JP2003169699A - METHOD FOR SCREENING SPECIFIC gamma-PROTEASE INHIBITOR - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for screening an inhibitor inhibiting protease acting on β-amyloid precursor protein of Alzheimer's disease, wherein the inhibitor inhibits specifically at the cleavage site of the protein. <P>SOLUTION: The multiple γ-protease-acting sites acting to β-amyloid precursor protein (APP) of Alzheimer's disease are found and, thereby, screening of a protease inhibitor becomes possible, wherein the protease inhibitor inhibits enzyme action of the γ-protease specifically at the cleavage site of the β-amyloid precursor protein, inhibits production of the aggregate and accumulative protein in Alzheimer's disease but does not inhibit other necessary enzyme actions of γ-protease. This method comprises screening γ-protease inhibitor inhibiting specifically at the γ-protease-acting site by using a mechanism of enzyme action of the γ-protease. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to a method for screening an inhibitor of a protease that acts on β-amyloid precursor protein of Alzheimer's disease, and more particularly, to the enzymatic action of γ-protease specific to the site of action of β-amyloid precursor protein. The present invention relates to a method for screening an inhibitor that inhibits

[0002]

2. Description of the Related Art Alzheimers disease
e: AD) is a progressive degenerative disease of the brain usually characterized by progressive atrophy in the frontal, temporal and occipital cortex. A major pathological change in Alzheimer's disease (AD) is senile plaque. The component of senile plaque is amyloid β protein (amyloid β p) having a molecular weight of about 4000.
rotein: Aβ). Aβ is produced by two-step cleavage from APP (amyloid precursor protein) which is a single transmembrane precursor. It is first cleaved at the extracellular domain (β-cleavage) and then within the cell membrane (approximately in the middle of the membrane) (γ-cleavage). As a result, two types of Aβ (Aβ40 and Aβ42) are generated. Of these, Aβ42 has a very high aggregation ability.

APP is a glycoprotein having a structure similar to that of a receptor that penetrates the membrane once, is composed of about 700 amino acids, and is abundant in the cell membrane and Golgi apparatus. The APP gene resides on chromosome 21 and is mainly composed of three isoforms, APP69, due to alternative RNA splicing.
5 (specifically expressed in nervous system), APP751, and AP
There is P770. Cleavage of APP is performed by a protease called secretase. The secretase contains Lys16-Leu inside Aβ.
Α-secretase, which cleaves, Me at the N-terminal site of Aβ
There are β-secretase that cleaves between t-Asp and γ-secretase that cleaves the C-terminal portion of Aβ in the transmembrane region of APP.

As a product of a causative gene of familial AD (FAD), three kinds of proteins, APP, presenilin 1 (presenilin 1: PS1) and 2 (PS2), have been identified so far. PS1 gene (14th
The chromosome 2) and PS2 genes (chromosome 1) encode 8-transmembrane membrane proteins consisting of 467 and 448 amino acids, respectively, and their gene products have very similar structures, In, it is mainly localized in the endoplasmic reticulum and the Golgi apparatus. Aβ deposited in the brain of Alzheimer's disease (AD) is excised from its precursor, APP, by a protease group such as β-secretase and γ-secretase, and presenilin is γ-.
It is considered to be secretase.

When an abnormality occurs in presenilin, the cleavage position of APP by γ-secretase shifts to the C-terminal side by 2 amino acids, and long Aβ which is not normally produced is produced. This Aβ (Aβ42) is normal Aβ
It tends to aggregate more easily than (Aβ40) and deposit in the brain. Many presenilin mutations found in familial AD are all known to enhance Aβ42 production. Presenilin also acts on the developmental process of the nervous system Notch
It has also been pointed out that it acts as a protease that cleaves the intracellular membrane domain of Notch in signaling. It has been shown that both γ-secretase activity and Notch cleavage activity disappear in PS1 and PS2 double knockout mice.

To date, proteins such as APP, PS1 and PS2 have been identified as products of the causative gene of familial AD (FAD), and mutations in these proteins increase the rate of secretion of Aβ42. It is known. That is, it is considered that senile plaque occurs early and AD occurs due to an increase in the secretion amount of Aβ42. Therefore, what is currently being developed as a therapeutic agent is an inhibitor of γ protease for the purpose of suppressing Aβ production. However, there are major problems with this strategy. This γ protease is known to be involved in the cleavage of Notch, which is essential for cell differentiation. This is actually
It has been revealed that administration of a γ-inhibitor to adult mice inhibits lymphocyte differentiation. Therefore, the development of preventive and therapeutic drugs that overcome these problems is currently awaited.

[0007]

The object of the present invention is to screen for inhibitors of proteases that act on the β-amyloid precursor protein of Alzheimer's disease, especially,
By inhibiting the enzymatic action of γ protease specifically at the cleavage site of β-amyloid precursor protein,
The present invention relates to providing a method for screening a protease inhibitor that inhibits the production of accumulated protein and does not inhibit the enzymatic action of other necessary γ proteases.

[0008]

Means for Solving the Problems As a result of intensive studies to solve the above-mentioned problems, the present inventors have found that 1) APP is at the center of the cell membrane (producing Aβ) and at a site near the cytoplasmic boundary of the cell membrane. 2) The latter cleavage site and the Notch cleavage site have commonality and are controlled by the same protease. 3) Therefore, Not
They found that there is a possibility of developing a drug that inhibits only the Aβ cleavage site without inhibiting the ch cleavage site, and completed the screening method for the inhibitor of the present invention.

That is, the cleavage site producing Aβ is almost in the middle of the membrane, which is 10 or 12 residues away from the cytoplasmic membrane boundary, but the cleavage site of Notch is 3 residues from the same site. The present inventor has focused on the C-terminal fragment (γ fragment) of APP that is produced at the same time as Aβ occurs. A cell-free measurement system was constructed using a membrane fraction prepared from a 1-day-old rat brain to produce a γ fragment,
The γ fragment was purified using immunoprecipitation, gel filtration and HPLC. Sequencing of the purified γ fragment identified fragments starting at 49 and 50 with Aβ numbers. That is, APP is cleaved from the cytoplasm of the membrane at the same time that it is fragmented by γ (sites of Aβ40 and 42).
Upon this purification, the APP family APLP1, A
It was also found that PLP2 was similarly cleaved at a few residues from the cytoplasmic boundary. That is, APP,
APLP1 and APLP2 are 2 or 3 from the cytoplasmic boundary in the membrane
It is cleaved at the residue site. It was considered that this cleavage was performed by the same protease. In addition, this disconnection is
1) Inhibition by a drug known as a γ-inhibitor 2) Cleavage site in the membrane (intramembrane topology)
Are very similar and are cleaved before valine (or leucine).

The screening method of the present invention for detecting an inhibitor that inhibits the enzymatic action of γ-protease in a cleavage site-specific manner of β-amyloid precursor protein allows β-amyloid precursor protein and γ protease to be present. It can be carried out by allowing a test substance to act on a biological sample and detecting a peptide fragment produced by the enzymatic action of γ protease.

That is, in the present invention, a test substance is allowed to act on a biological sample in which a β-amyloid precursor protein and a γ-protease are present, and the enzymatic action of the γ-protease is specific to the cleavage site of the β-amyloid precursor protein. A method for screening a specific γ-protease inhibitor, which comprises detecting an inhibitor that inhibits
The biological sample in which the β-amyloid precursor protein and the γ protease are present is an animal brain stably expressing the β-amyloid precursor protein, or an animal cell. The method for screening a specific γ-protease inhibitor (claim 2), the brain of an animal stably expressing the β-amyloid precursor protein, or the animal cell stably stabilizes the β-amyloid precursor protein. The method for screening a specific γ-protease inhibitor according to claim 2 (claim 3), characterized in that it is a brain or CHO cell of an expressing newborn rat,
The method for screening a specific gamma protease inhibitor according to any one of claims 1 to 3, wherein the gamma protease is gamma-secretase (claim 4).

Further, the present invention is characterized in that the inhibition of the enzymatic action of γ protease is the specific inhibition of the action site which cleaves β-amyloid precursor protein in the middle of the cell membrane. The method for screening a specific γ-protease inhibitor described above (claim 5) or the inhibition of the enzymatic action of γ-protease inhibits the action site that cleaves the β-amyloid precursor protein in the middle of the cell membrane, resulting in the cytoplasm of the cell membrane. The method for screening a specific γ-protease inhibitor according to claim 5 (claim 6) or the specific inhibition of γ-protease is an enzyme for a Notch cleavage site, characterized in that it does not inhibit at an action site for a cleavage site near the boundary. 7. A method for screening a specific γ protease inhibitor according to claim 5 or 6, which does not inhibit the action (claim 7) or the detection of the action site-specific inhibition of the β-amyloid precursor protein is carried out by detecting the cleavage fragment generated by the enzymatic action of the γ protease. A method for screening a specific γ-protease inhibitor (claim 8), or a cleaved fragment of β-amyloid precursor protein produced by the enzymatic action of γ-protease, wherein 10 to 12 residues are separated from the cytoplasmic membrane boundary. 9. A method for screening a specific γ-protease inhibitor according to claim 8, which is a cleavage fragment generated by cleavage in the middle of the cell or cleavage of 2 to 5 residues inside the cytoplasmic boundary of the cell membrane (claim 9). Or an inhibitor that specifically inhibits the site of action of the β-amyloid precursor protein detected by the screening method according to claim 1. Characterized by comprising the specific γ protease inhibitor consisting of (claim 10).

[0013]

BEST MODE FOR CARRYING OUT THE INVENTION
-Amyloid precursor protein (β-amyloid precursor p
rotein: APP amyloid β-protein (A
β), and the production of Aβ40 (40 residues) and Aβ42 (42 residues) proteins by the degradation of Aβ by a γ-protease such as γ-secretase, and AD by the aggregation / accumulation of the Aβ42 and the like in the brain. There is a mechanism called onset. The present invention basically comprises screening an inhibitor that inhibits the enzymatic action of the γ protease in order to block the mechanism of the onset of AD and develop a preventive / therapeutic drug for AD.

In the present invention, β-of γ protease is used.
The cleavage site of amyloid precursor protein (APP)
1) in the middle of the cell membrane (producing Aβ), 2) at multiple sites near the cytoplasmic boundary of the cell membrane, and 3)
The latter cleavage site and the Notch cleavage site have something in common,
Since it was confirmed that they were governed by the same protease, they found that there is a possibility that a drug that inhibits only the Aβ cleavage site without inhibiting the Notch cleavage site can be developed. Therefore, the present invention comprises screening a substance that inhibits the inhibition of the γ protease site-specifically. That is, in the screening method of the present invention, a test substance is allowed to act on a biological sample in which β-amyloid precursor protein and γ protease are present, and the enzymatic action of γ protease is changed to β-amyloid precursor protein (APP). It consists of detecting an inhibitor that specifically inhibits the cleavage site.

In the present invention, as a biological sample in which β-amyloid precursor protein (APP) and γ protease are present, β-amyloid precursor protein (AP
P) stably expressing animal brain such as newborn rat brain, or β-amyloid precursor protein (AP
Animal cells such as CHO cells stably expressing P) can be used. In the present invention, as the γ-protease that acts on β-amyloid precursor protein (APP), it is preferable to use γ-secretase that originally exists in tissues.

In the present invention, in order to detect the specific inhibition of the action site of the enzymatic action of γ protease, a cleavage fragment of β-amyloid precursor protein (APP) produced corresponding to the supposed cleavage site is used. This can be done by detecting. For the detection and purification of the cleaved fragment, purification and identification of the cleaved fragment are commonly used in this field, for example, centrifugation, gel filtration, RP-HPLC, mass spectrometry, immunoprecipitation, western blotting. ,
An amino acid analysis method or the like can be appropriately used. In the method for screening a specific γ protease inhibitor of the present invention, as a specific embodiment, inhibition of the enzymatic action of γ protease inhibits an action site that cleaves β-amyloid precursor protein in the middle of the cell membrane, Screening for inhibitors that do not block at the site of action for cleavage sites near the cytoplasmic border of the cell membrane, in other words,
In the inhibition of the enzymatic action of γ-protease, by screening for an inhibitor that inhibits the formation of an aggregation / accumulation protein such as Aβ42 which causes AD onset by cleavage, but does not inhibit the enzymatic action on the Notch cleavage site. It will be.

[0017]

The present invention will be described in more detail below with reference to examples, but the scope of the present invention is not limited to these examples. [Cleavage of β-Amyloid Precursor Protein (APP) and Detection of Cleavage Site] Precipitation of amyloid β-protein (Aβ) in the brain is invariable and indispensable in the early stage of pathogenesis of Alzheimer's disease (AD). It occurs as an event (Physiol. Rev. 81 (2001) 741-766). CT by cleavage of β-secretase in the extracellular domain
A 99-residue C-terminal fragment (CTF) called F99 or CTFβ is produced, and β
-Aβ is produced from amyloid precursor protein (APP). Then, CTFβ cleavage that occurs in the middle of the transmembrane domain (transmembrane region) by γ-secretase mainly produces a 40-residue protein (Aβ40) or a 42-residue protein (Aβ42). The aggregation ability of Aβ42 is higher, and it is
It is the first type that accumulates in the brain (Biochemistry
32 (1993), 4693-4697; Neuron 13 (1994), 45-53; Ann. Neur
ol. 40 (1995), 294-299).

Intracellular membrane γ-secretase (se of CTFβ
The cleavage by cretase is a 59- or 57-residue cytoplasmic C-terminal fragment called CTFγ.
terminal fragment) (CTF41-99 or 43
-99; APP712-770 or 714-770 according to the numbering of APP770 isoform)
To produce. However, CTFγ has not yet been identified in either brain homogenates or cell lysates, presumably due to instability in vivo. To characterize the γ-cleavage, the characteristics of another product, CTFγ, obtained from this unique cleavage was investigated.

(Cell fractionation of newborn rat brain and Chinese hamster ovary (CHO) cells stably expressing APP751 (GenBank Accession No. X06989))
Newborn (1 day old) rat brain was homogenized with a glass Teflon (registered trademark) in 8 volumes of buffer H (20 mM Hepes, 150 mM NaCl, 10% glycerol, 5 mM EDTA, pH 7.4). Crush on ice 20 times to homogenize. All subsequent procedures were performed at 4 ° C unless indicated. The nuclear fraction was homogenized at 2,500 xg (brain) or 8
Collect by centrifuging at 00 × g (CHO cells) for 10 min, then 0.4% Triton X-1 in buffer H.
A short wash was performed with and without addition of 00. Post-nuclear supernatant is 100,000
It was separated into a cytoplasmic fraction and a membrane fraction by centrifugation at × g for 1 hour. Each fraction was adjusted to equal volume for Western blotting.

(Preparation of Active Membrane Fraction) From the post-nuclear supernatant obtained from the newborn rat brain prepared by the above method, the membrane fraction F2P was spun at 10,000 × g for 15 minutes, and further , Then 10 more to get F3P
Spin for 1 hour at 000 xg. From CHO cells stably expressing APP751, 100,000 x
F2P was obtained by centrifugation at g for 1 hour.
The membrane fraction once washed with the buffer was mixed with a mixture of protease inhibitors (5 mM EDTA, 5 mM EGTA,
5 mM 1,10-phenanthroline, 10 μM bestatin, 1
0 μM amastatin, 10 μM leupeptin, 10 μM
Incubated in buffer H containing antipain, 50 μM TLCK) at 37 ° C. for the indicated times. The putative CTFγ, soluble and membrane bound, was added to the reaction mixture for 10
Separation was achieved by centrifugation at 000 xg for 1 hour. As described above for Aβ (J. Biol. Chem. 2
71 (1996) 22908-22914), 16.5% containing 8M urea.
Tricine gel with C4, UT-18, or 6E10
Western blotting was performed using (Senetek PLC (Maryland, Mo)). Assumed CTFγ and Aβ
F2P obtained from CHO cells stably expressing APP751 was analyzed by CHAP
S, CHAPSO, TritonX-100, NP-40, Tween-2
Incubated for 30 minutes at 37 ° C. in the presence of 1% of various detergents including 0, Tween-80, and Brij-35. The reaction mixture was Western-blotted with UT-18 and 6E10.

(Purification of putative CTFγ) As described above, the membrane fractions F2P and F3P were treated in buffer H containing a mixture of protease inhibitors at 37 ° C. for 90 minutes (F
3P) or 120 minutes (F2P) incubation.
After centrifugation at 100,000 xg for 1 hour, the supernatant and precipitate were pooled separately. Assumed CTF bound to the membrane
γ (CTFγ) was extracted twice from the precipitate with 0.1 M Na ₂ CO ₃ . Adjust the pH to 7.4, 0.5%
After the addition of NP-40, the membrane-bound CTFγ was purified in the same manner as the soluble CTFγ. Prepared CTF
γ is protein G-Sepharose CL-4B (Amersham Pha
Prepurified with rmacia Biotech (Uppsala Sweden) for 1 hour and incubated with C4 for 1 hour. protei
Immunoprecipitates were harvested by overnight incubation with n G-Sepharose CL-4B. It was determined that 95% or more of CTFγ was immunoprecipitated. Beads are TS
-T buffer (150 mM NaCl, 50 mM Tris
-HCL, pH 7.6, 0.1% Tween20) 3 times, 10
Once in mM Tris-HCL (pH 7.6) and cold H
It was washed once with ₂ O and then extracted with 1% TFA / 30% acetonitrile.

After drying in a vacuum centrifuge, the sample is placed in 5
Superdex75 (H, pre-equilibrated with 6M guanidine hydrochloride in 0 mM Tris-HCl, pH 7.6)
R10 / 30) column (Amersham Pharmacia Biotech) was used for gel filtration. When measured by Western blot using UT-18, the molecular mass is ˜9 kDa.
One clear peak was observed. The pooled fractions of CTFγ were collected on an Aquapore RP300 column (2.1 x 30
mm; Applied Biosystems (manufactured by (Foster City, CA)) reverse phase (RP) -HPLC (model 1090M; Hewlett-P)
It was further purified by ackard (Waldbronn, Germany). Purification was carried out in 0.1% trifluoroacetic acid (TFA),
Isocratic elution (isocratic elution) containing 5% acetonitrile for 3.5 minutes, then a linear gradient containing 5-55% acetonitrile in 0.1% TFA for 12 minutes at 0.2 ml / min. Flow rate was used.

(Amino acid sequence of purified CTFγ and mass spectrum analysis) Each RP-HPLC fraction was analyzed by Ap
plied Biosystems Model 494
cLC or Model492Protein Sequ
Sequencing was performed by the Encer. Mass spectrum analysis is performed by the Voyager RP equipped with a delayed extraction function.
MA (made by PerSeptive Biosystems (Framingham, MA))
LDI-TOF MS (matrix-assisted laser desorp
tion / ionization time-of-flight mass spectrometry)
Was performed using. Peptide mass was measured in delayed extraction mode using 2-mercaptobenzothiazole as matrix. For the mass calibration, human angiotensin I and bovine insulin (both manufactured by Sigma (St. Louis, MO)) were used as external standards.

(Preparation of antibody) In order to detect the putative CTFγ, two antibodies, C4 and UT-18, which were prepared for the 30- and 20-C-terminal residues of APP, were used (Bi
ochm. Biophys. Res. Commun. 160 (1989) 1296-1301; J. B.
iol. Chem. 274 (1999), 2243-2254). In this study, C4
Was used for immunoprecipitation and UT-18 for Western blotting. In the cell-free system, in order to detect the generated Aβ, 6E10 (the epitope is Aβ
1-16) was used.

(Analysis of Results) Newborn rat brain and APP7
In CHO cells stably expressing 51, CT
Fγ was detected at a position of ˜6.5 kDa (FIGS. 1A and 1).
B). Most CTFγ in vivo and in cultured cells
Was partially present in the membrane fraction in the cytoplasmic fraction, but could hardly be confirmed in the nuclear fraction (FIGS. 1A and 1B). CTFγ can be generated by using a cell-free system. From newborn rat brain, membrane fraction F2P (2,500 × g-10,000 ×
g) and F3P (10,000 × g to 100,000 ×)
g) was prepared. For CHO cells stably expressing APP751, 800 × g to 100,000 ×
F2P was obtained by precipitation with g. After incubating at 37 ° C. for the indicated period, 100,00 of the reaction mixture was added to obtain the soluble and membrane fractions separately.
Centrifuge at 0xg.

Then, each was inspected by Western blotting using UT-18. We consistently observed the time-dependent production of CTFγ in these membrane preparations (FIGS. 1C and 1D). F compared to F2P fraction
3P produced twice the amount of CTFγ. On the contrary,
Membranes from APP-transfected CHO cells produced 10 times more CTFγ. CTFγ migrated as a single band at 5 kDa when separated on a urea-free 16.5% tricine gel. However, in the 16.5% gel containing 8 M urea, a dark 6.5-kDa band and a light 7.5-kDa band were separated (Anal. Biochem. 237 (1996), 24-29). Nearly half of the CTFγ produced in the cell-free system was membrane bound and could be extracted with 0.1M Na ₂ Co ₃ but not 2M NaCl.

The specific protease (s) responsible for this cleavage have some similar properties to γ-secretase which produces Aβ from CTFβ. The production of CTFγ in the preparation of membranes from newborn rat brain
L-685,458 aspartyl protease transition state analog (aspartyl protease transitions
It is inhibited by a tate analogue) inhibitor in a dose-dependent manner (Biochemistry 39 (2000), 8698-8704) (Fig. 1
E). Two other γ-secretase inhibitors, pepstatin A and DFK167
(Substrate-based difluoroketone inhibitors (J. Med.
Chem.41 (1998), 6-9; J. Biol. Chem.276 (2001), 481-48
7) also suppressed this generation, but to a lesser extent.
Other protease inhibitors, including common serine proteases, cysteine proteases, and inhibitors such as metalloproteases, did not affect cleavage. Furthermore, "CTFγ production has a similar sensitivity / resistance profile to detergents to that of Aβ (Fig. IF).

Despite the reported production of CTFγ in the cell-free system (J. Biol. Chem. 27
6 (2001), 481-487), the generated fragment has never been confirmed by sequencing or mass spectrometry. Therefore, CTFγ produced from the F2P and F3P fractions of newborn rat brain was purified. Partially purified CTF
From preliminary mass spectral analysis of γ, CTF50-9
9 (APP721-770) and some incomplete N-terminal fragments were indicated, suggesting that certain aminopeptidases in preparation act on CTFγ. Therefore, during the production of CTFγ or immunoprecipitation, a high concentration of aminopeptidase inhibitor (5 mM 1,10-
phenanthroline, 10 μM bestatin and 10 μM am
astatin) was included. Compared to soluble CTFγ, which appeared as a single band at 6.5 kDa, membrane-bound CTFγ produced a band at 7.5 kDa in addition to the 6.5 kDa band (FIG. 2B).

Immunoprecipitation was used as the first step in the purification of both soluble CTFγ and membrane bound CTFγ. Immunoprecipitation with C4 was quite effective for CTFγ purification (lanes 1, 2 in Figures 2A and 2B,
3). The immunoprecipitated CTFγ was then gel-filtered in the presence of 6M guanidine hydrochloride (lane 4 of Figures 2A and 2B). Soluble CTF by Western blot
Both γ and membrane-bound CTFγ have M _r ˜9 kD
It was proved to elute as one peak in a.
Fractionated CTFγ was further purified by reverse phase (RP) -HPLC (lane 5 in Figures 2A and 2B). Both soluble CTFγ and membrane-bound CTFγ,
A similar RP-HPLC profile was seen: one large peak (peak s2 in Figure 2C and peak m2 in Figure 2D) and several smaller peaks with shoulders (s3 and m3 respectively). . The large peaks (s2 or m2) were most strongly labeled by UT-18, the shoulders (s3 or m3) were less pronounced, while the other small peaks were mostly not positive for UT-18. . In RP-HPLC, 6.5-kDa and 7.5- in the putative CTFγ bound to the membrane.
Failed to separate it of kDa.

All fractions positive for UT-18 and some with slower elution (see FIGS. 2C and 2D)
Were subjected to amino acid sequencing and major fractions were also subjected to mass spectral analysis. The results are summarized in Table 1. CTF5 is present in large peaks (s2 and m2) in both soluble and membrane bound CTFγ.
The 0-99 (APP721-770) fragment was included. N
Fragment cleaved at the terminal residues 1 to 3, C
TF51-99 (APP722-770), CTF52
-99 (APP723-770), and CTF53-
99 (APP724-770) was unequivocally identified by both mass spectral analysis and sequencing. However, the proportion of these fragments is 10 of CTF50-99.
% Or less. The 6.5- and 7.5-kDa proteins of peak m2 gave only one N-terminal sequence. The 7.5-kDa band may be a phosphorylated version of the 6.5-kDa protein (EMBO J.
13 (1994), 1114-1122; Mol Med3 (1997), 111-123), and it was not possible to detect a sign indicating it in the mass spectrum.

[0031]

[Table 1]

A small amount of CTF49-99 (APP720-
770) fragment was present in the s3 / m3 peak. C in these two peaks by mass spectrometry
A weaker signal was identified, corresponding to the mass of TF48-99 (data not shown). RP-HPL
Retention times appear to be extended for longer fragments of the C column. Therefore, sequencing was performed to determine if the small peaks at s4-s8 and m4-m8 contained sufficient amounts to detect authentic CTFγ (Table 1).

Table 1 shows the sequence and mass spectrometric results of soluble and membrane bound cCTF separated by RP-HPLC. The samples were subjected to 10 cycles of N-terminal sequencing. Spectral analysis is based on MALDI-TOF-MS (matrix-assisted lase
r desorption / ionization time-of-flight mass spectr
geometry) using 2-mercaptobenzothiazole as matrix and in delayed extraction mode. The table shows, for each peak, "N-terminal sequence", "amount (pmol) ^a ", "mass (MH ⁺ ) ^b ",
"Fragment ^c " is shown. In the table, a is an estimate from initial yield in sequence analysis; b is average mass; numbers in parentheses of c indicate identified N-terminal residues; d is unidentified; e is Not determined; f may include these fragments up to the carboxy terminus of the precursor protein; g, rat APLP1 sequence not available in the database, and S6 1
It is shown that when 8 cycles of N-terminal sequencing were performed, the sequence obtained therefrom was found to be the same as that of mAPLP1 (602-619). )

The results were surprising. Peak s4
The peptide of / m4 has an N-terminal sequence corresponding to residues 716-725 of rat β-amyloid precursor protein 2 (rAPLP2), and has residues L715 and V of APLP2.
It was shown that there is a break in the cell membrane between 716 (Fig. 2).
E). The peptide at peak s5 / m5 contains mAp
LP1 (604-613) and mAPLP1 (605
An N-terminal sequence corresponding to (-614) was found (Fig. 2E).

The peak s6 / m6 peptide contains mAp
An N-terminal sequence corresponding to LP1 (602-611) was found. No clear sequence was confirmed in the peaks s7 / m7 and s8 / m8. Again, the real CTF
No sequence corresponding to γ was confirmed. In the future, APP
The fragment (“CTFγ) generated from
Shown as (cCTF). APLP1 and APLP2
CCTF co-purification of C
It can be explained by a cross-reaction of 4: C4 was made to the 30 C-terminal residue of APP where there is high homology throughout the β-amyloid precursor protein family.

By comparison of the cleavage sites of APP, APLP1 and APLP2, the two located inside the cytoplasmic border of the membrane
A new cleavage site of 5 residues can be identified (Fig. 2E).
Therefore, this intracellular cleavage is distinctly different from the γ-cleavage that produces Aβ that occurs in the middle of the transmembrane domain (9 residues or more inside the membrane boundary). γ-
Cleavage is believed to have little sequence specificity (J. Biol. Chem. 274 (1999), 11914-11923). However, it has not been clarified why APP cleaves near the middle of the transmembrane domain, whereas Notch1 cleaves immediately within the cytoplasmic border of the membrane (Nature 393 (1998), 382-386). . This contradiction has never been properly explained. The conserved V1744 in the Notch family is very important in Notch1 cleavage (Nature393 (1998), 382-38.
6). The corresponding Val residue is present in APP and APLP2 but not in APLP1. The low specificity of this new type of cleavage seen in APLP1 may be related to the substitution of Val for Leu.

On the other hand, V717 and L723 of APP
Appears to be conserved among all four transmembrane proteins. Although the importance of these two residues is unclear, it should be noted that mutations occurring in L723 (L723P) cause familial AD (Ann. Neurol. 47 (2000), 249-253). ). This residue is γ
Remote from the cleavage site and therefore differently AP
It is considered that it affects the processing of P.
When presenilin is deficient, APL
It should be noted that the CTF of P1 is accumulated (Ne
uron 21 (1998), 1213-1221). Overall, these findings strongly suggest that Notch cleavage is more similar to this unique type of cleavage than gamma cleavage.

Then, what is the relationship between this newly identified cleavage and γ cleavage? The first possibility is that CTFγ (CTF41-99 and 43-99) is produced first, but is rapidly degraded by certain aminopeptidases and / or endopeptidases and therefore complete This means that long CTFγ cannot be detected. However, this possibility is not conceivable because the reaction mixture always contained high concentrations of aminopeptidase inhibitors and other types of protease inhibitors.

The second possibility is CT.
Fβ is first cleaved at this unique site, resulting in CTF50-
99 or 49-99. The corresponding "long Aβ" (Aβ1-49 or 1-4
8) is then rapidly degraded by a particular carboxypeptidase. Since the hydrophobic environment is unfavorable for hydrolysis, it is reasonable to assume that cleavage will occur first near the edges of the membrane. Produced "long A
It is believed that β'slide out of the membrane into the lumen.
This is because the cytoplasmic side lacks poly-base amino acid residues that stop it. Following this, it is considered that the degradation by carboxypeptidase gradually progresses,
Various lengths (Aβ1-38, -39, -40, -42,
-43) Aβ is produced.

The third possibility is that Aβ
The γ-cleavage that produces γ and the newly discovered type of cleavage occur in two related but separate proteolytic pathways, where the latter resembles that of Notch1. This observation is consistent with the finding that Aβ production and Notch1 cleavage are differentially affected by mutations in presenilin and certain protease inhibitors (Nat. Cell Biol. 2 (2000), 205). -
211); Nat. Cell Biol. 3 (2001), 507-511). Finally, whether there are other type 1 transmembrane proteins that cause similar cleavage, and this type of intracellular membrane cleavage is common in releasing cytoplasmic (and ectodomain) domains from single-pass proteins. It is important to investigate whether it is a mechanism.

[Description of Drawings] FIG. Assumed CTFγ in vivo and in vitro. A and B: Postulated CTF in newborn rat brain (A) and CHO cells stably expressing APP751 (B).
Cellular fraction of γ (“CTFγ”). The cytoplasmic (C), membrane (M), and nuclear (N) fractions were adjusted to equal amounts for Western blotting. CFTα (CTF83)
Was herein defined as the C4-positive and 6E10-negative fragment. C and D: Production of CTFγ in a cell-free system in membrane purified from rat brain. Post-nuclear (pos.
tnuclear) supernatant was F2P (10,000 xg for 15 minutes, C) and F3P (100,000 xg for 1 hour,
Fractionated in D). Membrane fractions were incubated in a mixture of protease inhibitors at 37 ° C for the indicated times.
The reaction mixture was separated into supernatant (C and D upper panels) and pellets (C and D lower panels) by centrifugation at 100,000 xg for 1 hour, both UT-1.
Western blotting using 8.

Inhibitor L- for CTFγ and Aβ production
Effect of 685,485 and surfactant. E, F2P from newborn rat brain was incubated for 2 hours in the presence of L-685,458 (concentrations shown above lanes). After centrifugation at 100,000 xg for 1 hour, the supernatant (upper panel) and precipitate (lower panel) were probed separately with UT-18. F, APP75
F2P from CHO cells stably expressing 1
1% of the various surfactants shown at the top of the lane
Was incubated for 30 minutes at 37 ° C. in the presence of the following and then Western blotted using UT-18 (top panel) and 6E10 (bottom panel). Con +,
-Indicates incubation at 37 ° C. or 0 ° C. for 30 minutes without using a surfactant, respectively. CT
Fβ is defined herein as the C4-positive and 6E10-positive fragments.

FIG. Purification of putative CTFγ and APP,
Cleavage sites in the plasma membrane of APLP1, APLP2, and Notch. Soluble (A and C) and membrane-bound (B and D) CTF
γ was prepared from the membrane fraction of newborn rat brain (lane 1 of A and B). CTFγ was effectively immunoprecipitated with C4 (lanes 2 of A and B are samples after immunoprecipitation). 1
After extraction with% TFA / 30% acetonitrile (lane 3), CTFγ was purified by gel filtration (lane 4) and further using RP-HPLC (lane 5). Recovery of CTFγ was observed by Western blot using UT-18. The arrow indicates CTFγ
Indicates. C and D: RP-HPLC profile for soluble and membrane bound CTFγ, and Western blot of HPLC fractions (inset). E: APP, APL
Comparison of intracellular cleavage sites for P1, APLP2, and Notch1. The upper figure shows the cleavage site of APP. The Aβ portion is indicated by a thick thick line. The shaded area represents the transmembrane domain (APP700-723).

C4 and UT-18 are listed for the last 30 and 20 residues of APP, respectively.
The area where intracellular membrane cleavage may occur is enlarged and 1
It is indicated by a letter code. The letter (K
KK, RKR, RKKK, RKRRR) represent basic residues that play a role of a transmembrane termination sequence. APP 721 site, rAPLP2 716 site, mNotch
The letter at the 1744 site of 1 indicates an important Val residue in the cleavage of Notch1. 717 and 723 sites of APP, 712 and 718 sites of rAPLP2, and mAPLP1
600,606 sites of mNotch1 1740,1
The letters at the 746 site indicate the residues conserved between the four sequences around the cleavage site. The 3-digit or 4-digit number represents the residue number. APLP1, APLP
The portions corresponding to 2 and Notch are aligned in a line. Arrows represent cleavage sites (major or secondary).

[0045]

EFFECTS OF THE INVENTION Currently being developed as a therapeutic drug for Alzheimer's disease, β-amyloid protein (A
β) It is an inhibitor of γ protease for the purpose of suppressing production, and this γ protease is an N that is essential for cell differentiation.
It is known to be involved in otch cleavage, and inhibition of this γ-protease had an inherent problem of also inhibiting Notch cleavage. In the present invention, β-amyloid precursor protein (APP) for Alzheimer's disease
It has been confirmed that there are multiple sites of action of γ-protease against γ-protease, and it does not inhibit the enzymatic action of γ-protease necessary for living organisms such as Notch cleavage, and does not inhibit the action of β-amyloid protein (Aβ) It has been found that it is possible to specifically inhibit the enzyme, and the screening method of the present invention made it possible to develop an inhibitor that specifically inhibits such enzyme action. The screening method of the present invention is an effective means for developing a preventive or therapeutic drug for Alzheimer's disease.

[0046]

[Sequence list]                                SEQUENCE LISTING <110> JAPAN SCIENCE AND TECHNOLOGY CORPORATION <120> Screening method of specific gamma protease inhibitor <130> A011P41 <140> <141> <160> 4 <170> PatentIn Ver. 2.1 <210> 1 <211> 16 <212> PRT <213> Homo sapiens <400> 1 Val Ile Ala Thr Val Ile Val Ile Thr Leu Val Met Leu Lys Lys Lys   1 5 10 15 <210> 2 <211> 16 <212> PRT <213> Rattus norvegicus <400> 2 Ala Ile Ala Thr Val Ile Val Ile Ser Leu Val Met Leu Arg Lys Arg   1 5 10 15 <210> 3 <211> 17 <212> PRT <213> Mus musculus <400> 3 Gly Gly Gly Ser Leu Ile Val Leu Ser Leu Leu Leu Leu Arg Lys Lys   1 5 10 15 Lys <210> 4 <211> 18 <212> PRT <213> Mus musculus <400> 4 Val Leu Leu Phe Phe Val Gly Cys Gly Val Leu Leu Ser Arg Lys Arg   1 5 10 15 Arg Arg

[Brief description of drawings]

FIG. 1 shows the results of Western blotting of putative CTFγ in vivo and in vitro in Examples of the present invention. A and B are cell fractions of putative CTFγ in neonatal rat brain (A) and CHO cells stably transfected with APP751 (B), and C and D are cell-free CTFγ in the membrane purified from rat brain. , E and F show the effects of γ-secretase inhibitor and detergent on the production of CTFγ and Aβ.

FIG. 2 shows purified putative CTF according to an embodiment of the present invention.
Results of Western blotting of γ (A, B) and R
It is a figure which shows a P-HPLC profile figure (C, D). E is APP, APLP1, APLP2, and No
It is a figure which shows the comparison of the intracellular cleavage site of tch1.

Claims (10)

[Claims] 1. A test substance is allowed to act on a biological sample in which β-amyloid precursor protein and γ protease are present to inhibit the enzymatic action of γ protease specifically at the cleavage site of β-amyloid precursor protein. A method for screening a specific γ protease inhibitor, which comprises detecting an inhibitor. 2. The biological sample in which the β-amyloid precursor protein and the γ protease are present is an animal brain or animal cells stably expressing the β-amyloid precursor protein. The method for screening a specific γ protease inhibitor according to claim 1. 3. An animal brain stably expressing a β-amyloid precursor protein or an animal cell is a newborn rat brain or CHO cell stably expressing a β-amyloid precursor protein. The method for screening a specific γ protease inhibitor according to claim 2, wherein 4. The method for screening a specific γ protease inhibitor according to any one of claims 1 to 3, wherein the γ protease is γ-secretase. 5. The inhibition of the enzymatic action of γ protease is β
-The method for screening a specific γ protease inhibitor according to any one of claims 1 to 4, which is a specific inhibition of an action site that cleaves an amyloid precursor protein in the middle of a cell membrane. 6. The inhibition of the enzymatic action of γ protease is β
Screening for a specific gamma protease inhibitor according to claim 5, characterized in that it inhibits the site of action that cleaves the amyloid precursor protein in the middle of the cell membrane, but not at the site of action for the cleavage site near the cytoplasmic boundary of the cell membrane. Method. 7. The specific inhibition of γ protease is Not
The method for screening a specific γ protease inhibitor according to claim 5 or 6, which does not inhibit the enzymatic action on the ch cleavage site. 8. The method according to claim 1, wherein the site-specific inhibition of β-amyloid precursor protein is detected by detecting a cleavage fragment produced by the enzymatic action of γ protease. For screening a specific γ protease inhibitor of the above. 9. A cleavage fragment of a β-amyloid precursor protein produced by the enzymatic action of γ protease has 10 to 12 residues from the cytoplasmic side membrane boundary, and the cleavage is in the middle of the cell membrane or inside the cytoplasmic boundary of the cell membrane. The method for screening a specific γ protease inhibitor according to claim 8, which is a cleavage fragment generated by cleavage of 2 to 5 residues. 10. A specific γ protease inhibitor, which comprises an inhibitor that specifically inhibits the action site of the β-amyloid precursor protein detected by the screening method according to any one of claims 1 to 9. Agent.