A new structural theme in C2-symmetric HIV-1 protease inhibitors: ortho-Substituted P1/P1′ side chains

Article abstract of DOI:10.1016/j.bmc.2006.03.045

In this report, the rapid syntheses of 24 novel C₂-symmetric HIV-1 protease inhibitors are described. Two ortho-iodobenzyloxy containing C-terminal duplicated inhibitors served as starting materials for microwave-enhanced palladium(0)-catalyzed

Full text of DOI:10.1016/j.bmc.2006.03.045

Bioorganic & Medicinal Chemistry 14 (2006) 5303–5315
A new structural theme in C2-symmetric HIV-1 protease inhibitors:
ortho-Substituted P1/P10 side chains
a
a
b
b
a
´
Johan Wannberg, Yogesh A. Sabnis, Lotta Vrang, Bertil Samuelsson, Anders Karlen,
Anders Hallberga and Mats Larheda,*
aOrganic Pharmaceutical Chemistry, Department of Medicinal Chemistry, Uppsala University, BMC,
Box 574, SE-751 23 Uppsala, Sweden
bMedivir AB, Lunastigen 7, SE-141 44 Huddinge, Sweden
Received 28 December 2005; revised 21 March 2006; accepted 24 March 2006
Available online 18 April 2006
Abstract—In this report, the rapid syntheses of 24 novel C2-symmetric HIV-1 protease inhibitors are described. Two ortho-iodo-
benzyloxy containing C-terminal duplicated inhibitors served as starting materials for microwave-enhanced palladium(0)-catalyzed
carbon–carbon bond forming reactions (Suzuki, Sonogashira, Heck, and Negishi). Highly potent inhibitors equipped with ortho-
functionalized P1/P10 side chains as the structural theme were identified. Computational efforts were applied to study the binding
mode of this class of inhibitors and to establish structure–activity relationships. The overall orientation of the inhibitors in the active
site was reproduced by docking which suggested three possible conformations of the P1/P10 groups of which two seem more
plausible.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
have employed the stereochemistry of L-mannitol for
the straightforward synthesis of linear C2-symmetric,
C-terminal duplicated inhibitors.6 The carbohydrate-
based scaffold in these inhibitors, containing a central
dihydroxyethylene transition-state isoster unit, has been
utilized in several studies with the intent to optimize the
binding of the P1/P10 and P2/P20 side chains to the C2-
symmetric HIV-1 protease.7,8 In particular, palladi-
um(0)-catalyzed coupling chemistry has proven to be
effective in modifying halide functionalized benzyl ether
P1/P10 side chains.9,10 Guidance in the design of new
inhibitors with novel substituents has come from crystal
structures of HIV-1 protease coupled with different
inhibitors.1,6,9
The HIV-1 protease is an essential enzyme for HIV-1
replication and constitutes an important target in the
treatment of HIV/AIDS.1–3 Efficient combination thera-
pies using inhibitors of the reverse transcriptase and
protease enzymes have led many to reevaluate HIV-1
infections from a terminal condition to a chronic-but-
manageable disease in the developed world. Unfortu-
nately, the emergence of drug resistant viral strains
and severe treatment-related adverse effects limit the
benefits of current anti-HIV/AIDS drugs for many pa-
tients.4 Furthermore, less than one in ten patients infect-
ed with HIV-1 in low- and middle-income countries
have access to proper treatments. These important
shortcomings highlight the need for new, cost effective
anti-HIV/AIDS drugs with unique properties.5
In our systematic mapping of the P1/P10 groups interact-
ing with S1/S10 sites, we recently published Mo(CO)6-
mediated aminocarbonylations in order to investigate
the effects of ortho-functionalizations on the activities
of these types of inhibitors in an expedient way.11 It
was initially speculated that bulky ortho-substituents
might collide with the protease surface. However, to
our surprise we found that compounds with large anilide
substituents in the ortho-position of the P1/P10
benzyloxy group showed nanomolar inhibition
(Fig. 1).11 Nevertheless, because of the presence of a
We are presently engaged in a program aiming for the
development of simple and inexpensive carbohydrate-
based HIV-1 protease inhibitors.6 In the project, we
Keywords: HIV-1 protease inhibitors; Microwave; High-speed
synthesis; 3D QSAR CoMFA; Palladium; Catalysis; Docking studies.
*
Corresponding author. Tel.: +46 18 471 4667; fax: +46 18 471
4474; e-mail: mats@orgfarm.uu.se
0968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.03.045

5304
J. Wannberg et al. / Bioorg. Med. Chem. 14 (2006) 5303–5315
The introduction of a non-functionalized vinyl group
P1’
by substitution of an aryl halide can be achieved by
different palladium-catalyzed methods, the most com-
mon perhaps being the Stille cross-coupling with trib-
utyl(vinyl)tin or the Heck reaction using ethene under
pressure.14 The usage of toxic tin reagents or the need
for a reactive gas are obvious drawbacks of these
strategies. A more attractive alternative would be to
use a vinylboronic acid derivative in a Suzuki reaction.
Free, monomeric vinylboronic acid is relatively unsta-
ble but the bench stable derivative 2,4,6-trivinylcyclo-
triboroxane–pyridine complex has been described as
a useful coupling partner.15 This commercially avail-
able reagent was utilized in a double Suzuki coupling
to produce vinyl derivative 4a in 57% yield after only
15 min of irradiation. The swift and straightforward
Suzuki-microwave methodology was also applied to
produce 4e (65%), 4f (44%) and 4j (19%) from the cor-
responding boronic acids and parent 3. Offsetting the
advantages of the Suzuki reaction are some detrac-
tions. With non-gaseous olefins and acetylenes, direct
Heck- and Sonogashira reactions are performed in
high atom-efficiency avoiding the usage of organome-
tallic reactants. The difficulty in introducing benzylic
groups by the Suzuki methodology is another
drawback. Therefore, continued modifications of the
P1/P10 side chains were conducted with complementa-
ry reaction protocols.
H
N
R
P2
O
H
N
OH
O
O
O
H
N
N
H
N
H
O
O
OH O
O
P2’
R
N
H
1 (R=H) Ki, 20 nM
2 (R=Cl) Ki, 8.5 nM
P1
Figure 1.
large number of amide bonds, these inhibitors were rath-
er unattractive from a drug development perspective.
The aim of the present study was to further evaluate the
concept of ortho-substitution in the P1/P10 positions by
replacing the amide functionality with vinyl, alkynyl,
acyl or aryl substituents while achieving similar or high-
er potencies. We envisioned the use of microwave-assist-
ed chemoselective carbon–carbon bond forming
reactions to introduce the alternative substituents.12 In
order to further reduce the peptide character it was
decided to also exploit the well-known N-(1S,2R)-2-hy-
droxy-1-indanyl moiety as a P2/P20 group (e.g., like in
indinavir13).
Here, we report the rapid preparation and biological
evaluation of 24 novel HIV-1 protease inhibitors. The
low nanomolar activity values of the inhibitors with
these rather large P1/P10 ortho-groups lead us to further
investigate the nature of their binding by employing
computational techniques. As a result, docking studies
were performed in an attempt to identify the possible
binding mode of these inhibitors and a CoMFA model
was derived based on the proposed binding
conformation.
The acetylene units of 4b were incorporated by a proce-
dure where 3 was initially reacted with trimethylsilylacet-
ylene in a Sonogashira coupling followed by basic
removal of the silyl group to give 4b in 37% two-step
yield. In another run, the TMS–acetylene containing
inhibitor 4g was isolated (36%). The methyl ketone deriv-
ative 4c was obtained by a two-step sequence. First, a
regioselective internal Heck coupling of 2-hydroxyethyl
vinyl ether with 3 afforded the 1,3-dioxolane protected
ketone.16 After an acidic deprotection of the 1,3-dioxo-
lane, methyl ketone 4c could be isolated in an excellent
85% overall yield from 3. Reactions of methyl acrylate
and styrene with 3 under standard Heck conditions
produced methyl cinnamate derivative 4d and stilbene
derivative 4i in 55% and 65% yield, respectively. Last
in the series, phenylacetylene and 3 were coupled via
Sonogashira chemistry to give 4h in a satisfying 93%
yield (Table 1).
2. Results and discussion
2.1. Chemistry
The synthesis of C2-symmetrical aryl palladium precur-
sor 3 was carried out according to the previously pub-
lished strategy.6,9,11 With aryl iodo substrate 3 in
hand, an abundance of possibilities to explore palladi-
um(0)-catalyzed carbon–carbon bond forming reactions
in order to produce new HIV-1 protease inhibitors were
recognized.14 From a synthetic perspective, the major
goal was to identify robust conditions for decoration
of non-protected precursor 3 without undesired water
elimination of the diol. A second goal was to keep reac-
tion times to a minimum, preferably 30 min or shorter
under non-inert conditions. Specific coupling reactants
were selected according to commercial availability and
in order to achieve some degree of diversity in the size
and flexibility of the P1/P10-modified inhibitors. In addi-
tion, the styrene and benzofuran type of ortho-substitu-
ents might be viewed as isosteres to the active ortho-
anilides in Figure 1.
The aminoindanol moiety present in indinavir was used
as P2/P20 group in a second series of inhibitors (Table 2).
The same synthetic route as for 3 was exploited but in
this case the bis-lactone 611 was stirred with (1S,2R)-1-
amino-2-indanol in 1,2-dichloroethane at 50 ꢁC for
4 h, conveniently affording bis-aryl-iodide 7 in 55%
isolated yield and in high purity (Scheme 1).
In an initial attempt to evaluate the potential of indanol-
amine containing inhibitors originating from 7, com-
pound 8 was synthesized by aminocarbonylation17,18
of 7 with aniline as the nucleophile and Mo(CO)6 as car-
bon monoxide source (Scheme 2).19,20 Rewardingly,
ortho-anilide 8 proved to be an equipotent inhibitor to
valine containing 1 (1, Ki = 20 nM; 8, Ki = 18 nM). This

J. Wannberg et al. / Bioorg. Med. Chem. 14 (2006) 5303–5315
Table 1. Synthesis and biological evaluation of inhibitors 4a–k
5305
R
I
OH
O
O
O
OH
O
O
O
H
N
H
N
H
N
H
N
[Pd], microwaves
N
H
N
H
N
H
N
H
O
O
OH
O
O
O
OH O
R
I
4a-j
3
Compound
Reactant
—
Time (min)
—
Temp (ꢁC)
R
Isolated yield (%)
—
Ki (nm)
3
—
4.3
I
B
57a
36b,c
85d
55e
65a
44a
37b
93b
65e
19a
—
0.41
x pyridine
4a
4b
4c
4d
4e
4f
15
15
30
5
100
120
100
120
120
120
120
100
120
120
—
O
B
O
B
O
17
Si
O
OH
1100
760
290
1900
860
88
O
O
O
O
O
S
S
30
30
15
15
5
(HO)2B
(HO)2B
Si
4g
4h
4i
Si
120
1600
O
O
4j
30
(HO)2B
4k
—
—
0.4f
H
a Pd(OAc)2, [(t-Bu)3PH]BF4, K2CO3, DME, and H2O.
b Pd(PPh3)2Cl2, CuI, Et2NH, and CH3CN.
c Then K2CO3 (satd) in MeOH, rt, 2 h.
d Pd(OAc)2, DPPP, TlOAc, (i-Pr)2EtN, DMF, and H2O.
e Pd2(OAc)2(P(o-tol)3)2, (i-Pr)2EtN, DMF, and H2O.
f Data from Ref. 6.
result encouraged us to proceed with the second series as
planned (Scheme 2).
pling of 7 with benzylzinc bromide delivered 9g in a
disappointing 25% yield but without detected epimeriza-
tion or water elimination. To prepare 9h, 2-bromothi-
azole was first reacted with the 2,4,6-trivinyl-
cyclotriboroxane–pyridine complex in a Suzuki reac-
tion. After a quick workup, the formed 2-vinylthiazole
was subsequently coupled with 7 under Heck conditions
to produce 9h in 62% isolated yield.
The identical protection group free strategy of Suzuki,
Heck, and Sonogashira couplings was used for the syn-
thesis of the aminoindanol series of inhibitors. Generally
similar yields were obtained for these compounds (Table
2) compared to those in Table 1. Further, a Negishi cou-

5306
J. Wannberg et al. / Bioorg. Med. Chem. 14 (2006) 5303–5315
Table 2. Synthesis and biological evaluation of inhibitors 9a–m
I
R
OH
OH
OH
O
OH O
O
O
H
N
H
N
[Pd], microwaves
N
H
N
H
O
OH
O
O
OH O
OH
OH
7
9a-l
I
R
Compound
Reactant
—
Time (min)
—
Temp (ꢁC)
R
Isolated yield (%)
Ki (nm)
7
—
33
I
51a
14
B
x pyridine
9a
15
100
O
B
O
B
O
53b,c
90d
13
Si(Me)3
OH
9b
9c
15
30
120
100
O
190
O
O
O
9d
9e
9f
5
30
30
5
120
120
120
100
120
120
100
120
57e
53a
29a
25f
62e
44b
72b
63e
45a
660
100
450
14
O
O
S
S
(HO)2B
(HO)2B
BrZn
9g
9h
9i
N
N
S
10
15
15
5
220
21
S
Si(Me)3
Si(Me)3
9j
200
4.1
9k
O
O
9l
30
—
120
—
220
0.2g
(HO)2B
9m
—
H
a Pd(OAc)2, [(t-Bu)3PH]BF4, K2CO3, DME, and H2O.
b Pd(PPh3)2Cl2, CuI, Et2NH, and CH3CN.
c Then K2CO3 (satd) in MeOH, rt, 4 h.
d Pd(OAc)2, DPPP, TlOAc, (i-Pr)2EtN, DMF, and H2O.
e Pd2(OAc)2(P(o-tol)3)2, (i-Pr)2EtN, DMF, and H2O.
f Pd(OAc)2, [(t-Bu)3PH]BF4, and THF.
g Data from Ref. 6.
2.2. Biological results
inhibition constants (Ki-values) were determined (Tables
1 and 2). In the case of P2 valine methylamides, the small
ortho-substituents on the P1/P10 benzyloxy groups gave
the most potent inhibitors in this series with activities
The synthesized compounds were tested for their ability
to inhibit HIV-1 protease in an enzyme-based assay and

J. Wannberg et al. / Bioorg. Med. Chem. 14 (2006) 5303–5315
5307
stilbene derivative 9k (4.1 nM), which also was the most
active compound in the second series. The TMS-ethynyl
derivative 9i was more potent than diphenylacetylene 9j.
The benzofuran derivative 9l (220 nM) exhibited similar
activity to the vinyl thiazole 9h (220 nM). In order to
rationalize the structure–activity relationships and to
elucidate the binding mode of especially the ortho-func-
tionalized aminoindanols, docking studies along with
3D QSAR CoMFA were employed.
L-Mannonic-γ-
lactone
OH
HO
O
HO
O
HO
a
OH
O
O
O
5, 53%
6, 80%
O
HO
I
2.3. Computational studies
b
2.3.1. Molecular modeling. Three-dimensional structure
building and all modeling was performed using the
SYBYL program package, version 6.921 and
FLO96.22 The FLO+ module in FLO96, (also called
QXP, Quick eXPlore), was used to dock the molecules
into the active site of the HIV-1 protease. FLO96
search algorithms were derived from the method of
Monte Carlo perturbation using a modified version of
the AMBER force field. The molecular modeling is dis-
cussed in detail in Section 4.
O
O
O
O
O
O
I
c
I
OH
OH
O
O
H
N
2.3.2. Protein structure and docking results. We chose
PDB23 entry 1EC0, resolved at 1.79 A, as a starting
˚
N
H
O
OH O
OH
point for our docking studies since the co-crystallized
P1/P10 ortho-fluorobenzyloxy inhibitor in this X-ray
crystal structure most closely resembled the compounds
under investigation.24 After minimization of the protein,
I
7, 55%
˚
Scheme 1. Reagents and conditions: (a) HNO3, 90 ꢁC, 5 h; (b)
o-iodobenzyl-2,2,2-trichloroacetimidate, BF3 Æ Et2O, rt, 18 h; (c)
(1S,2R)-1-amino-indan-2-ol, 50 ꢁC, 4 h.
the resulting structure had a heavy atom rmsd of 0.26 A
compared to the heavy atoms of the starting structure.
A low rmsd suggested no significant change from the
original crystal structure.
H
N
2.3.3. Common structural features of the inhibitor com-
plexes. Most of the docked inhibitors were bound to the
enzyme active site in an extended conformation. Thus,
when they were superimposed upon one another, their
functional elements aligned quite well. The dihedral an-
gles in the backbone of all the inhibitors were main-
tained and the contacts between the backbone atoms
of all the inhibitors and the protease were consistent.
The hydrogen bonds were typically seen between the
backbone atoms of the enzyme and the inhibitor. It
has previously been shown that C2-symmetric inhibitors
bind to the protease in an asymmetric fashion.25 We
experienced similar trends with the investigated class
of inhibitors 4 and 9 where one of the hydroxyls of
the diol was positioned between the D25/250 carboxyls
of the protease, within hydrogen bonding distance to
at least one carboxylate oxygen of each aspartate. Due
to the lack of large P1/P10 ortho-substituted inhibitors
reported in the literature, focus was applied on detailing
the binding of these groups in the S1/S10 pockets.
Although the docked backbone along with the P1/P10
benzyloxy groups of these inhibitors showed interac-
tions reported by Unge’s group earlier,9,24 the various
ortho-substituents on these benzyloxy groups could be
found positioned in three different areas of the enzyme.
When studying the different docking conformations it
was seen that all inhibitors could place their ortho-sub-
stituents in these areas with similar energies making it
difficult to select the most probable binding mode for
Aniline, Pd(OAc)2
Mo(CO)6, DBU
O
OH
OH
O
O
H
N
7
THF, 110 oC, 15 min
microwaves
N
H
O
OH O
HO
O
N
H
8, 45%
Scheme 2. Synthesis of 8 by aminocarbonylation of 7 with aniline.
of 4.3, 0.41, and 17 nM for 3, 4a and 4b, respectively. The
P1/P10 ketone 4c, gave a great reduction in the potency
(Ki = 1100 nM). The incorporation of aromatic substitu-
ents was not well tolerated as can be seen from the de-
crease in their potencies, with the biphenyl derivative
being the least active (1900 nM, 4f). Table 2 shows the
P2/P20 aminoindanol series where small lipophilic
ortho-substitutions of the P1/P10 benzyloxy side chains
(7, 9a, and 9b) yielded inhibitors with low nanomolar
activities. Polar acyclic substitutions in 9c and 9d provid-
ed decreased activities of 190 nM and 660 nM. When
comparing 9f (450 nM) with 9g (14 nM), it was clear that
the introduction of a flexible one-carbon linker between
the two phenyl rings was highly advantageous. Exchang-
ing the phenyl in 9f to a thiophene moiety furnished im-
proved potency (100 nM, 9e). Methyl cinnamate 9d and
vinyl thiazole 9h resulted in lower activity compared to

5308
J. Wannberg et al. / Bioorg. Med. Chem. 14 (2006) 5303–5315
the two series of inhibitors. Interestingly, upon closer
visual inspection it was revealed that two of these
observed conformations corresponded to the binding
mode of co-crystallized inhibitors reported earlier8,27
and is discussed below.
The potent stilbene derivative 9k was used as a represen-
tative compound to describe the observed binding
modes. In conformation A, the terminal phenyl rings
could be seen occupying the space between F53/530
and P81/810 in a fashion similar to the para-substituted
pyridines in the P1 and P10 positions of 1EC28,9
(Fig. 2). The binding mode in conformation B was con-
sistent with the X-ray structure of the macrocyclic tri-
peptide analogs found in PDB entry, 1D4K26 (Fig. 2).
In the third unprecedented binding mode, the phenyl
ring was observed lying in the grove formed by the
hydrophobic side chains of R8/80, L10/100, L23/230,
and V82/820. Figure 3I shows the various conformations
for the ortho-substituted stilbene analog 9k and Figure
3II illustrates the corresponding hydrogen bonding net-
work for conformation A, including structural waters.
To find out if any quantitative correlation existed be-
tween these structures and their biological activities,
we performed a CoMFA analysis. The inhibitors with
conformation A have a more close structural resem-
blance to the corresponding co-crystallized ligand (BEJ
in 1EC2),8,9 as a result we decided to work with confor-
mation A. PLS analysis with LOO cross-validation
yielded a three-component model with a reasonable q2
of 0.422 and a standard error of prediction of 0.86.
The non-cross-validated PLS analysis gave a conven-
tional r2 of 0.83. The steric and electrostatic contribu-
tions amounted to 48% and 52%, respectively. These
analyses suggested a substantial correlation between
the structures and their respective biological activities.
Although these CoMFA analyses pertain only to
Figure 2. (a) Compound 9k in conformation A over the symmetric P1/
P10 pyridine substituted inhibitor co-crystallized with HIV-1 protease
(PDB entry: 1EC2). (b) Inhibitor 9k in conformation B over the
macrocylic inhibitor co-crystallized with HIV-1 protease (PDB entry:
1D4K). Carbon atoms of 9k are colored orange, all other atoms are
colored according to Sybyl atom types. Only polar hydrogens are
displayed for clarity reasons.
B
A
C
I
II
Figure 3. (I) Connolly surface color-coded with cavity depth is shown with the three probable binding regions for the P1 and P10 substituents which
are colored in green (conformation A), orange (conformation B) and purple (conformation C). (II) Hydrogen bonding network in conformation A
between representative compound 9k and the protease with structural waters. The carbon atoms of the inhibitor are colored dark grey while all other
atoms are colored according to Sybyl atom types. Only polar hydrogens are displayed for clarity reasons

J. Wannberg et al. / Bioorg. Med. Chem. 14 (2006) 5303–5315
5309
conformation A, we believe that any of the modes dis-
cussed above could be a plausible binding mode for
these inhibitors.
4.2. HIV protease inhibition
The HIV-1 protease was cloned and heterologously ex-
pressed in Escherichia coli and purified as described else-
where.27 The Ki values for the synthesized compounds
were determined from two individual measurements by
a fluorometric assay.28 Assay variability was checked
by inclusion a known inhibitor and the standard devia-
tion for the enzyme assays was 50% of the mean.
3. Conclusions
In this investigation, the synthesis, biological evaluation
and docking of 24 novel HIV-1 protease inhibitors were
presented. Knowledge gained from a previous study
using carbonylation methodology guided the design of
this new set of more drug-like compounds. Two series
of microwave-accelerated Suzuki, Heck, Sonogashira,
and Negishi reactions were performed to rapidly modify
the aryl iodide core structure and it was demonstrated
that potent inhibitors containing ortho-substituted P1/
P10 side chains could be isolated. In general, small lipo-
philic 1-2 carbon-tethered aromatic ortho-substitutions
of the P1/P10 benzyloxy side chains provided inhibitors
of high potency as compared to sterically demanding
biaryl analogs. Docking studies were found to be in
good agreement with earlier reported interactions in
similar compounds co-crystallized with the protein.
The ortho-substituents in P1/P10 showed three plausible
binding modes of which two coincided with the confor-
mations of known crystal structures in the PDB. The
alignment obtained from the docking studies enabled
us to derive a 3D QSAR model for the present set of
inhibitors.
4.3. Chemistry
4.3.1. General procedure for the Suzuki couplings. Synthe-
ses of 4a, 4e, 4f, 4j, 9a, 9e, 9f, and 9l. A mixture of 3 or 7
(0.050 mmol), boronic acid (0.30 mmol), Pd(OAc)2
(1.1 mg, 0.0050 mmol), [(t-Bu)3PH]BF4 (3.0 mg,
0.010 mmol) and K2CO3 (41.5 mg, 0.30 mmol), H2O
(0.30 mL) and 1,2-dimethoxyethane (1.0 mL) in a 2.0 mL
microwave vial was irradiated to 100 ꢁC for 15 min. The
organic layer of the reaction mixture was filtered through
Celite and the solvent evaporated under reduced pressure.
The residue was purified as indicated below.
4.3.2. General procedure for the Sonogashira couplings.
Syntheses of 4g, 4h, 9i, and 9j. A mixture of 3 or 7
(0.050 mmol), alkyne (0.20 mmol), Pd(PPh3)2Cl2
(1.8 mg, 0.0025 mmol), CuI (1.9 mg, 0.010 mmol), dieth-
ylamine (0.50 mL), and acetonitrile (0.50 mL) in a
2.0 mL microwave vial was irradiated at the indicated
temperature for 15 min. The reaction mixture was fil-
tered through Celite and the solvent evaporated under
reduced pressure. The residue was purified as indicated
below.
4. Experimental
4.1. General information
4.3.3. N1,N6-Bis[(1S)-2-methyl-1-(methylcarbamoyl)pro-
pyl]-(2R,3R,4R,5R)-2,5-bis(2-vinylbenzyloxy)-3,4-dihydr-
oxyhexane-1,6-diamide (4a). The 2,4,6-trivinylcyclo-
triboroxane–pyridine complex (24.1 mg, 0.10 mmol)
was used as a boronic acid equivalent. The residue was
purified by silica gel flash chromatography (3–5% MeOH
in CHCl3) to give 4a as a colorless solid in 57% yield
1H and 13C NMR spectra were recorded on a Varian
Mercury-400 spectrometer or a Jeol JNM-EX400 spec-
trometer at 400 and 100.5 MHz, respectively. Chemical
shifts were referenced directly to tetramethylsilane (sol-
vent mixtures) or indirectly via the residual solvent sig-
nal (pure solvents). All microwave-assisted synthesis
was carried out in a Smith/EmrysTM Synthesizer sin-
gle-mode microwave cavity producing controlled irra-
diation at 2450 MHz (Biotage AB, Uppsala, Sweden).
Thin-layer chromatography was performed on silica
gel 60 F-254 plates (E. Merck) and visualized with
UV light and ninhydrin. Flash column chromatogra-
phy was performed using silica gel 60 (0.040–
0.063 mm) (E. Merck). RP-LC–MS analysis of reaction
mixtures and pure products were performed using a
Gilson HPLC system with a Chromolith SpeedROD
RP-18e column (50 · 4.6 mm) and a Finnigan AQA
1
(19.1 mg). H NMR (400 MHz, CDCl3/CD3OD 1:1): d
7.57 (dd, J = 7.7, 1.7 Hz, 2H), 7.39–7.23 (m, 6H), 7.08
(dd, J = 17.4, 11.0 Hz, 2H), 5.72 (dd, J = 17.4, 1.5 Hz,
2H), 5.35 (dd, J = 11.0, 1.5 Hz, 2H), 4.78 (d,
J = 11.5 Hz, 2H), 4.69 (d, J = 11.5 Hz, 2H), 4.26 (d,
J = 5.7 Hz, 2H), 4.20 (AA0 part of AA0BB0, 2H), 4.10
(BB0 part of AA0BB0, 2H), 2.74 (s, 6H), 2.29–2.17 (m,
2H), 0.94 (d, J = 6.9 Hz, 6H), 0.86 (d, J = 6.9 Hz, 6H).
13C NMR (100.5 MHz, CDCl3/CD3OD 1:1): d 172.8,
172.4, 137.9, 134.39, 134.37, 130.3, 129.3, 128.3, 126.4,
116.8, 80.6, 72.1, 71.1, 58.9, 30.7, 26.1, 19.6, 17.7. Anal.
Calcd for C36H50N4O8+0.3H2O: C, 64.32; H, 7.59; N,
8.33. Found: C, 63.92; H, 7.49; N, 8.17.
quadropole mass spectrometer using
a 4 mL/min
CH3CN/H2O gradient (0.05% HCOOH) and detection
by UV (DAD) and MS (ESI+). Preparative RP-LC–
MS purifications were performed by UV-triggered frac-
tion collection with a similar Gilson–Finnigan AQA
system and a Zorbax SB C8 column (150 · 21.2 mm),
using a 15 mL/min CH3CN/H2O gradient (0.05%
HCOOH) with UV (214, 255 nm) and MS (ESI+)
detection. Exact molecular masses were determined
on Micromass Q-Tof2 mass spectrometer equipped
with an electrospray ion source.
4.3.4. N1,N6-Bis[(1S)-2-methyl-1-(methylcarbamoyl)pro-
pyl]-(2R,3R,4R,5R)-2,5-bis(2-ethynylbenzyloxy)-3,4-di-
hydroxyhexane-1,6-diamide (4b). Sonogashira coupling
at 120 ꢁC with trimethylsilylacetylene. To the crude
product was added 2 mL of a saturated solution of
K2CO3 in MeOH and the mixture was stirred at room
temperature for 2 h. The solvent was evaporated and
the residue was purified by silica gel flash chromatogra-

5310
J. Wannberg et al. / Bioorg. Med. Chem. 14 (2006) 5303–5315
1
(21.5 mg, colorless solid). H NMR (400 MHz, CDCl3/
phy (5% MeOH in CHCl3) to give 12.0 mg of 4b (36%,
light yellow solid). 1H NMR (400 MHz, CDCl3/CD3OD
1:1): d 7.59–7.24 (m, 4H), 7.39 (dt, J = 7.6, 1.8 Hz, 2H),
7.31 (dt, J = 7.6, 1.7 Hz, 2H), 4.88 (d, J = 12.0 Hz, 2H),
4.80 (d, J = 12.0 Hz, 2H), 4.28–4.21 (m, 4H), 4.18–4.15
(m, 2H), 3.59 (s, 2H), 2.75 (s, 6H), 2.27–2.16 (m, 2H),
0.94 (d, J = 6.9 Hz, 6H), 0.87 (d, J = 6.9 Hz, 6H). 13C
CD3OD 1:1): d 8.04 (d, J = 15.9 Hz, 2H), 7.67–7.63 (m,
2H), 7.47–7.36 (m, 6H), 6.42 (d, J = 15.9 Hz, 2H), 4.86
(d, J = 11.7 Hz, 2H), 4.68 (d, J = 11.7 Hz, 2H), 4.27–
4.20 (m, 4H), 4.17–4.14 (m, 2H), 3.75 (s, 6H), 2.74 (s,
6H), 2.28–2.18 (m, 2H), 0.94 (d, J = 6.9 Hz, 6H), 0.87
(d, J = 6.9 Hz, 6H). 13C NMR (100.5 MHz, CDCl3/
CD3OD 1:1): d 172.9, 172.3, 168.5, 142.5, 136.6, 134.2,
130.84, 130.75, 129.6, 127.4, 119.9, 80.9, 71.8, 70.8,
58.9, 52.2, 30.5, 26.2, 19.7, 17.7. Anal. Calcd for
C40H54N4O12+0.5H2O: C, 60.67; H, 7.00; N, 7.08.
Found: C, 60.5; H, 6.9; N, 7.0.
NMR (100.5 MHz, CDCl3/CD3OD 1:1):
d 173.0,
172.5, 139.8, 133.5, 129.6, 129.2, 128.6, 122.2, 83.3,
81,7, 81.0, 72.2, 71.5, 59.0, 30.9, 26.2, 19.6, 17.9. Anal.
Calcd for C36H46N4O8+0.25H2O: C, 64.80; H, 7.02; N,
8.40. Found: C, 64.7; H, 7.0; N, 8.1.
4.3.5. N1,N6-Bis[(1S)-2-methyl-1-(methylcarbamoyl)pro-
pyl]-(2R,3R,4R,5R)-2,5-bis(2-acetylbenzyloxy)-3,4-di-
hydroxyhexane-1,6-diamide (4c). A 2.0 mL microwave
4.3.7. N1,N6-Bis[(1S)-2-methyl-1-(methylcarbamoyl)pro-
pyl]-(2R,3R,4R,5R)-2,5-bis(2-thiophen-3-ylbenzyloxy)-3,4-
dihydroxyhexane-1,6-diamide (4e). Suzuki coupling with
3-thienylboronicacid. Theresiduewaspurifiedbyprepara-
tive RP-LC–MS (30-90% MeCN in H2O, 25 min gradient)
to give 4e in 65% yield (25.3 mg, colorless solid). 1H NMR
(400 MHz,CDCl3/CD3OD1:1):d7.58–7.55(m, 2H),7.41–
7.34 (m, 10H), 7.18 (dd, J = 4.8, 1.7 Hz, 2H), 4.62 (s, 4H),
4.24 (d, J = 5.6 Hz, 2H), 4.12 (AA0 part of AA0BB0, 2H),
4.02 (BB0 part of AA0BB0, 2H), 2.72 (s, 6H), 2.28-2.16 (m,
2H), 0.89 (d, J = 6.9 Hz, 6H), 0.82 (d, J = 6.9 Hz, 6H). 13C
NMR (100.5 MHz, CDCl3, CD3OD 1:1): d 172.9, 172.4,
141.2, 137.3, 134.9, 130.5, 130.3, 129.4, 128.8, 128.1, 126.1,
123.8,81.3,72.2,71.5,58.9,30.7,26.2,19.7,17.7.Anal.Calcd
for C40H50N4O8S2+0.5H2O: C, 60.97; H, 6.52; N, 7.11.
Found:C,60.95;H,6.54;N,6.98.
vial was charged with
Pd(OAc)2 (2.2 mg,
3 (43.3 mg, 0.050 mmol),
0.010 mmol), 1,3-bis(diph-
enylphosphino)propane (dppp) (4.4 mg, 0.010 mmol),
2-hydroxyethyl vinyl ether (0.045 mL, 0.50 mmol),
TlOAc (28.1 mg, 0.10 mmol), (i-Pr)2EtN (0.035 mL,
0.20 mmol) and 1.0 mL of DMF/H2O (9:1). The vial
was capped with a Teflon septum and irradiated with
microwaves to 100 ꢁC for 30 min. The dioxolane was
deprotected by addition of 2 mL of 0.5 M HCl in
MeCN/H2O (1:1). After stirring for 30 min at room tem-
perature, the reaction mixture was poured onto 40 mL
of 2 M HCl (aq) and extracted with 3· 40 mL CH2Cl2.
The combined organic layers were washed with brine
(40 mL), dried with MgSO4 and evaporated to dryness.
The residue was purified by silica gel flash chromatogra-
phy (2–4% MeOH in CHCl3) to give 29.5 mg of 4c (85%)
as a colorless solid. 1H NMR (400 MHz, CDCl3/
CD3OD 1:1): d 7.87 (dd, J = 7.8, 1.6 Hz, 2H), 7.72
(dd, J = 7.8, 1.5 Hz, 2H), 7.58 (dd, J = 7.6, 1.5 Hz,
2H), 7.46 (dd, J = 7.6, 1.5 Hz, 2H), 4.96 (d,
J = 13.2 Hz, 2H), 4.90 (d, J = 13.2 Hz, 2H), 4.30–4.25
(m, 2H), 4.21 (AA0 part of AA0BB0, 2H), 4.12 (BB0 part
of, J = 7.5, 1.9 Hz, 2H), 7.36–7.30 (m, 6H), 7.27–7.22
(m, 4H), 7.05 (d, J = 16.2, 2H), 4.84 (d, J = 11.8 Hz,
2H), 4.76 (d, J = 11.8 Hz, 2H), 4.24–4.14 (m, 6H), 2.68
(s, 6H), 2.21–2.10 (m, 2H), 0.86 (d, J = 6.8 Hz, 6H),
0.79 (d, J = 6.8 Hz, 6H). 13C NMR (100.5 MHz, CDCl3,
CD3OD 1:1): d 172.8, 172.3, 137.9, 137.5, 134.7, 131.9,
130.5, 129.4, 129.2, 128.4, 128.1, 127.3, 126.4, 125.6,
80.5, 72.3, 71.0, 58.8, 30.4, 26.1, 19.6, 17.5. Anal. Calcd
for C48H58N4O8+0.5H2O: C, 69.63; H, 7.18; N, 6.77.
Found: C, 69.6; H, 7.1; N, 6.7.
4.3.8. N1,N6-Bis[(1S)-2-methyl-1-(methylcarbamoyl)pro-
pyl]-(2R,3R,4R,5R)-2,5-bis(2-phenylbenzyloxy)-3,4-dihydr-
oxyhexane-1,6-diamide (4f). Suzuki coupling with
phenylboronic acid. The residue was purified by prepara-
tive RP-LC–MS (30–90% MeCN in H2O, 25 min gradi-
1
ent) to give 4f in 44% yield (16.9 mg, colorless solid). H
NMR (400 MHz, CDCl3/CD3OD 1:1): d 7.62–7.58 (m,
2H), 7.41–7.27 (m, 18H), 4.58 (s, 4H), 4.21 (d,
J = 5.6 Hz, 2H), 4.01–3.96 (m, 4H), 2.70 (s, 6H), 2.26–
2.15 (m, 2H), 0.86 (d, J = 6.8 Hz, 6H), 0.79 (d,
J = 6.8 Hz, 6H). 13C NMR (100.5 MHz, CDCl3/CD3OD
1:1): d 172.8, 172.4, 142.5, 141.1, 134.8, 130.6, 129.7,
129.6, 128.8, 128.6, 128.1, 127.9, 81.5, 72.4, 71.2, 58.8,
30.6, 26.1, 19.7, 17.7. Anal. Calcd for C44H54N4O8: C,
68.91; H, 7.10; N, 7.31. Found: C, 68.91; H, 7.10; N,
7.20.
4.3.9. N1,N6-Bis[(1S)-2-methyl-1-(methylcarbamoyl)pro-
pyl]-(2R,3R,4R,5R)-2,5-bis(2-trimethylsilanylethynylben-
zyloxy)-3,4-dihydroxyhexane-1,6-diamide (4g). Sonogashira
coupling at 120 ꢁC with trimethylsilylacetylene. The residue
was purified by preparative RP-LC–MS (30–100% MeCN
in H2O, 25 min gradient) to give 4g in 37% yield (14.9 mg,
yellow solid). 1H NMR (400 MHz, CDCl3/CD3OD 1:1): d
7.81 (q, 4.7 Hz, 2H), 7.55 (dd, J = 7.7, 1.7 Hz, 2H), 7.47
(dd, J = 7.6, 1.7 Hz, 2H), 7.37 (dt, J = 7.6, 1.6 Hz, 2H),
7.28 (dt, J = 7.6, 1.6 Hz, 2H), 4.89 (d, J = 12.6 Hz, 2H),
4.83 (d, J = 12.6 Hz, 2H), 4.31–4.25 (m, 4H), 4.19–4.16
(m, 2H), 2.73 (s, 6H), 2.32–2.21 (m, 2H), 0.92 (d,
J = 6.9 Hz, 6H), 0.83 (d, J = 6.9 Hz, 6H), 0.25 (s, 18H).
13C NMR (100.5 MHz, CDCl3/CD3OD 1:1): d 172.8,
172.3, 139.6, 133.1, 129.4, 128.3, 128.2, 122.3, 102.9, 99.9,
81.6, 72.6, 71.4, 58.7, 30.4, 26.1, 19.7, 17.6, 0.1. Anal. Calcd
4.3.6. N1,N6-Bis[(1S)-2-methyl-1-(methylcarbamoyl)pro-
pyl]-(2R,3R,4R,5R)-2,5-bis(2-((E)-2-methoxycarbonylvi-
nyl)benzyloxy)-3,4-dihydroxyhexane-1,6-diamide (4d). A
2.0 mL microwave vial was charged with 3 (43.3 mg,
0.050 mmol), Herrmann’s catalyst (trans-di-(l-acetato)-
bis[o-(di-o-tolylphosphino)benzyl]dipalladium(II)) (2.4 mg,
0.0025 mmol), methyl acrylate (0.045 mL, 0.50 mmol),
(i-Pr)2EtN (0.035 mL, 0.20 mmol), and 1.0 mL DMF/
H2O (9:1). The vial was capped with a Teflon septum
and irradiated with microwaves to 120 ꢁC for 5 min.
The reaction mixture was filtered through Celite and
the solvent removed under reduced pressure. The residue
was purified by preparative RP-LC–MS (20–70% MeCN
in H2O, 25 min gradient) to give 4d in 55% yield

J. Wannberg et al. / Bioorg. Med. Chem. 14 (2006) 5303–5315
5311
for C42H62N4O8Si2: C, 62.50; H, 7.74; N, 6.94. Found: C,
62.2; H, 7.8; N, 6.8.
172.2, 155.4, 154.8, 134.7, 130.5, 130.3, 129.6, 129.3,
129.1, 129.0, 125.1, 123.5, 121.7, 111.5, 106.2, 80.9,
72.4, 71.4, 58.7, 30.3, 26.1, 19.5, 17.4. Anal. Calcd for
C48H54N4O10: C, 68.07; H, 6.43; N, 6.61. Found: C,
67.83; H, 6.52; N, 6.46.
4.3.10. N1,N6-Bis[(1S)-2-methyl-1-(methylcarbamoyl)pro-
pyl]-(2R,3R,4R,5R)-2,5-bis(2-phenylethynylbenzyloxy)-3,4-
dihydroxyhexane-1,6-diamide (4h). Sonogashira coupling
at 100 ꢁC with phenylacetylene. The residue was purified
by preparative RP-LC–MS (30-90% MeCN in H2O,
25 min gradient) to give 4h in 93% yield (37.8 mg, color-
4.3.13. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-iodobenzyloxy)-3,4-dihydroxyhexane-
1,6-diamide (7). The bis-lactone 6 (2.00 g, 3.30 mmol)
was dissolved in 200 mL of 1,2-dichloroethane.
(1S,2R)-1-Amino-2-indanol (2.00 g, 13.4 mmol) was
added to the solution and the reaction mixture was stir-
red at 50 ꢁC for 5 h. The solvent was removed under re-
duced pressure and the residue was purified by silica gel
chromatography (5% MeOH in CHCl3) to give 7 (1.63 g,
55%) as a light yellow solid. 1H NMR (400 MHz,
CDCl3): d 7.79 (dd, J = 7.9, 1.2 Hz, 2H), 7.44–7.37 (m,
4H), 7.31 (ddd, J = 7.9, 7.6, 1.2 Hz, 2H), 7.26–7.18 (m,
6H), 6.99 (ddd, J = 7.9, 7.4, 1.7 Hz, 2H), 5.31 (dd,
J = 8.8, 5.2 Hz, 2H), 4.74 (d, J = 11.8 Hz, 2H), 4.70 (d,
J = 11.8 Hz, 2H), 4.63 (app dt, J = 5.4, 2.5 Hz, 2H),
4.34 (AA0 part of AA0BB0, 2H), 4.26 (BB0 part of
AA0BB0, 2H), 3.08 (dd, J = 16.6, 5.6 Hz, 2H), 2.88 (dd,
J = 16.6, 2.5 Hz, 2H). 13C NMR (100.5 MHz, CDCl3):
d 171.4, 141.0, 139.8, 139.7, 139.2, 130.2, 128.7, 128.5,
127.2, 125.5, 124.5, 99.0, 82.3, 72.7, 71.5, 58.0, 39.4.
Anal. Calcd for C38H38I2N2O8+H2O: C, 49.47; H,
4.37; N, 3.04. Found C, 49.3; H, 4.6; N, 3.1.
1
less solid). H NMR (400 MHz, CDCl3/CD3OD 1:1): d
7.57–7.50 (m, 8H), 7.38–7.28 (m, 10H), 4.95 (d,
J = 12.3 Hz, 2H), 4.90 (d, J = 12.3 Hz, 2H), 4.29–4.26
(m, 2H), 4.23-4.19 (m, 4H), 2.69 (s, 6H), 2.24–2.13 (m,
2H), 0.87 (d, J = 6.9 Hz, 6H), 0.81 (d, J = 6.9 Hz, 6H).
13C NMR (100.5 MHz, CDCl3/CD3OD 1:1): d 172.8,
172.3, 139.1, 132.8, 132.1, 129.2, 129.1, 128.9, 128.8,
128.6, 123.5, 122.8, 94.8, 87.2, 81.3, 72.6, 71.5, 58.8,
30.5, 26.1, 19.6, 17.5. Anal. Calcd for C48H54N4O8: C,
70.74; H, 6.68; N, 6.87. Found: C, 70.6; H, 6.8; N, 6.8.
4.3.11. N1,N6-Bis[(1S)-2-methyl-1-(methylcarbamoyl)pro-
pyl]-(2R,3R,4R,5R)-2,5-bis(2-((E)-styryl)benzyloxy)-3,4-
dihydroxyhexane-1,6-diamide (4i). A 2.0 mL microwave
vial was charged with 3 (43.3 mg, 0.050 mmol), Herr-
mann’s catalyst (trans-di-(l-acetato)-bis[o-(di-o-toly-
lphosphino)benzyl]dipalladium(II)) (2.4 mg, 0.0025 mmol),
styrene (0.055 mL, 0.50 mmol), (i-Pr)2EtN (0.035 mL,
0.20 mmol) and 1.0 mL of DMF/H2O (9:1). The vial
was capped with a Teflon septum and irradiated with
microwaves to 120 ꢁC for 5 min. The reaction mixture
was filtered through Celite and the solvent removed under
reduced pressure. The residue was purified by silica gel
flash chromatography (0–3% MeOH in CHCl3) to give
4i in 65% yield (26.8 mg, colorless solid). 1H NMR
(400 MHz, CDCl3/CD3OD 1:1): d 7.66 (dd, J = 7.8,
1.8 Hz, 2H), 7.57–7.52 (m, 4H), 7.46 (d, J = 16.2, 2H),
7.39 (dd, J = 7.5, 1.9 Hz, 2H), 7.36–7.30 (m, 6H), 7.27–
7.22 (m, 4H), 7.05 (d, J = 16.2, 2H), 4.84 (d,
J = 11.8 Hz, 2H), 4.76 (d, J = 11.8 Hz, 2H), 4.24–4.14
(m, 6H), 2.68 (s, 6H), 2.21–2.10 (m, 2H), 0.86 (d,
J = 6.8 Hz, 6H), 0.79 (d, J = 6.8 Hz, 6H). 13C NMR
(100.5 MHz, CDCl3/CD3OD 1:1): d 172.8, 172.3, 137.9,
137.5, 134.7, 131.9, 130.5, 129.4, 129.2, 128.4, 128.1,
127.3, 126.4, 125.6, 80.5, 72.3, 71.0, 58.8, 30.4, 26.1,
19.6, 17.5. Anal. Calcd for C48H58N4O8+0.5H2O: C,
69.63; H, 7.18; N, 6.77. Found: C, 69.6; H, 7.1; N, 6.7.
4.3.14. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-phenylcarbamoylbenzyloxy)-3,4-dihydr-
oxyhexane-1,6-diamide (8). A 2.0 mL microwave vial
was charged with 7 (45.2 mg, 0.050 mmol), Pd(OAc)2
(2.2 mg, 0.01 mmol), Mo(CO)6 (26.4 mg, 0.10 mmol),
aniline (91 lL, 1.0 mmol), DBU (0.75 mL, 0.50 mmol)
and dry THF (1.5 mL). The vial was immediately
capped with a Teflon septum and irradiated with micro-
waves to 110 ꢁC for 15 min. After cooling, the reaction
mixture was filtered and the solvent was removed under
reduced pressure. The residue was purified by prepara-
tive RP-LC–MS (20-70% MeCN in H2O, 25 min gradi-
ent) to produce 8 in 45% (20.1 mg) yield. 1H NMR
(400 MHz, CDCl3): d 8.22 (s, 2H), 7.58-7.06 (m, 28H),
5.25 (dd, J = 8.6, 5.2 Hz, 2H), 4.90 (d, J = 11.6 Hz,
2H), 4.76 (d, J = 11.6 Hz, 2H), 4.53 (app dt, J = 5.4,
2.2 Hz, 2H), 4.09 (AA0 part of AA0BB0, 2H), 4.05 (BB0
part of AA0BB0, 2H), 3.08 (dd, J = 16.6, 5.6 Hz, 2H),
2.89 (dd, J = 16.6, 2.2 Hz, 2H). 13C NMR (100.5 MHz,
CDCl3/CD3OD 1:1): d 172.9, 169.6, 141.0, 140.9,
138.9, 137.2, 135.6, 131.1, 130.7, 129.4, 128.9, 128.6,
128.4, 127.5, 125.7, 125.2, 124.9, 121.4, 81.1, 73.4,
71.5, 71.2, 58.1, 40.1. HRMS (M+1): Calcd: 891.3605.
Found: 891.3607.
4.3.12. N1,N6-Bis[(1S)-2-methyl-1-(methylcarbamoyl)pro-
pyl]-(2R,3R,4R,5R)-2,5-bis(2-benzofuran-2-ylbenzyloxy)-
3,4-dihydroxyhexane-1,6-diamide (4j). Suzuki coupling
with 2-benzofuranylboronic acid. Product purified by
preparative RP-LC–MS (30–90% MeCN in H2O,
25 min gradient) to give 8.0 mg of 4j (19%, colorless sol-
1
id). H NMR (400 MHz, CDCl3/CD3OD 1:1): d 7.87
(dd, J = 7.6, 1.8 Hz, 2H), 7.76 (q, J = 4.7 Hz, 2H),
7.64–7.59 (m, 4H), 7.52 (d, J = 8.1 Hz, 2H), 7.47–7.37
(m, 4H), 7.29 (dt, J = 7.3, 1.6 Hz, 2H), 7.23 (dt,
J = 7.3, 1.3 Hz, 2H), 7.10 (d, J = 1.1 Hz, 2H), 4.97 (d,
J = 12.1 Hz, 2H), 4.90 (d, J = 12.1 Hz, 2H), 4.31 (AA0
of AA0XX0, 2H), 4.22 (d, J = 5.4 Hz, 2H), 4.17 (XX0
of AA0XX0, 2H), 2.69 (s, 6H), 2.22–2.11 (m, 2H), 0.79
(d, J = 6.8 Hz, 6H), 0.71 (d, J = 6.8 Hz, 6H). 13C
4.3.15. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-vinylbenzyloxy)-3,4-dihydroxyhexane-
1,6-diamide (9a). The 2,4,6-trivinylcyclotriboroxane–
pyridine complex (24.1 mg, 0.10 mmol) was used as a
boronic acid equivalent. The residue was purified by sil-
ica gel flash chromatography (3–5% MeOH in CHCl3)
to give 9a as a colorless solid in 51% yield (17.8 mg).
1H NMR (400 MHz, CDCl3): d 7.50 (dd, J = 7.7,
1.6 Hz, 2H), 7.34–7.18 (m, 14H), 7.11 (dd, J = 7.3,
NMR (100.5 MHz, CDCl3/CD3OD 1:1):
d 172.7,

5312
J. Wannberg et al. / Bioorg. Med. Chem. 14 (2006) 5303–5315
1.2 Hz, 2H), 6.99 (dd, J = 17.4, 11.0 Hz, 2H), 5.59 (dd,
J = 17.4, 1.5 Hz, 2H), 5.28 (dd, J = 8.7, 5.1 Hz, 2H),
5.23 (dd, J = 11.0, 1.5 Hz, 2H), 4.81 (d, J = 11.4 Hz,
2H), 4.73 (d, J = 11.4 Hz, 2H), 4.65 (app dt, J = 5.4,
2.3 Hz, 2H), 4.23 (AA0 part of AA0BB0, 2H), 4.19 (BB0
part of AA0BB0, 2H), 3.10 (dd, J = 16.6, 5.6 Hz, 2H),
2.90 (dd, J = 16.6, 2.4 Hz, 2H). 13C NMR (100.5 MHz,
CDCl3): d 171.6, 141.1, 139.7, 137.4, 133.8, 133.6,
130.1, 129.1, 128.5, 128.1, 127.1, 126.3, 125.5, 124.2,
117.0, 82.1, 72.5, 72.1, 71.3, 58.1, 39.5. Anal. Calcd for
C42H44N2O8+H2O: C, 69.79; H, 6.41; N, 3.88. Found:
C, 69.93; H, 6.34; N, 3.81.
201.7, 171.9, 141.1, 139.9, 137.5, 136.6, 132.6, 130.1,
129.8, 128.4, 128.1, 127.1, 125.5, 124.2, 83.3, 72.6,
72.4, 71.2, 58.3, 39.6, 29.2. Anal. Calcd for
C42H44N2O10: C, 68.46; H, 6.02; N, 3.80. Found C,
68.21; H, 6.16; N, 3.56.
4.3.18. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-((E)-2-methoxycarbonylvinyl)benzyl-
oxy)-3,4-dihydroxyhexane-1,6-diamide (9d). A 2.0 mL
microwave vial was charged with
0.050 mmol), Herrmann’s catalyst
5
(45.2 mg,
(2.4 mg,
0.0025 mmol), methyl acrylate (0.045 mL, 0.50 mmol),
(i-Pr)2EtN (0.035 mL, 0.20 mmol) and 1.0 mL of
DMF/H2O (9:1). The vial was capped with a Teflon sep-
tum and irradiated with microwaves to 120 ꢁC for 5 min.
The reaction mixture was filtered through Celite and the
solvent removed under reduced pressure. The residue
was purified by silica gel flash chromatography (3–5%
MeOH in CHCl3) to give 9d in 57% yield (23.8 mg, col-
4.3.16. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-ethynylbenzyloxy)-3,4-dihydroxyhex-
ane-1,6-diamide (9b). Sonogashira coupling at 120 ꢁC
with trimethylsilylacetylene. To the crude product was
added 2 mL of a saturated solution of K2CO3 in
MeOH and the mixture was stirred at room temperature
for 2 h. The solvent was evaporated and the residue
was purified by silica gel flash chromatography (5%
MeOH in CHCl3) to give 18.6 mg of 9b (53%, light yel-
1
orless solid). H NMR (400 MHz, CDCl3): d 8.01 (d,
J = 15.9 Hz, 2H), 7.60–7.55 (m, 2H), 7.39–7.30 (m,
8H), 7.23 (dd, J = 5.0, 1.2 Hz, 4H), 7.20–7.08 (m, 2H),
6.27 (d, J = 15.9 Hz, 2H), 5.33 (ddd, J = 8.7, 5.1,
1.2 Hz, 2H), 4.97 (d, J = 11.4 Hz, 2H), 4.73 (d,
J = 11.4 Hz, 2H), 4.70–4.64 (m, 2H), 4.28 (m, 4H),
3.55 (s, 6H), 3.12 (dd, J = 16.5, 5.5 Hz, 2H), 2.94 (dd,
J = 16.5, 2.1 Hz, 2H). 13C NMR (100.5 MHz, CDCl3):
d 171.7, 167.7, 141.9, 141.0, 140.0, 135.9, 134.0, 130.6,
130.4, 129.2, 128.4, 127.04, 127.01, 125.4, 124.2, 119.7,
81.6, 72.7, 71.7, 71.1, 58.0, 51.9, 39.6. Anal. Calcd for
C46H48N2O12: C, 67.30; H, 5.89; N, 3.41. Found: C,
67.2; H, 6.2; N, 3.4.
1
low solid). H NMR (400 MHz, CDCl3): d 7.49 (dd,
J = 7.4, 1.8 Hz, 2H), 7.41 (dd, J = 7.5, 1.8 Hz, 2H),
7.36–7.14 (m, 14H), 5.33 (ddd, J = 8.7, 5.1, 1.2 Hz,
2H), 4.90 (d, J = 11.5 Hz, 2H), 4.84 (d, J = 11.5 Hz,
2H), 4.70 (app dt, J = 5.4, 2.2 Hz, 2H), 4.31 (AA0 part
of AA0BB0, 2H), 4.26 (BB0 part of AA0BB0, 2H), 3.13
(dd, J = 16.5, 5.6 Hz, 2H), 3.08(s, 2H), 2.95 (dd,
J = 16.5, 2.2 Hz, 2H). 13C NMR (100.5 MHz, CDCl3):
d 171.7, 141.1, 139.8, 138.8, 133.3, 129.4, 129.2, 128.6,
128.5, 127.1, 125.5, 124.4, 121.8, 82.8, 82.4, 81.6, 72.6,
72.5, 71.4, 58.1, 39.5. Anal. Calcd for C42H40N2O8:
C, 71.98; H, 5.75; N, 4.00. Found: C, 71.8; H, 5.9; N,
3.9.
4.3.19. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-(3-thiophenyl)benzyloxy)-3,4-dihydroxy-
hexane-1,6-diamide (9e). Suzuki coupling with 3-thi-
enylboronic acid. Product purified by preparative RP-
LC–MS to give 21.7 mg of 9e (53%, colorless solid).
1H NMR (400 MHz, CDCl3): d 7.46 (d, J = 7.1, 2H),
7.35–7.16 (m, 18H), 7.13 (dd, J = 4.9, 1.6 Hz, 2H),
7.03 (d, J = 7.4 Hz, 2H), 5.27 (dd, J = 8.8, 5.0 Hz, 2H),
4.72–4.65 (m, 4H), 4.64–4.58 (m, 2H), 4.12 (m, 4H),
3.06 (dd, J = 16.6, 5.5 Hz, 2H), 2.86 (dd, J = 16.6,
2.3 Hz, 2H). 13C NMR (100.5 MHz, CDCl3): d 171.8,
141.1, 140.7, 139.7, 137.0, 134.2, 130.0, 128.9, 128.7,
128.5, 127.9, 127.2, 125.9, 125.5, 124.1, 123.4, 82.8,
72.5, 71.9, 71.6, 58.2, 39.4. Anal. Calcd for
C46H44N2O8S2+H2O: C, 66.17; H 5.55; N, 3,35. Found:
C, 65,96; H, 5,61; N, 3,31.
4.3.17. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-acetylbenzyloxy)-3,4-dihydroxyhexane-
1,6-diamide (9c). A 2.0 mL microwave vial was charged
with
0.010 mmol),
7
(45.2 mg, 0.050 mmol), Pd(OAc)2 (2.2 mg,
1,3-bis(diphenylphosphino)propane
(dppp) (4.4 mg, 0.010 mmol), 2-hydroxyethyl vinyl ether
(0.045 mL, 0.50 mmol), TlOAc (28.1 mg, 0.10 mmol), (i-
Pr)2EtN (0.035 mL, 0.20 mmol) and 1.0 mL of DMF/
H2O (9:1). The vial was capped with a Teflon septum
and irradiated with microwaves to 100 ꢁC for 30 min.
The dioxolane was deprotected by addition of 2 mL of
0.5 M HCl in MeCN/H2O (1:1). After stirring for
30 min at room temperature, the reaction mixture was
poured onto 40 mL of 2 M HCl (aq) and extracted with
2· 40 mL CH2Cl2. The combined organic layers were
washed with brine (40 mL), dried with MgSO4 and evap-
orated to dryness. The residue was purified by silica gel
flash chromatography (3–5% MeOH in CHCl3) to give
33.2 mg of 9c (90%) as a colorless solid. 1H NMR
(400 MHz, CDCl3): d 7.76 (dd, J = 7.7, 1.5 Hz, 2H),
7.62 (dd, J = 7.7, 1.5 Hz, 2H), 7.53–7.46 (m, 4H), 7.40
(dt, J = 7.6, 1.4 Hz, 2H), 7.26–7.13 (m, 8H), 5.33 (dd,
J = 8.6, 5.1 Hz, 2H), 5.03 (d, J = 12.6 Hz, 2H), 4.97 (d,
J = 12.6 Hz, 2H), 4.75–4.69 (m, 2H), 4.33 (AA0 part of
AA0BB0, 2H), 4.27 (BB0 part of AA0XX0, 2H), 3.14
(dd, J = 16.6, 5.4 Hz, 2H), 2.98 (dd, J = 16.6, 2.2 Hz,
2H), 2.55 (s, 6H). 13C NMR (100.5 MHz, CDCl3): d
4.3.20. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-phenylbenzyloxy)-3,4-dihydroxyhexane-
1,6-diamide (9f). Suzuki coupling with phenylboronic
acid. Product purified by preparative RP-LC–MS to
give 11.8 mg of 9f (29%, colorless solid). 1H NMR
(400 MHz, CDCl3): d 7.52–7.47 (m, 2H), 7.39–7.23 (m,
20H), 7.21–7.14 (m, 4H), 6.97 (d, J = 7.4 Hz, 2H), 5.25
(dd, J = 8.7, 5.0 Hz, 2H), 4.69–4.58 (m, 6H), 4.05 (AA0
part of AA0XX0, 2H), 4.01 (AA0 part of AA0XX0, 2H),
3.08 (dd, J = 16.6, 5.4 Hz, 2H), 2.91 (dd, J = 16.6,
2.0 Hz, 2H). 13C NMR (100.5 MHz, CDCl3): d 171.8,
142.1, 141.1, 140.6, 139.7, 134.0, 130.4, 129.5, 129.2,
128.5, 127.9, 127.6, 127.2, 125.5, 124.0, 83.1, 72.4,

J. Wannberg et al. / Bioorg. Med. Chem. 14 (2006) 5303–5315
5313
71.7, 71.5, 58.2, 39.5. Anal. Calcd for C50H48N2O8+-
H2O: C, 72.97; H, 6.12; N, 3.40. Found: C, 73.0; H,
6.1; N, 3.3.
71.8, 70.8, 58.2, 39.6. HRMS (M+1): Calcd: 871.2835.
Found: 871.2833.
4.3.23. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-trimethylsilanylethynylbenzyloxy)-3,4-
dihydroxyhexane-1,6-diamide (9i). Sonogashira coupling
at 120 ꢁC with trimethylsilylacetylene. Product purified
by silica gel flash chromatography (1–2% MeOH in
CHCl3) to give 9i in 44% yield (18.7 mg, yellow solid).
1H NMR (400 MHz, CDCl3): d 7.49 (dd, J = 7.3,
1.9 Hz, 2H), 7.42 (dd, J = 7.3, 2.0 Hz, 2H), 7.34–7.11
(m, 14H), 5.26 (ddd, J = 8.3, 5.0, 1.4 Hz, 2H), 4.95 (d,
J = 12.1 Hz, 2H), 4.87 (d, J = 12.1 Hz, 2H), 4.74 (app
dt, J = 5.2, 1.7 Hz, 2H), 4.32-4.28 (m, 4H), 3.14 (dd,
J = 16.6, 5.4 Hz, 2H), 2.98 (dd, J = 16.5, 1.9 Hz, 2H),
0.24 (s, 18H). 13C NMR (100.5 MHz, CDCl3): d 171.6,
141.2, 139.6, 138.8, 133.0, 129.1, 128.54, 128.47, 128.3,
127.1, 125.5, 124.0, 122.5, 102.6, 100.0, 83.8, 72.7,
72.4, 71.4, 58.6, 39.7, 0.1. Anal. Calcd for C48H56N2O8-
Si2+0.5H2O: C, 67.50; H, 6.73; N, 3.28. Found: C, 67.2;
H, 6.7; N, 3.2.
4.3.21. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-benzylbenzyloxy)-3,4-dihydroxyhexane-
1,6-diamide (9g). To an oven-dried 2.0 mL microwave
vial was added 7 (45.2 mg, 0.050 mmol), Pd(OAc)2,
(1.1 mg, 0.0050 mmol) and [(t-Bu)3PH]BF4 (3.0 mg,
0.010 mmol). The vial was capped with a septum and
evacuated and backfilled with N2 three times. Benzylzinc
bromide (0.5 M in THF, 1.0 mL, 0.50 mmol) was added
by syringe and the resulting mixture was irradiated with
microwaves to 100 ꢁC for 15 min. The reaction mixture
was poured onto 50 mL of cold 2 M HCl (aq) and
extracted with 2· 50 mL CH2Cl2. The combined organic
layers were washed with brine and dried with MgSO4
and the solvent removed under reduced pressure. The
residue (containing starting material, mono- and bis-
coupled product) was purified by RP-LC–MS to give
10.4 mg of 9g (25%) as a colorless solid. 1H NMR
(400 MHz, CDCl3): d 7.33 (dd, J = 7.4, 1.7 Hz, 2H),
7.30–7.13 (m, 20H), 7.09–7.04 (m, 6H), 5.29 (ddd,
J = 8.7, 5.0, 1.4 Hz, 2H), 4.69–4.63 (m, 6H), 4.09–4.04
(m, 8H), 3.11 (dd, J = 16.6, 5.4 Hz, 2H), 2.93 (dd,
J = 16.6, 2.0 Hz, 2H). 13C NMR (100.5 MHz, CDCl3):
d 171.9, 141.1, 140.3, 139.7, 139.4, 134.9, 131.0, 129.8,
129.0, 128.8, 128.7, 128.6, 127.2, 127.0, 126.4, 125.6,
124.0, 82.6, 72.5, 71.8, 71.7, 58.3, 39.5, 38.6. Anal. Calcd
for C52H52N2O8+2H2O: C, 71.87; H, 6.50; N, 3.22.
Found: C, 71.78; H, 6.29; N, 3.06.
4.3.24. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-phenylethynylbenzyloxy)-3,4-dihydroxy-
hexane-1,6-diamide (9j). Sonogashira coupling at 100 ꢁC
with phenylacetylene. Product purified by silica gel flash
chromatography (0–2% MeOH in CHCl3) to give 9j in
72% yield (30.6 mg, colorless solid). 1H NMR
(400 MHz, CDCl3): d 7.55–7.52 (m, 2H), 7.50–7.47 (m,
4H), 7.44–7.41 (m, 2H), 7.38–7.28 (m, 12H), 7.22–7.19
(m, 4H), 7.09–7.07 (m, 4H), 5.24 (dd, J = 8.5, 4.9 Hz,
2H), 4.98 (d, J = 12.0 Hz, 2H), 4.91 (d, J = 12.0 Hz,
2H), 4.66 (app dt, J = 5.2, 2.1 Hz, 2H), 4.33 (AA0 part
of AA0BB0, 2H), 4.28 (BB0 part of AA0BB0, 2H), 3.08
(dd, J = 16.6, 5.5 Hz, 2H), 2.89 (dd, J = 16.6, 2.1 Hz,
2H). 13C NMR (100.5 MHz, CDCl3): d 171.8, 141.0,
139.6, 138.2, 132.6, 131.7, 128.9, 128.84, 128.76, 128.6,
128.43, 128.42, 127.1, 125.4, 124.1, 122.8, 122.7, 94.5,
87.0, 82.7, 72.5, 72.4, 71.5, 58.3, 39.5. Anal. Calcd for
C54H48N2O8: C, 76.04; H, 5.67; N, 3.28. Found: C,
75.8; H, 5.8; N, 3.3.
4.3.22. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-((E)-2-thiazol-2-yl-vinyl)-benzyloxy)-
3,4-dihydroxyhexane-1,6-diamide (9h). A mixture of
2,4,6-trivinylcyclotriboroxane–pyridine complex (0.120 g,
0.5 mmol), 2-bromothiazole (0.045 mL, 0.5 mmol),
Pd(OAc)2 (5.6 mg, 0.025mmol), [(t-Bu)3PH]BF4 (14.5 mg,
0.050 mmol), K2CO3 (138 mg, 1.0 mmol), 0.3 mL H2O
and 1 mL DME was heated by microwaves to 100 C
for 15 min in a 2 mL process vial. The reaction mixture
was loaded onto a 2 cm silica pad, which was subse-
quently washed with CHCl3. The solvent was evaporat-
ed and the crude 2-vinylthiazole was dissolved in 2 mL
of MeCN/H2O (10:1) and the mixture was transferred
to a process vial loaded with 7 (45.3 mg, 0.050 mmol),
(i-Pr)2EtN (0.052 mL, 0.30 mmol) and Herrmann’s cata-
lyst (4.7 mg, 0.0050 mmol) and heated for 10 min at
120 C. The reaction mixture was filtered through Celite
and the solvent removed under reduced pressure. The
residue was purified by silica gel flash chromatography
(2–5% MeOH in CHCl3) to give 9h in 62% yield
4.3.25. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-((E)-styryl)benzyloxy)-3,4-dihydroxy-
hexane-1,6-diamide (9k). A 2.0 mL microwave vial was
charged with 7 (45.2 mg, 0.050 mmol), Herrmann’s
catalyst (2.4 mg, 0.0025 mmol), styrene (0.055 mL,
0.50 mmol), (i-Pr)2EtN (0.035 mL, 0.20 mmol) and
1.0 mL of DMF/H2O (9:1). The vial was capped with
a Teflon septum and irradiated with microwaves to
120 ꢁC for 5 min. The reaction mixture was filtered
through Celite and the solvent removed under reduced
pressure. The residue was purified by silica gel flash
chromatography (2–5% MeOH in CHCl3) to give 9k
in 63% yield (27.0 mg, colorless solid). 1H NMR
(400 MHz, CDCl3): d 7.61 (dd, J = 7.7, 1.6 Hz, 2H),
7.47–7.43 (m, 2H), 7.40 (d, J = 16.1 Hz, 2H), 7.36–7.18
(m, 20H), 7.10–6.98 (m, 4H), 6.89 (d, J = 16.1 Hz,
2H), 5.22 (dd, J = 8.6, 5.0 Hz, 2H), 4.88 (d,
J = 11.7 Hz, 2H), 4.72 (d, J = 11.7 Hz, 2H), 4.53 (app
dt, J = 5.3, 2.0 Hz, 2H), 4.25 (AA0 part of AA0BB0,
2H), 4.22 (BB0 part of AA0BB0, 2H), 3.02 (dd,
J = 16.6, 5.5 Hz, 2H), 2.79 (dd, J = 16.6, 2.1 Hz, 2H).
1
(26.8 mg, colorless solid). H NMR (400 MHz, CDCl3):
d 7.84 (d, J = 16.0 Hz, 2H), 7.68–7.64 (m, 2H), 7.62 (d,
J = 7.7 Hz, 2H), 7.39–7.12 (m, 14H), 7.09 (d,
J = 16.0 Hz, 2H), 5.32 (dd, J = 8.6, 4.9 Hz, 2H), 4.94
(d, J = 11.0 Hz, 2H), 4.71 (d, J = 11.0 Hz, 2H), 4.67–
4.61 (m, 2H), 4.38 (AA0 part of AA0BB0, 2H), 4.32
(BB0 part of AA0BB0, 2H), 3.12 (dd, J = 16.7, 5.4 Hz,
2H), 2.92 (dd, J = 16.7, 2.1 Hz, 2H). 13C NMR
(100.5 MHz, CDCl3): d 171.8, 166.9, 143.5, 140.9,
140.1, 135.5, 135.0, 131.1, 130.9, 129.2, 129.0, 128.3,
127.1, 126.2, 125.4, 124.3, 121.9, 118.9, 81.4, 72.7,

5314
J. Wannberg et al. / Bioorg. Med. Chem. 14 (2006) 5303–5315
13C NMR (100.5 MHz, CDCl3): d 171.6, 140.8, 139.8,
137.2, 137.0, 133.8, 131.6, 130.3, 129.1, 128.9, 128.6,
128.4, 128.1, 127.8, 127.1, 126.9, 126.1, 125.4, 125.0,
124.1, 81.2, 72.6, 71.8, 71.3, 58.0, 39.5. Anal. Calcd for
while the rest of the protein was restrained by a force
2
constant of 100 kJ/mol/A . A 1000-step modified Po-
˚
lak–Ribiere conjugate gradient minimization with mod-
ified AMBER force field was carried out. A solvation
model was not used; however, important structural
waters were included. Once the minimization was com-
plete, the active site was defined as the collection of ami-
2
C54H52N2O8 þ H2O: C, 74.63; H, 6.19; N, 3.22. Found
3
C 74.6, H 6.2, N 3.3.
˚
4.3.26. N1,N6-Bis[(1S,2R)-2-hydroxy-1-indanyl]-(2R,3R,
4R,5R)-2,5-bis(2-benzofuran-2-ylbenzyloxy)-3,4-dihydr-
oxyhexane-1,6-diamide (9l). Suzuki coupling with 2-ben-
zofuranylboronic acid. Product purified by preparative
RP-LC–MS to give 19.9 mg of 9l (45%, colorless
solid). 1H NMR (400 MHz, CDCl3): d 7.80 (dd,
J = 7.7, 1.7 Hz, 2H), 7.56 (dd, J = 7.6, 1.8 Hz, 2H),
7.52–7.48 (m, 4H), 7.40 (dt, J = 7.6, 1.8 Hz, 2H), 7.33
(dt, J = 7.5, 1.8 Hz, 2H), 7.29–7.13 (m, 10H), 7.03–6.97
(m, 4H), 6.90 (d, J = 7.5 Hz, 2H), 5.21 (dd, J = 8.6,
5.1 Hz, 2H), 5.00 (d, J = 12.0 Hz, 2H), 4.91 (d,
J = 12.0 Hz, 2H), 4.61–4.56 (m, 2H), 4.29–4.25 (m,
4H), 3.02 (dd, J = 16.5, 5.5 Hz, 2H), 2.79 (dd, J = 16.5,
2.3 Hz, 2H). 13C NMR (100.5 MHz, CDCl3): d 171.7,
154.9, 154.6, 140.9, 139.5, 133.8, 130.1, 129.13, 129.12,
129.0, 128.9, 128.4, 127.1, 125.4, 124.8, 124.0, 123.2,
121.4, 111.4, 105.8, 82.2, 72.5, 72.2, 71.6, 58.2, 39.4.
no acids enclosed within a 12.0 A radius sphere centered
on the bound ligand, which was placed in the cavity
after the protein minimization. The protein now com-
prised of 128 amino acids and five structurally impor-
tant water molecules.
4.4.2. Protein conformational flexibility and docking
methodology. NMR studies have previously confirmed
that the protease flaps are indeed flexible, and suggest
that this flexibility may be important for inhibitor bind-
ing.29 As a consequence, we choose to dock the ligands
by taking into account this protein flexibility. Flap resi-
dues, I47/470, G48/480, G49/490, I50/500, and G51/510,
along with the residues constituting the flexible 80 s
loop, T80/800, P81/810, V82/820, N83/830, I84/840, and
other critical residues, R8/80, G27/270, A28/280, and
D29/290, were tethered. They were allowed to move free-
2
1
˚
˚
Anal. Calcd for C54H48N2O10 þ H2O: C, 72.55; H,
ly up to 0.5 A and then a quadratic 20 kJ/mol/A penalty
was imposed.
2
5.52; N, 3.13. C, 72.50; H, 5.69; N, 3.25.
4.4. Computational studies
The ligands were docked into the active site by Full-
dock+, which is a thorough near systematic ligand dock-
ing module with contact potency scoring. Here, we
carried out repeated independent docking of each ligand
to get energy and geometry convergence of best docking
modes followed by merging of these results and a local
Monte Carlo docking search on the pooled ensemble.
A total of 10 runs of systematic docking were performed
and the top 40 solutions for each ligand were saved.
Each of these 40 solutions was then subjected to local
Monte Carlo perturbation cycles and was minimized
by 300 steps of PRCG method. The conformations were
randomly generated in such a way that each new confor-
mation generated differed from the previous one by a
4.4.1. Preparation of ligands, protein coordinates, and
definition of active sites. The structures were prepared in
MOL2 format using the sketcher module of SYBYL6.9
and were then superimposed on the inhibitor N,N-[2,5-
O-di-2-fluoro-benzyl-glucaryl]-di-[1-amino-indan-2-ol]
(BED) co-crystallized with HIV1PR24 (PDB entry
1EC0). The Tripos force field was used for the energy
calculations, and a dielectric constant of 4.0 was em-
ployed. Partial charges were calculated by the method
of Gasteiger and Huckel. The minimization was run
until they converged to a maximum derivative of
˚
0.05 kcal/mol/A, and the final coordinates were stored
in database. Care was taken to remove the non-polar
hydrogen atoms before docking the inhibitors.
˚
similarity distance of 1.0 A and had an energy window
˚
of not more than 10.0 kJ/mol/A. The maximum move-
ment of an atom in any single time step was limited to
˚
0.1 A and hydrogen vibrations were damped by assign-
ing an atomic weight of 10.
Reference protein coordinates used for docking were
taken from the X-ray structure of HIV1PR in complex
˚
with BED (PDB entry: 1EC0) resolved at 1.79 A. All
the crystallization water molecules along with the native
ligand were removed from the HIV1PR, except the
structural waters referred to in this article as WAT00,
WAT01, WAT02, WAT03, and WAT04 (Fig. 3). Only
polar hydrogen atoms were added in a smart way using
standard QXP geometries. Protonation states were
assumed to be those most common at pH 7, that is,
lysines, arginines, aspartates, and glutamates were con-
sidered in the ionized form except for D25/250 which
were kept protonated and neutral. Partial charges were
calculated using bond dipole moments. Despite the
hydrogen optimization, there still exist a few structural
discrepancies in assigning the side chains of an X-ray
4.4.3. Data set and 3D QSAR studies. The data set com-
prised of 24 docked conformations of the inhibitors,
which were used to generate the CoMFA alignment.
The CoMFA fields were generated using default settings
as implemented in Sybyl 7.0. A 3D cubic lattice was cre-
ated with a grid spacing of 2 A and extending 4 A units
beyond the docked inhibitors in all directions. An sp3
carbon with +1 charge was used as probe atom to pro-
duce steric and electrostatic field energies. Cut-offs for
both steric and electrostatic fields were set to 30 kcal/
mol, employing a distance dependent dielectric (1/r2)
for calculating electrostatics and a 6–12 Lennard–Jones
potential for sterics. The CoMFA descriptor fields were
used as independent variables in the PLS (partial least
squares)30 regression analysis with measured pKi
(ꢀlogKi) values as the dependent variable. A LOO
˚
˚
˚
structure resolved at around 2.0 A, so it was decided
to refine the crystal structure by restrained minimiza-
tion. Here, the hydrogens were allowed to move freely

J. Wannberg et al. / Bioorg. Med. Chem. 14 (2006) 5303–5315
5315
12. Larhed, M.; Moberg, C.; Hallberg, A. Acc. Chem. Res.
2002, 35, 717–727.
(leave-one-out) validation, and a maximum of six com-
ponents were used for all cross-validation calculations
to determine the q2 (cross-validated correlation coeffi-
cient), the PRESS value (standard error of prediction)
and the r2 (non-validated correlation coefficient).
13. Dorsey, B. D.; Levin, R. B.; McDaniel, S. L.; Vacca, J. P.;
Guare, J. P.; Darke, P. L.; Zugay, J. A.; Emini, E. A.;
Schleif, W. A.; Quintero, J. C.; Lin, J. H.; Chen, I.-W.;
Holloway, M. K.; Fitzgerald, P. M. D.; Axel, M. G.;
Ostovic, D.; Anderson, P. S.; Huff, J. R. J. Med. Chem.
1994, 37, 3443–3451.
14. Negishi, E.-i., Ed. Handbook of Organopalladium Chem-
istry for Organic Synthesis; Wiley-Interscience: New York,
2002; Vol. 1.
15. Kerins, F.; O’Shea, D. F. J. Org. Chem. 2002, 67, 4968–
4971.
16. Vallin, K. S. A.; Larhed, M.; Johansson, K.; Hallberg, A.
J. Org. Chem. 2000, 65, 4537–4542.
17. Beller, M.; Cornils, B.; Frohning, C. D.; Kohlpaintner, C.
W. J. Mol. Catal. A: Chem. 1995, 104, 17–85.
18. Skoda-Foldes, R.; Kollar, L. Curr. Org. Chem. 2002, 6,
1097–1119.
Acknowledgments
We express our sincere gratitude to the Swedish Re-
search Council, the Swedish Foundation for Strategic
Research, Knut and Alice Wallenberg’s Foundation
and Medivir AB for financing this research. We thank
Mr. Gunnar Wikman and Ms. Riina Arvela for linguis-
tic advice and Biotage AB for providing the microwave
synthesizer.
19. Wannberg, J.; Larhed, M. J. Org. Chem. 2003, 68, 5750–
5753.
20. Morimoto, T.; Kakiuchi, K. Angew. Chem., Int. Ed. 2004,
43, 5580–5588.
21. Sybyl; 6.9 ed. ; Tripos Inc., S. L., Missouri.
22. McMartin, C.; Bohacek, R. S. J. Comput. Aided Mol. Des.
1997, 11, 333–344.
23. Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.;
Bhat, T. N.; Weissig, H.; Shindyalov, I. N.; Bourne, P. E.
Nucleic Acids Res. 2000, 28, 235–242.
24. Lindberg, J.; Pyring, D.; Lo¨wgren, S.; Rosenquist, A.;
Zuccarello, G.; Kvarnstro¨m, I.; Zhang, H.; Vrang, L.;
Classon, B.; Hallberg, A.; Samuelsson, B.; Unge, T. Eur.
J. Biochem. 2004, 271, 4594–4602.
25. Dreyer, G. B.; Boehm, J. C.; Chenera, B.; DesJarlais, R.
L.; Hassell, A. M.; Meek, T. D.; Tomaszek, T. A., Jr.;
Lewis, M. Biochemistry 1993, 32, 937–947.
26. Martin, J. L.; Begun, J.; Schindeler, A.; Wickramasin-
ghe, W. A.; Alewood, D.; Alewood, P. F.; Bergman,
D. A.; Brinkworth, R. I.; Abbenante, G.; March, D.
R.; Reid, R. C.; Fairlie, D. P. Biochemistry 1999, 38,
7978–7988.
References and notes
1. Brik, A.; Wong, C. H. Org. Biomol. Chem. 2003, 1, 5–14.
2. Huff, J. R.; Kahn, J. Adv. Protein Chem. 2001, 56, 213–251.
3. Swanstrom,R.;Erona,J.Pharmacol.Ther.2000,86,145–170.
4. Clavel, F.; Hance, A. J. N. Engl. J. Med. 2004, 350, 1023–1035.
5. De Clercq, E. J. Med. Chem. 2005, 48, 1297–1313.
6. Alterman, M.; Bjo¨rsne, M.; Muhlman, A.; Classon, B.;
¨
Kvarnstro¨m, I.; Danielson, H.; Markgren, P. O.; Nillroth,
U.; Unge, T.; Hallberg, A.; Samuelsson, B. J. Med. Chem.
1998, 41, 3782–3792.
7. Pyring, D.; Lindberg, J.; Rosenquist, A.; Zuccarello, G.;
Kvarnstrom, I.; Zhang, H.; Vrang, L.; Unge, T.; Classon,
B.; Hallberg, A.; Samuelsson, B. J. Med. Chem. 2001, 44,
3083–3091.
8. Andersson, H. O.; Fridborg, K.; Lo¨wgren, S.; Alterman,
M.; Muhlman, A.; Bjo¨rsne, M.; Garg, N.; Kvarnstro¨m, I.;
¨
Schaal, W.; Classon, B.; Karlen, A.; Danielsson, U. H.;
¨
Ahlsen, G.; Nillroth, U.; Vrang, L.; Oberg, B.; Samuels-
son, B.; Hallberg, A.; Unge, T. Eur. J. Biochem. 2003, 270,
1746–1758.
27. Danielson, H.; Lindgren, M. T.; Markgren, P. O.;
Nillroth, U. Adv. Exp. Med. Biol. 1998, 436, 99–103.
28. Nillroth, U.; Vrang, L.; Markgren, P. O.; Hulten, J.;
Hallberg, A.; Danielson, U. H. Antimicrob. Agents Che-
mother. 1997, 41, 2383–2388.
29. Torchia, D. A.; Ishima, R. Pure Appl. Chem. 2003, 75,
1371–1381.
30. Wold, S. Quant. Struct.-Act. Relat. 1991, 10, 191–193.
9. Alterman, M.; Andersson, H. O.; Garg, N.; Ahlsen, G.;
Lo¨vgren, S.; Classon, B.; Danielson, U. H.; Kvarnstro¨m,
I.; Vrang, L.; Unge, T.; Samuelsson, B.; Hallberg, A.
J. Med. Chem. 1999, 42, 3835–3844.
10. Ersmark, K.; Larhed, M.; Wannberg, J. Curr. Opin. Drug
Discovery Dev. 2004, 7, 417–427.
11. Wannberg, J.; Kaiser, N. F. K.; Vrang, L.; Samuelsson, B.;
Larhed, M.; Hallberg, A. J. Comb. Chem. 2005, 7, 611–617.