Discovery of Nanomolar Dengue and West Nile Virus Protease Inhibitors Containing a 4-Benzyloxyphenylglycine Residue

Article abstract of DOI:10.1021/acs.jmedchem.5b01441

The dengue virus (DENV) and West Nile Virus (WNV) NS2B-NS3 proteases are attractive targets for the development of dual-acting therapeutics against these arboviral pathogens. We present the synthesis and extensive biological evaluation of inhibitors that

Full text of DOI:10.1021/acs.jmedchem.5b01441

Article
pubs.acs.org/jmc
Discovery of Nanomolar Dengue and West Nile Virus Protease
Inhibitors Containing a 4‑Benzyloxyphenylglycine Residue
Mira A. M. Behnam,† Dominik Graf,† Ralf Bartenschlager,‡,∥ Darius P. Zlotos,§ and Christian D. Klein*,†
†Medicinal Chemistry, Institute of Pharmacy and Molecular Biotechnology IPMB, Heidelberg University, Im Neuenheimer Feld 364,
69120 Heidelberg, Germany
‡Department of Infectious Diseases, Molecular Virology, Heidelberg University, Im Neuenheimer Feld 345, 69120 Heidelberg,
Germany
∥German Center for Infection Research, Heidelberg University, Im Neuenheimer Feld 345, 69120 Heidelberg, Germany
§Department of Pharmaceutical Chemistry, The German University in Cairo, New Cairo City, 11835 Cairo, Egypt
S
* Supporting Information
ABSTRACT: The dengue virus (DENV) and West Nile Virus
(WNV) NS2B-NS3 proteases are attractive targets for the
development of dual-acting therapeutics against these arboviral
pathogens. We present the synthesis and extensive biological
evaluation of inhibitors that contain benzyl ethers of 4-
hydroxyphenylglycine as non-natural peptidic building blocks
synthesized via a copper-complex intermediate. A three-step
optimization strategy, beginning with fragment growth of the
C-terminal 4-hydroxyphenylglycine to the benzyloxy ether,
followed by C- and N-terminal optimization, and finally
fragment merging generated compounds with in vitro affinities
in the low nanomolar range. The most promising derivative
reached Ki values of 12 nM at the DENV-2 and 39 nM at the
WNV proteases. Several of the newly discovered protease inhibitors yielded a significant reduction of dengue and West Nile virus
titers in cell-based assays of virus replication, with an EC50 value of 3.4 μM at DENV-2 and 15.5 μM at WNV for the most active
analogue.
Both DENV and WNV belong to the Flaviviridae family,
together with the hepatitis C virus (HCV) and several other
pathogenic viruses. There are four related, yet antigenically
distinct serotypes of dengue (DENV 1−4), in addition to an
emerging fifth serotype.7 The flavivirus genome consists of a
positive single-stranded RNA, which is translated into a
precursor polyprotein. The latter is then cleaved co- and
post-translationally, resulting in a small number of structural
and nonstructural (NS) proteins. The NS3 protein contains a
serine protease domain, whose activity depends on the
formation of a noncovalent complex with the NS2B protein
as cofactor.8−10 The NS2B-NS3 complex is responsible for the
proteolytic processing of the viral polyprotein. It is hence
crucial for virus replication and represents a promising target
for the development of antiviral drugs. Protease inhibition was
shown to be a successful strategy in the treatment of HIV and
HCV infections,11−13 with most approved candidates being
substrate-mimetics. While DENV and WNV proteases have
similarities in their binding pockets that can potentially be used
to develop dual inhibitors, they also share the challenges of
INTRODUCTION
■
Dengue virus (DENV) is the arthropod-borne virus (arbovirus)
with the utmost impact on public health. Infection with DENV
can cause a spectrum of symptoms ranging from the most
common dengue fever, an acute self-limiting febrile illness, to
the more severe and potentially fatal dengue hemorrhagic fever
(DHF) and dengue shock syndrome (DSS). With an estimate
of 50−100 million infections annually1 and a recent analysis
suggesting a higher infection burden of 390 million cases in
2010,2,3 dengue is prioritized by the World Health Organization
(WHO) among 17 neglected tropical diseases. A related,
medically relevant arboviral pathogen is the West Nile virus
(WNV). Infection with WNV is sometimes asymptomatic but
can develop into flu-like illness and fatal WNV neuroinvasive
disease.4 Global warming, climate changes, increased global
travel, urbanization, and insufficient vector control cause a
continuous habitat expansion of the disease-transmitting
mosquito vectors (Aedes and Culex spp.). This results in the
rapid spread of infections beyond their original geographical
boundaries.5,6 The development of antiviral chemotherapeutics
or vaccines against dengue and WNV remains a critical, unmet
medical need.
Received: September 16, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 1
a
Spheres represent Rink amide resin particles. (a) H2O/ACN, DIPEA, Fmoc-OSu; (b) HCl; (c) NaOH, CuSO4·5H2O; (d) NaOH, MeOH,
respective alkylating agent, rt, 18 h; (e) HCl or 8-hydroxyquinoline; (f) piperidine, DMF; (g) HATU, DIPEA, DMF; (h) Fmoc-Lys(Boc)-OH,
HATU, DIPEA, DMF; (i) Fmoc-Arg(Pbf)-OH, HATU, DIPEA, DMF; (j) Cap-COOH, HATU, DIPEA, DMF; (k) TFA, H2O, TIPS (95:4:1 v/v, 2
h, rt).
possessing a solvent-exposed, topologically shallow active site
and a selectivity for substrates containing basic amino acids
(arginine and lysine) at P1 and P2.14,15 These highly polar
groups account for the insufficient membrane permeability and
poor pharmacokinetic profile of substrate-based inhibitors,
despite their promising in vitro binding affinities. Both
proteases are inhibited by the polypeptide aprotinin,16,17 in
addition to various classes of nonpeptidic and peptidic
inhibitors.17−23
sequence (Nle-Lys-Arg-Arg). This sequence contains the
favored amino acids for the S1−S4 subsites of the protease
binding cavity.16,27,28 For DENV-2, we found the benzoyl-
capped retro-tripeptide lacking one arginine (Bz-Arg-Lys-Nle-
NH2) to be superior in terms of inhibitory activity, and this was
chosen as the standard sequence for our subsequent work.29
Our efforts to enhance activity and membrane permeability
span between optimizing the N-terminal capping group and,
more recently, the C-terminal residue. At the N-terminus,
thiazolidinedione-capped peptide hybrids bearing hydrophobic
substituents showed promising in vitro affinities at DENV-2
protease but displayed much lower inhibition at WNV
Potent peptidic inhibitors were obtained for the DENV-224,25
and WNV26 proteases by attaching a C-terminal, serine-reactive
electrophilic functionality to an N-benzoyl capped tetrapeptide
B
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 1. Inhibitory Activity of Various N-Benzoyl-Capped Tripeptide Sequences against DENV and WNV Proteases
DENV
IC50 (μM)
WNV
a
b
c
d
no.
sequence
Bz-Arg-Lys-L-Phg-NH2
%
%
IC50 (μM)
I31
23
24
25
26
27
28
29
30
31
32
33
95.0
99.3
96.6
98.0
99.9
100.1
16.4
31.0
77.0
26.8
29.4
23.5
3.323
3.790
9.695
2.032
0.935
0.367
n.d.
39.3
99.0
79.2
75.1
88.0
101.2
11.4
27.5
91.7
19.9
71.3
64.0
58.10
3.780
n.d.
Bz-Arg-Lys-D-Phg-NH2
Bz-Arg-Lys-(4-OH)-L-Phg-NH2
Bz-Arg-Lys-(4-OH)-D-Phg-NH2
n.d.
Bz-Arg-Lys-(4-benzyloxy)-L-Phg-NH2
Bz-Arg-Lys-(4-benzyloxy)-D-Phg-NH2
Bz-Lys-(4-benzyloxy)-D-Phg-NH2
Bz-Arg-(4-benzyloxy)-D-Phg-NH2
Bz-Lys-Arg-(4-benzyloxy)-D-Phg-NH2
Bz-D-Lys−D-Arg-(4-benzyloxy)-D-Phg-NH2
Bz-D-Arg-D-Lys-(4-benzyloxy)-L-Phg-NH2
Bz-D-Lys-D-Arg-(4-benzyloxy)-L-Phg-NH2
n.d.
0.728
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
a
b
Percent inhibition of DEN NS2B-NS3 protease serotype 2 (inhibitor 50 μM, substrate 50 μM). IC50 values against DEN NS2B-NS3 protease
c
d
serotype 2 at 50 μM substrate concentration. Percent inhibition of WNV NS2B-NS3 protease (inhibitor 50 μM, substrate 50 μM). IC50 values
against WNV NS2B-NS3 protease at 50 μM substrate concentration. n.d. = IC50 was not determined for compounds with less than 95% inhibition.
All measurements were carried out in triplicate. Standard deviations were below 10%. Phg is phenylglycine.
protease.30 At the C-terminus, exchanging norleucine for
phenylglycine resulted in a 4-fold enhancement of IC50 at
DENV-2 protease. Merging both activity-enhancing fragments
provided inhibition of dengue protease in the upper nanomolar
range.31 A similar range of inhibition at the WNV protease
could be achieved by fine-tuning of the nonpolar substitutent
on the thiazolidinedione scaffold.32 For the sequence (Bz-Arg-
Lys-Phg-NH2), (4-amidino)-L-phenylalanine was found to be a
superior mimetic of arginine at dengue protease, resulting in 8-
fold higher potency. However, the improvement in activity did
not extend to WNV protease.33
As our initial attempts at C-terminal amino acid modification
proved quite rewarding, we decided to further explore and
optimize this region through fragment growing, starting from
phenylglycine. In parallel, the smaller capping groups at the N-
terminus was explored. Optimization at the N-terminus was
considered crucial for two reasons: first, to refrain from the
previous thiazolidinedione scaffold and its limited capability for
the development of potent dual inhibitors; and second, to allow
for sufficient structural expansion to extensively probe the
binding site around the C-terminal residue, while achieving an
optimum balance between molecular size and high antiviral
activity in cell culture.
cine.35,36 Briefly, D- or L-(4-hydroxy)-phenylglycine was treated
with one equivalent of NaOH, followed by one equivalent of
CuSO4 at 50 °C. This temperature was selected after multiple
rounds of optimization and is essential for quantitative complex
formation and final product purity. After deprotonation of the
hydroxyl group using another equivalent of NaOH, the
phenolate of the Cu-complex was alkylated using the respective
alkyl bromide, followed by acidification with HCl to release the
unprotected 4-hydroxyphenylglycine ethers (3a−9a, 11a, 13a−
22a, and 108a) in 32−68% overall yield. For the 4-
trifluoromethoxybenzyl and 2-trifluoromethylbenzyl analogues
10a and 12a, the yield was 12% and 20%, respectively, because
of the gelatinous nature of the copper-complex that led to loss
of material during filtration. The copper complex of 21a was
soluble after HCl acidification, and an alternative workup based
on copper complexation by 8-quinolinol37 was used instead. All
products were obtained in high purity, and no subsequent
purification steps were required. As opposed to the traditional
strategy for amino acid protection, such as esterification of the
carboxylic acid and N-Boc formation, the Cu-complex approach
requires fewer steps, a simple workup, and no chromatographic
purification, rendering it the method of choice for O-alkylation
of hydroxyphenylglycine. All hydroxyphenylglycine derivatives
were N-Fmoc protected using the procedure by Dener et al.38
to give 1, 2, 3b−22b, and 108b, then coupled to Rink amide
resin using HATU/DIPEA. Peptide synthesis proceeded
according to the Fmoc-protocol as previously described39
until the desired sequence was obtained. A final wash with 25%
piperidine in DMF was used for compounds 24 and 25 to
cleave any esterification on the unprotected phenolic hydroxyl
group.
We now report the synthesis and biological evaluation of a
new series of dual inhibitors of DENV-2 and WNV proteases
comprising substituted benzyl and phenacyl ethers of 4-
hydroxyphenylglycine as C-terminal building blocks.
RESULTS AND DISCUSSION
■
Chemistry. The structures of all target molecules are shown
in Tables 1−4. Synthetic routes are outlined in Scheme 1,
except for compounds 23−25 and 28−33, which contain
alternative peptide sequences. The 4-hydroxyphenylglycine
analogues (3a−22a) were obtained by O-alkylation of the
commercially available (4-hydroxy)-D-phenylglycine and (4-
hydroxy)-L-phenylglycine using appropriately substituted alkyl
halogenides according to the procedure first described by
Structure−Activity Relationships. (SAR). Four series of
peptide hybrids (Tables 1−4) were synthesized and evaluated
in fluorimetric assays against DENV and WNV proteases and
thrombin as a potential off-target. Furthermore, selected
compounds were assessed in a HPLC-based DENV or WNV
protease assay to exclude false-positive values.
Choice of Sequence. The C-terminal residue modifica-
tions recently reported by us revealed steep structure−activity
relationships and substantial improvement in activity with Bz-
Lys-Arg-Phg-NH2 (I) showing the highest affinity to DENV
protease (IC50 = 3.32 μM). The corresponding D-phenylglycine
Wunsch et al. for O-benzylation of tyrosine.34 The approach
̈
involves the formation of a copper complex as a protecting
group for the amino acid moiety prior to the alkylation step and
was also successfully applied for stereospecific protection of a
phenolic hydroxyl group in tyrosine and hydroxyphenylgly-
C
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 2. Inhibitory Activity of N-Benzoyl-Capped Tripeptides with the General Structure (Bz-Arg-Lys-X-NH2), Where X
Represents Ethers of (4-Hydroxy)-D-phenylglycine as C-Terminal Amino Acid, against DENV and WNV Proteases
a
b
IC50 values against DEN NS2B-NS3 protease serotype 2 at 50 μM substrate concentration. Percent inhibition of WNV NS2B-NS3 protease
c
(inhibitor 50 μM, substrate 50 μM). IC50 values against WNV NS2B-NS3 protease at 50 μM substrate concentration. n.d. = IC50 was not
determined for compounds with less than 98% inhibition. Percent inhibition of thrombin (inhibitor 25 μM, substrate 50 μM) did not exceed 30%
(cf. Supporting Information for details). All measurements were carried out in triplicate. Standard deviations were below 10%.
analogue (23) was equipotent to the L-stereoisomer.31
Encouraged by this success, we allocated our research efforts
to extensively explore the impact of structural modification of
the C-terminal phenylglycine by 4-hydroxy substitution (Table
1). The sequence incorporating (4-hydroxy)-D-phenylglycine
Attempts to shorten the peptidic inhibitors by omitting one
basic group in 28 and 29 proved detrimental to the activity at
both targets. Switching the position of lysine and arginine in 30
was associated with moderately lower inhibition as opposed to
the alternative structure with basic amino acids in D-
configuration (31), which displayed a pronounced loss of
activity. To obtain more insight into the orientation of the ether
sequence in the binding cavity of both targets, we investigated
an interchange of C- and N-terminal residues through the
synthesis of sequences 32 and 33.40 Both molecules were far
less active than 27. Docking studies of phenylglycine-containing
peptide hybrids on DENV-231,32 and WNV NS2B-NS332
suggested the positioning of the C-terminal phenylglycine in
the S1 pocket and the N-capping group in S3 and S4 pockets for
both targets. These findings together with the experimental
results suggest an accommodation of the 4-hydroxyphenylgly-
cine ether in the S1 subsite of the binding cavity in both DENV-
2 and WNV proteases, rather than an exchange of its binding
location with that of the N-terminal moiety.
(25) showed slightly higher affinity to DENV protease (IC50
=
2.03 μM). Consequently, this amino acid was then used as an
anchor to extend the structure by fragment growth through the
metabolically stable ether linkage. Interestingly, O-benzylation
of 25 to give the first ether 27 led to substantial improvement
of inhibitory potency. In comparison to reference I, 27
displayed a 9-fold increase in inhibitory potential at DENV
protease, in addition to a remarkable enhancement of inhibition
at WNV protease by a factor of 80, representing a very
promising dual inhibition profile for both targets of interest.
While our earlier findings suggested little contribution of the
stereochemistry at the C-terminal position, which we showed
for L- and D-norleucine31 and here for L- and D-phenylglycine (I
and 23), this changed with the incorporation of the hydroxyl
group, where a clear preference for D-configuration could be
observed for the hydroxyphenylglycine analogues (25 vs 24)
and for the corresponding benzyl ethers (27 vs 26).
Variation of the C-Terminal Ether. After evaluating the
peptide sequence, we proceeded to further probe the binding
area surrounding the benzyl ether moiety in 27 (Table 2). The
diverse choice of ether substituents covered a range of
D
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 3. Inhibitory Activity of N-Capped Tripeptides Possessing the General Structure (Cap-Arg-Lys-X-NH2), X Being the C-
Terminal Amino Acid 4-(Benzyloxy)-D-phenylglycine, against DENV and WNV Proteases
a
b
IC50 values against DEN NS2B-NS3 protease serotype 2 at 50 μM substrate concentration. Percent inhibition of WNV NS2B-NS3 protease
c
(inhibitor 50 μM, substrate 50 μM). IC50 values against WNV NS2B-NS3 protease at 50 μM substrate concentration. n.d. = IC50 was not
determined for compounds with less than 98% inhibition. Percent inhibition of thrombin (inhibitor 25 μM, substrate 50 μM) did not exceed 30%
(cf. Supporting Information for details). All measurements were carried out in triplicate. Standard deviations were below 10%.
properties with a focus on hydrophobic groups to counter-
balance the high polarity of the basic side chains. Among the
benzyl substituted analogues, 109 with a polar electron-
donating amino group in the para position had the weakest
activity at DENV protease, while 38 with a 4-trifluoromethyl
moiety showed the highest inhibitory potential. Exchanging the
E
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
trifluoromethyl substituent by the structurally similar trifluor-
omethoxy group in 39 was associated with loss of potency.
Concerning the position of substitution, the ranking of activity
goes in the order para > meta > ortho. Extending the ether
methylene linker by a carbonyl group in the phenacyl derivative
50 caused a decrease of activity at DENV-2 protease, as
opposed to WNV, where it resulted in a slight improvement in
inhibitory potential. A more pronounced dissimilarity between
the two targets was seen for 34, which had the highest
inhibitory activity at WNV (IC50 = 213 nM) of this series. It
appears that the WNV binding site does not tolerate bulky
groups such as tert-butyl and naphthyl, with the corresponding
analogues (48 and 49, respectively) being relatively active at the
DENV-2 but not on the WNV protease. For the purpose of
dual inhibition, the 4-CF3-benzyl ether 38 (IC50 = 69 nM at
DENV-2 and 224 nM at WNV) was identified as the most
favorable sequence with a 5-fold and 3-fold improvement of
potency at DENV-2 and WNV, respectively, in comparison to
the unsubstituted parent compound 27.
Optimization of the N-Terminal Cap. The fine-tuning of
the activity at the C-terminal residue was accompanied by a
search for an optimal N-terminal cap. This was accomplished
by analyzing the inhibitory effect of a diverse collection of small
to intermediately sized groups appended to our standard
sequence 27 instead of the benzoyl moiety (Table 3). Size
restriction in this region was an essential criterion to match the
established requirement for activity at the C-terminus,
represented by the substituted hydroxyphenylglycine ether
scaffold. This would allow for the activity optimization to be
less affected by insufficient solubility of the developed
inhibitors, a problem that became evident while attempting to
incorporate bulkier scaffolds as caps, which moreover caused a
decrease of target affinity.
WNV protease, with nearly no improvement at DENV-2
protease.
Fragment Merging. Proteases as targets for drug discovery
often encompass distinct subpockets engaging in specific
interactions with the side chains of a substrate or a ligand.
This makes them attractive examples for the application of
fragment-based techniques.41,42 Among the various strategies of
this approach, our interest has been directed at fragment-
growing and fragment-merging.43,44 This was first realized by
extending the hydroxyphenylglycine sequence 25 to the ethers
shown in Table 2. In fragment-merging, we intended to use the
activity-optimized fragments with functionalities fitting at
different binding positions of the protease targets to generate
a series of high affinity inhibitors. For this purpose, selected
tripeptides from Tables 2 and 3 were merged, thus
incorporating optimal P1 and cap substituents into a single
peptidic scaffold. Fragment merging was not limited to the
most active sequences. It was also applied to fragments that
generated a satisfactory enhancement of in vitro potency, while
allowing to modulate the lipophilic character of the merged
inhibitors, and consequently their membrane permeability
needed for higher antiviral effect in cellular assays. In
accordance with the concept of the merging strategy, all
generated peptide hybrids (Table 4) displayed a higher
inhibitory potential than their parent fragments.
The most potent inhibitor 83 was obtained by merging the
fragments 38 (4-CF3-benzyl ether) and 56 (thiazole cap),
showing the highest enhancement in inhibitory potency at both
DENV and WNV proteases for optimization at the C-terminus
and N-terminus, respectively. Indeed, 83 (IC50 = 18 nM, Ki =
12 nM at dengue, IC50 = 50 nM, Ki = 39 nM at WNV) is
characterized by a noticeable improvement in binding affinity in
comparison to 27 (IC50 = 367 nM, Ki = 307 nM at dengue, IC50
= 728 nM, Ki = 536 nM at WNV) by a factor of 20 at DENV-2
and 14 at WNV. The tripeptide hybrid 83 represents thus the
We started our search with bioisosteric replacements of the
phenyl ring. For DENV-2, improvement of the inhibition could
be seen with pyridine (51), thiophene (53 and 54), and
thiazole (55−57), with preferred linkage at position two (53
and 56). Cycloalkyl groups generally resulted in lower
inhibition; progressive decrease of the ring size (58−60)
minimally enhanced the activity, with the exception of the
cyclopropyl analogue (61). With regard to phenyl ring
substitution, the meta-formyl derivative (69) showed the
highest potency. Introducing an unsaturated ethylene linker
in the trans-cinnamic acid analogue (72) maintained the
activity, whereas the saturated analogue (74) was characterized
by a nearly 3-fold loss of activity. Thiophene-based annelated
rings such as thienothiophenes (75−76) or benzothiophene
(77) were less active than thiophene by a factor of 4 or 5. Ring
substitution at the thiophene ring (79−82) generated activity
in the same range as the benzoyl substituted reference 27. For
both DENV-2 and WNV proteases, the thiazole ring present in
56 (IC50 = 99 nM at DENV-2 and 158 nM at WNV) and the
thiophene scaffold in 53 (IC50 = 109 nM at DENV-2 and 296
nM at WNV) were the most promising caps for dual inhibition
purposes. As already observed for the C-terminal ethers, some
discrepancy in the inhibition results could be seen between the
two targets. For WNV protease, this was particularly the case
for the furan, cyclobutane, and 4-chloro-trans-cinnamic acid
analogues (52, 60, and 73 respectively), which generated a
more pronounced inhibition as opposed to their action on
DENV-2 protease. Again, a twice higher potency was observed
for the thiophene-substituted derivatives 79, and 81−82 at
most potent dual inhibitor in this series, followed by 86 (IC50
=
28 nM, Ki = 19 nM at DENV-2, IC50 = 117 nM, Ki = 93 nM at
WNV) obtained by combining 38 (4-CF3-benzyl ether) with
53 (thien-2-yl cap).
Merged molecules inherited the established ranking of the
activity and target selectivity linked to their composing entities.
All composites maintained an excellent selectivity against
thrombin. Generally, compounds showed less inhibition at
WNV; this was particularly noticed for sequences with t-butyl
ethers (93, and 95−96) or annelated caps (96, 98−100, and
102−103). To confirm this, we determined the inhibitory
potential at WNV for the peptide hybrid combining both
features (96), which, as expected, showed a selectivity factor of
100 between the two proteases. The finding sheds light at
differences between both active sites.
Sequences 104−105 and 96 were fluorescent owing to their
bithiophene and thienothiophene N-capping moieties as was
seen also for peptide 79 (bithiophene cap) and bioisosteric 82
(phenylthiophene cap). Fluorescent compounds were tested in
HPLC-based dengue and WNV protease assays to avoid
interference with the fluorimetric assays.
Altogether, fragment growing and merging optimization led
to a class of tripeptide hybrids which are highly potent dual
inhibitors of DENV and WNV proteases. Members of this class
have binding affinities in the low nanomolar range without
incorporation of highly reactive electrophiles. The most active
derivative 83 noticeably outperforms the initial lead I (IC50
3.32 μM, Ki = 2.12 μM at DENV-2, IC50 = 58.10 μM, Ki =
=
F
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 4. Inhibitory Activity of Merged Sequences against
DENV and WNV Proteases
Table 4. continued
c
protease (inhibitor 50 μM, substrate 50 μM). IC50 values against
WNV NS2B-NS3 protease at 50 μM substrate concentration. n.d. =
IC50 was not determined for compounds with less than 98% inhibition
except 96. All measurements were carried out in triplicate. Standard
deviations were below 10%.
48.70 μM at WNV) showing 180-fold higher inhibition at
dengue and more than a 1000-fold higher activity at WNV.
Selectivity and Binding Mode. To study the potential of
the compounds as thrombin inhibitors and determine their
selectivity, a screen (25 μM) against thrombin was performed,
but no significant inhibition was observed for any compound. A
screen of the most promising compounds (Table 4) at 25 μM
against trypsin also showed no significant inhibition of this off-
target, with the sole exception of 91 (59% inhibition).
Complete thrombin and trypsin inhibition data are provided
in the Supporting Information.
Tryptophan quenching assays and competition with
aprotinin45 were used to study the binding mode of selected
compounds (104 and 83) at the dengue and West Nile virus
proteases. The results clearly indicate a binding of the
compounds in the active site of the enzymes. Details are
given in the Supporting Information. A Cheng−Prussof plot46
for compound 83 at the DENV-2 and WNV proteases revealed
a competitive inhibition mechanism (Figure S2).
Docking Studies. Docking studies were performed with
AutoDock Vina47 using the X-ray structures of DENV-3 NS2B-
NS3 (3U1I)48 and WNV NS2B-NS3 (2FP7).10 With the lack
of a crystal structure for DENV-2 protease in complex with an
inhibitor and considering the high degree of conservation of the
binding site between serotypes 2 and 3, the 3U1I structure was
considered adequate for docking simulations at DENV.
Attempts to dock compounds representing a gradient of
activities at the target proteases provided no significant
differences in the binding mode. As previously reported,33
this could be attributed to the high affinity of the present data
set, which makes docking simulations unable to determine fine
differences in binding that lead to the observed variation of
affinity. Considering this aspect and the similarity of our
compounds, the binding mode analysis was focused on the
most active analogue, compound 83 (Figure 1). In general, the
ligand assumed a nearly cyclic orientation in the binding site of
either protease. At both DENV and WNV, the obtained
binding poses came in agreement with previously reported
docking studies from our group,31,32 with the phenylglycine
ring placed in the S1 pocket, the arginine in the S2 pocket, and
the N-cap in the S3 pocket with potential extension to the S4
pocket in case of larger capping moieties. While these residues
did not significantly vary in their orientation in the binding
pockets, this was not the case for the lysine side chain, which
occupies the space between S1, S3, and S4 pockets, interacting
either with the protein or the ligand. The same observation was
made before.31,32 With respect to the newly introduced benzyl
ether, this group was modeled to bind in a previously
unexplored binding area between the S1 and the S3 pockets.
The benzyl ring is placed close to residue 155 (Val155 in
DENV-3 or Ile155 in WNV), with the potential for a π−σ
interaction between the phenyl ring of the ether and the alkyl
side chain of residue 155. This amino acid is conserved in all
DENV serotypes and was previously implicated in explaining
the lower affinity of aprotinin at WNV protease in comparison
a
IC50 values against DEN NS2B-NS3 protease serotype 2 at 50 μM
b
substrate concentration. Percent inhibition of WNV NS2B-NS3
G
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 1. Docking studies of compound 83 at DENV-3 and WNV proteases. The protein surface is colored by amino acid hydrophobicity, with blue
indicating hydrophilic and orange indicating hydrophobic regions. The ligand is shown in stick representation. The positions of the specificity
pockets, the catalytic triad, and residue 155 are indicated. (a) Docking pose of the ligand−protein complex for DENV protease. (b) Docking pose of
the ligand−protein complex for WNV protease.
to DENV protease.48 The underlying reason for this difference
Table 5. Results of the Dengue Virus Titer Reduction and
is the bulkier side chain of isoleucine in WNV as opposed to
valine in DENV, which would allow aprotinin to bind deeper in
the binding site and create a H-bond interaction with Gly153.48
Analogously, this may explain the lower tolerance of the WNV
binding site for bulkier benzyl ethers or larger N-terminal caps
of the inhibitors presented here.
Viability Assays for Selected Compounds
Antiviral Activity in Cell-Culture and Cytotoxicity
Evaluation. To confirm the antiviral effects of the developed
tripeptide hybrids on DENV-2 and WNV replication, 12
compounds in addition to our previously reported reference I31
were chosen for virus titer reduction assays using the Huh-7
human hepatoma cell line infected with DENV-2, and 3
compounds together with the reference I were additionally
tested in Huh-7 cells infected with WNV.31 The selected
molecules are characterized by variant protease inhibitory
activity, overall size, and lipophilicity. Compounds were first
assessed for cytotoxicity by cell viability assays, and the half-
maximum cytotoxic concentrations (CC50) were determined
(Tables 5 and 6). The viability testing was used to define the
upper limit of concentrations for each compound in the plaque
assay to eliminate the influence of cytotoxic effects. All selected
tripeptides displayed a significant reduction of virus titer and
were thus used for EC50 determination. Antiviral activity was
obtained in a very promising range, with EC50 values
predominantly between 3 and 20 μM at DENV-2, and between
15 and 23 μM at WNV. Furthermore, the compounds displayed
favorable therapeutic indices with the CC50 to EC50 ratio
usually higher than 10. At DENV-2, the most potent
compounds 104 and 90 showed EC50 values of 3.42 μM and
4.06 μM, respectively, and a promising CC50 to EC50 ratio (SI
(selectivity index) > 25). At WNV, compound 104 displayed
the lowest EC50-value of 15.5 μM and a favorable selectivity
index (SI > 7).
In contrast to the relatively straightforward SAR discussion of
protease inhibition in vitro, analysis of the cellular antiviral
activity is complicated by additional aspects, such as membrane
permeability and metabolic stability. For DENV-2, antiviral
cellular activity of the sequences bearing small aromatic rings at
the N-terminus, such as the thiophene-capped 88 and 90, seem
to be sensitive to the lipophilic character of the benzyl ether
moiety. As an experimental indicator for lipophilicity, we here
consider the retention times (tR) of the compounds on RP-18
HPLC columns.49 For instance, the more lipophilic 3,4-
dichlorobenzyl substituted 90 (retention time tR = 3.22 min)
showed 14-fold higher anti-DENV effect in cells (EC50 = 4.06
μM) than the 3-methoxybenzyl analogue 88 (retention time tR
= 2.71 min; EC50 = 54.63 μM), despite having comparable
activities in vitro. The lower antiviral activity of the thiazole-
capped 83 (EC50 = 20.35 μM) compared to the thiophene
analogue 86 (EC50 = 7.12 μM) could be explained by
metabolism in the hepatic Huh-7 cells used for testing,
reflecting thus the higher stability of the thiophene ring against
H
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 6. Results of the West Nile Virus Titer Reduction and
Viability Assays for Selected Compounds
Figure 2. Correlation between EC50 values (μM) from the DENV-2
plaque assay and IC50 values (nM) at the DENV-2 NS3-NS2B. The
compound tR (min) is indicated in parentheses.
oxidative biotransformation in comparison to thiazole.50
Consistent with the obtained results at the isolated DENV-2
and WNV proteases in vitro, the tested compounds were less
active at WNV than at DENV-2. Lipophilicity of the inhibitors
represented by their retention time (tR) was found to correlate
with the observed antiviral activity. This could be justified by
the shorter infection cycle in the WNV assay in comparison to
the DENV-2 assay, where factors affecting membrane
permeability such as lipophilicity are expected to exert higher
influence on the activity of the compounds than metabolic
stability. Consequently, compound 104 (retention time tR =
of the compounds that displayed detectable permeability are
listed in Table 7. These include derivatives 93, 104, and 105.
Table 7. Passive Membrane Permeability, Determined by
PAMPA, of Selected Compounds
a
b
c
d
no.
CAcc (μM)
CDon (μM)
Pe (10−6 cm/s)
R (%)
93
1.8 0.3
3.4 0.3
4.1 1.3
197.9 0.1
195.4 0.3
192.3 1.9
0.33
0.65
0.79
0.44
1.15
2.5
104
105
3.28 min) displayed higher anti-WNV effect in cells (EC50
=
15.5 μM) than the slightly more polar derivatives; 83 (retention
time tR = 3.11 min; EC50-value of 23.36 μM) or 90 (retention
time tR = 3.22 min; EC50-value of 23.31 μM).
a
Concentration of the compound in the acceptor plate after 5 h of
b
incubation. Concentration of the compound in the donor plate after
c
5 h of incubation. Permeability of the compound calculated according
In an attempt to model the relationship between in vitro and
cellular activities, we generated a plot of the antiviral activity on
DENV-2 replication in cells (EC50 in μM) versus their
inhibitory activity on DENV-2 protease (IC50 in nM) for the
compounds in Table 5 (excluding I). A linear correlation can be
obtained with an adjusted R-square of 0.945 and Pearson’s r
coefficient of 0.975 (Figure 2). However, the variance of
activities is relatively small, and therefore any conclusions from
this “QSAR” remain limited. In accordance with the discussion
above, 88 and 83 were considered as outliers and were not
included in the correlation. A certain trend evident from the
plot is that inhibition in the plaque assay appears to be related
to higher lipophilicity of the inhibitors, which is here expressed
as an experimental surrogate (retention time tR on RP-
HPLC).49
Taken together, the antiviral activity of this new class of
inhibitors is a significant improvement in comparison to the
previously published I; a nearly 25-times lower EC50 at DENV-
2 could be achieved with 104. Furthermore, for the first time a
dual antiviral profile in titer reduction assays was observed, with
compound 104 showing at least 7-fold lower EC50 at WNV
than compound I.
to the literature.52 Mass retention of the compound calculated
d
according to the literature.52
The weak permeability could be linked to the high polarity of
the compounds caused by the incorporation of two basic
residues. The results of the PAMPA assay may partially explain
the difference between the IC50-values at the isolated viral
proteases and the EC50 values in viral titer reduction assays for
DENV-2 and WNV. A slight improvement in passive
permeability could be achieved with compounds 93, 104, and
105 in comparison to the analogues published previously.33
Considering the peptidic nature of our inhibitors and their
sensitivity to metabolic clearance, particularly hepatic metabo-
lism,53 the metabolic stability of three selected compounds (83,
88, and 104) was assessed using liver microsomes from rats.33
Liver microsomes contain phase 1 metabolic enzymes that
perform hydrolysis, reduction, and oxidation of the compounds
to generate polar metabolites.54 Compounds were chosen with
the intention to obtain an understanding of the higher activity
of compound 104 in cell-based assays in comparison to
analogues such as 83 and 88, which were more active in vitro at
the DENV-2 and WNV proteases. Samples were incubated for
30 min at 37 °C, and the loss of parent compound was
monitored. Testosterone was used as the reference compound
and showed a half-life time of 63 min. The half-lives of the
tested compounds were 83, 45 min; 88, 93 min; and 104, 175
min.
Membrane Permeability and Metabolic Stability. The
passive membrane permeability of all compounds tested in
DENV-2 or WNV plaque assays (Tables 5 and 6) was further
evaluated using the parallel artificial membrane permeability
assay (PAMPA).51,52 Compounds 27, 77, 83, 86, 88, 90, 97,
100, and 101 were not permeable at detectable range, as was
previously reported for the reference compound I.33 The results
I
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Biotech (Germany) and was of an average capacity of 0.65 mmol/g.
The reaction progress was determined by thin layer chromatography
on Merck silica gel plates 60 F254 (UV detection).
In general, the ether-containing inhibitors displayed a
remarkable increase in metabolic stability in comparison to
the phenylglycine-based inhibitors (half-life 18−22 min).33 The
improvement in metabolic stability could be attributed to the
difference in the C-terminal or P2 residues.
For compound 104, the improved properties of this
derivative with regard to membrane permeability and metabolic
stability could provide an explanation for its superior
performance in dengue and WNV cell-based assays, in
comparison to analogues such as compound 83 that showed
higher inhibitory activity in vitro.
NMR spectra were recorded on Varian NMR instruments at 300 or
500 MHz, 300 K in acetone-d6, CDCl3, CD3OD, DMSO-d6/NaOD, or
D2O. For the preparation of unprotected amino acid NMR samples in
DMSO-d6/NaOD, 0.1 mL of NaOD and 0.5−0.6 mL of DMSO-d6
were added to the amino acid, the mixture was centrifuged to remove
undissolved material (if any), and the clear solution was used for NMR
analysis. Chemical shifts (δ) are given in parts per million (ppm) in
reference to residuals of nondeuterated solvents as internal standard.
(1H NMR: acetone-d6 δ = 2.05 ppm, CDCl3 δ = 7.26 ppm. CD3OD: δ
= 4.87 ppm, DMSO-d6 δ = 2.50 ppm. D2O: δ = 4.79 ppm. 13C NMR
(APT): acetone-d6 δ = 29.84 ppm, CDCl3 δ = 77.16 ppm. CD3OD: δ
= 49.00 ppm, DMSO-d6 δ = 39.52 ppm.) Coupling constants (J) are
given in hertz (Hz). Multiplicity is reported as s (singlet), d (doublet),
t (triplet), dd (doublet of doublet), td (triplet of doublet), and m
(multiplet).
HR-ESI mass spectra were measured on a Bruker microTOF-Q II
instrument. All tested target compounds were obtained with a purity of
at least 95%, unless otherwise indicated. Analysis was carried out using
50 μM samples in water/acetonitrile (1:1). Purity of the peptide
hybrids was determined by HPLC on a Jasco HPLC system with a
DAD detector on an RP-18 column (ReproSil-Pur-ODS-3, Dr. Maisch
GmbH, Germany, 5 μm, 50 mm × 2 mm) using method A. Purity of
CONCLUSIONS
■
We here report a straightforward approach toward 4-
hydroxyphenylglycine ethers as novel non-natural building
blocks of compounds with enhanced inhibitory action against
the DENV-2 and WNV proteases. The benzoyl-capped benzyl
ether analogue 27 showed a promising initial improvement of
potency against both target proteases that was achieved by
fragment growth from the previously described phenylglycine
derivative I. The 27 sequence was then optimized through
simultaneous C-terminal and N-terminal modifications. On the
basis of SAR analysis, selected activity-enhancing fragments
were merged to provide a series of highly affine inhibitors. All
molecules showed negligible activity on the off-targets
thrombin and trypsin. Compounds 83 (thiazole cap, 4-CF3-
benzyl ether), 85 (thiazole cap, 2,6-di-Cl-benzyl ether), 86
(thiophene cap, 4-CF3-benzyl ether), and 89 (thiophene cap, 4-
Br-benzyl ether) displayed a dual inhibition profile at both
proteases. However, compounds such as 95 (3-Cl-thiophene
cap, 4-tert-butyl benzyl ether), 96 (thienothiophene cap, 4-tert-
butyl benzyl ether), 98 (thienothiophene cap, 4-Cl-benzyl
ether), and 102 (benzothiophene cap, 2-CF3-benzyl ether)
were more selective toward dengue protease. The findings
provide valuable insight into the molecular recognition
preferences of both targets. For the C-terminal ether, small
para-substitutents such as 4-CF3 caused effective inhibition of
both viral proteases. Bulkier groups, such as 4-tert-butyl, were
less tolerated at WNV. Small heterocycles as N-capping
moieties yielded dual inhibitors whereas annelated caps showed
decreased activity at WNV. The effect of the protease inhibitors
on viral replication was validated in cellular assays, which
showed a reduction of dengue and West Nile virus titers and
favorable selectivity indices. The derivative 104 (bithiophene
cap, 3-OCH3-benzyl ether) possessed the highest antiviral
activity against DENV-2 and WNV, and displayed an improved
permeability and metabolic stability in comparison to those of
our previously reported inhibitors. The combination of low-
nanomolar in vitro affinities and promising activity in cell
culture offers a starting point for future improvements that will
be focused on the pharmacokinetic profile of this compound
class.
DENV and WNV FRET (Forster (fluorescence) resonance energy
̈
transfer) substrates was determined by LC-MS analysis carried out
using an Agilent HPLC system with MWD detector combined with
the Bruker microTOF-Q II instrument on a RP-18 column (ReproSil-
Pur-ODS-3, Dr. Maisch GmbH, Germany, 5 μm, 50 mm × 2 mm)
using method B. The conditions used for method A were as follows:
water (0.1% TFA); eluent B, acetonitrile (0.1% TFA); injection
volume, 10 μL; flow rate, 1 mL/min; and gradient, 1% B (0.2 min),
100% B (7 min), 100% B (8 min), 1% B (8.1 min), and 1% B (10
min). UV-detection was performed at 254 nm. The conditions used
for method B were as follows: water (0.1% HCO2H); eluent B,
acetonitrile (0.1% HCO2H); injection volume, 10 μL; flow rate, 0.4
mL/min; and gradient, 5% B (1 min), 95% B (10 min), 5% B (10.1
min), and 5% B (12 min). UV-detection was performed at 214, 254,
and 280 nm. Chromatograms recorded at 254 nm were used for purity
assessment.
General Procedure for the Synthesis of (4-Hydroxy)-phenyl-
glycine Ethers. A solution of the amino acid (1 equiv) in 1 M NaOH
(1 equiv) and a solution of CuSO4·5H2O (0.67 equiv) in water (10
mL) were warmed to 50 °C under stirring. Both solutions were
combined, and the reaction mixture was stirred for 30 min at 50 °C.
After cooling on an ice−water bath, the blue precipitate of the amino
acid Cu-complex separated immediately. The precipitate was isolated,
washed with water, and dried. The Cu-complex was dissolved in
methanol (25 mL) and 1 M NaOH (1 equiv). The corresponding
benzyl or phenacyl bromide (1.1 equiv) was added, and the mixture
was stirred at room temperature overnight. The insoluble Cu-complex
of the resulting ether was collected by filtration, washed with MeOH,
and then with water to remove excess of the unreacted starting
material. Finally, 1 M HCl (2 equiv) was added to release the product
from the Cu-complex. The precipitated product was washed with
water and dried under reduced pressure.34 The crude product was
used for subsequent synthetic steps without further purification. An
alternative workup using 8-quinolinol37 was employed to obtain 21a
due to the solubility of the Cu-product complex after acidification. To
the resulting solution, 8-hydroxyquinoline (0.5 equiv) in acetone was
added, and the mixture was stirred at room temperature for 3 h. The
solution was alkalized with 1 M NaOH, and the green precipitate of
copper 8-hydroxyquinolate was removed by filtration. Acetone was
evaporated under reduced pressure, and the solution was acidified
using 1 M HCl. After lyophilization, the solid was dissolved in EtOH/
ACN (1:1) to allow the removal of the insoluble NaCl by filtration.
The solvent was evaporated to afford the amino acid product, without
further purification.
EXPERIMENTAL SECTION
■
General. The unprotected (4-hydroxy)-D-phenylglycine and (4-
hydroxy)-L-phenylglycine, the carboxylic acids used as N-terminal
capping groups, and all other chemicals used for synthesis were
obtained from Sigma-Aldrich (Germany) and Alfa Aesar, Johnson
Matthey (Germany) and were of analytical grade. All solvents were
used as obtained from the commercial sources. The protected amino
acids were purchased from Orpegen (Germany) and Carbolution
Chemicals (Germany). HATU was purchased from Carbolution
Chemicals (Germany). The Rink amide resin was purchased from Iris
J
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(4-Benzyloxy)-L-phenylglycine (3a). Compound 3a was obtained
according to the general procedure from (4-hydroxy)-L-phenylglycine
(1003 mg, 6 mmol) and benzyl bromide (0.78 mL, 6.6 mmol) as a
128.36 (d, J = 32.3 Hz), 128.08, 128.06, 127.97, 126.22 (d, J = 1.5 Hz),
125.47 (d, J = 3.8 Hz), 122.62, 113.98, 68.34, 60.24. HRMS (ESI): m/
z [M − H]− calcd for C16H13F3NO3, 324.0853; found, 324.0853.
[4-(4-Trifluoromethoxy)benzyloxy]-D-phenylglycine (10a). Com-
pound 10a was obtained according to the general procedure from (4-
hydroxy)-D-phenylglycine (502 mg, 3 mmol) and 4-(trifluoromethoxy)
benzyl bromide (0.58 mL, 3.3 mmol) as a beige solid (131 mg, 12%
yield). 1H NMR (300 MHz, DMSO-d6/NaOD) δ 7.55 (d, J = 8.6 Hz,
2H), 7.36 (d, J = 8.0 Hz, 2H), 7.26 (d, J = 8.6 Hz, 2H), 6.83 (d, J = 8.6
Hz, 2H), 5.08 (s, 2H), 3.92 (s, 1H). 13C NMR (because of the
coupling with the fluorine atom, some of the split signals in 13C NMR
(APT) of fluorinated amino acids could not be detected) (APT, 75
MHz, DMSO-d6/NaOD) δ 176.11, 156.41, 142.85, 140.07, 139.05,
137.17, 133.97, 129.56, 127.96, 121.22, 113.96, 68.38, 60.33. HRMS
(ESI): m/z [M − H]− calcd for C16H13F3NO4, 340.0802; found,
340.0801.
1
beige solid (920 mg, 52% yield). H NMR (300 MHz, DMSO-d6/
NaOD) δ 7.54−7.38 (m, 5H), 7.32 (d, J = 8.7 Hz, 2H), 7.04 (d, J = 8.8
Hz, 2H), 5.15 (s, 2H), 4.31 (s, 1H); 13C NMR (75 MHz, DMSO-d6/
NaOD) δ 179.64, 155.48, 134.95, 133.75, 127.22, 126.82, 126.55,
126.53, 113.75, 113.75, 68.76, 58.22. HRMS (ESI): m/z [M − H]−
calcd for C15H14NO3, 256.0979; found, 256.0973.
(4-Benzyloxy)-D-phenylglycine (4a). Compound 4a was obtained
according to the general procedure from (4-hydroxy)-D-phenylglycine
(1003 mg, 6 mmol) and benzyl bromide (0.78 mL, 6.6 mmol) as a
beige solid (920 mg, 52% yield). 1H NMR (300 MHz, D2O/NaOD) δ
7.45−7.29 (m, 5H), 7.26 (d, J = 8.6 Hz, 2H), 6.83 (d, J = 8.7 Hz, 2H),
5.04 (s, 2H), 3.94 (s, 1H); 13C NMR (APT, 75 MHz, DMSO-d6/
NaOD) δ 176.45, 156.54, 138.77, 137.55, 128.53, 127.84, 127.81,
127.64, 113.90, 69.22, 60.22. HRMS (ESI): m/z [M − H]− calcd for
C15H14NO3, 256.0979; found, 256.0973.
[4-(2-Bromo)benzyloxy]-D-phenylglycine (11a). Compound 11a
was obtained according to the general procedure from (4-hydroxy)-D-
phenylglycine (1003 mg, 6 mmol) and 2-bromobenzyl bromide (1800
[4-(4-Cyano)benzyloxy]-D-phenylglycine (5a). Compound 5a was
obtained according to the general procedure from (4-hydroxy)-D-
phenylglycine (502 mg, 3 mmol) and 4-bromomethyl benzonitrile
1
mg, 6.6 mmol) as a beige solid (1250 mg, 56% yield). H NMR (300
MHz, DMSO-d6/NaOD) δ 7.65 (dd, J = 7.9, 1.1 Hz, 1H), 7.54 (dd, J
= 7.6, 1.6 Hz, 1H), 7.40 (td, J = 7.5, 1.2 Hz, 1H), 7.34−7.24 (m, 3H),
6.84 (d, J = 8.7 Hz, 2H), 5.06 (s, 2H), 3.95 (s, 1H). 13C NMR (APT,
75 MHz, DMSO-d6/NaOD) δ 176.50, 156.50, 139.04, 136.36, 136.31,
132.82, 130.36, 130.26, 128.14, 128.04, 122.94, 113.95, 69.27, 60.31.
HRMS (ESI): m/z [M − H]− calcd for C15H13BrNO3, 334.0084;
found, 334.0083.
1
(0.60 mL, 3.3 mmol) as a beige solid (600 mg, 63% yield). H NMR
(300 MHz, DMSO-d6/NaOD) δ 7.83 (d, J = 8.5 Hz, 2H), 7.61 (d, J =
8.4 Hz, 2H), 7.27 (d, J = 8.6 Hz, 2H), 6.83 (d, J = 8.7 Hz, 2H), 3.95 (s,
2H), 3.74 (s, 1H). 13C NMR (APT, 75 MHz, DMSO-d6/NaOD) δ
176.62, 156.19, 143.43, 139.01, 132.54, 128.16, 127.98, 127.91, 127.05,
118.99, 117.92, 113.98, 110.47, 70.21, 60.24. HRMS (ESI): m/z [M −
H]− calcd for C16H13N2O3, 281.0932; found, 281.0939.
[4-(2-Trifluoromethyl)benzyloxy]-D-phenylglycine (12a). Com-
pound 12a was obtained according to the general procedure from
(4-hydroxy)-D-phenylglycine (2006 mg, 12 mmol) and 2-(trifluor-
omethyl) benzyl bromide (2.01 mL, 13.2 mmol) as a beige solid (850
[4-(4-Bromo)benzyloxy]-D-phenylglycine (6a). Compound 6a was
obtained according to the general procedure from (4-hydroxy)-D-
phenylglycine (502 mg, 3 mmol) and 4-bromobenzyl bromide (825
1
mg, 3.3 mmol) as a light brown solid (726 mg, 65% yield). H NMR
1
mg, 20% yield). H NMR (300 MHz, DMSO-d6/NaOD) δ 7.77 (d, J
(300 MHz, DMSO-d6/NaOD) δ 7.56 (d, J = 8.1 Hz, 3H), 7.38 (d, J =
8.2 Hz, 2H), 7.24 (d, J = 7.5 Hz, 1H), 6.94−6.82 (m, 2H), 5.15 (s,
1H), 5.05 (s, 2H). 13C NMR (APT, 75 MHz, DMSO-d6/NaOD) δ
175.69, 153.67, 137.10, 135.15, 131.48, 129.80, 120.89, 68.52, 57.64.
HRMS (ESI): m/z [M − H]− calcd for C15H13BrNO3, 334.0084;
found, 334.0063.
= 7.8 Hz, 1H), 7.73−7.65 (m, 2H), 7.59−7.52 (m, 1H), 7.27 (d, J =
8.7 Hz, 2H), 6.82 (d, J = 8.7 Hz, 2H), 3.95 (s, 2H), 3.55 (s, 1H). 13C
NMR (because of the coupling with the fluorine atom, some of the
split signals in 13C NMR (APT) of fluorinated amino acids could not
be detected) (APT, 75 MHz, DMSO-d6/NaOD) δ 176.53, 156.37,
139.10, 135.39 (d, J = 1.8 Hz), 133.03, 130.50, 128.85, 128.06, 127.00
(d, J = 30.4 Hz), 126.37, 126.35, 126.27, 126.20, 122.74, 113.90, 71.92
(d, J = 1.0 Hz), 60.28. HRMS (ESI): m/z [M − H]− calcd for
C16H13F3NO3, 324.0853; found, 324.0869.
[4-(4-Fluoro)benzyloxy]-D-phenylglycine (7a). Compound 7a was
obtained according to the general procedure from (4-hydroxy)-D-
phenylglycine (502 mg, 3 mmol) and 4-fluorobenzyl bromide (0.41
1
mL, 3.3 mmol) as a beige solid (500 mg, 53% yield). H NMR (300
[4-(2-Chloro)benzyloxy]-D-phenylglycine (13a). Compound 13a
was obtained according to the general procedure from (4-hydroxy)-D-
phenylglycine (1003 mg, 6 mmol) and 2-chlorobenzyl bromide (0.86
MHz, DMSO-d6/NaOD) δ 7.46 (dd, J = 8.6, 5.7 Hz, 2H), 7.26 (d, J =
8.6 Hz, 2H), 7.22−7.12 (m, 2H), 6.83 (d, J = 8.7 Hz, 2H), 5.02 (s,
2H), 3.94 (s, 1H). 13C NMR (because of the coupling with the
fluorine atom, some of the split signals in 13C NMR (APT) of
fluorinated amino acids could not be detected) (APT, 75 MHz,
DMSO-d6/NaOD) δ 176.49, 156.52, 138.81, 133.86, 129.96 (d, J = 8.2
Hz), 127.95 (d, J = 1.4 Hz), 115.41 (d, J = 21.3 Hz), 114.03, 68.61,
60.30. HRMS (ESI): m/z [M − H]− calcd for C15H13FNO3, 274.0885;
found, 274.0888.
[4-(4-Chloro)benzyloxy]-D-phenylglycine (8a). Compound 8a was
obtained according to the general procedure from (4-hydroxy)-D-
phenylglycine (1003 mg, 6 mmol) and 4-chlorobenzyl bromide (1356
mg, 6.6 mmol) as a light brown solid (1000 mg, 51% yield). 1H NMR
(300 MHz, DMSO-d6/NaOD) δ 7.48−7.39 (m, 4H), 7.25 (d, J = 8.5
Hz, 2H), 6.82 (d, J = 8.7 Hz, 2H), 5.05 (s, 2H), 3.90 (s, 1H). 13C
NMR (APT, 75 MHz, DMSO-d6 /NaOD) δ 171.01, 156.21, 136.63,
132.27, 129.40, 128.47, 127.80, 113.82, 68.36, 60.26. HRMS (ESI): m/
z [M − H]− calcd for C15H13ClNO3, 290.0589; found, 290.0589.
[4-(4-Trifluoromethyl)benzyloxy]-D-phenylglycine (9a). Com-
pound 9a was obtained according to the general procedure from (4-
hydroxy)-D-phenylglycine (502 mg, 3 mmol) and 4-(trifluoromethyl)
benzyl bromide (0.51 mL, 3.3 mmol) as a beige solid (320 mg, 30%
yield). 1H NMR (300 MHz, DMSO-d6/NaOD) δ 7.73 (d, J = 8.1 Hz,
2H), 7.64 (d, J = 8.0 Hz, 2H), 7.26 (d, J = 8.6 Hz, 2H), 6.84 (d, J = 8.7
Hz, 2H), 3.94 (s, 1H), 3.53 (s, 2H). 13C NMR (because of the
coupling with the fluorine atom, some of the split signals in 13C NMR
(APT) of fluorinated amino acids could not be detected) (APT, 75
MHz, DMSO-d6/NaOD) δ 176.43, 156.29, 142.44, 138.96, 129.27,
1
mL, 6.6 mmol) as a beige solid (960 mg, 49% yield). H NMR (300
MHz, DMSO-d6/NaOD) δ 7.59−7.53 (m, 1H), 7.51−7.45 (m, 1H),
7.40−7.33 (m, 2H), 7.28 (d, J = 8.6 Hz, 2H), 6.84 (d, J = 8.7 Hz, 2H),
5.10 (s, 2H), 3.95 (s, 1H). 13C NMR (APT, 75 MHz, DMSO-d6/
NaOD) δ 176.41, 156.45, 139.09, 134.77, 132.67, 130.22, 130.18,
129.93, 129.54, 127.99, 127.54, 113.89, 67.00, 60.28. HRMS (ESI): m/
z [M − H]− calcd for C15H13ClNO3, 290.0589; found, 290.0596.
[4-(3-Chloro)benzyloxy]-D-phenylglycine (14a). Compound 14a
was obtained according to the general procedure from (4-hydroxy)-D-
phenylglycine (1003 mg, 6 mmol) and 3-chlorobenzyl bromide (0.87
1
mL, 6.6 mmol) as a beige solid (880 mg, 45% yield). H NMR (300
MHz, DMSO-d6/NaOD) δ 7.47 (s, 1H), 7.41−7.32 (m, 3H), 7.26 (d,
J = 8.6 Hz, 2H), 6.83 (d, J = 8.7 Hz, 2H), 5.06 (s, 2H), 3.93 (s, 1H).
13C NMR (APT, 75 MHz, DMSO-d6/NaOD) δ 176.41, 156.38,
140.30, 138.97, 133.30, 130.60, 127.99, 127.82, 127.29, 126.22, 114.03,
68.40, 60.31. HRMS (ESI): m/z [M − H]− calcd for C15H13ClNO3,
290.0589; found, 290.0588.
[4-(2,6-Dichloro)benzyloxy]-D-phenylglycine (15a). Compound
15a was obtained according to the general procedure from (4-
hydroxy)-D-phenylglycine (1003 mg, 6 mmol) and 2,6-dichlorobenzyl
bromide (1583 mg, 6.6 mmol) as a beige solid (700 mg, 32% yield).
1H NMR (300 MHz, DMSO-d6/NaOD) δ 7.57−7.52 (m, 2H), 7.48−
7.40 (m, 1H), 7.30 (d, J = 8.6 Hz, 2H), 6.88 (d, J = 8.7 Hz, 2H), 5.17
(s, 2H), 3.95 (s, 1H). 13C NMR (APT, 75 MHz, DMSO-d6/NaOD) δ
176.83, 156.71, 139.42, 136.12, 132.07, 131.56, 128.86, 127.89, 113.65,
K
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
64.99, 60.25. HRMS (ESI): m/z [M − H]− calcd for C15H12Cl2NO3,
324.0200; found, 324.0208.
yield). 1H NMR (300 MHz, DMSO-d6/NaOD) δ 7.56 (d, J = 8.5 Hz,
2H), 7.30 (d, J = 8.5 Hz, 2H), 7.21 (d, J = 8.6 Hz, 2H), 6.89 (d, J = 8.6
Hz, 2H), 3.95 (s, 2H), 3.74 (s, 1H). 13C NMR (APT, 75 MHz,
DMSO-d6/NaOD) δ 204.39, 173.99, 157.78, 133.50, 131.38, 130.18,
129.93, 127.46, 127.25, 126.81, 117.90, 114.31, 73.22, 60.34. HRMS
(ESI): m/z [M − H]− calcd for C16H13BrNO4, 362.0033; found,
362.0022.
[4-(3,4-Dichloro)benzyloxy]-D-phenylglycine (16a). Compound
16a was obtained according to the general procedure from (4-
hydroxy)-D-phenylglycine (502 mg, 3 mmol) and 3,4-dichlorobenzyl
bromide (0.48 mL, 3.3 mmol) as a beige solid (490 mg, 45% yield). 1H
NMR (300 MHz, DMSO-d6/NaOD) δ 7.68 (d, J = 1.7 Hz, 1H), 7.63
(d, J = 8.3 Hz, 1H), 7.42 (dd, J = 8.3, 1.9 Hz, 1H), 7.26 (d, J = 8.6 Hz,
2H), 6.83 (d, J = 8.6 Hz, 2H), 5.06 (d, J = 6.1 Hz, 1H), 3.92 (s, 1H).
13C NMR (APT, 75 MHz, DMSO-d6/NaOD) δ 176.40, 156.42,
136.11, 131.15, 130.76, 129.40, 127.87, 127.85, 127.84, 113.87, 70.92,
60.23. HRMS (ESI): m/z [M − H]− calcd for C15H12Cl2NO3,
324.0200; found, 324.0207.
Nε-Boc-[4-(4-amino)benzyloxy]-D-phenylglycine (108a). Com-
pound 108a was obtained according to the general procedure from
(4-hydroxy)-D-phenylglycine (251 mg, 1.5 mmol) and 107b (the
synthesis is described in the Supporting Information) (472 mg, 1.65
1
mmol) as a yellow solid (200 mg, 33% yield). H NMR (300 MHz,
DMSO-d6/NaOD) δ 7.22 (d, J = 8.6 Hz, 2H), 7.18 (d, J = 8.5 Hz,
2H), 6.97 (d, J = 8.5 Hz, 2H), 6.79 (d, J = 8.7 Hz, 2H), 4.79 (s, 2H),
3.94 (s, 1H), 1.32 (s, 9H). 13C NMR (APT, 75 MHz, DMSO-d6/
NaOD) δ 176.60, 156.76, 138.00, 128.01, 127.62, 120.45, 113.81,
74.90, 69.74, 60.05, 28.74. HRMS (ESI): m/z [M − H]− calcd for
C20H13N2O5, 371.1612; found, 371.1614.
General Procedure for the Synthesis of Nα-Fmoc-Protected
Amino Acids. A solution of the amino acid or its hydrochloride salt
(1 equiv) and DIPEA (2−3 equiv) in 20 mL of acetonitrile/water
(1:1) is stirred at room temperature for 20 min, then Fmoc-OSu (0.9−
0.95 equiv) is added. After 30 min, most of the solids have dissolved.
The reaction progress was monitored for the disappearance of Fmoc-
OSu by TLC (solvent system: ethyl acetate). When the reaction was
completed (1.5−2h), the solution was acidified to pH 2 using 1 M
aqueous HCl, 10 mL of water were added, and the mixture was
allowed to stir for an additional hour. Finally, the resulting precipitate
was collected by filtration, washed with water, dried under reduced
pressure, and used as crude product without further purification for
solid phase peptide synthesis.38
Nα-Fmoc-(4-hydroxy)-L-phenylglycine (1). Following the general
procedure, 4-(hydroxy)-L-phenylglycine (668 mg, 4 mmol), DIPEA
(1.39 mL, 8 mmol), and Fmoc-OSu (1.21 g, 3.6 mmol) were reacted
together to give 1 as a white solid (950 mg, 68% yield). 1H NMR (300
MHz, acetone-d6) δ 7.85 (d, J = 7.5 Hz, 2H), 7.73 (d, J = 7.3 Hz, 2H),
7.40 (t, J = 7.4 Hz, 2H), 7.35−7.27 (m, 4H), 7.09 (d, J = 7.5 Hz, 1H),
6.85 (d, J = 8.6 Hz, 2H), 5.26 (d, J = 7.6 Hz, 1H), 4.38−4.28 (m, 2H),
4.26−4.19 (m, 1H). HRMS (ESI): m/z [M − H]−calcd for
C23H18NO5, 388.1190; found, 388.1185.
[4-(3-Methoxy)benzyloxy]-D-phenylglycine (17a). Compound 17a
was obtained according to the general procedure from (4-hydroxy)-D-
phenylglycine (1003 mg, 6 mmol) and 3-methoxybenzyl chloride
1
(0.92 mL, 6.6 mmol) as a light brown solid (930 mg, 48% yield). H
NMR (300 MHz, DMSO-d6/NaOD) δ 7.33−7.20 (m, 2H), 6.99 (d, J
= 7.8 Hz, 2H), 6.95−6.77 (m, 4H), 5.04 (s, 1H), 3.74 (s, 3H), 3.46 (s,
2H). 13C NMR (APT, 75 MHz, DMSO-d6/NaOD) δ 172.68, 159.42,
151.36, 139.20, 129.65, 128.11, 126.79, 126.17, 119.67, 116.50, 113.18,
113.05, 69.11, 59.71, 55.15. HRMS (ESI): m/z [M − H]− calcd for
C16H16NO4, 286.1085; found, 286.1062.
[4-(3-Methyl)benzyloxy]-D-phenylglycine (18a). Compound 18a
was obtained according to the general procedure from (4-hydroxy)-D-
phenylglycine (1504 mg, 9 mmol) and 3-methylbenzyl bromide (1.34
1
mL, 9.9 mmol) as a beige solid (1680 mg, 61% yield). H NMR (300
MHz, DMSO-d6/NaOD) δ 7.28−7.16 (m, 5H), 7.10 (d, J = 7.1 Hz,
1H), 6.82 (d, J = 8.7 Hz, 2H), 4.99 (s, 2H), 3.93 (s, 1H), 2.30 (s, 3H).
13C NMR (APT, 75 MHz, DMSO-d6/NaOD) δ 176.40, 156.68,
138.72, 137.79, 137.54, 128.53, 128.27, 127.92, 124.82, 113.97, 69.35,
60.30, 21.20. HRMS (ESI): m/z [M − H]− calcd for C16H16NO3,
270.1136; found, 270.1132.
[4-(4-tert-Butyl)benzyloxy]-D-phenylglycine (19a). Compound
19a was obtained according to the general procedure from (4-
hydroxy)-D-phenylglycine (2006 mg, 12 mmol) and 4-tert-butylbenzyl
bromide (2.65 mL, 13.2 mmol) as a beige solid (2100 mg, 50% yield).
1H NMR (300 MHz, DMSO-d6/NaOD) δ 7.38 (d, J = 8.5 Hz, 2H),
7.33 (d, J = 8.5 Hz, 2H), 7.24 (d, J = 8.6 Hz, 2H), 6.81 (d, J = 8.7 Hz,
2H), 5.00 (s, 2H), 3.93 (s, 1H), 1.26 (s, 9H). 13C NMR (APT, 75
MHz, DMSO-d6/NaOD) δ 176.52, 156.70, 150.40, 138.60, 134.61,
127.94, 127.66, 125.38, 113.97, 69.09, 60.31, 34.48, 31.35. HRMS
(ESI): m/z [M − H]− calcd for C19H22NO3, 312.1605; found,
312.1606.
Nα-Fmoc-(4-hydroxy)-D-phenylglycine (2). Following the general
procedure, 4-(hydroxy)-D-phenylglycine (668 mg, 4 mmol), DIPEA
(1.39 mL, 8 mmol), and Fmoc-OSu (1214 mg, 3.6 mmol) were
1
reacted together to give 2 as a white solid (1000 mg, 72% yield). H
NMR (300 MHz, acetone) δ 7.85 (d, J = 7.5 Hz, 2H), 7.73 (d, J = 7.4
Hz, 2H), 7.40 (t, J = 7.4 Hz, 2H), 7.36−7.24 (m, 4H), 7.10 (d, J = 7.5
Hz, 1H), 6.85 (d, J = 8.6 Hz, 2H), 6.37 (s, 1H), 5.27 (d, J = 7.7 Hz,
1H), 4.40−4.28 (m, 2H), 4.28−4.17 (m, 1H); 13C NMR (75 MHz,
acetone) δ 172.65, 158.30, 156.57, 156.55, 145.09, 144.98, 129.82,
128.50, 127.94, 126.23, 126.18, 120.77, 116.22, 67.37, 58.45, 47.96.
HRMS (ESI): m/z [M − H]−calcd for C23H18NO5, 388.1190; found,
388.1196.
Nα-Fmoc-(4-benzyloxy)-L-phenylglycine (3b). Following the gen-
eral procedure, 3a (588 mg, 2 mmol), DIPEA (1.05 mL, 6 mmol), and
Fmoc-OSu (607 mg, 1.8 mmol) were reacted together to give 3b as a
beige solid (750 mg, 87% yield). 1H NMR (300 MHz, CDCl3) δ 7.74
(d, J = 8.8 Hz, 3H), 7.57 (d, J = 7.0 Hz, 1H), 7.42−7.30 (m, 9H), 7.17
(s, 2H), 6.95 (d, J = 8.7 Hz, 2H), 5.74 (d, J = 6.6 Hz, 1H), 5.35 (d, J =
6.7 Hz, 1H), 5.05 (s, 2H), 4.37 (d, J = 6.5 Hz, 2H), 4.25−4.15 (m,
1H). HRMS (ESI): m/z [M − H]−calcd for C30H24NO5, 478.1660;
found, 478.1645.
[4-(Naphth-2-yl)methoxy]-D-phenylglycine (20a). Compound 20a
was obtained according to the general procedure from (4-hydroxy)-D-
phenylglycine (1003 mg, 6 mmol) and 2-bromomethyl naphthalene
(1460 mg, 6.6 mmol) as a light yellow solid (1400 mg, 68% yield). 1H
NMR (300 MHz, DMSO-d6/NaOD) δ 7.97−7.87 (m, 4H), 7.55 (dd,
J = 8.4, 1.7 Hz, 1H), 7.52−7.48 (m, 2H), 7.26 (d, J = 8.4 Hz, 2H), 6.88
(d, J = 8.7 Hz, 2H), 5.22 (s, 2H), 3.93 (s, 1H). 13C NMR (APT, 75
MHz, DMSO-d6/NaOD) δ 176.34, 156.67, 138.87, 135.29, 132.99,
132.69, 128.26, 127.98, 127.96, 127.81, 126.53, 126.29, 126.24, 125.84,
114.10, 69.49, 60.33. HRMS (ESI): m/z [M − H]− calcd for
C19H16NO3, 306.1136; found, 306.1131.
(4-Phenacyloxy)-D-phenylglycine (21a). Compound 21a was
obtained according to the general procedure from (4-hydroxy)-D-
phenylglycine (502 mg, 3 mmol) and 2-bromoacetophenone (657 mg,
3.3 mmol) as a brown solid (484 mg, 50% yield). 1H NMR (300 MHz,
DMSO-d6/NaOD) δ 7.93−7.78 (m, 2H), 7.36 (dd, J = 8.2, 1.2 Hz,
1H), 7.29−7.17 (m, 2H), 6.72 (d, J = 8.5 Hz, 2H), 6.04 (d, J = 8.4 Hz,
2H), 4.45 (s, 1H), 3.74 (s, 2H). 13C NMR (APT, 75 MHz, DMSO-d6/
NaOD) δ 200.37, 178.70, 169.42, 131.18, 130.03, 129.11, 128.77,
127.21, 127.19, 126.34, 126.32, 125.71, 117.84, 73.81, 60.80. HRMS
(ESI): m/z [M − H]− calcd for C16H14NO4, 284.0928; found,
284.0929.
Nα-Fmoc-(4-benzyloxy)-D-phenylglycine (4b). Following the gen-
eral procedure, 4a (441 mg, 1.5 mmol), DIPEA (0.78 mL, 4.5 mmol),
and Fmoc-OSu (455 mg, 1.35 mmol) were reacted together to give 4b
1
as a beige solid (300 mg, 99% yield). H NMR (300 MHz, CDCl3) δ
7.77 (t, J = 7.6 Hz, 3H), 7.66−7.54 (m, 3H), 7.44−7.30 (m, 9H), 6.97
(d, J = 8.3 Hz, 2H), 5.77 (d, J = 6.3 Hz, 1H), 5.34 (d, J = 6.4 Hz, 1H),
5.05 (s, 2H), 4.39 (d, J = 6.6 Hz, 1H), 4.26−4.16 (m, 1H). HRMS
(ESI): m/z [M − H]− calcd for C30H24NO5, 478.1660; found,
478.1667.
[4-(4-Bromo)phenacyloxy]-D-phenylglycine (22a). Compound
22a was obtained according to the general procedure from (4-
hydroxy)-D-phenylglycine (1003 mg, 6 mmol) and 2,4′-dibromo
acetophenone (2001 mg, 6.6 mmol) as a yellow solid (900 mg, 37%
L
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Nα-Fmoc-[4-(4-cyano)benzyloxy]-D-phenylglycine (5b). Following
the general procedure, 5a (223 mg, 0.7 mmol), DIPEA (0.37 mL, 2.1
mmol), and Fmoc-OSu (213 mg, 0.63 mmol) were reacted together to
mL, 3 mmol) and Fmoc-OSu (320 mg, 0.95 mmol) were reacted
together to give 12b as a beige solid (500 mg, 96% yield). H NMR
1
(300 MHz, CDCl3) δ 7.90 (s, 1H), 7.78−7.68 (m, 4H), 7.62−7.51 (m,
2H), 7.43 (d, J = 7.7 Hz, 1H), 7.40−7.29 (m, 3H), 7.22 (d, J = 8.5 Hz,
1H), 7.18 (d, J = 3.2 Hz, 2H), 6.93 (dd, J = 16.7, 8.3 Hz, 2H), 5.73 (d,
J = 6.9 Hz, 1H), 5.36 (d, J = 6.8 Hz, 1H), 5.27 (s, 2H), 4.43−4.36 (m,
2H), 4.25−4.16 (m, 1H). HRMS (ESI): m/z [M − H]−calcd for
C31H23F3NO5, 546.1534; found, 546.1538.
1
give 5b as a beige solid (270 mg, 85% yield). H NMR (300 MHz,
CDCl3) δ 7.74 (t, J = 7.8 Hz, 3H), 7.67 (d, J = 8.2 Hz, 2H), 7.60−7.55
(m, 1H), 7.53 (d, J = 8.2 Hz, 2H), 7.45−7.29 (m, 4H), 7.17 (s, 2H),
6.92 (dd, J = 13.5, 8.5 Hz, 2H), 5.75 (d, J = 6.6 Hz, 1H), 5.35 (d, J =
6.6 Hz, 1H), 5.11 (s, 2H), 4.39 (d, J = 6.2 Hz, 2H), 4.25−4.15 (m,
1H). HRMS (ESI): m/z [M − H]−calcd for C31H23N2O5, 503.1612;
found, 503.1605.
Nα-Fmoc-[4-(2-chloro)benzyloxy]-D-phenylglycine (13b). Follow-
ing the general procedure, 13a (492 mg, 1.5 mmol), DIPEA (0.78 mL,
4.5 mmol), and Fmoc-OSu (455 mg, 1.3 mmol) were reacted together
to give 13b as a light brown solid (680 mg, 98% yield). 1H NMR (300
MHz, CDCl3) δ 7.75 (d, J = 7.3 Hz, 2H), 7.63−7.48 (m, 3H), 7.44−
7.30 (m, 5H), 7.28 (d, J = 3.6 Hz, 2H), 7.22 (d, J = 8.9 Hz, 2H), 6.98
(d, J = 8.4 Hz, 2H), 5.79 (d, J = 6.6 Hz, 1H), 5.34 (d, J = 6.8 Hz, 1H),
5.16 (s, 2H), 4.46−4.33 (m, 2H), 4.25−4.16 (m, 1H). HRMS (ESI):
m/z [M − H]−calcd for C30H23ClNO5, 512.1259; found, 512.1268.
Nα-Fmoc-[4-(3-chloro)benzyloxy]-D-phenylglycine (14b). Follow-
ing the general procedure, 14a (328 mg, 1 mmol), DIPEA (0.52 mL, 3
mmol), and Fmoc-OSu (304 mg, 0.9 mmol) were reacted together to
Nα-Fmoc-[4-(4-bromo)benzyloxy]-D-phenylglycine (6b). Follow-
ing the general procedure, 6a (261 mg, 0.7 mmol), DIPEA (0.37 mL,
2.1 mmol), and Fmoc-OSu (213 mg, 0.63 mmol) were reacted
1
together to give 6b as a beige solid (340 mg, 97% yield). H NMR
(300 MHz, CDCl3) δ 7.74 (t, J = 8.1 Hz, 2H), 7.57 (d, J = 7.0 Hz,
1H), 7.54−7.46 (m, 2H), 7.43−7.27 (m, 5H), 7.17 (s, 1H), 6.92 (dd, J
= 11.7, 7.6 Hz, 2H), 5.75 (d, J = 6.6 Hz, 1H), 5.35 (d, J = 6.8 Hz, 1H),
5.00 (s, 2H), 4.39 (d, J = 5.0 Hz, 2H), 4.26−4.15 (m, 1H). HRMS
(ESI): m/z [M − H]−calcd for C30H23BrNO5, 556.0765; found,
556.0741.
1
give 14b as a beige solid (450 mg, 97% yield). H NMR (300 MHz,
Nα-Fmoc-[4-(4-fluoro)benzyloxy]-D-phenylglycine (7b). Following
the general procedure, 7a (218 mg, 0.7 mmol), DIPEA (0.37 mL, 2.1
mmol), and Fmoc-OSu (213 mg, 0.63 mmol) were reacted together to
CDCl3) δ 7.88 (s, 1H), 7.74 (t, J = 8.7 Hz, 2H), 7.57 (d, J = 7.1 Hz,
1H), 7.48−7.31 (m, 6H), 7.30 (d, J = 1.2 Hz, 2H), 7.17 (d, J = 3.2 Hz,
2H), 6.98−6.88 (m, 2H), 5.74 (d, J = 6.7 Hz, 1H), 5.35 (d, J = 6.8 Hz,
1H), 5.02 (s, 2H), 4.39 (d, J = 5.9 Hz, 2H), 4.27−4.15 (m, 1H).
HRMS (ESI): m/z [M − H]−calcd for C30H23ClNO5, 512.1270;
found, 512.1240.
1
give 7b as a beige solid (300 mg, 96% yield). H NMR (300 MHz,
CDCl3) δ 7.74 (t, J = 8.1 Hz, 2H), 7.57 (d, J = 7.0 Hz, 1H), 7.54−7.46
(m, 2H), 7.43−7.27 (m, 5H), 7.17 (s, 1H), 6.92 (dd, J = 11.7, 7.6 Hz,
2H), 5.75 (d, J = 6.6 Hz, 1H), 5.35 (d, J = 6.8 Hz, 1H), 5.00 (s, 2H),
4.39 (d, J = 5.0 Hz, 2H), 4.26−4.15 (m, 1H). HRMS (ESI): m/z [M −
H]−calcd for C30H23FNO5, 496.1566; found, 496.1571.
Nα-Fmoc-[4-(2,6-dichloro)benzyloxy]-D-phenylglycine (15b). Fol-
lowing the general procedure, 15a (363 mg, 1 mmol), DIPEA (0.52
mL, 3 mmol), and Fmoc-OSu (321 mg, 0.95 mmol) were reacted
Nα-Fmoc-[4-(4-chloro)benzyloxy]-D-phenylglycine (8b). Following
the general procedure, 8a (656 mg, 2 mmol), DIPEA (1.05 mL, 6
mmol), and Fmoc-OSu (607 mg, 1.8 mmol) were reacted together to
1
together to give 15b as a beige solid (500 mg, 96% yield). H NMR
(300 MHz, CDCl3) δ 7.80−7.68 (m, 2H), 7.58 (d, J = 7.1 Hz, 1H),
7.44−7.39 (m, 1H), 7.37 (dd, J = 8.2, 1.6 Hz, 2H), 7.35−7.27 (m,
4H), 7.24 (dd, J = 8.1, 1.8 Hz, 1H), 7.20−7.11 (m, 2H), 7.08−6.95
(m, 2H), 5.75 (d, J = 6.7 Hz, 1H), 5.38 (d, J = 6.6 Hz, 1H), 5.27 (s,
2H), 4.42 (d, J = 6.9 Hz, 2H), 4.27−4.18 (m, 1H). HRMS (ESI): m/z
[M − H]−calcd for C30H22Cl2NO5, 546.0881; found, 546.0868.
Nα-Fmoc-[4-(3,4-dichloro)benzyloxy]-D-phenylglycine (16b). Fol-
lowing the general procedure, 16a (254 mg, 0.7 mmol), DIPEA (0.37
mL, 2.1 mmol), and Fmoc-OSu (224 mg, 0.67 mmol) were reacted
1
give 8b as a beige solid (596 mg, 64% yield). H NMR (300 MHz,
CDCl3) δ 7.74 (t, J = 8.2 Hz, 2H), 7.57 (d, J = 6.6 Hz, 1H), 7.44−7.28
(m, 9H), 7.17 (s, 2H), 6.98−6.87 (m, 2H), 5.75 (d, J = 6.4 Hz, 1H),
5.35 (d, J = 6.7 Hz, 1H), 5.01 (s, 2H), 4.39 (d, J = 5.7 Hz, 2H), 4.26−
4.13 (m, 1H). HRMS (ESI): m/z [M − H]−calcd for C30H23ClNO5,
512.1270; found, 512.1248.
Nα-Fmoc-[4-(4-trifluoromethyl)benzyloxy]-D-phenylglycine (9b).
Following the general procedure, 9a (181 mg, 0.5 mmol), DIPEA
(0.26 mL, 1.5 mmol), and Fmoc-OSu (160 mg, 0.48 mmol) were
1
together to give 16b as a light brown solid (360 mg, 98% yield). H
1
NMR (300 MHz, CDCl3) δ 7.74 (t, J = 7.8 Hz, 2H), 7.62−7.51 (m,
2H), 7.45 (d, J = 8.2 Hz, 1H), 7.43−7.27 (m, 5H), 7.23 (s, 1H), 7.17
(s, 2H), 6.91 (dd, J = 11.5, 8.6 Hz, 2H), 5.75 (d, J = 6.7 Hz, 1H), 5.35
(d, J = 6.7 Hz, 1H), 4.99 (s, 2H), 4.44−4.35 (m, 2H), 4.25−4.16 (m,
1H). HRMS (ESI): m/z [M − H]−calcd for C30H22Cl2NO5, 546.0881;
found, 546.0860.
reacted together to give 9b as a beige solid (243 mg, 93% yield). H
NMR (300 MHz, CDCl3) δ 7.74 (t, J = 8.1 Hz, 2H), 7.64 (d, J = 8.2
Hz, 2H), 7.55 (t, J = 8.7 Hz, 3H), 7.45−7.27 (m, 5H), 7.17 (s, 2H),
6.93 (dd, J = 12.3, 8.6 Hz, 2H), 5.75 (d, J = 6.5 Hz, 1H), 5.36 (d, J =
6.6 Hz, 1H), 5.11 (s, 2H), 4.39 (d, J = 5.9 Hz, 2H), 4.28−4.16 (m,
1H). HRMS (ESI): m/z [M − H]−calcd for C31H23F3NO5, 546.1534;
found, 546.1503.
Nα-Fmoc-[4-(3-methoxy)benzyloxy]-D-phenylglycine (17b). Fol-
lowing the general procedure, 17a (486 mg, 1.5 mmol), DIPEA (0.78
mL, 4.5 mmol), and Fmoc-OSu (455 mg, 1.35 mmol) were reacted
Nα-Fmoc-[4-(4-trifluoromethoxy)benzyloxy]-D-phenylglycine
(10b). Following the general procedure, 10a (110 mg, 0.29 mmol),
DIPEA (0.15 mL, 0.87 mmol), and Fmoc-OSu (93 mg, 0.28 mmol)
were reacted together to give 10b as a brown solid (135 mg, 86%
1
together to give 17b as a light brown solid (650 mg, 95% yield). H
NMR (300 MHz, CDCl3) δ 7.79−7.69 (m, 2H), 7.57 (d, J = 7.2 Hz,
1H), 7.45−7.26 (m, 6H), 7.18 (s, 2H), 7.04−6.91 (m, 4H), 6.87 (dd, J
= 7.8, 2.2 Hz, 1H), 5.72 (d, J = 6.5 Hz, 1H), 5.35 (d, J = 6.8 Hz, 1H),
5.03 (s, 2H), 4.37 (d, J = 14.2 Hz, 2H), 4.25−4.15 (m, 1H), 3.81 (s,
3H). HRMS (ESI): m/z [M − H]−calcd for C31H26NO6, 508,1766;
found, 508,1752.
1
yield). H NMR (300 MHz, CDCl3) δ 7.75 (d, J = 7.2 Hz, 2H), 7.62
(d, J = 7.7 Hz, 1H), 7.57 (d, J = 7.5 Hz, 1H), 7.45 (d, J = 8.6 Hz, 2H),
7.41−7.28 (m, 6H), 7.24 (d, J = 8.5 Hz, 2H), 6.96 (d, J = 8.4 Hz, 2H),
5.75 (d, J = 6.6 Hz, 1H), 5.35 (d, J = 6.8 Hz, 1H), 5.04 (s, 2H), 4.44−
4.36 (m, 2H), 4.25−4.17 (m, 1H). HRMS (ESI): m/z [M − H]−calcd
for C31H23F3NO6, 562.1483; found, 562.1456.
Nα-Fmoc-[4-(2-bromo)benzyloxy]-D-phenylglycine (11b). Follow-
ing the general procedure, 11a (559 mg, 1.5 mmol), DIPEA (0.78 mL,
4.5 mmol), and Fmoc-OSu (455 g, 1.35 mmol) were reacted together
to give 11b as a beige solid (450 mg, 60% yield). 1H NMR (300 MHz,
CDCl3) δ 7.74 (t, J = 8.3 Hz, 2H), 7.59 (dd, J = 7.9, 1.1 Hz, 2H), 7.52
(d, J = 7.8 Hz, 2H), 7.43−7.28 (m, 6H), 7.20 (dd, J = 7.9, 1.3 Hz, 2H),
6.95 (dd, J = 14.2, 8.3 Hz, 2H), 5.73 (d, J = 6.6 Hz, 1H), 5.36 (d, J =
6.8 Hz, 1H), 5.13 (s, 2H), 4.41 (d, J = 6.8 Hz, 2H), 4.26−4.16 (m,
1H). HRMS (ESI): m/z [M − H]−calcd for C30H23BrNO5, 556.0765;
found, 555.0744.
Nα-Fmoc-[4-(3-methyl)benzyloxy]-D-phenylglycine (18b). Follow-
ing the general procedure, 18a (616 mg, 2 mmol), DIPEA (1.05 mL, 6
mmol), and Fmoc-OSu (607 mg, 1.8 mmol) were reacted together to
1
give 18b as a beige solid (860 mg, 97% yield). H NMR (300 MHz,
CDCl3) δ 7.84 (s, 1H), 7.74 (t, J = 8.7 Hz, 2H), 7.57 (d, J = 7.1 Hz,
1H), 7.47−7.31 (m, 4H), 7.29 (d, J = 7.4 Hz, 2H), 7.24 (d, J = 3.4 Hz,
2H), 7.14 (d, J = 7.4 Hz, 2H), 7.02−6.90 (m, 2H), 5.73 (d, J = 6.8 Hz,
1H), 5.35 (d, J = 6.8 Hz, 1H), 5.01 (s, 2H), 4.46−4.33 (m, 2H), 4.25−
4.16 (m, 1H), 2.37 (s, 3H). HRMS (ESI): m/z [M − H]−calcd for
C31H26NO5, 492.1816; found, 492.1804.
Nα-Fmoc-[4-(4-tert-butyl)benzyloxy]-D-phenylglycine (19b). Fol-
lowing the general procedure, 19a (670 mg, 2 mmol), DIPEA (1.05
mL, 6 mmol), and Fmoc-OSu (1.21 g, 1.9 mmol) were reacted
Nα-Fmoc-[4-(2-trifluoromethyl)benzyloxy]-D-phenylglycine (12b).
Following the general procedure, 12a (362 mg, 1 mmol), DIPEA (0.52
M
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
1
together to give 19b as a beige solid (980 mg, 96% yield). H NMR
(300 MHz, CDCl3) δ 7.79−7.70 (m, 2H), 7.58 (d, J = 7.0 Hz, 1H),
7.42 (d, J = 8.4 Hz, 2H), 7.39−7.27 (m, 6H), 7.23 (s, 1H), 7.17 (s,
2H), 7.00−6.89 (m, 2H), 5.77 (d, J = 6.7 Hz, 1H), 5.35 (d, J = 6.8 Hz,
1H), 5.01 (s, 2H), 4.41 (d, J = 6.8 Hz, 2H), 4.26−4.17 (m, 1H), 1.33
(s, 9H). HRMS (ESI): m/z [M − H]−calcd for C34H32NO5, 534.2286;
found, 534.2262.
Nα-Fmoc-[4-(naphth-2-yl)methoxy]-D-phenylglycine (20b). Fol-
lowing the general procedure with the addition of 10 mL of DMSO
to improve the starting material solubility, 20a (138 mg, 0.4 mmol),
DIPEA (0.21 mL, 1.2 mmol), and Fmoc-OSu (130 mg, 0.39 mmol)
were reacted together to give 20b as a yellow solid (194 mg, 95%
yield). 1H NMR (300 MHz, CDCl3) δ 7.89−7.82 (m, 4H), 7.75 (d, J =
7.6 Hz, 2H), 7.60−7.53 (m, 2H), 7.49 (dd, J = 8.1, 4.4 Hz, 3H), 7.43−
7.28 (m, 6H), 7.01 (d, J = 8.1 Hz, 2H), 5.78 (d, J = 6.6 Hz, 1H), 5.34
(d, J = 6.8 Hz, 1H), 5.22 (s, 2H), 4.42−4.32 (m, 2H), 4.24−4.17 (m,
1H). HRMS (ESI): m/z [M − H]−calcd for C34H26NO5, 528.1816;
found, 528.1796.
coupling of the N-terminal cap to cleave any formed esters at the
unprotected hydroxyl group.
The resin loaded with the finished peptide or peptide hybrid was
washed 3× with diethyl ether and dried under reduced pressure. The
final product was cleaved off the resin with TFA/triisopropylsilane/
water solution (95:1:4, 1−2 mL per 100 mg resin), and the mixture
was shaken for at least 2 h. The cleavage solution was dispensed into
cold diethyl ether (20 mL per 100 mg of resin), and the resulting
precipitate was centrifuged (4000g, 10 min), washed with diethyl
ether, and dried under reduced pressure. All crude peptides were
̈
purified by preparative HPLC on an AKTA Purifier, GE Healthcare
(Germany), with an RP-18 pre and main column (Rephospher, Dr.
Maisch GmbH, Germany, C18-DE, 5 μm, 30 mm × 16 mm and 120
mm × 16 mm). The following conditions were used: eluent A, water
(0.1% TFA); eluent B, methanol (0.1% TFA); flow rate, 8 mL/min;
and gradient, 10% B (2.5 min), 100% B (23.5 min), 100% B (26 min),
10% B (26.1 min), and 10% B (30 min). Detection was performed at
214, 254, and 280 nm. After purification, methanol was evaporated,
and the peptides were freeze-dried in water and stored at −20 °C.
Expression and Purification of DENV and WNV Proteases.
DENV (serotype 2) and WNV NS2B-NS3 protease constructs
possessing a glycine-serine GGGGSGGGG linker, which covalently
connects the NS3 protease and NS2B cofactor domains, were
used.55,56 Both viral proteases were expressed and purified following
the protocol described by Steuer et al.57,58
FRET Substrates Synthesis. Dengue and WNV protease FRET
substrates, with the sequences 2-Abz-Nle-Lys-Arg-Arg-Ser-(3-NO2)-
Tyr-NH2 for DENV (Km 105 μM) and 2-Abz-Gly-Leu-Lys-Arg-Gly-
Gly-(3-NO3)-Tyr-NH2 for WNV (Km 212 μM), were synthesized by
solid phase peptide synthesis using Fmoc-strategy, as described for the
dengue FRET substrate.39 Purity was assessed by LC-MS using
method B.
DENV and WNV Protease Assays. DENV and WNV protease
assays were performed as previously described.30,39 The continuous
enzymatic assays were performed in black 96 well V-bottom plates
(Greiner Bio-One, Germany) on a BMG Labtech Fluostar OPTIMA
Microtiter fluorescence plate reader using excitation and emission
wavelengths of 320 and 405 nm, respectively. The inhibitors (final
concentration 50 μM, from 10 mM stock solutions in DMSO) were
preincubated for 15 min with the DENV protease (100 nM) or WNV
protease (150 nM) in the assay buffer (50 mM Tris-HCl pH 9,
ethylene glycol (10% v/v), and 0.0016% Brij 58). Subsequently, the
reaction was initiated by the addition of the FRET substrate (final
concentration 50 μM) to obtain a final assay volume 100 μL per well.
The enzymatic activity was determined as slope per second (relative
fluorescence units per second) and monitored for 15 min. Percentage
inhibition was calculated relative to a positive control (without the
inhibitor). All experiments were performed in triplicate and averaged.
HPLC-Based DENV and WNV Protease Assays. After perform-
ing the fluorimetric assay (15 min), the enzymatic reaction was
stopped by adding 10 μL of TFA (4%) to each well, and the samples
were cooled to 4 °C for 30 min. The HPLC analysis was carried out on
a Jasco HPLC system equipped with a RP-18 column Phenomenex
Luna C18(2) (5 μm, 150 × 3 mm) and a FP-2020 plus fluorescence
detector (excitation, 320 nm; emission, 405 nm). The following
conditions were used for DENV assay as previously described:30,39
flow rate, 1.0 mL/min; eluent A, water (0.1% TFA); eluent B,
acetonitrile (0.1% TFA); and gradient, 10% B (0 min), 10% B (1 min),
95% B (9.5 min), 95% B (9.6 min), 10% B (12.6 min), and 10% B (15
min). Different conditions were used for the WNV assay: flow rate, 1.2
mL/min; eluent A, water (0.1% TFA); eluent B, acetonitrile (0.1%
TFA); and gradient, 10% B (0 min), 20% B (1 min), 95% B (5 min),
95% B (6 min), 10% B (6.1 min), and 10% B (8 min). Percentage
inhibition was calculated relative to a positive control (without the
inhibitor). All experiments were performed in triplicate and averaged.
IC50 and Ki Determination. For determination of the half maximal
inhibitory concentrations (IC50), according to the compound percent
inhibition at 50 μM, eight inhibitor concentrations covering the range
0−1, 0−2, 0−6, or 0−50 μM were chosen for analysis. Determinations
were carried out in triplicate, and IC50-values were determined for a
Nα-Fmoc-4-(phenacyloxy)-D-phenylglycine (21b). Following the
general procedure, 21a (225 mg, 0.7 mmol), DIPEA (0.37 mL, 2.1
mmol), and Fmoc-OSu (213 mg, 0.63 mmol) were reacted together to
1
give 21b as a brown solid (220 mg, 69% yield). H NMR (300 MHz,
CDCl3) δ 8.06−7.96 (m, 2H), 7.75 (d, J = 7.3 Hz, 2H), 7.67−7.48 (m,
5H), 7.42−7.32 (m, 4H), 6.93 (d, J = 8.5 Hz, 2H), 6.81 (d, J = 8.3 Hz,
2H), 5.74 (d, J = 6.6 Hz, 1H), 5.34 (s, 1H), 5.29 (d, J = 6.3 Hz, 2H),
4.43−4.37 (m, 2H), 4.24−4.17 (m, 1H). HRMS (ESI): m/z [M −
H]−calcd for C31H24NO6, 506.1609; found, 506.1587.
Nα-Fmoc-[4-(4-bromo)phenacyloxy]-D-phenylglycine (22b). Fol-
lowing the general procedure, 22a (361 mg, 0.9 mmol), DIPEA (0.47
mL, 3 mmol), and Fmoc-OSu (273 mg, 0.81 mmol) were reacted
1
together to give 22b as a yellow solid (390 mg, 82% yield). H NMR
(300 MHz, CDCl3) δ 7.86 (dd, J = 8.6, 2.0 Hz, 2H), 7.75 (d, J = 7.5
Hz, 2H), 7.64 (dd, J = 8.4, 2.2 Hz, 2H), 7.57 (d, J = 6.2 Hz, 2H),
7.40−7.30 (m, 6H), 6.91 (d, J = 7.6 Hz, 2H), 5.75 (d, J = 4.9 Hz, 1H),
5.35 (d, J = 5.1 Hz, 1H), 5.20 (s, 2H), 4.40 (d, J = 6.5 Hz, 2H), 4.24−
4.18 (m, 1H). HRMS (ESI): m/z [M − H]−calcd for C31H23BrNO6,
584.0714; found, 584.0689.
Nα-Fmoc-Nε-Boc-[4-(4-amino)benzyloxy]-D-phenylglycine (108b).
Following the general procedure, 108a (82 mg, 0.2 mmol), DIPEA
(0.11 mL, 0.6 mmol), and Fmoc-OSu (64 mg, 0.19 mmol) were
reacted together to give 108b as a yellow solid (95 mg, 84% yield). 1H
NMR (300 MHz, CDCl3) δ 7.75 (d, J = 7.5 Hz, 2H), 7.58 (d, J = 7.2
Hz, 2H), 7.43−7.28 (m, 11H), 6.95 (d, J = 8.2 Hz, 2H), 6.53 (s, 1H),
5.35 (d, J = 6.4 Hz, 1H), 5.00 (s, 2H), 4.40 (d, J = 6.6 Hz, 2H), 4.24−
4.18 (m, 1H), 1.52 (s, 9H). HRMS (ESI): m/z [M − H]−calcd for
C35H33N2O7, 593.2293; found, 593.2294.
General Procedure for the Synthesis of Peptide Hybrids. All
peptide and peptide hybrid sequences were assembled by stepwise
solid-phase synthesis on Rink amide resin using the standard Fmoc-
strategy, as previously described.29−31,39 Solid phase synthesis was
done manually in plastic syringes equipped with a frit; all steps were
performed at room temperature under continuous shaking. The Rink
amide resin (loading capacity 0.64 mmol/g) was preswollen in DCM
for at least 30 min and then washed 3× with DMF. For Fmoc group
deprotection, a piperidine solution (25% in DMF) was added 2× for
10 and 5 min. Following each deprotection or coupling step, the resin
was washed 3× with DMF, 3× with DCM, and again 3× with DMF.
HATU/DIPEA were used for coupling steps. In detail, the coupling
solution contained the Nα-Fmoc-protected amino acid or N-terminal
carboxylic acid capping group (3 equiv), HATU (3 equiv), and DIPEA
(4 equiv) in DMF (∼1.0 mL per 100 mg of resin). The solution was
added to the resin, and the reaction was shaken for 45−90 min.
Afterward, the resin was washed as described before. Fmoc
deprotection and coupling steps were iteratively repeated until the
desired sequence was obtained. For the synthesis of peptide hybrids
sharing an identical amino acid sequence but differing in the N-
terminal cap, the Fmoc-protected peptide sequence was synthesized
on a large scale, split, and stored at 4 °C until use. For the sequences
containing 4-(OH)-phenylglycine (24−25), a final treatment with
piperidine solution (25% in DMF) for 30 min was carried out after
N
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
substrate concentration of 50 μM using Prism 5.0 (Graphpad
Software, Inc.). For DENV, an enzyme concentration of 30 nM was
used for IC50 determination, except for 83 and 86, where an enzyme
concentration of 15 nM was used. For WNV, IC50-values were
determined at an enzyme concentration of 75 nM, except for 83 where
an enzyme concentration of 40 nM was used. For the calculation of the
dissociation constant of the enzyme−inhibitor complex (Ki-value),
IC50-values at different substrate concentrations (50, 100, 150, and 200
μM) were plotted against their corresponding substrate concentrations
(Cheng−Prusoff equation),46,59 and linear regression (considering the
standard deviations) was performed using OriginPro 8.5 (OriginLab
Corp.).
mg/mL) were supplemented with NADPH (10 mM) in DPBS
(Dulbelcco’s phosphate buffered saline (Life Technologies, Ger-
many)) and preincubated at 37 °C for 15 min. The analyzed
compounds (100 μM) were added, and the samples were incubated at
37 °C for 30 min. Aliquots were removed at various time points (0, 2,
5, 10, 20, and 30 min). The reaction was terminated by the addition of
an equal volume of acetonitrile, and the samples were cooled on ice for
15 min before centrifugation (4,500g at 4 °C for 15 min). The
supernatants were used for further analysis. The loss of parent
compound was monitored by HPLC on a Jasco HPLC system with a
UV detector and RP-18 column (ReproSil-Pur-ODS, Dr. Maisch
GmbH, Germany, 3 μm, 50 × 2 mm) using the following method:
eluent A, water (+0.1% TFA); eluent B, acetonitrile (+0.1% TFA);
flow rate, 1 mL/min; and gradient, 1% B (0.2 min), 100% B (3.5 min),
100% B (4.5 min), 1% B (4.6 min), and 1% B (5 min). The metabolic
stability was determined by dividing the peak areas of the unaltered
parent compound in the metabolized sample by the peak areas of the
parent compound in the reference sample. The activity of the
microsomal preparations was verified by a positive control
(testosterone).
Cell Viability Assay. Huh-7 cells (104) per well were seeded into
96 well plates in 50 μL of DMEM supplemented with 10% FBS and
incubated overnight at 37 °C. Following this, the cells were infected
with WT DENV serotype 2 with an MOI of 1 or with WT WNV with
an MOI of 0.1 in the presence of the respective concentration of the
tested compound. Each concentration was assayed in triplicate. After
incubation for 48 h at 37 °C, the medium was harvested; the triplicates
were pooled and stored at −80 °C. Then, 50 μL of fresh DMEM was
added to the cells, and cell viability was determined using Cell-Titer
Glo Luminescent Viability Assay. A gradient of nontoxic concen-
trations was used for the determination of the virus yield reduction by
the plaque assay using Vero E6 cells.
Docking Studies. Autodock Vina 1.1.2 (AD Vina)47 was used to
perform the docking studies and investigate the binding mode of the
most active compound (83). All calculations were performed on an
Intel(R) Core(TM) i5-2450 M CPU at 2.50 GHz. The ligand was
initially prepared using Chem3D Pro (PerkinElmer). For DENV, the
crystal structure (3U1I)48 of the NS2B-NS3 protease of serotype 3 in
complex with a tetrapeptidyl aldehyde was extracted from the dimer,
and the sulfate ions and water molecules were removed. For the WNV,
the crystal structure (2FP7)10 of the NS2B-NS3 protease in complex
with the tetrapeptidyl aldehyde was used, and water molecules were
removed. Docking preparation was performed with Autodock Tools.60
Using default settings, polar hydrogen atoms were added, nonpolar
hydrogen atoms were merged, and Kollman charges were assigned to
the protein or Gasteiger charges to the ligand. AD Vina was used with
docking parameters at default values. The grid spacing was 1.0, and the
size of the docking grid was 25 Å for dengue and WNV. Visualization
was performed with UCSF Chimera.61
Virus Yield Reduction Assay. Vero E6 cells were seeded in 24
well plates with a density of 2.5 × 105 cells per well in DMEM
supplemented with 10% FBS. After overnight incubation at 37 °C, the
cells were infected with the harvested virus supernatant. The virus
containing medium was diluted with DMEM ranging from 10−1 to
10−6 for DENV and from 10−2 to 10−7 for WNV before infection. After
incubation of the cells with 100 μL of the virus dilution at 37 °C with
agitation for 1 h, the medium was removed, and 1 mL of plaque
medium was added. After further incubation for 7 days for DENV or 3
days for WNV at 37 °C, the cells were fixed with 5% (v/v)
formaldehyde for 2 h, stained with crystal violet, and plaques were
counted. EC50 values were calculated with OriginPro 8.5 using a
nonlinear dose−reponse curve fit.
The correlation between EC50 values (μM) from the DENV-2 viral
titer reduction assay and IC50 values (nM) at DENV-2 protease was
performed using OriginPro 8.5 (OriginLab Corp.).
ASSOCIATED CONTENT
* Supporting Information
■
S
Parallel Artificial Membrane Permeability Assay (PAMPA).
The permeability was evaluated for 12 compounds (27, 77, 83, 86, 88,
90, 93, 97,100, 101, 104, and 105), as described before,33 using a
precoated PAMPA plate system (BD Gentest, BD Bioscience,
Germany). Phosphate buffered saline (PBS) (Sigma-Aldrich, Ger-
many) was used for all experiments. Concentration determinations
were carried out on a Jasco HPLC system with a UV-detector and an
RP-18 column (ReproSil-Pur-ODS-3, Dr. Maisch GmbH, Germany, 5
μm, 50 mm × 2 mm). The conditions for the method used were
eluent A, water (0.1% TFA); eluent B, acetonitrile (0.1% TFA); and
gradient, 1% B (0.2 min), 100% B (3.5 min), 100% B (4.5 min), 1% B
(4.6 min), and 1% B (5 min). UV-detection was performed at 254 nm.
Six-point calibration curves (10, 25, 50, 100, 150, and 200 μM) were
generated for all analyzed compounds and references (amiloride,
caffeine, and phenytoin) wih correlation coefficients (R2) being at least
0.99. The PAMPA system was warmed to room temperature for 1 h.
In the donor plate, 300 μL of the 200 μM compound solutions in PBS
was dispensed in triplicate, and in the acceptor plate, 200 μL of PBS
buffer was added to all wells. The plates were combined, and the
system was incubated at room temperature for 5 h. After incubation,
100 μL samples from each of the donor and acceptor wells are
transferred into 96 well U-bottomed polypropylene plates (Greiner
Bio-One, Germany) and quantitatively analyzed by UV absorbance on
HPLC. The concentrations were determined from the previously
generated calibration curves. Permeability (Pe) and mass retention (R)
were calculated as described in the literature.51,52
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jmed-
chem.5b01441.
Synthesis of derivatives 107a and 107b, thrombin and
trypsin screening results, binding mode analysis by
tryptophan quenching assays, kinetic studies, HRMS
and HPLC data for all synthesized peptides, character-
ization of fluorescent compounds, NMR data for selected
peptides, in addition to 1H and 13C NMR spectra (PDF)
SMILES data (CSV)
AUTHOR INFORMATION
Corresponding Author
■
*Phone: +49-6221-544875. Fax: +49-6221-546430. E-mail: c.
klein@uni-heidelberg.de.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Heiko Rudy for measuring ESI high resolution
spectra, Lena Weigel for performing LC-MS analytics, and
Natascha Stefan for technical support. M.B. appreciates
financial support from the German Academic Exchange Service.
The project was sponsored by the Deutsche Forschungsge-
meinschaft, KL-1356/3-1.
Metabolic Stability in Liver Microsomes. Metabolic stability for
3 compounds (83, 88, and 104) were performed as described before,33
using pooled liver microsomes from male Sprague−Dawley rats
(Sigma-Aldrich, Germany). In short, liver microsomal proteins (0.2
O
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(17) Shiryaev, S. A.; Ratnikov; Boris, I.; Chekanov; Alexei, V.; Sikora,
S.; Rozanov; Dmitri, V.; Godzik, A.; Wang, J.; Smith; Jeffrey, W.;
Huang, Z.; Lindberg, I.; Samuel, Melanie A.; Diamond, Michael S.;
Strongin; Alex, Y. Cleavage Targets and the D-Arginine-based
Inhibitors of the West Nile Virus NS3 Processing Proteinase. Biochem.
J. 2006, 393, 503−511.
(18) Wu, H.; Bock, S.; Snitko, M.; Berger, T.; Weidner, T.; Holloway,
S.; Kanitz, M.; Diederich, W. E.; Steuber, H.; Walter, C.; Hofmann, D.;
Weißbrich, B.; Spannaus, R.; Acosta, E. G.; Bartenschlager, R.; Engels,
B.; Schirmeister, T.; Bodem, J.; Novel Dengue. Virus NS2B/NS3
Protease Inhibitors. Antimicrob. Agents Chemother. 2014, 59, 1100−
1109.
ABBREVIATIONS USED
■
Boc, tert-butoxycarbonyl; DENV, dengue virus; DIPEA, N,N-
diisopropylethylamine; Fmoc, 9-fluorenylmethoxycarbonyl;
Fmoc-OSu, N-(9-fluorenylmethoxycarbonyloxy)succinimide;
FRET, Forster (fluorescence) resonance energy transfer;
̈
HATU, 1-[bis(dimethylamino)methylene]-1H-1,2,3-triazolo-
[4,5-b]pyridinium 3-oxid hexafluorophosphate; HCV, hepatitis
C virus; PAMPA, parallel artificial permeability assay; Phg,
phenylglycine; RP-HPLC, reverse-phase high performance
liquid chromatography; THR, thrombin; tR, retention time;
WNV, West Nile virus
(19) Liu, H.; Wu, R.; Sun, Y.; Ye, Y.; Chen, J.; Luo, X.; Shen, X.; Liu,
H. Identification of Novel Thiadiazoloacrylamide Analogues as
Inhibitors of Dengue-2 Virus NS2B/NS3 Protease. Bioorg. Med.
Chem. 2014, 22, 6344−6352.
REFERENCES
■
(1) World Health Organization. Dengue and Severe Dengue, Fact
Sheet No. 117. http://www.who.int/mediacentre/factsheets/fs117/
en/ (accessed on Jun 9, 2015).
(20) Schuller, A.; Yin, Z.; Brian Chia, C. S.; Doan, D. N. P.; Kim, H.
̈
K.; Shang, L.; Loh, T. P.; Hill, J.; Vasudevan, S. G. Tripeptide
Inhibitors of Dengue and West Nile Virus NS2B−NS3 Protease.
Antiviral Res. 2011, 92, 96−101.
(2) Bhatt, S.; Gething, P. W.; Brady, O. J.; Messina, J. P.; Farlow, A.
W.; Moyes, C. L.; Drake, J. M.; Brownstein, J. S.; Hoen, A. G.; Sankoh,
O.; Myers, M. F.; George, D. B.; Jaenisch, T.; Wint, G. R. W.;
Simmons, C. P.; Scott, T. W.; Farrar, J. J.; Hay, S. I. The Global
Distribution and Burden of Dengue. Nature 2013, 496, 504−507.
(3) Mitka, M. Dengue More Prevalent Than Previously Thought.
JAMA 2013, 309, 1882−1882.
(4) Chancey, C.; Grinev, A.; Volkova, E.; Rios, M. The Global
Ecology and Epidemiology of West Nile Virus. BioMed Res. Int. 2015,
2015, 1.
(5) Nitsche, C.; Holloway, S.; Schirmeister, T.; Klein, C. D.
Biochemistry and Medicinal Chemistry of the Dengue Virus Protease.
Chem. Rev. 2014, 114, 11348−11381.
(6) Kilpatrick, A. M. Globalization, Land Use, and the Invasion of
West Nile Virus. Science 2011, 334, 323−327.
(7) Vasilakis, N.; Ooi, M.; Rabaa, M.; Mayer, S.; Widen, S.; Perera,
D.; Hickey, A.; Travassos Da Rosa, A.; Bossart, K.; Tesh, R.; Holmes,
E.; Cardos, J. The Daemon in the Forest - Emergence of a New Dengue
Serotype in Southeast Asia, III; International Conference on Dengue
and Dengue Haemorrhagic Fever, Bangkok, Thailand, Oct 21−23,
2013.
(8) Falgout, B.; Pethel, M.; Zhang, Y. M.; Lai, C. J. Both
Nonstructural Proteins NS2B and NS3 are Required for the
Proteolytic Processing of Dengue Virus Nonstructural Proteins. J.
Virol. 1991, 65, 2467−2475.
(9) Shiryaev, S. A.; Strongin, A. Y. Structural and Functional
Parameters of the Flaviviral Protease: A Promising Antiviral Drug
Target. Future Virol. 2010, 5, 593−606.
(21) Hammamy, M. Z.; Haase, C.; Hammami, M.; Hilgenfeld, R.;
Steinmetzer, T. Development and Characterization of New
Peptidomimetic Inhibitors of the West Nile Virus NS2B−NS3
Protease. ChemMedChem 2013, 8, 231−241.
(22) Lim, S. P.; Shi, P. Y. West Nile Virus Drug Discovery. Viruses
2013, 5, 2977−3006.
(23) Lim, S. P.; Wang, Q. Y.; Noble, C. G.; Chen, Y. L.; Dong, H.;
Zou, B.; Yokokawa, F.; Nilar, S.; Smith, P.; Beer, D.; Lescar, J.; Shi, P.
Y. Ten Years of Dengue Drug Discovery: Progress and Prospects.
Antiviral Res. 2013, 100, 500−519.
(24) Yin, Z.; Patel, S. J.; Wang, W. L.; Chan, W. L.; Ranga Rao, K. R.;
Wang, G.; Ngew, X.; Patel, V.; Beer, D.; Knox, J. E.; Ma, N. L.;
Ehrhardt, C.; Lim, S. P.; Vasudevan, S. G.; Keller, T. H. Peptide
Inhibitors of Dengue Virus NS3 Protease. Part 2: SAR Study of
Tetrapeptide Aldehyde Inhibitors. Bioorg. Med. Chem. Lett. 2006, 16,
40−43.
(25) Yin, Z.; Patel, S. J.; Wang, W. L.; Wang, G.; Chan, W. L.; Rao, K.
R. R.; Alam, J.; Jeyaraj, D. A.; Ngew, X.; Patel, V.; Beer, D.; Lim, S. P.;
Vasudevan, S. G.; Keller, T. H. Peptide Inhibitors of Dengue Virus
NS3 Protease. Part 1: Warhead. Bioorg. Med. Chem. Lett. 2006, 16, 36−
39.
(26) Knox, J. E.; Ma, N. L.; Yin, Z.; Patel, S. J.; Wang, W.-L.; Chan,
W.-L.; Ranga Rao, K. R.; Wang, G.; Ngew, X.; Patel, V.; Beer, D.; Lim,
S. P.; Vasudevan, S. G.; Keller, T. H. Peptide Inhibitors of West Nile
NS3 Protease: SAR Study of Tetrapeptide Aldehyde Inhibitors. J. Med.
Chem. 2006, 49, 6585−6590.
(27) Yusof, R.; Clum, S.; Wetzel, M.; Murthy, H. M. K.;
Padmanabhan, R. Purified NS2B/NS3 Serine Protease of Dengue
Virus Type 2 Exhibits Cofactor NS2B Dependence for Cleavage of
Substrates with Dibasic Amino Acids in Vitro. J. Biol. Chem. 2000, 275,
9963−9969.
(28) Li, J.; Lim, S. P.; Beer, D.; Patel, V.; Wen, D.; Tumanut, C.;
Tully, D. C.; Williams, J. A.; Jiricek, J.; Priestle, J. P.; Harris, J. L.;
Vasudevan, S. G. Functional Profiling of Recombinant NS3 Proteases
from All Four Serotypes of Dengue Virus Using Tetrapeptide and
Octapeptide Substrate Libraries. J. Biol. Chem. 2005, 280, 28766−
28774.
(29) Nitsche, C.; Behnam, M. A. M.; Steuer, C.; Klein, C. D. Retro
Peptide-Hybrids as Selective Inhibitors of the Dengue Virus NS2B-
NS3 Protease. Antiviral Res. 2012, 94, 72−79.
(30) Nitsche, C.; Schreier, V. N.; Behnam, M. A. M.; Kumar, A.;
Bartenschlager, R.; Klein, C. D. Thiazolidinone−Peptide Hybrids as
Dengue Virus Protease Inhibitors with Antiviral Activity in Cell
Culture. J. Med. Chem. 2013, 56, 8389−8403.
(10) Erbel, P.; Schiering, N.; D’Arcy, A.; Renatus, M.; Kroemer, M.;
Lim, S. P.; Yin, Z.; Keller, T. H.; Vasudevan, S. G.; Hommel, U.
Structural Basis for the Activation of Flaviviral NS3 Proteases from
Dengue and West Nile Virus. Nat. Struct. Mol. Biol. 2006, 13, 372−
373.
(11) Skorenski, M.; Sienczyk, M. Viral Proteases as Targets for Drug
Design. Curr. Pharm. Des. 2013, 19, 1126−1153.
(12) Au, J. S.; Pockros, P. J. Novel Therapeutic Approaches for
Hepatitis C. Clin. Pharmacol. Ther. 2014, 95, 78−88.
(13) De Clercq, E. Anti-HIV Drugs: 25 Compounds Approved
within 25 Years After the Discovery of HIV. Int. J. Antimicrob. Agents
2009, 33, 307−320.
(14) Cregar-Hernandez, L.; Jiao, G.-S.; Johnson, A. T.; Lehrer, A. T.;
Wong, T. A. S.; Margosiak, S. A. Small Molecule Pan-Dengue and
West Nile Virus NS3 Protease Inhibitors. Antiviral Chem. Chemother.
2011, 21, 209−218.
(15) Shi, P. Y.; Yin, Z.; Nilar, S.; Keller, T. Dengue Drug Discovery.
In Third World Diseases; Elliott, R., Ed.; Springer: Berlin, Germany,
2011; Vol. 7, pp 243−275.
(31) Behnam, M. A. M.; Nitsche, C.; Vechi, S. M.; Klein, C. D. C-
Terminal Residue Optimization and Fragment Merging: Discovery of
a Potent Peptide-Hybrid Inhibitor of Dengue Protease. ACS Med.
Chem. Lett. 2014, 5, 1037−1042.
(16) Mueller, N. H.; Yon, C.; Ganesh, V. K.; Padmanabhan, R.
Characterization of the West Nile Virus Protease Substrate Specificity
and Inhibitors. Int. J. Biochem. Cell Biol. 2007, 39, 606−614.
P
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(32) Bastos Lima, A.; Behnam, M. A. M.; El Sherif, Y.; Nitsche, C.;
Vechi, S. M.; Klein, C. D. Dual Inhibitors of the Dengue and West Nile
Virus NS2B−NS3 Proteases: Synthesis, Biological Evaluation and
Docking Studies of Novel Peptide-Hybrids. Bioorg. Med. Chem. 2015,
23, 5748−5755.
(52) Chen, X.; Murawski, A.; Patel, K.; Crespi, C.; Balimane, P. A
Novel Design of Artificial Membrane for Improving the PAMPA
Model. Pharm. Res. 2008, 25, 1511−1520.
(53) Craik, D. J.; Fairlie, D. P.; Liras, S.; Price, D. The Future of
Peptide-based Drugs. Chem. Biol. Drug Des. 2013, 81, 136−147.
(54) Yan, Z.; Caldwell, G. W. Metabolic assessment in liver
microsomes by co-activating cytochrome P450s and UDP-glycosyl-
transferases. Eur. J. Drug Metab. Pharmacokinet. 2003, 28, 223−232.
(55) Leung, D.; Schroder, K.; White, H.; Fang, N. X.; Stoermer, M.
J.; Abbenante, G.; Martin, J. L.; Young, P. R.; Fairlie, D. P. Activity of
Recombinant Dengue 2 Virus NS3 Protease in the Presence of a
Truncated NS2B Co-factor, Small Peptide Substrates, and Inhibitors. J.
Biol. Chem. 2001, 276, 45762−45771.
(56) Nall, T. A.; Chappell, K. J.; Stoermer, M. J.; Fang, N.-X.;
Tyndall, J. D. A.; Young, P. R.; Fairlie, D. P. Enzymatic Character-
ization and Homology Model of a Catalytically Active Recombinant
West Nile Virus NS3 Protease. J. Biol. Chem. 2004, 279, 48535−48542.
(57) Steuer, C.; Heinonen, K. H.; Kattner, L.; Klein, C. D.
Optimization of Assay Conditions for Dengue Virus Protease: Effect
of Various Polyols and Nonionic Detergents. J. Biomol. Screening 2009,
14, 1102−1108.
(58) Steuer, C.; Gege, C.; Fischl, W.; Heinonen, K. H.;
Bartenschlager, R.; Klein, C. D. Synthesis and Biological Evaluation
of α-Ketoamides as Inhibitors of the Dengue Virus Protease with
Antiviral Activity in Cell-Culture. Bioorg. Med. Chem. 2011, 19, 4067−
4074.
(59) Newton, P.; Harrison, P.; Clulow, S. A Novel Method for
Determination of the Affinity of Protein: Protein Interactions in
Homogeneous Assays. J. Biomol. Screening 2008, 13, 674−682.
(60) Morris, G. M.; Huey, R.; Lindstrom, W.; Sanner, M. F.; Belew,
R. K.; Goodsell, D. S.; Olson, A. J. AutoDock4 and AutoDockTools4:
Automated docking with selective receptor flexibility. J. Comput. Chem.
2009, 30, 2785−2791.
(33) Weigel, L. F.; Nitsche, C.; Graf, D.; Bartenschlager, R.; Klein, C.
D. Phenylalanine and Phenylglycine Analogs as Arginine Mimetics in
Dengue Protease Inhibitors. J. Med. Chem. 2015, 58, 7719−7733.
(34) Wunsch, E.; Fries, G.; Zwick, A. Peptide Synthesis. V.
̈
Preparation and Use of O-Benzyl-L-Tyrosine in the Synthesis of
Peptides. Chem. Ber. 1958, 91, 542−547.
(35) Dawei, M.; Hongqi, T. Stereoselective Synthesis of (S)-
(+)-αM4CPG, a Selective Antagonist of Metabotropic Glutamate
Receptors. Tetrahedron: Asymmetry 1996, 7, 1567−1570.
(36) Caplar, V.; Raza, Z.; Katalenic, D.; Zinic, M. Syntheses of Amino
Alcohols and Chiral C2-Symmetric Bisoxazolines Derived from O-
Alkylated R-4-Hydroxyphenylglycine and S-Tyrosine. Croat. Chem.
Acta 2003, 76, 23−36.
(37) Rzeszotarska, B.; Misiukiewicz, E.; Wiejak, S. A Process for
Preparing Nε-tert-Butoxycarbonyl-L-lysine. PL 194414, May 31, 2007.
(38) Dener, J. M.; Fantauzzi, P. P.; Kshirsagar, T. A.; Kelly, D. E.;
Wolfe, A. B. Large-Scale Syntheses of FMOC-Protected Non-
Proteogenic Amino Acids: Useful Building Blocks for Combinatorial
Libraries. Org. Process Res. Dev. 2001, 5, 445−449.
(39) Nitsche, C.; Klein, C. Fluorimetric and HPLC-Based Dengue
Virus Protease Assays Using a FRET Substrate. In Antiviral Methods
and Protocols; Gong, E. Y., Ed.; Humana Press: New York, 2013; Vol.
1030, pp 221−236.
(40) Chorev, M.; Goodman, M. Recent Developments in Retro
Peptides and Proteins  an Ongoing Topochemical Exploration.
Trends Biotechnol. 1995, 13, 438−445.
(41) Erlanson, D. A. Fragment-based Lead Discovery: A Chemical
Update. Curr. Opin. Biotechnol. 2006, 17, 643−652.
(61) Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.;
Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E. UCSF ChimeraA
visualization system for exploratory research and analysis. J. Comput.
Chem. 2004, 25, 1605−1612.
(42) Erlanson, D. A.; McDowell, R. S.; O’Brien, T. Fragment-based
Drug Discovery. J. Med. Chem. 2004, 47, 3463−3482.
(43) Hajduk, P. J.; Greer, J. A Decade of Fragment-based Drug
Design: Strategic Advances and Lessons Learned. Nat. Rev. Drug
Discovery 2007, 6, 211−219.
(44) Joseph-McCarthy, D.; Campbell, A. J.; Kern, G.; Moustakas, D.
Fragment-Based Lead Discovery and Design. J. Chem. Inf. Model. 2014,
54, 693−704.
(45) Bodenreider, C.; Beer, D.; Keller, T. H.; Sonntag, S.; Wen, D.;
Yap, L.; Yau, Y. H.; Shochat, S. G.; Huang, D.; Zhou, T.; Caflisch, A.;
Su, X.-C.; Ozawa, K.; Otting, G.; Vasudevan, S. G.; Lescar, J.; Lim, S.
P. A Fluorescence Quenching Assay to Discriminate between Specific
and Nonspecific Inhibitors of Dengue Virus Protease. Anal. Biochem.
2009, 395, 195−204.
(46) Yung-Chi, C.; Prusoff, W. H. Relationship between the
Inhibition Constant (KI) and the Concentration of Inhibitor Which
Causes 50 per Cent Inhibition (I50) of an Enzymatic Reaction.
Biochem. Pharmacol. 1973, 22, 3099−3108.
(47) Trott, O.; Olson, A. J. AutoDock Vina: Improving the speed and
accuracy of docking with a new scoring function, efficient
optimization, and multithreading. J. Comput. Chem. 2010, 31, 455−
461.
(48) Noble, C. G.; Seh, C. C.; Chao, A. T.; Shi, P. Y. Ligand-Bound
Structures of the Dengue Virus Protease Reveal the Active
Conformation. J. Virol. 2012, 86, 438−446.
(49) Giaginis, C.; Tsantili-Kakoulidou, A. Current State of the Art in
HPLC Methodology for Lipophilicity Assessment of Basic Drugs. A
Review. J. Liq. Chromatogr. Relat. Technol. 2007, 31, 79−96.
(50) Dalvie, D. K.; Kalgutkar, A. S.; Khojasteh-Bakht, S. C.; Obach,
R. S.; O’Donnell, J. P. Biotransformation Reactions of Five-Membered
Aromatic Heterocyclic Rings. Chem. Res. Toxicol. 2002, 15, 269−299.
(51) Kansy, M.; Senner, F.; Gubernator, K. Physicochemical High
Throughput Screening: Parallel Artificial Membrane Permeation Assay
in the Description of Passive Absorption Processes. J. Med. Chem.
1998, 41, 1007−1010.
Q
DOI: 10.1021/acs.jmedchem.5b01441
J. Med. Chem. XXXX, XXX, XXX−XXX