Studies on selectin blockers. 6. Discovery of homologous fucose sugar unit necessary for E-selectin binding

Article abstract of DOI:10.1021/jm9707481

We describe a mimic of the sugar unit of the E-selectin ligand, sialyl Lewis X (sLe(X)). Carbohydrates are entering the realm of rational drug design, aided by the growing understanding of the structure-function relationships. We investigated a new method

Full text of DOI:10.1021/jm9707481

2302
J . Med. Chem. 1998, 41, 2302-2307
Stu d ies on Selectin Block er s. 6. Discover y of Hom ologou s F u cose Su ga r Un it
Necessa r y for E-Selectin Bin d in g
Yasuyuki Hiramatsu, Hideki Moriyama, Takao Kiyoi, Takahiro Tsukida, Yoshimasa Inoue, and Hirosato Kondo*
Department of Medicinal Chemistry, Kanebo, New Drug Discovery Research Laboratories, 1-5-90 Tomobuchi-Cho,
Miyakojima-Ku, Osaka 534, J apan
Received November 3, 1997
We describe a mimic of the sugar unit of the E-selectin ligand, sialyl Lewis X (sLeX).
Carbohydrates are entering the realm of rational drug design, aided by the growing
understanding of the structure-function relationships. We investigated a new methodology
of preparing sLeX mimetics and developed a potent E-selectin blocker characterized by â-turn
dipeptides. Another characteristic point of this E-selectin blocker is that the six-membered
fucose ring was replaced with a five-membered fucose ring. Interestingly, it was found that
the five-membered fucose ring could also bind to a calcium ion on the E-selectin, which could
be an important role of the six-membered fucose ring. Especially, the L-Ser-D-Glu and D-Ser-
L-Glu derivatives 3a ,b showed 65-90-fold more potent inhibitory activities than the sulfated
LeX analogue 1. In addition, molecular dynamics (MD) studies indicated that the 2- and 3-OH
groups of the six-membered fucose ring, which were necessary for the calcium binding,
overlapped well with the 2- and 3-OH groups of the five-membered fucose ring. These new
findings could be useful for the design of new types of selectin blockers.
Ch a r t 1
In tr od u ction
Recently, some important vital functions of various
oligosaccharides have been clarified, such as in cancer
metastasis, anticancer activities, cell invasions by leu-
kocytes, and recognition probes for the bacteria and/or
viruses.1 In addition, synthetic2 and biological3 studies
of oligosaccharides have made remarkable progress. At
the same time, oligosaccharides have become one of the
best targets for new drug discovery. However, there are
still some problems in applying oligosaccharides as
drugs, because of the difficulties of large-scale synthesis
and absorption by oral administration. Current studies
of oligosaccharides and their biological activities focus
on their mimics, to overcome the problems described
above. Nevertheless, general methodologies and ap-
plications concerning sugar mimetics have not been
established yet.
We have studied the design and synthesis of sugar
mimetics, to develop a methodology for the preparaion
of a sugar mimetic with a simple structure. Thus, we
investigated a mimetic of a 3′-sulfo LeX derivative (1)
and have already found that compounds 2a ,b could
mimic compound 1; namely, the lactose unit of com-
pound 1 could be replaced with a simple dipeptide,
D-Ser-L-Glu or L-Ser-D-Glu, characterized by type II and/
or type II′ â-turn formation (Figure 1).4
It has been reported that sLeX itself is relatively
unstable in the blood stream, due to sensitivity to
glycosidases, i.e., fucosidase and sialidase.5 Although
any modifications of fucose on sLeX have been reported,6
there are few successful fucose mimetics. Therefore, the
search for novel fucose mimetics is very important for
the design of more useful selectin blockers. As part of
our sugar mimetic study, we next investigated mimics
of the six-membered fucose ring, as well as the applica-
tion of a type II and/or type II′ â-turn dipeptide. To
mimic the natural six-membered fucose ring necessary
for the binding to E-selectin, we focused on an unnatural
five-membered fucose ring, because it can be easily
derived from the six-membered fucose ring (L-fucose).
In this paper, we describe the synthesis of type II and/
or II′ â-turn dipeptides attached to the five-membered
fucose ring and their in vitro activities against E-selectin
binding. We also discuss an E-selectin/ligand complex
model.
Resu lts a n d Discu ssion
Ch em istr y. The five-membered fucose ring donor 7
was synthesized according to a previous paper.7 For the
synthesis of compounds 3a ,b, (Chart 1), we chose the
easily available Glu unit 4 and Ser unit 5. The coupling
of 4 and 5 in presence of the WSC, 1-ethyl-3-[3-
(dimethylamino)propyl]carbodiimide hydrochloride, and
HOBt, 1-hydroxy-1H-benzotriazole monohydrate, pro-
S0022-2623(97)00748-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 05/23/1998

Studies on Selectin Blockers
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 13 2303
F igu r e 1. Successful sLeX mimetics 1 and 2a ,b.
Sch em e 1a
a
Conditions: (a) WSC, HOBt, 80-88%; (b) SnCl2, AgOTf, TMU, 4 Å MS/CHCl3, 20-30%; (c) TFA/CHCl3; (d) WSC, HOBt, 43-47%
from 8a ,b; (e) Pd(OH)2/MeOH, 90-92%.
vided the Ser-Glu dipeptide 6a ,b in 80-88% yields. The
Ta ble 1. E-Selectin-Blocking Activities of Compounds 1, 2a ,b,
glycosylation of 6a ,b with the perbenzylfucose donor 7,
in the presence of SnCl2/AgOTf as a promoter, gave
mainly the R-glycosides 8a ,b. The deprotection of 8a ,b
under acidic conditions, followed by the coupling with
a branched alkyl chain, 2-tetradecylhexadecanoic acid
(9), according to the general method for an amide
linkage, afforded compounds 10a ,b in moderate yields.
Finally, compounds 10a ,b were transformed in good
yields, by the removal of the protecting groups, into the
desired compounds 3a ,b, as a sodium salt for 3a and
an arginine salt for 3b. (Scheme 1).
and 3a ,b
compd
*1
*2
E-selectin (IC50, µM)
In Vitr o Eva lu a tion of Com p ou n d s 3a ,b. Both
compounds 3a ,b were potent blockers of E-selectin-
sLeX binding, with IC50 values of 3.1 and 4.3 µM,
respectively (Table 1). Compounds 3a ,b were 65-90-
fold more potent blockers than compound 1 (IC50, 280
µM). Interestingly, both compounds 3a ,b showed simi-
lar activities to compounds 2a ,b, with IC50’s of 13 µM
for 2a and 5.5 µM for 2b. We have already reported
that 2a ,b could bind to E-selectin with type II and/or
II′ â-turn forms, respectively.4 Since the structures of
compounds 3a ,b are exactly the same as those of the
corresponding 2 forms, except for the fucose, 3a ,b could
also have the type II and II′ â-turn conformations such
as 2a ,b, respectively. These conformations would play
important roles in conserving the positions of the three
essential functional groups, e.g., the fucose, the branched
alkyl chain, and the negatively charged group of 3a ,b,
for binding to E-selectin. This description indicates that
the derivation of a pharmacophore conserving the
essential binding groups could be a useful methodology
3a
3b
L
D
L
3.1
4.3
D
1
2a
2b
280
13
5.5
L
D
L
D
to discover a new selectin blocker, and especially, the
type II and/or II′ â-turn dipeptides could be an ap-
plicable template for sugar mimetics. In addition, it is
interesting to note that the five-membered fucose ring
would work as well as the six-membered fucose ring.
To clarify the structural homology between the five-
membered fucose ring and the six-membered fucose
ring, we next performed a conformational analysis of
the five-membered fucose ring using a molecular dy-
namics (MD) calculation method and compared it with
that of the six-membered fucose ring.
Con for m a tion a l An a lysis of th e F ive-Mem ber ed
F u cose Rin g Usin g Molecu la r Dyn a m ics. It has
been reported that a five-membered ring occurs in two

2304 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 13
Hiramatsu et al.
F igu r e 2. Diagrammatic projections of the five-membered
ring fucose in two idealized twist conformations.
ranges of conformations, based on the φ (torsion angle
φ: C1-C2-C3-C4) and ψ (torsion angle ψ: O4-C1-
C2-C3) values, usually referred to as C(2)-endo for φ
< 0 and ψ > 0 and C(3)-endo for φ > 0 and ψ < 0, as
shown in Figure 2.8 Especially, in the case of the C(3)-
endo form (φ > 0 and ψ < 0), the 2-OH and 3-OH groups
of the five-membered fucose ring are directed toward
the equatorial positions and could coordinate with the
calcium ion of E-selectin. To obtain the φ-ψ plots of
conformers of the five-membered fucose ring, we inves-
tigated conformation using CAMDAS.9 First, we per-
formed an MD calculation to explore the conformational
space of the five-membered fucose ring. The calculation
was performed for 200 ps with an integral time step of
1 fs, the temperature in the MD calculation system was
maintained at 900 K, and the lengths of the covalent
bonds were fixed with the SHAKE algorithm.10 During
the calculation, a conformer was sampled for every 0.1
ps, and the clustering of similar conformers was carried
out. The criteria of the clustering involve 5° deviations
of the two torsion angles, φ and ψ. Namely, if the
deviations of the torsion angles of the two conformers
sampled are smaller than 5°, then each conformer is
defined as the same. Otherwise, the conformer sampled
here is added to the conformer’s list as a new one. Thus,
we investigated the φ-ψ plots of 40 conformers of the
five-membered fucose ring, in order of low energy. As
a result, 9 of the first 10 stable conformers showed C(3)-
endo forms, as shown in Figure 3. In addition, it was
found that the 2- and 3-OH groups of the five-membered
fucose ring in the C(3)-endo form overlapped well with
the corresponding 2- and 3-OH groups of the six-
membered fucose ring, which would be the essential
binding site for calcium on E-selectin (Figure 4). Ac-
cordingly, the five-membered fucose ring would coordi-
nate to the calcium ion and could mimic the six-
membered fucose ring.
F igu r e 3. φ-ψ plots of the 40 conformations of five-membered
ring fucose in order of low energy during the 200-ps MD
calculation. These numbers show the order of low-energy
conformation, and the bold numbers show the first 10 confor-
mations. The horizontal and vertical axes represent φ and ψ
values in degrees. φ: C1-C2-C3-C4. ψ: O4-C1-C2-C3. If
φ > 0 and ψ < 0, O2 and O3 are equatorial form; if φ < 0 and
ψ > 0, O2 and O3 are axial form.
F igu r e 4. Stereoview of superimposition of six-membered ring
fucose and the most stable conformer of five-membered ring
fucose obtained from CAMDAS. The thick line indicates five-
membered ring fucose, and the thin line indicates six-
membered ring fucose.
directed to Lys111 of E-selectin. On the other hand,
one of the branched alkyl chains was directed to a
hydrophobic surface consisting of Tyr44, Pro46, and
Tyr48, and another alkyl chain was directed to a shallow
cavity consisting of Ala9, Leu114, Lys111, Lys112, and
Lys113. In addition, the initial bound model of 3b
formed a type II′ â-turn conformation.
Next, we subjected the initial bound model of 3b to a
200-ps MD simulation with the explicit water solvent
and evaluated the dynamic behavior of the model in
solution. We analyzed the interaction mode between
the Glu in 3b and the basic residue on E-selectin, as
well as the conformation change of the peptide portion
in 3b. As a result, during the MD calculation 3b well-
conserved the type II′ â-turn conformation, and the
interaction distance between Lys111 of E-selectin and
the carboxylate of 3b was found as well, as shown in
Figure 5. Additionally, we analyzed the behavior of the
branched alkyl chains of 3b during the MD simulation
in solution. Since a direct estimation of the hydrophobic
force is difficult, we used the ratio of the solvent-
accessible surface area of the alkyl chain that was
hidden by the protein as an indication of the strength
of the hydrophobic interaction (see the Experimental
Section). As a result, the hidden surface area of each
alkyl chain was retained as the initial value and
fluctuated minimally during the MD simulation. Thus,
compound 3b interacted similarly with E-selectin, as
observed in the complex model4 of 2b/E-selectin. These
In itia l Mod el Con str u ction , MD Sim u la tion , a n d
An a lysis of th e In ter a ction betw een 3b a n d E-
Selectin . The structure of the lectin domain of E-
selectin was obtained from the Protein Data Bank
(indentity code: 1ESL). We constructed the initial
bound model of 3b to conserve the interactions of the
three functional groups with E-selectin, based on the
initial bound model of 2b /E-selectin,4 because the
structure of 3b is the same as that of the corresponding
2b, except for the fucose. The conformation of the five-
membered fucose ring in 3b corresponded to the lowest-
energy conformer obtained from the calculation. The
2- and 3-OH groups of the fucose were coordinated to
the calcium, and the carboxylate of the L-Glu was

Studies on Selectin Blockers
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 13 2305
F igu r e 5. Stereorepresentation of the models of E-selectin and 2b or 3b complex models in solution. These models are the
equilibrium structures during MD calculation (200 ps). The structures of 2b and 3b are indicated by thick lines. Thin lines indicate
the structure of E-selectin. Only the residues in the ligand-binding site on the protein are shown. Dashed lines indicate interactions
(hydrogen bond, electrostatic interaction, calcium coordination).
The coordinates of the system were recorded at every 0.2 ps
during the calculation, for the following analyses of the
trajectory. In the calculations with the explicit water a
dielectric constant, ꢀ ) 1, was used. During MD calculations,
data correlate with the in vitro activity of 3b. Namely,
it was found that the six-membered fucose ring of
compound 2b could be replaced with the five-membered
fucose ring. Moreover, we also constructed an initial
bound model of 3b/E-selectin and subjected this model
to a 200-ps MD simulation. It was found that the final
bound model of 3b/E-selectin was also similar to that
of the 2b/E-selectin complex model4 reported previously.
In conclusion, we succeeded in the design of sLeX
mimetics based on modified type II and/or II′ â-turn
dipeptides. Especially, the L-Ser-D-Glu 3a and D-Ser-
L-Glu 3b showed 65-90-fold more potent inhibitory
activities than the sulfated LeX analogue 1. In addition,
we discovered a homologous fucose sugar unit necessary
for E-selectin binding, which is characterized by a five-
membered fucose ring. Although the six-membered
fucose ring might be a substrate for a fucosidase in the
body, the five-membered fucose ring would not have this
problem. Therefore, these new findings well be useful
for the design of new types of selectin blockers, which
are resistant to hydrolytic enzymes, such as a fucosi-
dase. We are now in the progress of synthesizing a sLeX
analogue attached to a five-membered fucose ring.
all of the bond lengths were fixed by SHAKE algorithm.
A
nonbonded cutoff of 9 Å was used in all of the above calcula-
tions. The atoms coordinated to the calcium ion were fixed
by harmonic position constraints with a force constant of 50
kcal/(mol•Å). All calculations were performed using AMBER
4.0.1 with the force field published by Cornell et al.12 The
force-field parameters for the carbohydrate portion of 3a ,b
were derived from GLYCAM-93, a general force field for
carbohydrates that is suitable to AMBER.
An a lyses of th e Tr a jector y. Various analyses of the
trajectory from MD calculation in solution were performed
with several in-house programs. For the interactions observed
in the initial model complex, we calculated the distances of
each interaction from the sets of coordinates in the trajectory.
For the analysis of the behavior of the branched alkyl chains,
we calculated the “ratio of hidden surface area” as a measure
of hydrophobic interactions with the protein. This indicator
was calculated in the following way: for each set of coordinates
in the trajectory, we calculated the solvent-accessible surface
area of the alkyl chain portions. The area calculations were
performed in two ways: one was the area of free alkyl chains
(Afree), which was calculated after the deletion of the protein
coordinates. The other was the area of bound alkyl chains
(Acomplex), calculated in the presence of the protein. Then, the
ratio of hidden surface areas was calculated by the following
equation:
Exp er im en ta l Section
Sim u la tion of th e Com p lex in Solu tion Usin g Molec-
u la r Dyn a m ics. The models of complexes were solvated by
the addition of TIP3P water molecules11 which were added
within 25 Å from the center of each of the compounds 3a ,b.
The water molecules were then minimized until the root-mean-
square of the gradients was below 0.05 kcal/(mol•Å). Following
the minimization, the water molecules were equilibrated by a
5-ps MD calculation at a temperature of 298 K. An integral
time step of ∆t ) 2 fs was used. After that, all of the water
molecules, compounds 3a ,b, and the binding sites on E-selectin
(Ser6-Met10, Ser43-Tyr94, Trp76-Arg84, Val91-Val101,
Trp104-Ala115) were minimized until the root-mean-square
of gradients was below 0.05 kcal/(mol•Å). Part of the system
was then subjected to a 200-ps MD simulation at a tempera-
ture of 298 K. An integral time step of ∆t ) 2 fs was used.
ratio of hidden surface area (%) ) (Afree - Acomplex)/Afree
×
100
As indicated by the above equation, the ratio of hidden
surface area is the ratio of the solvent-accessible surface area
of the alkyl chains that was buried in the protein. A larger
percentage of hidden surface area reflects wider alkyl chain
contacts with the protein surface. The calculation of solvent-
accessible surface area was done by the methods of Richmond13
and Wesson and Eisenberg.14 Hydrogen atoms were not
included in the calculation.
In h ibition Assa y of E-Selectin -sLex Bin d in g. The

2306 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 13
Hiramatsu et al.
construction of the selectin-immunoglobulin was carried out
according to a previous paper.15
g), AgOTf (0.89 g, 3.46 mM), and SnCl2 (0.66 g, 3.48 mM) in
CH2Cl2 (10 mL) was stirred for 5 h under nitrogen atmosphere
at room temperature and cooled to -40-50 °C. TMU (1.0 g,
8.64 mM), 2,3,5-tri-O-benzyl-L-fucofuranosyl fluoride (7) (1.1
g, 2.5 mM) in CH2Cl2 (2 mL) and compound 6a (0.75 g, 1.71
mM) in CH2Cl2 (5 mL) were added to the mixture successively
and stirred for 1 h at the same temperature and for 19 h under
gradually returning to room temperature. The precipitate was
filtered off, and the filtrate was concentrated. The residue was
purified by thin-layer chromatography developing with 3:2
AcOEt/n-hexane to give 8b (0.45 g) as a white crystal: [R]D
A solution of sLex pentasaccharide ceramide, in a 1:1
mixture of methanol and distilled water, was pipetted into
microtiter plate wells (96 wells; Falcon PRO-BIND) at 100
pmol/50 µL/well and was adsorbed by evaporating the solvent.
The wells were washed twice with distilled water, blocked with
5% BSA (bovine serum albumin)-1 mM CaCl2/50 mM imida-
zole buffer (pH 7.2) for 1 h at room temperature, and washed
three times with 50 mM imidazole buffer (pH 7.2).
Separately, a 1:1 volumetric mixture of a 1:500 dilution in
1% BSA-1 mM CaCl2/50 mM imidazole buffer (pH 7.2) of
biotinylated goat F(ab′)2 anti-human IgG(γ)/streptavidin-
alkaline phosphatase (Zymed Lab Inc.) and a E-selectin-
immunoglobulin fusion protein (E-selectin-Ig) was incubated
at room temperature for 30 min to form a complex. The test
compounds were dissolved in distilled water (2a ,b; DMSO) at
10 mM and finally diluted by 1 mM CaCl2/50 mM imidazole
buffer (pH 7.2) to final concentrations of 200, 100, 50, 25, 12.5,
6.25, 3.13, and 1.56 µM, respectively. Reactant solutions were
prepared by incubating 30 µL of this solution at each concen-
tration with 30 µL of the above complex solution for 30 min at
room temperature. This reactant solution was then added to
the above microtiter wells at 50 µL/well and incubated at 37
°C for 45 min. The wells were washed three times with 50
mM imidazole buffer (pH 7.2) and distilled water, respectively,
followed by addition of p-nitrophenyl phosphate (1 mg/mL) and
0.01% MgCl2 in 1 M diethanolamine (pH 9.8) at 50 µL/well.
The reactant mixture was developed for 120 min at room
temperature, and absorbance at 405 nm was measured.
Percent binding was calculated by the following equation:
1
-19° (c ) 0.1, MeOH); mp 122-126 °C; H NMR (DMSO-d6)
δ 1.07 (d, 3H, J ) 6.2 Hz), 1.33 (s, 9H), 1.68-1.85 (m, 1H),
1.90-2.08 (m, 1H), 2.21-2.42 (m, 2H), 2.57 (d, 3H, J ) 4.5
Hz), 3.45-3.60 (m, 1H), 3.58-3.70 (m, 2H), 3.70-3.82 (m, 1H),
4.00 (d, 2H, J ) 4.5 Hz), 4.10-4.30 (m, 2H), 4.35-4.65 (m,
6H), 5.03 (3, 2H), 5.05 (bs, 1H), 6.93 (d, 1H, J ) 6.7 Hz), 7.15-
7.40 (m, 20H), 7.79 (d, 1H, J ) 4.4 Hz), 8.19 (d, 1H, J ) 7.6
Hz). Anal. Calcd for C48H59N3O11: C, H, N.
[N-(ter t-Bu toxyca r bon yl)-O-(2,3,5-tr i-O-ben zyl-r-L-fu -
cofu r a n osyl)-D-ser yl]-L-glu t a m ic a cid 1-m et h yla m id e
5-ben zyl ester (8b): yield 20.5% as a syrup; [R]D -21° (c)0.1,
MeOH); 1H NMR (DMSO-d6) δ: 1.07 (d, 3H, J ) 6.2 Hz), 1.36
(s, 9H), 1.65-1.85 (m, 1H), 1.90-2.12 (m, 1H), 2.32 (t, 2H, J
) 7.2 Hz), 2.57 (d, 3H, J ) 4.5 Hz), 3.45-3.70 (m, 3H), 3.78-
3.91 (m, 1H), 3.95-4.05 (m, 1H), 4.10-4.30 (m, 2H), 4.35-
4.72 (m, 6H), 5.02 (s, 2H), 5.09 (s, 1H), 7.16 (d, 1H, J ) 7.0
Hz), 7.20-7.45 (m, 20H), 7.79 (d, 1H, J ) 4.0 Hz), 8.17 (d, 1H,
J ) 7.9 Hz). Anal. Calcd for C48H59N3O11: C, H, N.
Gen er a l P r oced u r e for th e P r ep a r a tion of 10a ,b:
[N-(2-Tetr a d ecylh exa d eca n oyl)-O-(2,3,5-tr i-O-ben zyl-r-L-
fu cofu r a n osyl)-L-ser yl]-D-glu ta m ic Acid 1-Meth yla m id e
5-Ben zyl Ester (10a ). To a solution of 8a (0.4 g, 0.47 mM)
in CH2Cl2 (3 mL) was added TFA (3 mL) at 0 °C, and the
mixture was stirred for 2 h at room temperature. The reaction
mixture was concentrated, and the residue was dissolved in
CHCl3 (50 mL), washed with saturated sodium hydrogen
carbonate, and dried (MgSO4), and the solvent was removed
in vacuo. The residue was dissolved in DMF (30 mL),
2-tetradecylhexadecanoic acid (210 mg, 0.46 mM) was added
to the solution, and the mixture was dissolved by heating and
then cooled to room temperature. WSC (170 mg, 0.73 mM)
and HOBt (110 mg, 0.73 mM) were added to the solution. After
the mixture stirred for 22 h, CHCl3 (80 mL) was added to the
solution, washed with 1 N HCl, saturated sodium hydrogen
carbonate, and brine successively, dried (MgSO4), and concen-
trated. The residue was purified by thin-layer chromatogra-
phy developing with 20:1 CHCl3/methanol to give 10a (270 mg,
47%) as a white crystal: [R]D -14° (c ) 0.1, CHCl3); mp 127-
129 °C; 1H NMR (DMSO-d6) δ 0.75-0.95 (m, 6H), 1.06 (d, 3H,
J ) 6.1 Hz), 1.10-1.60 (m, 55H), 1.65-1.85 (m, 1H), 1.85-
2.10 (m, 1H), 2.10-2.25 (m, 1H), 2.25-2.40 (m, 2H), 2.55 (d,
3H, J ) 4.5 Hz), 3.49-3.60 (m, 1H), 3.60-3.70 (m, 2H), 3.70-
3.85 (m, 1H), 3.95-4.05 (m, 2H), 4.15-4.25 (m, 1H), 4.39-
4.72 (m, 7H), 5.01 (s, 2H), 5.04 (d, 1H, J ) 3.3 Hz), 7.15-7.40
(m, 20H), 7.84 (d, 1H, J ) 4.6 Hz), 8.11 (d, 1H, J ) 6.7 Hz,
8.28 (d, 1H, J ) 8.2 Hz). Anal. Calcd for C73H109N3O10: C, H,
N.
% binding ) (X - C/A - C) × 100
wherein X is the absorbance of wells containing the test
compounds at each concentration, C is the absorbance of wells
not containing the selectin-Ig and test compounds, and A is
the absorbance of control wells not containing the test com-
pounds. The results of inhibitory activities are presented in
Table 1 as IC50 values. The number of replicates is 2.
Gen er a l P r oced u r e for th e P r ep a r a tion of Dip ep tid es
6a ,b: N-(ter t-Bu toxyca r bon yl)-D-ser yl-L-glu ta m ic Acid
1-Meth yla m id e 5-Ben zyl Ester (6b). To the L-4 (10.0 g, 28.5
mM) was added 4 N HCl/1,4-dioxane (30 mL), and the solution
was stirred for 0.5 h at room temperature. The mixture was
concentrated, and the residue was dissolved in DMF (200 mL),
added to Et3N (5.4 mL), and then cooled to 0 °C. D-5 (5.9 g),
WSC (6.6 g, 34.2 mM), and HOBt (5.2 g, 34.2 mM) were added
to the residue, and this was stirred for 18 h at room temper-
ature. The solvent was removed under reduced pressure, and
the oily residue was dissolved in AcOEt (300 mL), washed with
0.1 N HCl, saturated sodium hydrogen carbonate and brine,
dried (MgSO4), and concentrated. The precipitate was filtered,
and recrystallization from AcOEt-n-hexane gave 6b (10.3 g,
82.5%) as a white crystal: [R]D -5° (c ) 0.1, CHCl3); mp 127-
128 °C; 1H NMR (DMSO-d6) δ 1.38 (s, 9H), 1.65-1.85 (m, 1H),
1.90-2.15 (m, 1H), 2.30-2.40 (m, 2H), 2.58 (d, 3H, J ) 4.6
Hz), 3.54 (t, 2H, J ) 5.6 Hz), 3.85-4.00 (m, 1H), 4.13-4.29
(m, 1H), 4.88 (t, 1H, J ) 5.6 Hz), 5.07 (s, 2H), 6.82 (d, 1H, J
) 6.9 Hz), 7.30-7.45 (m, 5H), 7.77 (d, 1H, J ) 4.6 Hz), 8.03
(d, 1H, J ) 8.3 Hz). Anal. Calcd for C21H31N3O7: C, H, N.
N-(ter t-Bu toxyca r bon yl)-L-ser yl-D-glu ta m ic a cid 1-m e-
th yla m id e 5-ben zyl ester (6a ): yield 80.6%; [R]D +5° (c )
[N-(2-Tetr a d ecylh exa d eca n oyl)-O-(2,3,5-tr i-O-ben zyl-
r-L-fu cofu r a n osyl)-D-ser yl]-L-glu ta m ic a cid 1-m eth yla -
m id e 5-ben zyl ester (10b): yield 43.3%; [R]D -19° (c ) 0.1,
CHCl3); mp 114-117 °C; 1H NMR (DMSO-d6) δ 0.75-0.95 (m,
6H), 1.06 (d, 3H, J ) 6.2 Hz), 1.10-1.50 (m, 52H), 1.65-1.85
(m, 1H), 1.88-2.12 (m, 1H), 2.15-2.30 (m, 1H), 2.30-2.42 (m,
2H), 2.55 (d, 3H, J ) 4.5 Hz), 3.45-3.72 (m, 3H), 3.80-3.95
(m, 1H), 3.95-4.05 (m, 2H), 4.16-4.30 (m, 1H), 4.38-4.72 (m,
7H), 5.00 (s, 2H), 5.10 (d, 1H, J ) 3.1 Hz), 7.15-7.45 (m, 20H),
7.84 (d, 1H, J ) 4.7 Hz), 8.16 (d, 1H, J ) 6.7 Hz), 8.26 (d, 1H,
J ) 8.1 Hz). Anal. Calcd for C73H109N3O10: C, H, N.
1
0.1, CHCl3); mp 124-127 °C; H NMR (DMSO-d6) δ 1.38 (s,
9H), 1.66-1.88 (m, 1H), 1.92-2.12 (m, 1H), 2.30-2.42 (m, 2H),
2.57 (d, 3H, J ) 4.6 Hz), 3.54 (t, 2H, J ) 5.6 Hz), 3.86-4.00
(m, 1H), 4.13-4.30 (m, 1H), 4.88 (t, 1H, J ) 5.7 Hz), 5.07 (s,
2H), 6.82 (d, 1H, J ) 6.9 Hz), 7.30-7.45 (m, 5H), 7.77 (d, 1H,
J
C
) 4.5 Hz), 8.03 (d, 1H, J ) 8.1 Hz). Anal. Calcd for
21H31N3O7; C, H, N.
Gen er a l P r oced u r e for th e P r ep a r a tion of 8a ,b by
Gen er a l P r oced u r e for th e P r ep a r a tion of 3a ,b: [N-(2-
Tetr a d ecylh exa d eca n oyl)-O-(r-L-fu cofu r a n osyl)-L-ser yl]-
D-glu ta m ic Acid 1-Meth yla m id e Sod iu m Sa lt (3a ). To a
solution of 10a (200 mg, 0.168 mM) in ethanol (30 mL) was
added 20% Pd(OH)2/C (200 mg), and the mixture was stirred
Glycosyla tion of 6a ,b w ith 2,3,5-Tr i-O-ben zylfu cose Do-
n or 7: [N-(ter t-Bu toxyca r bon yl)-O-(2,3,5-tr i-O-ben zyl-r-
L-fu cofu r a n osyl)-L-ser yl]-D-glu ta m ic Acid 1-Meth yla m id e
5-Ben zyl Ester (8a ). A mixture of 4 Å molecular sieve (1.0

Studies on Selectin Blockers
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 13 2307
Kurokawa, K.; Nakai, Y.; Achiha, T.; Kiyoi, T.; Kondo, H. A
Highly Practical Synthesis of Sulfated Lewis X: One-Pot, Two-
Step Glycosylation Using “Armed/Disarmed” Coupling and
Selective Benzoylation and Sulfation. J . Org. Chem. 1997, 62,
6876-6881.
for 5 h under hydrogen (3-4 atmospheric pressure). The
precipitate was filtered off, and the filtrate was concentrated
in vacuo. The residue was dissolved in methanol and water.
To the solution was added 1 H NaOH (1 equiv), and the
mixture stirred for 5 min. The reaction mixture was concen-
trated in vacuo. The residue was lyophilized with water to
give 3a (79 mg, 56.7%) as a white powder: mp 147-154 °C
(3) (a) Mulligan, M. S.; Lowe, J . B.; Larsen, R. D.; Paulson, J .; Zheng,
Z.; DeFrees, S.; Maemura, K.; Fukuda, M.; Ward, P. Protective
Effects of Sialylated Oligosaccharides in Immune Complex-
Induced Acute Lung Injury. J . Exp. Med. 1993, 178, 623-631.
(b) Springer, T. A.; Lasky, L. A. Sticky sugar for selectins. Nature
1991, 349, 196-197. (c) Fukushima, K.; Ohkura, T.; Kanai, M.;
Kuroki, M.; Matsuoka, Y.; Kobata, A.; Yamashita, K. Carbohy-
drate structures of a normal counterpart of the carcinoembryonic
antigen produced by colon epithelial cells of normal adults.
Glycobiology 1995, 5, 105-115. (d) Feizi, T.; Larkin, M. AIDS
and glycosylation. Glycobiology 1990, 1, 17-23. (e) Nguyen, M.;
Eilber, F. R.; DeFrees, S. Novel Synthetic Analogues of Sialyl
Lewis X Can Inhibit Angiogenesis in Vitro and in Vivo. Biochem.
Biophys. Res. Commun. 1996, 228, 716-723.
(4) Tsukida, T.; Hiramatsu, Y.; Tsujishita, H.; Kiyoi, T.; Yoshida,
M.; Kurokawa, K.; Moriyama, H.; Ohmoto, H.; Wada, Y.; Saito
T.; Kondo, H. Studies on Selectin Blockers. 5. Design, Synthesis,
and Biological profile of Sialyl Lewis X Mimetics Based on
Modified Serine-Glutamic Acid Dipeptides. J . Med. Chem. 1997,
40, 3534-3541.
1
dec; H NMR (DMSO-d6) δ 0.83-0.87 (m, 6H), 1.01 (d, 3H, J
) 6.3 Hz), 1.10-1.55 (m, 52H), 1.60-2.05 (m, 4H), 2.10-2.32
(m, 1H), 2.55 (d, 3H, J ) 3.5 Hz), 3.20-3.35 (m, 2H), 3.40-
3.65 (m, 3H), 3.66-3.76 (m, 1H), 3.77-3.86 (m, 1H), 3.86-
4.04 (m, 2H), 4.35 (s, 1H), 4.66 (d, 1H, J ) 4.2 Hz), 8.23 (d,
1H, J ) 7.4 Hz), 9.19 (d, 1H, J ) 5.8 Hz). Anal. Calcd for
C
45H84N3O10Na: C, H, N.
[N-(2-Tetr a d ecylh exa d eca n oyl)-O-(r-L-fu cofu r a n osyl)-
D-ser yl]-L-glu ta m ic a cid 1-m eth yla m id e L-a r gin in e sa lt
(3b): yield 82%; mp 175 °C decomp; 1H NMR (DMSO-d6) δ
0.75-0.95 (m, 6H), 1.08 (d, 3H, J ) 6.5 Hz), 1.1-1.5 (m, 54H),
1.5-2.2 (m, 9H), 2.55 (d, 3H, J ) 4.2 Hz), 3.6-3.8 (m, 5H),
3.9-4.05 (m, 1H), 4.1-4.2 (m, 1H), 4.2-4.4 (m, 1H), 4.69 (d,
1H, J ) 4.2 Hz), 7.47 (d, 1H, J ) 3.9 Hz), 8.51 (d, 1H, J ) 7.3
Hz), 8.85 (bs, 1H). Anal. Calcd for C45H85N3O10‚C5H12N4O2:
C, H, N.
(5) (a) Uchiyama, T.; Vassilev, V. P.; Kajimoto, T.; Wong, W.; Huang,
H.; Lin, C.-C.; Wong, C.-H. Design and Synthesis of sialyl Lewis
X Mimetics. J . Am. Chem. Soc. 1995, 117, 5395-5396. (b)
Uchiyama, T.; Woltering, T. J .; Wong, W.; Lin, C.-C.; Kajimoto,
T.; Takebayashi, M.; Weitz-Schmidt, G.; Asakura, T.; Noda, M.;
Wong, C.-H. Design and Synthesis of C-Linked Fucosides as
Inhibitors of E-Selectin. Bioorg. Med. Chem. 1996, 4, 1149-1165.
(6) (a) Dasgupta, F.; Rao, B. N. N. Anti-Adhesive Therapeutics: A
New Class of Antiinflammatory Agents. Exp. Opin. Invest. Drugs
1994, 3, 709-724. (b) Musser, J . H.; Rao, N.; Nashed, M.;
Dasgupta, F.; Abbas, S.; Nemetalla, A.; Date, V.; Foxall, C.; Asa,
D.; J ames, P.; Tyrell, D.; Brandley, B. K. Structure-Activity
Studies Based on the Sialyl Lewis X Epitope. In Trends in
Receptors Research; Claassen, V., Ed.; Elsevier: Amsterdam,
1993; pp 33-40. (c) Siuzdak, G.; Zheng, Z.-L.; Rhamphal, J . Y.;
Ichikawa, Y.; Nicolaou, K. C.; Gaeta, F. C. A.; Chatman, K. S.;
Wong, C.-H. Examination of the Sialyl Lewis X-Calcium
Complex by Electrospray Mass Spectrometry. Bioorg. Med.
Chem. Lett. 1994, 4, 2863-2866. (d) Ramphal, J . Y.; Zheng, Z.-
L.; Perez, C.; Walker, L. E.; DeFrees, S. A.; Gaeta, F. C. A.
Structure-Activity Relationships of Sialyl LewisX-Containing
Oligosaccharides. 1. Effect of Modifications of the Fucose Moiety.
J . Med. Chem. 1994, 37, 3459-3463.
Refer en ces
(1) (a) Berg, E. L.; Yoshino, T.; Rott, L. S.; Robinson, M. K.; Warnock,
R. A.; Kishimoto, T. K.; Picker, L. J .; Butcher, E. C. The
Cutaneous Lymphocyte Antigen is a Skin Lymphocyte Horming
Receptor for the Vascular Lectin Endothelial Cell-Leukocyte
Adhesion Molecule 1. J . Exp. Med. 1991, 174, 1461-1466. (b)
Picker, L. J .; Kishimoto, T. K.; Smith, C. W.; Warnock, R. A.;
Butcher, E. C. ELAM-1 is an Adhesion Molecule for Skin-
Homing T Cells. Nature 1991, 349, 796-799. (c) Lorant, D. E.;
Topham, M. K.; Whatley, R. E.; McEver, R. P.; Mclntyre, T. M.;
Prescott, S. M.; Zimmerman, G. A. Inflammatory Roles of
P-selectin. J . Clin. Invest. 1993, 92, 559-570. (d) Symon, F. A.;
Walsh, G. M.; Watson, S. R.; Wardlaw, A. J . Eosinophil Adhesion
to Nasal Polyp Endothelium is P-Selectin-Dependent. J . Exp.
Med. 1994, 80, 371-376. (e) Akbar, A. N.; Salmon, M.; J anossy,
G. The Synergy Between Naive and Memory T Cells During
Activation. Immunol. Today 1991, 12, 184-188. (f) Duijvestijn,
A. M.; Horst, E.; Pals, S. T.; Rouse, B. N.; Steere, A. C.; Picker,
L. J .; Meijer, C. J . L. M.; Butcher, E. C. High Endothelial
Differenciation in Human Lymphoid and Inflammatory Tissue
Defined by Monoclonal Antibody HECA-452. Am. J . Pathol.
1988, 130, 147-155. (g) Nguyen, M.; Eilber, F. R.; DeFrees, S.
Novel Synthetic Analogues of Sialyl Lewis X Can Inhibit
Angiogenesis in Vitro and in Vivo. Biochem. Biophys. Res.
Commun. 1996, 228, 716-723. (h) Kannagi, R. Carbohydrate-
Mediated Cell Adhesion Involved in Hematogenous Metastasis
of Cancer. Glycoconjugate J . 1997, 14, 577-584. (i) Selectin-
Carbohydrate Interactions During Inflammation and Metastasis.
Glycoconjugate J . 1997, 14, 585-591. (j) Austrup, F.; Vestweber,
D.; Borges, E.; Lo¨hning, M.; Bra¨uer, R.; Herz, U.; Renz, H.;
Hallmann, R.; Scheffold, A.; Radbruch, A.; Hamann, A. P- and
E-selectin mediate recruitment of T-helper-1 but not T-helper-2
cells into inflamed tissues. Nature 1997, 385, 81-83.
(7) Motoki, K.; Morita, M.; Kobayashi, E.; Uchida, T.; Akimoto, K.;
Fukushima, H.; Koezuka, Y. Immunostimulatory and Antitumor
Activities of Monoglycosyl ceramides having Various Sugar
Moieties. Biol. Pharm. Bull. 1995, 18, 1487-1491.
(8) Altona, C.; Sundaralingam, M. Conformational Analysis of the
Sugar Ring in Nucleosides and Nucleotides. A New Description
Using the Concept of Pseudorotation. J . Am. Chem. Soc. 1972,
94, 8205-8212.
(9) Tsujishita, H.; Hirono, S. CAMDAS: An automated conforma-
tional analysis system using molecular dynamics. J . Comput.-
Aided Mol. Des. 1997, 11, 305-315.
(2) For recent examples of elegant strategies for oligosaccharide
synthesis, see: (a) Kondo, H.; Ichikawa, Y.; Wong, C.-H. â-Sialyl
Phosphite and Phosphoramidite: Synthesis and Application to
the Chemoenzymatic Synthesis of CMP-Sialic Acid and Sialyl
Oligosaccharides. J . Am. Chem. Soc. 1992, 114, 8748-8750. (b)
Halcomb, R. L.; Huang, H.; Wong, C.-H. Solution- and Solid-
Phase Synthesis of Inhibitors of H. Pylori Attachment and
E-Selectin-Mediated Leukocyte Adhesion. J . Am. Chem. Soc.
1994, 116, 11315-11322. (c) Mehta, S.; Pinto, B. M. Novel
Glycosidation Methodology: The Use of Phenyl Selenoglycosides
as Glycosyl Donors and Acceptors in Oligosaccharide Synthesis.
J . Org. Chem. 1993, 58, 3269-3276. (d) Raghavan, S.; Kahne,
D. A One-Step Synthesis of the Ciclamycin Trisaccharide. J . Am.
Chem. Soc. 1993, 115, 1580-1581. (e) Yamada, H.; Harada, T.;
Takahashi, T. Synthesis of an Elicitor-Active Hexaglucoside
Analogue by a One-Pot, Two-Step Glycosidation Procedure. J .
Am. Chem. Soc. 1994, 116, 7919-7920. (f) Kanie, O.; Ito, Y.;
Ogawa, T. Orthogonal Glycosylation Strategy in Oligosaccharide
Synthesis. J . Am. Chem. Soc. 1994, 116, 12073-12074. (g)
Madsen, R.; Udodong, U.; Roberts, C.; Mootoo, D.; Mootoo, U.
D. R.; Konradsson, P.; Fraser-Reid, B. Studies related to
Synthesis of Glycophosphatidylinositol Membrane-Bound Pro-
tein Anchors. 6. Convergent Assembly of Subunits. J . Am. Chem.
Soc. 1995, 117, 1554-1565. (h) Tsukida, T.; Yoshida, M.;
(10) van Gunsteren, W. F.; Berendsen, H. J . C. Algorithms for
macromolecular Dynamics and Constraint Dynamics. Mol. Phys.
1977, 34, 1311-1327.
(11) J orgensen, W. L.; Chandrasekhar, J .; Madura, J .; Impey, R. W.;
Klein, M. L. Comparison of Simple Potential Functions for
simulating Liquid Water. J . Chem. Phys. 1983, 79, 926-935.
(12) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K.
M., J r.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell,
J . W.; Kollman, P. A. A Second Generation Force Field for the
Simulation of proteins, Nucleic Acids, and Organic Molecules.
J . Am. Chem. Soc. 1995, 117, 5179-5197.
(13) Richmond, T. J . Solvent Accessible Surface Area and Excluded
Volume in Proteins. J . Mol. Biol. 1984, 178, 63-89.
(14) Wesson, L.; Eisengerg, D. Atomic Solvation Parameters Applied
to Molecular Dynamics of Proteins in Solution. Protein Sci. 1992,
1, 227-235.
(15) Erbe, D. V.; Watson, S. R.; Presta, L. G.; Wolitzky, B. A.; Foxall,
C.; Brandley, B. K.; Lasky, L. A. P- and E-selectin use Common
Sites for Carbohydrate Ligand Recognition and Cell Adhesion.
J . Cell Biol. 1993, 120, 1227-1235.
J M9707481