P needs further investigation, but if adverse effects occur, it may be possible to take necessary measures, such as supplementing substance P.

[0055] When used alone as a therapeutic agent, a highly active flavoridin precursor or its active domain can be a promising candidate.Since the nucleic acid sequence (SEQ ID NO: 39) for expressing the flavoridin precursor is publicly known, a protein expressed using this can be used as an active ingredient, or can be introduced into the body using an AAV vector or the like.In addition, a protein expressed using the protease domain shown in SEQ ID NO: 24 or the region from base 571 to base 1233 of the base sequence shown in SEQ ID NO: 37, which is the nucleic acid sequence encoding this, can be used, or this region can be expressed in the body using a vector such as AAV.In addition, although further analysis is required, even if the content in crude snake venom is low, SVMPs with useful functions, such as SVMPs with high protease activity or those without cleavage activity against neuropeptides, or regions containing their metalloprotease domains, can also be used for treatment.

[0056] Figure 10 shows the amino acid sequence homology of the M domain of SVMP (SEQ ID NOS: 18-32). It shows the amino acid sequences of 11 SVMPs found in Habu snakes, as well as those of Calliophis bivirgatus, Pseudonaja textilis, Naja naja (Indian cobra), and Naja kaouthia (Siberian cobra) from the Elapidae family. The M domains of these SVMPs are highly conserved, and Elapidae SVMPs share a high sequence homology with the ancestral Habu snake svNP11_NaMP. Only the M domain shown in Figure 10 may be expressed in cells, or the full-length sequence of each SVMP (hub SVMPs: SEQ ID NOS: 1-9, 16, and 17; C. bivirgatus: SEQ ID NO: 33; P. textilis: SEQ ID NO: 34; N. naja: SEQ ID NO: 35; N. kaouthia: SEQ ID NO: 36) may be expressed. Based on the sequence information of the M domain, only the region necessary for protease activity may be expressed, or mutants may be prepared by introducing amino acid mutations that increase activity or mutations that do not cleave neuropeptides, and these may be used for treatment.

Claims

1. An agent for inhibiting the formation of amyloid fibrils, the active ingredient of which is a region including at least the metalloprotease domain of snake venom metalloprotease SVMP (snake venom metalloproteinase).

2. The amyloid fibril formation inhibitor according to claim 1, characterized in that the snake is a snake of the Viperidae family or the Elapidae family.

3. The amyloid fibril formation inhibitor according to claim 2, characterized in that the snake of the Viperidae family is a Habu, and the SVMP is at least one of HV1, flavorase, VMP-III_like, jerdonitin-like precursor, HR1a, H2pro, HR2a, flavoridin precursor, HR1b, elegantin precursor, and NaMP.

4. The amyloid fibril formation inhibitor according to claim 3, wherein the SVMP is a flavoridin precursor.

5. A pharmaceutical composition for treating Alzheimer's disease, comprising as an active ingredient the amyloid fibril formation inhibitor according to any one of claims 1 to 4.

6. A pharmaceutical composition comprising a gene therapy product for treating Alzheimer's disease, wherein said gene therapy product is a nucleic acid sequence encoding a region of a snake venom SVMP comprising at least the metalloprotease domain.

7. The pharmaceutical composition of claim 6, wherein said gene therapy product comprises a viral vector.

8. The pharmaceutical composition according to claim 7, characterized in that the viral vector is an adeno-associated virus (AAV).

9. The pharmaceutical composition according to any one of claims 6 to 8, wherein the SVMP is a gene encoding at least one of the amino acid sequences set forth in SEQ ID NOs: 1 to 9, 16 and 17.

10. The pharmaceutical composition according to claim 9, wherein the SVMP is a peptide represented by SEQ ID NO:1.

Images