α-Substituted norstatines as the transition-state mimic in inhibitors of multiple digestive vacuole malaria aspartic proteases

Article abstract of DOI:10.1016/j.bmc.2009.06.065

The impact of moving the P1 side-chain from the β-position to the α-position in norstatine-containing plasmepsin inhibitors was investigated, generating two new classes of tertiary alcohol-comprising α-benzylnorstatines and α-phenylnorstatines. Twelve α-s

Full text of DOI:10.1016/j.bmc.2009.06.065

Bioorganic & Medicinal Chemistry 17 (2009) 5933–5949
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
a-Substituted norstatines as the transition-state mimic in inhibitors
of multiple digestive vacuole malaria aspartic proteases
Kristina M. Orrling a, Melissa R. Marzahn b, Hugo Gutiérrez-de-Terán c, Johan Åqvist d,
Ben M. Dunn b, Mats Larhed a,
*
a Department of Medicinal Chemistry, Division of Organic Pharmaceutical Chemistry, Uppsala University, BMC, Box 574, 75 123 Uppsala, Sweden
b Biochemistry and Molecular Biology, College of Medicine, University of Florida, PO Box 100245, Gainsville, FL, USA
c Fundación Pública Galega de Medicina Xenómica, Hospital Clínico de Santiago de Compostela, 15705 Santiago de Compostela, Spain
d Department of Cell and Molecular Biology, BMC, Uppsala University, Box 596, 751 23 Uppsala, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
The impact of moving the P1 side-chain from the b-position to the
plasmepsin inhibitors was investigated, generating two new classes of tertiary alcohol-comprising
a-position in norstatine-containing
Received 7 May 2009
Revised 24 June 2009
Accepted 28 June 2009
Available online 3 July 2009
a-
benzylnorstatines and -phenylnorstatines. Twelve -substituted norstatines were designed, synthe-
a
a
sized and evaluated for their inhibitory potencies against plasmepsin II and the plasmepsin IV ortho-
logues (PM4) present in the digestive vacuole of all four Plasmodium species causing malaria in man.
New synthetic routes were developed for producing the desired a-substituted norstatines as pure ster-
Keywords:
Malaria
Plasmepsin
Inhibitors
Microwave-assisted synthesis
Molecular dynamics
Linear interaction energy mehtod
eoisomers. The best compounds provided Ki values in the nanomolar range for all PM4, with a best value
of 110 nM in PM4 from Plasmodium ovale. In addition, excellent selectivity over the closely related human
aspartic protease Cathepsin D was achieved. The loss of affinity toPlasmodium falciparum PM4, which was
experienced upon the move of the P1 substituent, was rationalized by the calculation of inhibitor–protein
binding affinities using the linear interaction energy method (LIE).
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
protease, HAP (aka PfPM3), are involved in the hemoglobin prote-
olysis in the DV.6,7 Furthermore, the solution of the first X-ray
Malaria is a major cause of death, illness and poverty for
approximately half of the world’s population.1 It is a parasite-in-
flicted disease, which is spread throughout the tropical regions of
the world by female mosquitoes of the genus Anopheles. Four spe-
cies of protozoal parasites of the Plasmodium genus cause malaria
in humans, Plasmodium falciparum (Pf), Plasmodium malariae
(Pm), Plasmodium ovale (Po) and Plasmodium vivax (Pv). Despite
considerable scientific advances and the development of new
drugs, malaria is reported to cause over one million fatal cases
annually, the majority being children under the age of five.1 The
constant emergence of parasitic strains resistant to the currently
available antimalarials implies an acute need for new therapeutic
agents with novel routes of action.2
In 1990, Goldberg et al. found that pepstatin A, a modified pep-
tide which inhibits most aspartic proteases, drastically reduced the
rate of the hemoglobin degradation in the parasite digestive vacu-
ole (DV).3 In subsequent studies, 10 aspartic proteases (plasmep-
sins) were identified to be encoded in the genome of Pf, the most
lethal malaria species.4,5 Four of these, plasmepsin I (PfPM1), II
(PfPM2) and IV (PfPM4) and the highly homologous histo-aspartic
inhibitor–enzyme cocrystal structure of PfPM2 in 1996 made
structure-based drug discovery efforts possible. Despite substan-
tial enzyme flexibility and difficult modeling, several groups have
since then reported on novel PfPM2 inhibitors which were found
to be active in nanomolar concentrations.8–10 Most of the recently
published inhibitors of the Pf DV plasmepsin have been peptidom-
imetic in structure, using different transition-state (TS) isosteres,
with lipophilic side-chains. In the three other malaria parasites
that infect humans, only the orthologues of PfPM4 have been iso-
lated from the DV (PmPM4, PoPM4 and PvPM4).11 This indicates
that PM4 might be the most relevant DV plasmepsin to target. As
the active site is highly conserved in these iso-enzymes, it has been
hypothesized that an inhibitor of PfPM4 would reduce the para-
sitaemia regardless of the infecting species.11–13 However, there
is evidence for redundancy of DV plasmepsins in Pf, which indi-
cates that all four Pf paralogues might have to be targeted to effec-
tively abate the parasite.14–16
The structural starting point for this study was the compound
KNI-10006, reported by Kiso in 2003 (Fig. 1).17 This norstatine-
based lead structure possesses nanomolar inhibition of all four
plasmepsins in the DV of Pf, and is remarkably potent in the case
of PfPM2 and PfPM4 with IC50 values of 39 nM and 15 nM, respec-
tively. The Ki value for PfPM2 was determined to be in the
* Corresponding author. Fax: +46 18 471 4882.
E-mail address: Mats.Larhed@orgfarm.uu.se (M. Larhed).
0968-0896/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2009.06.065

5934
K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
IC50 values:a
phenylnorstatine scaffold, as well as structures with an
norstatine scaffold (Fig. 2). In order to further reduce the
character of the inhibitors, the
-thioproline pseudoresidue in P10,
used in the KNI compounds, was complemented with compounds
incorporating the analogue with -configuration. In 2005, Beyer
a
-benzyl-
PfPM1 = 0.280 μM
PfPM2 = 0.039 μM
HAP = 0.690 μM
L
-peptidic
HO
O
O
O
O
L
O
PfPM4 = 0.015 μ
N
H
M
N
H
N
Ki values:b
D
OH
S
PfPM2 = 0.0005 μM
hCatD = 0.002 μM
et al. showed that the DV plasmepsins accommodated larger P10-
substituents better than small and constrained ones.22 In order to
investigate if this might hold true also for our new inhibitors of
class A, we decided to incorporate both phenylalanine and proline
moieties, providing examples of a big P10 side-chain and a small,
rigidified P10 group.
The specific objectives of this study were: (1) To identify adap-
tive inhibitors which are active against DV plasmepsins from all
four infecting species. (2) To design and synthesize plasmepsin
inhibitors devoid of any activity on hCatD. (3) To enhance the
membrane permeability by altering the TS-mimicking secondary
hydroxyl to a tertiary hydroxyl group by moving the P1-group
KNI-10006
IC50 value:a
PfPM2 = 0.030 μM
O
O
Ki value:c
HO
PmPM4 = 0.11 μM
N
H
N
H
N
OH
S
KNI-764
Ki values:d
*(S), PfPM4 = 12 μM
HIV-1 = 0.006 μM
*(R), PfPM4 = 48 μM
HO
O
O
H
N
N
N
H
N
O
∗
HIV-1 = 2.2 μ
M
H
O
from the b-position to the
a-position in the norstatine TS mimic
OH
EC50 values:d
and by keeping the molecular weight below 650 g/mol. (4) To elu-
cidate the impact of manipulating the stereochemistry of the thi-
oproline and of the tertiary hydroxyl group with regard to
activity on plasmepsins and selectivity profile. From a synthesis
point of view, this would be the major challenge.
*(S), HIV-1 = 1.9 μM
*(R), HIV-1 = >10 μM
Mzs27255
Figure 1. Phenylnorstatine and a-phenylnorstatine lead structures. (a) IC50 values
for plasmepsins reported by Nezami et al.,17 (b) Ki values reported by Kiso et al.,18
(c) Ki value for PmPM4 reported by Clemente et al.,13 (d) Ki values for HIV-1 refer to
the inhibition of HIV-1 protease and the EC50 values refer to the efficacy on infected
cells, according to Ekegren et al.19 The PfPM4 data of Mz27255 have not been
published elsewhere.
2. Results and discussion
2.1. Chemistry
0.5 1 nM range. However, having a good activity towards several
of aspartic proteases, KNI-10006 was also found to be an effica-
cious inhibitor (Ki = 2 nM)18 of the most closely related human
aspartic protease, cathepsin D (hCatD). Another disadvantage with
this compound was the low capacity to reduce parasite growth in
The tertiary alcohol in the herein reported
a-benzylnorstatines
and -phenylnorstatines originates from the epoxy acids 1a or 1b,
a
respectively. The novel, four-step synthetic route to these key
intermediates is outlined in Scheme 1. The acrylic acid 2a was pre-
pared by a Knoevenagel reaction, similar to the procedure previ-
ously reported for 2b.20 The reaction proceeded from the non-
commercially available diacid, which was produced from diethyl
2-(phenethyl)malonate by an ester hydrolysis. Attempts to per-
form the epoxidation directly on 2a,b were not successful. Thus,
a fast and convenient one-pot microwave-heated (MW) protocol
for esterification was developed, in order to generate the corre-
sponding ethyl ester (3a,b). The esters were then readily oxidized
by 3-chloroperbenzoic acid (mCPBA) to provide the epoxides
4a,b. The building blocks, 1a,b, were obtained by hydrolysis of
the ester. This last step proved to be more successful using etha-
nolic KOH under dry conditions than using aqueous NaOH, thus
avoiding concomitant epoxide opening by the hydroxide.
Pf-infected erythrocyte cultures (IC50 = 6.8 lM), a shortcoming
which partly might be explained by insufficient cell/vacuole mem-
brane permeability.18 A closely related compound, KNI-764, has
been co-crystallized in PmPM4, showing an unexpected binding
mode with the prime side of the compound binding to the non-
prime sub-pockets in the active site of the enzyme.13
Previous experiences with a tert-hydroxy containing TS mimic
for inhibition of another aspartic protease, HIV-1 protease, proved
the possibility of retaining high potency, while improving mem-
brane permeability in the Caco-2 cell assay.19–21 One of the HIV-
1
protease inhibitors comprising
Mzs27255,19 also showed some inhibitory activity on PfPM4
(Ki = 12 M) (Fig. 1). These findings prompted us to evaluate the
concept of utilizing masked tertiary alcohols for inhibition of
plasmepsins by synthesizing novel inhibitors with an -substi-
tuted norstatine core of type A (Fig. 2). In the phenylnorstatines
(e.g., KNI-10006 and KNI-764), the -hydroxyl is tethered to the
P1-phenyl by a two-carbon backbone linker, while in -phenyl-
norstatines, the benzyl group is moved to the -carbon adjacent
to the carbonyl and the linker is reduced to only one carbon. There-
a
tertiary alcohol, (S)-
l
Two of the P10 moieties, (R)- and (S)-5,5-dimethylthiazolidine
(L
- and
acids - and
described in Scheme 2, from
tively, by a novel MW-assisted procedure inspired by the thiazoli-
D
-Dmt), derive from the non-naturally occurring amino
-Dmt carboxylic acid (5). They were synthesized, as
- and -penicillamine (Pen), respec-
a
L
D
L
D
a
a
a
O
O
O
fore, we decided to synthesize and evaluate inhibitors with an
a-
a, b)
EtO
c)
OEt
OH
Bn
Bn
2a
β α
99%
e)
HO
O
O
O
O
O
O
O
∗
∗
O
d)
N
H
N
H
N
OH
OEt
OEt
OH
O
O
OH
R1
R2
R1
R1
R1
R1
Y
R2
2a: R1 = Bn
3a: R1 = Bn, 83%
4a: R1 = Bn, 81%
1a: R1 = Bn, 99%
P2'
2b: R1 = Ph
3b: R1 = Ph, 86%
4b: R1 = Ph, 87%
1b: R1 = Ph, 95%
P2
P1
P1'
A
Scheme 1. Synthesis of the P1 moieties. Reagents and conditions: (a) NaOH, H2O/
THF/MeOH (1:2:1.2), reflux, 30 min; (b) HCHO (aq), HNEt2, rt, 30 min then reflux,
3 h; (c) 2 or 2-benzylacrylic acid, SOCl2, EtOH, 80 °C (MW), 5 min; (d) mCPBA,
CH2Cl2, 18 h; (e) KOH (EtOH), rt, 15 min.
Figure 2. Generic structure A of the synthesized
with variable length of the tether to the P1-phenyl group, variations in the P10-
group and two tunable stereocentra ( ).
a-substituted norstatine inhibitors
*

K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
5935
HO
on the coupling products of 1b and
equipped with a conventional C8 column proved to be superior.
The absolute configuration of the quaternary -carbon in P1 of
L
- and
D
-6, where RP-HPLC
O
O
O
H
N
H
N
H2N
HS
a)
b)
*
*
*
OH
N
H
OH
a
S
S
epoxide (S)-11 was unambiguously established using X-ray crys-
tallographic analysis of a crystal prepared from the corresponding
inhibitor (S)-17 in a saturated MeOH solution. The absolute config-
uration of (S)-10 was determined by preparing the compound from
enantiomerically pure (S)-1a, synthesized and optically character-
ized as previously reported.20 The retention times on SP-HPLC and
RP-HPLC of the two epimers of 10 were compared with the reten-
tion times for the diastereomerically pure (S)-epoxide. 1H NMR and
Pen
L-
5
L- , 95%
L-6, 75%
D-6, 65%
Pen
D-
5
D- , 76%
Scheme 2. Microwave-assisted synthesis of P10P20 structures encompassing Dmt.
Reagents: (a) HCHO (aq), Pyr, 110 °C (MW), 5 min; (b) (1S,2R)-1-amino-2-indanol,
HOBt, EDC, CH2Cl2, rt, 1–3 h.
dine-synthesis developed by Howard-Lock et al.23 The elevated
temperature (110 °C), provided by MW heating of the sealed reac-
tion vial, resulted in a reduction of the reaction time from 16 h to
5 min.
The ring formation step was followed by a condensation reac-
tion with (1S,2R)-1-amino-2-indanol, using HOBt and EDC as cou-
pling reagents, furnishing (R)- and (S)-N-((1S,2R)-1-indanyl)-5,5-
13C NMR experiments indicated that the analogous
a-benzylnors-
tatine epimers (S)-9 and (R)-9 eluted with the same retention order
on RP-HPLC as 10, which corresponded to the reversed elution or-
der on SP-HPLC (Reprosil NR-R column). Based on the NMR shifts
and coupling constants and the results from chromatography
experiments, the first eluted stereoisomer using SP-HPLC was as-
signed as (R)-9 and the second eluted stereoisomer as (S)-9. 1H
NMR and 13C NMR data and retention times on SP-HPLC and RP-
HPLC of the corresponding inhibitors supported this hypothesis.
A similar procedure was applied when establishing the absolute
configuration of (S)-12 and (R)-12, but then in comparison with
the data of (S)-11 and (R)-11. The two other epimeric pairs 13,14
were denoted (A) and (B) according to their elution order on SP-
HPLC equipped with Reprosil NR-R column.
dimethylthiazolidine-4-carboxamide (L- and D-6) in 75% and 65%
yields after silica flash chromatography. The nitrogen in the thia-
zolidine was not protected prior to the amidation, thus reducing
the number of reaction steps by two. Although some dimer of
Dmt could be detected in the reaction mixture, it was produced
in very limited amounts according to LC–MS analysis.
The two other P10P20 motifs (
same coupling protocol from Boc-protected
tected -phenylalanine with (1S,2R)-1-amino-2-indanol (Scheme
L
-7 and
L
-8) were produced by the
-proline and Boc-pro-
L
The diastereomerically pure epoxides 9–14, were subsequently
opened with a large excess of aqueous ammonia in methanol, pro-
L
3). The amide coupling was quantitative in both cases, but due to
problems in the work-up after the deprotection, the overall yields
viding the N-unsubstituted
a-benzyl and a-phenylnorstatines as
synthetic intermediates. The volatiles were evaporated in vacuo
and the crude product was used without any further purification
in the next step. The final inhibitors (15–20) were afforded by
the condensation with the P2 group, 2-(2,6-dimethylphenoxy)ace-
tic acid (21), which had been synthesized according to a literature
procedure.24 PyPOB was employed as coupling agent and di-iso-
propyl-ethylamine (DIEA) as base. The stereopure inhibitors were
isolated by RP-HPLC in 4–67% yield, as summarized in Table 2.
were merely 29% for L-7 and 37% for L-8.
The next step in the synthesis was to couple the P1 epoxy acid
with the P10P20 building block by a second amidation reaction, as
depicted in Table 1. Due to the weak nucleophilic character of
the Dmt-nitrogen, the amide couplings between 1a,b and
-6 to yield epoxides 9–12 were extremely sluggish or did not
work at all with a set of different coupling reagents, such as
HOBt/EDC, HATU and PyBOP, used in various reaction media. Nev-
ertheless, with PyAOP, the aza-analogue of PyBOP, in CH2Cl2 or
DMF, the reactions proceeded reasonably well (Table 1), although
full conversion of 6 was never realized, despite more than a week
of stirring. The highest over-all yield for the amidation process was
L- and
D
2.2. Enzyme inhibition
The twelve peptidomimetic structures 15–20 were evaluated
for efficacy and selectivity by their ability to inhibit the paralogous
enzymes PfPM2 and PfPM4, the orthologues of the latter, PmPM4,
PoPM4, PvPM4, as well as hCatD. The results from the enzyme as-
says are presented as Ki values in Table 2.
Although only relatively minor structural modifications were
made on this series of compounds compared to the lead structure
KNI-10006 (Fig. 1), the drop in inhibitory activity on PfPM2, PfPM4
observed with the proline derivative
tion between the less sterically hindered
L
-7 (62%). For the condensa-
-8 and 1a, PyBOP was
L
used to afford epoxide 14. After the condensation with the P10P20
moieties, complete epimeric separation of all the epoxides (9–14)
was made possible with chromatography. The epoxy acid used in
the reaction dictated which separation method had to be applied
for successful separation of the diastereomers. When 1a was used,
epimeric separation was enabled by SP-HPLC equipped with a chi-
ral column (Reprosil NR-R). This method, however, did not apply
and hCatD was substantial.
a-Phenylnorstatines (S)- and (R)-16
was found to be more than 40,000 times less potent on PfPM2 than
KNI-10006. This assay value suggests that the extra sterical con-
gestion, caused by moving the P1 side-chain in the norstatine-scaf-
fold from the b-carbon to the
enzyme. The corresponding
was a moderately potent inhibitor of PfPM2 (Ki = 1.36
which might be explained by the increased flexibility of the P1
side-chain. Interestingly, the two most efficacious inhibitors of
a
-carbon, is not accepted by the
-benzylnorstatine analogue (S)-15
M), a fact
HO
HO
a
O
O
O
Boc
N
Boc
N
H
N
l
a)
a)
b)
OH
N
H
N
H
L
-Boc-Pro-OH
L
-Boc-7, quant.
L
-7, 29%
PfPM2 were (S)-17 (Ki = 0.44
l
M) and (S)-18 (Ki = 0.83
lM), both
featuring the new P10 group
D-Dmt and with (S)-configuration of
HO
HO
O
O
O
the quaternary
Generally, the affinities to PfPM4 were somewhat disappoint-
ing. With Ki values around 0.7 M, compounds (S)- and (R)-16
were the most potent inhibitors of PfPM4, an enzyme which clearly
prefers -phenylnorstatine (R1 = Ph) in the P1 position. The corre-
sponding -benzylnorstatine analogues (R1 = Bn), (S)- and (R)-15,
were very moderate inhibitors with Ki values of 5.46 and
3.11 M, respectively. The same trend could be deduced from the
a-carbon in P1.
Boc
HN
Boc
HN
b) H2N
OH
Ph
-Boc-Phe-OH
N
H
Ph
N
H
Ph
l
L
L
-Boc-8, quant.
L-8, 37%
a
a
Scheme 3. Synthesis of P10P20 structures encompassing proline and phenylalanine.
Reagents: (a) (1S,2R)-1-amino-2-indanol, HOBt, EDC, CH2Cl2, rt, 1 h; (b) HCl (4 M,
1,4-dioxane), rt, 5 h.
l

5936
K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
Table 1
Synthesis and separation of the epimeric pairs of epoxides 9–14
HO
O
O
Amide
Couplinga
6
L-
D-
L-
L-
∗
O
: R1 = Bn
: R1 = Ph
N
H
N
6
1a
1b
∗
7
8
R2
R1
Y
R2
9 14
(S)- and (R)- -
Cmpd
Y
R1
R2
From acid/amine
Coupling methoda
Separation methodb
Yieldc (%)
(S)-9
(R)-9
SCH2–
SCH2–
SCH2–
SCH2–
SCH2–
SCH2–
SCH2–
SCH2–
CH2CH2–
CH2CH2–
CH2Ph
CH2Ph
Bn
Bn
Ph
Ph
Bn
Bn
Ph
Ph
Bn
Bn
Bn
Bn
Me
Me
Me
Me
Me
Me
Me
Me
H
1a/
1a/
1b/
1b/
1a/
1a/
1b/
1b/
1a/
1a/
1a/
1a/
L
-6
-6
-6
-6
-6
-6
-6
1
1
2
2
1
1
2
2
1
1
3
3
1
1
2
2
1
1
2
2
1
1
1
1
24
25
31
45
77
49
50
34
61
62
31
29
L
(S)-10
(R)-10
(S)-11
(R)-11
(S)-12
(R)-12
(A)-13
(B)-13
(A)-14
(B)-14
L
L
D
D
D
D
-6
L
-7
-7
-8
-8
H
H
H
L
L
L
a
Coupling methods: (1) PyAOP, DIEA, in DCM, rt, 2–5 days; (2) As in (1), but in DMF; (3) PyBOP, DIEA, in DCM, rt, two days.
Separation methods: (1) column chromatography (silica, EtOAc/iHex) followed by SP-HPLC with a chiral column; (2) RP-HPLC with a C8 column.
Isolated yields calculated from amine 6–8 with the assumption that 1 was a racemate. The maximum yield is thus 100%.
b
c
Table 2
Synthesis of inhibitors 15–20 and biochemical data with the best value of each enzyme indicated in bolda
HO
HO
O
O
O
O
O
1. NH3 (aq), MeOH
∗
∗
O
∗
N
H
N
H
N
N
H
N
O
∗
2. 21, PyBOP,
DIEA,
CH2Cl2/DMF
OH
R1
R2
R2
R1
Y
Y
R2
R2
Cmpd
Structure
From
epoxide
Yieldb
PfPM2c Ki
M)
PfPM4 Ki
M)
PmPM4c Ki
M)
PoPM4c Ki
M)
PvPM4c Ki
M)
hCatDc Ki
(lM)
(%)
35
(
l
(
l
(l
(
l
(l
HO
O
O
O
O
O
N
H
N
H
N
N
(S)-15
(S)-9
1.4
7.4
3.1
5.5
7.9
4.5
6.5
0.37
>10.0
>10.0
OH
S
S
HO
O
O
O
O
O
N
H
N
H
(R)-15
(S)-16
(R)-16
(R)-9
57
62
37
>20
2.2
OH
HO
O
O
N
H
N
H
N
(S)-10
(R)-10
>20
>20
0.70
0.72
0.26
0.25
0.11
0.16
0.19
>10.0
>10.0
OH
S
HO
O
O
O
O
O
O
O
N
H
N
H
N
N
0.53
OH
S
S
HO
O
N
H
N
H
(S)-17
(S)-11
55
0.44
6.3
1.1
1.0
0.39
>10.0
OH

K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
5937
Table 2 (continued)
Cmpd
Structure
From
epoxide
Yieldb
(%)
PfPM2c Ki
M)
PfPM4 Ki
M)
PmPM4c Ki
M)
PoPM4c Ki
M)
PvPM4c Ki
M)
hCatDc Ki
(lM)
(
l
(
l
(
l
(
l
(
l
HO
O
O
O
O
O
O
N
H
N
H
N
N
(R)-17
(R)-11
(S)-12
(R)-12
(A)-13
(B)-13
(A)-14
(B)-14
67
52
57
47
4
7.6
10.7
1.6
0.88
0.34
0.16
0.12
2.2
>10.0
>10.0
>10.0
>10.0
n.d.
OH
S
HO
O
O
N
H
N
H
(S)-18
(R)-18
(A)-19
(B)-19
(A)-20
0.83
1.7
0.21
0.42
14.5
13.0
n.d.
n.d.
1.3
OH
S
S
HO
O
O
O
O
O
O
O
N
H
N
H
N
N
1.7
1.7
0.61
OH
HO
O
N
H
N
H
>20
>20
n.d.
n.d.
30.2
26.9
>20
>20
7.2
OH
HO
O
O
O
O
O
O
O
O
O
O
N
H
N
H
N
10.3
n.d.
n.d.
1.3
OH
HO
O
N
H
N
H
N
H
53
49
n.d.
n.d.
>10.0
>10.0
OH
HO
O
N
H
N
H
N
H
(B)-20
OH
a
R1 = Bn or Ph, R2 = Me or H, Y = SCH2–, CH2CH2– or CH2Ph.
Isolated yield.
n.d., not determined.
b
c
inhibitors incorporating the other stereoisomer in P10,
D
-Dmt. Com-
the best inhibitor of the
(Ki = 0.21 m), was found to be just barely lower. All four
nylnorstatines (16 and 18) were more active than the -benzyl-
norstatines (15 and 17). In fact, the four -phenylnorstatine
a
-substituted norstatines, (S)-18
paring the preference of the enzyme for the P10 pseudoresidue,
there was no significant difference between the pairs, but the over-
l
a-phe-
a
all tendency was that the compounds incorporating
L
-Dmt as P10
a
were more active than the corresponding -Dmt analogues (see
D
diastereomers were more or less equipotent, regardless of the ste-
epimers 15/16 and 17/18). The impact of the absolute configura-
tion of the TS-mimicking hydroxyl group was of subordinate
importance for the inhibition efficacy on PfPM4.
reochemistry in either the P10 or P1 position. On the other hand,
with R1 = Bn in P1,
substituent.
D
-configuration was preferred for the P10
There is no published inhibition data for PmPM4 of the parent
compound, KNI-10006, but KNI-764 (Fig. 1), encompassing identi-
The
inhibitor of PoPM4. The Ki value of 0.11
tency in the entire series of new inhibitors, not only for PoPM4, but
a
-phenylnorstatine (S)-16 was found to be the most active
lm represents the best po-
cal P1 and P10 groups, has a Ki value of 0.11 m.13 The potency of
l

5938
K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
HO
for all the assessed enzymes. Compound (S)-16 incorporated
in P10, which was the preferred P10 motif for all
-phenylnorsta-
tines. Nonetheless, if R1 = Bn, the compounds with -configuration
of the Dmt group ((R)- and (S)-17) were significantly more potent
and the Ki values were ꢀ1 M, compared to 6.5 M and 4.5 M for
(R)- and (S)-15, respectively.
L-Dmt
O
O
O
a
D
∗
∗
O
N
H
N
H
N
OH
Ph
S
l
l
l
P2
P1
P1'
P2'
All tested inhibitors were active on PvPM4. It was also the en-
zyme that exhibited the least variation in the Ki data, ranging from
Binding mode 1
Binding mode 2
Binding mode 3
S2'
S2'
S2
S1'
S1
S1
S1
S2
S2
S2'
S1'
S1'
0.12
slightly more potent (Ki = 0.12–0.19
tines, which also were active in the nanomolar range (Ki = 0.34–
l
M ((R)-18) to 2.2
lM ((A)-19). The
a-phenylnorstatines were
l
M) than the
a-benzylnorsta-
Figure 3. Proposed binding orientations of the herein reported a-phenylnorstatines
in the binding site of PM4 enzymes.
0.39 lM), except for compound (R)-15 (Ki = 2.2 lM).
As depicted in Table 2, the l-Pro P10 moiety found in (A)- and
(B)-19 was not well accommodated in any of the enzymes. The
two compounds with L
-Phe as P10 motif ((A)- and (B)-20) were only
tested on PfPM4, on which they were not active. Because of these
negative results, they were not further evaluated on the other
iso-enzymes.
None of the compounds were found to inhibit hCatD within
measurable concentrations, which however was limited to
(e.g., binding mode 3 in Fig. 3).28 Additionally, the binding mode
of (R)-phenylnorstatines to HIV-1 protease was found to differ in
the H-bond pattern of the TS-mimic as compared to (S)-derivatives,
such as KNI compounds.29 Therefore, we felt encouraged to inves-
tigate the binding poses of the newly designed compounds with an
10.0
lM, due to the inherent fluorescence of the inhibitors. Yet,
a-phenylnorstatine scaffold (i.e., R = Ph in Table 2). The four inhib-
this series of compounds bound at least 5 000 times less strongly
to hCatD than the lead compound KNI-10006. Inhibitor (S)-16
exhibited more than a 14- to 90-fold selectivity for PM4 over hCatD
and inhibitor (S)-17 >20,000 times more active on PfPM2 than
hCatD, compared to four times for KNI-10006. In addition, com-
pounds (S)-16, (R)-16, (S)-18, (R)-18, (A)-20 and (B)-20 were as-
sessed on the human aspartic protease drug-target b-amyloid
converting enzyme (BACE-1), but none of them proved to be active
itors (S)-16, (R)-16, (S)-18 and (R)-18 were systematically docked
into the active site of PmPM4, the enzyme of choice since the crys-
tal structure is pre-arranged to bind norstatine inhibitors. The
binding affinity estimations were instead conducted on PfPM4,
which allowed us to compare the results with previously experi-
mentally17 and computationally27 determined data for KNI-10006.
The docking results into PmPM4 indicated that only one binding
orientation could be considered for compounds (R)-16 and (R)-18
(binding mode 1, Fig. 3). This binding mode corresponds to the ori-
entation of KNI-764, observed by the X-ray crystallography in
PmPM4,13 and modeled for KNI-10006.10,27 Conversely, two differ-
ent binding orientations were isolated from the docking runs of the
in concentrations below 10 lm. This further confirmed our belief
that this series of plasmepsin inhibitors do not affect human aspar-
tic proteases to a considerable extent. Finally, all compounds, ex-
cept for (B)-19, were also tested for their efficacy to inhibit HIV-
1, either in an isolated protease assay or on an infected cell-line.
two (S)-a-phenylnorstatines ((S)-16 and (S)-18), featuring the
Only (R)-18 (Ki = 0.67
the isolated protease assay, and (S)-15 in the cell-based assay
(EC50 = 6.4 M).
l
M) and (B)-20 (Ki = 1.0
l
M) were active in
opposite absolute configuration of the TS mimic. The first orienta-
tion is analogous to the binding mode (1) (Fig. 3), as discussed be-
fore for the (R)-a-phenylnorstatines and also for the parent
l
From this dataset of measured Ki values in Table 2 some general
trends can be deduced: (1) There was no significant correlation be-
tween the efficacy for which the compounds inhibit PfPM2 and
PfPM4 or its orthologues. Similar patterns have previously been
observed both for compounds of pseudopeptidic character12,22,25
phenylnorstatine KNI compounds, whereas the second positioning
involves an alternative binding mode which was more frequently
observed for the two compounds considered, (S)-16 and (S)-18
(binding mode 2 in Fig. 3). In this second orientation, the P1 and
P10 inhibitor side chains interchange their positions in the corre-
sponding S1 and S10 binding pockets, with respect to the binding
mode experimentally observed for KNI-764.
In order to gain more knowledge about the ligand-protein inter-
actions and to choose between the two possible binding modes ob-
served for the compounds with (S)-a-phenylnorstatines, an MD
and for more drug-like peptidomimetics.8 (2) The
tines were stronger inhibitors of the four PM4 than the
norstatines. (3) The presence of
-Dmt in P10 reduced the
a-phenylnorsta-
a
-benzyl-
D
influence of the structural characteristics of P1 on the inhibition
of PM4 orthologues in Po, Pv and Pm. Thus, neither an alteration
of the length of the tether nor the stereochemistry of the
in P1 resulted in a change of the Ki values. In comparison, the com-
pounds encompassing -Dmt demonstrated a greater variation in
a
-carbon
simulation was performed for every binding pose selected at the
docking stage, as bound to PfPM4. Additionally, we considered
manual docking as starting pose for the MD simulation for every
compound in two alternative orientations. In the first orientation,
the compounds were built on the basis of KNI-764 as found in
the 2ANL structure (binding mode 1). This conformation could only
L
potency depending on the character of the P1 moiety.
2.3. Computational evaluation
be accomplished for the (R)-a-phenylnorstatines. The second ori-
As an attempt to explain the considerably lower affinities to
entation is depicted in binding mode (3) (Fig. 3), and corresponds
to a totally inverted binding orientation relative to the binding
mode (1). Even if this binding mode was not found in the docking
runs, we considered this orientation since other authors had sug-
gested that alternate binding orientations could exist for different
KNI inhibitors.28 Between two and six binding poses per inhibitor
were simulated, providing a total of 20 starting poses for further
evaluation.
Binding affinities were systematically calculated by the linear
interaction energy method (LIE) and the most energetically favored
binding pose of each molecule ((S)-16, (R)-16, (S)-18 and (R)-18), is
shown in Table 3. Although we did not attempt to explain the abso-
PfPM4 experienced for the
a-phenylnorstatines compared to the
lead structure KNI-10006, several computational methods were
employed. An exhaustive docking search was conducted, followed
by molecular dynamics (MD) and binding affinity estimations. This
computational protocol has been proven successful in the design of
other plasmepsin inhibitors.26
Even if the binding orientation of KNI-764 in PmPM4 has been
elucidated by X-ray crystallography (PDB entry 2ANL)13 (Fig. 3,
binding mode 1) and shown to explain the differences in affinity
for a series of KNI compounds,27 other alternative binding modes
of norstatine-based compounds have been proposed recently

K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
5939
Table 3
Experimental and calculated energetics (kcal/mol) of PfM4 inhibition of the four
a
-phenylnorstatines under study and a b-phenylnorstatine, KNI-10006
Cmpd
D
Gbind, exp (kcal/mol)
D
Gbind, LIE (kcal/mol)
Ligand–surrounding interactions (kcal/mol)
v
dw
el
v
dw
el
hVl—s ip
hV ip
hVl—s iw
hV iw
l—s
l—s
(R)-16
(S)-16
(R)-18
ꢁ8.44
ꢁ8.45
ꢁ7.92
ꢁ7.92
ꢁ10.7
ꢁ4.57 0.4
ꢁ5.71 0.5
ꢁ5.56 0.8
ꢁ8.99 0.3
ꢁ14.1 0.4
ꢁ76.7 1.1
ꢁ82.2 1.2
ꢁ83.2 0.1
ꢁ81.3 0.1
ꢁ77.2 0.1
ꢁ68.6 0.3
ꢁ69.6 0.1
ꢁ73.0 0.9
ꢁ80.1 0.3
ꢁ93.3 0.7
ꢁꢁ49.3 0.1
ꢁ49.1 0.1
ꢁ48.9 0.4
ꢁ48.1 0.3
ꢁ42.9 0.0
ꢁ68.6 0.3
ꢁ70.4 0.7
ꢁ74.9 1.4
ꢁ72.0 0.5
ꢁ69.4 0.6
(S)-18
KNI-10006a
a
Data reported by Gutierrez-de-Teran et al.27
lute binding affinities in our calculations, the equipotency of the
compound series was relatively correctly reproduced, apart from
the overestimation of the calculated affinity of compound (S)-18.
The data for compound KNI-10006 have been included for compar-
suggested at the docking stage (Fig. 4a and b), whereas for the
inhibitors with the (S)-configuration at the TS mimic ((S)-16, (S)-
18), binding mode (2) was the most favored one (Fig. 4c and d).
The third possibility, binding mode (3) (Fig. 3), showed least fa-
vored binding affinity estimates for the four studied compounds,
hence it was disregarded for further consideration. A detailed anal-
ysis of the ligand binding interactions followed in order to under-
stand the reduced affinity with respect to the original norstatine
scaffold in KNI-10006. Interestingly, the configuration of the thi-
ison.27 For all four
a-phenylnorstatines, the ligand binding affinity
was found mainly to be due to nonpolar contributions, which indi-
cated that polar interactions generally just counterbalanced
desolvation, except in the case of compound (S)-18 where they fa-
vor binding to the enzyme. In contrast, for the phenylnorstatines,
that is, the KNI compounds, previous studies have shown that both
polar and nonpolar contributions account for the comparatively
stronger ligand binding.27 Thus, these results suggest that the
oproline pseudoresidue in P10 (
L or D) exerted a great influence
on the orientation of the two carbonyl oxygens adjacent to it. Com-
pound (R)-16, with the same absolute configuration in P10 as the
KNI compounds, showed the most conserved binding pose with re-
spect to KNI-10006 (Fig. 4a). The main difference was found in the
H-bond network with the catalytic aspartates (Asp34 and Asp214).
The TS-mimicking a-hydroxyl interacted with Asp34 while the P1
carbonyl adjacent to it pointed in the opposite direction. In fact, the
interaction with the catalytic aspartates resembled the conforma-
introduction of the a-substituent in the norstatine scaffold impairs
the strong polar interactions of the catalytic aspartates with the
TS-mimicking hydroxyl and the central P1 carbonyl group, previ-
ously observed with norstatine-containing plasmepsin inhibitors.
The binding affinity calculations indicated that inhibitors (R)-16
and (R)-18 should bind to PfPM4 according the binding mode 1, as
Figure 4. Structures of four PfPM4 enzyme–inhibitor complexes, as average coordinates of the MD trajectories sampled for binding free energy calculations. The enzyme is
depicted with the prime side subsites to the left and the non-prime subsites to the right. From top, left corner: (a) (R)-16 in orange, (b) (R)-18 in pink, (c) (S)-16 in purple, (d)
(S)-18 in cyan. Yellow dashed lines indicate frequently observed hydrogen bonds.

5940
K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
tion of the (R)-phenylnorstatine derivative crystallized with HIV-1
protease by Brynda el al.29 In our simulations, this carbonyl was
interacting with Thr217 in the S1 subsite, as depicted in Figure
4a. Inhibitor (R)-18, with the inverse configuration in P10, that is,
tures reported in this contribution are the first drug-like peptidom-
imetic compounds which have been tested and proven to be
efficient inhibitors in the nanomolar range of the PM4 iso-en-
zymes. It can be stated, in general terms, that the inhibitors of
PfPM4 are active also on the DV plasmepsins of the other species.
Nonetheless, despite the small differences in the active sites of the
four PM4, variations in inhibitory activity were revealed. For exam-
ple, compound (R)-17 was found to be a weak inhibitor of PfPM4,
but provided sub-micromolar activities on both PoPM4 and PvPM4,
thus showing a >30-fold deviation depending on the enzyme
orthologue. Therefore, we believe it is important to measure the
inhibitory effect of new inhibitors in all four PM4 enzymes. The
experimental and computational information reported herein
might have implications for future design of new PM4 inhibitors.
D
-Dmt, exhibited the same binding pose in the prime subsites,
whereas the conformation of the P10P20 pseudoresidues in the
non-prime sites of the enzyme was slightly altered (Fig. 4b). Nev-
ertheless, the H-bond network was satisfied with the P1 carbonyl
interacting with Gly76 at the tip of the flap and the P10 carbonyl
with Thr217. For the two (S)-a-phenylnorstatines, the binding
mode in the S20 site was well conserved with respect to KNI-764,
while major differences were discovered in the other three pock-
ets. The two compounds had the carbonyl at P1 pointing towards
the tip of the flap and, depending on the configuration of the thi-
oproline, the carbonyl in P10 pointing towards the inner part of
the enzyme, interacting with Thr217 (P10 =
L
-Dmt, (S)-16) or also
towards the flap (P10 =
and d, respectively.
D-Dmt, (S)-18) as depicted in Figure 4c
4. Experimental
4.1. Chemistry
3. Conclusions
4.1.1. General information
The microwave experiments were conducted using Emrys Initi-
atorTM or Smith SynthesizerTM single mode cavity, producing con-
trolled irradiation at 2450 MHz in sealed reaction vials (capable
of withholding elevated pressure) dedicated for microwave pro-
cessing, under air and with magnetic stirring. The temperature of
the reaction mixture was measured using a built-in, on-line infra-
red temperature sensor. For column chromatography commer-
cially available Silica Gel 60 (particle size: 0.040–0.063 mm) was
used. Analytical thin layer chromatography was done using glass
sheets precoated with Silica Gel 60 F254 and visualization of com-
ponents was made by UV, I2 and/or Pancaldi’s solution followed
by heating. Analytical HPLC–MS was performed on a Gilson–Finn-
egan ThermoQuest AQA system equipped with a C18 (Onyx Mono-
lithic C18 (50 ꢂ 4.6 mm)) column, using CH3CN in 0.05% aqueous
HCOOH as mobile phase at a flow rate of 4 mL/min with UV (214
and 254 nm) and MS (ESI+ or ESIꢁ) detection. HPLC–MS was com-
plemented with RPLC on a Gilson HPLC Pump 312 system with a
LKB variable wavelength detector equipped with a C8 (Agilent Zor-
A new class of plasmepsin inhibitors encompassing an a-substi-
tuted norstatine TS-isostere scaffold was prepared. Variations of
the P10 pseudoresidue and alterations of the stereochemistry and
the size of the P1
a-substituent provided twelve stereopure
a-phe-
nylnorstatine and
a-benzylnorstatine test compounds. The investi-
gation afforded a handful of drug-like inhibitors of Pf-, Pm-, Po-
and PvPM4, with not only Ki values in the nanomolar range, but
also with high selectivity versus the human aspartic proteases
hCatD and BACE-1. Furthermore, the incorporation of the
motif in the inhibitors furnished a reduced sensitivity for structural
modifications of the -substituent in the norstatine core, that is,
the length of the P1 tether and the absolute configuration of the
-carbon. However, the best over-all inhibitor of PM4 was LL-Dmt
comprising (S)-16 (Ki = 0.70, 0.26, 0.11 and 0.16 M on Pf, Pm, Po
D-Dmt
a
a
l
and Pv, respectively), and with a more than 90- to 14-fold selectiv-
ity over hCatD. Note that the importance of the absolute configura-
tion of the
a-carbon seems to be very small, as proven by the
almost equipotent PM4 inhibition profile experienced with epi-
meric (R)-16.
bax SB-C8 (5
l
m, 50 ꢂ 4.6 mm)) column, using CH3CN in 0.05%
aqueous HCOOH or 0.10% aqueous TFA as mobile phase at a flow
rate of 2 mL/min with UV (220 nm) detection. Preparative RP–
HPLC was performed on a Gilson HPLC Pump 312 with a LKB sys-
tem variable wavelength detector equipped with a C8 (Agilent Zor-
Although the new inhibitors were most active in the PM4 ortho-
logue assays, two
counted for relatively good inhibitory potency against PfPM2
(Ki = 0.44 M and 0.88 M, respectively). Hence, the -Dmt moiety
was identified as a new promising P10 residue in the active site also
of PfPM2, being more potent compared than the ‘natural’ -Dmt
motif. Being highly active in all five malaria enzyme assays,
Dmt containing -phenylnorstatine (S)-18 can be stated as the
most adaptive inhibitor in the whole series.
The observed loss of affinity to PfPM4, due to the move of the P1
substituent to the -position in the norstatine scaffold, as com-
D-Dmt based compounds ((S)-17 and (A)-18) ac-
l
l
D
bax SB-C8 (5
lm, 50 ꢂ 21.2 mm)) column, using CH3CN in 0.05%
aqueous HCOOH or 0.10% aqueous TFA as mobile phase at a flow
rate of 5 mL/min with UV (220 or 230 nm) detection. Specific rota-
tions (aD) are reported in deg/dm and the concentration (c) is given
L
D
-
a
as g/100 mL in the specified solvent. 1H and 13C NMR spectra were
recorded on Varian Mercury Plus instruments at 24 °C, 1H NMR
spectra were recorded at 399.9 MHz and 13C NMR spectra at
100.5 MHz. The chemical shifts for 1H NMR and 13C NMR were ref-
erenced to TMS via the solvent signals (1H NMR at 7.26 ppm for
CDCl3 and at 3.31 for CD3OD and 13C NMR at 77.16 for CDCl3 and
49.00 for CD3OD). Exact molecular masses (HRMS) were deter-
mined on Micromass Q-Tof2 mass spectrometer equipped with
an electrospray ion source. Analytische Laboratorien, Lindlar, Ger-
many and Microkemi, Uppsala, Sweden, performed elemental anal-
yses. The crystallization and X-ray determination of inhibitor (S)-
17 were performed at the Latvian Institute of Organic Synthesis,
Latvia, Riga. CCDC724531 contain the Supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk. The purity for all new compounds were
established to be ꢃ98% by at least two HPLC systems, 1H NMR,
13H NMR and elemental analysis or HRMS. All starting materials,
a
pared to the b-substituted lead compound, KNI-10006, could be
explained by less favorable polar interactions with the enzyme.
In addition, a comprehensive computational exploration of the
binding poses of this novel class of a-phenylnorstatine derivatives
to PmPM4 and PfPM4 was carried out. The computational results
suggested that the compounds with the (R) configuration of the
TS-mimicking P1 present the same binding orientation as observed
for KNI-764, whereas the compounds with an inverted (S) config-
uration interchange the position of the P1 and P10 substituents in
their respective S1/S10 binding pockets.
To the best of our knowledge, there are presently only three
publications available reporting inhibitors of all four orthologues
of PM4.12,22,30 The structures included there are all large and of
pseudopeptidic character. Thus, the a-substituted norstatine struc-

K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
5941
reagents and solvents were purchased from commercial suppliers
and were used directly as received. Compounds 2b31 and 2124 were
synthesized according to literature procedures.
(275 mL) and mCPBA (77% in balance with 3-chlorobenzoic acid and
water, 48.240 g, 195.7 mmol) was added. The reaction mixture was
refluxed under stirring. After 4 h, additional mCPBA (10.0 g,
40.0 mmol) was added and the mixture refluxed for another 14 h.
The reaction mixture was washed with 10% Na2S2O3 (aq), sat HCO3
(aq) (4 ꢂ 150 mL), reduced under vacuum and purified (silica,
EtOAc/iHex, 1:20–1:2). to yield 4a as an uncolored syrup (3.487 g
(81%). 1H NMR (400 MHz, CDCl3): d = 7.32–7.26 (m, 2H), 7.22–7.17
(m, 3H), 4.23 (dd, J = 10.8, 7.2 Hz, 1H), 4.20 (dd, J = 10.8, 7.2 Hz,
1H), 3.05 (d, J = 5.9 Hz, 1H), 2.84–2.79 (m, 2H), 2.77 (d, J = 5.9 Hz,
1H), 2.43–2.33 (m, 1H), 2.06–1.86 (m, 1H), 1.30 ppm (t, J = 7.2 Hz,
3H); 13C NMR (100 MHz, CDCl3): d = 170.3, 141.1, 128.6, 128.5,
126.3, 61.8, 56.7, 52.3, 33.5, 31.3, 14.2 ppm; MS-ESI+: m/z 221,
[M+H]+.
4.1.2. Experimental details and spectroscopic data for
compounds 1a,b–4a,b
Compound 2a is essentially synthesized according to a litera-
ture procedure for the benzyl analogue 2b.20,31 Compounds 3a32
and 4a33 are known from the literature, but the microwave-as-
sisted esterification is a new procedure resulting in a higher yield.
There are no previously reported spectroscopic data for 3a and 4a.
The synthesis of enantiomerically pure epoxy acid 1b is described
in the literature,20 but results in a lower over-all yield (58%) com-
pared to the racemic method described herein (71%).
4.1.2.1. 2-Methylene-4-phenylbutyric acid (2a). NaOH (10.40 g,
260 mmol) was dissolved in water (25.0 mL) and diethyl 2-(phen-
ylethyl)malonate (13.09 g, 49.53 mmol), THF (50.0 mL) and MeOH
(30.0 mL) were added. The reaction mixture was refluxed under
stirring for 30 min. The solvents were evaporated under vacuum
and the residue was dissolved in water (400 mL) and acidified to
pH <1 by addition of 2 M HCl (aq). The reaction mixture was ex-
tracted with EtOAc (3 ꢂ 250 mL). The combined organic phases
were dried over MgSO4, filtered and evaporated to yield phenyl-
ethyl malonic acid (10.5 g, quantitative) as a white solid. The prod-
uct was used without any further purification in the next step. MS-
ESIꢁ: m/z 207, [M–H]+. Phenylethyl malonic acid (10.4 g,
49.5 mmol) was mixed with aqueous formaldehyde (37% wt,
22.0 mL, 290 mmol) and diethylamine (57 mL, 55 mmol) was
added dropwise. The reaction mixture was stirred at rt for
30 min until a white solid was formed and then refluxed for 3 h.
The reaction mixture was reduced under vacuum, acidified with
5 M HCl (aq) and extracted with EtOAc (3 ꢂ 100 mL). The com-
bined organic phases were dried over MgSO4, filtered, evaporated
and dried under vacuum over-night to yield 2a (8.728 g, 99%) as
a off-white solid. 1H NMR (400 MHz, CDCl3): d = 7.33–7.13 (m,
5H, Ar), 6.32 (s, 1H), 5.62 (s, 1H), 2.84 (d, J = 7.2 Hz, 1H), 2.82 (d,
J = 5.6 Hz, 1H), 2.64 (d, J = 5.6 Hz, 1H), 2.62 ppm (d, J = 7.2 Hz,
1H); 13C NMR (100 MHz, CDCl3): d = 172.4, 141.4, 139.4, 128.6,
128.5, 128.0, 126.2, 34.9, 33.7 ppm; MS-ESI– (m/z 175, M–H+).
4.1.2.5. 2-Benzyloxirane-2-carboxylic acid ethyl ester (4b).
Compound 4b was prepared according to the procedure for compound
4a using 3b (1.40 g, 7.36 mmol) and mCPBA (8.47 g, 36.8 mmol) to
yield 4b (1.32 g, 87%) as an uncoloured syrup.
4.1.2.6. 2-Phenethyloxirane-2-carboxylic acid (1a). Compound
4a (2.05 g, 9.29 mmol) was dissolved in EtOH (13 mL) and freshly
prepared 1.22 M ethanolic KOH (13.0 mL, 15.9 mmol) was added
under vigorous stirring. The reaction mixture was stirred at room
temperature and after 15 min a white solid was formed. The reac-
tion mixture was diluted with satd NaHCO3 (5 mL), acidified with
5 M HCl (aq) to pH 8, washed with EtOAc (3 ꢂ 75 mL), acidified
to pH <1 by addition of 5 M HCl (aq) and extracted with EtOAc
(3 ꢂ 150 mL). The combined organic phases were dried over
MgSO4, filtered and evaporated to yield 1a as a pale yellow syrup
(1.763 g, 99 %): 1H NMR (400 MHz, CDCl3): d = 7.33–7.27 (m, 2H),
7.23–7.18 (m, 3H), 3.11 (d, J = 5.7 Hz, 1H), 2.87 (d, J = 5.7 Hz, 1H),
2.85–2.81 (m, 2H), 2.43–2.33 (m, 1H), 2.06–1.86 ppm (m, 1H);
13C NMR (100 MHz, CDCl3): d = 175.1, 140.8, 128.7, 128.5, 126.4,
56.6, 52.6, 33.0, 31.2 ppm; MS-ESI–: m/z 191, [M–H]+. Anal. Calcd
for C11H12O3ꢄ1/4H2O: C, 61.16; H, 6.40. Found: C, 67.0; H, 6.4.
4.1.2.7. 2-Benzyloxirane-2-carboxylic acid (1b).20 Compound
1b was prepared according to the procedure for compound 1a
using 4b (1.30 g, 6.30 mmol) and 1.22 M ethanolic KOH (11 mL,
13.4 mmol) to yield 1b (1.15 g, 95%) as thick paste.
4.1.2.2. 2-Methylene-4-phenylbutyric acid ethyl ester (3a).32
Compound 2a (3.01 g, 17.08 mmol) and thionylchloride (1.6 mL,
22 mmol) were stirred for 2 min in a capped 10–20 mL Emrys reaction
vial. Absolute EtOH (10.00 mL) was added dropwise with a syringe.
The reaction mixture was irradiated with microwaves for 5 min at
80 °C. Another 1.6 mL (22 mmol) thionylchloride was added by syringe
and the vial was irradiated for an additional 5 min at 80 °C. The reaction
mixture was reduced under vacuum, diluted with EtOAc (150 mL),
washed with saturated NaHCO3 (aq) (3 ꢂ 50 mL), 0.5 M HCl (aq)
(50 mL) and brine. The organic phase was reduced and purified (silica,
EtOAc/iHex, 1:20) to yield 3a (2.895 g, 83%) as a clear oil. 1H NMR
(400 MHz, CDCl3): d = 7.32–7.26 (m, 2H), 7.22–7.17 (m, 3H), 6.17–6.16
(m, 1H), 5.55 (dd, J = 2.7, 1.4 Hz, 1H), 4.23 (q, J = 7.2 Hz, 2H), 2.82 (d,
J = 7.2 Hz, 1H), 2.79 (d, J = 5.6 Hz, 1H), 2.64 (d, J = 5.6 Hz, 1H), 2.62 (d,
J = 7.2 Hz, 1H), 1.32 ppm (t, J = 7.2 Hz, 3H); 13C NMR (100 MHz, CDCl3):
d = 167.3, 141.6, 140.3, 128.6, 128.5, 126.1, 125.2, 60.8, 35.1, 34.1,
14.4 ppm; MS-ESI+: m/z 205, [M+H]+.
4.1.3. Experimental details and spectroscopic data for
compounds
The synthesis of compounds
erature but with a lower isolated yield.23 The synthesis of the N-
L
-5,
D
-5,
L
-6,
D
-6, 7 and 8
L-5 and D-5, are described in the lit-
Boc derivative of compound L
-6 is described in the literature.34
The syntheses of 735, 8 and Boc-836 are reported in the literature,
but with alternative procedures. The spectroscopic data of the
known substances matched with the literature, when reported,
otherwise we have reported the data below. Analytical SP-HPLC
of
LKB system variable wavelength detector equipped with a (Repro-
sil Chiral NR-R (8
L-6 and D-6 was performed on a Gilson HPLC Pump 312 with
l
m, 250 ꢂ 4.6 mm)) column, using iPrOH in iHex
as mobile phase at a flow rate of 2 mL/min with UV (220 nm)
detection.
4.1.3.1. (R)-5,5-Dimethyl-1,3-thiazolidine-4-carboxylic acid (
L-
5). -Penicillamine (0.50 g, 3.50 mmol) was suspended in 1.50 mL
L
4.1.2.3. 2-Benzylacrylic acid ethyl ester (3b). Compound 3b was
prepared according to the procedure for compound 3a using 2b
(1.00 g, 6.17 mmol) and thionylchloride (1.5 + 0.5 mL, 21 + 7 mmol)
to yield 3b (1.00 g, 86%) as a clear oil.
aqueous formaldehyde (37% wt, 20.0 mmol), three drops of pyri-
dine and 2.5 mL ethanol in a 2–5 mL Emrys reaction vial. The vial
was irradiated by microwaves for 5 min (hold-time) at 110 °C.
After cooling for >1 h, the product had crystallized and could be fil-
tered off and washed with cold Et2O to give compound L-5 as white
4.1.2.4. 2-Phenethyloxirane-2-carboxylic acid ethyl ester
crystals (0.516 g, 95%). MS-ESI+: m/z 162, [M+H]+.
(4a).33 Compound 3a (4.00 g, 19.60 mmol) was dissolved in CH2Cl2

5942
K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
4.1.3.2. (S)-5,5-Dimethyl-1,3-thiazolidine-4-carboxylic acid (
5). Compound -5 was synthesized by the same procedure as -5
from -penicillamine (0.50 g 3.50 mmol), but 2 mL Et2O was added
to the reactions vial after MW-irradiation and cooling to rt, in order
D
-
concentrated and purified (silica, 3–5% MeOH in CH2Cl2) to give
Boc-8 (1.0 g, quantitative) as a white solid. 1H NMR (400 MHz,
CD3OD): d = 7.32–7.27 (m, 4H), 7.27–7.15 (m, 5H), 5.29 (d,
J = 5.1 Hz, 1H), 4.50 (dt, J = 5.1, 1.6 Hz, 1H), 4.41 (dd, J = 9.7, 5.0 Hz,
1H), 3.25 (dd, J = 13.8, 5.0 Hz, 1H), 3.13 (dd, J = 16.3, 5.1 Hz, 1H),
2.908 (dd, J = 16.3, 1.6 Hz, 1H), 2.907 (dd, J = 13.8, 9.7 Hz, 1H),
1.36 ppm (s, 9H); 13C NMR (100 MHz, CD3OD): d = 174.7 (2 C@O),
142.2, 141.7, 138.9, 129.4, 128.4, 127.8, 127.7, 126.1, 125.4, 80.7,
73.9, 58.6, 57.7, 40.8, 39.0, 28.6 ppm; MS-ESI+: m/z 347, [M+H]+.
D
L
D
to initiate the crystallization. The product,
D-5 was isolated as
white crystals (0.412 g, 76%). MS-ESI+: m/z 162, [M+H]+.
4.1.3.3. (R)-N-((1S,2R)-2-Hydroxy-1-indanyl)-5,5-dimethylthiaz-
olidine-4-carboxamide
(L
-6). Carboxylic acid
L-5 (0.454 g,
2.82 mmol) and (1S,2R)-1-amino-2-indanol (0.511 g, 3.43 mmol),
EDC (0.602 g, 3.14 mmol), HOBT (0.425 g, 3.14 mmol) were stirred
in dry CH2Cl2 (20 mL) for 3 h at rt. The reaction mixture was con-
4.1.3.8. (S)-N-((1S,2R)-2-Hydroxy-1-indanyl)-3-phenylpropana-
mide (8). Compound Boc-8 (1.00 g, 2.52 mmol) was dissolved in
dioxane (20 mL) and 4 M HCl (1,4-dioxane) (12.5 mL, 50 mmol)
was added. The mixture was stirred for 5 h at rt. The volatiles were
evaporated in vacuo and the residue was dissolved in DCM and ex-
tracted with satd NH4Cl (aq). The combined aqueous phases were
basified to pH 11 with 1 M NaOH (aq) and the product was ex-
tracted with EtOAc to give pure 8 (0.273 g, 37%) as a white solid.
MS-ESI+: m/z 247, [M+H]+.
centrated and purified (silica, 2% MeOH in CH2Cl2) to give
L
-6
-6: tR = 5.49 min (SP-
= ꢁ31.4 (c 0.98, MeOH: CH2Cl2
(0.608 g, 75%) as a white solid. Compound
L
HPLC, 15% iPrOH in iHex); ½a D22
ꢅ
(15:1)); 1H NMR (400 MHz, CD3OD): d = 7.29–7.17 (m, 4H), 5.36
(d, J = 5.4 Hz, 1H), 4.54 (dt, J = 5.4, 2.6 Hz, 1H), 4.31 (d, J = 9.4 Hz,
1H), 4.14 (d, J = 9.4 Hz, 1H), 3.57 (s, 1H), 3.15 (dd, J = 16.3, 5.4 Hz,
1H), 2.94 (dd, J = 16.3, 2.6 Hz, 1H), 1.68 (s, 3H), 1.38 ppm (s, 3H);
13C NMR (100 MHz, CD3OD): d = 171.4, 142.2, 141.8, 129.1, 127.9,
126.2, 125.5, 75.4, 73.8, 58.3, 57.9, 51.5, 40.8, 29.1, 28.1 ppm;
MS-ESI+: m/z 293, [M+H]+.
4.1.4. Experimental details and spectroscopic data for epimeric
epoxides (A)-9–14 and (B)-9–14
4.1.4.1. General procedure for amide coupling reactions. Cor-
responding carboxylic acid, corresponding amine and PyAOP were
suspended in dry CH2Cl2, stirred for a few minutes and thereafter
diisopropylethylamine was added. The reaction mixture was stir-
red at rt for 2–5 days. Evaporation afforded the crude products,
which were, in the case of 9,11,13, and 14, purified by column
chromatography (silica, EtOAc/iHex) prior to the epimeric separa-
tion. The epimeric separations of (R)-9/(S)-9, (R)-11/(S)-11, (A)-
13/(B)-13 and (A)-14/(B)-14 were performed on a Gilson HPLC sys-
4.1.3.4. (S)-N-((1S,2R)-2-Hydroxy-1-indanyl)-5,5-dimethylthiaz-
olidine-4-carboxamide
(D-6). Carboxylic acid
D-5 (0.335 g,
1.24 mmol) and (1S,2R)-1-amino-2-indanol (0.301 g, 1.87 mmol),
EDC (0.395 g, 2.06 mmol), HOBT (0.278 g, 2.06 mmol) were stirred
in dry CH2Cl2 (20 mL) for 1 h at rt. The reaction mixture was con-
centrated and purified (silica, 3% MeOH in CH2Cl2) to give
D
-6
-6: tR = 4.73 min (SP-
= +53.6 (c 1.37, MeOD); 1H NMR
(0.354 g, 65%) as a white solid. Compound
D
HPLC, 15% iPrOH in iHex); ½a D22
ꢅ
(400 MHz, CD3OD): d = 7.29–7.20 (m, 4H), 5.32 (d, J = 5.2 Hz, 1H),
4.59 (dt, J = 5.2, 2.0 Hz, 1H), 4.34 (d, J = 9.3 Hz, 1H), 4.15 (d,
J = 9.3 Hz, 1H), 3.56 (s, 1H), 3.16 (dd, J = 16.4, 5.2 Hz, 1H), 2.94
(dd, J = 16.4, 2.0 Hz, 1H), 1.67 (s, 3H), 1.39 ppm (s, 3H); 13C NMR
(100 MHz, CD3OD): d = 171.6, 141.99, 141.96, 129.1, 127.9, 126.3,
125.1, 75.2, 73.9, 58.9, 57.9, 51.7, 40.6, 29.3, 28.2 ppm; MS-ESI+:
m/z 293, [M+H]+.
tem equipped with a chiral (Reprosil Chiral NR-R (8 lm,
250 ꢂ 20 mm)) column, using iPrOH in iHex as mobile phase at a
flow rate of 5 mL/min with UV (230 nm) detection. The epimeric
separation was analyzed on a Gilson HPLC system equipped with
a chiral (Reprosil Chiral NR-R (8
l
m, 250 ꢂ 4.6 mm)) column, using
iPrOH in iHex as mobile phase at a flow rate of 2 mL/min or 1.5 mL/
min, as indicated, with UV (220 nm) detection. The epimeric sepa-
ration of (R)- and (S)-10 and (S)- and (R)-11 was performed on a
Gilson–Finnigan ThermoQuest AQA system C8 (Agilent Zorbax
4.1.3.5. (S)-1-(tert-Butoxycarbonyl)-N-((1S,2R)-2-hydroxy-1-inda-
nyl)pyrrolidine-5-carboxamide (Boc-7). Boc-(S)-proline (1.01 g,
4.67 mmol) and (1S,2R)-1-amino-2-indanol (0.837 g, 5.613 mmol),
PyBOP (2.673 g, 5.136 mmol), DIEA (1.60 mL, 9.346 mmol) were stir-
red in dry CH2Cl2 (80 mL) for 3 h at rt. The reaction mixture was con-
centrated and purified (silica, EtOAc/iHex, 1:1–5:1) to give Boc-7
(1.62 g, quantitative) as a white solid. 1H NMR (400 MHz, CD3OD):
d = 7.31–7.16 (m, 4H), 5.33–5.24 (m, 1H), 4.55 (s, 1H), 4.37–4.30 (m,
1H), 3.57–3.48 (m, 1H), 3.48–3.39 (m, 1H), 3.14 (dd, J = 16.4, 5.1 Hz,
1H), 2.97–2.88 (m, 1H), 2.34–2.18 (m, 1H), 2.18–2.06 (m, 1H), 2.06–
1.86 (m, 2H), 1.43 ppm (s, 9H); 13C NMR (100 MHz, CD3OD):
d = 175.7, 175.3, 142.2, 141.6, 129.0, 127.7, 126.1, 125.4, 81.6, 79.1,
73.6, 61.7, 58.5, 47.8, 32.6, 28.7, 24.6 ppm; MS-ESI+: m/z 347, [M+H]+.
SB-C8 (5
lm, 50 ꢂ 21.2 mm)) column, using CH3CN in 0.05% aque-
ous HCOOH as mobile phase at a flow rate of 15 mL/min with UV
(214 nm) detection and analyzed on a Gilson-Finnigan Thermo-
Quest AQA system equipped with a C18 (Onyx Monolithic C18
(50 ꢂ 4.6 mm)) column, using CH3CN in 0.05% aqueous HCOOH
as mobile phase at a flow rate of 4 mL/min with UV (214 and
254 nm) and MS (ESI+ or ESI–) detection. At room temperature,
the NMR spectra of (S)-9–12 and (R)-9–12 and (A)- and (B)-13 indi-
cate the presence of two conformers, due to the cis-trans-isomer-
ism of the thioproline and the proline P1–P10 amide bond in
various ratios (6:1–1:1), which has been confirmed by 1H NMR
experiments at a temperature range of 24–60 °C. The shifts at
24 °C of the major conformer are reported, except for (S)-9 and
(R)-12, for which all signals at 24 °C are reported.
4.1.3.6. (S)-N-((1S,2R)-2-Hydroxy-1-indanyl)pyrrolidine-5-car-
boxamide (7). Compound Boc-7 (0.730 g, 2.11 mmol) was dis-
solved in dioxane (11 mL) and 4 M HCl (1,4-dioxane) (10.5 mL,
42 mmol) was added. The mixture was stirred for 5 h at rt. The
reaction mixture was concentrated and purified (silica, EtOAc/
MeOH/NEt3, 200:2:1) to give 7 (0.150 g, 29%) as a white solid.
MS-ESI+: m/z 247, [M+H]+.
4.1.4.2. (4R)-N-((1S,2R)-2-Hydroxy-1-indanyl)-5,5-dimethyl-3-
(2-phenethyloxirane-2-carbonyl)thiazolidine-4-carboxamide
((R)-9 and (S)-9). Compound 9 was synthesized according to the
general procedure using
0.427 mmol), PyAOP (240 mg, 0.460 mmol) and DIEA (165
L
-6 (94.0 mg, 0.321 mmol), 1a (82.0 mg,
l,
l
0.964 mmol). It was purified (silica, EtOAc/iHex, 1:3–3:1) to give
9 (80.0 mg, 53%) as a diastereomeric mixture. The epimers were
then separated and further purified by chiral SP-HPLC to give (R)-
9 (18.5 mg) and (S)-9 (18.3 mg) as white solids (totally 36.8 mg,
25%). Compound (S)-9: tR = 14.3 min (SP-HPLC, 15% iPrOH in iHex,
2.0 mL/min); rotamer ratio (5:4); 1H NMR (400 MHz, CD3OD): ma-
4.1.3.7. (S)-1-(tert-Butoxycarbonylamino)-N-((1S,2R)-2-hydroxy-
1-indanyl)-3-phenylpropanamide (Boc-8). Boc-(S)-phenylalanine
(669 mg 2.52 mmol), (1S,2R)-1-amino-2-indanol (458 mg, 3.07
mmol), HOBt (380 mg, 2.81 mmol), EDC (539 mg, 2.81 mmol) were
stirred in dry CH2Cl2 (50 mL) for 1 h at rt. The reaction mixture was

K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
5943
jor conformer: d = 7.45–7.41 (m, 1H), 7.27–7.13 (m, 8H), 5.38 (dd,
J = 5.1, 1.2 Hz, 1H), 5.06 (d, J = 9.8 Hz, 1H), 4.92 (d, J = 9.8 Hz, 1H),
4.52 (dt, J = 5.1, 2.0 Hz, 1H), 4.50 (s, 1H), 3.15 (dd, J = 16.3, 5.1 Hz,
1H), 3.03 (d, J = 4.8 Hz, 1H), 2.932 (dd, J = 16.3, 2.0 Hz, 1H), 2.90
(d, J = 4.8 Hz, 1H), 2.86–2.79 (m, 2H), 2.35–2.27 (m, 1H), 2.15–
2.07 (m, 1H), 1.57 (s, 3H), 1.53 ppm (s, 3H); Minor conformer:
d = 7.32–7.27 (m, 1H), 7.27–7.13 (m, 8H), 5.38 (dd, J = 5.1, 1.2 Hz,
1H), 4.91 (d, J = 10.3 Hz, 1H), 4.87 (s, 1H), 4.71 (d, J = 10.3 Hz,
1H), 4.52 (dt, J = 5.1, 2.0 Hz, 1H), 3.14 (dd, J = 16.3, 5.1 Hz, 1H),
2.932 (dd, J = 16.3, 2.0 Hz, 1H), 2.928 (d, J = 5.0 Hz, 1H), 2.92 (d,
J = 5.0 Hz, 1H), 2.86–2.79 (m, 1H), 2.70 (ddd, J = 14.2, 11.8, 5.5 Hz,
1H), 2.46 (ddd, J = 14.2, 11.8, 5.4 Hz, 1H), 1.86 (ddd, J = 14.1, 11.7,
5.2 Hz, 1H), 1.58 (s, 3H), 1.54 ppm (s, 3H); 13C NMR (100 MHz,
CD3OD): major conformer: d = 171.0, 170.1, 142.4, 142.1, 141.60,
129.52, 129.46, 128.90, 127.1, 126.1, 125.9, 74.07, 74.0, 61.0,
58.8, 52.32, 52.1, 50.2, 41.07, 36.0, 31.7, 30.5, 25.1 ppm; minor con-
former: d = 171.6, 171.3, 142.4, 142.3, 141.58, 129.59, 129.3,
128.91, 127.82, 127.3, 126.2, 125.7, 74.10, 72.9, 61.3, 58.5, 52.31,
52.1, 49.5, 41.11, 36.6, 32.2, 31.8, 25.2 ppm; MS-ESI+: m/z 467,
[M+H]+. Anal. Calcd for C26H30N2O4Sꢄ2H2O: C, 62.13; H, 6.82; N,
5.57. Found: C, 63.29; H, 5.95; N, 5.45. Compound (R)-9:
tR = 8.8 min (SP-HPLC, 15% iPrOH in iHex, 2.0 mL/min); rotamer ra-
tio (4:1); 1H NMR (400 MHz, CD3OD) d = 7.41–7.35 (m, 1H), 7.30–
7.11 (m, 8H), 5.36 (d, J = 5.2 Hz, 1H), 4.95 (d, J = 10.4 Hz, 1H),
4.89 (d, J = 10.4 Hz, 1H), 4.53 (s, 1H), 4.50 (dt, J = 5.2, 5.2, 1.0 Hz,
1H), 3.14 (dd, J = 16.3, 5.2 Hz, 1H), 3.04 (d, J = 5.0 Hz, 1H), 2.929
(d, J = 5.0 Hz, 1H), 2.926 (dd, J = 16.3, 2.0 Hz, 1H), 2.87–2.69 (m,
2H), 2.54 (ddd, J = 14.4, 10.6, 5.8 Hz, 1H), 1.91 (ddd, J = 14.4, 10.3,
6.2 Hz, 1H), 1.62 (s, 3H), 1.50 ppm (s, 3H); 13C NMR (100 MHz,
CD3OD) d = 170.9, 170.3, 142.2, 142.1, 141.6, 129.53, 129.44,
129.0, 127.8, 127.2, 126.1, 125.8, 74.1, 73.9, 60.5, 58.5, 52.3, 52.0,
50.2, 41.0, 35.7, 31.9, 29.0, 25.7 ppm; MS-ESI+: m/z 467, [M+H]+.
Anal. Calcd for C26H30N2O4Sꢄ1/5 H2O: C, 65.42; H, 6.52; N, 5.97.
Found: C, 65.86; H, 5.86; N, 5.55.
40.3, 29.0, 25.3 ppm; MS-ESI+: m/z 453, [M+H]+; HRMS-ESI+: m/z
[M+H]+ calcd for C25H29N2O4S: 453.1848, found 453.1855.
4.1.4.4. (4S)-N-((1S,2R)-2-Hydroxy-1-indanyl)-5,5-dimethyl-3-
(2-phenethyloxirane-2-carbonyl)thiazolidine-4-carboxamide
((S)-11 and (R)-11). Compound 11 was synthesized according to
the general procedure using
(92.0 mg, 0.427 mmol), PyAOP (324 mg, 0.621 mmol) and DIEA
(182 l, 1.06 mmol). It was purified (silica, EtOAc/iHex, 1:3–3:1)
D-6 (94.0 mg, 0.322 mmol), 1a
l
to give 11 as a diastereomeric mixture. The epimers were then sep-
arated and further purified by chiral HPLC to give (S)-11 (57.6 mg)
and (R)-11 (39.8 mg) as white solids (totally 36.8 mg, 65%). Com-
pound (S)-11: tR = 9.4 min (SP-HPLC, 15% iPrOH in iHex, 1.5 mL/
min); rotamer ratio (3:1); 1H NMR (400 MHz, CD3OD): d = 7.29–
7.13 (m, 9H), 5.30 (d, J = 5.0 Hz, 1H), 4.96 (d, J = 10.3 Hz, 1H), 4.90
(d, J = 10.3 Hz, 1H), 4.59 (dt, J = 5.0, 1.6 Hz, 1H), 4.49 (s, 1H), 3.14
(dd, J = 16.5, 5.1 Hz, 1H), 2.97–2.95 (m, 1H), 2.91 (d, J = 5.0 Hz,
1H), 2.89 (d, J = 5.0 Hz, 1H), 2.84–2.67 (m, 3H), 2.52–2.43 (m,
1H), 1.96–1.86 (m, 1H), 1.62 (s, 3H), 1.52 ppm (s, 3H); 13C NMR
(100 MHz, CD3OD): d = 170.9, 170.7, 142.0, 141.9, 141.8, 129.51,
129.45, 129.1, 127.8, 127.2, 126.2, 125.3, 74.2, 73.9, 60.5, 59.1,
52.3 51.9 50.1, 40.6, 35.6, 31.9, 29.2, 25.6 ppm; MS-ESI+: m/z 467,
[M+H]+. Anal. Calcd for C26H30N2O4Sꢄ1/3H2O: C, 66.08; H, 6.54; N,
5.93. Found: C, 66.28; H, 6.55; N, 5.82. Compound (R)-11:
tR = 11.8 min (SP-HPLC, 15% iPrOH in iHex, 1.5 mL/min); rotamer
ratio (3:1); 1H NMR (CD3OD) d = 7.32–7.08 (m, 9H), 5.33 (d,
J = 5.0 Hz, 1H), 5.03 (d, J = 10.0 Hz, 1H), 4.93 (d, J = 10.0 Hz, 1H),
4.62 (dt, J = 5.0, 1.5 Hz, 1H), 4.44 (s, 1H), 3.14 (dd, J = 16.4, 5.0 Hz,
1H), 2.96 (d, J = 4.8 Hz, 1H), 2.94 (dd, J = 16.4, 1.5 Hz, 1H), 2.87 (d,
J = 4.8 Hz, 1H), 2.29–2.20 (m, 1H), 2.06–1.98 (m, 1H), 1.57 (s, 3H),
1.54 ppm (s, 3H); 13C NMR (100 MHz, CD3OD): d = 171.5, 170.1,
142.2, 141.94, 141.87, 129.58, 129.38, 129.0, 127.8, 127.1, 126.2,
125.4, 74.3, 73.8, 61.0, 59.2, 52.24 51.9 50.2, 40.6, 36.1, 31.8,
30.5, 24.8 ppm; MS-ESI+: m/z 467, [M+H]+. Anal. Calcd for
C26H30N2O4Sꢄ1/2H2O: C, 65.66; H, 6.57; N, 5.89. Found: C, 65.57;
H, 6.58; N, 5.89.
4.1.4.3. (4R)-N-((1S,2R)-2-Hydroxy-1-indanyl)-5,5-dimethyl-3-
(2-benzyloxirane-2-carbonyl)thiazolidine-4-carboxamide ((R)-
10 and (S)-10). Compound 10 was synthesized according to the
4.1.4.5. (4S)-N-((1S,2R)-2-Hydroxy-1-indanyl)-5,5-dimethyl-3-
(2-benzyloxirane-2-carbonyl)thiazolidine-4-carboxamide ((S)-
12 and (R)-12). Compound 12 was synthesized according to the
general procedure using
(68.2 + 50.0 mg, 0.383 + 0.280 mmol), PyAOP (202.0 + 101.0 mg,
0.388 + 0.194 mmol), DIEA (180 l, 1.05 mmol) and 1 mL DMF as
L-6 (96.8 mg, 0.33 mmol), 1b
l
general procedure using
0.798 mmol), PyAOP (415 mg, 0.796 mmol), DIEA (250
D
-6 (93.8 mg, 0.320 mmol), 1b (142 mg,
l,
co-solvent. The epimers were then separated and purified by RP-
HPLC to give (R)-10 (33.3 mg) and (S)-10 (23.4 mg) as white solids
(totally 56.7 mg, 38%). Compound (S)-10: tR = 1.79 min (RP-HPLC,
10–90%, 4.2 min); tR = 7.90 min (SP-HPLC, 15% iPrOH in iHex,
2.0 mL/min); rotamer ratio (7:4); 1H NMR (400 MHz, CD3OD):
d = 7.39–7.16 (m, 9H), 5.37 (d, J = 5.0 Hz, 1H), 4.83 (d, J = 10.3 Hz,
1H), 4.50 (dt, J = 5.0, 1.8 Hz, 1H), 4.22 (s, 1H), 4.07 (d, J = 10.3 Hz,
1H), 3.71 (d, J = 14.2 Hz, 1H), 3.14 (dd, J = 16.4, 5.0 Hz, 1H), 3.00
(d, J = 5.2 Hz, 1H), 2.97 (d, J = 5.2 Hz, 1H), 2.93 (dd, J = 16.4,
1.8 Hz, 1H), 2.58 (d, J = 14.2 Hz, 1H), 1.39 (s, 3H), 0.82 ppm (s,
3H); 13C NMR (100 MHz, CD3OD): d = 171.7, 170.8, 142.2, 141.5,
137.4, 130.9, 129.9, 129.0, 128.5, 127.8, 126.2, 125.5, 74.1, 73.1,
62.6, 58.3, 52.2, 50.1, 42.2, 41.2, 40.7, 30.7, 24.4 ppm; MS-ESI+:
m/z 453, [M+H]+; HRMS-ESI+: m/z [M+H]+ calcd for C25H29N2O4S:
453.1848, found 453.1844. Compound (R)-10: tR = 1.86 min (RP-
HPLC, 10–90%, 4.2 min); tR = 7.88 min (SP-HPLC, 15% iPrOH in iHex,
2.0 mL/min); rotamer ratio (10:1); 1H NMR (400 MHz,
CD3OD:CDCl3): d = 7.36–7.14 (m, 9H), 5.33 (dd, J = 5.2, 1.1 Hz,
1H), 4.79 (d, J = 10.2 Hz, 1H), 4.48 (dt, J = 5.2, 1.9 Hz, 1H), 4.39 (s,
1H), 4.04 (d, J = 10.2 Hz, 1H), 3.65 (d, J = 14.2 Hz, 1H), 3.13 (dd,
J = 16.4, 5.2 Hz, 1H), 3.09 (d, J = 5.1 Hz, 1H), 2.94 (d, J = 5.1 Hz,
1H), 2.91 (dd, J = 16.4, 1.9 Hz, 1H), 2.68 (d, J = 14.2 Hz, 1H), 1.41
(s, 3H), 1.40 ppm (s, 3H); 13C NMR (100 MHz, CD3OD:CDCl3):
d = 170.8, 170.2, 142.0, 141.4, 136.8, 130.6, 129.7, 128.9, 128.4,
127.8, 126.0, 125.6, 74.1, 73.8, 61.9, 58.3, 52.6, 51.5, 49.5, 41.0,
l
1.46 mmol) and 1 mL DMF as co-solvent. The epimers were then
separated and purified by RP-HPLC to give pure (S)-12 (36.2 mg)
and (R)-12 (24.8 mg) and mixed epimers as white solids (totally
82.8 mg, 57%). Compound (S)-12: tR = 2.09 min (RP-HPLC, 10–
90%, 4.2 min); tR = 6.90 min (SP-HPLC, 15% iPrOH in iHex, 2.0 mL/
min); rotamer ratio (6:1), 1H NMR (400 MHz, CD3OD): d = 7.34–
7.16 (m, 9H), 5.28 (d, J = 5.0 Hz, 1H), 4.80 (d, J = 10.2 Hz, 1H), 4.56
(dt, J = 5.0, 1.6 Hz, 1H), 4.38 (s, 1H), 4.07 (d, J = 10.2, 1H), 3.63 (d,
J = 14.2, 1H), 3.14 (ddd, J = 16.4, 5.0, 1.1 Hz, 1H), 2.932 (d,
J = 5.05 Hz, 1H), 2.926 (ddd, J = 16.4, 1.6, 0.5 Hz, 1H), 2.91 (d,
J = 5.05 Hz, 1H), 2.69 (d, J = 14.2 Hz, 1H), 1.42 (s, 3H), 1.41 ppm
(s, 3H); 13C NMR (100 MHz, CD3OD): d = 170.9, 170.7, 141.9,
141.8, 137.0, 130.7, 129.8, 129.1, 128.5, 127.8, 126.3, 125.2, 74.2,
73.9, 62.1, 59.0, 52.6, 51.6, 49.3, 40.6, 40.3, 29.1, 25.3 ppm; MS-
ESI+: m/z 453, [M+H]+; HRMS-ESI+: m/z [M+H]+ calcd for
C25H29N2O4S: 453.1848, found 453.1859. Compound (R)-12:
tR = 2.16 min (RP-HPLC, 10–90%, 4.2 min); tR = 6.92 min (SP-HPLC,
15% iPrOH in iHex, 2.0 mL/min); rotamer ratio (1:1), the lower shift
in brackets; 1H NMR (400 MHz, CD3OD): d = 7.35–7.18 (m, 9+9H),
5.37 (d, J = 5.0 Hz, 1H) (5.25 (d, J = 5.2 Hz, 1H)), 4.80 (d, J = 9.8 Hz,
1H) (4.63 (d, J = 10.6 Hz, 1H)), 4.61 (dt, J = 5.0, 1.5 Hz, 1H) (4.56
(dt, J = 5.2, 2.1 Hz, 1H)), 4.44 (d, J = 9.8 Hz, 1H) 4.58 (d,
J = 10.6 Hz, 1H), 4.35 (s, 1H) (4.23 (s, 1H)), 3.61 (d, J = 14.2, 1H)
(3.05 (d, J = 14.3, 1H)), 3.30 (d, J = 14.2, 1H) (2.61 (d, J = 14.3,

5944
K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
1H)), 3.20–3.12 (m, 2H), 3.00 (d, J = 4.9 Hz, 1H) (2.92 (d, J = 4.9 Hz,
1H)), 2.96 (dd, J = 16.4, 1.5 Hz, 1H) (2.94 (dd, J = 16.4, 2.1 Hz, 1H)),
2.69 (d, J = 4.9 Hz, 1H) (2.84 (d, J = 4.9 Hz, 1H)), 1.52 (s, 3 H), 1.49 (s,
3H), 1.38 (s, 3H), 0.84 ppm (s, 3H); 13C NMR (100 MHz, CD3OD):
d = 172.0 (171.7), 171.0 (170.0), 142.04 (141.99), 141.9, 137.2
(136.0), 131.1 (130.9), 129.9 (129.7), 129.13 (129.06), 128.5
(128.3), 127.83 (127.81), 126.2, 125.9 (125.4), 74.3 (73.9), 73.8
(72.9), 63.1 (61.5), 59.2 (58.9), 54.7 (52.9), 52.45 (52.42), 50.1
(49.5), 40.63 (40.61), 40.54 (39.5), 30.7 (30.3), 25.0 (24.8) ppm;
MS-ESI+: m/z 453, [M+H]+; HRMS-ESI+: [M+H]+ calcd for
C25H29N2O4S: 453.1848, found 453.1841.
found: 471.2287. Compound (B)-14: tR = 16.63 min (SP-HPLC, 20%
iPrOH in iHex, 2.0 mL/min); 1H NMR (400 MHz, CD3OD):
d = 7.31–7.09 (m, 14H), 5.31 (d, J = 5.1 Hz, 1H), 4.77 (dd, J = 10.1,
5.0 Hz, 1H), 4.54 (dt, J = 5.0, 2.1 Hz, 1H), 3.29 (dd, J = 13.9, 5.0 Hz,
1H), 3.14 (dd, J = 16.4, 5.2 Hz, 1H), 2.98 (dd, J = 13.9, 10.1 Hz, 1H),
2.92 (dd, J = 16.4, 2.1 Hz, 1H), 2.72 (d, J = 5.2 Hz, 1H), 2.67–2.61
(m, 2H), 2.50–2.42 (m, 1H), 2.27 (d, J = 5.2 Hz, 1H), 1.54–
1.47 ppm (m, 1H); 13C NMR (100 MHz, CD3OD): d = 173.5, 172.6,
142.5, 142.2, 141.8, 138.5, 130.4, 129.5, 129.38, 129.37, 129.0,
127.89, 127.85, 127.0, 126.2, 125.4, 73.8, 59.7, 58.7, 55.3, 53.6,
40.8, 38.6, 34.3, 31.9 ppm; MS-ESI+: m/z 471, [M+H]+; HRMS-ESI+:
[M+H]+ calcd for C29H31N2O4: 471.2284, found: 471.2289.
4.1.4.6.
(5S)-N-((1S,2R)-2-Hydroxy-1-indanyl)-1-(2-phenethy-
loxirane-2-carbonyl)pyrrolidine-5-carboxamide ((A)-13 and (B)-
13). Compound 13 was synthesized according to the general
4.1.5. Experimental details and spectroscopic data for
inhibitors (S)-15–18, (R)-15–18, (A)-19,20 and (B)-19,20
4.1.5.1. General procedure for epoxide ring opening reac-
tions. Epoxide, 500 equiv of aqueous NH3 (25%) and the equiva-
lent volume of methanol was stirred in room temperature over-
night. The volatiles were evaporated in vacuo and the residue
was used in the proceeding amide coupling without any further
purification.
procedure using
L-7 (70.0 mg, 0.284 mmol), 1a (60.01 mg,
0.3126 mmol), PyAOP (240 mg, 0.460 mmol) and DIEA (160
l
l,
0.94 mmol). It was purified (silica, EtOAc/MeOH, 99:1–98:2) to
give 13 as a diastereomeric mixture. The epimers were then sepa-
rated and further purified by chiral HPLC to give (A)-13 (36.6 mg)
and (B)-13 (37.3 mg) as white solids (totally 73.9 mg, 62%). Com-
pound (A)-13: tR = 12.5 min (SP-HPLC, 15% iPrOH in iHex, 1.5 mL/
min); rotamer ratio (4:1), 1H NMR (400 MHz, CD3OD): d = 7.40–
7.33 (m, 1H), 7.29–7.11(m, 9H), 5.31 (d, J = 5.0 Hz, 1H), 4.55–4.48
(m, 2H), 3.80–3.68 (m, 2H), 3.14 (dd, J = 16.4, 5.1 Hz, 1H), 2.98 (d,
J = 5.1 Hz, 1H), 2.91 (d, J = 5.0 Hz, 1H), 2.96–2.87 (m, 1H), 2.83–
2.70 (m, 2H), 2.28–2.13 (m, 1H), 2.12–1.94 (m, 2H), 1.94–
1.81 ppm (m, 2H); 13C NMR (100 MHz, CD3OD): d = 174.2, 170.9,
142.3, 142.1, 141.5, 129.47, 129.44, 128.9, 127.9, 127.1, 126.1,
125.6, 74.0, 62.1, 60.5, 58.7, 52.2, 49.1, 40.8, 35.6, 31.8, 30.2,
26.4 ppm; MS-ESI+: m/z 421, [M+H]+. Anal. Calcd for
C25H28N2O4ꢄ1/2 H2O: C, 69.91; H, 6.81; N, 6.52. Found: C, 69.89;
H, 6.83; N, 6.19. Compound (B)-13: tR = 17.5 min (SP-HPLC, 15%
iPrOH in iHex, 1.5 mL/min); rotamer ratio (4:3), 1H NMR
(400 MHz, CD3OD): d = 7.42–7.39 (m, 1H), 7.29–7.13 (m, 8H),
5.33 (dd, J = 5.1, 1.0 Hz, 1H), 4.81 (dd, J = 8.2, 1.3 Hz, 1H), 4.51
(dt, J = 5.1, 2.0 Hz, 1H), 3.89–3.76 (m, 2H), 3.15 (dd, J = 16.4,
5.1 Hz, 1H), 2.98 (d, J = 4.9 Hz, 1H), 2.93 (d, J = 16.4 Hz, 1H), 2.88
(d, J = 4.9 Hz, 1H), 2.85–2.79 (m, 1H), 2.79–2.63 (m, 1H), 2.24–
2.00 (m, 5H), 1.89–1.82 ppm (m, 1H); 13C NMR (100 MHz, CD3OD):
d = 174.5, 170.1, 142.44, 142.38, 141.6, 129.5, 129.4, 128.9, 127.9,
127.1, 126.0, 125.8, 74.0, 61.0, 58.8, 52.2, 48.6, 40.8, 36.1, 31.8,
30.3, 26.0 ppm; MS-ESI+: m/z 421, [M+H]+. Anal. Calcd for
C25H28N2O4 ꢄ H2O: C, 68.47; H, 6.90; N, 6.39. Found: C, 68.75; H,
6.90; N, 6.09.
4.1.5.2. General procedure for amide coupling reactions. Car-
boxylic acid, amine and PyBOP were dissolved in dry CH2Cl2, stir-
red for a few minutes and thereafter Hünig’s base (DIEA) was
added. The reaction mixture was stirred at room temp for 24 h.
Evaporation afforded the crude product, which was purified on
RP-HPLC and freeze-dried to give the final products. At room tem-
perature, the NMR spectra of (S)-15–17, (R)-17, (B)-18 and (A)-19
indicate the presence of two conformers, due to the cis-trans-isom-
erism of the thioproline and the proline P1–P10 amide bond in var-
ious ratios (>10:1–9:8), which has been confirmed by 1H NMR
experiments at a temperature range of 24–60 °C. The shifts at
24 °C of the major conformer are reported, except for (S)-16 and
(B)-18 for which all signals are reported.
4.1.5.3. (S)-(4S)-N-((1S,2R)-2-Hydroxy-1-indanyl)-3-(2-phenethyl-
3-(2-(2,6-dimethylphenoxy)acetamido)-2-hydroxypropanoyl)-5,5-
dimethylthiazolidine-4-carboxamide ((S)-15). Compound (S)-15
was synthesized according to the general procedure from (S)-9
(14.5 mg, 0.030 mmol), 21 (7.1 mg, 0.039 mmol), PyBOP (28.6 mg,
0.055 mmol) and DIEA (12.0 ll, 0.070 mmol). 0.30 mL of DMF was
added as co-solvent. It was purified (RP-HPLC) to give (S)-15 as a white
solid (9.83 mg, 51%). Compound (S)-15:tR = 17.83 (SP-HPLC, 15% iPrOH
in iHex); ½a 2D1
ꢅ
= +5.5 (c 0.181, MeOH); rotamer ratio (3:1); 1H NMR
(400 MHz, CD3OD): d = 7.40–7.37 (m, 1H), 7.29–7.10 (m, 8H), 7.04–
6.97 (m, 2H), 6.92 (dd, J = 8.4, 6.4 Hz, 1 H), 5.47 (d, J = 10.4 Hz, 1H),
5.33 (d, J = 5.0 Hz, 1H), 5.17 (d, J = 10.4 Hz, 1H), 4.59 (s, 1H), 4.50 (dt,
J = 5.0, 1.8 Hz, 1H), 4.23 (s, 2H), 3.87 (d, J = 13.7 Hz, 1H), 3.63 (d,
J = 13.7 Hz, 1H), 2.88 (dd, J = 16.3, 5.0 Hz, 1H), 2.85–2.72 (m, 2H), 2.27
(ddd, J = 13.8, 12.6, 4.7 Hz, 1H), 2.24 (s, 6H), 2.00 (ddd, J = 13.8, 12.6,
5.6 Hz, 1H), 1.63 (s, 3H), 1.53 ppm (s, 3H); 13C NMR (100 MHz, CD3OD):
d = 174.5, 171.9, 171.4, 156.0, 143.5, 142.3, 141.5, 131.8, 130.0, 129.5,
129.4, 128.8, 127.8, 126.9, 125.97, 125.95, 125.8, 80.1, 75.8, 74.1,
71.3, 58.9, 50.9, 50.6, 47.1, 40.9, 39.8, 30.8, 29.8, 25.5, 16.5 ppm; MS-
ESI+: m/z 646, [M+H]+; HRMS-ESI+: [M+H]+ calcd for C36H44N3O6S:
646.2951, found: 646.2963. Anal. Calcd for C36H43N3O6Sꢄ2/5 H2O: C,
66.21; H, 6.76; N, 6.43. Found: C, 65.89; H, 6.36; N, 6.09.
4.1.4.7.
(2S)-N-((1S,2R)-2-Hydroxy-1-indanyl)-2-(2-phenethy-
loxirane-2-carbonyl)amino-3-phenylpropylamide ((A)-14 and
(B)-14). Compound 14 was synthesized according to the general
procedure using
0.35 + 0.10 mmol), PyBOP (199 + 68 mg, 0.383 + 0.13 mmol) and
DIEA (180 l, 1.05 mmol). It was purified (silica, EtOAc/iHex, 1:3–
8 (94.5 mg, 0.32 mmol), 1a (67.4 + 20.0 mg,
l
3:1) to give 14 as a diastereomeric mixture. The epimers were then
separated and purified by chiral HPLC to give (A)-14 (23.4 mg) and
(B)-14 (21.7 mg) as white solids (totally 45.1 mg, 30%). Compound
(A)-14: tR = 12.08 min (SP-HPLC, 20% iPrOH in iHex, 2.0 mL/min);
1H NMR (400 MHz, CD3OD): d = 7.34–7.11 (m, 12H), 7.06–7.03
(m, 2H), 5.29 (d, J = 5.2 Hz, 1H), 4.81 (dd, J = 10.3, 4.8 Hz, 1H),
4.52 (dt, J = 5.2, 2.0 Hz, 1H), 3.32 (dd, J = 14.0, 4.8 Hz, 1H), 3.15
(ddd, J = 16.4, 5.2, 1.0 Hz, 1H), 2.99 (dd, J = 14.0, 10.4 Hz, 1H),
2.92 (dd, J = 16.4, 2.0 Hz, 1H), 2.84 (s, 2H), 2.50–2.40 (m, 1H),
2.29–2.19 (m, 1H), 1.54–1.47 ppm (m, 1H); 13C NMR (100 MHz,
CD3OD): d = 173.6, 172.4, 142.5, 142.2, 141.7, 138.5, 130.5, 129.6,
129.4, 129.3, 129.0, 127.90, 127.86, 127.0, 126.2, 125.4, 73.8,
59.9, 58.6, 55.7, 53.8, 40.8, 38.8, 34.4, 31.8 ppm; MS-ESI+: m/z
471, [M+H]+; HRMS-ESI+: [M+H]+ calcd for C29H31N2O4: 471.2284,
4.1.5.4. (R)-(4S)-N-((1S,2R)-2-Hydroxy-1-indanyl)-3-(2-phenethyl-
3-(2-(2,6-dimethylphenoxy)acetamido)-2-hydroxypropanoyl)-5,5-
dimethylthiazolidine-4-carboxamide ((R)-15). Compound (R)-15
was synthesized according to the general procedure from (R)-9
(14.5 mg, 0.030 mmol), 21 (7.1 mg, 0.039 mmol), PyBOP (28.6 mg,
0.055 mmol) and DIEA (12.0
ll, 0.070 mmol). It was purified (RP-
HPLC) to give (R)-15 as a white solid (10.3 mg, 53%). Compound (R)-

K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
5945
15: tR = 21.12 (SP-HPLC, 15% iPrOH in iHex); ½a D21
ꢅ
= ꢁ1.5 (c 0.027,
50.44, 48.9, 44.2, 41.0, 28.7, 25.6, 16.5 ppm; MS-ESI+: m/z 632,
[M+H]+; HRMS-ESI–: [M–H]+ calcd for C35H40N3O6S: 630.2638,
found 630.2640.
MeOH); rotamer ratio (>10:1); 1H NMR (400 MHz, CD3OD): d = 7.35
(d, J = 7.5 Hz, 1H), 7.290–6.86 (m, 11H), 5.39 (d, J = 10.8 Hz, 1H), 5.29
(d, J = 5.0 Hz, 1H), 5.17 (d, J = 10.8 Hz, 1H), 4.54 (s, 1H), 4.48 (dt,
J = 5.0, 1.8 Hz, 1H), 4.26 (d, 14.8 Hz, 1H), 4.14 (d, J = 14.8 Hz, 1H), 3.79
(d, J = 13.7 Hz, 1H), 3.75 (d, J = 13.7 Hz, 1H), 3.02 (dd, J = 16.4, 5.0 Hz,
1H), 2.84 (dd, J = 16.4, 1.8 Hz, 1H), 2.83–2.66 (m, 2H), 2.23–2.14 (m,
1H), 2.21 (s, 6H), 2.03–1.94 (m, 1H), 1.62 (s, 3H), 1.52 ppm (s, 3H);
13C NMR (100 MHz, CD3OD): d = 174.7, 172.0, 171.3, 156.0, 143.2,
141.9, 141.6, 131.8, 130.0, 129.5, 129.4, 128.8, 127.7, 127.0, 126.0,
125.8, 125.7, 79.8, 75.8, 74.0, 71.2, 58.9, 50.9, 50.7, 47.1, 40.9, 39.7,
30.9, 29.1, 25.7, 16.5 ppm; MS-ESI+: m/z 646, [M+H]+; HRMS-ESI+:
[M+H]+ calcd for C36H44N3O6S: 646.2951, found 646.2955. Anal. Calcd
for C36H43N3O6S: C, 66.95; H, 6.71; N, 6.51. Found: C, 65.71; H, 6.83; N,
6.33.
4.1.5.7. (S)-(4S)-N-((1S,2R)-2-Hydroxy-1-indanyl)-3-(2-phenethyl-
3-(2-(2,6-dimethylphenoxy)acetamido)-2-hydroxypropanoyl)-5,5-
dimethylthiazolidine-4-carboxamide ((S)-17). Compound (S)-17
was synthesized according to the general procedure from (S)-11
(37.9 mg, 0.081 mmol), 21 (17.6 mg, 0.098 mmol), PyBOP (63.4 mg,
0.12 mmol) and DIEA (30.6 ll, 0.18 mmol). It was purified (RP-HPLC)
to give (S)-17 as a white solid (28.7 mg, 55%). Compound (S)-17:
tR = 16.52 (SP-HPLC, 15% iPrOH in iHex); ½a D22
ꢅ
= +4.4 (c 0.313, MeOH);
1
rotamer ratio (>10:1); H NMR (400 MHz, CD3OD): d= 7.28–7.13 (m,
9H), 7.04–7.00 (m, 2H), 6.95 (dd, J = 8.4, 6.4 Hz, 1H), 5.35 (d,
J = 10.8 Hz, 1H), 5.22 (d, J = 10.8 Hz, 1H), 5.22 (dd, J = 4.9, 1.0 Hz, 1H),
4.53 (s, 1H), 4.46 (dt, J = 4.9, 1.7 Hz, 1H), 4.30 (d, J = 14.8 Hz, 1H), 4.23
(d, J = 14.8 Hz, 1H), 3.78 (d, J = 13.8 Hz, 1H), 3.73 (d, J = 13.8 Hz, 1H),
2.82 (dd, J = 16.5, 4.9 Hz, 1H), 2.74 (dd, J = 16.5, 1.7 Hz, 1H), 2.79–2.64
(m, 2H), 2.27 (s, 6H), 2.19 (ddd, J = 13.9, 12.1, 5.1 Hz, 1H), 1.97 (ddd,
J = 13.9, 12.1, 5.2 Hz, 1H), 1.63 (s, 3H), 1.54 ppm (s, 3H); 13C NMR
(100 MHz, CD3OD): d = 174.5, 171.8, 170.0, 155.9, 143.0, 142.0, 141.7,
131.9, 130.1, 129.5, 129.4, 129.0, 127.7, 127.0, 126.2, 125.8, 125.2,
80.1, 75.7, 73.9, 71.2, 59.5, 50.9, 50.7, 47.2, 40.1, 39.7, 30.9, 29.4, 25.6,
16.5 ppm; MS-ESI+: m/z 646, [M+H]+; HRMS-ESI+: [M+H]+ calcd for
C36H44N3O6S: 646.2951, found 646.2959. Anal. Calcd for
C36H43N3O6Sꢄ2/5 CHCl3: C, 63.04; H, 6.31; N, 6.06. Found: C, 63.45; H,
5.94; N, 5.78 (retrieved from the NMR sample).
4.1.5.5. (S)-(4S)-3-(2-Benzyl-3-(2-(2,6-dimethylphenoxy)acetam-
ido)-2-hydroxypropanoyl)-N-((1S,2R)-2-hydroxy-1-indanyl)-5,5-
dimethylthiazolidine-4-carboxamide ((S)-16). Compound (S)-16
was synthesized according to the general procedure from (S)-10
(24.0 mg, 0.053 mmol), 21 (12.8 mg, 0.071 mmol), PyBOP (46.4 mg,
0.089 mmol) and DIEA (30.0 ll, 0.175 mmol). It was purified (RP-
HPLC) to give (S)-16 as a white solid (20.9 mg, 62%). Compound
(S)-16: tR = 15.72 (SP-HPLC, 15% iPrOH in iHex); ½a D22
= ꢁ23.4 (c
ꢅ
0.82, MeOH); rotamer ratio (9:8); 1H NMR (400 MHz, CD3OD):
d = 7.37–7.33 (m, 2+2H), 7.29–7.07 (m, 7+7H), 7.05–6.88 (m,
3+3H), 5.35 (d, J = 10.4 Hz, 1H) (4.75 (d, J = 10.9 Hz, 1H)), 5.33 (d,
J = 4.9 Hz, 1H) (5.32 (d, J = 4.9 Hz, 1H)), 4.79 (d, J = 10.4 Hz, 1H)
(4.53 (d, J = 10.9 Hz, 1H)), 4.54 (dt, J = 4.9, 1.6 Hz, 1H) (4.46 (dt,
J = 4.9, 1.8 Hz, 1H)), 4.52 (s, 1H) (4.31 (s, 1H)), 4.41 (s, 2H) (4.17
(d, J = 14.8 Hz, 1H), 4.12 (d, J = 14.8 Hz, 1H)), 3.95(d, J = 13.7 Hz,
1H) (3.85 (d, J = 13.7 Hz, 1H)), 3.52 (d, J = 13.7 Hz, 1H) (3.70 (d,
J = 13.7 Hz, 1H)), 3.23 (d, J = 13.6 Hz, 1H) (3.20 (d, J = 13.1 Hz,
1H)), 3.12 (d, J = 14.0 Hz, 1H) (2.92 (d, J = 13.4 Hz, 1H)), 3.11 (dd,
J = 16.6, 4.9 Hz, 1H) (2.98 (dd, J = 16.3, 4.9 Hz, 1H)), 2.85 (dd,
J = 16.3, 2.0 Hz, 1+1H), 2.24 (s, 6H) (2.23 (s, 6H)), 1.57 (s, 3H)
(1.51 (s, 3H)), 1.37 ppm (s, 3H) (0.84 ppm (s, 3H)); 13C NMR
(100 MHz, CD3OD): major rotamer: d = 174.4, 172.9, 172.1, 156.0,
141.98, 141.73, 137.3, 132.0, 131.71, 130.07, 129.2, 128.9, 128.1,
127.78, 126.12, 125.8, 125.66, 82.5, 76.1, 74.0, 71.28, 59.1, 51.0,
50.5, 47.2, 43.7, 41.0, 29.9, 25.4, 16.6 ppm; minor rotamer:
d = 175.2, 173.1, 171.4, 156.0, 142.06, 141.7, 136.7, 131.9, 130.72,
130.05, 129.5, 129.5, 127.9, 127.83, 126.07, 125.9, 125.74, 80.3,
74.1, 74.5, 71.33, 58.7, 51.70, 50.5, 45.4, 43.7, 40.8, 30.9, 25.4,
16.4 ppm; MS-ESI+: m/z 632, [M+H]+; HRMS-ESI–: [M–H]+ calcd for
C35H40N3O6S: 630.2638, found 630.2647.
4.1.5.8. (R)-(4S)-N-((1S,2R)-2-Hydroxy-1-indanyl)-3-(2-(2,6-dim-
ethylphenoxyacetyl)amino-2-hydroxy-4-phenylbutanoyl)-5,5-
dimethyl-1,3-thiazolidine-4-carboxamide ((R)-17). Compound
(R)-17 was synthesized according to the general procedure from
(R)-11 (22.2 mg, 0.048 mmol), 21 (10.3 mg, 0.057 mmol), PyBOP
(37.2 mg, 0.071 mmol) and DIEA (24.4 ll, 0.14 mmol). It was puri-
fied (RP-HPLC) to give (R)-17 as a white solid (20.6 mg, 67%). Com-
pound (R)-17: tR = 22.15 (SP-HPLC, 15% iPrOH in iHex); ½a D21
= ꢁ1.5
ꢅ
(c 0.160, MeOH); rotamer ratio (5:1); 1H NMR (400 MHz, CD3OD):
d = 7.28–7.06 (m, 9H), 7.04–6.99 (m, 2H), 6.95 (dd, J = 8.4, 6.5 Hz,
1H), 5.43 (d, J = 10.6 Hz, 1H), 5.27 (dd, J = 8.4, 5.0 Hz, 1H), 5.18 (d,
J = 10.6 Hz, 1H), 4.57 (dt, J = 5.0, 1.3 Hz, 1H), 4.53 (s, 1H), 4.28 (s,
2H), 3.81 (dd, J = 13.7, 4.5 Hz, 1H), 3.66 (dd, J = 13.7, 4.6 Hz, 1H),
3.06 (dd, J = 16.5, 5.0 Hz, 1H), 2.85 (d, J = 16.5 Hz, 1H), 2.76 (dt,
J = 12.9, 4.9 Hz, 1H), 2.63 (ddd, J = 12.9, 12.9, 4.7 Hz, 1H), 2.25 (s,
6H), 2.16 (ddd, J = 13.7, 12.8, 4.7 Hz, 1H), 1.95 (ddd, J = 13.7, 12.9,
4.9 Hz, 1H), 1.63 (s, 3H), 1.55 ppm (s, 3H); 13C NMR (100 MHz,
CD3OD): d = 174.6, 172.0, 171.0, 155.9, 143.3, 142.0, 141.7, 131.7,
130.1, 129. 5, 129.4, 129.0, 127.8, 126.8, 126.1, 125.8, 125.4, 80.3,
76.0, 73.8, 71.3, 59.4, 50.7, 50.6, 47.2, 40.9, 40.3, 30.7, 29.9, 25.3,
16.5 ppm; MS-ESI+: m/z 646, [M+H]+; HRMS-ESI+: [M+H]+ calcd for
C36H44N3O6S: 646.2951, found 646.2957. Anal. Calcd for
C36H43N3O6Sꢄ2/5 CHCl3: C, 63.04; H, 6.31; N, 6.06; N, 5.90. Found:
C, 63.31; H, 5.94; N, 5.65 (retrieved from the NMR sample).
4.1.5.6. (R)-(4S)-3-(2-Benzyl-3-(2-(2,6-dimethylphenoxy)acet-
amido)-2-hydroxypropanoyl)-N-((1S,2R)-2-hydroxy-1-indanyl)-
5,5-dimethylthiazolidine-4-carboxamide ((R)-16). Compound
(R)-16 was synthesized according to the general procedure from
(R)-10 (18.7 mg, 0.040 mmol), 21 (8.6 mg, 0.048 mmol), PyBOP
(35.1 mg, 0.067 mmol) and DIEA (15.0 ll, 0.088 mmol). It was puri-
fied (RP-HPLC) to give (R)-16 as a white solid (8.11 mg, 37%). Com-
4.1.5.9. (A)-(4R)-3-(2-Benzyl-3-(2-(2,6-dimethylphenoxy)acet-
amido)-2-hydroxypropanoyl)-N-((1S,2R)-2-hydroxy-1-indanyl)-
5,5-dimethylthiazolidine-4-carboxamide ((S)-18). Compound
(S)-18 was synthesized according to the general procedure from
(S)-12 (36.2 mg, 0.080 mmol), 21 (17.5 mg, 0.097 mmol), PyBOP
pound (R)-16: tR = 19.40 (SP-HPLC, 15% iPrOH in iHex);
½
a 2D1
ꢅ
= ꢁ47.5 (c 0.356, MeOH); 1H NMR (400 MHz, CD3OD):
d = 7.37–7.21 (m, 6H), 7.17–7.08 (m, 2H), 7.03 (d, J = 7.2 Hz, 1H),
6.95 (d, J = 7.5 Hz, 2H), 6.90 (dd, J = 8.5, 6.2 Hz, 1H), 5.28 (d,
J = 5.1 Hz, 1H), 4.84 (d, J = 10.8 Hz, 1H), 4.47 (dt, J = 5.1, 1.7 Hz,
1H), 4.40 (s, 1H), 4.33 (d, J = 14.8 Hz, 1H), 4.26 (d, J = 14.8 Hz,
1H), 4.2 (d, J = 10.8 Hz, 1H), 3.76 (s, 2H), 3.19 (d, J = 13.5 Hz, 1H),
3.04 (dd, J = 16.4, 5.0 Hz, 1H), 2.95 (d, J = 13.5 Hz, 1H), 2.86 (ddd,
J = 16.4, 1.7, 0.6 Hz, 1H), 2.25 (s, 6H), 1.45 (s, 3H), 1.39 ppm (s,
3H); 13C NMR (100 MHz, CD3OD): d = 174.9, 172.1, 171.2, 156.1,
142.0, 141.5, 137.2, 131.8, 131.5, 130.1, 129.5, 128.8, 128.1,
127.8, 126.0, 125.70, 125.69, 81.1, 75.9, 74.1, 71.3, 58.7, 50.45,
(63.0 mg, 0.12 mmol) and DIEA (30.0 ll, 0.18 mmol). It was puri-
fied (RP-HPLC) to give (S)-18 as a white solid (26.3 mg, 52%). Com-
pound (S)-18: tR = 17.94 min (SP-HPLC, 15% iPrOH in iHex);
½
a 2D1 = +12.1 (c 0.942, MeOH); 1H NMR (400 MHz, CD3OD):
ꢅ
d = 7.31–7.14 (m, 9H), 7.03 (d, J = 7.4 Hz, 2H), 6.95 (dd, J = 8.3,
6.5 Hz, 1H), 5.19 (dd, J = 4.8, 0.6 Hz, 1H), 4.92 (d, J = 10.8 Hz, 1H),
4.47 (ddd, J = 4.9, 4.8, 1.6 Hz, 1H), 4.39 (s, 1H), 4.34 (d,
J = 14.8 Hz, 1H), 4.250 (d, J = 10.8 Hz, 1H), 4.245 (d, J = 14.8 Hz,

5946
K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
1H), 3.73 (s, 2H), 3.16 (d, J = 13.5 Hz, 1H), 2.95 (d, J = 13.5 Hz, 1H),
2.85 (dd, J = 16.5, 4.9 Hz, 1H), 2.76 (dd, J = 16.4, 1.6 Hz, 1H), 2.31 (s,
6H), 1.46 (s, 3H), 1.42 ppm (s, 3H); 13C NMR (100 MHz, CD3OD):
d = 174.6, 172.1, 171.6, 156.1, 142.0, 141.7, 137.1, 131.9, 131.5,
130.1, 129.5, 129.0, 128.1, 127.7, 126.2, 125.8, 125.2, 81.4, 75.9,
73.9, 71.3, 59.4, 50.6, 50.3, 48.9, 44.3, 40.1, 28.9, 25.6, 16.5 ppm;
MS-ESI+: m/z 632, [M+H]+; HRMS-ESI–: [M–H]+ calcd for
C35H40N3O6S: 630.2638, found 630.2636.
mmol), 21 (13.6 mg, 0.076 mmol), PyBOP (39.4 mg, 0.076 mmol)
and DIEA (21.6 l, 0.13 mmol). It was purified (RP-HPLC) to give
l
(B)-19 as a white solid (1.64 mg, 4.3%). Compound (B)-19: Rotamer
ratio (5:1) 1H NMR (400 MHz, CD3OD): d = 7.37–6.87 (m, 12H),
5.24 (d, J = 4.8 Hz, 1H), 4.49 (dt, J = 4.9, 1.5 Hz, 1H), 4.31 (d,
16.3 Hz, 1H), 4.29 (s, 1H), 4.22 (d, J = 16.3 Hz, 1H), 4.14–3.98 (m,
2H), 3.81 (d, J = 13.8 Hz, 1H), 3.69 (d, J = 13.8 Hz, 1H), 3.06 (dd,
J = 16.5, 4.9 Hz, 1H), 2.85 (d, J = 16.5 Hz, 1H), 2.82–2.78 (m, 1H),
2.75–2.66 (m, 1H), 2.37–2.30 (m, 1H), 2.22 (s, 6H), 2.18–2.11 (m,
2H), 2.02–1.91 ppm (m, 3H); 13C NMR (100 MHz, CD3OD):
d = 175.3, 174.7, 171.8, 156.0, 143.3, 141.5, 131.8, 130.04, 130.01,
129.46, 129.45, 128.8, 127.8, 126.9, 125.9, 125.9, 125.7, 79.3, 74.1,
71.2, 63.8, 58.9, 50.3, 46.9, 40.7, 39.3, 30.8, 29.9, 26.9, 16.5 ppm;
MS-ESI+: m/z 600, [M+H]+; HRMS-ESI+: [M+H]+ calcd for
C35H42N3O6: 600.3074, found 600.3069.
4.1.5.10. (B)-(4R)-3-(2-Benzyl-3-(2-(2,6-dimethylphenoxy)acet-
amido)-2-hydroxypropanoyl)-N-((1S,2R)-2-hydroxy-1-indanyl)-
5,5-dimethylthiazolidine-4-carboxamide ((R)-18). Compound
(R)-18 was synthesized according to the general procedure from
(R)-12 (16.3 mg, 0.036 mmol), 21 (7.8 mg, 0.043 mmol), PyBOP
(28.1 mg, 0.054 mmol) and DIEA (20.0 ll, 0.12 mmol). It was puri-
fied (RP-HPLC) to give (R)-18 as a white solid (12.9 mg, 57%). Com-
pound (R)-18: tR = 13.95 min (SP-HPLC, 15% iPrOH in iHex);
4.1.5.13. (A)-(2S)-N-((1S,2R)-2-Hydroxy-1-indanyl)-2-((2-phen-
ethyl-3-(2-(2,6-dimethylphenoxy)acetamido)-2-hydroxyprop-
anoyl)amino)-3-phenyl-propylamide ((A)-20). Compound (A)-20
was synthesized according to the general procedure for the epoxide
opening using (A)-14 (22.6 mg, 0.48 mmol) and then proceeding with
21 (10.4 mg, 0.058 mmol), HOBT (9.7 mg, 0.072 mmol) and EDC
(13.8 mg, 0.072 mmol) that was dissolved in freshly distilled CH2Cl2
and stirred at room temperature for 30 h. It was isolated (silica,
EtOAc/MeOH, 98:2–92.5:7.5) and further purified by RP-HPLC to give
½
a 2D2 = +11.1 (c 0.83, MeOH); rotamer ratio (3:2), 1H NMR
ꢅ
(400 MHz, CD3OD): major conformer: d = 7.33–7.11 (m, 9H), 7.05–
6.99 (m, 2H), 6.95 (m, 1H), 5.33 (d, J = 5.0 Hz, 1H), 5.29 (d,
J = 10.5 Hz, 1H), 4.66 (d, J = 10.5 Hz, 1H), 4.60 (dt, J = 5.0, 1.4 Hz,
1H), 4.44 (s, 1H), 4.25 (s, 2H), 3.77 (d, J = 13.6 Hz, 1H), 3.55 (d,
J = 13.6 Hz, 1H), 3.16 (d, J = 13.8 Hz, 1H), 3.11 (dd, J = 16.5, 5.0 Hz,
1H), 3.05 (d, J = 13.8 Hz, 1H), 2.90 (dd, J = 16.5, 1.4 Hz, 1H), 2.26 (s,
6H), 1.57 (s, 3H), 1.51 ppm (s, 3H); minor conformer: d = 7.33–7.11
(m, 9H), 7.05–6.99 (m, 2H), 6.95 (m, 1H), 5.22 (d, J = 5.2 Hz, 1H),
4.76 (d, J = 10.7 Hz, 1H), 4.69 (s, 1H), 4.57 (dt, J = 5.2, 2.4 Hz, 1H),
4.56 (d, J = 10.7 Hz, 1H), 4.30 (d, J = 14.9 Hz, 1H), 4.17 (d,
J = 14.9 Hz, 1H), 3.89 (d, J = 13.5 Hz, 1H), 3.58 (d, J = 13.5 Hz, 1H),
3.20 (d, J = 13.2 Hz, 1H), 3.02 (dd, J = 16.3, 5.2 Hz, 1H), 2.88 (d,
J = 13.2 Hz, 1H), 2.86 (dd, J = 16.3, 2.4 Hz, 1H), 2.27 (s, 6H), 1.34 (s,
3H), 0.81 ppm (s, 3H); 13C NMR (100 MHz, CD3OD): major con-
former: d = 174.3, 173.2, 172.0, 156.0, 142.0, 141.81, 136.5, 132.0,
131.7, 130.13, 129.3, 129.0, 127.9, 127.8, 126.18, 125.8, 125.4,
80.6, 76.1, 73.9, 71.3, 59.3, 50.7, 50.5, 47.5, 43.9, 40.4, 30.0, 25.1,
16.6 ppm; minor conformer: d = 174.8, 172.9, 171.9, 156.1, 142.3,
141.84, 137.3, 131.9, 131.7, 130.07, 129.4, 129.1, 128.0, 127.8,
126.19, 125.7, 125.4, 82.9, 76.1, 73.2, 71.2, 58.8, 51.9, 50.1, 45.7,
43.9, 40.6, 31.0, 25.4, 16.5 ppm; MS-ESI+: m/z 632, [M+H]+; HRMS-
ESI–: [M–H]+ calcd for C35H40N3O6S: 630.2638, found 630.2653.
(A)-20 (16.4 mg, 53%) as
a white solid. Compound (A)-20:
tR = 21.82 min (SP-HPLC, 15% iPrOH in iHex); ½a D21
= ꢁ3.8 (c 0.153,
ꢅ
MeOH); 1H NMR (400 MHz, CD3OD): d = 7.35–7.32 (m, 2H), 7.26–
7.06 (m, 10H) , 7.03–6.88 (m, 5H), 5.29 (d, J = 5.1 Hz, 1H), 4.87 (dd,
J = 10.1, 4.7 Hz, 1H), 4.51 (ddd, J = 5.2, 5.1, 2.0 Hz, 1H), 4.23 (d,
J = 14.9 Hz, 1H), 4.14 (d, J = 14.9 Hz, 1H), 3.68 (d, J = 13.7 Hz, 1H),
3.48 (d, J = 13.7 Hz, 1H), 3.32 (dd, J = 14.0, 4.7 Hz, 1H), 3.11 (dd,
J = 16.3, 5.1 Hz, 1H), 3.01 (dd, J = 14.0, 10.1 Hz, 1H), 2.91 (dd, J = 16.3,
2.0 Hz, 1H), 2.48–2.35 (m, 1H), 2.19 (s, 6H), 2.00–1.85 (m, 2H), 1.72–
1.63 ppm (m, 1H); 13C NMR (100 MHz, CD3OD): d = 176.3, 173.7,
172.2, 156.0, 143.2, 142.0, 141.7, 138.4, 131.8, 130.5, 130.1, 129.6,
129.4, 129.3, 129.0, 128.0, 127.8, 126.8, 126.1, 125.7, 125.4, 78.4,
74.0, 71.2, 58.8, 55.8, 47.8, 40.8, 40.1, 39.1, 30.3, 16.5 ppm; MS-ESI+:
m/z 650, [M+H]+; HRMS-ESI+: [M+H]+ calcd for C39H42N3O6:
650.3074, found 650.3076. Anal. Calcd for C39H43N3O6ꢄ1/4 H2O: C,
71.59; H, 6.70; N, 6.42. Found: C, 71.60; H, 6.84; N, 6.37.
4.1.5.11. (A)-(5R)-N-((1S,2R)-2-Hydroxy-1-indanyl)-1-(2-(2,6-dim-
ethylphenoxyacetyl)amino-2-hydroxy-4-phenylbutanoyl)pyrrol-
idine-5-carboxamide ((A)-19). Compound (A)-19 was synthesized
according to the general procedure from (A)-13 (27.5 mg, 0.065
mmol), 21 (13.7 mg, 0.076 mmol), PyBOP (39.5 mg, 0.076 mmol)
4.1.5.14. (B)-(2S)-N-((1S,2R)-2-Hydroxy-1-indanyl)-2-((2-phen-
ethyl-3-(2-(2,6-dimethylphenoxy)acetamido)-2-hydroxypropa-
noyl)amino)-3-phenyl-propylamide ((B)-20). Compound (B)-20
was synthesized according to the general procedure for the epoxide
opening using (B)-14 (20.9 mg, 0.48 mmol) and then proceeding
with 21 (9.3 mg, 0.051 mmol), HOBT (8.7 mg, 0.064 mmol) and
EDC (12.3 mg, 0.064 mmol) that was dissolved in freshly distilled
CH2Cl2 and stirred at room temperature for 30 h. It was isolated (sil-
ica, EtOAc/MeOH, 98:2–85.5:15) and further purified by RP-HPLC to
give (B)-20 (13.57 mg, 49%) as a white solid. Compound (B)-20:
and DIEA (21.7 ll, 0.13 mmol). It was purified (RP-HPLC) to give
(A)-19 as a white solid (17.8 mg, 45%). (A)-19: Rotamer ratio (8:1);
1H NMR (400 MHz, CD3OD): d = 7.38–6.85 (m, 12H), 5.24 (d,
J = 5.0 Hz, 1H), 4.55 (dd, J = 8.1, 5.6 Hz, 1H), 4.48 (dt, J = 5.0, 1.7 Hz,
1H), 4.29 (d, 14.8 Hz, 1H), 4.25 (d, J = 14.8 Hz, 1H), 4.13–4.06 (m,
1H), 4.05–3.97 (m, 1H), 3.85 (d, J = 13.5 Hz, 1H), 3.69 (d, J = 13.5 Hz,
1H), 3.03 (dd, J = 16.4, 5 Hz, 1H), 2.85 (dd, J = 16.4, 1.7 Hz, 1H), 2.82–
2.67 (m, 2H), 2.24–2.13 (m, 3H), 2.21 (s, 6H), 2.02–1.87 ppm (m,
3H); 13C NMR (100 MHz, CD3OD): d = 175.2, 174.7, 171.8, 156.0,
143.3, 142.1, 141.5, 131.8, 130.0, 129.5, 129.4, 128.8, 127.6, 126.9,
125.9, 125.7, 125.7, 79.3, 74.1, 71.2, 63.8, 58.9, 50.2, 46.9, 40.7, 39.3,
30.7, 29.9, 26.9, 16.5 ppm; MS-ESI+: m/z 600, [M+H]+; HRMS-ESI+:
[M+H]+ calcd for C35H42N3O6: 600.3074, found 600.3071. Anal. Calcd
for C35H41N3O6ꢄ1ꢄ1/4 HCOOH: C, 66.24; H, 6.67; N, 6.39. Found: C,
65.96; H, 6.83; N, 6.11.
tR = 15.35 min (SP-HPLC, 15% iPrOH in iHex); ½a D21
= ꢁ0.3 (c 0.321,
ꢅ
MeOH:CH2Cl2 (1:1)); 1H NMR (400 MHz, CD3OD:CDCl3): d = 7.35–
7.09 (m, 13H), 7.03–6.98 (m, 3H) , 6.95–6.91 (m, 1H), 5.25 (d,
J = 5.1 Hz, 1H), 4.80 (dd, J = 8.5, 6.1 Hz, 1H), 4.44 (ddd, J = 5.1, 5.0,
2.0 Hz, 1H), 4.21 (s, 2H), 3.68 (d, J = 13.7 Hz, 1H), 3.35–3.31 (m,
1H) (confirmed by HSQC experiment), 3.23 (dd, J = 13.8, 6.1 Hz,
1H), 3.10 (dd, J = 13.8, 8.5 Hz, 1H), 3.06 (dd, J = 16.4, 5.1 Hz, 1H),
2.87 (dd, J = 16.4, 2.0 Hz, 1H), 2.75 (dt, J = 13.1, 4.6 Hz, 1H), 2.47
(dt, J = 13.1, 4.7 Hz, 1H), 2.22 (s, 6H), 2.10 (ddd, J = 13.6, 12.8,
4.7 Hz, 1H), 1.84 ppm (ddd, J = 13.7, 12.8, 4.6 Hz, 1H); 13C NMR
(400 MHz, CD3OD:CDCl3): d = 176.0, 173.2, 171.9, 155.6, 142.9,
141.7, 141.3, 137.9, 131.4, 130.3, 130.0, 129.4, 129.22, 129.18,
128.8, 127.8, 127.7, 126.7, 125.9, 125.7, 125.2, 78.2, 73.6, 71.0,
58.5, 55.5, 47.6, 40.5, 40.0, 39.2, 30.4, 16.5 ppm; MS-ESI+: m/z 650,
4.1.5.12. (B)-(5R)-N-((1S,2R)-2-Hydroxy-1-indanyl)-1-(2-(2,6-dim-
ethylphenoxyacetyl)amino-2-hydroxy-4-phenylbutanoyl)pyrroli-
dine-5-carboxamide ((B)-19). Compound (B)-19 was synthesized
according to the general procedure from (B)-13 (26.0 mg, 0.062

K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
5947
[M+H]+; HRMS-ESI+: [M+H]+ calcd for C39H42N3O6: 650.3074, found
650.3071. Anal. Calcd for C39H43N3O6 ꢄ 1/4H2O: C, 71.59; H, 6.70; N,
6.42. Found: C, 71.64; H, 6.82; N, 6.42.
synthesized compounds were determined from two individual
measurements by a fluorometric assay.42 Assay variability was
checked by inclusion of a known inhibitor and the standard devia-
tion for the enzyme assays was 50% of the mean.
4.2. Enzyme assays and Ki determinations
4.2.4. In vitro anti-HIV activity assay
The in vitro anti-HIV activity was assayed in MT4 cells accord-
4.2.1. Plasmepsin inhibition assays
ing to a previously published procedure.42
4.2.1.1. Enzyme purification. All enzymes are expressed as inac-
tive proenzymes using the pET expression system. Each gene was
cloned into pET3a and transformed into BL21 DE3 pLysS E. coli
expression cells. In brief, the enzymes were expressed as inclusion
bodies as described by Westling et al.37 The proenzyme form was
recovered by dissolving the inclusion bodies in 8 M urea and
refolding was done by dialysis, all using conditions identical to
those described by Westling et al. Enzymes were purified to homo-
geneity by ion exchange chromatography.
4.2.5. BACE-1 inhibition assay
BACE-1 enzyme was prepared and the inhibition of BACE-1 was
determined as previously described using the Swedish mutant se-
quence Eu-EVNLDAEFK-Quencher as substrate.43 IC50 values were
determined using the following relationship:
%
inhibi-
tion = 100[I]/([I] + IC50).
4.3. Computational evaluation
4.2.1.2. Dissociation constant determination. The dissociation
constants (Ki) for the inhibitor–enzyme complexes were deter-
mined by monitoring the decrease in chromogenic substrate
hydrolysis due to the presence of inhibitors. All enzymes are ex-
pressed as proenzymes. The enzymes are self-activated and con-
verted to the mature, catalytically active form by incubating in a
0.1 M sodium acetate pH 4.5 buffer at 37 °C for 5 min prior to start-
ing the assay. Substrate hydrolysis was measured using a Varian
Cary50 spectrophotometer with an 18 cell sample handling sys-
tem. To determine the Ki value, a Michaelis–Menten curve was
generated using no less than five different substrate concentra-
tions. Two additional curves were generated with two different
concentrations of inhibitors. All three curves, obtained at different
concentrations of an inhibitor, were fitted simultaneously using
Sigmaplot10 (Systat Software Inc) to provide the value of Ki. Eq.
1 or 2 for non-tight binding and tight binding inhibitors, respec-
tively, gave the Ki value for the inhibitor.
4.3.1. Automated docking
Automated docking exploration was performed with GOLD ver-
sion 3.0.1,44 allowing full flexibility for the ligand while keeping
the protein fixed (i.e., only rotation of the protein hydroxyl groups
was considered). Each docking exploration consisted of 10 inde-
pendent runs, using the default genetic algorithm (GA) search
parameters and the Goldscore function for scoring. Parallel docking
runs were performed considering either Asp214 or Asp34 in the
protonated (neutral) state, while maintaining the other Asp nega-
tively charged. A hydrogen bond distance restraint was imposed
to guide docking between one of the catalytic aspartates and the
hydroxyl TS mimic of the inhibitor. The protein model used was
PmPM4, in the conformation adopted upon binding of KNI-764
(PDB entry 2ANL).13 Docking solutions were selected for post-pro-
cessing only if they satisfied the following conditions: (1) The exis-
tence of a minimum hydrogen bond network between protein and
ligands (minimum 3H bonds), and (2) complementarity of the four
inhibitor sidechains (P2 to P20) in the four protein binding pockets
(S2 to S2). The poses selected this way were translated into PfPM4
for the next modelling step by superimposing the resulting inhib-
itor–2ANL complex onto the crystal structure of PfPM4 (PDB entry
1LS5). Manual docking was alternatively performed by building the
inhibitor with Maestro, followed by energy minimization in Mac-
romodel version 9.5 M (Schrödinger, New York, 2007) (1000 itera-
tion steepest descendent steps, with H-bond restraints with the
protein) in the extended conformation. Two manual conformations
were considered: the first used as scaffold KNI-764 in PmPM4 (PDB
code 2ANL). The second one was a flipped conformation, in which
P1 and P2 substituents were interchanged with the P10 and P20,
respectively.
Vmax
m
m
¼ ð
Þ
ð1Þ
ð2Þ
Km
1
i
ð1 þ Þð1 þ K
Þ
S
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
mo
¼
ðE ꢁ I ꢁ K þ 2ðE ꢁ I ꢁ Kap þ 4EKapÞ
2E
4.2.2. hCatD inhibition assay
Human liver CatD was purchased from Sigma–Aldrich, Sweden.
The activities of hCatD were measured essentially as described ear-
lier,38 using a total reaction volume of 100
lL. hCatD was activated
by incubation in the assay reaction buffer (100 mM sodium acetate
buffer (pH 4.5), 10% glycerol, and 0.01% Tween 20) at 37 °C for
20 min. The reaction was initiated by the addition of 3 lM sub-
strate (DABCYL-Glu-Arg-Nle-Phe-Leu-Ser-Phe-Pro-EDANS, Ana-
Spec Inc, San Jose, CA, USA) and hydrolysis was recorded as the
increase in fluorescence intensity over a 10-min time period, dur-
ing which the rate increased linearly with time. Stock solutions of
inhibitors in DMSO were serially diluted in DMSO and added di-
rectly before addition of the substrate, giving a final DMSO concen-
tration of 1%. IC50 values were obtained by assuming competitive
4.3.2. Molecular dynamics and binding free energy calculations
Molecular dynamics (MD) simulations were done using the pro-
gram
was solvated with a simulation sphere of TIP3P waters47 of radius
20 Å, centred on the -carbon of the transition-state mimicking P1.
Q
45 and the OPLS force field there implemented.46 The system
a
The water surface of this sphere was subjected to radial and polar-
ization restraints in order to mimic bulk water at the sphere
boundary.45 Non-bonded interaction energies were calculated up
to a 10 Å cut-off, except for the ligand atoms for which no cut-off
was used. Beyond the cut-off, long-range electrostatics were trea-
ted with the local reaction field (LRF) multipole expansion meth-
od.48 Protein atoms outside the simulation sphere were
restrained to their initial positions, and only interacted with the
system through bonds, angles and torsions. For the ligandꢁprotein
simulations, a heating and equilibration procedure was applied be-
fore the data collection phase. The equilibration protocol started
with 1000 steps MD using very short time step (0.2 fs) at 1 K tem-
inhibition and fitting the Langmuir isotherm (vi/vo = 1/(1 + [I]/
IC50)) to the dose–response data (Grafit), where vi and vo are the
initial velocities of the inhibited and uninhibited reactions, respec-
tively, and [I] is the inhibitor concentration.39 Ki values were sub-
sequently calculated using Ki = IC50/(1 + [S]/Km)40 and Km values
were determined according to Michaelis–Menten.
4.2.3. HIV-1 protease inhibition assay
The HIV-1 protease was cloned and heterologously expressed in
E. coli and purified as described elsewhere.41 The Ki values for the

5948
K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
perature, coupled to a strong bath (0.2 fs bath coupling) with posi-
tional restraints on heavy atoms. Then the system was gradually
heated up to 300 K, relaxing the bath coupling to 100 fs and
increasing the timestep to 1.5 fs. During this equilibration process
positional restraints on the protein heavy atoms and the central
atom of the ligand were gradually released from 200 to 0 kcal/
(mol Å2). The production phase followed for 1.5–1.8 ns and col-
lecting energies at regular intervals of 15 fs. Stability was ad-
dressed by comparing the average potential energy values of
the first and second halves of the data collection period. Error
estimates in the calculated free energies were obtained from
the appropriate scaling of these errors from the water and protein
simulations, through the use of the linear interaction energy (LIE)
References and notes
1. World Malaria Report 2008, Geneva: World Health Organisation, 2008, 1.
2. White, N. J. J. Clin. Invest. 2004, 113, 1084.
3. Goldberg, D. E.; Slater, A. F.; Cerami, A.; Henderson, G. B. Proc. Natl. Acad. Sci.
U.S.A. 1990, 87, 2931.
4. Coombs, G. H.; Goldberg, D. E.; Klemba, M.; Berry, C.; Kay, J.; Mottram, J. C.
Trends Parasitol. 2001, 17, 532.
5. Global burden of disease: 2004 update Geneva: World Health Organization, 2004.
6. Dame, J. B.; Reddy, G. R.; Yowell, C. A.; Dunn, B. M.; Kay, J.; Berry, C. Mol.
Biochem. Parasitol. 1994, 64, 177.
7. Banerjee, R.; Liu, J.; Beatty, W.; Pelosof, L.; Klemba, M.; Goldberg, D. E. Proc. Natl.
Acad. Sci. U.S.A. 2002, 99, 990.
8. Ersmark, K.; Samuelsson, B.; Hallberg, A. Med. Res. Rev. 2006, 26, 626.
9. Luksch, T.; Chan, N. S.; Brass, S.; Sotriffer, C. A.; Klebe, G.; Diederich, W. E.
Chemmedchem. 2008, 3, 1323.
10. Hidaka, K.; Kimura, T.; Ruben, A. J.; Uemura, T.; Kamiya, M.; Kiso, A.; Okamoto,
T.; Tsuchiya, Y.; Hayashi, Y.; Freire, E.; Kiso, Y. Bioorg. Med. Chem. 2008, 16,
10049.
a
and b factors. The MD sampling of the free ligand was again
done in a TIP3P water sphere of the same radius as in the corre-
sponding protein simulation. The sphere was equilibrated with a
10.2 ps MD run at 300 K, with positional restraints on the heavy
atoms of the ligand (200 kcal/(mol Å2) for the first 0.2 ps and
10 kcal/(mol Å2) in the following 10 ps). MD followed for more
than 1.5 ns under the same conditions as for the bound state,
but keeping the initial position of the central atom of the ligand
fixed with 10 kcal/(mol Å2) to ensure homogeneous solvation.
Binding affinities were calculated using the linear interaction en-
ergy (LIE) method, described in detail elsewhere.49,50 Basically,
this approach estimates the ligand free energy of binding from
the difference in the ligandꢁsurrounding interaction energies in
both its bound and free state. The relationship between the ligand
intermolecular interaction energies and the free energy of binding
is given by Eq. 3:
11. Dame, J. B.; Yowell, C. A.; Omara-Opyene, L.; Carlton, J. M.; Cooper, R. A.; Li, T.
Mol. Biochem. Parasitol. 2003, 130, 1.
12. Li, T.; Yowell, C. A.; Beyer, B. B.; Hung, S.-H.; Westling, J.; Lam, M. T.; Dunn, B.
M.; Dame, J. B. Mol. Biochem. Parasitol. 2004, 135, 101.
13. Clemente, J. C.; Govindasamy, L.; Madabushi, A.; Fisher, S. Z.; Moose, R. E.;
Yowell, C. A.; Hidaka, K.; Kimura, T.; Hayashi, Y.; Kiso, Y.; Agbandje-
McKenna, M.; Dame, J. B.; Dunn, B. M.; McKenna, R. Acta Crystallogr., Sect. D
2006, 62, 246.
14. Omara-Opyene, A. L.; Moura, P. A.; Sulsona, C. R.; Bonilla, J. A.; Yowell, C. A.;
Fujioka, H.; Fidock, D. A.; Dame, J. B. J. Biol. Chem. 2004, 279, 54088.
15. Liu, J.; Gluzman, I. Y.; Drew, M. E.; Goldberg, D. E. J. Biol. Chem. 2005, 280, 1432.
16. Bonilla, J. A.; Bonilla, T. D.; Yowell, C. A.; Fujioka, H.; Dame, J. B. Mol. Microbiol.
2007, 65, 64.
17. Nezami, A.; Kimura, T.; Hidaka, K.; Kiso, A.; Liu, J.; Kiso, Y.; Goldberg, D. E.;
Freire, E. Biochemistry 2003, 42, 8459.
18. Kiso, A.; Hidaka, K.; Kimura, T.; Hayashi, Y.; Nezami, A.; Freire, E.; Kiso, Y.
J. Pept. Sci. 2004, 10, 641.
19. Ekegren, J. K.; Unge, T.; Safa, M. Z.; Wallberg, H.; Samuelsson, B.; Hallberg, A.
J. Med. Chem. 2005, 48, 8098.
20. Ekegren, J. K.; Gising, J.; Wallberg, H.; Larhed, M.; Samuelsson, B.; Hallberg, A.
Org. Biomol. Chem. 2006, 4, 3040.
v
dw
el
l—s
D
Gbind
¼
a
D
hVl—s i þ b
D
hV i þ Y
ð3Þ
21. Wu, X. Y.; Öhrngren, P.; Ekegren, J. K.; Unge, J.; Unge, T.; Wallberg, H.;
Samuelsson, B.; Hallberg, A.; Larhed, M. J. Med. Chem. 2008, 51, 1053.
22. Beyer, B. B.; Johnson, J. V.; Chung, A. Y.; Li, T.; Madabushi, A.; Agbandje-
McKenna, M.; McKenna, R.; Dame, J. B.; Dunn, B. M. Biochemistry 2005, 44,
1768.
23. Howard-Lock, H. E.; Lock, C. J. L.; Martins, M. L.; Smalley, P. S.; Bell, R. A. Can. J.
Chem. 1986, 64, 1215.
24. Yamansarova, E. T.; Kukovinets, A. G.; Kukovinets, O. S.; Zainullin, R. A.; Galin,
F. Z.; Kunakova, R. V.; Zorin, V. V.; Tolstikov, G. A. Russ. J. Org. Chem. 2001, 37,
246.
25. Westling, J.; Yowell, C. A.; Majer, P.; Erickson, J. W.; Dame, J. B.; Dunn, B. M. Exp.
Parasitol. 1997, 87, 185.
26. For a review, see: Bjelic, S.; Nervall, M.; Gutierrez-de-Terán, H.; Ersmark, K.;
Hallberg, A.; Åqvist, J. Cell. Mol. Life Sci. 2007, 64, 2285.
27. Gutiérrez-de-Terán, H.; Nervall, M.; Ersmark, K.; Liu, P.; Janka, L. K.; Dunn, B.;
Hallberg, A.; Åqvist, J. Biochemistry 2006, 45, 10529.
28. Hidaka, K.; Kimura, T.; Tsuchiya, Y.; Kamiya, M.; Ruben, A. J.; Freire, E.; Hayashi,
Y.; Kiso, Y. Bioorg. Med. Chem. Lett. 2007, 17, 3048.
v
dw
el
l—s
where hVl—s i and hV i denote, respectively, the Lennard–Jones
and electrostatic interactions between the ligand and its surround-
ings (l–s). These interactions are evaluated as energy averages (de-
noted by the open brackets) from the two separate MD simulations
of the free (solvated in water) and bound states. The difference (D)
between such averages for each type of potential is scaled by dif-
ferent coefficients,50 giving the polar and non-polar contributions
to the binding free energy. For the non-polar contribution, this
coefficient has been empirically set to
a = 0.181, while for the polar
contribution the scaling factor is dependent on the chemical nature
of the ligand. For ligands considered in this work the value is
b = 0.33.
29. Brynda, J.; Rezacova, P.; Fabry, M.; Horejsi, M.; Stouracova, R.; Sedlacek, J.;
Soucek, M.; Hradilek, M.; Lepsik, M.; Konvalinka, J. J. Med. Chem. 2004, 47, 2030.
30. Janka, L.; Clemente, J.; Vaiana, N.; Sparatore, A.; Romeo, S.; Dunn, B. M. Protein
Pept. Lett. 2008, 15, 868.
31. Liu, X. W.; Hu, X. E.; Tian, X. R.; Mazur, A.; Ebetino, F. H. J. Organomet. Chem.
2002, 646, 212.
32. Fraisse-Jullien, R.; Fréjaville, C. Bull. Soc. Chim. Fr. 1970, 219.
33. Eistetter, K.; Wolf, H. P. O. J. Med. Chem. 1982, 25, 109.
34. Mimoto, T.; Kato, R.; Takaku, H.; Nojima, S.; Terashima, K.; Misawa, S.;
Fukazawa, T.; Ueno, T.; Sato, H.; Shintani, M.; Kiso, Y.; Hayashi, H. J. Med. Chem.
1999, 42, 1789.
35. Tanimori, S.; Naka, T.; Kirihata, M. Synth. Commun. 2004, 34, 4043.
36. Thompson, W. J.; Fitzgerald, P. M. D.; Holloway, M. K.; Emini, E. A.; Darke, P. L.;
Mckeever, B. M.; Schleif, W. A.; Quintero, J. C.; Zugay, J. A.; Tucker, T. J.;
Schwering, J. E.; Homnick, C. F.; Nunberg, J.; Springer, J. P.; Huff, J. R. J. Med.
Chem. 1992, 35, 1685.
Acknowledgements
The authors from Uppsala University would like to acknowl-
edge the Swedish Foundation for Strategic Research, the Swedish
Research Council and Alice and Knut Wallenberg0s Foundation for
financial support, Professor Anders Hallberg for valuable intellec-
tual contributions, Mr. Jonas Sävmarker and Dr. Luke Odell for
purification of intermediates, Dr. Aleh Yahorau for performing
the HRMS experiments and Medivir AB, Huddinge, Sweden, for
conducting the inhibition assays with hCatD, HIV-1 protease and
BACE-1. HGT is an Isidro Parga Pondal Research fellow (Xunta de
Galicia, Spain) and acknowledges a bursary from mobility research
program of Xunta de Galicia, for the stay at Uppsala University.
Supercomputational centre of Galicia—CESGA has provided the
computer allocation time.
37. Westling, J.; Cipullo, P.; Hung, S. H.; Saft, H.; Dame, J. B.; Dunn, B. M. Protein Sci.
1999, 8, 2001.
38. Haque, T. S.; Skillman, A. G.; Lee, C. E.; Habashita, H.; Gluzman, I. Y.; Ewing, T. J.
A.; Goldberg, D. E.; Kuntz, I. D.; Ellman, J. A. J. Med. Chem. 1999, 42, 1428.
39. Copeland, R. A. In Enzymes: A Practical Introduction to Structure, Mechanism, and
Data Analysis; Copeland, R. A., Ed.; VCH: New York, NY, 1996; p 300.
40. Cheng, Y.; Prusoff, W. H. Biochem. Pharmacol. 1973, 22, 3099.
41. Danielson, U. H.; Lindgren, M. T.; Markgren, P.-O.; Nillroth, U. Adv. Exp. Med.
Biol. 1998, 436, 99.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2009.06.065.
42. Nillroth, U.; Vrang, L.; Markgren, P.-O.; Hulten, J.; Hallberg, A.; Danielson, U. H.
Antimicrob. Agents Chemother. 1997, 41, 2383.

K. M. Orrling et al. / Bioorg. Med. Chem. 17 (2009) 5933–5949
5949
43. Bäck, M.; Nyhlén, J.; Kvarnström, I.; Appelgren, S.; Borkakoti, N.; Jansson, K.;
46. Jorgensen, W. L.; Maxwell, D. S.; TiradoRives, J. J. Am. Chem. Soc. 1996, 118,
11225.
Lindberg, J.; Nyström, S.; Hallberg, A.; Rosenquist, A.; Samuelsson, B. Bioorg.
Med. Chem. 2008, 16, 9471.
44. Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R. J. Mol. Biol. 1997, 267,
47. Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L.
J. Chem. Phys. 1983, 79, 926.
727.
48. Lee, F. S.; Warshel, A. J. Chem. Phys. 1992, 97, 3100.
49. Åqvist, J.; Medina, C.; Samuelsson, J. E. Protein Eng. 1994, 7, 385.
50. Hansson, T.; Marelius, J.; Åqvist, J. J. Comput. Aided Mol. Des. 1998, 12, 27.
45. Marelius, J.; Kolmodin, K.; Feierberg, I.; Åqvist, J. J. Mol. Graph. Model. 1998, 16,
213.