The Mechanism of
-Secretase
MULTIPLE INHIBITOR BINDING SITES FOR TRANSITION STATE ANALOGS AND SMALL MOLECULE INHIBITORS*
Gaochao Tian
,
Smita V. Ghanekar ¶,
David Aharony ¶,
Ashok B. Shenvi ||,
Robert T. Jacobs ||,
Xiaodong Liu ** and
Barry D. Greenberg **
From the
Departments of
Lead Discovery,
¶Neuroscience,
**Molecular Sciences, and
||Chemistry, AstraZeneca Pharmaceuticals,
Wilmington, Delaware 19850
Received for publication, January 28, 2003
, and in revised form, April 24, 2003.
 |
ABSTRACT
|
---|
Transition state analogs pepstatin methylester (PME) and L685458 have been
shown to inhibit
-secretase non-competitively (Tian, G.,
Sobotka-Briner, C., Zysk, J., Liu, X., Birr, C., Sylvester, M. A., Edwards, P.
D., Scott, C. W., and Greenberg, B. D. (2002) J. Biol. Chem. 277,
3149931505). This unusual kinetics suggests physical separation of the
sites for substrate binding and catalysis with binding of the transition state
analogs to the catalytic site and not to the substrate binding site. Methods
of inhibitor cross-competition kinetics and competition ligand binding were
utilized to address whether non-transition state small molecule inhibitors,
which also display non-competitive inhibition of
-secretase, inhibit
the enzyme by binding to the catalytic site as well. Inhibitor
cross-competition kinetics indicated competitive binding between the
transition state analogs PME and L685458 and between small molecules
arylsulfonamides and benzodiazepines, but non-competitive binding between the
transition state analogs and the small molecule inhibitors. These results were
indicative of two inhibitor binding sites, one for transition state analogs
and the other for non-transition state small molecule inhibitors. The presence
of two inhibitor binding sites for two different classes of inhibitors was
corroborated by results from competition ligand binding using
[3H]L685458 as the radioligand. Although L685458 and PME displaced
the radioligand at the same concentrations as for enzyme inhibition,
arylsulfonamides and benzodiazepines did not displace the radioligand at their
Ki values, a result consistent with the presence
of two inhibitor binding sites. These findings provide useful insights into
the catalytic and regulatory mechanisms of
-secretase that may
facilitate the design of novel
-secretase inhibitors.
 |
INTRODUCTION
|
---|
Alzheimer's disease is characterized by formation of neurofibrillary
tangles and amyloid plaques in the regions of the central nervous system that
are involved in learning and memory
(1). It is believed that
accumulation of
A
1 in plaques or
as soluble aggregates initiates a pathological cascade leading to synaptic
dysfunction and neuronal toxicity, with neurodegeneration and dementia as the
final outcome (1,
2). Therefore, strategies to
reduce the level of brain A
are being aggressively pursued as an
approach likely to benefit Alzheimer's disease patients.
A
is produced as the result of sequential proteolysis of a type I
transmembrane protein APP by
- and
-secretases.
-Secretase
cleaves APP in its extracellular domain at a site close to the membrane
surface, a reaction that generates a membrane-bound APP C-terminal fragment of
99 amino acid residues (C99). A subsequent endoproteolysis within the
transmembrane domain of C99 by
-secretase produces A
. Whereas
-secretase, an aspartyl protease, has been well characterized
(36),
the identity and structure of
-secretase, also thought to be an
aspartyl protease
(712),
remains elusive, and its kinetic and catalytic mechanisms are poorly
understood. To a large extent, this is due to the highly complicated
structural organization of this unusual protease. In contrast to other known
proteases,
-secretase is composed of a high molecular weight
multicomponent complex of transmembrane proteins
(13,
14). Primarily due to this
structural complexity, the catalytic site and mechanism of action of
-secretase has not been unequivocally established. Early findings point
to presenilin 1 or 2 as the catalytic subunit of
-secretase
(15,
16). These multipass
transmembrane proteins contain two essential aspartate residues in putatively
adjacent transmembrane domains
(15) and can be cross-linked
by high affinity
-secretase inhibitors
(1618).
Recent advances have identified three additional proteins, nicastrin
(19), aph-1
(20,
21), and pen-2
(20,
22), in the same
multicomponent complex, whose co-expression with presenilin appears to be
critical for
-secretase activity
(1922).
However, the precise roles of these additional protein subunits in the
catalytic mechanism of
-secretase await further investigation.
Associated with the structural complexity of
-secretase is the
versatility of this protease in cleaving several type I transmembrane
proteins. In addition to APP processing,
-secretase is required for
proteolytic activation of Notch receptor
(2326),
a signaling molecule essential for embryonic development of all metazoan
species (27). Cleavage of
Notch in the transmembrane domain by
-secretase generates Notch
intracellular domain (NICD), which then translocates into the nucleus, where
it regulates gene transcription
(28,
29). The list of other
potential protein substrates for
-secretase has recently been expanded
to include ErbB4 (30,
31), E-cadherin
(32), and CD44
(33). However, the mechanisms
by which
-secretase reacts with these different substrates remains
unknown. Given the importance of
-secretase in the Notch pathway, as
well as its potential importance in pathways involving these other substrates,
it is highly desirable to develop inhibitors selective for APP cleavage.
Recent advances in understanding the kinetics and mechanisms of
-secretase activity suggest that this may be possible.
Transition state isosteres and small molecule inhibitors have been
described for
-secretase. Recently, PME and L685458, which are
transition state isosteres, have been shown to display non-competitive
inhibition of
-secretase
(34), suggesting an
unprecedented enzyme kinetic mechanism that involves physical separation of
substrate binding and catalysis
(34). Inhibitors that bind to
the catalytic site, as is normally expected for transition state isosteres,
should impact similarly on the catalytic activities of the enzyme with
different substrates. Inhibitors that target a substrate binding site or some
other modulatory sites on
-secretase may, on the other hand, display
substrate-selective effects
(34). Interestingly,
arylsulfonamides and benzodiazepines, which do not resemble transition state
isosteres, also display non-competitive inhibition of
-secretase
(34). This raised the question
of whether these small molecule inhibitors also affect
-secretase
activity by binding to the same site as the transition state isosteres or
interact with
-secretase at some other as yet unknown binding
sites.
To address this issue, we employed the methods of inhibitor
cross-competition kinetics and competition ligand binding to characterize
further the interactions of
-secretase with transition state analogs
(PME and L685458) and small molecule inhibitors (arylsulfonamide 1 and
benzodiazepine 2, also known as compound E) (see
Fig. 1). Data obtained from
inhibitor cross-competition kinetics suggested the presence on
-secretase of two inhibitor binding sites, one for binding of
transition state isosteres and the other for non-transition state small
molecule inhibitors. This observation was corroborated by results from
competition ligand binding with the radioligand [3H]L685458. These
findings provide useful insights into the catalytic and regulatory mechanisms
of
-secretase that may facilitate the design of novel, and possibly
substrate-selective, inhibitors.

View larger version (21K):
[in this window]
[in a new window]
|
FIG. 1. -Secretase inhibitors used in the current study. Chemical
structures of PME and L685458, which are transition state analogs of aspartyl
proteases, and small molecule inhibitors arylsulfonamide 1 and
benzodiazepine 2 (Compound E) are shown.
|
|
 |
EXPERIMENTAL PROCEDURES
|
---|
MaterialsCHAPS and CHAPSO were purchased from Pierce.
Phosphotidylcholine and phosphotidylethanolamide were purchased from Avanti
Polar Lipids. A
40 and R
A
40 were purchased from BIOSOURCE.
Biotin-4G8 was from Senetek PLC. Dynabeads (M-280) were purchased from IGEN.
Ruthenium-labeled G
R, C100, and detergent-solubilized human
-secretase in Tris-HCl at pH 8.4 were prepared as described previously
(34). Preparation of the
enzyme in a buffer of pH 6.5 was performed under the same conditions as
described for the enzyme preparation at pH 8.4
(34) except that Tris-HCl was
replaced with MES. PME, L685458, 1, and 2 were prepared as
reported previously (34).
Synthesis of [3H]L685458 [3H]L685458
was synthesized by tritiation, under catalytic conditions, of
{1S-benzyl-4R-[1-(1-carbamoyl-2-phenyl-ethylcarbamoyl)-1S-3-methyl-but-3-enylcarbamoyl]-2R-hy
droxy5-phenyl-pentyl}-carbamic acid tert-butyl ester (dehydroleucine
derivative of L685458). The desired dehydroleucine derivative of L685458 was
prepared by a method similar to that described for the synthesis of L685458
(35) except that
4,5-dehydro-Leu-Phe-NH2 was used instead of
Leu-Phe-NH2.
Enzyme and Inhibition KineticsEnzyme and inhibition
kinetics were performed using the methods as described previously
(34) with minor modifications.
All the reactions were run in 96-well microplates. Reactions with a final
volume of 280 µl were run in 25 mM MES, pH 6.5, containing C100
at a defined concentration, solubilized enzyme at 2040-fold dilution
from stock (prepared at either pH 8.4 or pH 6.5), inhibitor or inhibitors at
different combinations of final concentrations diluted from a stock in
Me2SO (final concentration of Me2SO was maintained at
5%), 1 mM EDTA, 1 mM dithiothreitol, 0.1 mg/ml bovine
serum albumin, 0.25% CHAPSO, 0.01% phosphotidylethanolamide, and 0.01%
phosphotidylcholine. The reactions were initiated by addition of enzyme, and
40-µl aliquots of the reaction mixture were quenched at six different times
by addition of 50 µl of 200 µM PME. A 50-µl solution
containing 0.2 µg/ml R
A
40 and 0.25 µg/ml biotin-4G8 was
added to each quenched reaction mixture. The solutions were shaken in 96-well
plates at 4 °C for >7 h. Subsequently, 50 µl of a solution
containing 0.018 µg/ml Ru-G
R and 62.5 µg/ml Dynabeads was added,
and the plates were incubated at room temperature for 1 h, at which time they
were measured for ECL counts in an IGEN ECL M8 analyzer.
Competition Ligand BindingIn a 96-well plate, 20 µl of
[3H]L685458 (final concentration 1 nM) was mixed with
170 µl of solubilized
-secretase at a final 10-fold dilution of the
stock (prepared at pH 6.5) and 10 µl of a test compound (with a final
concentration in the range of 100 pM to 30 µM) in
Me2SO or in Me2SO only. The mixture was incubated on a
plate shaker at room temperature for 30 min, and then filtered through Beckman
GF/B filters (soaked in 0.6% polyethyleneimmine) using a Packard
Filtermate-196 and washed six times with 5 mM Tris-HCl, pH 7.4.
After the filters dried, 30 µl of Microscint20 (Packard) was added to each
well, and the plates were counted on a Packard TopCount. To assay binding of
the radioligand itself, the same conditions were used as in the ligand
competition studies, except that 12 concentrations (between 0.4 pM
and 12 nM) of [3H]L685458 were used in the absence of
test compound.
Data AnalysisTime courses of
-secretase reaction as
monitored by ECL assay were analyzed by linear regression to obtain the slopes
in counts/min. These values were then converted to pM/min by an ECL
assay of an A
40 standard containing C100 at the same concentration as
used in the reaction.
IC50, the inhibitor concentration at which the enzyme activity
is inhibited by 50%, was obtained by fitting inhibition data to,
 | (Eq. 1) |
where %Act is calculated by dividing the ECL signal of a reaction
mixture in the presence of inhibitor with the ECL signal obtained in its
absence.
For inhibitor cross-competition analyses, the theoretical inhibition
kinetics of two inhibitors, I1 and I2, is given by,
 | (Eq. 2) |
where
0 is the initial rate in the absence of inhibitor,
Ki1 and Ki2 are inhibition constants
for I1 and I2, respectively, and
is the constant
defining the interaction between the two inhibitors. The value of
equals unity for binding of the two inhibitors to enzyme independently. An
value of <1 indicates that binding of one inhibitor facilitates the
binding of the other. A value of >1 indicates that binding of one inhibitor
makes the binding of the other more difficult. An infinitely large
indicates that bindings of the two inhibitors are mutually exclusive or
competitive. In the latter case, Equation
2 reduces to,
 | (Eq. 3) |
which describes competitive binding. The pattern of binding of two inhibitors
may be analyzed graphically by using the reciprocal of
Equation 2, which is a linear
function of 1/
versus [I1], as given by,
 | (Eq. 4) |
Changes in [I2] will have a slope effect if
is close to
unity but will have no slope effect if it is infinitely large. Therefore, if
the two inhibitors bind simultaneously to the enzyme, the reciprocal plots
will intercept on, or to the left of, the 1/
axis. If binding of one
inhibitor excludes the binding of the other, the reciprocal plots will be a
set of parallel lines.
Data of radioligand [3H]L685458 binding to detergent extracts of
HeLa cell membranes were analyzed by,
 | (Eq. 5) |
where [L685458]b is the concentration of bound L685458,
[L685458]bmax is the concentration of maximum bound ligand, and
Kd is the dissociation constant for the binding.
The ligand displacement data were analyzed by,
 | (Eq. 6) |
where [L685458]t is the total concentration of L685458,
and Ki is the dissociation constant for I.
 |
RESULTS
|
---|
Inhibitor Cross-competition KineticsThe cross-competition
method can provide information on whether enzyme inhibition in the presence of
two inhibitors may arise from simultaneous binding to independent sites or
mutually exclusive binding to a single site or overlapping sites on the
enzyme. In this study, all possible binary combinations of the four inhibitors
(PME, L685458, arylsulfonamide 1, and benzodiazepine 2, shown in
Fig. 1) were investigated using
this approach. The data were analyzed according to
Equation 2 to extract the kinetic
constants, and the reciprocal 1/
was plotted against [I1] to
visualize the slope effect from varying [I2].
Variation of [PME] in both directions in a binary combination was first
performed to validate the method of inhibitor cross-competition kinetics for
-secretase. As anticipated, variation of [PME] in one direction had no
effect on the slope of the reciprocal plot obtained from variation of [PME] in
the other direction, producing parallel plots as shown in
Fig. 2A. The same
parallel pattern was observed for the cross-competition between PME and
L685458 (Fig. 2B),
indicating competitive binding of PME and L685458. This result is consistent
with the fact that both PME and L685458 are transition state isosteres, and
therefore, are expected to bind to the same catalytic site of
-secretase. Cross-competition pairs formed between L685458 and either
of the small molecule inhibitors produced an intercepting, or non-competitive,
binding pattern (Fig. 3),
indicating that the small molecule inhibitors bind to one or two sites
different from the site for the transition state analogs. The pattern for the
kinetics of cross-competition between 1 and 2 was, however,
parallel (Fig. 4), suggesting
that it is a single site to which both 1 and 2 bind. The
relatively small size of the
constants (1.72.5) obtained for
the competitions between L685458 and the small molecule inhibitors 1 or
2 (Table I) indicated
that binding of L685458 had little effect on binding of the small molecule
inhibitors at a different site. The cross-competition between PME and 1
or 2 also displayed an intercepting, non-competitive pattern
(Fig. 5). However, the size of
the
values was relatively large (
= 717,
Table I), suggesting that
binding of PME affected somewhat the binding of 1 or 2 to the
other binding site, making it more difficult for 1 or 2 to
bind.
View this table:
[in this window]
[in a new window]
|
TABLE I Summary of kinetic inhibition constants obtained from inhibitor
cross-competition kinetics
Cross-competition kinetics was performed at pH 6.5, 22 °C, with a
combination of two inhibitors at various concentrations. The data were
analyzed by non-linear least squares analysis using Equations
2 or
3 to obtain
Ki1 and Ki2, the
inhibition constants for inhibitors 1 or 2, respectively.
|
|
Competition Ligand BindingCompetition ligand binding was
conducted to characterize further the interaction of different
-secretase inhibitors with the enzyme. [3H]L685458 was
synthesized and used as the radioligand. Binding to the enzyme was assayed by
measuring the counts associated with the enzyme complex captured on
filter-bottom 96-well microplates (see under "Experimental
Procedures"). The time courses for binding of [3H]L685458 (2
nM) and its displacement in the presence of cold L685458 yielded a
kon of 0.069 ± 0.01
nM1 min1
(or 0.14 ± 0.02 min1 at 2 nM
[3H]L685458) (data not shown) and a koff of
0.017 ± 0.006 min1 (data not shown), which
translate into a Kd of
0.3 nM.
This Kd value is in reasonable agreement with the
Kd and Ki values
from ligand binding and displacement experiments or IC50 values
obtained from enzyme inhibition (Table
II). An initial test indicated that maintaining a signal level
sufficient for statistically meaningful analysis required an enzyme
concentration that was no more than a 10-fold dilution of the stock, as
compared with the 2040-fold dilution used in inhibitor
cross-competition kinetics. All subsequent ligand binding assays were then
performed at this enzyme concentration. This initial test also indicated that
for ligand displacement, 1 nM [[3H]L685458] was
sufficient, and this concentration of radioligand was used for all the ligand
displacement experiments.
View this table:
[in this window]
[in a new window]
|
TABLE II Summary of constants for ligand binding and displacement and inhibition
of -secretase at different enzyme concentrations
Ligand binding and displacement experiments were performed at [E]
= 2.8 nM, and IC50 values were obtained with
[E] = 0.7 or 2.8 nM at pH 6.5, 22 °C. The
Kd was obtained by non-linear least squares
analysis of data depicted in Fig.
6A according to Equation
5, and the Ki values were obtained by
analyzing data depicted in Fig.
6B using Equation
6. The IC50 values were the results of
analyzing inhibition data using Equation
1.
|
|

View larger version (15K):
[in this window]
[in a new window]
|
FIG. 6. Ligand binding and competition. A, [3H]L685458
binding to solubilized -secretase. The circles represent
experimental data, and the solid line represents the theoretical
values calculated using Equation
5 with [[3H]L685458]bmax = 2.9 nM
and Kd = 1 nM
(Table II). B,
displacement of [3H]L685458 by C100 ( ), PME ( ), L685458
( ), 1 ( ), and 2 ( ) at pH 6.5, 22 °C. The
solid lines are theoretical values calculated using
Equation 6 with
[[3H]L685458]t = 1nM,
Kd = 1nM and the
Ki values listed in
Table II. C,
[3H]L685458 binding to solubilized -secretase in the absence
( ) and presence of 2 ( ) or PME ( ). The solid
lines represent the theoretical values calculated using
Equation 5 with parameters listed
in Table III.
|
|
The ligand binding curve obtained with a range of radioligand
concentrations is depicted in Fig.
6A. Analysis of this data set using a hyperbolic equation
(Equation 5) yielded a
[[3H]L685458]bmax of 2.8 nM (data not shown),
with an apparent Kd value of 1.0 ± 0.3
nM (Table II). This
apparent Kd is comparable with the
Ki value (1.9 ± 1.1 nM,
Table II) obtained from
displacement of the radioligand with the unlabeled ligand
(Fig. 6B). Since the
calculated maximum concentration of bound ligand is used to determine the
concentration of enzyme active sites, the value for
[[3H]L685458]bmax obtained in the ligand binding
experiments indicated that the active enzyme concentration was about 2.8
nM. The fact that this value was close to the apparent
Kd suggested that a partial depletion of free
ligand might have happened at some ligand concentrations used. Thus, the
Kd obtained in this manner (using
Equation 5) might have somewhat
underestimated the true affinities of the radioligand in these experiments, as
well as the Ki values to be obtained later (using
Equation 6) from subsequent
competition ligand binding experiments. However, under partial ligand
depletion conditions, such simple analyses would only have small perturbations
in the true Ki values and would not have altered
the conclusions regarding the sites of inhibitor binding. In support of this,
the Ki values of 1.01.9 nM
(Table II) obtained from the
ligand binding or competition were statistically indistinguishable from the
Ki values of 1.32.7 nM
(Table I) obtained from kinetic
inhibitor cross-competition, in which the enzyme concentration was
24-fold less than that used in ligand binding. Therefore, no data
treatment with quadratic (for ligand binding) or cubic equations (for ligand
competition) was applied to maintain the simplicity for data analysis.
View this table:
[in this window]
[in a new window]
|
TABLE III Summary of constants for ligand binding in the presence of 2 and
PME
These experiments were performed at [E] = 2.8 nM in the
absence or presence of 1 µM 2 or 30 µM PME.
Kd was obtained by non-linear least squares
analysis of data depicted in Fig.
6C according to Equation
5. Kd(calculated) was calculated
using Kd(calculated) =
Kd(L685458 alone)(1 +
[I]/Ki), where I is either 2 or PME and
Ki is the constant obtained from ligand
displacement experiments (Table
II).
|
|
There was no observed displacement of L685458 by C100, the substrate used
for kinetic studies, at concentrations up to 5 µM, a value much
greater than the Km for C100. This is consistent
with the previous finding that transition state analogs inhibit
-secretase non-competitively
(34).
Surprisingly, the Ki of 8.3 ± 2.3
µM (Table II) for
PME in the displacement of L685458 was significantly greater than the values
of 0.670.83 µM (Table
I) obtained from the kinetic cross-competition. Radioligand
binding performed in the presence of PME (30 µM) shifted the
observed Kd for radioligand accordingly without
perturbing Bmax (Fig.
6C and Table
III). The size of the observed Kd is
similar to the Kd calculated assuming binding of
PME and L685458 at the same site. This result appeared to provide further
support for the competitive binding between L685458 and PME, albeit suggesting
that this competitive ligand binding occurred with a significantly higher
Ki value than inhibitor cross-competition.
Recognizing that the only major difference in experimental conditions
between kinetic cross-competition and ligand binding or displacement was the
enzyme concentration (0.71.4 nM in kinetic inhibition and
2.8 nM in ligand binding), we assessed in more detail whether such
differences in enzyme concentration have any effect on the enzymatic and
inhibition kinetics, thereby causing the observed discrepancy in the
Ki values obtained from inhibitor
cross-competition and ligand displacement. As shown in
Fig. 7, the enzyme activity was
linear with [E] up to
1.5 nM and then leveled off at
higher enzyme concentrations. This perturbation in enzyme activity at higher
enzyme concentrations resulted in a 15-fold increase in the IC50
value for PME (Table II),
indicating that the inflated Ki value for PME
obtained from ligand displacement was likely due to reduced affinity of the
more concentrated enzyme for PME. The IC50 value for L685458,
however, was essentially identical at both enzyme concentrations, as were the
values for 1 or 2 (Table
II). These data suggested that the increased [E] might
have had a subtle structural effect in the binding site for the transition
state analogs that was particular for PME. Taking this perturbation into
account, the ligand competition results do appear to support the conclusion
from inhibitor cross-competition kinetics of L685458 and PME binding at the
same site. These results also demonstrate the importance of validating assay
conditions thoroughly when developing series of inhibitors or studying kinetic
mechanisms.
The competition ligand binding assay was subsequently applied to
investigate whether bound [3H]L685458 could be displaced with the
small molecule inhibitors 1 and 2. As shown in
Fig. 6B, 1 did
not reduce the bound radioligand at any of the inhibitor concentrations used
(up to 30 µM), indicating that 1 did not bind at the site
for L685458. On the other hand, 2 did show displacement of bound
[3H]L685458 (Fig.
6B). This observed ligand competition by 2 was
also supported by shifting of the observed Kd for
radioligand in the presence of 1 µM 2
(Fig. 6C and
Table III). However, the
Ki value of 130 ± 50 nM
(Table II) for displacement of
L685458 by 2 was much greater than the Ki
value of 1.22.5 nM (Table
I) obtained from inhibitor cross-competition kinetics or the
IC50 value of 0.6 nM from inhibition obtained at the
same enzyme concentration (2.8 nM)
(Table II). Therefore,
inhibition of
-secretase by 2 could not have occurred via
binding to the site for L685458 but rather via binding of 2 to a
different site. These data are consistent with the results from
cross-competition kinetics that 1 and 2 bind at a site different
from the site for the transition state isosteres PME and L685458.
 |
DISCUSSION
|
---|
The non-competitive inhibition of
-secretase by the transition state
isosteres PME and L685458, as reported previously
(34), and the lack of observed
displacement of L685458 by C100 as shown in the present study, are consistent
with a mechanism in which substrate binding and catalysis occur at different
locations on the enzyme and the catalysis requires a substrate movement after
binding at the substrate binding site that swings the scissile bond into the
catalytic site (34). This
mechanism is supported by co-purification of C83, an alternative substrate of
-secretase of APP-derived fragment, using affinity columns prepared
with derivatized transition state analogs
(36). Delineation of the site
on PS1 for binding of APP/C99 that is away from the proposed catalytic site on
PS1 has also led to the proposal for spatial dissociation between substrate
binding and catalysis that requires substrate movement for catalysis
(37). By this mechanism,
binding of the transition state analogs at the catalytic site would not
necessarily interfere with the binding of substrate to the enzymatic complex
(34).
The data presented in this study from inhibitor cross-competition kinetics
and competition ligand binding indicate the presence on
-secretase of
two inhibitor binding sites, one for the transition state isosteres PME and
L685458 and the other for non-transition state small molecule arylsulfonamide
and benzodiazepine inhibitors. It is conceivable that the site on
-secretase for binding transition state analogs is the site for
catalysis. The question remaining is how the activity of
-secretase is
inhibited by the small molecule inhibitors arylsulfonamides and
benzodiazepines, which also display non-competitive inhibition of
-secretase (34) but
bind at a site different from the site for transition state analogs (this
study).
There are at least three ways to inhibit an enzyme where substrate binding
and catalysis occur at different locations: (a) interfering with
substrate binding, (b) affecting catalysis, and (c) blocking
movement of substrate into the catalytic site. The first two are common enzyme
inhibition mechanisms, whereas the third is only applicable for enzymes with
separate substrate and catalytic sites. For such enzymes, catalysis requires
movement, either by translocation or by swinging
(34), of bound substrate into
the catalytic site. Blocking substrate movement, which would display a
non-competitive kinetic inhibition, may be accomplished either by inhibitor
binding at a site located between the sites of substrate binding and catalysis
to form a direct physical blockage or by a conformational change induced upon
binding of inhibitor at a remote site.
The non-competitive inhibition of
-secretase demonstrated for
arylsulfonamides and benzodiazepines
(34) rules out the possibility
that these inhibitors either bind at the substrate binding site or induce a
conformational change in the substrate binding site when bound at a remote
site. As a result, these small molecule inhibitors do not seem to inhibit
-secretase by interfering with substrate binding. Rather, the
non-competitive inhibition displayed by these small molecule inhibitors
(34) is consistent with
inhibition by directly affecting catalysis. However, it is difficult to
explain this in the context of their non-competitive cross-competition with
the transition state isosteres (Figs.
3 and
5), which presumably (although
not necessarily) bind at the catalytic site
(34). In particular, the fact
that the interaction factor
is near unity for cross-competition
between L685458 and 1 or 2
(Table I) suggests that binding
of the small molecule inhibitors causes little structural perturbation around
the catalytic site where L685458 binds. As such, for the mechanism of
affecting catalysis to be valid for the small molecule inhibitors, it is
necessary either that the catalytic site be so spacious that it could
accommodate both L685458 and a small molecule inhibitor simultaneously (the
physical mechanism, Fig.
8A) or that the catalytic site must undergo such a
delicate, partial structural distortion upon inhibitor binding at a remote
site that the catalysis is prevented, whereas the integrity of the catalytic
site for binding of L685458 is preserved
(Fig. 8A, the
conformational mechanism). Although these possibilities cannot be ruled out,
in the absence of additional data, these scenarios do not seem very
likely.

View larger version (34K):
[in this window]
[in a new window]
|
FIG. 8. Proposed mechanisms of inhibition of -secretase by small
molecule inhibitors. -Secretase is a high molecular weight
multicomponent complex containing presenilin, nicastrin, aph-1, and pen-2. For
simplicity, presenilin is depicted as the catalytic subunit containing
separate substrate binding and catalytic sites. The site for catalysis is also
the binding site for transition state isosteres PME or L685458. The location
for the second inhibitor site, a site for non-transition state small molecule
inhibitors, is undefined. A, inhibition of -secretase by
affecting catalysis. This may be achieved either by a conformational mechanism
where the catalytic site is distorted by inhibitor binding at a site away from
the catalytic site or by a physical mechanism in which the inhibitor occupies
a portion of the catalytic site, in both cases, without perturbing binding of
the transition state analog L685458. B, inhibition of
-secretase by blocking substrate movement. The catalytic reaction
proceeds with substrate binding at the substrate binding site followed by a
motion of the part that contains the scissile bond into the catalytic site and
subsequent bond cleavage (34).
The reaction is inhibited either through a physical mechanism where the
inhibitor binds at the region between the substrate binding and catalytic
sites to form a physical blockage to the substrate movement or through a
conformational mechanism by which the structure between the substrate and
catalytic sites is distorted (stretched as it is depicted here as an example)
such that the substrate movement into the catalytic site is no longer
possible.
|
|
As
-secretase may catalyze its reaction through separated substrate
binding and catalytic sites
(34,
36,
37, and this study), it is
possible that inhibition may be achieved by blocking the movement of substrate
into the catalytic site. Analogous to the mechanism of affecting catalysis,
blocking substrate movement may be achieved in at least two ways: 1) through a
physical mechanism, where the inhibitor occupies a site located between the
substrate binding site and the catalytic site to form a physical blockage to
the substrate motion or 2) by a conformational mechanism, in which a
structural distortion in the region between the substrate binding and
catalytic sites results from binding of inhibitor remotely
(Fig. 8B).
In addition to APP processing,
-secretase catalyzes endoproteolysis
of Notch receptor
(2326),
a fundamentally important signal transduction molecule crucial for the
development of all metazoans
(27). Because of its
biological importance, the cleavage activation of Notch by
-secretase
is likely to be tightly regulated and substrate-specific given that in
addition to APP, a number of other type I transmembrane proteins, such as
ErbB4 (30,
31), E-cadherin
(32), and CD44
(33), are also
-secretase substrates. Generally, substrate specificity of an enzyme
cannot be modified by affecting catalysis alone since all substrates access
the same catalytic machinery of the enzyme. For enzymes with substrate binding
and catalysis at the same site, the only way of modulating substrate
specificity is by modifying the affinity for different substrates. For enzymes
with separated substrate binding and catalytic sites, substrate specificity
may also be controlled by affecting substrate movement, if such a motion is
substrate-specific. This is straightforward conceptually if multiple substrate
binding sites exist, a concept that remains to be tested directly. It is a bit
more difficult to conceptualize whether only a single substrate binding site
exists, although still feasible in principle depending on the higher order
structural features associated with the movement of different substrates into
the catalytic site.
The feasibility of substrate selectivity for
-secretase is supported
by these studies and two other recent reports
(38,
39). Certain compounds
containing the same 3,5-difluorophenyl substituent as 2
(38) and certain
arylsulfonamides and benzodiazepines
(39) have been reported to
inhibit selectively APP processing over Notch cleavage by
-secretase.
Since arylsulfonamides and benzodiazepines do not interfere with substrate
binding (34), this observed
substrate specificity for APP over Notch supports the hypothesis of blocking
substrate movement as the mechanism of
-secretase inhibition. In
contrast, the transition state isostere L685458 has been shown to inhibit APP
and Notch processing with equal potency
(39), consistent with the
principle that substrate selectivity cannot result from affecting catalysis
alone. Further investigation is needed to determine precisely how
-secretase is inhibited by the small molecule arylsulfonamide and
benzodiazepine inhibitors and how these affect the processing of various
substrates.
The separation of substrate binding and catalysis potentially offers a
facile means for regulating and controlling the substrate specificity of
-secretase biologically. For drug discovery, separation of substrate
binding and catalysis presents an opportunity for developing
substrate-specific inhibitors. Such inhibitors for
-secretase, if
discovered, would be more beneficial in principle to Alzheimer patients,
sparing pathways that are otherwise unrelated to disease pathogenesis.
 |
FOOTNOTES
|
---|
* The costs of publication of this article were defrayed in part by the
payment of page charges. This article must therefore be hereby marked
"advertisement" in accordance with 18 U.S.C. Section 1734
solely to indicate this fact. 
To whom correspondence should be addressed: Dept. of Lead Discovery,
AstraZeneca Pharmaceuticals, 1800 Concord Pike, Wilmington, DE 19850. Tel.:
302-886-8137; Fax: 302-886-4983; E-mail:
gaochao.tian{at}astrazeneca.com.
1 The abbreviations used are: A
,
-amyloid; APP, amyloid precursor
protein; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate;
CHAPSO, 3-[(3-cholamidopropyl)dimethylammonio]-2-hydoxy-1-propanesulfonate;
MES, 4-morpholineethanesulfonic acid; C99, APP C-terminal fragment of 99 amino
acid residues; C100, recombinantly produced C99 containing an additional
methionine residue at its N terminus; ECL, electrochemiluminescence;
G
R, goat anti-rabbit; PME, pepstatin A methylester; R
A
40,
rabbit anti-A
40. 
 |
ACKNOWLEDGMENTS
|
---|
We thank Dr. Deborah S. Hartman for sponsoring the work described in this
study and Dr. Clay W. Scott for critical reading of the manuscript. We are
grateful for Drs. Mark A. Sylvester, Philip D. Edwards, Michael T. Klimas,
Thomas R. Simpson, James M. Woods, James Kang, Peter R. Bernstein, Robert D.
Dedinas, Bruce T. Dembofsky, Michael Balestra, Jingbo Yan, and James D.
Rosamond for contributing to the syntheses of compounds 1 and 2.
We are indebted to Dr. John Zysk for providing HeLa cell membranes used in
preparing detergent extracts of
-secretase.
 |
REFERENCES
|
---|
- Sisodia, S. S., Annaert, W., Kim, S. H., and De Strooper, B.
(2001) Trends Neurosci.
24, Suppl. 11,
S2S6[CrossRef][Medline]
[Order article via Infotrieve]
- Klein, W. L. (2002) Neurochem.
Int. 41,
345352[CrossRef][Medline]
[Order article via Infotrieve]
- Lin, X., Keels, G., Wu, S., Downs, D., Dash, A., and Tang, J.
(2000) Proc. Natl. Acad. Sci. U. S. A.
97,
14561460[Abstract/Free Full Text]
- Yean, R., Blenkowski, M. J., Shu, M. E., Miao, H., Tory, M. C.,
Pauley, A. M., Brashler, J. R., Stratman, N. C., Mathews, W. R., Buhl, A. E.,
Catrer, D. B., Tomasselli, A. G., Parodi, L. A., Heinricson, R. L., and
Gurney, M. E. (1999) Nature
402,
533537[CrossRef][Medline]
[Order article via Infotrieve]
- Vassar, R., Bennett, B. D., Babu-Khan, S., Maendiaz, E. A., Denis,
P., Teplow, D. B., Ross, S., Amarante, P., Loeloff, R., Luo, Y., Fisher, S.,
Fuller, J., Edenson, S., Lile, J., Jarosinski, M. A., Biere, A. L., Curran,
E., Burgess, T., Louis, J.-C., and Collins, F. (1999)
Science 286,
735741[Abstract/Free Full Text]
- Hong, L., Keels, G., Lin, X., Wu, S., Terzyan, S., Ghosh, A. K.,
Zhang, X. C., and Tang, J. (2000) Science
290,
150153[Abstract/Free Full Text]
- Zhang, L., Song, L., Terracina, G., Liu, Y., Pramanik, B., and
Parker, E. (2001) Biochemistry
40,
50495055[CrossRef][Medline]
[Order article via Infotrieve]
- Pinnix, I., Musunuru, U., Tun, H., Sridharan, A., Golde, T.,
Eckman, C., Ziani-Cherif, C., Onstead, L., and Sambamurti, K.
(2001) J. Biol. Chem.
276,
481487[Abstract/Free Full Text]
- Murphy, M. P., Hickman, L. J., Eckman, C. B., Uljon, S. N., Wang,
R., and Golde, T. E. (1999) J. Biol.
Chem. 274,
1191411923[Abstract/Free Full Text]
- Wolfe, M. S., Xia, W., Moore, C. L., Leatherwood, D. D.,
Ostazewski, B., Tahmati, T., Donkor, I. O., and Selkoe, D. J.
(1999) Biochemistry
38,
47204727[CrossRef][Medline]
[Order article via Infotrieve]
- Wolfe, M. S., and Haass, C. (2001) J. Biol.
Chem. 276,
54135416[Free Full Text]
- Evin, G., Sharples, R. A., Weidemann, A., Reinhard, F. B., Carbone,
V., Culvenor, J. G., Holsinger, R. M., Sernee, M. F., Beyreuther, K., and
Masters, C. L. (2001) Biochemistry
40,
83598368[CrossRef][Medline]
[Order article via Infotrieve]
- Yu, G., Chen, F., Levesque, G., Nishimura, M., Zhang, D.-M.,
Levesque, L., Rogaeva, E., Xu, D., Liang, Y., Duthie, M., St. George-Hyslop,
P. H., and Fraser, P. E. (1998) J. Biol.
Chem. 273,
1647016475[Abstract/Free Full Text]
- Li, Y.-M., Lai, M.-T., Xu, M., Huang, Q., DiMuzio-Mower, J.,
Sardana, M. K., Shi, X.-P., Yin, K. C., Shafer, J. A., and Gardell, S. J.
(2000) Proc. Natl. Acad. Sci. U. S. A.
97,
61386143[Abstract/Free Full Text]
- Wolfe, M. S., Xia, W. M., Ostaszewski, B. L., Diehl, T. S.,
Kimberly, W. T., and Selkoe, D. J. (1999)
Nature 398,
513517[CrossRef][Medline]
[Order article via Infotrieve]
- Li, Y. M., Xu, M., Lai, M. T., Huang, Q., Castro, J. L.,
DiMuzio-Mower, J., Harrison, T., Lellis, C., Nadin, A., Neduvelil, J. G.,
Register, R. B., Sardana, M. K., Shearman, M. S., Smith, A. L., Shi, X. P.,
Yin, K. C., Shafer, J. A., and Gardell, S. J. (2000)
Nature 405,
689694[CrossRef][Medline]
[Order article via Infotrieve]
- Esler, W. P., Kimberly, W. T., Ostaszewski, B. L., Diehl, S. T.,
Moore, C. L., Tsai, J.-Y., Rahmati, T., Xia, W., Selkoe, D. J., and Wolfe, M.
S. (2002) Nat. Cell Biol.
2,
428434[CrossRef]
- Seiffert, D., Bradley, J. D., Rominger, C. M., Rominger, D. H.,
Yang, F., Meredith, Jr., J. E., Wang, Q., Roach, A. H., Thompson, L. A.,
Spitz, S. M., Higaki, J. N., Prakash, S. R., Combs, A. P., Copeland, R. A.,
Arneric, S. P., Hartig, P. R., Robertson, D. W., Cordell, B., Stern, A. M.,
Olson, R. E., and Zaczek, R. (2002) J. Biol.
Chem. 275,
3408634091[Abstract/Free Full Text]
- Yu, G., Nishimura, M., Arawaka, S., Levitan, D., Zhang, L. L.,
Tandon, A., Song, Y. Q., Rogaeva, E., Chen, F. S., Kawaral, T., Supala, A.,
Levesque, L., Yu, H., Yang, D. S., Holmes, E., Millman, P., Liang, Y., Zhang,
D. M., Xu, D. H., Sato, C., Rogaev, E., Smith, M., Janus, C., Zhang, Y.,
Aebersold, R., Farrer, L., Sorbi, S., Bruni, A., Fraser, P., and St.
George-Hyslop, P. (2000) Nature
407,
4854[CrossRef][Medline]
[Order article via Infotrieve]
- Francis, R., McGrath, G., Zhang, J., Ruddy, D. A., Sym, M., Apfeld,
J., Nicoll, M., Maxwell, M., Hai, B., Ellis, M. C., Parks, A. L., Xu, W., Li,
J., Gurney, M., Myers, R. L., Himes, C. S., Hiebsch, R., Ruble, C., Nye, J.
S., and Curtis, D. (2002) Dev. Cell
3,
8597[Medline]
[Order article via Infotrieve]
- Lee, S. F., Shah, S., Li, H., Yu, C., Han, W., and Yu, G.
(2002) J. Biol. Chem.
277,
4501345019[Abstract/Free Full Text]
- Steiner, H., Winkler, E., Edbauer, D., Prokop, S., Basset, G.,
Yamasaki, A., Kostka, M., and Haass, C. (2002) J.
Biol. Chem. 277,
3906239065[Abstract/Free Full Text]
- Ye, Y., Lukinova, N., and Fortini, M. E. (1999)
Nature 389,
525529
- Struhl, G., and Greenwald, I. (1999)
Nature 389,
522525[CrossRef]
- De Strooper, B. Annaert, W., Cupers, P., Saftig, P., Craessaerts,
K., Mumm, J. S., Schroeter, E. H., Schrijvers, V., Wolfe, M. S., Ray, W. J.,
Goate, A., and Kopan, R. (1999) Nature
389,
518522
- Song, W., Nadeau, P., Yuan, M., Yang, X., Shen, J., and Yankner, B.
A. (1999) Proc. Natl. Acad. Sci. U. S. A.
96,
69596963[Abstract/Free Full Text]
- Artavanis-Tsakonas, S., Rand, M. D., and Lake, R. J.
(1999) Science
284,
770776[Abstract/Free Full Text]
- Schroeter, E. H., Kisslinger, J. A., and Kopan, R.
(1998) Nature
393,
382386[CrossRef][Medline]
[Order article via Infotrieve]
- Struhl, G., and Adachi, A. (1998) Mol.
Cell 6,
625636
- Ni, C.-Y., Murphy, M. P., Golde, T. E., and Carpenter, G.
(2001) Science
294,
21792181[Abstract/Free Full Text]
- Lee, H.-J., Jung, K.-M., Huang, Y. Z., Bennett, L. B., Lee, J. S.,
Mei, L., and Kim, T.-W. (2002) J. Biol.
Chem. 277,
63186323[Abstract/Free Full Text]
- Marambaud, P., Shioi, J., Serban, G., Georgakopoulos, A., Sarner,
S., Nagy, V., Baki, L., Wen, P., Efthimiopoulos, S., Shao, Z., Wisniewski, T.,
and Robakis, N. K. (2002) EMBO J.
21,
19481956[Abstract/Free Full Text]
- Lammich, S., Okochi, M., Takeda, M., Kaether, C., Capell, A.,
Zimmer, A.-K., Edbauer, D., Walter, J., Steiner, H., and Haass, C.
(2002) J. Biol. Chem.
277,
4475444759[Abstract/Free Full Text]
- Tian, G., Sobotka-Briner, C., Zysk, J., Liu, X., Birr, C.,
Sylvester, M. A., Edwards, P. D., Scott, C. W., and Greenberg, B. D.
(2002) J. Biol. Chem.
277,
3149931505[Abstract/Free Full Text]
- Shearman, M. S., Beher, D., Clarke, E. E., Lewis, H. D., Harrison,
T., Hunt, P., Nadin, A., Smith, A. L., Stevenson, G., and Castro, J. L.
(2000) Biochemistry
39,
86988704[CrossRef][Medline]
[Order article via Infotrieve]
- Esler, W. P., Kimberly, W. T., Ostaszewski, B. L., Ye, W. J.,
Diehl, T. S., Selkoe, D. J., Wolfe, M. S. (2002) Proc.
Natl. Acad. Sci. U. S. A. 99,
27202725[Abstract/Free Full Text]
- Annaert, W. G., Esselens, C., Baert, V., Boeve, C., Snellings, G.,
Cupers, P., Craessaerts, K., and De Strooper, B. (2001)
Neuron 32,
579589[Medline]
[Order article via Infotrieve]
- Teall, M. R., and Stortford, B. (January 31, 2002) U.
S. Patent, US 2002/0013315 A1
- Liu, X. F., Ghanekar, S. V., Jiang, Q., Aharony, D., and Greenberg,
B. D. (2002) Society for Neuroscience, 32nd Annual
Meeting, Orlando, November 27, 2002, Abstract
Viewer/Itinerary Planner Online, Program No. 296.2, Society for Neuroscience,
Washington, D. C.