JP7289510B2 - Dementia drug screening method - Google Patents

Description

TECHNICAL FIELD The present invention relates to a method for screening therapeutic agents for dementia that do not have serious side effects, and therapeutic agents for dementia.

The number of people with dementia worldwide was 47 million as of 2016 and is projected to well exceed 100 million by 2050. Therefore, research institutes and pharmaceutical companies around the world are working on the development of dementia therapeutic drugs, but they have not yet led to the development of fundamental therapeutic drugs.

Alzheimer's disease (AD) drugs currently approved are modulators of neurotransmitters such as acetylcholine, and since they are only symptomatic treatments, their effects are temporary and cannot prevent the progression of dementia. Therefore, drugs that can fundamentally treat dementia are being searched for.

The following hypotheses have been put forward as the cause of Alzheimer's disease. First, when amyloid β (Aβ) accumulates outside nerve cells in the brain and forms senile plaques, tau protein is phosphorylated and aggregated to cause neurofibrillary tangles. Next, it is believed that oligomers and neurofibrillary tangles generated in the process of Aβ accumulation induce neuronal dysfunction and cell death. Amyloid β is produced by cleavage of amyloid protein precursor (APP) by β-secretase to generate APP carboxy-terminal fragment (β-CTF), followed by cleavage of β-CTF by γ-secretase, resulting in oligomers. known to turn into Therefore, inhibitors of γ-secretase and β-secretase have been investigated as therapeutic agents for Alzheimer's disease (Patent Document 1, etc.).

However, since secretase has an important function of degrading various proteins in vivo, administration of secretase inhibitors has been confirmed to cause serious side effects such as skin cancer and liver damage. Therefore, secretase inhibitors have many problems as therapeutic agents for dementia.

International Publication No. 2013/108780 pamphlet

As described above, the current situation is that a true therapeutic drug for dementia that is free of serious side effects has not yet been developed. Accordingly, an object of the present invention is to provide a method for screening for a fundamental therapeutic agent for dementia that does not have serious side effects, and a therapeutic agent for dementia.

The present inventors have made intensive studies to solve the above problems. As a result, we identified NRBP1 as a substrate receptor for Cullin-RING ubiquitin ligase (CRL), which degrades BRI2 and BRI3, which inhibits amyloid precursor protein processing and oligomerization of amyloid β. As an indicator, the inventors have found that it is possible to screen dementia therapeutic agents without serious side effects, and have completed the present invention.
The present invention is shown below.

[1] A method for screening a therapeutic agent for dementia, comprising:
reacting BRI2 and/or BRI3 with NRBP1 in the presence or absence of a test compound;
comparing the amount of BRI2 and/or BRI3-NRBP1 complex in the presence and absence of the test compound; and
identifying a test compound for which said amount in the presence of the test compound is less than said amount in the absence of the test compound.
[2] The method of [1] above, further comprising the step of dimerizing NRBP1 with TSC22D4.
[3] The method of [1] or [2] above, further comprising the step of dimerizing NRBP1 with TSC22D3.
[4] The method according to any one of [1] to [3] above, wherein the dementia is Alzheimer's disease.
[5] The method according to any one of [1] to [3] above, wherein the dementia is familial British dementia and/or familial Danish dementia.
[6] A therapeutic agent for dementia comprising, as an active ingredient, a compound that inhibits the expression or function of one or more proteins selected from the group consisting of NRBP1, TSC22D3, and TSC22D4.
[7] The therapeutic agent for dementia according to [6] above, wherein the compound knocks down a gene selected from the group consisting of NRBP1 gene, TSC22D3 gene and TSC22D4 gene.
[8] The therapeutic agent for dementia according to [7] above, wherein the compound is siRNA of a gene selected from the group consisting of NRBP1 gene, TSC22D3 gene and TSC22D4 gene.

Conventional therapeutic drugs for dementia are modulators of neurotransmitters such as acetylcholine, and therefore cannot fundamentally stop the progression of symptoms, or secretase inhibitors that exhibit various functions in vivo. Therefore, serious side effects have become a problem.
In contrast, according to the method of the present invention, compounds that inhibit the degradation of BRI2 and BRI3, which inhibit the oligomerization and aggregation of amyloid β, which is thought to cause Alzheimer's disease, can be screened as therapeutic agents for dementia. Therefore, the antidementia drug found by the method of the present invention can directly inhibit the oligomerization and aggregation of amyloid β, and is therefore considered to exhibit no or almost no serious side effects. Theoretically, the drug for treating dementia found by the method of the present invention can also be a drug for treating not only Alzheimer's disease but also familial British dementia and familial Danish dementia.
Therefore, it can be said that the present invention has a very high industrial value as it can contribute to the treatment of dementia, for which the number of patients is expected to increase sharply in the future.

FIG. 1 is an immunoblot image showing the results of an experiment to clarify the relationship between the interaction between NRBP1 and Cul2 and TSC22DF. FIG. 2 is an immunoblot image showing the results of an experiment to clarify the relationship between NRBP1 dimerization and TSC22DF. FIG. 3 is an immunoblot image showing the results of experiments to clarify the relationship between NRBP1 and Cul2 stability and TSC22DF. FIG. 4 is a schematic diagram showing the principle of an experiment for searching substrates for NRBP1-ubiquitin ligase. FIG. 5 is an immunoblot image showing the results of experiments to demonstrate that BRI3 is a substrate for NRBP1-ubiquitin ligase. FIG. 6 is an immunoblot image showing the results of experiments to demonstrate that BRI3 is a substrate for NRBP1-ubiquitin ligase. FIG. 7 is an immunoblot image showing the results of experiments to demonstrate that BRI2 is a substrate for NRBP1-ubiquitin ligase. FIG. 8 is an immunoblot image showing the results of experiments to demonstrate that BRI2 is a substrate for NRBP1-ubiquitin ligase. FIG. 9 is an immunoblot image showing the results of an experiment to determine whether NRBP1 affects the stability of BRI2 and BRI3 in cells. FIG. 10 is an immunoblot image showing the results of an experiment to determine whether NRBP1 affects the stability of BRI2 and BRI3 in cells. FIG. 11 is an immunoblot image showing the results of experiments to examine the involvement of Cul2 and Cul4A in the ubiquitination of BRI2 and BRI3 by NRBP1-ubiquitin ligase in the presence of exogenously expressed NRBP1. FIG. 12 is an immunoblot image showing the results of experiments to examine the involvement of Cul2 and Cul4A in the ubiquitination of BRI2 and BRI3 by NRBP1-ubiquitin ligase in the presence of exogenously expressed NRBP1. FIG. 13 is an immunoblot image showing the results of experiments to examine the involvement of Cul2 and Cul4A in the ubiquitination of BRI2 and BRI3 by NRBP1-ubiquitin ligase in the absence of exogenously expressed NRBP1. FIG. 14 is an immunoblot image showing the results of experiments to examine the involvement of Cul2 and Cul4A in the ubiquitination of BRI2 and BRI3 by NRBP1-ubiquitin ligase in the absence of exogenously expressed NRBP1. FIG. 15 is an immunoblot image showing the results of an experiment to clarify the relationship between the interaction between NRBP1 and Cul4A and TSC22D3/TSC22D4. FIG. 16 is an immunoblot image showing the results of experiments to identify amino acid sequences important for dimerization of NRBP1. FIG. 17 is an immunoblot image showing the results of an experiment to determine whether NRBP1 dimers bind Cul2 and Cul4A to form a dimeric E3 ubiquitin ligase complex containing Cul2/Cul4A. FIG. 18 shows the amino acid sequences of NRBP1 mutants used in experiments to identify the amino acid sequences of NRBP1 that are important for binding to Cul2 or Cul4A. FIG. 19 is an immunoblot image showing the results of an experiment to investigate the interaction between each mutant of monomeric NRBP1 and Cul2 or Cul4A. FIG. 20 is an immunoblot image showing the ubiquitination tendency of BRI2 and BRI3 when wild-type NRBP1 or each mutant NRBP1 is co-expressed. FIG. 21 is a schematic representation of a Cullin-RING type E3 ubiquitin ligase complex containing NRBP1 dimers. FIG. 22 is an immunoblot image showing the binding ability of wild-type NRBP1 or monomeric NRBP1 to BRI2 and BRI3. FIG. 23 is an immunoblot image showing the results of an experiment to identify amino acid sequences important for binding to BRI3 in NRBP1. FIG. 24 is an immunoblot image showing the results of experiments to identify amino acid sequences important for binding to NRBP1 in BRI3. FIG. 25 is an immunoblot image showing the results of experiments to identify amino acid sequences important for binding to NRBP1 in BRI2 and BRI3. FIG. 26 is an immunoblot image showing changes in the amount of endogenous BRI2 and BRI3 by gene knockdown of NRBP1, TSC22D3 or TSC22D4. FIG. 27 is a graph showing changes in the amount of amyloid β due to gene knockdown of NRBP1, TSC22D3 or TSC22D4.

The method of the present invention will be explained step by step below.
1. NRBP1 Dimerization Step In this step, TSC22D4, TSC22D3, or a combination of TSC22D4 and TSC22D3 promotes dimerization of NRBP1. Implementation of this step is optional, and even NRBP1 monomers exhibit binding ability to BRI2 and/or BRI3, but NRBP1 dimers are more stable than NRBP1 monomers, so this step is performed. preferably.

Although TSC22D4 is known as a transcriptional repressor, our experimental findings show that it significantly increases the stability of NRBP1, promotes its dimerization, and prevents NRBP1 dimers from being degraded. Protect.

TSC22D3 protects T cells from IL2 deficiency-induced apoptosis through inhibition of FOXO3A transcriptional activity leading to downregulation of the pro-apoptotic factor BCL2L11. In macrophages, it plays an important role in the anti-inflammatory and immunosuppressive effects of glucocorticoids and IL10. According to our experimental findings, TSC22D3 induces a conformational change in NRBP1 to increase its affinity for Cul2 and Cul4A, which forms a Cullin-RING type E3 ubiquitin ligase complex with the NRBP1 dimer, Stabilizes the interaction of NRBP1 with Cul2 and Cul4A.

As described above, since TSC22D4 and TSC22D3 have different mechanisms of action on NRBP1, it is preferable to use TSC22D4 and TSC22D3 in combination.

The Cullin-RING type E3 ubiquitin ligase complex is composed of the scaffold protein Cullin, as well as the Rbx protein, adapter protein, and substrate receptor that have E2-binding properties. Eight types of Cullin subtypes, Cull, CuI2, CuI3, CuI4A, CuI4B, CuI5, CuI7, and PARC, have been reported so far in humans. Among these, Cul2 and Cul5 use Elongin BC as an adapter protein and the BC-box protein as a substrate receptor. On the other hand, Cul4A and Cul4B have been shown to use DDB1 as an adapter protein and an H-box protein as a substrate receptor.

NRBP1 (Nuclear Receptor Binding Protein 1) is a protein believed to be involved in intracellular transport between the endoplasmic reticulum and the Golgi apparatus, possibly through interaction with Rho-type GTPase. NRBP1 dimerizes through a LisH motif at its C-terminus. The inventors found that NRBP1 dimerizes and further associates with Cu12 and Cu14 to form a Cullin-RING type E3 ubiquitin ligase complex. Therefore, this step is preferably carried out in order to further promote the dimerization of NRBP1.

In this step, TSC22D4, or TSC22D4 and TSC22D3 may only be added to NRBP1, but the NRBP1 gene and TSC22D4 gene or TSC22D3 gene and TSC22D4 gene may be transformed into cells and co-expressed. The cells used here are not particularly limited as long as they are human-derived cells commonly used in transformation experiments, but for example, HeLa cells and HEK293T cells can be used.

2. Interaction Step In this step, BRI2 and/or BRI3 and NRBP1 are reacted in the presence or absence of a test compound. Hereinafter, "and/or" may be simply abbreviated as "/".

The test compound is not particularly limited as long as it is a compound to be tested for its therapeutic effect on dementia. Water-soluble compounds are preferable from the viewpoint of ease of testing, but in the case of compounds that are not sufficiently water-soluble, a water-miscible solvent such as ethanol may be added to water as the solvent within a range that does not inhibit the reaction. Alternatively, a surfactant may be used in combination.

The concentration of the test compound in the reaction solution is preferably adjusted as appropriate because the appropriate concentration of the test compound is unknown and also depends on the concentrations of NRBP1 and BRI2/BRI3, but for example, about 10 μM or more and 150 μM or less. should be adjusted within the range of

The BRI family belongs to the type II transmembrane protein family and includes BRI1-3. BRI2 shows systemic expression, whereas BRI1 and BRI3 have localized expression, and BRI3 is mainly expressed in the central nervous system, whereas BRI1 is expressed in bone and cartilage. BRI2 and BRI3 also have the ability to bind to the amyloid precursor protein (APP), whereby both proteins inhibit the cleavage of the N-terminal fragment of APP by β-secretase, and BRI2 binds to APP by γ-secretase. inhibits cleavage of the C-terminal fragment of Furthermore, BRI2 and BRI3 inhibit the oligomerization of amyloid-β.

The present inventors found that dimers of NRBP1 form a Cullin-RING type E3 ubiquitin ligase complex and function as substrate receptors for BRI2 and BRI3. The ubiquitin-proteasome system is a substrate-specific proteolytic system in which target proteins to which polyubiquitin chains are bound are degraded by proteasomes. Therefore, a compound that inhibits the binding of NRBP1 to BRI2 or BRI3 inhibits the formation of amyloid β and can be used as a therapeutic drug for dementia caused by amyloid β oligomers and aggregates.

Conditions for interaction between NRBP1 and BRI2/BRI3 are not particularly limited, and may be adjusted as appropriate. For example, a test compound may or may not be added to a solution containing NRBP1 and BRI2/BRI3, and incubated at 20° C. or higher and 40° C. or lower for about 30 minutes or longer and 50 hours or shorter. Also, NRBP1 may be a monomer or a dimer, but the NRBP1 dimer is preferred due to its higher stability.

Alternatively, the solution containing NRBP1 and BRI2/BRI3 may be a solution obtained by transforming cells with the NRBP1 gene and BRI2/BRI3 gene, culturing the cells, homogenizing the cells, and then removing insoluble components. In addition, for example, the NRBP1 dimerization step may be performed simultaneously with the interaction step by introducing TSC22D4 or TSC22D3 and TSC22D4 genes into the cells.

3. Comparison Step In this step, the amount of the BRI2/BRI3-NRBP1 complex formed in the interaction step is compared in the presence of the test compound and in the absence of the test compound.

The amount of BRI2/BRI3-NRBP1 complex may be measured qualitatively or quantitatively by a conventional method, preferably quantitatively. For example, the formation of a complex between NRBP1 and BRI2/BRI3 ^can be confirmed by immunoprecipitation, and the amount can be qualitatively compared by the intensity of the complex band by immunoblotting ^. It is preferable to use a system that can quantitatively evaluate the interaction between proteins based on the luminescence intensity.

For example, in NanoBit ^(R) , a gene encoding a protein in which one of NRBP1 and BRI2/BRI3 is fused with a luciferase-based Large Bit, and a gene encoding a protein in which the other is fused with Small Bit. Transform the cells with and culture the transformed cells in the presence or absence of the test compound. When NRBP1 and BRI2/BRI3 interact, Large Bit and Small Bit also interact to form a luciferase, so the interaction between NRBP1 and BRI2/BRI3 can be quantified as luminescence intensity. NanoBit ^(R) has the advantage that it can be easily measured using a general luminometer.

In NanoBRET ^(R) , cells are transformed with a gene encoding a protein in which one of NRBP1 and BRI2/BRI3 is luciferase as an energy transfer donor and the other is a fluorescent ligand-labeled protein as an energy transfer acceptor. Transformed cells are cultured in the presence or absence of the test compound. When NRBP1 and BRI2/BRI3 interact, the luciferase and the fluorescent ligand-labeled protein approach each other, BRET (Bioluminescence Resonance Energy Transfer) occurs, and a fluorescence signal is emitted. The interaction of NRBP1 and BRI2/BRI3 can be quantified as the intensity of this fluorescent signal.

4. Identifying step In this step, the amount of the BRI2/BRI3-NRBP1 complex in the presence of the test compound is less than the amount of the BRI2/BRI3-NRBP1 complex in the absence of the test compound, Identify as a dementia drug.

As described above, in the Cullin-RING type E3 ubiquitin ligase complex, the NRBP1 dimer acts as a receptor for degradation target proteins. By inhibiting the interaction of BRI3, the production of amyloid-β from the amyloid-β precursor protein is inhibited, and by inhibiting the oligomerization of amyloid-β, the progression of dementia associated with amyloid-β is fundamentally suppressed. is considered to hinder In addition, the therapeutic agent for dementia according to the present invention does not inhibit the action of in vivo enzymes, but inhibits the decomposition of BRI2/BRI3 that negatively regulates the progression of dementia associated with amyloid β. is not expected to cause serious side effects.

Two mutations in the BRI2 gene have also been identified as the cause of familial British dementia (FBD) and familial Danish dementia (FDD), in which reduction of normal BRI2 protein leads to increased amyloid-β levels. It is known that its production and oligomerization are accelerated, resulting in symptoms resembling Alzheimer's disease. Therefore, the therapeutic drug for dementia found by the method of the present invention is considered to be effective not only for Alzheimer's disease but also for familial British dementia and familial Danish dementia.

In addition, it is preferable to conduct a secondary evaluation as to whether or not the compound identified as the therapeutic agent for dementia in this step actually suppresses the ubiquitination of BRI2/BRI3 by NRBP1 ubiquitin ligase. Such a step increases the likelihood that the identified compound will actually show efficacy as a therapeutic agent for dementia. For example, whether ubiquitination of BRI2/BRI3 by NRBP1-ubiquitin ligase is actually inhibited in the presence or absence of the identified compounds can be determined by deubiquitination of ubiquitinated BRI2/BRI3 and proteasome It is confirmed by immunoprecipitation-western blot, or western blot alone in the presence of TR-TUBE (Trypsin-Resistant Tandem Ubiquitin-Binding Entity) which protects against degradation by cytotoxicity.

As described above, according to the experimental findings of the present inventors, dimerization of NRBP1 and formation of the Cullin-RING type E3 ubiquitin ligase complex promote amyloid β production, Promotes merization and E3 complex formation. Therefore, compounds that inhibit the expression or function of one or more proteins selected from the group consisting of NRBP1, TSC22D3 and TSC22D4 are considered effective as active ingredients of therapeutic agents for dementia.

Examples of the above compounds include those that knock down a gene selected from the group consisting of NRBP1 gene, TSC22D3 gene and TSC22D4 gene. Further examples include siRNA of genes selected from the group consisting of NRBP1 gene, TSC22D3 gene and TSC22D4 gene.

siRNA is a double-stranded nucleic acid consisting of an antisense strand having a base sequence complementary to the target mRNA and a sense strand complementary to the antisense strand, and is usually a short-chain RNA of about 21 to 25 mers. The end has a structure in which two bases TT protrude. The siRNA binds to the target mRNA and cleaves the target mRNA by causing RNA interference. For example, the following siRNA can be mentioned.

siRNA1: siRNA containing an antisense strand having one or more nucleotide sequences selected from SEQ ID NOs: 1 to 12 and a TT sequence;

siRNA2: In the base sequence defined in the above siRNA1, 1 or 2 bases have deleted, substituted and/or added bases, and the siRNA1 has affinity for the NRBP1 gene, TSC22D3 gene or TSC22D4 gene. equivalent or higher siRNA compared to;
siRNA3: having a nucleotide sequence with 90% or more sequence identity to the nucleotide sequence defined in the above siRNA1, and having an affinity for the NRBP1 gene, TSC22D3 gene or TSC22D4 gene that is equal to or higher than that of siRNA1 siRNA.

In the above siRNA2, the number of deletions, substitutions and/or additions, that is, the number of mutations is preferably one. In the siRNA3, the sequence identity is preferably 95% or more, more preferably 98% or more or 99% or more, and even more preferably 99.5% or more.

Although the form of the therapeutic agent for dementia according to the present invention is not particularly limited, it is preferably an injection. Specifically, when the active ingredient is siRNA, siRNA comprising an antisense strand complementary to a portion of mRNA corresponding to one or more genes selected from the group consisting of NRBP1 gene, TSC22D3 gene and TSC22D4 gene. , a solution that is isotonic or approximately isotonic with body fluids by dissolving in physiological saline, phosphate buffer, or the like can be used as an injection.

The dosage of the therapeutic agent for dementia according to the present invention may be appropriately adjusted according to the severity, age, body weight, sex, etc. of the patient. or less, preferably 1 ng/kg or more and 10 mg/kg or less. The administration frequency of the therapeutic agent for dementia according to the present invention may be adjusted as appropriate, and may be, for example, once or more per month and once or less per day.

Hereinafter, the present invention will be described in more detail with reference to examples, but the present invention is not limited by the following examples, and can be modified appropriately within the scope that can conform to the gist of the above and later descriptions. It is of course possible to implement them, and all of them are included in the technical scope of the present invention.

Example 1: Dimerization of NRBP1, Interaction with Cul2, and Relationship with TSC22DF FLAG-tagged NRBP1, Cul2, and TSC22DF members were co-expressed in HeLa cells in the combinations shown in FIG. Cells were then lysed with a buffer solution (pH 8.0) containing 50 mM Tris-HCl, 150 mM NaCl, 1 mM DTT, 0.5% Triton X-100, 10% glycerol, and protease inhibitor cocktail tablets (manufactured by Roche). After that, the cell lysate was obtained by centrifugation at 10,000 xg at 4°C.
Furthermore, in the experiment whose results are shown in FIGS. 1 and 2, an antibody against the FLAG tag was added to the cell lysate, incubated at 4° C. for 1 hour, and then protein A/G-agarose beads (manufactured by Santa Cruz) were added for 4 hours. °C for 1 hour. Agarose beads were washed five times with a buffer containing 40 mM Hepes-NaOH, 150-1000 mM NaCl, 1 mM DTT, 0.5% Triton X-100, and 10% glycerol (pH 7.9).
Cell lysates and immunoprecipitated proteins were subjected to SDS-PAGE, transferred to polyvinylidene difluoride membranes (Millipore) and analyzed by immunoblotting with antibodies. Immunoblots were visualized with either Western Lightning Plus-ECL (Perkin Elmer), Immobilon Western (Millipore), or SuperSignal West Femto chemiluminescent reagent (Pierce) according to the manufacturer's instructions.
As a result, proteins that interact with NRBP1 were detected in the beads immunoprecipitated using an antibody against the FLAG tag attached to NRBP1, and in the absence of TSC22DF members, as shown in the results shown in FIG. NRBP1 and Cul2 did not detectably interact. On the other hand, the presence of TSC22D3 markedly increased the association of NRBP1 with Cul2. Also, the effect of co-expression of TSC22D3 and each TSC22DF protein was analyzed. As a result, the association between NRBP1 and Cul2 was most significantly increased when TSC22D3 and TSC22D4 were co-expressed.

We also analyzed the effect of TSC22DF protein on NRBP1 dimerization by co-transfecting different tagged versions of NRBP1. As the results shown in Figure 2, the formation of NRBP1 dimers was most significantly enhanced when both TSC22D3 and TSC22D4 were overexpressed.

Furthermore, co-expression of TSC22D3 and TSC22D4 was also observed to increase the expression of NRBP1, suggesting that interaction with TSC22D4 may increase the stability of NRBP1. Indeed, TSC22D3 had little effect on the stability of NRBP1, whereas TSC22D4 significantly increased its stability, according to cycloheximide follow-up analysis. Co-expression of both TSC22D3 and TSC22D4 also increased the stability of Cul2, as the results shown in FIG.

Taken together, these experimental results suggest that TSC22D3 induces a conformational change in NRBP1 to increase its affinity for Cul2 or stabilizes the interaction between NRBP1 and Cul2, and TSC22D4 induces NRBP1 promoted the dimerization of and protected it from degradation. Thus, TSC22D3 and TSC22D4 act as molecular chaperones to promote NRBP1 dimerization and NRBP1-Cul2 binding through different mechanisms. It is considered that a more remarkable effect can be obtained by combined use.

Example 2: Demonstration that BRI3 and BRI2 are substrates of NRBP1-ubiquitin ligase Forms a RING-type E3 ubiquitin ligase complex.
In this experiment, a mutant NRBP1-IPL having I348A, P349A and L353A mutations in the Cul2-box was produced. NRBP1-IPL has alanine substitutions for amino acid residues at positions 348, 349 and 353, which were predicted to be important for binding to Cul2. Specifically, the ubiquitin chain binding probes TR-TUBE, TSC22D3, TSC22D4, and FLAG-tagged wild-type (WT) NRBP1 or NRBP1-IPL were co-expressed in HEK293T cells (Fig. 4). Immunoblotting results (Fig. 5) clearly showed ubiquitination in cells co-expressing wild-type NRBP1 and TR-TUBE, but not in cells co-expressing NRBP1-IPL and TR-TUBE. I didn't.
The interaction between exogenously expressed NRBP1 and BRI3 in HEK293T cells was also investigated by co-immunoprecipitation analysis. As shown in FIG. 6, ubiquitination and proteasomal degradation are expected to proceed more efficiently with wild-type NRBP1 than with NRBP1-IPL. , substantially less co-immunoprecipitation of BRI3 with wild-type NRBP1. This result is probably due to the formation of a Cullin-RING type E3 ubiquitin ligase complex from wild-type NRBP1 and the degradation of BRI3.
BRI3 is a member of the BRI gene family, which includes homologues BRI1, BRI2 and BRI3 in both mice and humans. BRI2 and BRI1 show 43.7% and 38.3% amino acid sequence identity with BRI3, respectively. Therefore, it was investigated whether BRI1 and BRI2 are substrates of NRBP1. As a result, co-immunoprecipitation analysis revealed that only BRI2, out of BRI2 and BRI1, binds to NRBP1 in HEK293T cells. In addition, as with BRI3, co-immunoprecipitation between BRI2 and wild-type NRBP1 was less than co-immunoprecipitation between BRI2 and NRBP1-IPL, as shown in FIG. Furthermore, as the results shown in Figure 8, substantially more ubiquitinated BRI2 was detected in cells co-expressing wild-type NRBP1 and TR-TUBE than in cells co-expressing NRBP1-IPL and TR-TUBE. was done.
Mutations in the BRI2 gene have 11 more amino acid residues than wild-type BRI2, resulting in the mutant proteins FBD-BRI2 and FBD-BRI2, which cause familial British dementia (FBD) and familial Danish dementia (FDD), respectively. Leads to the production of FDD-BRI2. Therefore, we investigated whether these two mutant proteins are also substrates of NRBP1. As a result, both FBD-BRI2 and FDD-BRI2 interacted with NRBP1-IPL similarly to wild-type BRI2, and approximately the same degree of ubiquitination was observed with the wild-type NRBP1 gene as with wild-type BRI2. Observed in cells expressing the TR-TUBE gene.
Next, it was examined whether NRBP1 affects the stability of BRI2 and BRI3 in cells. According to cycloheximide chase experiments, both BRI2 and BRI3 increased in cells overexpressing the wild-type NRBP1 gene compared to cells overexpressing the NRBP1-IPL gene, as results shown in FIG. was less stable. Also, as the results shown in FIG. 10, the stability of BRI2 and BRI3 was increased in NRBP1 knockdown cells compared to control cells.
Taken together, these results reveal that the NRBP1-containing ubiquitin ligase complex ubiquitinates both BRI2 and BRI3 and regulates their turnover within the cell.

Example 3: Involvement of Cul2 and Cul4A in the ubiquitination of BRI2 and BRI3 by NRBP1-ubiquitin ligase NRBP1 contains a Cul2-box in its amino acid sequence and associates with Cul2 in the presence of TSC22D3 and TSC22D4. We therefore investigated (1) whether Cul2 contributes to the NRBP1-dependent ubiquitination of BRI2 and BRI3 and (2) various dominant-negative dominant-negative studies that are catalytically inactive due to the lack of regions involved in the recruitment of E2 ubiquitin-conjugating enzymes. (DN) Cullin was used to investigate whether other Cullin subtypes exhibit similar effects. In these experiments, we injected HEK293T cells stably expressing wild-type NRBP1, TSC22D3 and TSC22D4 with DN-Cul2, DN-Cul3, DN-Cul4A, DN-Cul4B or DN-Cul5, and BRI2. Alternatively, BRI3 was exogenously expressed and an in vivo ubiquitination assay using HA-tagged TR-TUBE was performed.
As shown in Figure 11, ubiquitination of both BRI2 and BRI3 was dramatically reduced by co-expression of DN-Cul2 as well as DN-Cul4A. We also examined whether overexpression of Cul2 and Cul4A affects the ubiquitination of BRI2 and BRI3. As the results shown in Figure 12, the ubiquitination of BRI2 and BRI3 was clearly strongly enhanced by co-expression of either Cul2 or Cul4A. As shown in FIGS. 13 and 14, even in cells in which the exogenous NRBP1 gene was not expressed, these ubiquitination DN-Cul2-dependent and DN-Cul4A-dependent reductions, and Cul2-dependent Both enhancement of and Cul4A-dependent enhancement were observed.
These results demonstrate that the BC-box protein, NRBP1, cooperates not only with Cul2, previously known to act with BC-box-containing substrate receptors, but also surprisingly with Cul4A, leading to BRI2 and Regulates ubiquitination of BRI3.

Example 4: Role of NRBP1 dimerization in Cul2/Cul4A-type E3 complex formation Our above data support a role for both Cul2 and Cul4A in NRBP1-mediated ubiquitination of BRI2 and BRI3. Therefore, we investigated whether NRBP1 also forms a complex with Cul4A. As the results shown in Figure 15, NRBP1 and Cul4A did not interact intracellularly in the absence of exogenously expressed TSC22DF protein. However, the association of Cul4A and DDB1 with NRBP1 was significantly increased by co-expression of TSC22D3 and TSC22D4.
NRBP1 is known to homodimerize via amino acid residues 406-479. At positions 465-476 of NRBP1, the LXXL motif found as a sequence involved in the dimerization of the LIS1 protein ( LisH motif). To investigate whether the repeats of these Leu residues in NRBP1 are important for its dimerization, we investigated two NRBP1 substitution mutants, NRBP1 [L465E; L468E; L472E; L476E] (LisH -M1) and NRBP1 [L464E; L465E; L466E; L468E; L472E; L476E] (LisH-M2) and three deletion mutants: )], and NRBP1 [Δ(406-479)] were constructed. Co-immunoprecipitation analysis of each of these mutants with V5-tagged wild-type NRBP1 revealed that all of the LisH mutants had a reduced ability to interact with wild-type NRBP1, as shown in FIG. It was found that the LisH domain is important for the dimerization of NRBP1.

Since NRBP1 can homodimerize and also interact with both Cul2 and Cul4A, NRBP1 binds Cul2 and Cul4A to form a dimeric E3 ubiquitin ligase complex containing Cul2/Cul4A. There is a possibility of forming To investigate this possibility, differently tagged Cul2 and Cul4A were co-immunized with wild-type NRBP1 or LisH-M2, an NRBP1 mutant that is unable to form dimers, in the presence of TSC22D3 and TSC22D4. allowed to settle. As a result, as shown in FIG. 17, Cul2 and Cul4A were found to interact with exogenously expressed wild-type NRBP1, but not with LisH-M2. Remarkably, both Cul2 and Cul4A were found to be able to form homodimers in the presence of wild-type NRBP1. These findings suggest that only dimeric, but not monomeric, NRBP1 can bind Cul2 and Cul4A to form a dimeric E3 ubiquitin ligase complex containing Cul2/Cul4A.

Example 5: Contribution of NRBP1 dimerization ability and binding ability with Cul2 and Cul4A to the ubiquitination of BRI2 and BRI3 In the CRL4A ubiquitin ligase, the WD40 motif and a short α- A helical motif is involved in the interaction with the adapter protein DDB1 that bridges the interaction between DCAF and Cul4A. NRBP1 does not have a WD40 motif, but there is a potential H-box motif sequence at positions 331-343 of NRBP1, which contains the canonical BC-box sequence. Of note, a potential H-box sequence is also present in the BC-box sequence of FEM1B, another substrate receptor for CRL2. To clarify the functional significance of these H-box sequences in NRBP1 and FEM1B, and at the same time to identify NRBP1 mutants with selectively impaired binding to Cul2 or Cul4A, the present inventors further Three NRBP1 mutants were constructed. One of them is NRBP1 [H339E] (H- box-M1), and the other is NRBP1 [L336P; H339E] (H-box-M2 ). The remaining one is H-box-M3 in which the NRBP1 H-box is replaced with a sequence almost identical to the FEM1B H-box (positions 595-607) (FIG. 18). To prevent exogenously expressed NRBP1 with wild-type or these mutant H-box sequences from forming dimers with endogenous NRBP1, we modified the amino acid sequence of positions 406-479 to A monomeric version of the NRBP1 construct that was deleted to prevent dimerization was used for the analysis.
As the results shown in FIG. 19, co-immunoprecipitation analysis of HA-tagged Cul2 or Cul4A with FLAG-tagged monomeric NRBP1 mutants revealed that NRBP1 [Δ(406-479)] and monomeric H-box-M3 was shown to associate with both Cul2 and Cul4A. In contrast, monomeric H-box-M1 showed reduced binding to Cul4A while bound to Cul2, whereas monomeric H-box-M2 showed binding to both Cul2 and Cul4A. Connectivity was significantly impaired. As expected, binding to Cul2 was selectively impaired in the monomeric form of NRBP1-IPL.
These results suggest that potential H-box sequences of NRBP1 and FEM1B are involved in association with Cul4A.

NRBP1 is able to dimerize and bind both Cul2 and Cul4A in the presence of TSC22D3 and TSC22D4, thus forming monomeric CRL2 or CRL4A, homodimeric CRL2 or CRL4A, and CRL2/CRL4A in the cell. Various NRBP1-containing CRL complexes can be formed, including CRL4A heterodimeric forms. To investigate whether dimerization of NRBP1 and association with Cul2 and Cul4A is required for ubiquitination of BRI2 and BRI3 by NRBP1-ubiquitin ligases, BRI2 or BRI3 were transformed into TSC22D3, TSC22D4, HA-tagged TR- In vivo ubiquitination assays were performed using HEK293T cells co-expressed with TUBE and wild-type or mutant NRBP1.
As the results shown in FIG. 20, clearly increased ubiquitinated BRI2 and BRI3 accumulation was observed in cells co-expressing wild-type NRBP1, whereas ubiquitinated BRI2 and BRI3 accumulation -Attenuated or not observed in cells co-expressing IPL, H-box-M1, H-box-M2, and H-box-M3. NRBP1-IPL and H-box-M1 showed slightly more accumulation of ubiquitinated BRI2 and BRI3 than H-box-M2 and LisH-M2 because the former were at least partially associated with endogenous NRBP1 in cells. NRBP1-IPL or H-box-M1 and each dimerization partner could bind to Cul4A or Cul2 and Cul2 or Cul4A, respectively, and exert E3 ubiquitin ligase activity. guessed.
These results suggest that not only dimerization, but also binding to both Cul2 and Cul4A are important for NRBP1 function, and that only heterodimeric NRBP1-ubiquitin ligases containing CRL2 and CRL4A shown in Figure 21 are efficient. suggest that BRI2 and BRI3 can be directed to ubiquitin-dependent proteolysis.
Furthermore, H-box-M3 had impaired ability to ubiquitinate BRI2 and BRI3, although it retained binding properties to both Cul2 and Cul4A. The difference in the /H-box sequence is thought to contribute to the E3 ubiquitin ligase activity of NRBP1 to some extent.

Next, we investigated whether dimerization of NRBP1 affects substrate recognition. Either BRI2 or BRI3 were co-expressed in HEK293T cells with TSC22D3, TSC22D4 and FLAG-tagged wild-type or NRBP1 with LisH-M2 mutations and co-immunoprecipitation analysis with an anti-FLAG antibody was performed. To prevent ubiquitination and degradation of BRI2 and BRI3, the cells were treated with a proteasome/protease inhibitor (“MG132” BostonBiochem). As shown in FIG. 22, wild-type NRBP1 and LisH-M2 exhibited similar binding capacities to BRI2 and BRI3. This result suggests that dimerization of NRBP1 has little effect on its affinity for BRI2 and BRI3.

Example 6 Identification of Critical Positions for NRBP1 and BRI2/BRI3 Binding They were tested for their ability to associate with VSV-G-tagged BRI2 or BRI3.
As the results shown in FIG. 23, the C-terminal deletion mutant Δ(328-535) was unable to bind to BRI3, and the N-terminal deletion mutants Δ(2-72), Δ(2-327) , and Δ(2-405) were all able to bind. This result indicates that the N-terminus of NRBP1 is not required for binding to BRI3 and that amino acid residues 406-535 are sufficient for interaction. However, the NRBP1 deletion mutant Δ(406-535) was still able to bind to BRI3, suggesting that at least two regions among positions 328-405 and 406-535 contribute to the interaction. was suggested. Similar results were obtained with BRI2.

We further investigated the interaction of a series of VSV-G-tagged BRI3 and BRI2 deletion mutants with FLAG-tagged NRBP1.
As the results shown in FIG. 24, BRI3 internal deletion mutants such as Δ(91-120), Δ(121-150), and Δ(181-210) were unable to bind to NRBP1. Further analysis using a series of mutants with smaller deletions indicates that the two central portions of BRI3, positions 121-140 and 191-210, are important for the interaction.
Furthermore, as shown in the results shown in FIG. 25, truncated deletion mutants consisting of positions 61-135 and 81-210 or positions 61-136 and 77-210 of BRI3 or BRI2 were added to NRBP1. possessed the ability to connect. The lumenal portion of each protein is important and the C-terminal peptide portion secreted after cleavage by furin protease is not required for interaction.

Example 7: Confirmation that NRBP1 and TSC22D3/TSC22D4 are involved in the degradation and APP processing of BRI2 and BRI3. We investigated whether knockdown of endogenous NRBP1 or TSC22D3/TSC22D4 affects amyloid precursor protein (APP) processing and production of amyloid-β from neurons.
control siRNA, siRNA targeting NRBP1 (SEQ ID NOS: 1-4), siRNA targeting TSC22D3 (SEQ ID NOS: 5-8), siRNA targeting TSC22D4 (SEQ ID NOS: 9-12), or siRNAs (SEQ ID NOs: 1-12) targeting NRBP1, TSC22D3 and TSC22D4 were transfected. Twenty-four hours after transfection, medium was changed, cells were incubated for an additional 24 hours, and cell lysates and conditioned medium were prepared.
As the results shown in Figure 26, knockdown of either NRBP1, TSC22D3 or TSC22D4 by RNAi resulted in increased amounts of endogenous BRI2 and BRI3, and moderately reduced levels of sAPPβ but not sAPPα in the culture medium. bottom. Also, as the results shown in FIG. 27, knocking down either NRBP1, TSC22D3 or TSC22D4 significantly decreased the levels of secreted Aβ40 and Aβ42. These results suggest that the amyloidogenic pathway in neuronal cells is fine-tuned by the endogenous levels of these proteins.
Furthermore, as the results shown in FIG. 26, secreted IDE was slightly increased, but BACE1 in cell lysates was clearly decreased by gene knockdown of NRBP1, TSC22D3 or TSC22D4. Remarkably, as the results shown in FIG. 27, simultaneous knockdown of all three genes using low concentrations of siRNA compared to knockdown of genes encoding either NRBP1, TSC22D3 or TSC22D4. and caused a further decrease in secreted amyloid-β.
These results suggest that NRBP1 and the chaperone proteins TSC22D3/TSC22D4 cooperate to regulate the turnover of intracellular BRI2 and BRI3, thereby controlling the production of amyloid β.

Claims (5)

A method for screening therapeutic agents for dementia, comprising:
reacting BRI2 and/or BRI3 with NRBP1 in the presence or absence of a test compound;
comparing the amount of BRI2 and/or BRI3-NRBP1 complex in the presence and absence of the test compound; and
identifying a test compound for which said amount in the presence of the test compound is less than said amount in the absence of the test compound. 2. The method of claim 1, further comprising dimerizing NRBP1 with TSC22D4. 3. The method of claim 1 or 2, further comprising the step of dimerizing NRBP1 with TSC22D3. The method according to any one of claims 1 to 3, wherein the dementia is Alzheimer's disease. A method according to any one of claims 1 to 3, wherein said dementia is familial English dementia and/or familial Danish dementia.

Images