Prolyl oligopeptidase inhibition by N-acyl-pro-pyrrolidine-type molecules

Article abstract of DOI:10.1021/jm800944x

Three novel, N-acyl-pro-pyrrolidine-type, inhibitors of prolyl oligopeptidase (POP) with nanomolar activities were synthesized and their binding analyzed to the host enzyme in the light of X-ray diffraction and molecular modeling studies. We were interested in the alteration in the binding affinity at the S3 site as a function of the properties of the N-terminal group of the inhibitors. Our studies revealed that, for inhibitors with flat aromatic terminal groups, the optimal length of the linker chain is three C-C bonds, but this increases to four C-C bonds if there is a bulky group in the terminal position. Molecular dynamics calculations indicate that this is due to the better fit into the binding pocket. A 4-fold enhancement of the inhibitor activity upon replacement of the 4-CH₂ group of the proline ring by CF₂ is a consequence of a weak hydrogen bond formed between the fluorine atom and the hydroxy group of Tyr473 of the host enzyme. There is notably good agreement between the calculated and experimental free energies of binding; the average error in the IC₅₀ values is around 1 order of magnitude.

Full text of DOI:10.1021/jm800944x

7514
J. Med. Chem. 2008, 51, 7514–7522
Prolyl Oligopeptidase Inhibition by N-Acyl-pro-pyrrolidine-type Molecules†
Ka´roly Ka´nai,‡ Pe´ter Ara´nyi,‡ Zsolt Bo¨cskei,‡ Gyo¨rgy Ferenczy,‡ Veronika Harmat,§ Ka´lma´n Simon,‡ Sa´ndor Ba´tori,‡
Ga´bor Na´ray-Szabo´,§,¶ and Istva´n Hermecz‡,*
CHINOIN, Ltd, H-1045 Budapest, To´ u. 1-5, Hungary, Protein Modeling Group, Hungarian Academy of Sciences, Eo¨tVo¨s Lora´nd UniVersity,
H-1117 Budapest, Pa´zma´ny Pe´ter u. 1a, Hungary, and Co-operation Research Center for Molecular Design, Semmelweis UniVersity,
H-1062 Budapest, Rippl-Ro´nai u. 37, Hungary
ReceiVed July 28, 2008
Three novel, N-acyl-pro-pyrrolidine-type, inhibitors of prolyl oligopeptidase (POP) with nanomolar activities
were synthesized and their binding analyzed to the host enzyme in the light of X-ray diffraction and molecular
modeling studies. We were interested in the alteration in the binding affinity at the S3 site as a function of
the properties of the N-terminal group of the inhibitors. Our studies revealed that, for inhibitors with flat
aromatic terminal groups, the optimal length of the linker chain is three C-C bonds, but this increases to
four C-C bonds if there is a bulky group in the terminal position. Molecular dynamics calculations indicate
that this is due to the better fit into the binding pocket. A 4-fold enhancement of the inhibitor activity upon
replacement of the 4-CH2 group of the proline ring by CF2 is a consequence of a weak hydrogen bond
formed between the fluorine atom and the hydroxy group of Tyr473 of the host enzyme. There is notably
good agreement between the calculated and experimental free energies of binding; the average error in the
IC50 values is around 1 order of magnitude.
Introduction
seven-bladed ꢀ propeller domain regulates access to the active
site, which is in a large cavity at the interface of the two
domains,21,22 but recent investigations revealed that an opening-
closing mechanism of domains is more likely.23 The crystal
structure of the complex with a reversible transition state
inhibitor, 124 (Z-Pro-prolinal), reveals that the prolinal residue
is covalently bonded to the active site serine, with the proline
and the phenyl moiety of benzyloxycarbonyl groups accom-
modated in subsites S2 and S3, respectively.21 Subsite S3 is
more open than subsite S1, which allows the binding of larger
residues. The wide subsite S3 is very hydrophobic in nature.
The structures of POP (both native and inactive mutant forms)
complexes with substrates and substrate analogues have shed
light on various structural aspects of the catalytic mechanism,
and substrate binding.25-28
Prolyl oligopeptidase (POPa) (EC 3.4.21.26), previously
called prolyl endopeptidase or postproline cleaving enzyme, a
cytosolic serine protease, degrades a variety of proline-contain-
ing peptides by cleaving the peptide bond on the carboxy side
of proline residues.1 Its active site consists of a catalytic
Ser554His680Asp641 triad that is common in all serine pro-
teases. The enzyme is selective for oligopeptides not longer than
about 30 amino acid residues.2 It has been identified and purified
from various mammalian tissues, such as the brain, liver, muscle
and kidney.3-10
POP degrades proline-containing neuropeptides involved in
the processes of memory and learning, such as vasopressin,
substance P and thyrotropin-releasing hormone.11 Its activity
in plasma correlates with different stages of depression,12 and
it has also been suggested that the enzyme participates in the
control of blood pressure through the degradation of bradykinin,
and angiotensin I and II.13 POP also plays a central role in mood
stabilization, learning and memory modulating inositol con-
centration.14,15a Inhibitors of POP are potential therapeutic
agents15 in the treatment for dysfunction of the memory system,
for example, Alzheimer’s disease,16 and for the improvement
of general cognitive behavior in the elderly.17
Most currently available inhibitors are derivatives of the
transition-state analogue inhibitor, for example, 1 and the
reversible 217 (SUAM-1221) (Figure 1). The effects of modi-
fication of the five-membered rings, the shape and polarity of
the N-terminal moiety and the linker connecting the N-terminal
group and the proline ring on the inhibitor potency have been
extensively studied in Vitro.29
In our POP program, substitution of the 4-phenylbutanoyl of
2 with hexanoyl resulted in 3a with activity similar to that of
2.30-32 Substitution of the 4-CH2 group of the proline moiety
of 3a with a sulfur atom resulted in a derivative 3b that was 4
times more active. Introduction of carbonyl-containing N-
heterocycles into the end of the alkyl chain of compounds 3
led to novel types of POP inhibitors 4 with nanomolar
activities.33,34 A comparative molecular field analysis study on
44 derivatives of 4 containing phthalimido (e.g., 5) or saccharino
moieties as N-terminal group suggested some steric and
electronic modifications on these N-heterocyclic end-groups35
(Figure 2).
The C-terminal catalytic domain of POP belongs in the R/ꢀ
hydrolase fold family.18,19 Sequence homology can be detected
with lipases,20 while it is structurally unrelated to the chymot-
rypsin and subtilisin serine protease families. X-ray protein
crystallographic investigations suggested that its N-terminal
†
Atomic coordinates and structure factors have been deposited in the
Protein Data Bank with accession codes 3EQ7, +EQ8 and 3EQ9,
respectively.
* To whom correspondence should be addressed. Tel: +36-136-90-151.
Fax: +36-137-05-597. E-mail: istvan.hermecz@gmail.com.
‡
CHINOIN Ltd.
Protein Modeling Group, Eo¨tvo¨s Lora´nd University.
Co-operation Research Center for Molecular Design, Semmelweis
§
However, as the interactions at subsite S3 of POP are
nonspecific and difficult to predict, we decided to map these
interactions in the structures of the complexes for derivatives
6, 8, and 11. We determined the structures of three POP inhibitor
¶
University.
a Abbreviations: POP prolyl oligopeptidase; MD Molecular Dynamics;
DAST (diethylamino)sulfur trifluoride.
10.1021/jm800944x CCC: $40.75
2008 American Chemical Society
Published on Web 11/12/2008

Prolyl Oligopeptidase Inhibition
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7515
Compounds 14 were prepared by alkylation of quinoxalin-2-
one 1339 with ethyl ω-bromoalkanoates in DMF in the presence
of NaH, followed by acidic hydrolysis.
Replacement of methylene by a CF2 moiety may enhance
the biological activity,40 and we therefore decided to synthesize
and investigate 6. The starting difluoro derivative 16b was
prepared from (2S,4R)-1-(tert-butoxycarbonyl)-4-hydroxypyr-
rolidine-2-carboxylic acid 1741 as outlined in Scheme 2.
Compound 17 was coupled with pyrrolidine by the classical
mixed anhydride method to give 18. Subsequent Swern oxida-
tion of the secondary hydroxy group of 17 provided the
corresponding oxo derivative 18, which could be transformed
to the difluoro derivative via a known protocol with DAST.42
A deprotective step with HCl/EtOAc gave access to 16b.
Compounds 5-11 exhibit cis-trans isomerism about the
prolyl amide bond in solution.43
Results and Discussion
From the aspect of inhibitor design, it is important to know
the main driving forces that influence the position of the
N-terminal group at subsite S3 of the enzyme. A further question
is why the optimal linker chain length is three C-C bonds for
some N-terminal groups and four C-C bonds for others. The
4-fold enhancement of the inhibitor activity upon substitution
of the 4-CH2 of the P2 proline ring by CF2, a group of similar
size, was a further interesting result. It was decided to attempt
to clarify the structural basis of this finding. We carried out
X-ray crystallographic studies of enzyme-inhibitor complexes,
as representatives of the inhibitor families (6, 8 and 11), in order
to try to answer these questions.
Figure 1
The Binding Pattern at Subsites S1 and S2 is Well-
Defined. The accommodation of molecules 6, 8 and 11 in their
complex structures with POP is very similar to that of 1. There
are only minor differences in the interacting side chain
conformations of POP, in spite of the shape diversity of the
N-terminal residue and the P2 proline ring of these inhibitors.
Figure 3 depicts the superposition of inhibitors at the active
site, while Table 1 lists contacting residues. The complex
structures preserve the hydrogen bonds tying the carbonyl
oxygen atoms of the linker between P3 and P2 groups, and the
proline moiety (P2) to the Trp595 and Arg643 side chains,
respectively.
In the substrate specificity pocket (S1), the C-terminal
pyrrolidine moieties of the studied inhibitors form several
contacts with various side chains and also with the active site
serine, Ser554 (Table 1, Figure 3). However, as they are unable
to bind covalently to the active site serine, the five-membered
ring of the C-terminal pyrrolidine can be slightly rotated about
the N-C bond, while it retains its stacking interaction with the
indole ring of Trp595. Subsite S2, lined by Cys255, Tyr473,
Phe476, Trp595 and Arg643, has a slightly polar character and
is partially open toward the cavity of the enzyme. The
hydrophobic side chain of Pro forms relatively few contacts
with it.
We found that replacement of the 4-CH2 moiety by 4-CF2
results in a 4 times lower IC50 value (Table 2, 5 vs 6).
Comparison of the crystal structures of the complex formed
between POP and molecule 6 and the other complexes reveals
a weak hydrogen bond between the fluorine atom (cis to the
proline carboxyl) of 6 and the hydroxy group of Tyr473 (3.5
Å, Figure 3). We suggest that this extra hydrogen bond is
responsible for the stronger inhibitory effect of 6. The other
fluorine atom is solvent-exposed.
Figure 2
complexes (POP-6, POP-8 and POP-11) and carried out a
molecular dynamics-based study on the binding of these
inhibitors and their homologues 5, 7, 9, and 10. Below, we report
on our results and attempt to find a relationship between inhibitor
structures and activities.
Chemistry
All compounds investigated in this paper were synthesized
by acylation of the known prolylpyrrolidinine 16a36 or its
difluoro derivative 16b with acids 12,37 14 and 15,38 using the
standard mixed anhydride N-acylation method (Scheme 1).

7516 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Ka´nai et al.
Scheme 1a
a Reagents and conditions: a, Br(CH2)nCO2Et/NaH/DMF; b, HCl/∆; c, Piv-Cl, Et3N, 16a or 16b.
Scheme 2a
Table 1. Number of Contacts (d < 4.0 Å) Formed by POP and
Inhibitors 6, 8 and 11 in the Crystal Structures
inhibitor
residue
POP residue
Phe476
6
8
11
P1
P1
P1
P1
P1
P1
P1
P1
P2
P2
P2
P2
P2
P2
P2
P3
P3
P3
P3
P3
P3
P3
P3
P3
P3
P3
3
2
1
5
6
4
1
1
Ser554
Asn555
Val580
Val588
Trp595
Tyr599
Val644
Cys255
Tyr473
Phe476
Ile478
Trp595
Arg643
His680
Phe173
Ser174
Met235
Arg252
Gly254
Cys255
Ile591
4
3
1
a Reagents and conditions: a, Piv-Cl, Et3N, pyrrolidine; b, (COCl)2/
DMSO; c, DAST/CH2Cl2; d, HCi(g)/EtOAc.
1
11
1
1
1
1
4
1
7
9
1
2
14
1
2
1
1
2
1
2
4
1
22
2
4
2
5
8
13
6
6
9
4
7
1
1
4
3
3
3
1
Ala594
Trp595
Arg643
H-bonds to water
1
2
3
1
4
3
1
several contacts with Phe173 (see Figure 4A and Table 1). The
planes of their rings are in very similar orientations: the plane
angle with the P3 group’s plane of 6 is -4.9° for 8, and -2.2°
for 1. However, the positions of these ring systems in the
complexes differ significantly from that of the benzyl group in
1 (Figure 4A). While the phenyl group in 1 is close to the wall
of subsite S3, it is mainly stabilized by Ala594 and Cys255
besides Met235 and Ile591, forming few contacts with Phe173.
The ring systems of the imido and amido group-containing
inhibitors 6 and 8, respectively, with more polar characters, are
shifted closer to Phe173. The vicinity of the NCO group to
Arg643 and the similarity of the conformations of the linker
chains of 6 and 8, and of the position of the ring N atoms,
suggest that the N-terminal groups of these molecules participate
in a similar hydrogen-bonding network, though waters mediating
these hydrogen bonds could not be localized in electron density
maps. Possibly for better binding of the phenyl group in 8, the
imido group in that inhibitor is shifted as compared with those
in 6, but the guanidino group in Arg643 is also shifted. (The
distances of the guanidine NH1 and NH2 atoms from the imido
Figure 3. Binding of the P1-P2 moiety of the inhibitors in the crystal
structures. The ribbon model of the protein is colored gray, while the
inhibitor molecules and POP binding sites of the P1-P2 moieties are
magenta, red and green for the POP-6, POP-8 and POP-11 complexes,
respectively. Hydrogen bonds are shown as shaded lines. The Figure
was made with the program MOLSCRIPT.67
There Are Three Types of Inhibitor Binding at Subsite
S3. At the border of the two domains of POP, the substrate
binding groove broadens, generating the flat and open subsite
S3, lined by residues Phe173, Met235, Gly236, Ala594 and
backbone atoms of Ser174 below and by Gly254, Cys255 and
side chain carbon atoms of Arg252 on one side. At the border
of subsites S2 and S3, besides the hydrogen bond with Trp595,
plane-to-plane contacts with the guanidino group of Arg643
stabilize the P2-P3 amide bond for all inhibitors examined.
The linker between the P2 amide nitrogen and the ring system
of the N-terminal group is three C-C bonds long for inhibitors
6 and 8, similarly as for 1. These molecules with three linker
C-C bonds stack on the flat bottom of subsite S3, forming

Prolyl Oligopeptidase Inhibition
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7517
Table 2. Experimental and Calculated Binding Free Energies and their Components (kcal/mol)
b
c
d
code
X
n
∆Eele a
∆Evdw
∆Econf
IC50
3.0
∆Gexp
∆Gcalc
∆Gexp - ∆Gcalc
5
CH2
CF2
CH2
CH2
CH2
CH2
CH2
1
1
2
1
2
1
2
19.4
--
-21.0
--
0.3
--
-10.6
-11.8
-9.1
-10.9
--
+0.3
--
6e
7
0.81
40
0.88
59
18.0
13.6
23.0
36.1
20.8
-21.4
-26.5
-22.2
-26.3
-25.3
0.8
6.1
1.7
6.2
7.2
-11.6
-13.6
-9.0
-2.1
-7.7
+2.5
+2.3
+0.1
-1.9
-1.4
8e
9
-11.3
-8.9
10
-
41
-4.0f
-9.1
11e
a ∆Eele ) Eele(L - E) - Eele(L - W). b ∆Evdw ) Evdw(L - E) - Evdw(L - W). c IC50 values are in nM. d ∆Gcalc ) ∆G4 in eq 1. e X-ray structure
available. f No inhibition at 0.1 µM. Estimated ∆G corresponds to an IC50 of 25 µM.
Figure 4. Binding at subsite S3 in the crystal structures. Inhibitors with three C-C bond linkers in their P3 moiety (A) bind in significantly
different orientations from those with four C-C linkers (B). The ribbon model of the protein is colored gray, while 1 covalently bound to Ser554
of Protein Data Bank entry 1QFS is shown in dark-gray as reference. The inhibitor molecules and POP binding sites of the S3 moieties of the
POP-6, POP-8 and POP-11 complexes are colored magenta, red and green, respectively. Hydrogen bonds are shown as shaded lines. The Figure
was made with the program MOLSCRIPT.67
oxygen of 6 are 4.4 and 3.7 Å, while the amido one of 8 are
4.3 and 4.8 Å, respectively. The plane angles of guanidine and
oxo groups are 70.7° and 49.0° for 6 and 8, respectively, with
oxy group-guanidine plane distances of 3.5 Å in both com-
plexes.) This relates to the hydrogen-bonded connection with
Arg643. Although the ring systems in 6 and 8 have very
different sizes and shapes, their imido moieties are still bonded
in a similar way. This suggests that the building-up of the
hydrogen-bonding network, including the imido oxygen in the
inhibitor and the guanidino group in Arg643, is a very important
factor for stabilization of the complex.
are polar, replacement of the phenyl group by a polar aromatic
group would facilitate further favorable interactions with the
binding site.
The complexes with inhibitors 6 and 8 reveal the reason for
the different optimal linker chain length in molecules 10 and
11. These also possess the imido oxygen at the junction of their
N-terminal moiety and the linker chain, but they cannot adopt
the same position as can 6 and 8 in their complexes, because
this requires a flat N-terminal group stacked tightly to the bottom
of subsite S3. The N-terminal group in 10 and 11 is sterically
demanding. Its large substituents (i.e., phenyl groups) lie out
of the plane of the central imido ring of the N-terminal group.
Indeed, the crystal structure of the POP-11 complex shows the
N-terminal group in a different position, with contacts different
from those of the other complex structures (cf. Table 1). Its
connecting chain with the extra CH2 group bends toward the
“wall” of subsite S3 (cf. Figure 4B and Table 1). The hydrogen
bond of the imido oxygen is lost and the hydantoin moiety
contacts the side chains of Met235 from one side and Cys255
from the other. Its plane is rotated by 41.9°, allowing the
formation of a hydrogen bond with the carbonyl oxygen of
The N-heterocyclic (aromatic) moieties in 6 and 8 cover
different regions of subsite S3. While that in 6 binds between
the side chains of Met235 and Phe173, those in 8 are
accommodated mainly by the side chain of Met235 and are close
to the edge of the “wall” at the side of subsite S3 and contact
Cys255. The phenyl group in 8 undergoes a stacking interaction
with the amide plane of the Phe173-Ser174 peptide bond. The
other side of the phenyl group is solvent-exposed due to the
broadening of subsite S3. As the nearby groups of the protein

7518 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Ka´nai et al.
contrast with molecules with other P3 groups, where those with
the shorter chain (n ) 1) exhibit higher binding affinities.
Structures 10 and 11 have the largest P3 group, and thus, they
are unable to simultaneously adopt a low energy conformation
and undergo favorable interactions with the enzyme when the
alkyl chain is short (n ) 1, 10). The insertion of an extra
methylene group (n ) 2, 11) allows the P3 group to fit better
into the binding pocket. Although the conformational strain is
still significant, the lack of the highly unfavorable electrostatic
interactions demonstrated by the MD simulation in 10 (cf. Table
2) means that 11 binds to the enzyme with increased affinity as
compared with 10.
Conclusions
Our X-ray diffraction study has shed light on various aspects
of the binding of N-acyl-pro-pyrrolidine-type inhibitors of POP.
Analysis of the structures of the complexes provides an efficient
tool in drug design. The enhanced inhibitory activity of the
γ-CF2-substituted proline-containing derivatives proved to be
due to the formation of a new hydrogen bond and a subsequent
slight shift of the N-acyl-pro-pyrrolidine moiety. The crystal
structures available allowed us to study the change in binding
power at the S3 site, depending on the size, shape and polarity
of the N-terminal group of the inhibitors. Inhibitors with flat
ring systems are stacked on the bottom of the S3 site. For these
inhibitors, the optimal linker chain length is three C-C bonds.
Insertion of an imido moiety into the junction of the ring system
allows hydrogen bond formation with water molecules of the
first solvent shell of the enzyme, indicated by a shift of the
ring position in the complex as compared with the hydrophobic
benzyloxycarbonyl group. We have provided an explanation for
the change in optimal length of the linker chain for a derivative
with a bulky N-terminal group. The binding site of this group
overlaps only partially with those of the flat ones.
Figure 5. Calculated versus experimental binding free energies. The
indicated line corresponds to perfect matching of the calculated and
experimental values.
Gly254. A dipole-dipole interaction is also established between
the carbonyl groups of the N-terminal group and Gly254 in
reverse orientation. One of the phenyl rings lies in the extension
of subsite S3, packed between the side chain of Met235 and
side chain carbon atoms of Arg252, while the other is accom-
modated above subsite S3 by Arg252 side chain carbon atoms
and Cys255.
Molecular Dynamics (MD). Calculations were performed
to study the structural and energetic features of some representa-
tive complexes. The calculated free energies of binding, together
with the experimental values, are presented in Table 2 and
Figure 5. The mean unsigned error is 1.37 kcal/mol, which
corresponds to an error of approximately 1 order of magnitude
in IC50. This agreement between the calculated and experimental
values is remarkably good with regard to the simple model
applied (see Computations within the Experimental Section).
The pyrrolidine head of the ligands resides at the hydrophobic
S1 binding site, while other regions are partially exposed to
water in the central cavity. In the course of the MD simulations
of the enzyme-ligand complexes, the pro-pyrrolidine moiety
exhibited little flexibility and adopted similar conformations in
all complexes. Moreover, the bound conformations of these
fragments are very close to that obtained in water.
For all the studied molecules, a favorable van der Waals and
an unfavorable electrostatic energy component (cf. Table 2) were
found to accompany the change of the environment from water
to enzyme. The conformational strain proved to be significant
in the binding of molecules 8, 10 and 11. These are the structures
with the largest N-terminal groups. In spite of the open nature
of the S3 binding site, the bulky substituents are unable to adopt
a low-energy conformation. The excess conformational energy
is close to the higher limit of the conformational strain that
typically occurs in ligand binding.44
It is interesting that in the higher homologue of 8, that is, 9,
the conformational strain almost disappears (cf. Table 2). This
can occur because the increased length of the alkyl chain gives
more flexibility to the molecule and allows the P3 group to
penetrate further into the large water-filled pocket of the
ꢀ-propeller region. In spite of the decreased conformational
strain, 9 binds to POP with lower affinity than does 8, since
the latter interacts more favorably with the enzyme.
Experimental Section
Chemistry. Compounds 12,37 1538 and 16a36 were prepared
according to literature methods. All other reagents and solvents
were purchased from Aldrich or Fluka and were used as received.
Solvents were evaporated under reduced pressure on a Bu¨chi RE111
rotavapor. Flash chromatography was performed on Merck Kie-
selgel gel 60 (230-400 mesh) silica, with CH2Cl2/MeOH as mobile
phase. Melting points were determined on a Bu¨chi-530 melting point
apparatus and were uncorrected. Proton nuclear magnetic resonance
(1H NMR) spectra were recorded on a Bruker Avanence DRX-
200 instrument [200 MHz, δ ) 0 (TMS), in DMSO-d6 or CDCl3
as solvent]. Chemical shifts are expressed in ppm, and signal
multiplicities are described as s (singlet), d (doublet), t (triplet), q
(quartet) or br (broad). Combustion analysis was performed on a
Carlo Erba Mod 1110 for C, H and N, while F was determined by
titration after Scho¨niger oxidation. Combustion analyses for C, H,
N and F gave results within 0.4% of theory.
tert-Butyl (2S,4R)-4-Hydroxy-2-(pyrrolidinocarbonyl)pyrro-
lidine-1-carboxylate. Compound 1741 (6.92 g, 30.0 mmol) was
reacted with pyrrolidine (2.13 g, 30.0 mmol) according to general
procedure (see below) (18 h), and the reaction residue was purified
by flash chromatography to yield 6.48 g (76%) of tert-butyl (2S,4R)-
4-hydroxy-2-(pyrrolidinocarbonyl)pyrrolidine-1-carboxylate as a
1
white powder; mp 106-107 °C. H NMR (DMSO-d6): δ 1.30 +
1.37 (s + s, 9H), 1.62-1.97 (m, 5H), 2.06 (m, 1H), 3.19-3.56
(m, 6H), 4.26 (br s, 1H), 4.42 (m, 1H), 5.00 (d, 1H, exchangeable
with D2O). Anal. (C14H24N2O4) C, H, N.
tert-Butyl (2S)-4-Oxo-2-(pyrrolidinocarbonyl)pyrrolidine-1-
carboxylate (18). To a solution of oxalyl chloride (3.81 g, 30 mmol)
in CH2Cl2 (120 mL) was gradually added DMSO (4.26 mL, 60
mmol), followed by a solution of tert-butyl (2S,4R)-4-hydroxy-2-
(pyrrolidinocarbonyl)pyrrolidine-1-carboxylate (5.68 g, 20.0 mmol)
A comparison of molecules 10 and 11 reveals better binding
of the latter, which has a longer alkyl chain (n ) 2). This is in

Prolyl Oligopeptidase Inhibition
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7519
in CH2Cl2 (30 mL) at -78 °C under an argon atmosphere. After
stirring for 90 min at this temperature, Et3N was added and the
reaction mixture was left to warm up to room temperature. The
mixture was then diluted with CH2Cl2 (50 mL), washed successively
with water and brine, dried over MgSO4, filtered and evaporated
in Vacuo. The residue was purified by flash chromatography to give
2-{3-[(2S)-4,4-Difluoro-2-(pyrrolidinocarbonyl)pyrrolidin-1-
yl]-3-oxopropyl}isoindole-1,3(2H)-dione (6). Compound 6 (970
mg) was prepared in 73% yield from 12a37 (723 mg, 3.3 mmol)
and 16b (794 mg, 3.3 mmol) according to general method, mp 129
1
°C. H NMR (DMSO-d6): δ 1.72-1.91 (m, 4H), 2.30-3.05 (m,
4H), 3.19-3.42 (m, 4H), 3.57 (m, 1H), 3.74 (m, 1H), 3.93 (m,
1H), 4.08 (m, 1H), 4.74 + 4.99 (two dd, 1H), 7.86 (m, 4H). Anal.
(C20H21F2N3O4) C, H, N, F.
1
4.5 g (79.7%) of 18 as a white solid; mp 148-150 °C. H NMR
(DMSO-d6): δ 1.36 + 1.40 (s + s, 9H) 1.67-1.96 (m, 4H), 2.36
(d, 1H), 3.01 (m, 1H), 3.14-3.68 (m, 4H), 3.74 (m, 2H), 4.86 (t,
1H). Anal. (C14H22N2O4) C, H, N.
2-{4-Oxo-4-[(2S)-2-(pyrrolidinocarbonyl)pyrrolidin-1-yl]bu-
tyl}-1H-isoindole-1,3(2H)-dione (7). Compound 7 (850 mg) was
prepared in 74% yield from 12b37 (699 mg, 3.0 mmol) and 16a
(614 mg, 3.0 mmol) according to general method, mp 115-116
(2S)-4,4-Difluoro-2-(pyrrolidinocarbonyl)pyrrolidine (16b).
To a solution of 18 (2.26 g, 8 mmol) in dry benzene (20 mL),
DAST (2.44 g, 15.1 mmol) was carefully added at -5 °C under an
argon atmosphere. After 30 min, the reaction mixture was warmed
up to room temperature and was stirred overnight. The mixture
was then poured onto ice-water and the separated aqueous layer
was extracted 3 times with EtOAc. The combined organic phases
were washed successively with water and brine, dried over MgSO4,
filtered and evaporated under reduced pressure. The residue was
purified by column chromatography (silica gel, 30% n-hexane in
acetone) to give 1.24 g (52%) of tert-butyl (2S)-4,4-difluoro-2-
(pyrrolidinocarbonyl)pyrrolidine-1-carboxylate as an off-white solid;
mp 106-107 °C. 1H NMR (DMSO-d6): δ 1.36 + 1.39 (s + s,
9H), 1.71-1.96 (m, 4H), 2.30 (m, 1H), 2.87 (m, 1H), 3.21-3.62
(m, 4H), 3.75 (m, 2H), 4.61 (m, 1H). Anal. (C14H22F2N2O3) C, H,
N, F.
1
°C. H NMR (DMSO-d6): δ 1.62-2.18 (m, 10H), 2.26 (m, 2H),
3.18-3.54 (m, 6H), 3.50 (m, 2H), 4.30 + 4.55 (two dd, 1H), 7.85
(m, 4H). Anal. (C21H25N3O4) C, H, N.
1-{3-Oxo-3-[(2S)-2-(pyrrolidinocarbonyl)pyrrolidin-1-yl]pro-
pyl}-3-phenylquinoxalin-2(1H)-one (8). Compound 8 (153 mg)
was prepared in 43% yield from 14a (247 mg, 0.8 mmol) and 16a
(146 mg, 0.8 mmol) according to general method, mp 223-226
°C. 1H NMR (CDCl3): δ 1.87-2.30 (m, 8H), 2.83 (m, 2H),
3.42-3.85 (m, 6H), 4.57 (m, 1H), 4.70 (m, 1H), 4.75 (m, 1H),
7.34 (t, 1H), 7.48-7.64 (m, 5H), 7.95 (d, 1H), 8.31 (m, 2H). Anal.
(C26H28N4O3) C, H, N.
1-{4-Oxo-4-[(2S)-2-(pyrrolidinocarbonyl)pyrrolidin-1-yl]bu-
tyl}-3-phenylquinoxalin-2(1H)-one (9). Compound 9 (140 mg) was
prepared in 38% yield from 14b (230 mg, 0.78 mmol) and 16a
(160 mg, 0.88 mmol) according to general method, mp 159-160
°C. 1H NMR (CDCl3): δ 1.83-2.38 (m, 10H), 2.53 (m, 2H),
3.35-3.90 (m, 6H), 4.42 (m, 2H), 4.68 (m, 1H), 7.36 (t, 1H), 7.48
(m, 3H), 7.59 (t, 1H), 7.67 (d, 1H), 7.94 (d, 1H), 8.30 (m, 2H).
Anal. (C27H30N4O3) C, H, N.
To a solution of tert-butyl (2S)-4,4-difluoro-2-(pyrrolidinocar-
bonyl)pyrrolidine-1-carboxylate (1.0 g, 0.33 mmol) in EtOAc (3
mL), 3 M HCl in EtOAc (1.0 mL) was added dropwise at room
temperature. After stirring for 20 h, the precipitated crystals were
filtered off, washed with EtOAc, dried and used in the following
step without further purification.
3-{3-Oxo-3-[(2S)-2-(pyrrolidinocarbonyl)pyrrolidin-1-yl]pro-
pyl}-5,5-diphenylimidazolidine-2,4-dione (10). Compound 10 (273
mg) was prepared in 36% yield from 15a38
3-(2-Oxo-3-phenylquinoxalin-1(2H)-yl)propanoic acid (14a).
To a suspension of NaH (220 mg, 5.4 mmol) in dry MeCN (20
mL) under argon, 1339 (999 mg, 4.5 mmol) was carefully added,
followed by the dropwise addition of a MeCN solution of ethyl
2-bromopropionate (825 mg, 4.5 mmol). The reaction mixture was
heated at reflux for 12 h. After cooling to room temperature, the
solvent was removed under reduced pressure, and the residue was
extracted with EtOAc. The combined organic solvent was washed
successively with water and brine, and dried over anhydrous
MgSO4. Removal of the solvent, followed by flash chromatographic
purification gave 1.05 g (72%) of the ethyl ester of 14a. The crude
ethyl ester of 14a (0.72 g, 2 mmol) in 6 N HCl (25 mL) was stirred
at reflux for 10 h, and the reaction mixture was then concentrated
under reduced pressure to give 14a (370 mg, 28%) as a white solid;
(518 mg, 1.6 mmol)
and 16a (337 mg, 1.6 mmol) according to the general method, mp
129 °C (recrystallized from aqueous MeOH).
1H NMR (DMSO-
d
6): δ 1.70-2.19 (m, 8H), 2.41-2.71 (m, 2H), 3.19-3.69 (m, 8H),
4.41 + 4.43 (two dd, 1H), 7.36 (m, 10H), 9.64 (s, 1H). Anal.
(C
27H30N4O4) C, H, N.
3-{4-Oxo-4-[(2S)-2-(pyrrolidinocarbonyl)pyrrolidin-1-yl]bu-
tyl}-5,5-diphenylimidazolidine-2,4-dione (11). Compound (646
mg) 11 was prepared from 15b38 (700 mg, 2.07 mmol) and 16a
(430 mg, 2.1 mmol) according to general method, mp 230-232
°C. 1H NMR (DMSO-d6): δ 1.70-1.95 (m, 10H), 2.04 (m, 1H),
2.17 (t, 1H), 3.15-3.58 (m, 8H), 4.41-4.44 (m, 1H), 7.37 (m,
10H), 9.62 + 9.65 (s + s, 1H). Anal. (C28H32N4O4) C, H, N.
Biological Assays. After removal of the cerebellum, the whole
brains of male rats (Sprague-Dawley, 180-220 g) were homog-
enized in 0.1 M Tris-HCl, 1 mM EDTA buffer, pH ) 7.5. The
homogenate was centrifuged and the 40 000g supernatant was used
at 300 times final dilution in the reaction mixture. Enzyme reactions
were performed at room temperature for 15 min in the presence of
62.5 nM Z-glycyl-prolyl-7-amino-4-methylcoumarin as a highly
specific synthetic substrate of POP. The inhibitory effects of the
compounds were tested at inhibitor concentrations of 100-0.001
nM. The formation of 7-amino-4-methylcoumarin was detected by
spectrofluorometry at an excitation wavelength of 370 nm and an
emission wavelength of 440 nm. The concentrations of the
compounds producing 50% inhibition (IC50) were calculated by
curve fitting of the % inhibition versus inhibitor concentration (M)
plot, using the Hill equation.
X-ray Crystallography. POP purified from porcine muscle was
purchased from La´szlo´ Polga´r in the Institute of Enzymology,
Hungarian Academy of Sciences.45,46 Crystals were grown by
cocrystallization at 4 °C, using the hanging drop vapor diffusion
method. The reservoir composition was as described by Fu¨lo¨p et
al.21 Drops were prepared by mixing equal amounts of protein
solution (10 mg/mL) and the reservoir solution containing the
inhibitor (1 mM). The largest crystals were obtained by micro-
seeding. Data were collected at room temperature, using a Rigaku
R-AXIS IIC image plate detector attached to a Rigaku RU-H2R
1
mp 174-177 °C. H NMR (DMSO-d6 + CDCl3): δ 2.77 (t, 2H),
4.60 (t, 2H), 7.38 (t, 1H), 7.52 (m, 3H), 7.60 (t, 1H), 7.62 (d, 1H),
7.91 (d, 1H), 8.30 (m, 2H), 12.35 (br s, 1H). Anal. (C17H14N2O3)
C, H, N.
4-(2-Oxo-3-phenylquinoxalin-1(2H)-yl)butanoic acid (14b).
Compound 14b (350 mg, 57%) was prepared from 1339 (999 mg,
4.5 mmol) and ethyl 3-bromobutyrate (1053 mg, 5.4 mmol)
1
similarly as for 14a (18 h), as a white solid; mp 172-176 °C. H
NMR (DMSO-d6): δ 1.94 (m, 2H), 2.42 (t, 2H), 4.33 (t, 2H), 7.38
(t, 1H), 7.50 (m, 3H), 7.68 (m, 2H), 7.86 (d, 1H), 8.26 (m, 2H),
12.17 (br s, 1H). Anal. (C18H16N2O3) C, H, N.
General Procedure for Preparation of Compounds 5-11. The
synthesis of 5 is described to illustrate the general method. A
solution of 12a37 (876 mg, 4 mmol) in CH2Cl2 (10 mL) was cooled
to -15 °C under argon and treated with NEt3 (445 mg, 4.4 mmol),
followed by pivaloyl chloride (431 mg, 4.4 mmol), and the mixture
was stirred for 30 min. A solution of 16a (818 mg, 4 mmol) in
CH2Cl2 (5 mL) was added and the reaction mixture was allowed
to warm up to room temperature. After stirring for 5 h, the mixture
was washed successively with water and brine, and dried over
anhydrous MgSO4. Removal of the solvent, followed by flash
chromatographic purification, gave 5 (1.16 g, 83%) as a white solid;
mp 121-122 °C. 1H NMR (DMSO-d6): δ 1.72-2.14 (m, 8H), 2.63
(m, 2H), 3.18-3.59 (m, 6H), 3.75 (m, 2H), 4.49 + 4.63 (two dd,
1H), 7.86 (m, 4H). Anal. (C20H23N3O4) C, H, N.

7520 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Ka´nai et al.
Table 3. X-ray Diffraction Data Collection and Refinement Statistics for POP Complexes with Inhibitors 6, 8 and 11, Respectively
6
8
11
Data Collection
P212121
a ) 72.44, b ) 101.44, c ) 112.25 a ) 72.44, b ) 101.57, c ) 112.17 a ) 72.55, b ) 101.46, c ) 112.33
Space group
P212121
P212121
Cell dimensions (Å)
Resolution range (Å)
Completeness (%)
2.89-74.53
96.5
2.73-34.38
93.5
2.47-34.50
94.1
No. of reflections, redundancy
17989 2.7
21070 4.2
28405 3.4
Refinement
R factor
Rfree (5% of data)
0.1779
0.2570
0.1768
0.2533
0.1744
0.2438
No. of atoms (protein/inhibitor/water) 5574/29/52
5656/33/99
5675/36/132
Average B factors (Å2)
POP main chain
POP side chain
Inhibitor molecule
Water molecules
39.44
41.31
46.35
41.91
35.04
38.06
58.43
40.24
24.58
28.10
41.12
30.51
rms Deviation from Ideal Geometry
Bond lengths (Å)
Bond angles (deg)
Dihedral angles (deg)
0.003
0.809
28.14
0.002
0.718
27.99
0.006
1.182
27.95
rotating anode generator from one crystal for each complex. The
resolutions of data sets were 3.0, 2.8 and 2.5 Å for the complexes
with 6, 8 and 11, respectively. Data processing was carried out
with the programs MOSFLM47 (complexes with inhibitors 8 and
11), BioteX48 (POP-6 complex) and SCALA.49 The scaled dif-
fraction intensities were converted to structure factor amplitudes
by using the program Truncate.50 The structures were isostructural
with those of POP complexes and mutants to be found in the Protein
Data Bank. Rigid body fitting was carried out with the program
AMoRe51 of the Collaborative Computing Project 4 suite,52 with
the protein part of Protein Data Bank53 entry 1QFS. The model
building steps were carried out with program O54 and the resulting
model was refined with X-PLOR, version 3.851.55 During the
refinement, we used torsion angle dynamics (slow cooling from
4000 K), resolution-dependent weighting, overall anisotropic B-
factor refinement, bulk solvent correction and grouped B-factor
refinement. Atomic coordinates have been deposited in the Protein
Data Bank with accession codes 3EQ7, 3EQ8 and 3EQ9 for 6, 8
and 11, respectively. Table 3 shows data collection and refinement
statistics.
or enzyme atom were deleted. Evaporation of water was prevented
by half-harmonic constraints. Residues within 18 Å from the center
of the sphere were allowed to move, while residues in an outer 2
Å shell were fixed. The rest of the system was not included in the
simulations. The AMBER force field of Cornell et al.58 was applied.
Charges were calculated by the multiple fitting procedure59,60 from
an ab initio 6-31G* wave function obtained with the Gaussian 98
program.61 A nonbonded cutoff of 15 Å was used and the SHAKE
algorithm62,63 was applied for bonds involving hydrogen atoms.
NVT simulations with coupling to an external heat bath64 were
performed.
The ligand-enzyme systems were first gradually heated from
10 to 298 K. We applied positional constraints on the heavy atoms
at low temperature and the constraints were gradually released in
the course of heating. Simulations were then performed for 100 ps
with a time step of 1 fs at a temperature of 298 K, using the SYBYL
suite of programs.65 Data collection was performed in the final 70
ps.
The experimental ∆G values were obtained as ∆G ) RT ln Ki,
where Ki, the dissociation constant of a complex, was calculated66
as Ki ) IC50/(1 + S/Km), where S is the substrate concentration
and Km is the Michaelis constant of the substrate.
Computations. Molecular dynamics simulations were performed
to study the binding of the ligands in Table 2 and to make estimates
of the free energies of binding. A modified version of the Linear
Interaction Energy (LIE) method of Åqvist et al.56 was applied.
The free energy change accompanying ligand binding was calcu-
lated as
´
Acknowledgment. The authors are grateful to Drs Agnes
Papp-Behr, Edit Susan and David Durham for valuable contri-
butions to the synthetic work, the biological evaluations of
compounds, and linguistic improvement of the manuscript,
respectively. This work was supported in part by a grant from
the National Scientific Research Foundation (OTKA) NK67800,
F67937 and NI68466 (V.H., G.N.-S.).
∆G ) 1⁄2{Eele(L - E) - Eele(L - W)} + {Evdw(L - E) -
E
vdw(L - W)} + {Esolute(LE) - Esolute(LW)} (1)
where Eele(L - E) is the electrostatic interaction energy between
the ligand and the enzyme, and Eele(L - W) is that between the
ligand and water. The van der Waals components are designated
analogously. Esolute(LE) is the internal energy of the ligand bound
to the enzyme, and Esolute(LW) is the internal energy of the free
ligand in water. This equation differs from that of Åqvist et al. by
the presence of the last term, which can be identified as the ligand
strain accompanying complex formation. In the equation, the factor
1/2 for the electrostatic term corresponds to the ‘ideal’ value of
the LIE method, and it was found that the experimental values could
be reasonably reproduced without multiplying the van der Waals
term by a factor different from 1.
References
(1) Yoshimoto, T.; Orlowski, R. C.; Walter, R. Postproline cleaving
enzyme: identification as serine protease using active site specific
inhibitors. Biochemistry 1977, 16, 2942–2948.
(2) Moriyama, A.; Nakanishi, M.; Sasaki, M. Porcine muscle prolyl
endopeptidase and its endogenous substrates. J. Biochem. (Tokyo)
1988, 104, 112–117.
(3) Walter, R.; Shlank, H.; Glass, J. D.; Schwartz, I. L.; Kerenyi, T. D.
Leucylglycinamide released from oxytocin by human uterine enzyme.
Science 1971, 173, 827–829.
(4) Koida, M.; Walter, R. Post-proline cleaving enzyme. Purification of
this endopeotidase by affinity chromatography. J. Biol. Chem. 1976,
251, 7593–7599.
The setup of the MD simulations was as follows. The ligand
and the enzyme/ligand systems were immersed in a previously
equilibrated TIP3P57 water droplet with 20 Å radius. Water
molecules within 2.5 Å heavy atom separations from any ligand
(5) Yoshimoto, T.; Ogita, K.; Walter, R.; Koida, M.; Tsuru, D. Postproline
cleaving enzyme. Synthesis of a new fluorogenic substrate and
distribution of the endopeptidase in rat tissues and body fluids of man.
Biochem. Biophys. Acta 1979, 569, 184–192.

Prolyl Oligopeptidase Inhibition
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7521
(6) Orlowski, M.; Wilk, E.; Pearce, S.; Wilk, S. Purification and properties
of prolyl endopeptidase from rabbit brain. J. Neurochem. 1979, 33,
461–469.
(25) Fu¨lo¨p, V.; Szeltner, Z.; Polga´r, L. Catalysis of serine oligopeptidases
is controlled by a gating filter mechanism. EMBO Rep. 2000, 1, 277–
281.
(7) Knisatschek, H.; Bauer, K. Characterization of ‘thyroliberin-deami-
dating enzyme’ as a post-proline-cleaving enzyme. Partial purification
and enzyme-chemical analysis of the enzyme from anterior pituitary
tissue. J. Biol. Chem. 1979, 254, 10936–10943.
(8) Kato, T.; Okada, M.; Nagatsu, T. Distribution of postproline cleaving
enzyme in human brain and peripheral tissues. Mol. Cell. Biochem.
1980, 32, 117–121.
(9) Kalwant, S.; Porter, A. G. Purification and characterization of human
brain prolyl endopeptidase. Biochem. J. 1991, 276, 237–244.
(10) Goossens, F.; De Meester, I.; Vanhoof, G.; Scharpe´, S. Distribution
of prolyl oligopeptidase in human peripheral tissues and body fluid.
Eur. J. Clin. Chem. Clin. Biochem. 1996, 34, 17–22.
(11) Kovacs, G. L.; Bohus, B.; Versteeg, D. H. G.; De Kloet, E. R.; De
Wied, D. Effect of oxytocin and vasopressin on memory consolidation:
sites of action and catecholaminergic correlates after local microin-
jections into limbic-midbrain structures. Brain Res. 1979, 175, 303–
314.
(26) Fu¨lo¨p, V.; Szeltner, Z.; Renner, V.; Polga´r, L. Structures of prolyl
oligopeptidase substrate/inhibitor complexes. Use of inhibitor binding
for titration of the catalytic histidine residue. J. Biol. Chem. 2001,
276, 1262–1266.
(27) Szeltner, Z.; Rea, D.; Juha´sz, T.; Renner, V.; Mucsi, Z.; Orosz, G.;
Fu¨lo¨p, V.; Polga´r, L. Substrate-dependent competency of the catalytic
triad of prolyl oligopeptidase. J. Biol. Chem. 2002, 277, 44597–44605.
(28) Szeltner, Z.; Rea, D.; Renner, V.; Fu¨lo¨p, V.; Polga´r, L. Electrostatic
effects and binding determinants in the catalysis of prolyl oligopep-
tidase. Site-specific mutagenesis at the oxyanion binding site. J. Biol.
Chem. 2002, 277, 42613–42622.
(29) (a) 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. Pyrrolidinyl pyridone and pyrazinone analogues as potent
inhibitors of prolyl oligopeptidase (POP). Bioorg. Med. Chem. Lett.
2008, 18, 4360–4363. (b) Jarho, E. M.; Vena¨la¨inen, J. I.; Poutiainen,
S.; Leskinen, H.; Vepsna¨la¨inen, J.; Christiaans, J. A. M.; Forsberg,
M. M.; Ma¨nnisto¨, P. T.; Walle´n, E. A. A. 2(S)-(Cycloalk-1-enecar-
bonyl)-1-(4-phenylbutanoyl)pyrrolidines and 2(S)-(aroyl)-1-(4-phe-
nylbutanoyl)pyrrolidines as prolyl oligopeptidase inhibitors. Bioorg.
Med. Chem. 2007, 15, 2024–2031. (c) Jarho, E. M.; Vena¨la¨inen, J. I.;
Juntunen, J.; Yli-Kokko, A. L.; Vepsna¨la¨inen, J.; Christiaans, J. A. M.;
Forsberg, M. M.; Ja¨rvinen, T.; Ma¨nnisto¨, P. T.; Walle´n, E. A. A. An
introduction of a pyridine group into the structure of prolyl oligopep-
tidase inhibitors. Bioorg. Med. Chem. Lett. 2006, 16, 5590–5593. (d)
Vena¨la¨inen, J. I.; Walle´n, E. A.; A.; Poso, A.; Garc´ıa-Horsman, J. A.;
Ma¨nnisto¨, P. T. Synthesis and characterization of the novel fluorescent
prolyl oligopeptidase inhibitor 4-fluoresceinthiocarbonyl-6-aminoca-
proyl-L-prolyl-2(S)-hydroxyacetyl)pyrrolidine. J. Med. Chem. 2005,
48, 7093–7095. (e) Jarho, E. M.; Walle´n, E. A. A.; Christiaans,
J. A. M.; Forsberg, M. M.; Vena¨la¨inen, J. I.; Ma¨nnisto¨, P. T.; Gynther,
J.; Poso, A. Dicarboxylic acid azacycle L-prolyl-pyrrolidine amides
as prolyl oligopeptidase inhibitors and three-dimensional quantitative
structure-activity relationship of the enzyme - inhibitor interactions.
J. Med. Chem. 2005, 48, 4772–4782. (f) Jarho, E. M.; Vena¨la¨inen,
J. I.; Huuskonen, J.; Christiaans, J. A. M.; Garc´ıa-Horsman, J. A.;
Forsberg, M. M.; Ja¨rvinen, T.; Gynther, J.; Ma¨nnisto¨, P. T.; Walle´n,
E. A. A. A cyclopent-2-enecarbonyl group mimics proline at the P2
position of prolyl oligopeptidase inhibitors. J. Med. Chem. 2004, 47,
5605–5607. (g) Walle´n, E. A. A.; Christiaans, J. A. M.; Jarho, E. M.;
Forsberg, M. M.; Vena¨la¨inen, J. I.; Ma¨nnisto¨, P. T.; Gynther, J. New
prolyl oligopeptidase inhibitors developed from dicarboxylic acid bis(L-
prolyl-pyrrolidine) amides. J. Med. Chem. 2003, 46, 4543–4551. (h)
Walle´n, E. A. A.; Christiaans, J. A. M.; Saarinen, T. J.; Jarho, E. M.;
Forsberg, M. M.; Vena¨la¨inen, J. I.; Ma¨nnisto¨, P. T.; Gynther, J.
Conformationally rigid N-acyl-5-alkyl-L-prolyl-pyrrolidines as prolyl
oligopeptidase inhibitors. Bioorg. Med. Chem. 2003, 11, 3611–3619.
(i) Walle´n, E. A. A.; Christiaans, J. A. M.; Forsberg, M. M.;
Vena¨la¨inen, J. I.; Ma¨nnisto¨, P. T.; Gynther, J. Dicarboxylic acid bis(L-
prolyl-pyrrolidine) amides as prolyl oligopeptidase inhibitors. J. Med.
Chem. 2002, 45, 4581–4584. (j) Walle´n, E. A. A.; Christiaans, J. A. M.;
Saario, S. M.; Forsberg, M. M.; Vena¨la¨inen, J. I.; Paso, H. M.;
Ma¨nnisto¨, P. T.; Gynther, J. 4-Phenylbutanoyl-2(S)-acylpyrrolidines
and 4-phenylbutanoyl-L-prolyl-2(S)-acylpyrrolidines as prolyl oli-
gopeptidase inhibitors. Bioorg. Med. Chem. 2002, 10, 2199–2206.
(12) Maes, M.; Goossens, F.; Scharpe´, S.; Meltzer, H. Y.; D’Hondt, P.;
Cosyns, P. Lower serum prolyl endopeptidase enzyme activity in
major depression: further evidence that peptidases play a role in
the pathophysiology of depression. Biol. Psychiatry 1994, 35, 545–
552.
(13) Welches, W. R.; Brosnihan, K. B.; Ferrario, C. M. A comparison of
properties and enzymatic activities of three angiotensin processing
enzymes: angiotensin converting enzyme, prolyl endopeptidase and
neutral endopeptidase 24.11. Life. Sci. 1993, 52, 1461–1480.
(14) (a) Brandt, I.; Scharpe´, S.; Lambeir, A. M. Suggested function for
prolyl oligopeptidase: A puzzling paradox. Clin. Chim. Acta 2007,
377, 50–61. (b) Williams, R. S. B. Pharmacogenetics in model systems:
Defining a common mechanism of action for mood stabilizers. Prog.
Neuropsychopharmacol. Biol. Psychiatry 2005, 29, 1029–1037. (c)
Williams, R. S. B.; Cheng, L.; Mudge, A. W.; Harwood, A. J. A
common mechanism of action for three mood-stabilizing drugs. Nature
2002, 417, 292–295. (d) 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, 5813–5820.
(15) (a) Garc´ıa-Horsman, J. A.; Ma¨nnisto¨, P. T.; Vena¨la¨inen, J. I. On the
role of prolyl oligopeptidase in health and disease. Neuropeptidase
2007, 41, 1–24. (b) Gass, J.; Khosla, C. Prolyl endopeptidase. Cell.
Mol. Life Res. 2007, 64, 345–355. (c) Williams, R. S. B. Prolyl
oligopeptidase and bipolar disorder. Clin. Neurosci. Res. 2004, 4, 233–
242.
(16) Cacabelos, R.; Alvarez, A.; Lombardi, V.; Ferna´ndez-Novoa, L.; Corzo,
L.; Pe´rez, P.; Laredo, M.; Pichel, V.; Herna´ndez, A.; Varela, M.;
Figuero, J., Jr.; Windisch, M.; Vigo, C. Pharmacological treatment of
Alzheimer disease: from psychotropic drugs and cholinesterase inhibi-
tors to pharmacogenomics. Drugs Today 2000, 36, 415–499.
(17) De Nanteuil, G.; Portevin, B.; Lepagnol, J. Prolyl endopeptidase
inhibitors: a new class of memory enhancing drugs. Drugs Future
1998, 23, 167–179.
(18) Vanhoof, G.; Goossens, F.; Hendriks, L.; De Meester, I.; Hendriks,
D.; Vriend, G.; Van Broeckhoven, C.; Scharpe´, S. Cloning and
sequence analysis of the gene encoding human lymphocyte prolyl
endopeptidase. Gene 1994, 149, 363–366.
(19) Goossens, F.; De Meester, I.; Vanhoof, G.; Hendriks, D.; Vriend, G.;
Scharpe´, S. The purification, characterization and analysis of primary
and secondary structure of prolyl endopeptidase from human lym-
phocytes. Evidence that the enzyme belongs to the R/ꢀ hydrolase fold
family. Eur. J. Biochem. 1995, 233, 432–441.
(30) Ka´nai, K.; Erdõ, S.; Susa´n, E.; Fehe´r, M.; Sipos, J.; Poda´nyi, B.;
Szappanos, A.; Hermecz, I. Prolyl endopeptidase inhibitors: N-acyl
derivatives of L-thioproline-pyrrolidine. Bioorg. Med. Chem. Lett. 1997,
7, 1701–1704.
(20) Polga´r, L. Structural relationship between lipases and peptidases of
the prolyl oligopeptidase family. FEBS Lett. 1992, 311, 281–284.
(21) Fu¨lo¨p, V.; Bo¨cskei, Zs.; Polga´r, L. Prolyl oligopeptidase: an unusual
ꢀ-propeller domain regulates proteolysis. Cell 1998, 94, 161–170.
(22) Bo¨cskei, Zs.; Fuxreiter, M.; Na´ray-Szabo´, G.; Szabo´, E.; Polga´r, L.
Crystallization and preliminary X-ray analysis of porcine muscle prolyl
oligopeptidase. Acta Crystallogr. 1998, D54, 1414–1415.
(23) (a) Szeltner, Z.; Polga´r, L. Structure, function and biological relevance
of prolyl oligopeptidase. Curr. Protein Pept. Sci. 2008, 9, 96–107.
(b) Shan, L.; Mathews, I. I.; Khosla, C. Structural and mechanistic
analysis of two prolyl endopeptidases: Role of interdomain dynamics
in catalysis and specificity. Proc. Natl. Acad. Sci. U.S.A. 2005, 102,
3599–3604. (c) Fuxreiter, M.; Magyar, Cs.; Juha´sz, T.; Szeltner, Z.;
Polga´r, L.; Simon, I. Flexibility of prolyl oligopeptidase: Molecular
Dynamics and molecular framework analysis of the potential substrate
pathways. Proteins 2005, 60, 504–512.
(31) Fehe´r, M.; Ka´nai, K.; Poda´nyi, B.; Hermecz, I. Conformational analysis
of two inhibitors of prolyl endopeptidase: comparison of 4-phenyl-
butyryl- and octanoyl-prolyl-pyrrolidine. Quant. Struct.-Act. Relat.
1997, 16, 136–141.
´
(32) Ka´nai, K.; Feher, M.; Lopata, A.; Poda´nyi, B.; Nova´k, I.; Susa´n, E.;
Hermecz, I.; Ara´nyi, P. Molecular modeling and CoMFA studies on
prolyl endopeptidase inhibitors. Acta Pharm. Hung. 1999, 69, 240-
246; Chem. Abstr. 2000, 133, 86.
(33) Hermecz, I.; Ka´nai, K. Prolyl endopeptidase inhibitors. Il Farmaco
2000, 55, 188–190.
(34) Ka´nai, K.; Erdõ, S.; Szappanos, A.; Bence, J.; Hermecz, I.; Szvoboda,
´
Gy.; Ba´tori, S.; He´ja, G.; Balogh, M.; Horva´th, A.; S´ıpos, J.; Barta-
(24) Yoshimoto, T.; Kawahara, K.; Matsubara, F.; Kado, K.; Tsuru, D.
Comparison of inhibitory effects of prolinal-containing peptide deriva-
tives on prolyl endopeptidases from bovine brain and Flavobacterium.
J. Biochem. 1985, 98, 975–979.
Bodor, V.; Pa´rka´ny, Zs.; Lakics, V.; Mola´r, P. Preparation of proline
pyrrolidine amide derivatives as prolyl endopeptidase inhibitors. WO
97/07,116. Chem. Abstr. 1997, 126, 238661.

7522 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Ka´nai et al.
(35) Fehe´r, M.; Ka´nai, K.; Hermecz, I.; Lopata, A.; Nova´k, I. A 3D-QSAR
analysis of prolyl endopeptidase inhibitors. Quant. Struct.-Act. Relat.
1999, 18, 456–463.
(36) Arai, H.; Nishioka, H.; Niwa, S.; Yamanaka, I.; Tanaka, Y.; Yoshinaga,
K.; Kobayashi, N.; Miura, N.; Ikade, Y. Synthesis of prolyl endopep-
tidase inhibitors and evaluation of their structure-activity relationships:
in vitro inhibition of prolyl endopeptidase from canine brain. Chem.
Pharm. Bull. 1993, 41, 1583–1588.
(52) Collaborative Computing Project 4 The CCP4 suite: programs for
protein crystallography Acta Cryst. 1994, D50, 760-763.
(53) Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T. N.;
Weissig, H.; Shindyalov, I. N.; Bourne, P. E. The Protein Data Bank.
Nucleic Acids Res. 2000, 28, 235–242.
(54) Jones, T. A.; Zou, J.-Y.; Cowan, S. W.; Kjeldgaard, M. Improved
methods for building protein models in electron density maps and the
location of errors in these models. Acta Crystallogr. 1994, A47, 110–
119.
(37) Chapman, J. M.; Cocolas, G. H.; Hall, I. H. Hypolipidemic activity
of phthalimide derivatives. 1. N-substituted phthalimide derivatives.
J. Med. Chem. 1979, 22, 1399–1402.
(55) Brunger, A. T. XPLOR: A System for X-ray crystallography and NMR
Version 3, Yale University, New Haven, CT, 1992.
(38) (a) Shaffer, J. W.; Steinberg, E.; Krimsley, V.; Winstead, M. B.
Substitution in the hydantoin ring. VII. N-propionic acid and its ethyl
ester and N-3-(2-cyanoethyl) derivatives. J. Med. Chem. 1968, 11,
462–466. (b) Sayo, H.; Hosokawa, M. Result of spin immunoassay
for simultaneous measurement of phenytoin and phenobarbital in serum
compared with those of liquid chromatography. Chem. Pharm. Bull.
1998, 33, 2866–2870.
(39) Nishio, T. (2 + 2) photocycloaddition of the carbon-nitrogen double
bond of quinoxalin-2(1H)-ones to electron-deficient olefins. J. Org.
Chem. 1984, 49, 827–832.
(40) (a) Hagmann, W. K. The many roles for fluorine in medicinal
chemistry. J. Med. Chem. 2008, 51, 4359–4369. (b) Purser, S.; Moore,
P. R.; Swallow, S.; Gouverneur, V. Fluorine in medicinal chemistry.
Chem. Soc. ReV. 2008, 37, 320–330. (c) Shah, P.; Westwell, A. D.
The role of fluorine in medicinal chemistry. J. Enzyme Inhib. Med.
Chem. 2007, 22, 527–540. (d) Mu¨ller, K.; Faeh, C.; Diederich, F.
Fluorine in pharmaceuticals: Looking beyond intuition. Science 2007,
317, 1881–1886. (e) Yoder, N. C.; Kumar, K. Fluorinated amino acids
in protein design and engineering. Chem. Soc. ReV. 2002, 31, 335–
341. (f) O’Hagan, D.; Rzepa, H. S. Some influences of fluorine in
bioorganic chemistry. Chem. Commun. 1997, 645–652.
(56) Åqvist, J.; Medina, C.; Samuelsson, J.-E. A new method for predicting
binding affinity in computer-aided drug design. Protein Eng. 1994, 7,
385–391.
(57) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.;
Klein, M. L. Comparison of simple potential functions for simulating
liquid water. J. Chem. Phys. 1983, 79, 926–935.
(58) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.,
Jr.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.;
Kollman, P. A. A second generation force field for the simulation of
proteins, nucleic acids, and organic molecules. J. Am. Chem. Soc. 1995,
117, 5179–5197.
(59) Ferenczy, G. G. Charges derived from distributed multipole series.
J. Comput. Chem. 1991, 12, 913–917.
´
(60) Chipot, C.; Angya´n, J. G.; Ferenczy, G. G.; Scheraga, H. A.
Transferable net atomic charges from a distributed multipole analysis
for the description of electrostatic properties: a case study of saturated
hydrocarbons. J. Phys. Chem. 1993, 97, 6628–6636.
(61) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, Jr. J. A.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels,
¨
A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone,
(41) Postels, H.-T.; Ko¨nig, W. A. What is the structure of peptide antibiotic
Tu¨ 1718 B? previously proposed structure disproved by synthesis.
Liebigs Ann. Chem. 1992, 1281–1287.
V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.;
Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Koma´romi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara,
A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-Gordon, M.;
Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.6; Gaussian,
Inc.: Pittsburgh PA, 1998.
´
(42) (a) Demange, L.; Me´nez, A. E.; Dugave, C. Practical synthesis of
Boc and Fmoc protected 4-fluoro and 4-difluoroprolines from trans-
4-hydroxyroproline. Tetrahedron Lett. 1998, 39, 1169–1172. (b)
Middleton, W. J. New fluorinating reagents. Dialkylaminosulfur
fluorides. J. Org. Chem. 1975, 40, 574–578.
(43) Poda´nyi, B.; Bokotey, S.; Ka´nai, K.; Fehe´r, M.; Hermecz, I. Investiga-
tions on flexible prolyl-endopeptidase inhibitor in solution by NMR
techniques. Magn. Reson. Chem. 1999, 37, 346–352.
(44) Bostro¨m, J.; Norrby, P.-O.; Liljefors, T. Conformational energy
penalties of protein-bound ligands. J. Comput.-Aided Mol. Des. 1998,
12, 383–396.
(45) Polga´r, L. pH-Dependent mechanism in the catalysis of prolyl
oligopeptidase from pig muscle. Eur. J. Biochem. 1991, 197, 441–
447.
(62) Ryckaert, J.-P.; Ciccotti, G.; Berendsen, H. J. C. Numerical integration
of the Cartesian equations of motion of a system with constraints:
molecular dynamics of n-alkanes. J. Comput. Phys. 1977, 23, 327–
341.
(63) van Gunsteren, W. F.; Berendsen, H. J. C. Algorithms for macromo-
lecular dynamics and constraint dynamics. Mol. Phys. 1977, 34, 1311–
1327.
(64) Berendsen, H. J. C.; Postma, J. P. M.; van Gunsteren, W. F.; DiNiola,
A.; Haak, J. R. Molecular dynamics with coupling to an external bath.
J. Chem. Phys. 1984, 81, 3684–3690.
(65) SYBYL Version 6.7, Tripos. Inc., St. Louis, MO.
(66) Cheng, Y.-C.; Prusoff, W. H. Relationship between inhibition constant
(Ki) and the concentration of inhibitor which causes 50% inhibition
(IC50) of an enzymatic reaction. Biochem. Pharmacol. 1973, 22, 3099–
3108.
(46) Polga´r, L. Prolyl oligopeptidases. Methods Enzymol. 1994, 244, 188–
200.
(47) Leslie, A. G. W. Molecular data processing. In Crystallographic
Computing 5: From Chemistry to Biology, Moras, D., Podjarny, A.,
Thierry, J., Eds.; Oxford University Press: Oxford, 1991; chapter 4,
pp 50-61.
(48) BioteX, Version 1.0, Molecular Structure Corporation, The Woodlands,
TX, 1995.
(49) Evans, P. R. Proc. of CCP4 Study Weekend on Data Collection &
Processing. Data Reduction 1993, 114–122.
(50) French, S.; Wilson, K. On the treatment of negative intensity
observations. Acta Crystallogr. 1978, A34, 517–525.
(51) Navaza, J. AMoRe: an automated package for molecular replacement.
Acta Crystallogr. 1994, A50, 157–163.
(67) Kraulis, P. J. MOLSCRIPT: a program to produce both detailed and
schematic plots of protein structures. J. Appl. Crystallogr. 1991, 24,
945–949.
JM800944X