Peptide-Boronic Acid Inhibitors of Flaviviral Proteases: Medicinal Chemistry and Structural Biology

Article abstract of DOI:10.1021/acs.jmedchem.6b01021

A thousand-fold affinity gain is achieved by introduction of a C-terminal boronic acid moiety into dipeptidic inhibitors of the Zika, West Nile, and dengue virus proteases. The resulting compounds have K_i values in the two-digit nanomolar range

Full text of DOI:10.1021/acs.jmedchem.6b01021

Subscriber access provided by Olson Library | Northern Michigan University
Brief Article
Peptide-Boronic Acid Inhibitors of Flaviviral
Proteases: Medicinal Chemistry and Structural Biology
Christoph Nitsche, Linlin Zhang, Lena F. Weigel, Jonas Schilz, Dominik
Graf, Ralf Bartenschlager, Rolf Hilgenfeld, and Christian D. Klein
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01021 • Publication Date (Web): 28 Nov 2016
Downloaded from http://pubs.acs.org on November 29, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 9
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Peptide-Boronic Acid Inhibitors of Flaviviral Proteases: Me-
dicinal Chemistry and Structural Biology
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
§,
†
§
†
Christoph Nitsche,†,‡ Linlin Zhang, Lena F. Weigel, Jonas Schilz, Dominik Graf, Ralf Barten-
∥
§,
,†
schlager,∥,⊥ Rolf Hilgenfeld, Christian D. Klein*
∥
†Medicinal Chemistry, IPMB, Heidelberg University, INF-364, 69120 Heidelberg, Germany
§Institute of Biochemistry, Center for Structural and Cell Biology in Medicine, University of Lübeck, Ratzeburger
Allee 160, 23562 Lübeck, Germany
∥German Center for Infection Research (DZIF), sites Hamburg - Lübeck - Borstel and Heidelberg
⊥Department of Infectious Diseases, Molecular Virology, Heidelberg University, INF-345, 69120 Heidelberg, Germa-
ny
ABSTRACT: A thousand-fold affinity gain is achieved by introduction of a C-terminal boronic acid moiety into dipeptidic
inhibitors of the Zika, West Nile and dengue virus proteases. The resulting compounds have Ki values in the two-digit
nanomolar range, are not cytotoxic, and inhibit virus replication. Structure-activity relationships and a high resolution X-
ray co-crystal structure with West Nile virus protease provide a basis for the design of optimized covalent-reversible in-
hibitors aimed at emerging flaviviral pathogens.
INTRODUCTION
also allowed the co-crystallization with the target prote-
ases, yielding the first X-ray structures of a small-
molecule, drug-like inhibitor covalently bound to West
Nile and Zika virus proteases that enable the rational
design of further derivatives with even higher potency. It
should also be noted that haptenization and subsequent
allergic reactions are not expected to be an issue with
boronic acids, since they bind to biomacromolecules in a
reversible-covalent manner.
Previous work by several groups7-12 had indicated that
compounds containing basic residues which resemble the
polybasic substrate recognition profile of the flaviviral
proteases display modest affinities in the micromolar
range. In comparison, extension of these compounds
towards the N-terminus and in the area of the P1 residue,
as reported recently, resulted in peptidic inhibitors of the
DENV and WNV proteases with activity in the lower na-
nomolar range.13-16 However, the size (Mr ≈ 750 Da) and
basicity of these compounds present a problematic liabil-
ity for further optimization and pharmacokinetics. We
therefore considered it desirable to obtain alternative and
smaller compounds in which these design factors are in a
more attractive range. Considering a previous report,17 in
which a single boronic acid–tetrapeptide analog was re-
ported as promising (Ki = 0.043 µM) but not studied in
more detail, we set out to explore the potential of reversi-
ble-covalent dipeptide-boronic acid inhibitors of flaviviral
proteases and elucidate their structure-activity relation-
ships.
Infections with flaviviruses such as dengue, West Nile,
and Zika virus are an increasing global health concern.
With spreading habitats of the transmitting mosquitoes,
infections and epidemics caused by these flaviviruses are
a potential threat for previously unaffected regions such
as Europe and Northern America. The recent Zika virus
outbreak in South America has been declared as Public
Health Emergency of International Concern by the WHO.
Enzyme inhibitors with a covalent binding mode1 cur-
rently experience a remarkable revival in medicinal chem-
istry within a large variety of disease and target classes.
The beta-lactam antibiotics, fosfomycin, aspirin and pro-
ton-pump inhibitors are long-established examples of
drugs with a covalent binding mode and a convincing
clinical activity and safety profile. Newly developed cova-
lent inhibitors entered clinical practice, for example, in
the kinase2, 3 and proteasome fields,4, 5 and have recently
shown promise as inhibitors of serine carbapenemases.6
From a chemical perspective, these compounds include
potentially reactive functional groups such as boronic
acid, acrylamide or epoxide that are otherwise considered
highly unusual in drugs. The introduction of as-yet un-
derexplored elements and building blocks into medicinal
chemistry obviously represents a highly valuable strategy.
We here present an example in which the inclusion of a
boronic acid moiety into peptidic inhibitors of flaviviral
proteases led to a thousand-fold increase of affinity. The
resulting compounds show efficacy in cell-culture models
of dengue (DENV) and West Nile virus (WNV) replication
and are not cytotoxic. The high affinity of the compounds
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 9
RESULTS AND DISCUSSION
1
The chemistry behind the present SAR dataset involved
stereoselective preparation of the key intermediate 9 by
Matteson homologation using the (+)-pinanediol ester as
described previously (SI).18-21 The further synthesis is
based on a chiral-pool approach for the preparation of
benzoyl-capped unnatural amino acids 10–15 (Scheme S2,
SI), followed by fusion to precursor 9 (Scheme S1) and
subsequent multi-step conversion of the side chain bro-
mine to a guanidine function (Scheme 1).
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The syntheses of the monopeptide-boronic acids 1 and
2 from 9 are outlined in the supporting information
(Scheme S1, SI). The dipeptide-boronic acids 3–8, contain-
ing various aromatic bioisosters for the basic P2 side
chain, were prepared from acids 10-15 according to
Scheme 1. Coupling towards precursor 9 was successful in
moderate yields, using the isobutyl mixed anhydride ap-
proach. The bromine intermediates were converted into
the corresponding azides, the compounds were hydro-
genated and finally transformed into the di(Boc)-
protected guanidines.
The two guanidino groups in compounds 6–8 could be
introduced simultaneously within one synthetic step after
parallel reduction of the nitro and azido groups. This
combined transformation crucially reduced the number
of synthetic steps and improved the yield. For final depro-
tection, all remaining Boc groups were cleaved using TFA.
The boronic acids 1-8 were subsequently released from
pinanediol with an excess of phenylboronic acid in a two-
phase mixture of water and diethyl ether.
Scheme 1. Synthesis of dipeptide-boronic acids.a
aReagents and conditions: (a) NMM, IBCF, DIPEA THF,
DCM, –15 °C; (b) NaN3, DMF, 100 °C; (c) H2, Pd/C, MeOH; (d)
bis-Boc-pyrazole-1-carboxamidine, DMAP, MeOH; (e) TFA,
DCM, HCl; (f) PhB(OH)2, water, diethyl ether, HCl (aq.). The
dipeptide 8 was synthesized in analogy to 7 from precursor 15
with a tBu group in para-position. Step (c) was performed
under acidic conditions (aqueous HCl) during synthesis of
compounds 5–8. All final compounds were isolated as hydro-
chloride salts.
ACS Paragon Plus Environment

Page 3 of 9
Journal of Medicinal Chemistry
Table 1. Activity data. Inhibitory activity of peptidyl-amides and -boronic acids against viral proteases and inhi-
bition of dengue and West Nile virus replication by the peptidyl-boronic acids in cell culture.
1
2
Inhibition of
Inhibition of
Inhibition of
Cellular activity
of the boronic acids (X=B(OH)2)
3
4
5
6
7
8
9
Cpd. ID
Structure
DENV protease[a]
WNV protease[b]
ZIKV protease[c]
for X =
B(OH)2
X = replacement of the terminal Arg/Orn carboxyl group;
Ki and IC50 values in µM; % inhibition values at [inhibitor] = 50 µM
Cyto-
Antiviral effect
EC50 (µM)[e]
toxicity[d]
X = CONH2
2.0%
X = B(OH)2
8.3%
X = CONH2
18%
X = B(OH)2
30%
X = B(OH)2
IC50 > 50
CC50 (µM)
> 100
DENV
WNV
33%
1
Bz-Orn-X
7.5%
12%
2
Bz-Arg-X
9.7%
51.9%
9.6%
18%
IC50 > 50
> 100
19%
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Bz-(4-NH2)Phe-
Arg-X
IC50 = 7.4
Ki = 4.2
IC50 = 7.4
Ki = 6.0
44%
(3.1 µM)
EC50
1.5
=
3
5.2%
6.6%
13%
IC50 > 40
10.1
Bz-[4-
(CH2NH2)]Phe-
Arg-X
IC50 = 0.066[f]
IC50 = 0.11[g]
EC50
30
=
EC50
38
=
IC50 = 0.25
Ki = 0.04
4
20%
> 100
Ki = 0.051[f]
Ki = 0.082[g]
Bz-(3-guani–
dinyl)Phg-Arg-X
IC50 = 0.26
Ki = 0.22
IC50 = 0.14
Ki = 0.12
5
6
7
9.8%
22%
33%
34%
> 100
≥ 100
> 100
24%
47%
38%
IC50 = 1.2
IC50 = 1.9
IC50 = 0.83
Bz-(3-guani–
dinyl)Phe-Arg-X
IC50 = 0.11
Ki = 0.091
IC50 = 0.31
Ki = 0.30
EC50
48
=
IC50 = 211
Bz-(4-guani–
dinyl)Phe-Arg-X
IC50 = 87
Ki = 29
IC50 = 0.036[f]
IC50 = 0.071[h]
EC50
25
=
EC50 = 18
Ki = 0.027[f]
Ki = 0.065[h]
4-tBuBz-(4-
guanidinyl)Phe-
Arg-X
IC50 = 0.11[f]
IC50 = 0.21
Ki = 0.16
EC50
19
=
8
IC50 = 145
34%
> 100
30%
Ki = 0.078[f]
IC50 = 2.1
All measurements were performed in triplicate. Standard deviations are always below 10%. Percentage inhibition values de-
termined at 50 µM inhibitor concentration. Ki values were calculated from measurements at substrate concentrations of 50, 100,
150, and 200 µM. [a] Dengue serotype 2 NS2B-NS3 protease (100 nM unless stated otherwise). Substrate: 50 µM Abz-Nle-Lys-Arg-
Arg-Ser-3-(NO2)Tyr (Km = 105 µM). [b] WNV NS2B-NS3 protease (150 nM unless stated otherwise). Substrate: 50 µM Abz-Gly-Leu-
Lys-Arg-Gly-Gly-3-(NO2)Tyr (Km = 212 µM). [c] ZIKV NS2B-NS3 protease (5 nM). Substrate: 10 µM Bz-Nle-Lys-Lys-Arg-AMC (Km
=
18 µM). [d] Cytotoxicity was assayed up to a maximum concentration of 100 µM. [e] Plaque assay results for DENV and WNV. Per-
centage inhibition values at 50 µM or at the highest non-toxic inhibitor concentration. [f] 50 nM DENV protease. [g] 100 nM WNV
protease. [h] 75 nM WNV protease.
As shown in Table 1, significant activity is restricted to
compounds that contain the reversible-covalent boronic
acid building block at the C-terminus, with the corre-
sponding amides being practically inactive. Analogs of the
most active compounds 4 and 7 with a C-terminal carbox-
ylic acid were also inactive (Table S2). The second im-
portant factor is the basicity of the C-terminal residue:
The ornithine-boronic acid derivative 1 is significantly less
active than its arginine congener 2 at DENV protease, but
slightly more active at WNV protease. Further extension
by one additional, basic-lipophilic residue such as ami-
nomethyl-phenylalanine (4) and guanidino-phenylalanine
(6, 7, 8) lead to a dramatic increase in activity. A clear
preference for a para substitution of the basic moiety is
evident for all three viral proteases, revealing compounds
4 and 7 as the most active ones with nanomolar affinity. A
more flexible phenylalanine moiety in P2 is preferred
compared to a truncated and rigified phenylglycine scaf-
fold (5 vs. 6).
contain additional, specific molecular recognition ele-
ments.22
It is noteworthy that the affinity of the dipeptidic com-
pound 7 (Ki = 0.027 µM) is in the same range as that of the
previously reported tetrapeptidic inhibitor Bz-Nle-Lys-
Arg-Arg-B(OH)2 (reported Ki = 0.043 µM) against DENV
protease. The boronic acid derivatives also compare fa-
vorably with the recently reported, much larger N-
terminally capped, non-natural tripeptides that reach IC50
values of 18 and 50 nM at DENV and WNV protease, re-
spectively (compound 83 in the reference).15
While in case of DENV and WNV proteases the selec-
tivity between compounds 4 and 7 is limited, for ZIKV
protease a remarkable preference towards compound 4
with the 4-aminomethyl-phenylalanine side chain in P2 is
evident. This indicates a very pronounced preference for
lysine mimetics in P2 and may guide the way towards
highly affine ZIKV protease ligands.
A compelling comparison can be made for the most ac-
tive boronic acids 4 and 7 and their respective amide
congeners. For both compounds, replacement of the am-
ide group by a reversible-covalent boronic acid function
increases the affinity by a factor of 1000.
The reversible-covalent building block, however, always
requires the presence of specific recognition elements, as
evidenced by the relative inactivity of the monopeptide-
boronic acids 1 and 2. This is in agreement with a recent
study that showed protein targets to be remarkably re-
sistant against small electrophilic fragments which do not
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 9
The broad inhibitory effect of these compounds against bonds with the main-chain oxygen of Asp82* and with
1
2
3
4
5
6
7
8
all analyzed viral proteases gives hope for a route towards
pan-antiflaviviral drugs.
Asn84* Oδ1, both from the NS2B polypeptide chain. The-
se interactions presumably contribute to the stabilization
of the closed form of the complex.
The cellular activities reflect the high in-vitro poten-
cies, with the boronic acid 7 showing the highest activity
against dengue protease in-vitro and DENV replication in
cell culture. A single compound (3) had cytotoxic effects,
probably due to the aniline moiety, with all other com-
pounds being not cytotoxic up to 100 µM. Although this
indicates good target selectivity, the inhibition data on
trypsin (cf. SI) show that further improvements are re-
quired. In contrast, selectivity against thrombin is already
relatively good. (cf. SI)
Remarkably, the phenyl group of the P2 residue inter-
acts with the imidazole of the catalytic His51 through π
stacking (Fig. 2; distance ∼3.5 Å). The Nε atom of the P1
guanidinium group is involved in a 2.97±0.08-Å hydrogen
bond with the carbonyl oxygen of Tyr130, and the Nη2
donates a "charged" hydrogen bond to Asp129 Oδ1
(2.97±0.22 Å; average and standard deviation calculated
from the 3 monomers of the complex in the asymmetric
unit of the crystal). Interestingly, the Nη1 forms a
3.18±0.11-Å hydrogen bond with the carbonyl oxygen of
the inhibitor's P2 residue, drawing the P1 guanidinium
group close to the main chain of the compound. Moreo-
ver, there may be a weak (parallel) interaction between
the guanidinium group and the phenyl group of Tyr161
(Fig. 2).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In contrast to a recently solved, lower resolution dimer-
ic structure of compound 4 in complex with ZIKV prote-
ase (PDB code 5LC0), the present monomeric WNV crys-
tal structure shows n0 evidence for specific contacts be-
tween two protease units, and no shape complementarity,
as opposed to the crystal structure of the Zika virus NS2B-
NS3 protease.23 The overall observed inhibitor orientation
in the active site is quite similar for both crystal structures
(cf. SI Figure S1) and also in close correlation to an NMR
structure of derivative 8 with DENV-2 protease.24
The boron atom of compound 4 forms a 1.59±0.01-Å co-
valent bond with Oγ of the active-site Ser135 (Fig. 2, upper
right inset). During refinement, additional positive Fo–Fc
density emerged near the two hydroxyl groups of the
boronic acid and was unambiguously assigned to a cyclic
di-ester formed between the boronic acid and atoms O1
and O2 of glycerol, leading to the formation of a 5-
membered ring (Fig. 2, lower left inset). Boronic acids
tend to form esters with aliphatic diols, especially if 5- or
6-membered rings can be formed,25 and glycerol was pre-
sent in our enzyme preparation during purification, crys-
tallization, and cryoprotection of crystals.
Figure 1. Overall structure of the WNV NS2B-NS3 protease
in complex with compound 4, PDB code 5IDK. A distinctive
feature of the closed (active) form of the enzyme as found
here is that the NS2B chain (red) is wrapped around the NS3
protease domain (green). The Gly4-Ser-Gly4 linker between
NS2B and NS3 is not defined by electron density. Chain ter-
mini are labeled N*, C* for NS2B and N, C for NS3. Colors of
inhibitor atoms: C, magenta; N, blue; O, dark red; B, yellow.
Active-site residues are dark blue and labeled.
The presence of alcohols (e.g., TRIS) and polyols in bio-
logical and biochemical assays leads to formation of an
equilibrium mixture of boronate esters and boronic acids.
For example, the formation of a TRIS ester under assay
conditions was detected for a representative compound of
the current dataset by mass spectrometry (data not
shown). The kinetics and dynamics of these ester/acid
equilibria are not yet well characterized, neither for the
currently presented compounds nor for the clinically used
boronic acids. Affinity data for boronic acid electrophiles
must therefore be interpreted with caution, because the
composition of buffers and biological media influences
the concentration of the "active", free boronic acid.
A 1.50-Å crystal structure of 4 in complex with the
WNV NS2B-NS3 protease revealed the enzyme in the
closed (active) conformation, with the NS2B chain
wrapped around the NS3 domain (Fig. 1, PDB code 5IDK).
The amide backbone of the inhibitor forms only two hy-
drogen bonds to the protease (P1 NH...Gly151 O, P3
O...Gly153 NH), whereas the P1 and P2 sidechains and the
P3 benzoyl cap are neatly accommodated in the corre-
sponding subsites of the enzyme (Fig. 2). The terminal
NH2 group of the P2 sidechain is involved in hydrogen
Negatively charged and chiral, the tetrahedral boronate
unit is a close mimic of the transition state of peptide
hydrolysis. The cyclic-ester oxygens (O1 and O2) accept
hydrogen bonds from the amide of Gly133 (2.85±0.05 Å;
ACS Paragon Plus Environment

Page 5 of 9
Journal of Medicinal Chemistry
towards the protein backbone. Creation of additional
part of the oxyanion hole) and the Nε2 of His51 (2.67±0.04
1
2
3
4
Å), respectively. The five-membered ring occupies the S1'
site, but the hydroxy group (O3 of glycerol) is oriented
towards bulk solvent.
protein backbone – inhibitor interactions would be highly
attractive since this is expected to impede resistance de-
velopment and increase the likelihood of broad-spectral
activity against proteases from multiple flaviviruses.
5
6
7
8
An unresolved issue is the relatively weak activity of the
compounds in the virus replication assay, and the lack of
correlation between the cellular and the biochemical
activity of the compounds. We assume that the high po-
larity of the compounds, not only due to the basic side
chains but also to the boronic acid moiety, impedes
membrane passage, and thus, cellular uptake of the com-
pounds. In this respect, the boronic acid moiety, while
contributing significantly to target affinity, probably has a
detrimental effect. Therefore, while the value of the pre-
sent compounds for structural and biochemical studies is
high – such as successfully shown for the protease of the
newly emerged Zika virus23 – modifications will be re-
quired to reach higher antiviral efficacy in cells. We en-
visage that this can be reached via two strategies: First,
the boronic acid might be modified to a prodrug, for ex-
ample by esterification with diols, yielding compounds
that resemble the inhibitor-glycerol complex as found in
the present X-ray structure. Second, the boronic acid
might be replaced by other electrophilic functional
groups with lower polarity. Alternative functional groups
such as the fluoromethylketone moiety would further-
more allow the extension of the inhibitors towards the
primed sites. This may increase selectivity towards fla-
viviral proteases in relation to eukaryotic off-targets such
as trypsin and thrombin, which remains problematic for
the compounds presented here.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Compound 4 in the substrate-binding site of WNV
NS2B-NS3 protease. The Fo-Fc difference density, calculated
from the final coordinates with the inhibitor omitted and
contoured at 4σ, is shown as blue grid. Residues of the NS2B
chain are indicated by an asterisk (*). H-bonds are shown as
dashed lines. Upper right inset: The tetrahedral boronate;
Ser135 in green, boron in yellow, oxygens and carbons of the
glycerol moiety in red, Fo-Fc difference density as blue grid.
Lower left inset: Structure of the glycerol boronate bound to
Ser135.
In conclusion, the current study not only contributes to
our understanding of flaviviral proteases and their sub-
strate and inhibitor binding mechanisms, but also sup-
ports the ongoing search for novel therapeutics to treat
infections with dengue, West Nile, Zika and other emerg-
ing flaviviruses.
The structure presented here has the highest resolution
of all flavivirus protease structures in complex with a
small-molecule inhibitor published thus far and will ena-
ble us and others to further optimize target affinity and
selectivity of flaviviral protease inhibitors.
Some key consequences for drug design shall be sum-
marized here: Several backbone atoms of the inhibitor are
not engaged in interactions with the protease and may
therefore be replaced in order to reduce the peptidic na-
ture of the inhibitor. This relates in particular to the P2
main-chain amide, which could be replaced by a meth-
ylene group or a similar functionality, allowing the crea-
tion of peptidomimetics with increased drug-likeness. A
ring closure between the P2 main-chain carbonyl and the
P1 side-chain appears possible and would reduce the con-
formational degrees of freedom of the inhibitor, thus
further increasing affinity.
EXPERIMENTAL SECTION
General Procedures. Detailed information are provided
in the Supporting Information. In brief, all reagents and
solvents were used as obtained from commercial sources
unless otherwise indicated. Sensitive reagents were gen-
erally handled under nitrogen atmosphere. The reaction
progress was monitored by TLC. Normal phase chroma-
tographic purification was performed on a Biotage Isolera
One purification system using silica gel and UV monitor-
ing. RP-phase purifications were executed using an ÄKTA
purifier (GE Healthcare) system with Reprospher C18-DE,
Dr. Maisch (5 μm) 30 x 16 mm, and 120 x 16 mm pre- and
main columns and UV detection. NMR spectra were rec-
orded on Varian Mercury Plus (300 MHz) and Varian
NMR System 500 (500 MHz) instruments at 300 K. Mass
spectrometry was performed on a Bruker micrOTOF-Q II
(ESI) instrument. Analytical HPLC was performed on a
Jasco HPLC system with UV detector and ReproSil-Pur-
ODS-3, Dr. Maisch GmbH, Germany, 5 µm, 50 x 2 mm
The structure also highlights significant opportunities
to further extend these inhibitors in a rational way, not
only with respect to target affinity but also in light of
selectivity: Most prominently, the glycerol ester indicates
that inhibitors can be extended towards the prime-sites of
the substrate-binding cleft, which so far remain complete-
ly unexploited. The P2 and P3 regions offer significant
potential for the design of additional specific interactions
with side chains of the target and, more importantly,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 9
and TFA (1 ml) overnight. After solvent evaporation, the
column. Active compounds 3-8 were of ≥95% purity ac-
cording to HPLC analysis.
1
2
3
4
5
6
7
8
residue was triturated with hydrogen chloride (1 ml, 4 M
in dioxane) and the solvent was evaporated again to ob-
tain a crude intermediate (120 mg, quant.). HRMS (ESI):
m/z [M+H]+ calcd for C32H46BN6O4: 589.3674, found:
589.3672. The residue was stirred overnight in a two-
phase mixture of water and diethyl ether containing phe-
nylboronic acid (109 mg, 0.9 mmol). Afterwards, the
phases were separated and the aqueous phase was ex-
tracted with diethyl ether for three times. The aqueous
phase was concentrated and directly used for purification
by preparative HPLC (methanol/water 0.1% TFA). The
excess of methanol was evaporated and hydrochloric acid
(1 ml, 1N) was added before freeze-drying. Compound 4
was obtained as colorless solid (15 mg, 16%). 1H-NMR (500
MHz, D2O): δ = 1.38-1.54 (m, 4H), 2.65 (m, 1H), 3.13 (t, J =
6.8 Hz, 2H), 3.17 (dd, J = 13.9, 9.5 Hz, 1H), 3.33 (dd, J = 13.8,
6.4 Hz, 1H), 4.12 (s, 2H), 4.96 (dd, J = 9.5, 6.5 Hz, 1H), 7.38
(m, 4H), 7.27 (m, 2H), 7.61 (m, 3H) ppm; 13C-NMR (125
MHz, D2O): δ = 25.8, 27.1, 36.3, 40.9, 42.7, 52.8, 127.1, 128.7,
129.1, 129.9, 131.4, 132.5, 132.6, 137.2, 170.9, 174.8 ppm;
HRMS (ESI): m/z [M–H2O+H]+ calcd for C22H30BN6O3:
437.2471, found: 437.2463.
Synthesis of precursor 9. Compound 9 was synthesized
from allyl bromide in 4 steps according to the literature21
1
(5.60 g, 64%, final step). H-NMR (300 MHz, CDCl3): δ =
0.85 (s, 3H), 0.98 (t, J = 7.4 Hz, 2H), 1.10 (d, J = 10.9 Hz,
1H), 1.30 (s, 3H), 1.39 (s, 3H), 1.80-2.07 (m, 5H), 2.23 (m,
1H), 2.33 (m, 1H), 3.45 (t, J = 6.9 Hz, 2H), 4.27 (dd, J = 8.7,
2.0 Hz, 1H) ppm; 13C-NMR (75 MHz, CDCl3): δ = 24.0, 26.5,
27.1, 27.6, 28.6, 35.5, 36.3, 38.1, 39.5, 51.3, 77.7, 85.6 ppm;
MS (ESI): m/z: 323.1 [M+Na]+.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of precursor amino acid 11. Compound 11
was synthesized from L-4-cyanophenylalanine in 5 steps
following established synthetic procedures (442 mg, 93%,
1
final step). H-NMR (300 MHz, CDCl3): δ = 7.72 (d, J = 7.1
Hz, 2H), 7.38-7.61 (m, 4H), 7.32 (d, J = 8.2 Hz, 1H), 7.05-
7.16 (m, J = 6.8 Hz, 2H), 6.69 (br s, 1H), 5.03-5.16 (m, 1H),
4.29 (s, 1H), 3.23-3.53 (m, 2H), 1.46 (s, 6H) ppm; HRMS
(ESI): m/z [M–H]– calcd for C22H25N2O5: 397.1769, found:
397.1770.
Synthesis of 4. A solution of 11 (200 mg, 0.5 mmol) and
N-methylmorpholine (60 ꢀl, 0.55 mmol) was dissolved in
THF and cooled to -15 °C, before IBCF (130 ꢀl, 1 mmol)
was added. Afterwards a solution of 9 (220 mg, 0.6 mmol)
in dichloromethane was added dropwise, followed by a
slightly delayed addition of DIPEA (190 ꢀl, 1.1 mmol) in
dichloromethane at -15 °C over 30 min. Afterwards, the
reaction mixture was allowed to warm up to room tem-
perature and stirred overnight. The solvent was evapo-
rated and the residue was purified by flash chromatog-
raphy (cyclohexane/ethyl acetate) to obtain a crude in-
termediate (59%). HRMS (ESI): m/z [M+Na]+ calcd for
C36H49BBrN3NaO6: 732.2796, found: 732.2789. The inter-
mediate (320 mg, 0.45 mmol) was treated with sodium
azide (39 mg, 0.6 mmol) in DMF at 100 °C for 1 h. After-
wards, the solvent was evaporated and the residue was
purified by flash chromatography (cyclohexane/ethyl
acetate) to obtain a crude product (79%). HRMS (ESI):
m/z [M+Na]+ calcd for C36H49BN6NaO6: 695.3705, found:
695.3712. A mixture of this intermediate (220 mg, 0.33
mmol) and palladium on carbon (10%, 35 mg) in metha-
nol was stirred at room temperature under hydrogen
atmosphere (1 bar) overnight. The mixture was filtered
using celite, the solvent was evaporated and the crude
product was directly used for the next synthetic step (200
mg, 94%). HRMS (ESI): m/z [M+H]+ calcd for
C36H52BN4O6: 647.3981, found: 647.3957. A solution of the
crude intermediate (195 mg, 0.3 mmol), bis-Boc-pyrazole-
1-carboxamidine (116 mg, 0.375 mmol) and DMAP (10 mg,
0.075 mmol) in methanol was stirred 2 days at room tem-
perature. The solvent was evaporated and the residue was
purified by flash chromatography (cyclohexane/ethyl
acetate) to obtain a crude product (175 mg, 66%). HRMS
(ESI): m/z [M+H]+ calcd for C47H70BN6O10: 889.5249,
found: 889.5228. The intermediate (160 mg, 0.18 mmol)
was treated with a mixture of dichloromethane (10 ml)
ASSOCIATED CONTENT
Supporting Information. Detailed biochemical and chemi-
cal procedures, structures of all intermediates, biochemical
and analytical data, details of the crystallisation, X-ray data
collection, and refinement. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: c.klein@uni-heidelberg.de. Phone: ++49-6221-
544875.
Present Address
‡C.N.: Australian National University, Research School of
Chemistry, Canberra, ACT 2601, Australia
Funding Sources
The project was supported by the Deutsche Forschungsge-
meinschaft, KL-1356/3-1, and by the German Center for Infec-
tion Research (DZIF), 8029801911.
ACKNOWLEDGMENT
We thank Dr. Tom Sundermann, Heiko Rudy, Tobias Tim-
mermann, Natascha Stefan and Yasmin Gül for support. Dr.
Mascha Jäkel kindly provided (+)-pinanediol. Dr. Mira A. M.
Behnam and Dr. Daizong Lin were valuable partners for
discussions.
ABBREVIATIONS
DENV, dengue virus; WNV, West Nile virus; ZIKV, Zika
virus; NS2B, non-structural protein 2B; NS3, non-structural
protein 3.
ACS Paragon Plus Environment

Page 7 of 9
Journal of Medicinal Chemistry
12.
Leung, D.; Schroder, K.; White, H.; Fang, N.-X.;
PDB ID CODES
1
2
3
4
5
6
7
8
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.
Atomic coordinates of the WNV NS2B-NS3 protease in com-
plex with compound 4 will be released upon acceptance of
the article, PDB code 5IDK.
REFERENCES
13.
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.
1.
Smith, A. J.; Zhang, X.; Leach, A. G.; Houk, K. N.
Beyond picomolar affinities: Quantitative aspects of
noncovalent and covalent binding of drugs to proteins. J.
Med. Chem. 2009, 52, 225-233.
9
14.
Bastos Lima, A.; Behnam, M. A. M.; El Sherif, Y.;
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
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.
2.
Li, D.; Ambrogio, L.; Shimamura, T.; Kubo, S.;
Takahashi, M.; Chirieac, L. R.; Padera, R. F.; Shapiro, G. I.;
Baum, A.; Himmelsbach, F.; Rettig, W. J.; Meyerson, M.;
Solca, F.; Greulich, H.; Wong; K, K. BIBW2992, an
irreversible EGFR/HER2 inhibitor highly effective in
preclinical lung cancer models. Oncogene 2008, 27, 4702-
4711.
15.
Behnam, M. A. M.; Graf, D.; Bartenschlager, R.;
Zlotos, D. P.; Klein, C. D. Discovery of nanomolar dengue
and West Nile virus protease inhibitors containing a 4-
benzyloxyphenylglycine residue. J. Med. Chem. 2015, 58,
9354-9370.
3.
Weisner, J.; Gontla, R.; van der Westhuizen, L.;
Oeck, S.; Ketzer, J.; Janning, P.; Richters, A.; Mühlenberg, T.;
Fang, Z.; Taher, A.; Jendrossek, V.; Pelly, S. C.; Bauer, S.; van
Otterlo, W. A. L.; Rauh, D. Covalent-allosteric kinase
inhibitors. Angew. Chem. Int. Ed. 2015, 54, 10313-10316.
16.
Weigel, L. F.; Nitsche, C.; Graf, D.; Bartenschlager,
R.; Klein, C. D. Phenylalanine and phenylglycine analogues as
arginine mimetics in dengue protease inhibitors. J. Med.
Chem. 2015, 58, 7719-7733.
4.
Dubiella, C.; Baur, R.; Cui, H.; Huber, E. M.; Groll,
17.
Yin, Z.; Patel, S. J.; Wang, W.-L.; Wang, G.; Chan,
M. Selective inhibition of the immunoproteasome by
structure-based targeting of a non-catalytic cysteine. Angew.
Chem. Int. Ed. 2015, 54, 15888-15891.
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.
5.
Adams, J.; Palombella, V. J.; Sausville, E. A.;
Johnson, J.; Destree, A.; Lazarus, D. D.; Maas, J.; Pien, C. S.;
Prakash, S.; Elliott, P. J. Proteasome inhibitors: A novel class
of potent and effective antitumor agents. Cancer Res. 1999,
59, 2615-2622.
18.
synthesis. J. Org. Chem. 2013, 78, 10009-10023.
19. Matteson, D. S.; Ray, R. Directed chiral synthesis
Matteson, D. S. Boronic esters in asymmetric
with pinanediol boronic esters. J. Am. Chem. Soc. 1980, 102,
7590-7591.
6.
Hecker, S. J.; Reddy, K. R.; Totrov, M.; Hirst, G. C.;
Lomovskaya, O.; Griffith, D. C.; King, P.; Tsivkovski, R.; Sun,
D.; Sabet, M.; Tarazi, Z.; Clifton, M. C.; Atkins, K.; Raymond,
A.; Potts, K. T.; Abendroth, J.; Boyer, S. H.; Loutit, J. S.;
Morgan, E. E.; Durso, S.; Dudley, M. N. Discovery of a cyclic
boronic acid β-lactamase inhibitor (rpx7009) with utility vs
class A serine carbapenemases. J. Med. Chem. 2015, 58, 3682-
3692.
20.
Matteson, D. S.; Sadhu, K. M. Boronic ester
homologation with 99% chiral selectivity and its use in
syntheses of the insect pheromones (3S,4S)-4-methyl-3-
heptanol and exo-brevicomin. J. Am. Chem. Soc. 1983, 105,
2077-2078.
21.
Wityak, J.; Earl, R. A.; Abelman, M. M.; Bethel, Y. B.;
Fisher, B. N.; Kauffman, G. S.; Kettner, C. A.; Ma, P.;
McMillan, J. L. Synthesis of thrombin inhibitor DuP 714. J.
Org. Chem. 1995, 60, 3717-3722.
7.
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.
22.
Jöst, C.; Nitsche, C.; Scholz, T.; Roux, L.; Klein, C. D.
Promiscuity and selectivity in covalent enzyme inhibition:
A systematic study of electrophilic fragments. J. Med. Chem.
2014, 57, 7590-7599.
8.
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.
23.
Lei, J.; Hansen, G.; Nitsche, C.; Klein, C. D.; Zhang,
L.; Hilgenfeld, R. Crystal structure of Zika virus NS2B-NS3
protease in complex with a boronate inhibitor. Science 2016,
353, 503-505.
9.
Nitsche, C.; Behnam, M. A. M.; Steuer, C.; Klein, C.
24.
Chen, W. N.; Nitsche, C.; Pilla, K. B.; Graham, B.;
D. Retro peptide-hybrids as selective inhibitors of the dengue
virus NS2B-NS3 protease. Antiviral Res. 2012, 94, 72-79.
10.
Jensen, C. M.; Gan, C. H.; Gupta, P. K.; Xu, W.-J.; Young, P.
R.; Fairlie, D. P. Potent cationic inhibitors of West Nile virus
NS2B/NS3 protease with serum stability, cell permeability
and antiviral activity. J. Med. Chem. 2008, 51, 5714-5721.
Huber, T.; Klein, C. D.; Otting, G. Sensitive NMR approach
for determining the binding mode of tightly binding ligand
molecules to protein targets. J. Am. Chem. Soc. 2016, 138,
4539-4546.
Stoermer, M. J.; Chappell, K. J.; Liebscher, S.;
25.
Türker, L. Some boric acid esters of glycerol - An ab
initio treatment. Indian J. Chem. 2006, 45A, 1339-1344.
11.
Schüller, 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.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 9
TOC graphic
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
8
ACS Paragon Plus Environment

Page 9 of 9
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
TOC GRAPHIC
1212x319mm (72 x 72 DPI)
ACS Paragon Plus Environment