β-Amino acid substitutions and structure-based CoMFA modeling of hepatitis C virus NS3 protease inhibitors

Article abstract of DOI:10.1016/j.bmc.2008.04.005

In an effort to develop a new type of HCV NS3 peptidomimetic inhibitor, a series of tripeptide inhibitors incorporating a mix of α- and β-amino acids has been synthesized. To understand the structural implications of β-amino acid substitution, the P₁, P₂, and P₃ positions of a potent tripeptide scaffold were scanned and combined with carboxylic acid and acyl sulfonamide C-terminal groups. Inhibition was evaluated and revealed that the structural changes resulted in a loss in potency compared with the α-peptide analogues. However, several compounds exhibited μM potency. Inhibition data were compared with modeled ligand-protein binding poses to understand how changes in ligand structure affected inhibition potency. The P₃ position seemed to be the least sensitive position for β-amino acid substitution. Moreover, the importance of a proper oxyanion hole interaction for good potency was suggested by both inhibition data and molecular modeling. To gain further insight into the structural requirements for potent inhibitors, a three-dimensional quantitative structure-activity relationship (3D-QSAR) model has been constructed using comparative molecular field analysis (CoMFA). The most predictive CoMFA model has q² = 0.48 and r_pred² = 0.68.

Full text of DOI:10.1016/j.bmc.2008.04.005

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 5590–5605
b-Amino acid substitutions and structure-based CoMFA modeling
of hepatitis C virus NS3 protease inhibitors
Johanna Nurbo,a Shane D. Peterson,a Go¨ran Dahl,b U. Helena Danielson,b
a
a,
*
´
Anders Karlen and Anja Sandstro¨m
aDepartment of Medicinal Chemistry, Organic Pharmaceutical Chemistry, Uppsala University, BMC, Box 574,
SE-751 23 Uppsala, Sweden
bDepartment of Biochemistry and Organic Chemistry, Uppsala University, BMC, Box 576, SE-751 23 Uppsala, Sweden
Received 24 August 2007; revised 28 March 2008; accepted 1 April 2008
Available online 6 April 2008
Abstract—In an effort to develop a new type of HCV NS3 peptidomimetic inhibitor, a series of tripeptide inhibitors incorporating a
mix of a- and b-amino acids has been synthesized. To understand the structural implications of b-amino acid substitution, the P1, P2,
and P3 positions of a potent tripeptide scaffold were scanned and combined with carboxylic acid and acyl sulfonamide C-terminal
groups. Inhibition was evaluated and revealed that the structural changes resulted in a loss in potency compared with the a-peptide
analogues. However, several compounds exhibited lM potency. Inhibition data were compared with modeled ligand–protein bind-
ing poses to understand how changes in ligand structure affected inhibition potency. The P3 position seemed to be the least sensitive
position for b-amino acid substitution. Moreover, the importance of a proper oxyanion hole interaction for good potency was sug-
gested by both inhibition data and molecular modeling. To gain further insight into the structural requirements for potent inhib-
itors, a three-dimensional quantitative structure–activity relationship (3D-QSAR) model has been constructed using comparative
molecular field analysis (CoMFA). The most predictive CoMFA model has q2 = 0.48 and rp2red ¼ 0:68.
ꢀ 2008 Elsevier Ltd. All rights reserved.
1. Introduction
of the virally encoded HCV NS3 protein. The NS3 pro-
tein is a bifunctional enzyme with both helicase/NTPase
and protease activities. The NS3 protease, in complex
with the NS4A co-factor, is responsible for proteolytic
processing at four sites of the HCV polyprotein, making
it crucial for viral replication.3 Additionally, recent re-
sults indicate that the NS3 protease is involved in the
cellular mechanisms of viral persistence.4 Thus, inhibi-
tion of the protease may have dual effects: both inhibi-
tion of viral replication as well as restoration of the
host immune response. Indeed, HCV NS3 protease
inhibitors have shown proof-of-concept in clinical trials
in the cases of BILN 2061 (ciluprevir),5 VX-950 (telapre-
vir),6 SCH 503034 (boceprevir),7 and ITMN-1918
(Fig. 1).
Hepatitis C Virus (HCV) is the major cause of chronic
liver disease and has an estimated world prevalence of
2–3%. Consequently, HCV is also the primary indica-
tion for liver transplantations.1 Pegylated interferon-a
and ribavirin constitute standard therapy for people in-
fected with HCV, but the sustained virological response
rate is only roughly 55% and is least effective for the
most prevalent genotype (genotype 1).2 Thus, consider-
ing the seriousness of the impact of HCV infection on
global health, the development of more efficient thera-
pies is extremely important. This is reflected by the rap-
idly increasing number of patents and publications from
both academic and industrial groups devoted to HCV-
related research.
Although ciluprevir has been subjected to clinical test-
ing,5 it has since been removed due to cardiac toxicity.9
Telaprevir and boceprevir are currently in phase II tri-
als10,11 while InterMune is currently conducting phase
1b clinical trials on ITMN-191.12 Ciluprevir and
ITMN-191 are non-covalent product-based inhibi-
tors,13,14 containing a C-terminal carboxylic acid and
carboxylic acid bioisostere (acyl sulfonamide), respec-
Of several possible drug targets for HCV therapy, one
thoroughly studied and promising target is the protease
Keywords: Hepatitis C; HCV; NS3; Protease inhibitor; b-Amino acid;
3D-QSAR; CoMFA; Docking.
*
Corresponding author. Tel.: +46 18 471 4957; fax +46 18 471 4474;
e-mail: Anja.Sandstrom@orgfarm.uu.se
0968-0896/$ - see front matter ꢀ 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.04.005

J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
5591
The insertion of an additional methylene group alters
the electronic conditions of the amide bonds and adds
additional degrees of conformational freedom. More-
over, peptide recognition and resistance to proteolysis
may be affected. Although b-peptides are widely known
to be more resistant to proteolysis,19 studies have shown
that peptides comprising a mix of a- and b-amino acids
may also be more resistant to proteolytic degradation.20
O
H
N
O
N
O
H
N
O
H
N
N
N
O
NH
O
N
HN
O
S
O
N
VX-950 (telaprevir)
N
O
O
NH
O
OH
O
N
O
H
The use of b-amino acids in the development of pepti-
dase/protease inhibitors has previously been successful
in both reducing proteolysis and retaining po-
tency.18,21–24 b-Amino acid substitutions have also been
explored as an approach to obtain more selective serine
protease inhibitors.25 Additionally, it has been possible
to design ligands more resistant to cleavage by exchang-
ing native a-amino acid residues with a b-amino acid
variant.26,27
F
BILN 2061 (ciluprevir)
N
O
O
H
O
N
O
N
N
O
O
O
O
NH
S
O
NH2
NH
O
NH
O
N
O
O
N
H
H
O
SCH 503034 (boceprevir)
ITMN-191
The accessibility of b-amino acids, their potentially en-
hanced resistance to proteolysis, and the fact that
they, to the best of our knowledge, had not been pre-
viously used in HCV protease inhibitor design inspired
us to explore their use in our ongoing HCV research
endeavors. We began the study by selecting a tripep-
tide inhibitor whose corresponding b-amino acids are
commercially available and/or synthetically accessible.
A b-amino acid scan was performed, replacing the ori-
ginal a-amino acid with its b-amino acid counterpart
in the P1, P2, and P3 positions. Inhibitors comprising
both the carboxylic acid and acyl sulfonamide C-ter-
minal groups were prepared and subjected to inhibi-
tion studies. To study how the incorporation of b-
amino acids would affect important protein–ligand
interactions, molecular docking studies were per-
formed. Based on the ligand poses, a Comparative
Molecular Field Analysis28 (CoMFA) model was cal-
culated, also including inhibitors of the HCV NS3
protease previously synthesized by our group.
Figure 1. Structures of the HCV NS3 protease inhibitors that have
entered clinical trials.
tively. Telaprevir and boceprevir are serine trap inhibi-
tors15 containing an electrophilic functional group (a-
ketoamide), capable of forming a reversible covalent
bond with the enzyme.
While some inhibitors of the HCV NS3 protease have
been found using high throughput screening,16 the
majority have been discovered using peptidomimetic de-
sign with the natural peptide substrate as the starting
point. The development of peptidomimetics has
emerged as an attractive approach for overcoming the
limitations inherent to peptides, seeking to improve their
bioavailability and stability. Peptide truncation, the use
of non-natural amino acids, side chain extension, and
cyclizations are peptidomimetic techniques that have
been used to reduce the peptide character of HCV
NS3 protease inhibitors in order to produce more
drug-like and potent compounds, as exemplified by the
compounds subjected to clinical trials (Fig. 1).
2. Results
As seen in Figure 1, all compounds currently in clinical
trials are comprised of a-amino acids. However, back-
bone modifications, such as b-amino acid insertion,
may be useful in reducing some negative attributes asso-
ciated with peptide-like compounds.17 b-Amino acids
differ from their corresponding a-variants in the position
where the side chain is attached to the peptide back-
bone. In the case of a b-amino acid, an extra methylene
group is inserted into the backbone and the side chain is
attached to one of the two methylene carbons, depend-
ing on the type of b-amino acid desired.18 Figure 2
shows a comparison of a-, b2-, and b3-amino acids.
2.1. Chemistry
The b-amino acid-comprising tripeptide compounds in-
cluded in this study were generated by P1–P3 building
block assembly followed by C-terminal modifications.
The b-amino acid-based P1 building blocks 3 and 4 were
prepared by C-alkylation essentially following a proce-
dure by Mertin and coworkers29 outlined in Scheme 1.
The hemiaminal 1-dibenzylamino-cyclopropanol,29 was
prepared from 1-ethoxy-1-trimethylsiloxycyclopro-
pane,30 and was treated with TiCl4 followed by the addi-
tion of the corresponding ketene silyl acetals to give
benzyl-protected 1 and 229 in reasonable yields.
The benzyl-protected building blocks were deprotected
by catalytic hydrogenation with 10% palladium on car-
bon and were isolated as the hydrochloride salts 3 and 4.
The methyl ester 2 was quantitatively deprotected in 3 h,
whereas 1 required 40 h under hydrogen atmosphere to
be fully deprotected.
O
O
N
H
N
H
N
H
O
β2-Alanine
β3-Alanine
α-Alanine
Figure 2. Depiction of a-, b2- and b3-amino acid backbone structures.

5592
J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
BzlO
ylamino)-methylene]-N- methylmethanaminium hexa-
OTBDMS
fluorophosphate N-oxide (HBTU) in the presence of
N,N-diisopropylethylamine (DIEA) to give dipeptide
10. Coupling of the P2 building block 8 with Boc-L-va-
line was conducted under the same conditions to give
11. Also, 9 was allowed to react with commercially
available Boc-L-b-homovaline and yielded 12. Methyl
esters 10–12 were subsequently hydrolyzed using LiOH,
producing carboxylic acids 13–15. The synthesis of
intermediate 13 has been previously described by oth-
ers.31 The dipeptide 13 was coupled with P1 building
blocks 3 or 4 to give methyl ester-protected tripeptides
16 and 17. Dipeptides 14 and 15 were also further cou-
pled with P1 building blocks to afford 18 and 19. Depro-
tection with LiOH of methyl esters 16–19 gave
carboxylic acids 20–23.
O
or
OTBDMS
O
O
a
(Bzl)2N
OH
(Bzl)2N
O
R
1 R=OBzl (55%)
2 R=H (62%)
O
b
HCl x H2N
O
R
3 R=OH (99%)
4 R=H (99%)
The benzene sulfonamide compounds were prepared as
described in Scheme 4. The corresponding carboxylic
acids 21–23 were preactivated with N-[(dimethylamino)-
1H-1,2,3-triazolo-[4,5-b]pyridine-1-yl-methylene]-N-
methylmethanaminium hexafluorophosphate N-oxide
(HATU) and DIEA followed by the addition of benzene
sulfonamide, 4-dimethylaminopyridine (DMAP) and
1,8-diazabicyclo[5.4.0]undec-7-ene (DBU).32
Scheme 1. Reagents: (a) TiCl4, DCM; (b) H2, Pd/C 10%, HCl, MeOH.
The P2 building block 8 (Scheme 2) was prepared start-
ing from commercially available L-b-homohydroxypro-
line hydrochloride according to a method for the
corresponding a-amino acid building block.31 Treatment
with Boc-anhydride in the presence of aqueous NaOH
gave Boc-protected 5 that was further arylated with 4-
chloro-7-methoxy-2-phenylquinoline to give the ether
6. The carboxylic acid was converted to methyl ester
with Cs2CO3 and methyl iodide to afford 7, which was
further Boc-deprotected to give the dihydrochloride 8.
Finally, the a-hydroxy-b-amino acid-containing com-
pound 20 was oxidized using Dess–Martin reagent to
give the a-ketoacid 27 as depicted in Scheme 5.
2.2. Molecular modeling
The target compounds 21–27, containing a mix of a-
and b-amino acids, were generated in a parallel fashion
as described in Schemes 3–5. The assembly of carboxylic
acids 21–23 is described in Scheme 3. The methyl ester 9,
previously described in the literature,31 was coupled to
Boc-L-valine using N-[(1H-benzotriazole-1-yl)-(dimeth-
Docking studies were performed with the inhibitors pre-
sented herein as well as those from our previous work on
non-covalent proline-based inhibitors.33–35 Ligand
structures in their predicted binding conformation were
used to develop CoMFA models. By using default
O
N
OH
OH
O
a
b
O
HCl x HN
O
BocN
O
BocN
OH
OH
OH
5 (96%)
6 (86%)
O
O
N
HCl x N
O
O
c
d
O
BocN
O
HCl x HN
O
O
7 (90%)
8 (99%)
Scheme 2. Reagents: (a) Boc2O, NaOH, 1,4-dioxane, H2O; (b) 4-Chloro-7-methoxy-2-phenylquinoline, KOt-Bu, DMSO; (c) MeI, Cs2CO3, DMF; (d)
HCl/1,4-dioxane.

J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
5593
O
O
O
O
HCl x N
N
O
N
H
OH
O
l
O
l=0
l=1
O
O
a
HCl x HN
N
H
N
m
l
m
O
OR
O
O
O
9 m=0
8 m=1
10 l=0, m=0, R=Me (89%)
13 l=0, m=0, R=H (94%)
11 l=0, m=1, R=Me (72%)
14 l=0, m=1, R=H (97%)
12 l=1, m=0, R=Me (87%)
15 l=1, m=0, R=H (97%)
b
b
b
O
O
HCl x H2N
O
n
R'
N
n=0
n=1, R'=OH
n=1, R'=H
O
a
O
N
H
N
O
l
m
N
H
O
O
n
OR''
R'
O
16 l=0, m=0, n=1, R'=OH, R''=Me (67%)
20 l=0, m=0, n=1, R'=OH, R''=H (97%)
17 l=0, m=0, n=1, R'=H, R''=Me (81%)
21 l=0, m=0, n=1, R'=H, R''=H (98%)
18 l=0, m=1, n=0, R''=Me (65%)
22 l=0, m=1, n=0, R''=H (67%)
b
b
b
b
19 l=1, m=0, n=0, R''=Me (76%)
23 l=1, m=0, n=0, R''=H (96%)
Scheme 3. Reagents: (a) HBTU, DIEA, DMF; (b) LiOH, THF, MeOH, H2O.
O
O
N
N
N
O
O
a
O
21-23
N
a
H
N
O
H
N
O
S
20
O
l
m
O
N
O
O
O
H
n
N
N
OH
O
O
O
H
O
H
O
24 l=0, m=0, n=1 (58%)
25 l=0, m=1, n=0 (60%)
26 l=1, m=0, n=0 (63%)
27 (36%)
Scheme 5. Reagents: (a) Dess–Martin periodinane, DCM.
Scheme 4. Reagents: (a) Benzenesulfonamide, HATU, DIEA, DMAP,
DBU, DMF.
2.3. Biochemical evaluation
CoMFA parameters, it was possible to calculate a 7-
component model with internal (q2) and external
ðr2predÞ predictive abilities of 0.31 and 0.56, respectively,
and a fit (r2) of 0.99. This model was further refined
using our CoMFA parameter optimization methodol-
Target compounds 21–27 were tested in an HCV NS3
protease enzymatic assay comprising the full-length
NS3 protein.37 Structures and inhibition data (expressed
as Ki values) are shown in Table 1.
ogy,36 resulting in
a
4-component model with
q2 = 0.48, rp2red ¼ 0:68, and r2 = 0.97. Settings for the im-
proved model differed from the default settings in steric
and electrostatic cutoff (60 and 15, respectively), trans-
form (Indicator), switch function (No), volume averag-
ing type (Box) and the drop electrostatics setting (No).
3. Discussion
In an effort to further reduce the peptide character of
HCV NS3 protease inhibitors and to provide insight
into the structural requirements for potency, we have

5594
J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
Table 1. Structures and inhibition constants for tripeptide inhibitors of HCV NS3
Scaffold
Compound
R
Ki SD (lM)
a-Amino acid reference compounds
O
N
O
28
29
0.32 0.03a
O
OH
O
O
S
H
N
N
O
O
0.055 0.007a
R
N
H
N
H
O
O
O
b-Amino acid-comprising compounds
O
O
21
24
27
18
14
5
7
N
OH
O
O
O
S
O
N
H
H
N
N
O
R
N
H
O
O
O
O
1.0 0.6
O
OH
O
O
22
25
61 13
N
OH
O
O
O
S
O
O
N
N
H
1.8 0.5
O
N
N
H
O
R
O
H
O
O
23
26
17
3
N
OH
O
O
O
H
N
O
S
H
N
N
H
1.2 0.4
R
N
O
O
O
O
35
a Inhibition data reported earlier by Ro¨nn et al.
explored the possibility of substituting the more com-
monly used a-amino acids with their b3-counterparts
using a previously developed tripeptide inhibitor as a
reference. Thus, we have substituted the a-amino acid
in each position with the corresponding b3-amino acid.
Inhibition analysis, docking, and 3D-QSAR studies
have been performed to understand the effect of this
structural modification on potency.
found to be potent inhibitors of the HCV NS3 protease
and were selected partly due to the building block acces-
sibility. Both Boc-L-b-homovaline and L-b-homo-
hydroxyproline were obtained commercially, while (1-
aminocyclopropyl)-acetic acid methyl ester (4) was ob-
tained by synthesis. Although previous studies by our
research group indicated that the acyl sulfonamide moi-
ety in the C-terminal position is beneficial for potency
and selectivity,35,38–40 inhibitor scaffolds with both car-
boxylic acid and acyl sulfonamide C-terminal groups
were studied. Insertion of a b-amino acid in the inhibitor
The tripeptide scaffolds, 2831,35 and 2935 (Table 1), se-
lected for b-amino acid modification, were previously

J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
5595
backbone will result in longer, more flexible inhibitors
with potentially different structure–activity
relationships.
in the C-terminal carboxylic acid series, rendering inhib-
itors 23 and 21 with Ki values of 17 and 18 lM, respec-
tively. Use of a b-amino acid in the P2 position as in
compound 22 resulted in a significant loss of potency,
yielding a Ki value of 61 lM.
According to observations from docking, reference
inhibitors 28 and 29 form hydrogen-bonding interac-
tions as depicted in Figure 3. These interactions have
been observed between the natural cleavage product
and the protease in the protein crystal structure41 and
are consistent with other published structure–activity
relationship studies. Thus, the product peptide in the
crystal structure forms an extended anti-parallel b-
strand with the protease backbone, allowing the NH
and carbonyl moieties of the P3 residue to form hydro-
gen bonds with the A157 residue in the protease domain.
The hydrogen bonds between the C-terminal carboxylic
acid and the backbone amides of G137 and S139 have
also been noted in the crystal structure. G137 and
S139 form what is commonly referred to as the oxyanion
hole42,43 and is an important binding site for the oxyan-
ion in natural substrate cleavage for serine proteases. In
contrast to the natural cleavage product in the crystal
structure, our docking model indicates that the NH of
the inhibitors P1 residue tends to form hydrogen-bond-
ing interactions with the R155 backbone rather than
the K136 side chain (Fig. 3). This hydrogen bond is a
common feature in other published crystal structures
of protease inhibitors.44,45
b-Amino acid substitutions in the P3 position provided
lM inhibitors using both carboxylic acid and acyl sul-
fonamide C-terminal groups. Docking studies were con-
ducted to understand the experimental inhibition data
and showed a close similarity in interactions seen in
the corresponding reference inhibitors (Fig. 3). Substitu-
tion in the P3 position did not generally prevent the
inhibitors from forming a similar hydrogen bond net-
work with A157 observed in the reference inhibitor
but may have introduced some ligand strain. Compared
with the P3 side chain of the reference inhibitors, the va-
line side chain of both P3 b-amino acid inhibitors was
found in closer proximity to the C159 side chain and
may be a sign of steric clash. Since the compounds with
a b-amino acid in the P3 position provided lM inhibi-
tors using both carboxylic acid and acyl sulfonamide
C-terminal groups, this position shows potential for fur-
ther structure optimization and studies of stability,
selectivity, and pharmacokinetic properties.
According to docking studies, insertion of a b-amino
acid in the P2 position resulted in the disruption of the
P1 NH to R155 carbonyl hydrogen bond for both the
acyl sulfonamide and the carboxylic acid compounds
22 and 25. In the case of 22, where a carboxylic acid
C-terminal group was used, interactions with the oxyan-
ion hole were also lost. This was likely not the case for
25, where the acyl sulfonamide C-terminal was used.
Docking experiments indicated hydrogen bonds be-
tween the sulfonyl oxygens and the Q41 and S139 side
chains, which may have helped position the acyl sulfon-
amide carbonyl oxygen in the oxyanion hole. Although
25 may interact favorably with residues in the oxyanion
hole, docking suggests disruption of hydrogen bonding
with A157 to a greater extent than was observed in the
corresponding carboxylic acid inhibitor 22. These obser-
vations may explain the loss in potency for carboxylic
acid 22 and show why acyl sulfonamide 25 remained
fairly potent.
The P1, P2, and P3 positions were explored by substituting
b3-amino acids in place of the original a-amino acids. All
of the b-amino acid-comprising compounds 21–26 (Table
1) were found to be less potent than their a-amino acid
counterpart (28 or 29) but lM potency was still achieved.
As in previous studies, inhibitors with an acyl sulfon-
amide C-terminal group were found to be of higher po-
tency than their corresponding carboxylic acid
analogues (compare 24, 25, 26 with 21, 22, and 23). In
the acyl sulfonamide series, b-amino acid substitution in
the P3 and P2 positions were most successful, rendering
inhibitors 25 and 26 with Ki values of 1.8 lM and
1.2 lM, respectively. The corresponding compound with
b-amino acid in P1 position (24) had a Ki value of 14 lM.
In contrast to the acyl sulfonamide series, b-amino acid
substitution in the P3 and P1 position was best tolerated
Molecular docking indicated that insertion of a b-amino
acid in the P1 position did not preclude the formation of
hydrogen bonds between the terminal carbonyl and the
G137 and S139 NHs. Although 21 adopted a conforma-
tion permitting hydrogen bonding with G137 and S139
with less strain than was seen in 24, its hydrogen bond
to R155 may be disrupted to a greater extent. Inhibition
measurements showed that 24 exhibited the greatest loss
in potency. The corresponding reference inhibitor 29 has
a Ki of 0.055 lM, while 24 is 250 times less potent
(14 lM). This dramatic loss of potency may be attrib-
uted to the size of the bulky acyl sulfonamide and the
resulting ligand strain necessary to form important
interactions in the oxyanion hole.
HN
NH2
O
R155
NH
OH
S139
N
O
NH
O
O
HN
O
S138
G137
H
N
N
O
HO
O
NH
O
N
O
O
OH
O
H
HN
O
K136
NH
O
NH
A157
H2N
To sum up, a general loss in potency was seen for inhib-
itors 21–26. Docking suggests that the 20- to 60-fold loss
Figure 3. Protein–ligand hydrogen-bonding interactions for reference
compound 28, as proposed by docking.

5596
J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
in potency seen for inhibitors 21, 23, 25, and 26 is most
likely due to unfavorable conditions for the formation
of hydrogen bonds between the inhibitor backbone
and A157 or R155. The more dramatic loss in potency
of compounds 22 and 24 may be explained by poor oxy-
anion hole interactions.
lowed for interaction with the oxyanion hole without
the introduction of any notable ligand strain, allowing
for better interactions with R155 and A157.
Based on the results obtained from docking, it was pos-
sible to develop 3D-QSAR models using CoMFA. In
addition to the compounds presented in this work, the
CoMFA model included 47 previously published non-
covalent proline-based compounds comprising various
C-terminal groups and different P1 side chains (struc-
tures available in the Supplementary data). Due to the
structural similarity in P2 and P3, we have assumed that
the inhibitors bind in a similar fashion. Docking was
used to predict the biologically active conformation
and also suggested a similar binding mode for this set
of compounds. Although it was possible to derive an
acceptable CoMFA model (q2 = 0.31, rp2red ¼ 0:56) using
only default settings, predictive performance was en-
hanced by adjusting settings (q2 = 0.48, rp2red ¼ 0:68).
Figure 5 shows a comparison of experimental versus
predicted activities for compounds included in the exter-
nal validation set, where the general improvement in
predictive performance of the improved CoMFA model
can be seen. The high degree of predictive accuracy of
the CoMFA model suggests a clear correlation between
ligand structure and measured inhibitory effects and
supports information obtained from docking.
In our past publications we have speculated that com-
pounds with an electrophilic C-terminal a-ketoacid
group and a neighboring cyclopropane P1 side chain
may not function as a covalent serine-trap inhibitor
due to the steric hindrance introduced by the bulky
a,a-disubstituted side chain.38,39,46 Instead, the high
potencies observed for this type of inhibitor were thus
expected to be due to the carboxylic acid (positioned
as in a b-amino acid), similar to a non-covalent prod-
uct-based inhibitor. Thus, we were surprised to see the
loss in potency for compound 21 comprising a b-amino
acid with a free carboxylic acid in the C-terminus. Dock-
ing indicated that hydrogen bonding to the oxyanion
hole in the P1b-amino acid inhibitors required some
backbone ligand strain. This inspired the modification
of the P1 cyclopropane-based b-amino acid in com-
pound 21 to its a-ketoacid analogue 27, supposed to
be a non-covalent inhibitor as discussed above. Inhibi-
tion studies showed 27 (Ki = 1.0 lM) to be roughly 20-
fold more potent than 21 (Ki = 18 lM). This may indi-
cate that it is the keto function adjacent to the P1-side
chain and not the terminal carbonyl that is important
for the non-covalent oxyanion hole interactions. The
proposed binding pose of 27 is shown in Figure 4, where
it can be seen that the a-keto moiety is positioned appro-
priately for efficient interaction with the oxyanion hole
as in a non-covalent product-based inhibitor. This al-
In addition to its value as a predictive tool, CoMFA
fields may be visualized and overlaid with the protein
to show regions correlated with changes in activity.
CoMFA models are developed without knowledge of
the protein structure and correspondence of contour
surfaces with the protein structure supports structure–
Figure 4. Depiction of the hydrogen-bonding network observed in
docking studies of 27. The P3 valine residue is seen lying anti-
parallel to the protease, forming hydrogen bonds with A157. The P1
NH is located in close proximity to the R155 carbonyl while the
a-keto C-terminal group is seen hydrogen bonding with the NHs of
G137 and S139. The terminal carboxylic acid moiety may partic-
ipate in hydrogen-bonding interactions with G137. Hydrogen bond
Figure 5. Experimental versus predicted pKi values for test set
compounds showing a comparison of the predictive quality of the
default and improved CoMFA models. The placement of points
relative to the line gives an indication of the predictive accuracy of each
model. The experimental and predicted pKi values are available in the
supplementary material.
˚
distances varied between 2.8 A and 3.4 A.
˚

J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
5597
activity relationships. Figure 6 shows the steric and elec-
trostatic contour maps overlaid with the protein surface.
Some of the more notable features include the red sur-
face between the S139 and Q41 side chains, the blue sur-
face near the R155 backbone carbonyl and the green
surfaces near the M485 and K136 side chains. Greater
potency is correlated with the presence of electron with-
drawing groups near red surfaces and in this case corre-
sponds with the acyl sulfonamide C-terminal group in
many of the inhibitors of the training set. In the docking
poses of nearly all C-terminal acyl sulfonamides, one of
the sulfone oxygens is located in or near this red surface.
The model expects the presence of electropositive groups
in the ligand near blue regions. In this case, the blue re-
gion near the R155 backbone carbonyl and the P1 back-
bone nitrogen may indicate the importance of hydrogen
bonding between these groups. The green surfaces near
M485 and K136 indicate the importance of hydrophobic
interactions in these areas.
inhibitor backbone results in greater conformational
flexibility. Upon inhibitor binding, this would result in
a greater loss of entropy and may have also contributed
to the observed loss of potency. Despite these detrimen-
tal effects on potency for this particular series, several
inhibitors were still of lM potency. This study has
shown that it is possible to produce inhibitors of modest
potency by inserting b-amino acids into the peptide
backbone and has shown that optimization of the b-
amino acid backbone will likely enhance potency. Fur-
ther ligand optimization will be required to obtain more
potent inhibitors.
4. Conclusion
A series of tripeptide protease inhibitors comprising b-
amino acids have been synthesized and evaluated in a
full-length HCV NS3 assay. Although experimental data
showed a loss in potency, inhibitors with lM potency
were obtained. This study has provided important infor-
mation regarding structural requirements in the P1, P2,
and P3 positions for inhibitors comprising carboxylic
acid and acyl sulfonamide C-terminal groups. First,
the importance of good oxyanion hole interactions has
been confirmed. Also, the P3-position seems to be least
affected by b-amino acid substitution, something that
can be exploited in future inhibitor optimization. Molec-
ular docking has provided a qualitative understanding
of which protein–ligand interactions that may be af-
fected at each position. A predictive CoMFA model
has been derived using docked ligand poses, allowing
the prediction of potency for future compounds and
suggesting a correlation between changes in ligand
structure and observed potency.
Unfortunately, inhibition studies showed that introduc-
tion of b-amino acids resulted in an overall loss in po-
tency. Docking indicated that, depending on the
insertion position of the b-amino acid, some electro-
static interactions may have been affected. In such cases,
ligand strain may have prevented the inhibitor from
forming the hydrogen bonds seen in the relevant refer-
ence compound. This would result in unsatisfied hydro-
gen-bond partners being buried within the protein,
which is known to be detrimental to potency.47 Addi-
tionally, insertion of an extra methylene group to the
5. Experimental
5.1. Chemistry
Reagents and solvents were obtained commercially and
used without further purification. TLC was performed
using aluminium sheets precoated with silica gel 60
F254 (Merck) with visualization of spots using UV-detec-
tion and/or ninhydrin treatment followed by heating.
Column chromatography was performed using silica
gel 60 (40–63 lm) (Merck). LC-MS was performed on
a Gilson-Finnigan AQA system in ESI mode using a
Chromolith SpeedROD RP-18e 4.6 · 50 mm column
(Merck) and a CH3CN/H2O linear gradient with
0.05% HCOOH. Preparative RP-HPLC was performed
on an ACE-Phenyl, 5 lm, (21.2 · 150 mm) column or
a Zorbax SB-C8, 5 lm, (21.2 · 150 mm) column using
a CH3CN/H2O linear gradient with 0.1% TFA. Selected
fractions were pooled and lyophilized. The purity of the
inhibitors was determined by analytical RP-HPLC in
the following systems (UV detection at 220 nm): column
1 (ACE 5 C18, 50 · 4.6 mm, H2O/MeCN gradient with
25 mM NH4OAc, pH 6.3) and column 2 (Thermo
Hypersil C4, 50 · 4.6 mm, 5 lm, H2O/MeCN gradient
with 0.1% TFA). NMR spectra were recorded on a Var-
ian Mercury plus spectrometer (1H at 399.8 MHz, 13C at
Figure 6. Steric and electrostatic contour maps derived from our
improved CoMFA model and overlaid with the protein surface and 27.
Electrostatic maps are shown in red and blue; red regions are those in
which the presence of electronegative groups are associated with
improved potency and blue contour maps indicate the opposite. Steric
maps are shown in green and yellow; the green areas are those, where
the presence of steric bulk is associated with greater potency and the
yellow areas are those, where steric bulk has a negative impact on
potency.

5598
J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
100.5 MHz) at ambient temperature unless otherwise
stated. Chemical shifts are reported as d values in ppm
and are indirectly referenced to TMS via the solvent sig-
nal (1H: CHCl3 d 7.26, CHD2OD d 3.31, DMSO-d5 d
2.50; 13C: CDCl3 d 77.16, CD3OD d 49.00, DMSO-d6
d 39.50). Exact molecular masses were determined on
a Micromass Q-Tof2 mass spectrometer equipped with
an electrospray ion source. Elemental analyses were per-
formed by Analytische Laboratorien, Lindlar,
Germany.
was stirred under H2 atmosphere for 40 h. The catalyst
was filtered off and the solvent was removed by rotary
evaporation to give 3 (0.15 g, 99%) as colorless syrup.
1H NMR (DMSO-d6): d 8.34 (br s, 3H), 6.35 (d,
J = 5.3 Hz, 1H), 4.01 (d, J = 5.3 Hz, 1H), 3.33 (s, 3H),
1.05–0.95 (m, 2H), 0.95–0.87 (m, 1H), 0.87–0.76 (m,
1H). 13C NMR (DMSO-d6): d 171.1, 71.6, 52.1, 36.1,
8.8, 7.8. Anal. Calcd for C6H12ClNO: C, 39.68; H,
6.66; N, 7.71. Found: C, 39.42; H, 6.68; N, 7.71.
5.1.4. (1-Aminocyclopropyl)-acetic acid methyl ester
hydrochloride (4). To a solution of 2 (0.49 g, 1.6 mmol)
in MeOH (30 mL) was added Pd/C 10% (49 mg). The
reaction mixture was stirred under H2 atmosphere for
3 h. The catalyst was filtered off and HCl in EtOH was
added to pH 3. The solvent was removed by rotary
evaporation to give 4 (0.26 g, 99%) as a colorless syrup.
1H NMR (CDCl3): d 8.70 (br s, 3H), 3.79 (s, 3H), 2.77
(s, 2H), 1.48–1.41 (m, 2H), 0.81–0.74 (m, 2H). 13C
NMR (CDCl3): d 171.3, 52.6, 39.3, 31.7, 10.6. Anal.
Calcd for C6H12ClNO2: C, 43.51; H, 7.30; N, 8.46.
Found: C, 43.32; H, 7.20; N, 8.42.
5.1.1.
Benzyloxy-(1-dibenzylamino-cyclopropyl)-acetic
acid methyl ester (1). A solution of 1-dibenzylamino-
cyclopropanol (1.0 g, 4.0 mmol) in dry DCM (60 mL)
was stirred at ꢀ78ꢁC under N2 atmosphere. TiCl4
(0.48 mL, 4.4 mmol) was slowly added and the reaction
mixture was stirred for 30 min. (Z)-2-Benzyloxy-1-meth-
oxy-1-(tert-butyldimethylsiloxy)ethene48
(ꢁ8 mmol),
prepared from benzyloxyacetic acid methyl ester and
used as crude product, was added dropwise. After 1 h
the reaction mixture was poured into saturated aqueous
NaHCO3 (50 mL) and extracted three times with dieth-
ylether (30 mL). The organic layers were dried with
MgSO4 and concentrated by rotary evaporation. The
resulting product was purified by column chromatogra-
phy (gradient elution, i-hexane/EtOAc 12:1 followed by
i-hexane/EtOAc 9:1) to give 1 (0.91 g, 55%) as a color-
5.1.5. (2S,4R)-2-Carboxymethyl-4-hydroxy-pyrrolidine-
1-carboxylic acid tert-butyl ester (5). A solution of L-b-
homohydroxyproline hydrochloride (0.36 g, 2.0 mmol)
in 1,4-dioxane (4.0 mL) was treated with 1.0 M aqueous
NaOH (4.0 mL). Di-tert-butyldicarbonate (0.65 g,
3.0 mmol) in 1,4-dioxane (4.0 mL) was added and the
reaction mixture was stirred at room temperature over-
night. The dioxane was evaporated and the aqueous
phase was acidified to pH 2 with 10% aqueous KHSO4.
The acidic solution was extracted with EtOAc
(3 · 10 mL), dried with MgSO4 and concentrated by ro-
tary evaporation to give 5 (0.47 g, 96%) as a colorless
1
less oil. H NMR (CDCl3): d 7.42–7.30 (m, 5H), 7.23–
7.13 (m, 10H), 4.68 (d, J = 11.6 Hz, 1H), 4.38 (d,
J = 11.6 Hz, 1H), 4.33 (s, 1H), 3.86 (d, J = 13.7 Hz,
1H), 3.73 (s, 3H), 3.73 (d, J = 13.7 Hz, 1H), 0.90–0.84
(m, 1H), 0.67–0.61 (m, 1H), 0.58–0.51 (m, 1H), 0.46–
0.39 (m, 1H). 13C NMR (CDCl3): d 172.2, 140.2,
137.3, 129.0, 128.4, 128.3, 127.9, 127.8, 126.6, 79.5,
72.2, 56.9, 51.7, 45.6, 15.0, 12.3. HRMS calcd for
C27H30NO3 (M+H+) 416.2226, found: 416.2222.
1
syrup which was used without further purification. H
NMR (DMSO-d6, 70ꢁ C): d 11.89 (br s, 1H), 4.69 (br
s, 1H), 4.22–4.17 (m, 1H), 4.08–3.99 (m, 1H), 3.30–
3.25 (m, 2H), 2.90–2.72 (m, 1H), 2.28 (dd, J = 9.1,
15.2 Hz, 1H), 2.05–1.98 (m, 1H), 1.85–1.76 (m, 1H),
1.41 (s, 9H). HRMS calcd for C11H20NO5 (M+H+)
246.1341, found: 246.1343.
5.1.2. (1-Dibenzylamino-cyclopropyl)-acetic acid methyl
ester (2)29. A solution of 1-dibenzylamino-cyclopropa-
nol (1.0 g, 4.0 mmol) in dry DCM (60 mL) was stirred
at ꢀ78 ꢁC under N2 atmosphere. TiCl4 (0.48 mL,
4.4 mmol) was slowly added and the reaction mixture
was stirred for 30 min. 1-(tert-Butyl-dimethylsilyloxy)-
1-methoxyethene (1.8 mL, 8.0 mmol) was added drop-
wise. After 1 h the reaction mixture was poured into sat-
urated aqueous NaHCO3 (50 mL) and extracted three
times with diethylether (30 mL). The organic layers were
dried with MgSO4 and concentrated by rotary evapora-
tion. The resulting product was purified by column
chromatography (gradient elution, i-hexane/EtOAc 9:1
followed by i-hexane/EtOAc 4:1) to give 2 (0.77 g,
5.1.6. (2S,4R)-2-Carboxymethyl-4-(7-methoxy-2-phenyl-
quinolin-4-yloxy)-pyrrolidine-1-carboxylic acid tert-butyl
ester (6). To a solution of 5 (0.20 g, 0.82 mmol) in dry
DMSO (3.0 mL) was added KOtBu (0.23 g, 2.1 mmol)
in portions. The reaction mixture was stirred under N2
atmosphere for 2.5 h followed by the addition of a solu-
tion of 4-chloro-7-methoxy-2-phenylquinoline (0.24 g,
0.90 mmol) in dry DMSO (4.0 mL). The reaction mix-
ture was stirred for additional 22 h and then cooled to
0 ꢁC. Thereafter cold H2O (25 mL) was added and the
suspension was washed with diethyl ether (5 · 20 mL),
acidified to pH 4.7 using 1.0 M aqueous HCl, filtered
and dried to give 6 (0.34 g, 86%) as beige solid. 1H
NMR (DMSO-d6, 70 ꢁC): d 12.0 (br s, 1H), 8.28–8.24
(m, 2H), 7.94 (d, J = 9.1 Hz, 1H), 7.57–7.47 (m, 3H),
7.44 (s, 1H), 7.39 (d, J = 2.6 Hz, 1H), 7.16 (dd, J = 2.6,
9.1 Hz, 1H), 5.54–5.49 (m, 1H), 4.28–4.20 (m, 1H),
3.94 (s, 3H), 3.81 (dm, J = 12.5 Hz, 1H), 3.66 (dd,
J = 4.1, 12.5 Hz, 1H), 2.85 (dm, J = 15.3 Hz, 1H),
2.58–2.50 (m, 2H), 2.32–2.23 (m, 1H), 1.38 (s, 9H).
1
62%) as a colorless oil. H NMR (CDCl3): d 7.16–7.26
(m, 10H), 3.76 (s, 4H), 3.70 (s, 3H), 2.61 (s, 2H), 0.57–
0.53 (m, 2H) 0.52–0.48 (m, 2H). 13C NMR (CDCl3): d
173.2, 140.1, 128.9, 128.0, 126.7, 56.4, 51.6, 41.2, 36.3,
15.1. MS (M+H+) 310.0.
5.1.3. (1-Amino-cyclopropyl)-hydroxy-acetic acid methyl
ester hydrochloride (3). To a solution of 1 (0.35 g,
0.84 mmol) in MeOH (10 mL) was added 4 M HCl in
1,4-dioxane (1.5 mL). Coevaporation with MeOH gave
a white solid. To the residual were added MeOH
(25 mL) and Pd/C 10% (70 mg). The reaction mixture

J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
5599
HRMS calcd for C27H31N2O6 (M+H+) 479.2182, found:
479.2173.
1H), 5.52–5.48 (m, 1H), 4.69 (dd, J = 7.6, 10.1 Hz,
1H), 4.65 (dm, J = 11.5 Hz, 1H), 4.08–4.01 (m, 2H),
3.93 (s, 3H), 3.74 (s, 3H), 2.78 (ddm, J = 7.5, 14.2 Hz,
1H), 2.36 (ddd, J = 4.3, 10.1, 14.2 Hz, 1H), 1.99 (dh,
J = 6.7, 9.0 Hz, 1H), 1.20 (s, 9H), 1.01 (d, J = 6.7 Hz,
3H), 0.97 (d, J = 6.7 Hz, 3H). 13C NMR (CD3OD):
174.2, 173.5, 163.1, 161.9, 161.4, 158.0, 157.9, 152.3,
141.5, 130.5, 129.8, 129.1, 124.4, 119.3, 116.4, 107.4,
100.1, 80.3, 78.1, 59.5, 59.4, 56.0, 54.2, 52.8, 35.9, 31.7,
28.5, 19.4, 19.1. MS (M+H+) 578.3.
5.1.7. (2S,4R)-2-Methoxycarbonylmethyl-4-(7-methoxy-
2-phenyl-quinolin-4-yloxy)-pyrrolidine-1-carboxylic acid
tert-butyl ester (7). To a solution of 6 (0.25 g 0.52 mmol)
in dry DMF (15 mL) was added Cs2CO3 (0.20 g,
0.63 mmol). The reaction mixture was stirred under N2
atmosphere for 10 min followed by the addition of
MeI (0.049 mL, 0.78 mmol). After 6 h the reaction mix-
ture was diluted with EtOAc (30 mL), washed with 5%
aqueous NaHCO3 (2 · 15 mL), H2O (15 mL), and brine
(15 mL). The organic layer was dried with MgSO4 and
concentrated by rotary evaporation. The resulting prod-
uct was purified by column chromatography (i-hexane/
5.1.10. Compound 11. To a suspension of 8 (0.13 g,
0.28 mmol) in dry DMF (4.0 mL) were added Boc-L-va-
line (0.12 g, 0.55 mmol), HBTU (0.13 g, 0.33 mmol), and
DIEA (0.19 mL, 1.1 mmol). The resulting solution was
stirred at room temperature for 5 h, diluted with EtOAc
(30 mL), and washed with 5% aqueous NaH-
CO3(10 mL), 35 mM aqueous NaHSO4 (2 · 10 mL),
H2O (10 mL), and brine (10 mL). The organic layer
was dried with MgSO4 and concentrated by rotary evap-
oration. The resulting product was purified by column
chromatography (gradient elution, i-hexane/EtOAc 2:1
followed by i-hexane/EtOAc 1:1) to give 11 (0.12 g,
1
EtOAc 2:1) to give 7 (0.23 g, 90%) as white foam. H
NMR (DMSO-d6, 70 ꢁC): d 8.27–8.24 (m, 2H), 7.93
(d, J = 9.1 Hz, 1H), 7.57–7.46 (m, 3H), 7.43 (s, 1H),
7.39 (d, J = 2.5 Hz, 1H), 7.15 (dd, J = 2.5, 9.1 Hz, 1H),
5.53–5.48 (m, 1H), 4.31–4.23 (m, 1H), 3.94 (s, 3H),
3.81 (dm, J = 12.2 Hz, 1H), 3.66 (dd, J = 3.8, 12.2 Hz,
1H), 3.63 (s, 3H), 2.89 (dd, J = 4.1, 15.3 Hz, 1H), 2.65
(dd, J = 8.3, 15.3 Hz, 1H), 2.57–2.49 (m, 1H), 2.30–
2.21 (m, 1H), 1.37 (s, 9H). HRMS calcd for
C28H33N2O6 (M+H+) 493.2339, found: 493.2349.
1
72%) as white foam. H NMR (CD3OD): d (ꢁ5:1 mix-
ture of rotamers, major rotamer reported) 8.08 (d,
J = 9.2 Hz, 1H), 8.06–8.00 (m, 2H), 7.57–7.46 (m, 3H),
7.38 (d, J = 2.5 Hz, 1H), 7.19 (s, 1H), 7.09 (dd, J = 2.5,
9.2 Hz, 1H), 5.45–5.40 (m, 1H), 4.56 (dm, J = 12.2 Hz,
1H), 4.56–4.50 (m, 1H), 4.04 (d, J = 8.8 Hz, 1H), 3.96
(dd, J = 3.7, 12.2 Hz, 1H), 3.94 (s, 3H), 3.66 (s, 3H),
2.99 (dd, J = 3.7, 16.2 Hz, 1H), 2.80 (dd, J = 7.5,
16.2 Hz, 1H), 2.67 (ddm, J = 7.6, 14.2 Hz, 1H), 2.36
(ddd, J = 4.5, 9.2, 14.2 Hz, 1H), 1.97 (dh, J = 6.8,
8.8 Hz, 1H), 1.21 (s, 9H), 0.95 (d, J = 6.8 Hz, 3H),
0.93 (d, J = 6.8 Hz, 3H). 13C NMR (CD3OD): d (ꢁ5:1
mixture of rotamers, major rotamer reported) 174.0,
173.2, 163.2, 162.2, 161.4, 157.9, 152.2, 141.5, 130.5,
129.8, 129.1, 124.5, 119.3, 116.5, 107.3, 100.2, 80.3,
77.8, 59.9, 56.0, 54.6, 54.3, 52.1, 37.6, 37.3, 31.8, 28.5,
19.6, 18.9. HRMS calcd for C33H42N3O7 (M+H+)
592.3023, found: 592.3035.
5.1.8. [(2S,4R)-4-(7-Methoxy-2-phenyl-quinolin-4-yloxy)-
pyrrolidin-2-yl]-acetic acid methyl ester dihydrochloride
(8). To a solution of 7 (0.21 g, 0.43 mmol) in anhydrous
1,4-dioxane (1.5 mL) was added 4.0 M HCl in 1,4-diox-
ane (2.2 mL). The reaction mixture was stirred at room
temperature for 8 h. Coevaporation with MeOH gave 8
(0.20 g, 99%) as yellow solid. 1H NMR (CD3OD): d 8.47
(d, J = 9.3 Hz, 1H), 8.13–8.07 (m, 2H), 7.80–7.70 (m,
3H), 7.67 (s, 1H), 7.61 (d, J = 2.3 Hz, 1H), 7.50 (dd,
J = 2.3, 9.3 Hz, 1H), 6.00–5.95 (m, 1H), 4.43–4.32 (m,
1H), 4.08 (s, 3H), 3.96 (dm, J = 13.1 Hz, 1H), 3.88
(dm, J = 13.1 Hz, 1H), 3.78 (s, 3H), 3.09 (dd, J = 3.8,
18.0 Hz, 1H), 3.00 (dd, J = 9.4, 18.0 Hz, 1H), 2.80
(ddm, J = 6.0, 14.9 Hz, 1H), 2.44–2.32 (m, 1H). 13C
NMR (CD3OD): d 172.3, 167.4, 166.7, 158.4, 143.7,
134.0, 133.2, 130.8, 130.1, 126.5, 122.1, 116.0, 102.3,
100.6, 80.7, 57.1, 56.7, 52.8, 51.8, 37.6, 36.4. HRMS
calcd for C23H25N2O4 (M+H+) 393.1814, found:
393.1811.
5.1.11. Compound 12. To a suspension of (2S,4R)-4-(7-
Methoxy-2-phenyl-quinolin-4-yloxy)-pyrrolidin-2-car-
boxylic acid methyl ester dihydrochloride (9) (0.13 g,
0.28 mmol) in dry DMF (4.0 mL) were added Boc-L-b-
homovaline (0.13 g, 0.55 mmol), HBTU (0.13 g,
0.33 mmol) and DIEA (0.19 mL, 1.1 mmol). The result-
ing solution was stirred at room temperature for 3 h, di-
luted with EtOAc (30 mL), and washed with 5%
aqueous NaHCO3 (10 mL), 35 mM aqueous NaHSO4
(2 · 10 mL), H2O (10 mL), and brine (10 mL). The or-
ganic layer was dried with MgSO4 and concentrated
by rotary evaporation. The resulting product was puri-
fied by column chromatography (gradient elution, i-hex-
ane/EtOAc 2:1 followed by i-hexane/EtOAc 1:1) to give
5.1.9. Compound 1031. To a suspension of (2S,4R)-4-(7-
Methoxy-2-phenyl-quinolin-4-yloxy)-pyrrolidin-2-car-
boxylic acid methyl ester dihydrochloride (9) (0.13 g,
0.28 mmol) in dry DMF (4.0 mL) were added Boc-L-va-
line (0.12 g, 0.55 mmol), HBTU (0.13 g, 0.33 mmol), and
DIEA (0.19 mL, 1.1 mmol). The resulting solution was
stirred at room temperature for 5 h, diluted with EtOAc
(30 mL), and washed with 5% aqueous NaHCO3
(10 mL), 35 mM aqueous NaHSO4 (2 · 10 mL), H2O
(10 mL), and brine (10 mL). The organic layer was dried
with MgSO4 and concentrated by rotary evaporation.
The resulting product was purified by column chroma-
tography (gradient elution, i-hexane/EtOAc 2:1 followed
by i-hexane/EtOAc 1:1) to give 10 (0.14 g, 89%) as white
1
12 (0.14 g, 87%) as white foam. H NMR (CD3OD): d
(ꢁ4:1 mixture of rotamers, major rotamer reported)
8.10–8.00 (m, 3H), 7.56–7.46 (m, 3H), 7.39 (d,
J = 2.5 Hz, 1H), 7.23 (s, 1H), 7.14 (dd, J = 2.5, 9.2 Hz,
1H), 5.59–5.53 (m, 1H), 4.67–4.62 (m, 1H), 4.14 (dd,
J = 3.7, 11.8 Hz, 1H), 4.08 (dm, J = 11.8 Hz, 1H), 3.94
(s, 3H), 3.80–3.75 (m, 1H), 3.75 (s, 3H), 2.75 (ddm,
1
foam. H NMR (CD3OD): d 8.08 (d, J = 9.1 Hz, 1H),
8.05–8.01 (m, 2H), 7.56–7.46 (m, 3H), 7.37 (d,
J = 2.5 Hz, 1H), 7.18 (s, 1H), 7.08 (dd, J = 2.5, 9.1 Hz,

5600
J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
J = 8.1, 14.0 Hz, 1H), 2.53–2.48 (m, 2H), 2.40 (ddd,
J = 4.7, 9.0, 14.0 Hz, 1H), 1.83–1.70 (m, 1H), 1.40 (s,
9H), 0.86 (d, J = 6.7 Hz, 3H), 0.80 (d, J = 6.7 Hz, 3H).
13C NMR (CD3OD): d (ꢁ4:1 mixture of rotamers, ma-
jor rotamer reported) 173.9, 172.6, 163.2, 162.0, 151.3,
157.9, 152.3, 141.3, 130.6, 129.8, 129.0, 124.2, 119.5,
116.4, 107.5, 100.0, 79.8, 78.0, 59.3, 56.0, 54.3, 54.0,
52.9, 38.3, 25.8, 32.9, 28.8, 19.9, 18.2. HRMS calcd for
C33H42N3O7 (M+H+) 592.3023, found: 592.3019.
5.1.14. Compound 15. To a solution of 12 (0.13 g,
0.23 mmol) in THF (10 mL) and MeOH (1.5 mL) was
added a solution of LiOH (0.055 g, 2.3 mmol) in H2O
(4.0 mL). The reaction mixture was stirred at room tem-
perature for 2.5 h and neutralized with 1.0 M aqueous
HCl. The organic solvents were removed by rotary evap-
oration and the remaining aqueous phase was acidified
with 1.0 M aqueous HCl to pH 3 and extracted with
EtOAc (3 · 20 mL). The organic layers were washed
with brine (20 mL), dried with MgSO4, and concen-
trated by rotary evaporation to give 15 (0.12 g, 97%)
5.1.12. Compound 1331. To a solution of 10 (0.13 g,
0.23 mmol) in THF (10 mL) and MeOH (1.5 mL) was
added a solution of LiOH (0.055 g, 2.3 mmol) in H2O
(4.0 mL). The reaction mixture was stirred at room tem-
perature for 2 h and neutralized with 1.0 M aqueous
HCl. The organic solvents were removed by rotary evap-
oration, H2O (10 mL) was added, and the aqueous
phase was acidified with 1.0 M aqueous HCl to pH 3
and extracted with EtOAc (3 · 40 mL). The organic lay-
ers were washed with brine (20 mL), dried with MgSO4,
and concentrated by rotary evaporation to give 13
(0.12 g, 94%) as white solid. 1H NMR (CD3OD): d
8.14 (d, J = 9.2 Hz, 1H), 8.08–8.02 (m, 2H), 7.60–7.51
(m, 3H), 7.40 (d, J = 2.5 Hz, 1H), 7.28 (s, 1H), 7.14
(dd, J = 2.5, 9.2 Hz, 1H), 5.59–5.54 (m, 1H), 4.70–4.61
(m, 2H), 4.08–4.03 (m, 2H), 3.96 (s, 3H), 2.82 (ddm,
J = 7.4, 14.1 Hz, 1H), 2.47–2.38 (m, 1H), 2.00 (dh,
J = 6.7, 9.0 Hz, 1H), 1.19 (s, 9H), 1.03 (d, J = 6.7 Hz,
3H), 0.97 (d, J = 6.7 Hz, 3H). 13C NMR (CD3OD): d
175.2, 174.0, 163.7, 163.0, 160.9, 157.9, 150.8, 140.1,
131.1, 129.9, 129.2, 124.8, 119.8, 116.4, 106.2, 100.5,
80.3, 78.7, 59.4, 56.1, 54.1, 52.8, 36.1, 31.8, 28.5, 19.6,
19.1. MS (M+H+) 564.3.
1
as white solid. H NMR (CD3OD): d (ꢁ3:2 mixture of
rotamers, major rotamer reported) 8.18 (d, J = 9.2 Hz,
1H), 8.08–8.01 (m, 2H), 7.65–7.55 (m, 3H), 7.43 (d,
J = 2.5 Hz, 1H), 7.35 (s, 1H), 7.22 (dd, J = 2.5 9.2 Hz,
1H), 5.66–5.61 (m, 1H), 4.61–4.59 (m, 1H), 4.20 (m,
2H), 3.98 (s, 3H), 3.81–3.74 (m, 1H), 2.78 (ddm,
J = 7.9, 14.1 Hz, 1H), 2.60–2.35 (m, 3H), 1.86–1.75 (m,
1H), 1.40 (s, 9H), 0.86 (d, J = 6.8 Hz, 3H), 0.81 (d,
J = 6.8 Hz, 3H). 13C NMR (CD3OD): d (ꢁ3:2 mixture
of rotamers, major rotamer reported) 175.5, 173.4,
164.5, 164.2, 160.3, 157.9, 149.7, 138.9, 131.6, 130.1,
129.3, 125.0, 120.2, 116.3, 105.4, 100.7, 79.8, 78.9,
59.5, 56.3, 54.1, 53.9, 38.4, 36.1, 32.8, 28.8, 19.9, 18.2.
HRMS calcd for C32H40N3O7 (M+H+) 578.2866, found:
578.2876.
5.1.15. Compound 16. To a solution of 13 (0.10 g,
0.18 mmol) and 3 (0.048 g, 0.27 mmol) in dry DMF
(2.0 mL) were added HBTU (0.081 g, 0.21 mmol) and
DIEA (0.12 mL, 0.71 mmol). The resulting solution
was stirred at room temperature for 3 h, diluted with
EtOAc (15 mL), and washed with aqueous NaOAc buf-
fer pH 4 (2 · 10 mL), 5% aqueous NaHCO3 (10 mL),
and brine (10 mL). The organic layer was dried with
MgSO4 and concentrated by rotary evaporation. The
resulting product was purified by column chromatogra-
phy (gradient elution, i-hexane/EtOAc 1:1 followed by i-
hexane/EtOAc 1:3) to give 16 (0.82 g, 67%) as white so-
lid. 1H NMR (CD3OD): d (1:1 mixture of diasteromers)
8.13–8.04 (m, 3H), 7.58–7.48 (m, 3H), 7.42–7.39 (m,
1H), 7.273 & 7.268 (s, 1H), 7.12–7.07 (m, 1H), 5.58–
5.52 (m, 1H), 4.60–4.51 (m, 2H), 4.10–4.03 (m, 2H),
3.96 (s, 3H), 3.89 & 3.84 (s, 1H), 3.73 & 3.72 (s, 3H)
2.73–2.61 (m, 1H), 2.45–2.36 (m, 1H), 2.04–1.94 (m,
1H), 1.25 & 1.24 (s, 9H), 1.25–1.09 (m, 2H), 1.03–0.94
(m, 6H), 0.93–0.73 (m, 2H). 13C NMR (CD3OD): d
(1:1 mixture of diasteromers) 176.0 & 174.7, 174.4 &
174.2, 173.9 & 173.8, 163.0, 161.9, 161.2, 157.9, 152.2,
141.3, 130.5, 129.8, 129.0, 124.4, 119.2, 116.4, 107.3,
100.0, 80.3, 78.0, 76.5 & 74.5, 60.4 & 60.2, 59.5 &
59.4, 56.0, 54.4, 52.6, 37.0 & 36.5, 36.2 & 36.0, 31.7,
28.6, 19.8 & 19.0, 19.1 & 19.0, 12.8, 12.4 & 11.8. HRMS
calcd for C37H47N4O9 (M+H+) 691.3343, found:
691.3331.
5.1.13. Compound 14. To a solution of 11 (0.099 g,
0.17 mmol) in THF (10 mL) and MeOH (1.5 mL)
was added a solution of LiOH (0.040 g, 1.7 mmol) in
H2O (4.0 mL). The reaction mixture was stirred at
room temperature for 4 h and neutralized with 1.0 M
aqueous HCl. The organic solvents were removed by
rotary evaporation, H2O (10 mL) was added, and
the aqueous phase was acidified with 1.0 M aqueous
HCl to pH 3 and extracted with EtOAc (3 · 40 mL).
The organic layers were washed with brine (20 mL),
dried with MgSO4, and concentrated by rotary evapo-
ration to give 14 (0.093 g, 97%) as white solid. 1H
NMR (CD3OD): d (ꢁ5:1 mixture of rotamers, major
rotamer reported) 8.11 (d, J = 9.2 Hz, 1H), 8.05–8.01
(m, 2H), 7.59–7.48 (m, 3H), 7.39 (d, J = 2.5 Hz, 1H),
7.23 (s, 1H), 7.11 (dd, J = 2.5, 9.2 Hz, 1H), 5.54–
5.43 (m, 1H), 4.58 (dm, J = 12.2 Hz, 1H), 4.58–4.50
(m, 1H), 4.04 (d, J = 8.8 Hz, 1H), 3.99 (dd, J = 3.3,
12.2 Hz, 1H), 3.95 (s, 1H), 3.04 (dd, J = 3.3,
16.3 Hz, 1H), 2.76–2.67 (m, 2H), 2.39 (ddd, J = 4.5,
9.3, 14.0 Hz, 1H), 1.98 (dh, J = 6.8, 8.8 Hz, 1H),
1.21 (s, 9H), 0.95 (d, J = 6.8 Hz, 1H), 0.94 (d,
J = 6.8 Hz, 1H). 13C NMR (CD3OD): d (ꢁ5:1 mixture
of rotamers, major rotamer reported) 174.8, 173.9,
163.4, 162.7, 161.1, 157.9, 151.4, 140.8, 130.8, 129.9,
129.0, 124.7, 119.5, 116.5, 106.7, 100.3, 80.3, 78.1,
59.9, 56.1, 54.7, 54.3, 37.8, 37.4, 31.9, 28.5, 19.6,
19.0. HRMS calcd for C32H40N3O7 (M+H+)
578.2866, found: 578.2855.
5.1.16. Compound 17. To a solution of 13 (0.11 g,
0.20 mmol) and 4 (0.050 g, 0.30 mmol) in dry DMF
(2.0 mL) were added HBTU (0.092 g, 0.24 mmol) and
DIEA (0.14 mL, 0.81 mmol). The resulting solution
was stirred at room temperature for 3 h, diluted with
EtOAc (20 mL), and washed with aqueous NaOAc buf-
fer pH 4 (5 · 8 mL), 5% aqueous NaHCO3 (8 mL), and

J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
5601
brine (8 mL). The organic layer was dried with MgSO4
and concentrated by rotary evaporation. The resulting
product was purified by column chromatography (gradi-
ent elution, i-hexane/EtOAc 1:1 followed by i-hexane/
uct was purified by column chromatography (gradient
elution, i-hexane/EtOAc 1:2 followed by i-hexane/
1
EtOAc 1:3) to give 19 (0.086 g, 76%) as white solid H
NMR (CD3OD): d (ꢁ 2:1 mixture of rotamers, major
rotamer reported) 8.10–8.01 (m, 3H), 7.57–7.47 (m,
3H), 7.41 (d, J = 2.5 Hz, 1H), 7.30 (s, 1H), 7.15 (dd,
J = 2.5, 9.1 Hz, 1H), 5.62–5.57 (m, 1H), 4.61–4.55 (m,
1H), 4.14 (dd, J = 4.3, 11.7 Hz, 1H), 4.06 (dm,
J = 11.7 Hz, 1H), 3.96 (s, 3H), 3.82–3.74 (m, 1H), 3.66
(s, 3H), 2.73–2.65 (m, 1H), 2.63–2.51 (m, 1H), 2.55
(dd, J = 5.5, 15.2 Hz, 1H), 2.44 (dd, J = 7.8, 15.2 Hz,
1H), 1.81–1.71 (m, 1H), 1.57–1.49 (m, 1H), 1.49–1.42
(m, 1H), 1.41 (s, 9H), 1.25–1.14 (m, 2H), 0.86 (d,
J = 6.8 Hz, 3H), 0.81 (d, J = 6.8 Hz, 3H). 13C NMR
(CD3OD): d (ꢁ 2:1 mixture of rotamers, major rotamer
reported) 175.3, 174.3, 172.9, 163.3, 162.1, 161.4, 158.0,
152.3, 141.3, 130.6, 129.8, 129.0, 124.2, 119.5, 116.4,
107.5, 100.1, 79.9, 77.6, 60.3, 56.0, 54.3, 54.1, 53.0,
38.6, 36.0, 34.4, 33.1, 28.8, 19.7, 18.2, 18.0, 17.4. HRMS
calcd for C37H47N4O8 (M+H+) 675.3394, found:
675.3413.
1
EtOAc 1:2) to give 17 (0.11 g, 81%) as white solid. H
NMR (CD3OD): d 8.09 (d, J = 9.2 Hz, 1H), 8.07–8.02
(m, 2H), 7.57–7.47 (m, 3H) 7.38 (d, J = 2.6 Hz, 1H),
7.23 (s, 1H), 7.08 (dd, J = 2.6, 9.2 Hz, 1H), 5.55–5.49
(m, 1H), 4.57 (dm, J = 11.8 Hz, 1H), 4.55 (dd, J = 7.4,
9.5 Hz, 1H), 4.05 (dd, J = 3.9, 11.8 Hz, 1H), 4.04 (d,
J = 8.9 Hz, 1H) 3.94 (s, 3H), 3.66 (s, 3H), 2.73 (d,
J = 15.7 Hz, 1H), 2.64 (ddm, J = 7.4, 13.9 Hz, 1H),
2.47 (d, J = 15.7 Hz, 1H), 2.36 (ddd, J = 4.4, 9.5,
13.9 Hz, 1H), 1.99 (dh, J = 6.7, 8.9 Hz, 1H), 1.23 (s,
9H), 1.01 (d, J = 6.7 Hz, 3H), 0.96 (d, J = 6.7 Hz, 3H),
0.91–0.76 (m, 4H).13C NMR (CD3OD): d 174.3, 173.9,
173.6, 163.2, 162.1, 161.4, 157.9, 152.2, 141.4, 130.6,
129.8, 129.1, 124.4, 119.3, 116.5, 107.3, 100.1, 80.3,
78.1, 60.4, 59.6, 56.0, 54.4, 52.1, 41.2, 36.1, 31.7, 31.0,
28.6, 19.7, 19.1, 14.1, 13.4. HRMS calcd for
C37H47N4O8 (M+H+) 675.3394, found: 675.3383.
5.1.17. Compound 18. To a solution of 14 (0.086 g,
0.15 mmol) and the hydrochloric salt of 1-amino-cyclo-
propanecarboxylic acid methyl ester (0.034 g,
0.22 mmol) in dry DMF (3.0 mL) were added HBTU
(0.068 g, 0.18 mmol) and DIEA (0.10 mL, 0.60 mmol).
The resulting solution was stirred at room temperature
for 3 h diluted with EtOAc (30 mL), and washed with
aqueous NaOAc buffer pH 4 (3 · 12 mL) and brine
(10 mL). The organic layer was dried with MgSO4 and
concentrated by rotary evaporation. The resulting prod-
uct was purified by column chromatography (i-hexane/
EtOAc 1:3) to give 18 (0.065 g, 65%) as white solid.
1H NMR (CD3OD): d (ꢁ5:1 mixture of rotamers, major
rotamer reported) 8.17 (d, J = 9.3 Hz, 1H), 8.08–8.01
(m, 2H), 7.62–7.53 (m, 3H), 7.42 (d, J = 2.5 Hz, 1H),
7.33 (s, 1H), 7.17 (dd, J = 2.5, 9.3 Hz, 3H), 5.56–5.50
(m, 1H), 4.60 (dm, J = 12.6 Hz, 3H), 4.60–4.55 (m,
1H), 4.04 (d, J = 8.9 Hz, 1H), 4.01 (dd, J = 8.8,
12.6 Hz, 1H), 3.98 (s, 3H), 3.59 (s, 3H), 3.04 (dd,
J = 3.9, 14.5 Hz, 1H), 2.71 (ddm, J = 8.0, 14.5 Hz,
1H), 2.50–2.41 (m, 2H), 2.00 (dh, J = 6.6, 8.9 Hz, 1H),
1.51–1.46 (m, 2H), 1.23 (s, 9H) 1.13–1.08 (m, 2H),
0.97 (d, J = 6.6 Hz, 3H), 0.95 (d, J = 6.6 Hz, 3H). 13C
NMR (CD3OD): d (ꢁ5:1 mixture of rotamers, major
rotamer reported) 174.4, 174.2, 173.9, 163.2, 162.3,
161.3, 157.9, 152.1, 141.3, 130.6, 129.8, 129.1, 124.5,
119.3, 116.5, 107.2, 100.2, 80.4, 77.8, 59.9, 56.0, 55.3,
54.2, 52.9, 39.5, 37.0, 34.2, 31.9, 28.6, 19.7, 19.0, 17.8,
17.7. HRMS calcd for C37H47N4O8 (M+H+) 675.3394,
found: 675.3398.
5.1.19. Compound 20. To a solution of 16 (0.074 g,
0.11 mmol) in THF (10.0 mL) and MeOH (1.5 mL)
was added a solution of LiOH (0.026 g, 1.1 mmol) in
H2O (4.0 mL). The reaction mixture was stirred at room
temperature for 3 h and neutralized with 1.0 M aqueous
HCl. The organic solvents were removed by rotary evap-
oration, H2O (10 mL) was added, and the aqueous
phase was acidified with 1.0 M aqueous HCl to pH 3
and extracted with EtOAc (3 · 15 mL). The organic lay-
ers were washed with brine (10 mL), dried with MgSO4,
and concentrated by rotary evaporation to give 20
1
(0.070 g, 97%) as white solid. H NMR (CD3OD): d
(1:1 mixture of diasteromers) 8.12 (d, J = 9.2 Hz, 1H),
8.09–8.03 (m, 2H), 7.59–7.50 (m, 3H), 7.40 (d,
J = 2.5 Hz, 1H), 7.28 (s, 1H), 7.12 & 7.11 (dd, J = 2.5,
9.2 Hz, 1H), 5.59–5.53 (m, 1H), 4.63–4.53 (m, 2H),
4.10–4.00 (m, 2H), 3.95 (s, 3H), 4.06 (detected by
HSQC) & 3.81 (s, 1H), 2.74–2.63 (m, 1H), 2.51–2.38
(m, 1H), 2.04–1.93 (m, 1H), 1.24 (s, 9H), 1.04–0.73 (m,
4H), 1.01 (d, J = 6.8 Hz, 3H), 0.96 (d, J = 6.8 Hz, 3H).
13C NMR (CD3OD): d (1:1 mixture of diasteromers)
176.3 & 175.8, 176.0 & 174.9, 174.0 & 173.9, 163.6,
162.8, 160.9, 157.9, 151.0, 140.2, 131.0, 129.9, 129.2,
124.7, 119.6, 116.4, 106.4, 100.4, 80.3, 78.6 & 78.5,
77.1 & 73.7, 60.5 & 60.2, 59.6, 56.1, 54.4, 37.2 & 36.3,
36.2 & 36.1, 31.7, 28.6, 19.74 & 19.67, 19.09 & 19.07,
13.1
&
12.9, 12.0
&
11.1. HRMS calcd for
C36H45N4O9 (M+H+) 677.3187, found: 677.3199.
5.1.20. Compound 21. To a solution of 17 (0.097 g,
0.14 mmol) in THF (7.0 mL) and MeOH (1.5 mL) was
added a solution of LiOH (0.035 g, 1.4 mmol) in H2O
(2.0 mL). The reaction mixture was stirred at room tem-
perature overnight and neutralized with 1.0 M aqueous
HCl. The organic solvents were removed by rotary evap-
oration, H2O (10 mL) was added, and the aqueous
phase was acidified with 1.0 M aqueous HCl to pH 3
and extracted with EtOAc (3 · 35 mL). The organic lay-
ers were washed with brine (40 mL), dried with MgSO4
and concentrated by rotary evaporation to give 21
(0.093 g, 98%) as white solid. 1H NMR (CD3OD): d
5.1.18. Compound 19. To a solution of 15 (0.097 g,
0.17 mmol) and the hydrochloric salt of 1-amino-cyclo-
propanecarboxylic acid methyl ester (0.038 g,
0.25 mmol) in dry DMF (3.0 mL) were added HBTU
(0.077 g, 0.20 mmol) and DIEA (0.12 mL, 0.67 mmol).
The resulting solution was stirred at room temperature
for 4 h, diluted with EtOAc (30 mL), and washed with
aqueous NaOAc buffer pH 4 (4 · 12 mL) and brine
(15 mL). The organic layer was dried with MgSO4 and
concentrated by rotary evaporation. The resulting prod-

5602
J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
8.11 (d, J = 9.2 Hz, 1H), 8.07–8.03 (m, 2H), 7.58–7.49
(m, 3H) 7.40 (d, J = 2.5 Hz, 1H), 7.25 (s, 1H), 7.10
(dd, J = 2.5, 9.2 Hz, 1H), 5.55–5.53 (m, 1H), 4.59 (dm,
J = 11.8 Hz, 1H), 4.56 (dd, J = 7.6, 9.8 Hz, 1H), 4.06
(dd, J = 3.7, 11.8 Hz, 1H), 4.05 (d, J = 8.9 Hz, 1H)
3.96 (s, 3H), 2.79 (d, J = 15.7 Hz, 1H), 2.67 (ddm,
J = 7.6, 14.1 Hz, 1H), 2.38 (d, J = 15.7 Hz, 1H), 2.38
(ddd, J = 4.0, 9.8, 14.1 Hz, 1H), 1.99 (dh, J = 6.8,
8.9 Hz, 1H), 1.23 (s, 9H), 1.02 (d, J = 6.8 Hz, 3H),
0.97 (d, J = 6.8 Hz, 3H), 0.90–0.76 (m, 4H). 13C NMR
(CD3OD): d 175.3, 174.3, 173.9, 163.3, 162.3, 161.3,
157.9, 151.8, 141.0, 130.7, 129.8, 129.1, 124.5, 119.4,
116.5, 107.0, 100.2, 80.3, 78.2, 60.4, 59.6, 56.0, 54.5,
41.4, 36.1, 31.7, 31.0, 28.6, 19.7, 19.1, 14.2, 13.5. HRMS
calcd for C36H45N4O8 (M+H+) 661.3237, found:
661.3242. RP-HPLC purity (column 1: 99%, column 2:
99%).
major rotamer reported) 8.09 (d, J = 9.2 Hz, 1H), 8.08–
8.02 (m, 2H), 7.58–7.48 (m, 3H), 7.41 (d, J = 2.5 Hz,
1H), 7.31 (s, 1H), 7.16 (dd, J = 2.5, 9.2 Hz, 1H), 5.63–
5.57 (m, 1H), 4.61–4.55 (m, 1H), 4.14 (dd, J = 4.2,
11.7 Hz, 1H), 4.05 (dm, J = 11.7 Hz, 1H), 3.96 (s, 3H),
3.81–3.75 (m, 1H), 2.75–2.67 (m, 1H), 2.61–2.52 (m,
2H), 2.43 (dd, J = 7.7, 15.2 Hz, 1H), 1.81–1.71 (m,
1H), 1.58–1.50 (m, 1H), 1.58–1.50 (m, 1H), 1.41 (s,
9H), 1.26–1.10 (m, 2H), 0.86 (d, J = 6.8 Hz, 1H), 0.80
(d, J = 6.8 Hz, 1H). 13C NMR (CD3OD): d (ꢁ2:1 mix-
ture of rotamers, major rotamer reported) 176.0,
175.1, 172.8, 163.4, 162.5, 161.0, 158.0, 151.6, 140.6,
130.8, 129.9, 129.0, 124.3, 119.6, 116.3, 107.0, 100.2,
79.9, 77.9, 60.3, 56.1, 54.2, 54.1, 38.5, 36.0, 34.3, 33.1,
28.8, 19.9, 18.2, 17.9, 17.3. HRMS calcd for
C36H45N4O8 (M+H+) 661.3237, found: 661.3258. RP-
HPLC purity (column 1: 98%, column 2: 99%).
5.1.21. Compound 22. To a solution of 18 (0.057 g,
0.084 mmol) in THF (7.0 mL) and MeOH (1.5 mL)
was added a solution of LiOH (0.020 g, 0.84 mmol) in
H2O (2.0 mL). The reaction mixture was stirred at room
temperature for 4 h and neutralized with 1.0 M aqueous
HCl. The organic solvents were removed by rotary evap-
oration, H2O (10 mL) was added, and the aqueous
phase was acidified with 1.0 M aqueous HCl to pH 3
and extracted with EtOAc (3 · 25 mL). The organic lay-
ers were washed with brine (20 mL), dried with MgSO4
and concentrated by rotary evaporation. The resulting
product was purified by preparative RP-HPLC
(MeCN/H2O gradient with 0.1% TFA) to give the
5.1.23. Compound 24. A solution of 21 (0.030 g,
0.045 mmol), HATU (0.021 g, 0.055 mmol), and DIEA
(0.032 mL, 0.18 mmol), in dry DMF (2.0 mL) was stir-
red at room temperature for 1.5 h. Benzenesulfonamide
(0.029 g, 0.18 mmol), DMAP (0.022 g, 0.18 mmol) and
DBU (0.028 mL, 0.18 mmol) in dry DMF (1.0 mL) were
added and the mixture was stirred at room temperature
overnight, diluted with EtOAc (25 mL), and washed
with aqueous NaOAc buffer pH 4 (2 · 12 mL), 5% aque-
ous NaHCO3 (12 mL) and brine (15 mL). The organic
layer was dried with MgSO4 and concentrated by rotary
evaporation. The resulting product was purified by pre-
parative RP-HPLC (MeCN/H2O gradient with 0.1%
TFA) to give the TFA salt of 24 (0.024 g, 58%) as white
1
TFA salt of 22 (0.044 g, 67%) as white solid. H NMR
1
(CD3OD): d (ꢁ5:1 mixture of rotamers, major rotamer
reported) 8.33 (d, J = 9.3 Hz, 1H), 8.09–8.04 (m, 2H),
7.74–7.67 (m, 3H), 7.55 (s, 1H), 7.49 (d, J = 2.4 Hz,
1H), 7.34 (dd, J = 2.4, 9.3 Hz, 1H), 5.71–5.67 (m, 1H),
4.71 (dm, J = 12.2 Hz, 1H), 4.66–4.58 (m, 1H), 4.06–
4.00 (m, 1H), 4.03 (s, 3H), 3.98 (d, J = 8.7 Hz, 1H),
3.04 (dd, J = 3.8, 14.6 Hz, 1H), 2.77 (ddm, J = 7.6,
14.6 Hz, 1H), 2.57–2.49 (m, 2H), 1.97 (dh, J = 6.7,
8.7 Hz, 1H), 1.51–1.47 (m, 2H), 1.17 (s, 9H), 1.12–
1.08, (m, 2H), 0.97 (d, J = 6.7 Hz, 3H), 0.95 (d, J = 6.7
Hz, 3H). 13C NMR (CD3OD): d (ꢁ5:1 mixture of rota-
mers, major rotamer reported) 175.8, 174.2, 174.1,
167.5, 166.1, 158.6, 158.0, 144.7, 134.6, 133.4, 130.7,
129.8, 126.4, 121.5, 116.2, 102.1, 101.3, 80.8, 80.1,
60.0, 56.8, 55.2, 53.9, 39.2, 37.3, 34.0, 31.8, 28.5, 19.7,
19.2, 17.7, 17.6. HRMS calcd for C36H45N4O8
(M+H+) 661.3237, found: 661.3238. RP-HPLC purity
(column 1: 99%, column 2: 99%).
solid. H NMR (CD3OD): d 8.35 (d, J = 9.3 Hz, 1H),
8.10–8.07 (m, 2H), 8.02–7.99 (m, 2H) 7.76–7.69 (m,
3H), 7.66–7.61 (m, 1H), 7.60 (s, 1H) 7.58–7.53 (m,
2H), 7.52 (d, J = 2.4 Hz, 1H), 7.38 (dd, J = 2.4, 9.3 Hz,
1H), 5.78–5.74 (m, 1H), 4.76 (dm, J = 12.3 Hz, 1H),
4.55 (dd, J = 7.5, 10.0 Hz, 1H), 4.08 (dd, J = 3.4,
12.3 Hz, 1H), 4.05 (s, 3H), 3.96 (d, J = 9.0 Hz, 1H),
2.73 (d, J = 15.7 Hz, 1H), 2.70 (ddm, J = 7.5, 14.3 Hz,
1H), 2.45–2.35 (m, 1H), 2.36 (d, J = 15.7 Hz, 1H), 1.97
(dh, J = 6.7, 9.0 Hz, 1H), 1.17 (s, 9H), 1.03 (d,
J = 6.7 Hz, 3H), 0.98 (d, J = 6.7 Hz, 3H), 0.83–0.78
(m, 2H), 0.75–0.69 (m, 2H). 13C NMR (CD3OD): d
174.1, 174.0, 171.3, 167.9, 166.5, 158.4, 158.0, 143.9,
141.0, 134.8, 133.3, 133.6, 130.8, 130.0, 129.9, 129.2,
126.4, 121.9, 116.2, 102.5, 100.6, 81.3, 80.1, 60.3, 59.7,
56.9, 54.2, 43.0, 36.3, 31.6, 30.8, 28.5, 19.6, 19.3, 13.7,
13.6. HRMS calcd for C42H50N5O9S (M+H+)
800.3329, found: 800.3323. RP-HPLC purity (column
1: >99%, column 2: 99%).
5.1.22. Compound 23. To a solution of 19 (0.073 g,
0.11 mmol) in THF (7.0 mL) and MeOH (1.5 mL) was
added a solution of LiOH (0.026 g, 1.1 mmol) in H2O
(2.0 mL). The reaction mixture was stirred at room tem-
perature for 5 h and neutralized with 1.0 M aqueous
HCl. The organic solvents were removed by rotary evap-
oration and the remaining aqueous phase was acidified
with 1.0 M aqueous HCl to pH 3 and extracted with
EtOAc (3 · 30 mL). The organic layers were washed
with brine (25 mL), dried with MgSO4 and concentrated
by rotary evaporation to give 23 (0.069 g, 96%) as white
solid. 1H NMR (CD3OD): d (ꢁ2:1 mixture of rotamers,
5.1.24. Compound 25. A solution of 22 (0.025 g,
0.038 mmol), HATU (0.017 g, 0.045 mmol), and DIEA
(0.026 mL, 0.15 mmol) in dry DMF (2.0 mL) was stirred
at room temperature for 1.5 h. Benzenesulfonamide
(0.024 g, 0.15 mmol), DMAP (0.019 g, 0.18 mmol), and
DBU (0.023 mL, 0.15 mmol) in dry DMF (1.0 mL) were
added and the mixture was stirred at room temperature
overnight, diluted with EtOAc (25 mL), and washed
with aqueous NaOAc buffer pH 4 (2 · 12 mL), 5% aque-
ous NaHCO3 (12 mL), and brine (15 mL). The organic
layer was dried with MgSO4 and concentrated by rotary

J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
5603
evaporation. The resulting product was purified by pre-
parative RP-HPLC (MeCN/H2O gradient with 0.1%
TFA) to give the TFA salt of 25 (0.021 g, 60%) as white
MgSO4, and concentrated by rotary evaporation. The
resulting product was purified by preparative RP-HPLC
(MeCN/H2O gradient with 0.1% TFA) to give the TFA
salt of 27 (0.010 g, 36%) as white solid. 1 NMR (DMSO-
d6): d 8.88 (s, 1H), 8.25–8.13 (m, 3H), 7.71–7.57 (m, 4H),
7.53–7.46 (m, 1H), 7.25–7.17 (m, 1H), 7.03 (d,
J = 7.9 Hz, 1H), 5.81–5.75 (m, 1H), 4.57–4.51 (m, 1H),
4.47–4.41 (m, 1H), 3.96 (s, 3H), 3.96–3.91 (m, 1H), 3.87–
3.81 (m, 1H), 2.59–2.52 (m, 1H), 2.27–2.18 (m, 1H),
1.93–1.82 (m, 1H), 1.64–1.58 (m, 1H), 1.47–1.41 (m, 1H),
1.29–1.22 (m, 1H), 1.19 (s, 9H), 1.13–1.06 (m, 1H), 0.90
(d, J = 6.8 Hz, 3H), 0.87 (d, J = 6.8 Hz, 3H). HRMS calcd
for C36H43N4O9 (M+H+) 675.3030, found: 675.3035. RP-
HPLC purity (column 1: 96%, column 2: 96%).
1
solid. H NMR (CD3OD): d 8.37 (d, J = 9.3 Hz, 1H),
8.10–8.05 (m, 2H), 8.03–7.97 (m, 2H), 7.78–7.64 (m,
4H), 7.60 (s, 1H), 7.57–7.53 (m, 2H), 7.52 (d,
J = 2.5 Hz, 1H), 7.38 (dd, J = 2.5, 9.2 Hz, 1H), 5.76–
5.71 (m, 1H), 4.77–6.67 (m, 1H), 4.61 (dm,
J = 12.7 Hz, 1H), 4.13 (d, J = 8.2 Hz, 1H), 4.12–4.07
(m, 1H), 4.05 (s, 3H), 2.78 (ddm, J = 6.9, 14.6 Hz,
1H), 2.63 (dd, J = 7.7, 15.3 Hz, 1H), 2.48 (dd, J = 3.0,
15.3 Hz, 1H), 2.30 (ddd, J = 3.8, 10.6, 14.6 Hz, 1H),
2.10 (dh, J = 6.7, 8.2 Hz, 1H), 1.54–1.46 (m, 1H),
1.40–1.33 (m, 1H), 1.20 (s, 9H), 1.15–1.07, (m, 2H),
1.06 (d, J = 6.7 Hz, 3H), 1.04 (d, J = 6.7 Hz, 3H). 13C
NMR (CD3OD): d 175.6, 174.5, 173.5, 167.7, 166.4,
158.4, 158.2, 144.2, 140.7, 134.8, 133.9, 133.7, 130.7,
129.9, 129.9, 129.2, 126.4, 121.8, 116.2, 102.3, 100.9,
80.4, 80.3, 59.5, 56.9, 55.7, 54.1, 42.3, 38.5, 36.0, 32.1,
28.5, 19.9, 19.8, 19.4, 19.0. HRMS calcd for
C42H50N5O9S (M+H+) 800.3329, found: 800.3328. RP-
HPLC purity (column 1: >99%, column 2: 98%).
5.2. Inhibition analysis
Protease inhibition of full-length HCV NS3 protein was
determined essentially as described previously.37 The
hydrolysis of a depsipeptide substrate, Ac-Asp-Glu-As-
p(EDANS)-Glu-Glu-Abu-w-[COO]Ala-Ser-Lys(DAB-
CYL)-NH2 (AnaSpec, San Jose, CA, USA), was
measured continuously over time with a fluorescence
plate reader (Fluoroskan Ascent Labsystems, Stock-
holm, Sweden). Enzyme at a concentration of 1 nM
was pre-incubated in 50 mM HEPES, pH 7.5, 10 mM
DTT, 40% (w/v) glycerol, 0.1% n-octyl-b-D-glucoside,
3.3% (v/v) DMSO, 25 lM NS4A peptide cofactor
(KKGSVVIVGRIVLSGK), and inhibitor. The reaction
was started by the addition of 0.5 lM substrate and the
initial rate velocities were used to determine the Ki using
nonlinear regression analysis.49
5.1.25. Compound 26. A solution of 23 (0.030 g,
0.045 mmol), HATU (0.021 g, 0.055 mmol) and DIEA
(0.032 mL, 0.18 mmol), in dry DMF (2.0 mL) was stirred
at room temperature for 1.5 h. Benzenesulfonamide
(0.029 g, 0.18 mmol), DMAP (0.022 g, 0.18 mmol) and
DBU (0.028 mL, 0.18 mmol) in dry DMF (1.0 mL) were
added and the mixture was stirred at room temperature
overnight, diluted with EtOAc (25 mL), and washed with
aqueous NaOAc buffer pH 4 (2 · 12 mL), 5% aqueous
NaHCO3 (12 mL), and brine (15 mL). The organic layer
was dried with MgSO4 and concentrated by rotary evap-
oration. The resulting product was purified by prepara-
tive RP-HPLC (MeCN/H2O gradient with 0.1% TFA)
to give the TFA salt of 26 (0.026 g, 63%) as white solid.
1H NMR (CD3OD): d 8.38 (d, J = 9.3 Hz, 1H), 8.09–
8.04 (m, 2H), 8.03–7.98 (m, 2H) 7.76–7.65 (m, 4H),
7.61–7.54 (m, 3H), 7.53 (d, J = 2.5 Hz, 1H), 7.43 (dd,
J = 2.5, 9.3 Hz, 1H), 5.89–5.84 (m, 1H), 4.24–4.19 (m,
2H), 4.06 (s, 3H), 3.76–3.68 (m, 1H), 2.71 (ddm, J = 6.8,
13.8 Hz, 1H), 2.69–2.65 (ddd, J = 4.0, 10.4, 13.8 Hz,
1H), 2.02–1.91 (m, 1H), 1.48–1.39 (m, 2H), 1.41 (s, 9H),
1.15–1.06 (m, 2H), 0.95, (d, J = 6.8 Hz, 3H), 0.89 (d,
J = 6.8 Hz, 3H). 13C NMR (CD3OD): d 174.9, 173.7,
172.9, 167.8, 166.6, 158.2, 158.1, 143.6, 140.5, 134.8,
133.9, 133.3, 130.8, 129.9, 129.9, 129.2, 126.4, 122.0,
116.0, 102.2, 100.5, 81.2, 79.3, 61.2, 57.0, 54.8, 54.4,
38.3, 35.8, 35.7, 32.6, 28.8, 20.0, 19.6, 19.2, 19.0. HRMS
calcd for C42H50N5O9S (M+H+) 800.3329, found:
800.3304. RP-HPLC purity (column 1: 99%, column 2:
99%).
5.3. Molecular docking
FLO+50 was used for all docking calculations and was
chosen partly due to its ability to allow protein flexibility,
allowing a better approximation of protein–ligand inter-
actions.51 Like the assay used for inhibition analysis, the
active site used in docking is comprised of both the prote-
ase and helicase domains of the NS3 protein and is derived
from the crystal structure of Yao et al. (PDB code
1CU1).41 This active site has been used previously and de-
scribed in greater detail in our earlier work.35
Ligand conformations within the active site were explored
using limited Monte Carlo perturbation and simulated
annealing was used to explore local conformational
space. To limit conformational change induced by simu-
2
˚
lated annealing, an energy penalty of 20 kJ/(molA ) was
applied when the similarity distance between two sequen-
˚
tial conformations differed by more than 0.2 A. Residues
R155 and K136 of the active site were allowed full confor-
mational freedom while movement of all other residues by
2
˚
˚
5.1.26. Compound 27. Compound 20 (0.024 g,
0.35 mmol) was dissolved in dry DCM (1.5 mL), stirred
under N2 atmosphere, and cooled to 0 ꢁC. Dess–Martin
periodinane (0.022 g, 0.53 mmol) in dry DCM (1.5 mL)
was added and the resulting mixture was stirred at room
temperature for 4 h. The mixture was cooled to 0 ꢁC and
quenched with saturated aqueous Na2S2O4 (2.0 mL) and
saturated aqueous NaHCO3 (2.0 mL). The reaction mix-
ture was extracted with DCM (3 · 10 mL), dried with
more than 0.2 A was penalized by 20 kJ/(molA ). Zero-
order bonds to hold the ligand in the vicinity of the cata-
lytic site of the protein were used.
5.4. Comparative molecular field analysis
A 3D-QSAR model was derived using comparative
molecular field analysis (CoMFA)28 using all of our pre-
viously synthesized and published proline-based inhibi-

5604
J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
tors33–35 as well as the b-amino acid- comprising pep-
tides presented herein. Docking was used to propose
bioactive conformation and provide molecular align-
ment. Docking poses for subsequent use in CoMFA
were selected based on docking score as well as subjec-
tive analysis. Prior to use in CoMFA, ligands were sub-
jected to 10 steps of minimization using the Tripos force
field52 and partial atomic charges were calculated using
6. Lin, K.; Perni, R. B.; Kwong, A. D.; Lin, C. Antimicrob.
Agents Chemother. 2006, 50, 1813.
7. Malcolm, B. A.; Liu, R.; Lahser, F.; Agrawal, S.;
Belanger, B.; Butkiewicz, N.; Chase, R.; Gheyas, F.; Hart,
A.; Hesk, D.; Ingravallo, P.; Jiang, C.; Kong, R.; Lu, J.;
Pichardo, J.; Prongay, A.; Skelton, A.; Tong, X.; Ven-
katraman, S.; Xia, E.; Girijavallabhan, V.; Njoroge, F. G.
Antimicrob. Agents Chemother. 2006, 50, 1013.
8. Condroski, K. R.; Zhang, H.; Seiwart, S. D.; Ballard, J.
A.; Bernat, B. A.; Brandhuber, B. J.; Andrews, S. W.;
Josey, J. A.; Blatt, L. M. Digestive Disease Week, May 20–
25, Los Angeles 2006, Poster T1794.
the Gasteiger-Huckel53,54 method. A carbon sp3 atom
¨
with a +1 charge and the default CoMFA region were
used for all CoMFA modeling.
9. Hinrichsen, H.; Benhamou, Y.; Wedemeyer, H.; Reiser,
M.; Sentjens, R. E.; Calleja, J. L.; Forns, X.; Erhardt, A.;
Croenlein, J.; Chaves, R. L.; Yong, C.-L.; Nehmiz, G.;
Steinmann, G. G. Gastroenterology 2004, 127, 1347.
10. Press Release at the Vertex Website: http://inves-
tors.vrtx.com/releasedetail.cfm?ReleaseID=248679
(Accessed March, 2008).
11. Press Release at the Schering-Plough Website: http://
www.schering-plough.com/schering_plough/news/
release.jsp?releaseID=1064540 (Accessed March, 2008).
12. Press Release at the InterMune Website: http://phx.cor-
porate-ir.net/phoenix.zhtml?c=100067&p=irol-newsArti-
cle&ID=1092155&highlight= (Accessed March, 2008).
13. Llinas-Brunet, M.; Bailey, M.; Fazal, G.; Goulet, S.;
Halmos, T.; Laplante, S.; Maurice, R.; Poirier, M.;
Poupart, M. A.; Thibeault, D.; Wernic, D.; Lamarre, D.
Bioorg. Med. Chem. Lett. 1998, 8, 1713.
In total, 54 proline-based inhibitors of HCV NS3 prote-
ase were available for use in CoMFA. To allow for mod-
el validation using external prediction, the compounds
were divided into training and test sets. This was done
using the Selector module in Sybyl. CoMFA fields and
biological activity data (expressed as pKi) were used as
descriptors and hierarchical clustering was used to select
the 36 (two-thirds) training set compounds. The remain-
ing 18 compounds were withheld from the CoMFA
model for use as an external validation set. Sybyl 7.255
was used for training and test set selection as well as
CoMFA modeling. The structures and inhibition data
for all compounds included in the training and test sets
are available as Supplementary material. Predictive per-
formance was enhanced by optimizing CoMFA
parameters.36
14. Steinkuhler, C.; Biasiol, G.; Brunetti, M.; Urbani, A.;
Koch, U.; Cortese, R.; Pessi, A.; De Francesco, R.
Biochemistry 1998, 37, 8899.
15. Bianchi, E.; Pessi, A. Biopolymers 2002, 66, 101.
16. Sudo, K.; Yamaji, K.; Kawamura, K.; Nishijima, T.;
Kojima, N.; Aibe, K.; Shimotohno, K.; Shimizu, Y.
Antivir. Chem. Chemother. 2005, 16, 385.
17. Aguilar, M. I.; Purcell, A. W.; Devi, R.; Lew, R.;
Rossjohn, J.; Smith, A. I.; Perlmutter, P. Org. Biomol.
Chem. 2007, 5, 2884.
Acknowledgments
We gratefully acknowledge support from the Knut and
Alice Wallenberg’s Foundation. We also thank Dr Aleh
Yahorau, Department of Pharmaceutical Biosciences,
Uppsala University, for help with HRMS analyses.
18. Steer, D. L.; Lew, R. A.; Perlmutter, P.; Smith, A. I.;
Aguilar, M.-I. Curr. Med. Chem. 2002, 9, 811.
19. Hook, D. F.; Gessier, F.; Noti, C.; Kast, P.; Seebach, D.
ChemBioChem 2004, 5, 691.
Supplementary data
20. Hook, D. F.; Bindschaedler, P.; Mahajan, Y. R.; Sebesta,
R.; Kast, P.; Seebach, D. Chem. Biodivers 2005, 2, 591.
21. Lew, R. A.; Boulos, E.; Stewart, K. M.; Perlmutter, P.;
Harte, M. F.; Bond, S.; Aguilar, M.-I.; Smith, A. I. J.
Pept. Sci. 2000, 6, 440.
22. Steer, D. L.; Lew, R. A.; Perlmutter, P.; Smith, A. I.;
Aguilar, M. I. J. Pept. Sci. 2000, 6, 470.
23. Lew, R. A.; Boulos, E.; Stewart, K. M.; Perlmutter, P.;
Harte, M. F.; Bond, S.; Reeve, S. B.; Norman, M. U.;
Lew, M. J.; Aguilar, M.-I.; Smith, I. FASEB J. 2001, 15,
1664.
Structures and inhibition data for training and test set
compounds and experimental and predicted pKi values
used to produce Figure 6, are available as supplemen-
tary material. Supplementary data associated with this
article can be found, in the online version, at
doi:10.1016/j.bmc.2008.04.005.
References and notes
24. Steer, D.; Lew, R.; Perlmutter, P.; Smith, A. I.; Aguilar,
M.-I. Biochemistry 2002, 41, 10819.
25. Bromfield, K. M.; Quinsey, N. S.; Duggan, P. J.; Pike, R.
N. Chem. Biol. Drug Des. 2006, 68, 11.
26. Cardillo, G.; Gentilucci, L.; Qasem, A. R.; Sgarzi, F.;
Spampinato, S. J. Med. Chem. 2002, 45, 2571.
27. Sagan, S.; Milcent, T.; Ponsinet, R.; Convert, O.; Tasseau,
O.; Chassaing, G.; Lavielle, S.; Lequin, O. Eur. J.
Biochem. 2003, 270, 939.
28. Cramer, R. D., III; Patterson, D. E.; Bunce, J. D. J. Am.
Chem. Soc. 1988, 110, 5959.
29. Mertin, A.; Thiemann, T.; Hanss, I.; De Meijere, A.
Synlett 1991, 87.
30. Salaun, J.; Marguerite, J. Org. Synth. 1985, 63, 147.
¨
31. Campbell, J. A.; Good, A. WO 02/060926, 2002.
1. Shepard, C. W.; Finelli, L.; Alter, M. J. Lancet Infect. Dis.
2005, 5, 558.
2. Pearlman, B. L. Am. J. Med. 2004, 117, 344.
3. Bartenschlager, R.; Lohmann, V. J. Gen. Virol. 2000, 81,
1631.
4. Gale, M.; Foy, E. M. Nature 2005, 436, 939.
5. Lamarre, D.; Anderson, P. C.; Bailey, M.; Beaulieu, P.;
Bolger, G.; Bonneau, P.; Boes, M.; Cameron, D. R.;
Cartier, M.; Cordingley, M. G.; Faucher, A.-M.; Goud-
reau, N.; Kawai, S. H.; Kukolj, G.; Lagace, L.; LaPlante,
S. R.; Narjes, H.; Poupart, M.-A.; Rancourt, J.; Sentjens,
R. E.; St. George, R.; Simoneau, B.; Steinmann, G.;
Thibeault, D.; Tsantrizos, Y. S.; Weldon, S. M.; Yong, C.-
`
L.; Llinas-Brunet, M. Nature 2003, 426, 186.

J. Nurbo et al. / Bioorg. Med. Chem. 16 (2008) 5590–5605
5605
`
32. Llinas-Brunet, M.; Gorys, V. J. WO 03/064456, 2003.
33. Ro¨nn, R.; Gossas, T.; Sabnis, Y. A.; Daoud, H.; Akerb-
J. A.; Tung, R. D.; Van Drie, J. H.; Wei, Y. Bioorg. Med.
Chem. Lett. 2004, 14, 1441.
˚
lom, E.; Danielson, U. H.; Sandstro¨m, A. Bioorg. Med.
Chem. 2007, 15, 4057.
43. Narjes, F.; Brunetti, M.; Colarusso, S.; Gerlach, B.; Koch,
U.; Biasiol, G.; Fattori, D.; De Francesco, R.; Matassa, V.
G.; Steinkuehler, C. Biochemistry 2000, 39, 1849.
44. Tsantrizos, Y. S.; Bolger, G.; Bonneau, P.; Cameron, D.
34. Ro¨nn, R.; Lampa, A.; Peterson, S. D.; Gossas, T.;
˚
´
Akerblom, E.; Danielson, U. H.; Karlen, A.; Sandstro¨m,
A. Bioorg. Med. Chem. 2008, 16, 2955.
`
R.; Goudreau, N.; Kukolj, G.; LaPlante, S. R.; Llinas-
˚
35. Ro¨nn, R.; Sabnis, Y. A.; Gossas, T.; Akerblom, E.;
Danielson, U. H.; Hallberg, A.; Johansson, A. Bioorg.
Med. Chem. 2006, 14, 544.
Brunet, M.; Nar, H.; Lamarre, D. Angew. Chem., Int. Ed.
2003, 42, 1356.
45. Di Marco, S.; Rizzi, M.; Volpari, C.; Walsh, M. A.;
Narjes, F.; Colarusso, S.; De Francesco, R.; Matassa, V.
G.; Sollazzo, M. J. Biol. Chem. 2000, 275, 7152.
´
36. Peterson, S. D.; Schaal, W.; Karlen, A. J. Chem. Inf.
Model 2006, 46, 355.
37. Poliakov, A.; Hubatsch, I.; Shuman, C. F.; Stenberg, G.;
Danielson, U. H. Protein Expr. Purif. 2002, 25, 363.
˚
38. Johansson, A.; Poliakov, A.; Akerblom, E.; Wiklund, K.;
˚
46. Poliakov, A.; Sandstro¨m, A.; Akerblom, E.; Danielson, U.
H. J. Enzyme Inhib. Med. Chem. 2007, 22, 191.
47. Fersht, A. R.; Shi, J. P.; Knill-Jones, J.; Lowe, D. M.;
Wilkinson, A. J.; Blow, D. M.; Brick, P.; Carter, P.; Waye,
M. M. Y.; Winter, G. Nature 1985, 314, 235.
48. Mukaiyama, T.; Shiina, I.; Iwadare, H.; Saitoh, M.;
Nishimura, T.; Ohkawa, N.; Sakoh, H.; Nishimura, K.;
Tani, Y.-I.; Hasegawa, M.; Yamada, K.; Saitoh, K. Chem.
Eur. J. 1999, 5, 121.
49. GOSA; BioLog: Ramonville, France.
50. McMartin, C.; Bohacek, R. S. J. Comput. Aided Mol. Des.
1997, 11, 333.
51. Teague, S. J. Nat. Rev. Drug Disc. 2003, 2, 527.
52. Clark, M.; Cramer, R. D., III; Van Opdenbosch, N. J.
Comput. Chem. 1989, 10, 982.
Lindeberg, G.; Winiwarter, S.; Danielson, U. H.; Samu-
elsson, B.; Hallberg, A. Bioorg. Med. Chem. 2003, 11,
2551.
˚
39. Poliakov, A.; Johansson, A.; Akerblom, E.; Oscarsson,
K.; Samuelsson, B.; Hallberg, A.; Danielson, U. H.
Biochim. Biophys. Acta 2004, 1672, 51.
˚
¨
40. Ortqvist, P.; Peterson, S. D.; Akerblom, E.; Gossas, T.;
Sabnis Yogesh, A.; Fransson, R.; Lindeberg, G.; Daniel-
´
son, U. H.; Karlen, A.; Sandstro¨m, A. Bioorg. Med. Chem.
2007, 15, 1448.
41. Yao, N.; Reichert, P.; Taremi, S. S.; Prosise, W. W.;
Weber, P. C. Structure 1999, 7, 1353.
42. Perni, R. B.; Pitlik, J.; Britt, S. D.; Court, J. J.; Courtney,
L. F.; Deininger, D. D.; Farmer, L. J.; Gates, C. A.;
Harbeson, S. L.; Levin, R. B.; Lin, C.; Lin, K.; Moon, Y.-
C.; Luong, Y.-P.; O’Malley, E. T.; Rao, B. G.; Thomson,
53. Gasteiger, J.; Marsili, M. Tetrahedron 1980, 36, 3219.
54. Purcell, W. P.; Singer, J. A. J. Chem. Eng. Data 1967, 12,
235.
55. Sybyl; 7.2 Ed.; Tripos Inc.: St. Louis, MO.