P2-substituted N-acylprolylpyrrolidine inhibitors of prolyl oligopeptidase: Biochemical evaluation, binding mode determination, and assessment in a cellular model of synucleinopathy

Article abstract of DOI:10.1021/jm301060g

We have investigated the effect of regiospecifically introducing substituents in the P2 part of the typical dipeptide derived basic structure of PREP inhibitors. This hitherto unexplored modification type can be used to improve target affinity, selectivity, and physicochemical parameters in drug discovery programs focusing on PREP inhibitors. Biochemical evaluation of the produced inhibitors identified several substituent types that significantly increase target affinity, thereby reducing the need for an electrophilic warhead functionality. Pronounced PREP specificity within the group of Clan SC proteases was generally observed. Omission of the P1 electrophilic function did not affect the overall binding mode of three representative compounds, as studied by X-ray crystallography, while the P2 substituents were demonstrated to be accommodated in a cavity of PREP that, to date, has not been probed by inhibitors. Finally, we report on results of selected inhibitors in a SH-SY5Y cellular model of synucleinopathy and demonstrate a significant antiaggregation effect on α-synuclein.

Full text of DOI:10.1021/jm301060g

Article
pubs.acs.org/jmc
P2-Substituted N‑Acylprolylpyrrolidine Inhibitors of Prolyl
Oligopeptidase: Biochemical Evaluation, Binding Mode
Determination, and Assessment in a Cellular Model of
Synucleinopathy
Pieter Van der Veken,*,† Vilmos Fulop,‡ Dean Rea,‡ Melanie Gerard,§ Roos Van Elzen,∥ Jurgen Joossens,†
̈
̈
Jonathan D. Cheng,⊥ Veerle Baekelandt,# Ingrid De Meester,∥ Anne-Marie Lambeir,∥
and Koen Augustyns†
†Laboratory of Medicinal Chemistry, Department of Pharmaceutical Sciences, University of Antwerp, Universiteitsplein 1, B-2610
Antwerp, Belgium
‡School of Life Sciences, University of Warwick, Coventry CV4 7AL, U.K.
§Laboratory of Biochemistry, Interdisciplinary Research Facility Life Sciences, KU Leuven KULAK, B-8500 Kortrijk, Belgium
∥Laboratory of Medical Biochemistry, Department of Pharmaceutical Sciences, University of Antwerp, Universiteitsplein 1, B-2610
Antwerp, Belgium
⊥Fox Chase Cancer Center, 333 Cottman Avenue, Philadelphia, Pennsylvania 19111, United States
#Laboratory for Neurobiology and Gene Therapy, KU Leuven, Kapucijnenvoer 33, B-3000 Leuven, Belgium
S
* Supporting Information
ABSTRACT: We have investigated the effect of regiospecifi-
cally introducing substituents in the P2 part of the typical
dipeptide derived basic structure of PREP inhibitors. This
hitherto unexplored modification type can be used to improve
target affinity, selectivity, and physicochemical parameters in
drug discovery programs focusing on PREP inhibitors.
Biochemical evaluation of the produced inhibitors identified
several substituent types that significantly increase target
affinity, thereby reducing the need for an electrophilic
“warhead” functionality. Pronounced PREP specificity within
the group of Clan SC proteases was generally observed.
Omission of the P1 electrophilic function did not affect the overall binding mode of three representative compounds, as studied
by X-ray crystallography, while the P2 substituents were demonstrated to be accommodated in a cavity of PREP that, to date, has
not been probed by inhibitors. Finally, we report on results of selected inhibitors in a SH-SY5Y cellular model of synucleinopathy
and demonstrate a significant antiaggregation effect on α-synuclein.
However, not all results of these experiments have been
equivocal, and so far, no clear picture has emerged that defines
the enzyme’s position in the known pathways of peptide
processing and degradation.8 Over the past decade, functional
PREP research has broadened its scope. It now no longer
exclusively focuses on the enzyme’s hydrolase activity but also
addresses potential intracellular actions that affect signal
transduction, protein secretion, aging, neuronal development,
and differentiation.9−20 Several of these studies indicate that
protein−protein interactions underlay the role of PREP in the
central nervous system, and the list of proteins that have been
found to interact with PREP includes the growth associated
protein GAP43, β-tubulin, and α-synuclein. However, our own
INTRODUCTION
■
Prolyl oligopeptidase (EC 3.4.21.29, PREP, PO, POP) was first
discovered as an oxytocin degrading, post-proline cleaving
peptidase.1 Apart from oxytocin, numerous other peptides have
since been reported as substrates for PREP in vitro and in
vivo.2,3 Since PREP activity in the brain is high and several
neuropeptides are cleaved by the enzyme, a physiological role
in neuropeptide processing and turnover is widely predicted.
Several PREP inhibitors developed during the 1990s were
found to have a positive effect on memory, learning, and
cognition in animal models of Alzheimer’s disease and brain
injury, and at least two of them have been tested in humans.2,6
Nonetheless, to date, no PREP inhibitors have been approved
for therapeutic use. The in vivo effect of these inhibitors was
ascribed to the prolonged action of specific neuropeptides,
known to be involved in memory, learning, and cognition.7
Received: July 20, 2012
Published: November 2, 2012
© 2012 American Chemical Society
9856
dx.doi.org/10.1021/jm301060g | J. Med. Chem. 2012, 55, 9856−9867

Journal of Medicinal Chemistry
Article
observation that active site inhibitors of PREP can also affect
these protein−protein interactions remains enigmatic and is a
topic that calls for the mechanistic reinvestigation of in vivo
effects reported earlier.20 The recent resurgence in functional
PREP research, resulting from these findings, has already led to
the production of PREP knockout mice and a new burst in
inhibitor development.11−13
Most reported synthetic PREP inhibitors, extensively
reviewed in 2010 by Lawandi et al., possess a dipeptde or in
some cases tripeptide derived structure.21 Relevant examples
are shown in Figure 1.22 These compounds typically interact
generally alleviate DPP-selectivity issues, it also decreases the
aqueous solubility of PREP inhibitors.
In an ongoing effort to design new generations of PREP
inhibitors with improved activity, selectivity, and biopharma-
ceutical profile, we studied PREP−inhibitor complex crystal
structures reported in the literature and present in the PDB.24
In all of these structures, there is space to introduce
substituents in the P2 position that could extend beyond the
S2 binding site to probe hitherto unexplored regions of the
enzyme. We decided to test this approach using the
prolylpyrrolidine scaffold, since this structure forms the basis
of several potent PREP and DPP IV inhibitors that have
demonstrated in vivo activity in relevant animal models.4−8,25
We predicted that grafting substituents onto the P2-prolyl
residue of this scaffold could be used to (1) increase the
enzyme−inhibitor affinity and thus reduce the need for the
potentially problematic “warhead” functionality, (2) divert
compounds away from the generic dipeptide-derived structure
of most existing PREP inhibitors, hopefully leading to improved
selectivity over identified and as yet unidentified potential off-
targets of PREP inhibitors, and (3) introduce functionalities
that modify physicochemical compound parameters, such as
aqueous solubility. In this paper we describe the synthesis,
potency, and specificity of the series of P2-substituted PREP
inhibitors that were designed to meet these goals, along with
the X-ray crystal structures of PREP in complex with a selection
of these compounds. In addition, screening results are reported
for selected compounds in a cellular model of synucleinopathy.
Figure 1. Relevant literature examples of PREP-inhibitors.22a−d
DESIGN AND SYNTHESIS OF INHIBITORS
■
On the basis of the existing PDB structures of PREP−inhibitor
complexes and subsequent docking experiments, we identified
the 4-position of the P2-prolyl residue as the preferential
position to introduce substituents (Figure 2). Synthetically, we
with three substrate binding sites of the enzyme: the S1 pocket
which accommodates the substrate’s P1-proline residue, the S2
which is less well-defined and known to accommodate several
types of residues, and the S3 that is usually filled by an aliphatic
spacer attached to an aromatic group. Additionally, a substantial
fraction of compounds reported are characterized by the
presence of a P1 “warhead”: an electrophilic functional group
that is capable of covalently interacting, either reversibly or
irreversibly, with PREP’s catalytically active serine hydroxyl
group. Warhead types that have been built into PREP inhibitors
include aldehyde, keto, carbonitrile, and boronate functional
groups. Warhead functionalities do not necessarily limit the
druglikeness of inhibitors (as demonstrated, for example, by the
marketed DPP IV inhibitors vildagliptin and saxagliptin) and
have the potential to significantly increase target affinity.23
Nonetheless, the presence of these electrophiles could also be
expected to have a negative influence on inhibitor stability and
might perturb the selectivity profile of the inhibitors. The
selectivity issue certainly deserves consideration, since a
substantial number of proline-selective proteases, phylogeneti-
cally related to PREP, are present in eukaryotic genomes.23
According to the classification of Rawlings and Barrett, these
are grouped in Clan SC and include the dipeptidyl peptidases
(DPPs: DPP IV, DPP II, DPP8, and DPP9) and fibroblast
activation protein (FAP) that has mixed endopeptidase and
dipeptidyl peptidase activity. As the DPPs use a protonated,
positively charged free P2-α-amine group as an important
determinant for recognition of substrates and inhibitors, PREP
inhibitors (possessing a nonbasic, acylated P2-α-amine group)
harbor greater undesired reactivity toward FAP than DPPs.
While acylation of the P2 amine function is indeed known to
Figure 2. Generic overview of target compound structures.
opted to exploit the 1,3-dipolar Huisgen addition (“click”
chemistry) for substituent introduction while also taking
advantage of the metabolically stable triazole ring for linking
the substituent to the inhibitor core. From design studies, a
relative cis-positioning of the substituent and the proline
carboxylate residue (2S,4S configuration) was predicted to be
optimal. This configuration in our docking model also allowed
hydrogen bond formation between the enzyme’s backbone and
the N3 of the triazole ring. Nonetheless, we also decided to
synthesize a limited number of the trans analogues (possessing
the 2S,4R configuration) to validate initial observations. For the
N-acyl substituent of the P2-prolyl residue, either a
benzyloxycarbonyl (“Cbz” or “Z”) or a 4-phenylbutanoyl
residue was selected. Although probably suboptimal, their
ease of introduction and commercial availability were
considered as prime factors at this stage of research. Finally,
9857
dx.doi.org/10.1021/jm301060g | J. Med. Chem. 2012, 55, 9856−9867

Journal of Medicinal Chemistry
Article
as the P1 residue, a pyrrolidine ring not equipped with a
warhead function was chosen.
Scheme 2. Synthesis of Target Products Carrying a P1-
Carbonitrile Warhead
a
Synthesis of target compounds started with the single
(2S,4S) and (2S,4R) diastereomers of N-Boc-4-azidoproline
5a and 5b using a published procedure.26 These were coupled
to pyrrolidine using TBTU. Deprotection of the Boc group was
followed by reaction with benzyloxycarbonyl chloride or 4-
phenylbutanoyl chloride to afford intermediates 7a,b and 8.
Final products 9a−v and 10 were obtained using the Cu(I)-
catalyzed variant of the 1,3-dipolar Huisgen cycloaddition
between these intermediates and a selection of alkynes
(Scheme 1). This reaction exclusively yielded one regioisomeric
a
Scheme 1. Synthesis of Target Products 9a−v and 10
a
Reagents and conditions: (a) prolinamide hydrochloride, TBTU,
Et3N, DMF, rt, 8 h; (b) (i) TFA, DCM, rt, quant; (ii) CbzCl, Et3N,
DCM, 0 °C (R1 = Z) or phenylbutanoyl chloride, Et3N, DCM, 0 °C
(R1 = phenylbutanoyl chloride); (c) trifluoroacetic anhydride,
pyridine, DCM, 0 °C, 1 h; (d) N-Boc-propargylamine, CuI, Et3N
(trace), MeOH, rt, 24 h.
alkyne and carbonitrile function were observed in the reaction
mixture.
a
Reagents and conditions: (a) pyrrolidine, TBTU, Et3N, DMF, rt, 8 h;
(b) (i) TFA, DCM, rt, 30 min; (ii) CbzCl, Et3N, DCM, 0 °C, 2 h (R1
= Cbz) or 4-phenylbutanoyl chloride, Et3N, DCM, 0 °C, 2 h (R1 =
phenylbutanoyl chloride); (c) “alkyne”, CuI, Et3N (trace), MeOH, rt,
24 h. (d) If “alkyne” = trimethylsilylacetylene: KF, MeOH, rt, 48 h. If
“alkyne” = N-Boc-propargylamine: TFA, DCM, rt, 30 min.
RESULTS AND DISCUSSION
■
All synthesized molecules were evaluated as inhibitors of PREP,
FAP, and the dipeptidyl peptidases DPP IV, DPPII, and DPP9.
Although not reported, assay results for DPP8 can be expected
to be comparable to data obtained in experiments with DPP9,
taking into account the high degree of homology between these
enzymes.28 The prolylpyrrolidine-based PREP inhibitor 15,
reported earlier by Yoshimoto et al., and its 4-phenylbutanoyl-
containing analogue 16 (SUAM-1221), taken from Atack et al.,
were chosen as appropriate reference compounds.29 These
molecules have the same backbones that are present in the
target compounds, but they lack the substituent at the 4-
position of the P2-proline residue. Hence, they can directly
serve to evaluate the effect and impact of this substitution type.
The IC50 values for PREP inhibition of the reference
compounds obtained under our assay conditions are situated
in the higher nanomolar range and are summarized in Table 1.
Additionally, the (2S,4S)-4-azidoproline-containing intermedi-
ates 7a and 7b from Scheme 1 are present in this table.
Inhibitory potency for these compounds is comparable to
inhibitory potency obtained for the two reference compounds,
and this already indicates that some additional space is available
in the region of the PREP active center where the P2-proline
residue is accommodated.
triazole product (the 1,4-disubstituted compound), as exten-
sively documented by earlier literature studies of this reaction
type.27
Although not incorporating the strategic ambitions set out in
the Introduction, we decided to prepare a limited number of
analogues carrying a P1-carbonitrile warhead group. We
included these compounds in order to serve as additional
reference compounds for validating some of the hypotheses
made for target compounds 9a−v.
The synthesis of these compounds was accomplished in a
way that was similar to the strategy followed for the preparation
of target compounds 9a−v, but this time only analogues
containing the (2S,4S) isomer of N-Boc-4-azidoproline (5a),
expected to yield active molecules, were used. To this end, the
latter was first coupled to prolinamide hydrochloride. Then acid
deprotection of the Boc group was followed by introduction of
a 4-phenylbutanoyl (R1) substituent to yield compound 12.
Dehydration of the primary amide group delivered 13,
containing the carbonitrile warhead group. Since this
compound already contains all necessary determinants for
PREP binding, it was also evaluated as an inhibitor of the
enzyme. Finally, for putting in place the triazole ring of target
compound 14 (Scheme 2), the same experimental procedure
was used as for the corresponding P1-pyrrolidine analogues.
Under these conditions, no products of the potentially
interfering, tetrazine forming [3 + 2]-cycloaddition of the
Inhibitory potential of the prolylpyrrolidine reference
structure toward FAP and DPPs appears to be very limited
from these data, raising expectations for obtaining at least a
comparable degree of selectivity with respect to these proteins
for the target compounds. Another observation that can already
be made from the data in Table 1 and that proves to be valid as
well for the target compounds (IC50 data summarized in Table
2) is that compounds with R1= phenylbutanoyl display a
9858
dx.doi.org/10.1021/jm301060g | J. Med. Chem. 2012, 55, 9856−9867

Journal of Medicinal Chemistry
Article
Table 1. Inhibitory Potential of Reference Compounds 15 and 16 and Azide-Substituted Analogues 7a,b toward PREP, FAP,
DPP IV, DPPII, and DPP9
limited but consistently higher PREP affinity than their Z-
derivatized analogues. This difference explains why within the
series of target compounds, a higher number of inhibitors of the
former type were prepared. In cases where both the Z and
phenylbutanoyl analogues of inhibitors with an identical R2-
substituent were synthesized, the corresponding assay data are
presented in subsequent entries in Table 2. Compounds in
Table 2 are furthermore grouped according to the structural
and electronic properties of their triazole susbtituents (R2),
mainly encompassing (1) aliphatic, (2) aromatic, and (3) basic
substituent types of different steric impact.
these findings, we determined the crystal structure of PREP
complexed with an inhibitor from the halogenated subset.
Monofluorinated compound 9i was selected because, although
not the most potent in the series, replacement of a single
hydrogen atom by a very small fluorine atom already resulted in
a substantial change in enzyme affinity, with little additional
potency gained from additional changes. Specifically, we were
wondering whether F−H hydrogen bonding with the enzyme
could be involved in this (crystal structures discussed below).
Another goal of this study was to investigate whether typical
solubility-enhancing (basic) groups could be added to the 4-
position of the P2-proline ring while retaining or improving the
inhibitory potential of the parent prolylpyrrolidine structure.
Compounds 9q−v represent the effort undertaken in this
direction. In general, the assay data show that the introduction
of a range of protonatable groups in this part of the inhibitors is
tolerated. While most compounds possess IC50 values that are
close to the reference compounds’ (15 and 16), the N,N-
dimethylaminomethyl-containing analogue 9r clearly stands out
in this subset, outperforming the inhibitory potential of both
the corresponding reference molecule and its nonmethylated
counterpart 9q by an order of magnitude. With respect to the
selectivity discussion, morpholine and piperazine-susbtituted
analogues 9s−u were found to display substantial DPP9
affinity. This is highly surprising, given the N-acylation of the α-
amine function of the P2-prolyl residue in these inhibitors. A
protonated amine function, known to be engaged in salt-
bridging with the target protein, is present in all DPP inhibitors
reported to date, either as the P2-α-amino group in peptide
derived inhibitors or in a position that is topologically
equivalent to the latter in non-peptide-derived compounds.
Proper investigation of these findings and comparison to SAR
data we have generated in the past for individual DPPs will
follow in due course.28
Compound 9a, prepared from 9b by desilylation using CsF
in MeOH, represents the simplest example (R2= H) in the
series and offers the opportunity to evaluate whether the
triazole ring (present in all target molecules) is indeed involved
in stabilizing interactions with the enzyme as observed during
docking-based design studies. Assay results for this inhibitor
indicate this not to be the case. The presence of an aliphatic
substituent on the triazole moiety, on the other hand, can lead
to an increase in inhibitory potential with respect to the
unsubstituted 9a and the corresponding reference compounds
15 and 16 in Table 1. This is demonstrated by compound 9c,
bearing a small R2-cyclopropyl group. The size of the
substituent, however, does not seem to be of prime importance,
since even the N-Boc-derivatized inhibitors 9e and 9f (used as
the precursor for deprotected, basic target compound 9q)
display a potency improvement with respect to the reference
compounds that is comparable to the result obtained for 9c.
For the set of inhibitors with an aromatic substituent
(compounds 9g−p), attaching an unsubstituted phenyl or
pyridine ring at the triazole moiety does not lead to significant
alterations in enzyme affinity with respect to the reference
compounds (maximal change corresponding to a factor of 2).
Halogenation of the aromatic ring (compounds 9i−m), on the
other hand, clearly has substantial impact and can lead to an
increase in PREP affinity of about 1 order of magnitude. The
origin of this jump in affinity is not clear and does not appear to
be directly correlated to either the nature of the halogen or its
position on the aromatic ring. In order to further interpret
As explained in the Introduction, a limited number of
(2S,4R) containing analogues were also prepared in this study
in order to validate initial assumptions on the expected binding
mode of the target compounds (represented Table 3).
Although explained more fully by the crystal structures
9859
dx.doi.org/10.1021/jm301060g | J. Med. Chem. 2012, 55, 9856−9867

Journal of Medicinal Chemistry
Article
Table 2. Inhibitory Potential of P2-(4S)-Triazoloproline-Based Compounds 9a−v toward PREP, FAP, DPP IV, DPPII, and
b
DPP9
9860
dx.doi.org/10.1021/jm301060g | J. Med. Chem. 2012, 55, 9856−9867

Journal of Medicinal Chemistry
Article
Table 2. continued
a
b
“n.d.” means “not determined”. For compounds 9g−i and 9p−r, an IC50 for inhibition of FAP could not be determined because of compound
precipitation under the FAP assay conditions. For compound 9v, an IC50 for inhibition of DPP IV, DPP II, and DPP9 could not be determined
because of limited compound availability.
Table 3. Inhibitory Potential of P2-(2S,4R)-Azido- and Triazoloproline-Based Compounds 8 and 10a,b toward PREP, FAP,
b
DPP IV, DPPII, and DPP9
a
b
‘”n.d.” means “not determined”. Compound 10a precipitated under the FAP-assay conditions.
(discussed below), the lower affinity of these molecules is
revealed by comparing the sterically hindered compound 10b
with its (2S,4S) counterpart 9f, in support of our original
hypothesis.
Finally, although not originally part of the intention of this
study, we evaluated the possibility of further increasing the
affinity of P2-substituted target compounds by introduction of a
carbonitrile warhead. As a reference compound, carbonitrile 17
(KYP-2047), reported earlier by Jarho et al., was synthesized
and evaluated under our assay conditions (Table 4).30
The PREP assay data for reference compound 16 and its
carbonitrile-containing analogue 17 clearly illustrate the
potential of warhead introduction to increase enzyme affinity.
A similar trend can be observed when comparing assay results
of 7b and that of inhibitor 13, with the latter roughly being 2
orders of magnitude more potent. The increase in affinity is less
pronounced (a factor of 3) when comparing inhibitors 9f and
14. This could be because accommodation of the N-Boc-
aminomethyltriazole substituent in 14 impedes optimal
orientation of the carbonitrile group for reaction with the
serine hydroxyl group. To get a clearer view on this matter,
compound 14 was also selected for cocrystallization with PREP.
Crystal Structures of PREP−Inhibitor Complexes. The
overall binding mode of the acylated prolylpyrrolidine moiety
of the three novel PREP complexes with compounds 13, 14,
9861
dx.doi.org/10.1021/jm301060g | J. Med. Chem. 2012, 55, 9856−9867

Journal of Medicinal Chemistry
Article
Table 4. Inhibitory Potential of P1-(2S)-Cyanoproline-Based Compounds 13−15
Figure 3. Crystal structures of PREP−inhibitor complexes: (A) Z-Pro-prolinal (PDB code 1qfs); (B) compound 13 (PDB code 4BCB); (C)
compound 14 (PDB code 4BCC); (D) compound 9i (PDB code 4BCD). PREP and inhibitor carbon atoms are colored gray and green, respectively.
Hydrogen bonds are drawn as thin lines. The σA weighted 2mFo − ΔFc electron density using phases from the final model is contoured at the 1σ
level, where σ represents the rms electron density for the unit cell. Contours more than 1.4 Å from any of the displayed atoms have been removed for
clarity. The figure was drawn using Molscript31 and rendered with Raster 3D.32
and 9i (Figure 3) closely resembles that of the Z-Pro-prolinal
(ZPP) complex determined previously.24a
The P1 pyrrolidine ring occupies the snug P1 pocket, and
additional specificity is provided by a ring stacking interaction
with Trp595. The carbonitrile of compounds 13 and 14,
equivalent to the aldehyde group of ZPP, is stabilized by
hydrogen bonds with the main chain NH of Asn555 that
follows the catalytic Ser (as observed with all α/β-hydrolase
enzymes) and with the phenolic OH of Tyr473.33 Hydrogen
bonds are made between the inhibitor P2 carbonyl and the
9862
dx.doi.org/10.1021/jm301060g | J. Med. Chem. 2012, 55, 9856−9867

Journal of Medicinal Chemistry
Article
guanidino group of Arg643 and between the P3 carbonyl and
the side chain NH of Trp595 (Figure 3B,C). The consistent
increase in inhibitor potency when changing a P3-Z function
for a 4-phenylbutanoyl residue (for example, 9f vs 9e, and 9h vs
9g) can be understood when comparing the crystal structure of
PREP in complex with ZPP with the structures for complexes
with the novel compounds. The longer phenylbutanoyl
substituent is bound snugly into this P3 pocket, while the Z
group that is two carbons shorter would not fit this pocket as
efficiently, resulting in the diminished inhibitory activity.
In the complex containing compound 13, there is clear
electron density for the azide group of the P2 substituent
(Figure 3B), and there is a hydrogen bond between the
terminal nitrogen of this group and the phenolic OH of Tyr473
that also stabilizes the oxyanion during substrate hydrolysis.
The IC50 of this inhibitor (0.003 μM) is characterized by a 2-
fold increase in potency over the equivalent compound 17 that
has a hydrogen atom at this position (IC50 = 0.006 μM) and by
a 3-fold increase over compound 14 (IC50 = 0.009 μM). While
no additional direct interactions are made with compound 14,
the tert-butyl group of this inhibitor lies only 3.8 Å away from
the OH group of Tyr471. This suggests that incorporating
some additional polar functionality in this part of the inhibitor
could yield enhanced binding. Also, the carbonyl part of the
Boc group is bound to a solvent glycerol molecule, suggesting
that expansion of this part of compound 14 could displace this
glycerol for further enhanced binding (Figure 3C).
Unlike the Boc-protected aminomethyl substituent of
compound 14 that binds close to the edge of the enzyme
cavity, the 4-fluorophenyl substituent of compound 9i points
into the center of the enzyme cavity (Figure 3D). Despite the
differences in binding mode of these substituents, the very
similar IC50 values for compounds 9f and 9i suggest that this
differential binding does not have a large effect on potency. In
addition, the binding mode determined for 9i cannot explain
the increased PREP affinity when compared to its non-
fluorinated analogue 9h, as neither the fluorine atom nor the
phenyl ring is engaged in specific interactions with the enzyme.
Nonetheless, it is clear from these results that the inhibitors’
P2 substituents that are the subject of this report can in general
be expected to be accommodated in a region of PREP that
hitherto has not been probed by inhibitors (Figure 4). With
regard to the three complexed inhibitors, only the P2-azido
substituent in 13 was found to provide additional enthalpic
stabilization of the crystallized enzyme−inhibitor complex.
However, the potential involvement of stabilizing interactions
mediated by other P2 substituents in compounds 9a−v cannot
be excluded. In a more general view, further compound
optimization by applying the insights provided by this
crystallographic study (as exemplified for compound 14)
seems a realistic goal. Finally, it is worth mentioning that the
dimensions of the P2-substituent-accommodating cavity are
large enough to allow for very bulky substituent moieties.
These could include fluorophores, liganded metal ions, and
related functionalities that typically serve as reporter groups, for
example, in activity-based probes for proteases.35
High-Content Analysis of the Effect of PREP Inhibitors
on α-Synuclein Aggregation and Apoptosis in a Cellular
Synucleinopathy Model. α-Synuclein (α-SYN) is mainly a
neuronal protein, accounting in nonpathological conditions for
up to 1% of the protein content in the cytosol. The functional
characterization of the protein is hitherto far from complete:
several hypotheses are currently investigated, but there is
mounting evidence that α-SYN is functioning as a chaperone in
the formation of SNARE complexes and in this way is involved
in vesicular trafficking and the cellular functioning of the neural
Golgi apparatus. The soluble form of α-SYN can aggregate into
insoluble fibrils in pathological conditions characterized by
intracellular Lewy bodies, such as Parkinson’s disease, Lewy
body dementia, and multiple system atrophy.36 Recently, PREP
was reported to accelerate the aggregation of α-SYN and the
formation of fibrils in a cell-free in vitro model.20,37 Moreover,
small-molecule inhibitors of PREP were demonstrated to
reverse the aggregation process in the same model. Very
recently, reference 17 was reported to also reduce α-SYN
aggregation in cellulo and in vivo.38 These findings, together
with literature evidence for the interaction between α-SYN and
PREP in cells with high expression of α-SYN, led us to evaluate
a number of our PREP inhibitors in a model of synucleinopathy
in SH-SY5Y5 human neuroblastoma cells that overexpress α-
SYN.39 Compounds 9f, 9i, 13, 14 were selected. These
molecules are among the most potent and selective novel
inhibitors we identified during this study.
Prior to the start of the aggregation experiments, cellular
permeation potential was determined. By reliance on a standard
protocol, the SH-SY5Y5 cells were incubated with different
concentrations of compounds 9i, 9f, 13, 14, and 16 for 24 h.
The supernatant was subsequently washed away, and remaining
PREP activity in the cells was evaluated after lysis, using a
standard colorigenic activity assay (Figure 5). The data
obtained indicate optimal but still moderate membrane
permeability for compounds 13 and 14. The latter inhibitors
were therefore forwarded to the aggregation experiments.
Reference 16 was not able to significantly reduce intracellular
PREP activity under these conditions (data not shown) and
was omitted in the cellular α-Syn aggregation experiments.
In the α-SYN aggregation experiments, the SH-SY5Y
neuroblastoma cells overexpressing α-SYN were exposed to
oxidative stress conditions using H2O2 and FeCl2. As described
earlier, these conditions led to the formation of α-SYN
aggregates, detected by thioflavin S staining, and to apoptosis,
measured as nuclear condensation and fragmentation. Both
parameters (aggregate formation and nucleus condensation/
fragmentation) were used as end points to evaluate the effect of
PREP inhibitors. After fixation, the cells were analyzed with an
Figure 4. Electrostatic surface of prolyl oligopeptidase−inhibitor
complexes. Carbon atoms are colored cyan for ZPP, yellow for
compound 13, salmon for compound 14, and magenta for compound
9i. The figure was drawn with PyMOL.34
9863
dx.doi.org/10.1021/jm301060g | J. Med. Chem. 2012, 55, 9856−9867

Journal of Medicinal Chemistry
Article
reduce α-SYN aggregate formation in this model is
considerable.
Both 13 and 14 also significantly reduced the number of cells
in late apoptosis. Although apoptosis in this model has been
shown to be positively correlated to occurrence of α-SYN
aggregates, several other, yet unidentified factors appear to be
influencing the apoptotic processes as well.39 Therefore, care
should be taken in interpreting this end point and in comparing
it to the effect on α-SYN aggregation. Nonetheless, since
synucleinopathies are characterized by important neuronal
apoptotic processes, we decided to include this parameter in
our analysis as an additional criterion to select compounds for
potential evaluation in more advanced preclinical models of
synucleinopathy. Again, the maximal effect on apoptosis of 13
and 14 is roughly comparable to what has been reported earlier
for tacrolimus and 17 in this model.
Figure 5. Membrane permeability of inhibitors 9f, 9i, 13, and 14
determined by measuring residual PREP activity in SH-SY5Y cells after
incubation with different inhibitor concentrations for 24 h, followed by
washing away of the supernatant and lysis of the cells. Residual activity
is defined as the % of PREP activity relative to untreated cells that
remains after incubation and washing, measured using the colorigenic
substrate Z-Gly-Pro-p-nitroanilide.
CONCLUSION
■
We have investigated the effect on affinity of regioselectively
introduced substituents in the S2-binding part of the typical
dipeptide-derived basic structure of PREP inhibitors. Such
modifications could be used to modify or tune the target
affinity, the selectivity, and the physicochemical parameters in
drug discovery programs focusing on PREP inhibitors.
Biochemical evaluation of the produced inhibitors identified
several substituent types that are able to significantly increase
target affinity, thereby reducing the need for an electrophilic
“warhead” functionality. Pronounced PREP specificity within
the group of Clan SC proteases was generally observed for the
novel molecules, comparable to the profile of several dipeptide-
derived molecules that were reported earlier. Nonetheless, the
presence of the additional substituent could also be expected to
increase PREP selectivity with respect to other potential targets
recognizing the dipeptidic architecture of PREP inhibitors or
delay metabolization by dipeptide-processing enzymes. Fur-
thermore, the structures of several inhibitor−PREP complexes
were determined by X-ray crystallography. Finally, we studied
selected compounds in a cellular model of synucleinopathy.
Our results demonstrate that PREP inhibitors exert a significant
antiaggregation potential on α-synuclein in SH-SY5Y cells, an
effect that had been reported earlier in the same model for
reference compound 17. The experiments also suggest that the
compounds influence late apoptosis in these cells. In
conclusion, the obtained data call for extended investigation
of these novel, selective, and high affinity PREP inhibitors in
automated image acquisition and analysis protocol. To facilitate
the quantitative analysis of α-SYN inclusion formation in this
cell culture model, a high-content multiparametric method was
used.38
Next, the concentration dependence of the effect on α-SYN
aggregation was investigated for the two most active
compounds, 13 and 14. The corresponding data, summarized
in Figure 6, suggest EC50 values in the low micromolar region
for the molecules. The observation that compound 13
consistently displays a somewhat lower effective concentration
than 14 is in line with the measured difference in IC50 for both
as inhibitors of PREP. Altogether, the cell permeability, the low
micromolar EC50 values, and the maximal reduction in cell
numbers displaying α-SYN aggregation are equivalent to the
results obtained earlier for compound 17 in this model.38 The
maximal reduction in α-SYN aggregation is also comparable to
that induced by tacrolimus, the most potent compound
identified so far in this model.39 The latter, however, has an
EC50 in the low nanomolar range. Taking into account the
moderate cell permeability of the compounds (Figure 5), these
data indicate that the intrinsic potential of PREP inhibitors to
Figure 6. Effect of PREP inhibitors 13 and 14 on α-SYN aggregation and late apoptosis. SH-SY5Y cells were incubated with the inhibitors for 3 days
under oxidative stress and then analyzed as described. The percentages of cells with α-SYN aggregates and of cells in late apoptosis are expressed as
percentages of the number of cells recorded under control conditions (1% DMSO). Statistical significance compared to control (1% DMSO) was
determined using ANOVA, followed by Dunnett’s post-test with a significance level of 5%. This is represented as follows: (∗) p < 0.05; (∗∗) p <
0.01; (∗∗∗) p < 0.001.
9864
dx.doi.org/10.1021/jm301060g | J. Med. Chem. 2012, 55, 9856−9867

Journal of Medicinal Chemistry
Article
2.9−3.01 (m, 1H, CH2); 3.38−3.51 (m, 2H, CH2); 3.56−3.67 (quin, J
= 7.2 Hz, 1H, CH2); 3.88−4.06 (m, 2H, CH2); 4.2 (t, J = 7.6 Hz, 1H,
CH2); 4.85 (t, J = 7.6 Hz, 1H, CH2); 5.43 (t, J = 7.6 Hz, 1H, CH2);
7.12 (t, J = 8 Hz, 3H, CHar); 7.18 (d, J = 7.2 Hz, 2H, CHar); 7.26 (t, J
= 8 Hz, 2H, CHar); 7.81 (dt, J = 8 Hz, J = 2 Hz, 2H, CHar); 8.24 (s,
1H, CHar). ESI-MS+: m/z = 476.2 [M + H]+. Purity determination:
(1) RP-HPLC, tR = 30.06 min, purity = 99%; (2) LC/MS, tR = 16.2
min, purity = 98%.
both in vitro and in vivo models of synucleinopathy. Such
experiments could be instrumental to validate PREP as a
druggable target for this class of diseases.40 Additional effort to
apply the design principles presented in this publication for
obtaining PREP inhibitors with optimized cell permeability
would, however, be desirable.
1-((2S,4S)-4-(4-((Dimethylamino)methyl)-1H-1,2,3-triazol-1-
yl)-2-(pyrrolidine-1-carbonyl)pyrrolidin-1-yl)-4-phenylbutan-
1-one (9r). A procedure identical to the one used for the preparation
of compound 9c was applied using azide 7b (0.28 mmol, 100 mg, 1
equiv), N,N-dimethylprop-2-yn-1-amine (0.34 mmol, 28 mg, 1.2
equiv), CuI (0.028 mmol, 5.3 mg, 0.1 equiv), and Et3N (trace). After
chromatography, the target product was obtained as a colorless oil that
transformed upon standing into an amorphous solid. 1H NMR
(CDCl3, 400 MHz): δ 1.81−2.14 (m, 6H, CH2); 2.2−2.31 (m, 8H,
CH2, CH3); 2.36−2.44 (m, 1H, CH2); 2.67 (t, J = 7.6 Hz, 2H, CH2);
2.81−2.92 (m, 1H, CH2); 3.34−3.41 (m, 2H, CH2); 3.48 (s, 2H,
CH2); 3.54−3.62 (m, 1H, CH2); 3.86−3.97 (m, 2H, CH2); 4.17 (t, J =
7.6 Hz, 1H, CH2); 4.79 (t, J = 7.6 Hz, 1H, CH2); 5.28 (quin, J = 7.6
Hz, 1H, CH); 7.16 (d, J = 7.6 Hz, 2H, CHar); 7.28 (t, J = 7.6 Hz, 3H,
CHar); 7.91 (d, J = 7.6 Hz, 1H, CHar). ESI-MS+: m/z = 439.6 [M +
EXPERIMENTAL SECTION
■
Amino acids and TBTU were purchased from Novabiochem. Other
reagents were obtained from Sigma-Aldrich or Acros and used as such
unless otherwise specified. Characterization of intermediates and final
compounds was done using NMR spectroscopy and mass spectrom-
etry. Final purity was determined using HPLC analysis. 1H NMR (400
MHz) and 13C NMR (100 MHz) spectra were recorded on a Bruker
Advance III Ultrashield 400 spectrometer. CDCl3, CD3OD, or
DMSO-d6 were used as the solvents. Chemical shifts in the spectra
are given in ppm and coupling constants (J) in Hz. ES mass spectra
were obtained with an Esquire 3000Plus ion trap mass spectrometer
from Bruker Daltonics, using the direct infusion mode.
Purity of final products was determined using HPLC analysis. (1)
LC/MS chromatograms were recorded on an Agilent 1100 series
HPLC system equipped with an Alltech Prevail C18 column (2.1 mm
× 50 mm × 3 μm) connected to an Esquire 3000Plus ion trap mass
spectrometer from Bruker Daltonics. A 5−100% acetonitrile/water, 20
min gradient, was used with a flow rate of 0.2 mL/min. Formic acid
(0.1%) was added to both solvents. (2) In addition, reversed phase
HPLC chromatograms were recorded on a Gilson instrument
equipped with an Ultra Sphere ODS column (4.6 mm × 250 mm ×
5 μm) and a UV detector. A 10−100% acetonitrile, 35 min gradient,
was used with a flow rate of 1 mL/min. Then 0.1% trifluoroacetic acid
was added to both solvents. An indicated purity of 100% indicates that
no other peaks in the chromatogram occur. All final products reported
in this publication were determined to have purities of ≥95%.
General Experimental and Analytical Data for Key Products
9c, 9i, 9r, and 14. Detailed experimental and analytical data for all
intermediates and final products in the manuscript can be found in the
Supporting Information.
H]+, 461.3 [M + Na]+. Purity determination: (1) RP-HPLC, tR
=
13.98 min, purity = 100%; (2) LC/MS, tR = 18.3 min, purity = 99%.
tert-Butyl ((1-((3S,5S)-5-((S)-2-Cyanopyrrolidine-1-carbonyl)-
1-(4-phenylbutanoyl)pyrrolidin-3-yl)-1H-1,2,3-triazol-4-yl)-
methyl)carbamate (14). A procedure identical to the one used for
the preparation of compound 9c was applied using azide 13 (0.26
mmol, 100 mg, 1 equiv), N-Boc-propargylamine (0.34 mmol, 50 mg,
1.2 equiv), CuI (0.026 mmol, 5.0 mg, 0.1 equiv), and Et3N (trace) and
using dichloromethane (5 mL) as the solvent instead of methanol.
After chromatography, the target product was obtained as a colorless
oil that transformed upon standing into an amorphous solid. 1H NMR
(CDCl3, 400 MHz): δ 1.44 (s, 9H, CH3); 1.92−2.06 (m, 3H, CH2);
2.18−2.33 (m, 5H, CH2); 2.39−2.5 (m, 1H, CH2); 2.66 (quin, J = 8
Hz, 2H, CH2); 2.84−2.95 (m, 1H, CH2); 3.58−3.68 (m, 1H, CH);
3.88 (t, J = 8.2 Hz, 1H, CH2); 3.91−3.98 (m, 1H, CH); 4.17 (t, J = 8.2
Hz, 1H, CH2); 4.34−4.49 (m, 2H, CH2); 4.7 (t, J = 8 Hz, 1H, CH);
4.79−4.86 (m, 1H, CH); 5.13 (br s, 1H, NH) 5.26 (t, J = 8 Hz, 1H,
CH); 7.18 (t, J = 8 Hz, 3H, CHar); 7.26 (m, 2H, CHar); 7.76 (s, 1H,
CHar). ESI-MS+: m/z = 536.3 [M + H]+. Purity determination: (1)
RP-HPLC, tR = 26.2 min, purity = 98.4%; (2) LC/MS, tR = 18.3 min,
purity = 98.8%.
1-((2S,4S)-4-(4-Cyclopropyl-1H-1,2,3-triazol-1-yl)-2-(pyrroli-
dine-1-carbonyl)pyrrolidin-1-yl)-4-phenylbutan-1-one (9c).
Azide 7b (0.28 mmol, 100 mg, 1 equiv) was dissolved in methanol.
To this solution was added ethynylcyclopropane (0.34 mmol, 23 mg,
1.2 equiv), followed by CuI (0.028 mmol, 5.3 mg, 0.1 equiv) and Et3N
(trace). The reaction mixture was stirred for 24 h at room
temperature, after which thin-layer chromatography was used to
demonstrate complete consumption of the starting material. The
volatile components were then evaporated, and the target compound
was isolated after flash chromatography (hexanes to hexanes/ethyl
actetate, 1:1), mainly for removing copper residues. The target
compound was obtained as a colorless oil that transformed upon
ASSOCIATED CONTENT
* Supporting Information
■
S
(1) General synthetic procedures, (2) compound character-
ization data, (3) enzymatic assay conditions, (4) crystallo-
graphic data, and (5) the SH-SY5Y cellular assay conditions.
This material is available free of charge via the Internet at
http://pubs.acs.org.
1
standing into an amorphous solid. H NMR (CDCl3, 400 MHz): δ
0.97 (m, 2H, CH2); 1.42 (m, 2H, CH2); 1.76−2.12 (m, 6H, CH2);
2.24−2.39 (m, 3H, CH2 + CH); 2.67 (t, J = 7.2 Hz, 2H, CH2); 2.79−
2.93 (m, 1H, CH2); 3.18−3.30 (m, 1H, CH2); 3.36−3.45 (m, 2H,
CH2); 3.54−3.68 (m, 1H, CH2); 3.82 (t, 1H, J = 9.6 Hz, CH2); 3.94
(q, J = 8 Hz, 1H, CH2); 4.13 (t, J = 9.6 Hz, 1H, CH2); 4.8 (t, J = 8 Hz,
1H, CH2); 5.29 (quin, J = 8 Hz, 1H, CH2); 7.14−7.21 (m, 3H, CHar);
7.24−7.31 (m, 2H, CHar); 7.71 (s, 1H, CHar). ESI-MS+: m/z = 422.3
[M + H]+. Purity determination: (1) RP-HPLC, tR = 18.44 min, purity
= 97.31%; (2) LC/MS, tR = 13.4 min, purity = 96%.
AUTHOR INFORMATION
Corresponding Author
*Phone: +32-3-2652708. E-mail: pieter.vanderveken@ua.ac.be.
■
Notes
The authors declare no competing financial interest.
1-((2S,4S)-4-(4-(4-Fluorophenyl)-1H-1,2,3-triazol-1-yl)-2-
(pyrrolidine-1-carbonyl)pyrrolidin-1-yl)-4-phenylbutan-1-one
(9i). A procedure identical to the one used for the preparation of
compound 9c was applied using azide 7b (0.28 mmol, 100 mg, 1
equiv), 1-ethynyl-4-fluorobenzene (0.34 mmol, 40 mg, 1.2 equiv), CuI
(0.028 mmol, 5.3 mg, 0.1 equiv), and Et3N (trace). After
chromatography, the target product was obtained as a colorless oil
that transformed upon standing into an amorphous solid. H NMR
(CDCl3, 400 MHz): δ 1.82−2.12 (m, 6H, CH2); 2.32 (t, J = 7.2 Hz,
2H, CH2); 2.34−2.51 (m, 1H, CH2); 2.68 (t, J = 7.6 Hz, 2H, CH2);
ACKNOWLEDGMENTS
■
This work received support from the EU FP7-HEALTH-2007-
B, proposal No. 223077 (nEuroPro), the Flanders Research
Foundation (FWO, Project G.0114.08), BOF GOA (Research
Council of University of Antwerp, Belgium), and the Hercules
Foundation. Crystallographic data were collected at beamline
IO4 at Diamond Light Source, U.K., and we acknowledge the
support of Ralf Flaig. Nicole Lamoen and Sophie Lyssens are
1
9865
dx.doi.org/10.1021/jm301060g | J. Med. Chem. 2012, 55, 9856−9867

Journal of Medicinal Chemistry
Article
(12) Warden, C. H.; Fisler, J. S.; Espinal, G.; Graham, J.; Havel, P. J.;
Perroud, B. Maternal influence of prolyl endopeptidase on fat mass of
adult progeny. Int. J. Obes. 2009, 33 (9), 1013−1016.
(13) Perroud, B.; Alvarado, R. J.; Espinal, G. M.; Morado, A. R.;
Phinney, B. S.; Warden, C. H. In vivo multiplex quantitative analysis of
3 forms of alpha melanocyte stimulating hormone in pituitary of prolyl
endopeptidase deficient mice. Mol. Brain 2009, 2 (1), 14.
(14) Schulz, I.; Gerhartz, B.; Neubauer, A.; Holloschi, A.; Heiser, U.;
Hafner, M.; Demuth, H. U. Modulation of inositol 1,4,5-triphosphate
concentration by prolyl endopeptidase inhibition. Eur. J. Biochem.
2002, 269 (23), 5813−5820.
(15) Schulz, I.; Zeitschel, U.; Rudolph, T.; Ruiz-Carrillo, D.; Rahfeld,
J. U.; Gerhartz, B.; Bigl, V.; Demuth, H. U.; Rossner, S. Subcellular
localization suggests novel functions for prolyl endopeptidase in
protein secretion. J. Neurochem. 2005, 94 (4), 970−979.
thanked for their excellent technical assistance. The Laboratory
of Medicinal Chemistry and Laboratory of Medical Biochem-
istry, University of Antwerp, are partners of the Antwerp Drug
Discovery Network (ADDN).
ABBREVIATIONS USED
■
DPP, dipeptidyl peptidase; FAP, fibroblast activation protein;
POP, prolyl oligopeptidase; PO, prolyl oligopeptidase; PREP,
prolyl oligopeptidase; SNARE, soluble NSF attachment
receptor protein; α-SYN, α-synuclein; ZPP, N-benzyloxycarbo-
nylprolylprolinal
ADDITIONAL NOTE
■
Prolyl oligopeptidase, formerly known as prolyl endopeptidase,
has received several abbreviations (PREP, PEP, POP, PO). In
this paper we refer to it as PREP, which relates to the unique
name of the PREP gene as found in gene and protein databases.
(16) Rossner, S.; Schulz, I.; Zeitschel, U.; Schliebs, R.; Bigl, V.;
Demuth, H. U. Brain prolyl endopeptidase expression in aging, APP
transgenic mice and Alzheimer’s disease. Neurochem. Res. 2005, 30 (6−
7), 695−702.
(17) Moreno-Baylach, M. J.; Felipo, V.; Mannisto, P. T.; Garcia-
Horsman, J. A. Expression and traffic of cellular prolyl oligopeptidase
are regulated during cerebellar granule cell differentiation, maturation,
and aging. Neuroscience 2008, 156 (3), 580−585.
(18) Agirregoitia, N.; Bizet, P.; Agirregoitia, E.; Boutelet, I.; Peralta,
L.; Vaudry, H.; Jegou, S. Prolyl endopeptidase mRNA expression in
the central nervous system during rat development. J. Chem.
Neuroanat. 2010, 40 (1), 53−62.
(19) Myohanen, T. T.; Garcia-Horsman, J. A.; Tenorio-Laranga, J.;
Mannisto, P. T. Issues about the physiological functions of prolyl
oligopeptidase based on its discordant spatial association with
substrates and inconsistencies among mRNA, protein levels, and
enzymatic activity. J. Histochem. Cytochem. 2009, 57 (9), 831−848.
(20) Brandt, I.; Gerard, M.; Sergeant, K.; Devreese, B.; Baekelandt,
V.; Augustyns, K.; Scharpe, S.; Engelborghs, Y.; Lambeir, A. M. Prolyl
oligopeptidase stimulates the aggregation of alpha-synuclein. Peptides
2008, 29 (9), 1472−1478.
(21) Lawandi, J.; Gerber-Lemaire, S.; Juillerat-Jeanneret, L.;
Moitessier, N. Inhibitors of prolyl oligopeptidases for the therapy of
human diseases: defining diseases and inhibitors. J. Med. Chem. 2010,
53 (9), 3423−3438.
REFERENCES
■
(1) Walter, R. Partial purification and characterization of post-proline
cleaving enzyme: enzymatic inactivation of neurohypophyseal
hormones by kidney preparations of various species. Biochim. Biophys.
Acta 1976, 422 (1), 138−158.
(2) Garcia-Horsman, J. A.; Mannisto, P. T.; Venalainen, J. I. On the
role of prolyl oligopeptidase in health and disease. Neuropeptides 2007,
41 (1), 1−24.
(3) Brandt, I.; Scharpe, S.; Lambeir, A. M. Suggested functions for
prolyl oligopeptidase: a puzzling paradox. Clin. Chim. Acta 2007, 377
(1−2), 50−61.
(4) Umemura, K.; Kondo, K.; Ikeda, Y.; Kobayashi, T.; Urata, Y.;
Nakashima, M. Pharmacokinetics and safety of JTP-4819, a novel
specific orally active prolyl endopeptidase inhibitor, in healthy male
volunteers. Br. J. Clin. Pharmacol. 1997, 43 (6), 613−618.
(5) Morain, P.; Robin, J. L.; De Nanteuil, G.; Jochemsen, R.; Heidet,
V.; Guez, D. Pharmacodynamic and pharmacokinetic profile of S
17092, a new orally active prolyl endopeptidase inhibitor, in elderly
healthy volunteers. A phase I study. Br. J. Clin. Pharmacol. 2000, 50
(4), 350−359.
(22) (a) Z-Pro-prolinal was originally described in the following:
Friedman, T. C.; Orlowski, M.; Wilk, S. Prolyl endopeptidase:
inhibition in vivo by N-benzyloxycarbonyl-prolyl-prolinal. J. Neuro-
chem. 1984, 42 (1), 237−241. (b) JTP-4819 was originally reported in
the following: Wallen, E. A. A.; Christiaans, J. A. M.; Saario, S. M.;
Forsberg, M. M.; Venalainen, J. I.; Paso, H. M.; Mannisto, P. T.;
Gynther, J. 4-Phenylbutanoyl 2(S)-acylpyrrolidines and 4-phenyl-
butanoyl-L-prolyl-2(S)-acylpyrrolidines as prolyl oligopeptidase inhib-
itors. Bioorg. Med. Chem. 2002, 10 (7), 2199−2206. (c) Compound 3
was reported in the following: Haffner, C. D.; Diaz, C. J.; Miller, A. B.;
Reid, R. A.; Madauss, K. P.; Hassel, A.; Hanlon, M. H.; Porter, D. J. T.;
Becherer, J. D.; Carter, L. H. Bioorg. Med. Chem. Lett. 2008, 18, 4360−
4363. (d) Compound 4 was reported in the following: Tran, T.;
Quan, C.; Edosada, C. Y.; Mayeda, M.; Wiesmann, C.; Sutherlin, D.;
Wolf, B. B. Synthesis and structure−activity-relationship of N-acyl-Gly,
N-acyl-Sar and N-blocked boroPro inhibitors of FAP, DPP IV and
POP. Bioorg. Med. Chem. Lett. 2007, 17 (5), 1438−1442.
(23) Van der Veken, P.; Haemers, A.; Augustyns, K. Prolyl peptidases
related to dipeptidyl peptidase IV: potential of specific inhibitors in
drug discovery. Curr. Top. Med. Chem 2007, 7 (6), 621−635.
(24) (a) Fulop, V.; Szeltner, Z.; Renner, V.; Polgar, L. Structures of
prolyl oligopeptidase substrate/inhibitor complexes. Use of inhibitor
binding for titration of the catalytic histidine residue. J. Biol. Chem.
2001, 276 (2), 1262−1266 (corresponding PDB entry, 1E8N).
(b) Fulop, V.; Bocskei, Z.; Polgar, L. Prolyl oligopeptidase: an unusual
beta-propeller domain regulates proteolysis. Cell 1998, 94 (2), 161−
170 (corresponding PDB entry, 1QFS). (c) Kanai, K.; Aranyi, P.;
Bocskei, Z.; Ferenczy, G.; Harmat, V.; Simon, K.; Batori, S.; Naray-
Szabo, G.; Hermecz, I. Prolyl oligopeptidase inhibition by N-acyl-Pro-
(6) Morain, P.; Lestage, P.; De Nanteuil, G.; Jochemsen, R.; Robin, J.
L.; Guez, D.; Boyer, P. A. S 17092: a prolyl endopeptidase inhibitor as
a potential therapeutic drug for memory impairment. Preclinical and
clinical studies. CNS Drug. Rev. 2002, 8 (1), 31−52.
(7) Bellemere, G.; Morain, P.; Vaudry, H.; Jegou, S. Effect of S
17092, a novel prolyl endopeptidase inhibitor, on substance P and
alpha-melanocyte-stimulating hormone breakdown in the rat brain. J.
Neurochem. 2003, 84 (5), 919−929.
(8) Jalkanen, A. J.; Puttonen, K. A.; Venalainen, J. I.; Sinerva, V.;
Mannila, A.; Ruotsalainen, S.; Jarho, E. M.; Wallen, E. A.; Mannisto, P.
T. Beneficial effect of prolyl oligopeptidase inhibition on spatial
memory in young but not in old scopolamine-treated rats. Basic Clin.
Pharmacol. Toxicol. 2007, 100, 132−138.
(9) Williams, R. S.; Eames, M.; Ryves, W. J.; Viggars, J.; Harwood, A.
J. Loss of a prolyl oligopeptidase confers resistance to lithium by
elevation of inositol (1,4,5) trisphosphate. EMBO J. 1999, 18 (10),
2734−2745.
(10) Williams, R. S.; Cheng, L.; Mudge, A. W.; Harwood, A. J. A
common mechanism of action for three mood-stabilizing drugs. Nature
2002, 417 (6886), 292−295.
(11) (a) Di Daniel, E.; Glover, C. P.; Grot, E.; Chan, M. K.;
Sanderson, T. H.; White, J. H.; Ellis, C. L.; Gallagher, K. T.; Uney, J.;
Thomas, J.; Maycox, P. R.; Mudge, A. W. Prolyl oligopeptidase binds
to GAP-43 and functions without its peptidase activity. Mol. Cell.
Neurosci. 2009, 41 (3), 373−382. (b) Szeltner, Z.; Morawski, M.;
Juhasz, T.; Szamosi, I.; Liliom, K.; Csizmok, V.; Tolgyesi, F.; Polgar, L.
GAP43 shows partial co-localisation but no strong physical interaction
with prolyl oligopeptidase. Biochim. Biophys. Acta 2010, 1804, 2162−
76.
9866
dx.doi.org/10.1021/jm301060g | J. Med. Chem. 2012, 55, 9856−9867

Journal of Medicinal Chemistry
Article
pyrrolidine-type molecules. J. Med. Chem. 2008, 51, 7514−7522
(corresponding PDB entries, 3EQ7 and 3EQ8). (d) Reference 22c
(corresponding PDB entry, 3DDU).
(36) Ostrerova-Golts, N.; Petrucelli, L.; Hardy, J.; Lee, J. M.; Farer,
M.; Wolozin, B. The A53T alpha-synuclein mutation increases iron-
dependent aggregation and toxicity. J. Neurosci. 2000, 20, 6048−6055.
(37) (a) Lambeir, A. M. Interaction of prolyl oligopeptidase with
alpha-synuclein. CNS Neurol. Disord.: Drug Targets 2011, 10, 349−354.
(b) Van Elzen, R.; Lambeir, A. M. Structure and function relationship
in prolyl oligopeptidase. CNS Neurol. Disord.: Drug Targets 2011, 10,
297−305.
(25) (a) Venalainen, J. I.; Garcia-Horsman, J. A.; Forsberg, M. M.;
Jalkanen, A.; Wallen, E. A. A.; Jarho, E. M.; Christiaans, J. A. M.;
Gynther, J.; Mannisto, P. T. Binding kinetics and duration of in vivo
action of prolyl oligopeptidase inhibitors. Biochem. Pharmacol. 2006,
71 (5), 683−692. (b) Zhai, W.; Cardell, M.; De Meester, I.; Augustyns,
(38) Myohanen, T. T.; Hannula, M. J.; Van Elzen, R.; Gerard, M.;
̈
̈
́
K.; Hillinger, S.; Inci, I.; Arni, S.; Jungraithmayr, W.; Scharpe, S.;
Van Der Veken, P.; García-Horsman, J. A.; Baekelandt, V.; Mannisto,
̈
̈
Weder, W.; Korom, S. Ischemia/reperfusion injury: the role of CD26/
dipeptidyl-peptidase IV inhibition in lung transplantation. Transplant.
Proc. 2006, 38 (10), 3369−3371.
P.; Lambeir, A. M. A prolyl oligopeptidase inhibitor, KYP-2047,
reduces α-synuclein protein levels and aggregates in cellular and
animal models of Parkinson’s disease. Br. J. Pharmacol. 2012, 166,
1097−1113.
(26) (a) Webb, T. R.; Eigenbrot, C. Conformationally restricted
arginine analogs. J. Org. Chem. 1991, 56, 3009−3016. (b) in Haitao, J.;
Gomez-Vidal, J. A.; Martasek, P.; Roman, L.; Silverman, R. B.
Conformationally restricted dipeptide amides as potent and selective
neuronal nitric oxide synthase inhibitors. J. Med. Chem. 2006, 49,
6253−6264.
(39) Gerard, M.; Deleersnijder, A.; Daniels, V.; Schreurs, S.; Munck,
S.; Reumers, V.; Pottel, H.; Engelborghs, Y.; Van den Haute, C.;
Taymans, J. M.; Debyser, Z.; Baekelandt, V. Inhibition of FK506
binding proteins reduces alpha-synuclein aggregation and Parkinson’s
disease-like pathology. J. Neurosci. 2010, 30, 2454−2463.
(40) Lambeir, A. M. Translational research on prolyl oligopeptidase
inhibitors: the long road ahead. Expert Opin. Ther. Pat. 2011, 21, 977−
981.
(27) Rostovstsev, V. V.; Green, L. G.; Fokin, V. F.; Sharpless, K. B. A
stepwise Huisgen cycloaddition process: Cu-I-catalyzed regioselective
ligation of azides and terminal alkynes. Angew. Chem., Int. Ed. 2002, 41
(14), 2596−2599.
(28) (a) Van der Veken, P.; De Meester, I.; Dubois, V.; Soroka, A.;
Van Goethem, S.; Maes, M. B.; Brandt, I.; Lambeir, A. M.; Chen, X.;
Haemers, A.; Scharpe, S.; Augustyns, K. Inhibitors of dipeptidyl
peptidase 8 and dipeptidyl peptidase 9. Part 1: Identification of
dipeptide derived leads. Bioorg. Med. Chem. Lett. 2008, 18 (14), 4154−
8. (b) Van Goethem, S.; Van der Veken, P.; Dubois, V.; Soroka, A.;
Lambeir, A. M.; Chen, X.; Haemers, A.; Scharpe, S.; De Meester, I.;
Augustyns, K. Inhibitors of dipeptidyl peptidase 8 and dipeptidyl
peptidase 9. Part 2: Isoindoline containing inhibitors. Bioorg. Med.
Chem. Lett. 2008, 18, 4159−4162. (c) Van der Veken, P.; Soroka, A.;
Brandt, I.; Chen, Y. S.; Maes, M. B.; Lambeir, A. M.; Chen, X.;
Haemers, A.; Scharpe, S.; Augustyns, K.; De Meester, I. Irreversible
inhibition of dipeptidyl peptidase 8 by dipeptide-derived diaryl
phosphonates. J. Med. Chem. 2007, 50, 5568−5570. (d) Van
Goethem, S.; Matheeussen, V.; Joossens, J.; Lambeir, A. M.; Chen,
X.; De Meester, I.; Haemers, A.; Augustyns, K.; Van der Veken, P.
Structure−activity relationship studies on isoindoline inhibitors of
dipeptidyl peptidases 8 and 8 (DPP8, DPP9): Is DPP8-selectivity
attainable? J. Med. Chem. 2011, 54, 5737−5746.
(29) (a) Yoshimoto, T.; Tsuru, D.; Yamamoto, N.; Ikezawa, R.;
Furukawa, S. Agric. Biol. Chem. 1991, 55 (1), 37−41. (b) Atack, J. R.;
Suman-Chauhan, N.; Dawson, G.; Kulagowski, J. J. In vitro and in vivo
inhibition of prolyl endopeptidase. Eur. J. Pharmacol. 1991, 205 (2),
157−163.
(30) Jarho, E. M.; Venailainen, E. I.; Huuskonen, J.; Christiaans, J. A.
M.; Forsberg, M. M.; Jairvinen, T.; Gynther, T.; Mannisto, P. T.;
Wallen, E.A.A. A cyclopent-2-enecarbonyl group mimics proline at the
̀
P2 position of prolyl oligopeptidase inhibitors. J. Med. Chem. 2004, 47
(23), 5605−5607.
(31) Kraulis, P. J. MolScript: a program to produce both detailed and
schematic plots of protein structures. J. Appl. Crystallogr. 1991, 24,
946−950.
(32) Merritt, E. A.; Murphy, M. E. P. Raster3D version 2.0. A
program for photorealistic molecular graphics. Acta Crystallogr. D
1994, 50, 869−873.
(33) DeLano, W. L. The PyMOL User’s Manual; DeLano Scientific,
LLC: Palo Alto, CA, 2002.
(34) (a) Szeltner, Z.; Rea, D.; Renner, V.; Juliano, L.; Fulop, V.;
̈
̈
Polgar
́
, L. Electrostatic environment at the active site of prolyl
oligopeptidase is highly influential during substrate binding. J. Biol.
Chem. 2003, 278, 48786−48793. (b) Rea, D.; Fulop, V. Prolyl
̈
̈
oligopeptidase structure and dynamics. CNS Neurol. Disord.: Drug
Targets 2011, 10, 306−310.
(35) Sabido, E.; Tarrago, T.; Niessen, S.; Cravatt, B. F.; Giralt, E.
Activity-based probes for measuring postproline protease activity.
ChemBioChem 2009, 10, 2361−2366.
9867
dx.doi.org/10.1021/jm301060g | J. Med. Chem. 2012, 55, 9856−9867