Carbohydrate-protein interactions at interfaces: synthesis of thiolactosyl glycolipids and design of a working model for surface plasmon resonance.

Article abstract of DOI:10.1039/b210672h

Thiolactosyl lipids designed for carbohydrate-protein binding studies have been synthesised. One representative was selected for binding studies with a plant lectin RCA120, the agglutinin from Ricinus communis. The interactions were measured quantitatively in real time using a BIAcore surface plasmon resonance instrument. Removal of much of the galactose from the thiolactosyl lipid in situ with beta-galactosidase showed that the lectin binding was highly specific. A dissociation constant KD = 8.77 x 10(-8) M was measured for 1-[2-[2-(2-[beta-D-galactopyranosyl-(1-->4)-1-thio-beta-D -glucopyranosyl]ethoxy)ethoxy]ethoxy]octadecane 30 which is four orders of magnitude greater than that determined for binding to lactose in solution. A concentration of lactose of > 80 mM was required to block the lectin binding to thiolactosyl lipid in a neomembrane.

Full text of DOI:10.1039/b210672h

View Article Online / Journal Homepage / Table of Contents for this issue
Carbohydrate–protein interactions at interfaces: synthesis of
thiolactosyl glycolipids and design of a working model for surface
plasmon resonance
Peter Critchley,* M. Nicolas Willand, Atvinder K. Rullay and David H. G. Crout
Department of Chemistry, University of Warwick, Coventry, UK CV4 7AL
Received 29th October 2002, Accepted 7th February 2003
First published as an Advance Article on the web 24th February 2003
Thiolactosyl lipids designed for carbohydrate-protein binding studies have been synthesised. One representative was
selected for binding studies with a plant lectin RCA120, the agglutinin from Ricinus communis. The interactions were
measured quantitatively in real time using a BIAcore surface plasmon resonance instrument. Removal of much of
the galactose from the thiolactosyl lipid in situ with β-galactosidase showed that the lectin binding was highly specific.
A dissociation constant KD = 8.77 × 10Ϫ8 M was measured for 1-{2-[2-(2-[β--galactopyranosyl-(1
4)-1-thio-β--
glucopyranosyl]ethoxy)ethoxy]ethoxy}octadecane 30 which is four orders of magnitude greater than that determined
for binding to lactose in solution. A concentration of lactose of >80 mM was required to block the lectin binding to
thiolactosyl lipid in a neomembrane.
comparison of polyvalent and monovalent interactions can
be made. In most polyvalent systems, the number of ligand–
Introduction
A primary event in many biological processes involved in
cell–cell recognition/adhesion is the specific attachment of
biological polymers (normally proteins) to glycolipids or glyco-
proteins frequently located in cell membranes.1,2
receptor interactions is unknown and therefore the magnitude
of the cooperativity factor α cannot be calculated.5 The co-
operativity term modifies the valency term of the equation
quantifying the interaction. However, examples of such
cooperative effects are rare.5
The specific recognition of saccharides by lectins has been
the subject of many studies despite the fact that the function of
many lectins (particularly those from plants) is still unknown.3,4
Many factors determine whether specific molecular recog-
nition can occur at an interface, and whether the amount of
binding is sufficient to trigger a relevant biological cascade such
as mast cell degranulation, cell division, or differentiation.
The binding of saccharides to lectins in solution in a 1:1
complex, although specific, reveals fairly low binding constants
of less than 104 MϪ1.2 However, when the same saccharides
are present in liposomes or when they are immobilised at an
interface in a way that makes possible significant interaction,
binding constants of 106 MϪ1 or higher are observed.1,5 Part of
the explanation for this difference in the observed binding con-
stant is the fact that at an interface it is possible to form clusters
of ligands, provided that the ligand molecules are not rigidly
bound by very short covalent linkers to the surface. Such
clustering makes possible multivalent associations with proteins
that have more than one binding site. The strength of binding is
thereby increased significantly.6
Cooperative effects can, in principle, occur in a polyvalent
interaction because the free energy of the interaction between a
ligand and a lectin may be positive, neutral or negative com-
pared with the free energy of monovalent interaction between
a ligand and a lectin. The term cooperativity was introduced
into the biochemical literature to explain the behaviour of
multi-subunit or multi-chain enzyme assemblies, particularly
allosteric enzymes. Alteration in the conformation of one chain
resulting from the binding of substrate, allosteric activator or
inhibitor may change the conformation of a neighbouring
chain and so influence the binding of a second molecule of
substrate. The cooperativity effect can be seen as a sigmoidal
plot of rate against substrate concentration for positive co-
A further complication when examining binding events at
interfaces is that non-specific binding of proteins occurs at
apolar sites.7 Recent work has shown that the introduction of
polyethyleneglycol (PEG) derivatives or other polar molecules
at an interface can significantly reduce the non-specific binding
and promote specific interactions.8
The presentation of a ligand held at an interface can influ-
ence the strength of interaction with an appropriate lectin on a
scale from 0–100%.9 Rigid conformations are likely to militate
against ligand–lectin interactions and result in enthalpically
diminished binding. While the concept is easy to describe it is
difficult to quantitate. Examples of antibody binding to antigens
are usually monovalent although potentially the two arms of
the antibody have potential for bivalent binding. Entropies of
translation, rotation and solvation all play a role. Ligands held
at surfaces have already sacrificed some degrees of freedom
that would have contributed to the free energy of interaction in
solution. At an interface one is dealing essentially with two-
dimensional chemistry and the entropic cost of association is
less than for the interaction of the same species diffusing in
three dimensions.
There are many unexplained examples in the biological
literature of the effect of different lipid structures, that are
either part of a glycolipid or are part of the phospholipid
microenvironment, changing the degree of interaction between
the glycoside (bound to the lipid) and a lectin10,11 or enzyme.12
Described below are syntheses of model compounds
designed for use in experiments to investigate some of the
factors that may control the specific molecular interactions
between carbohydrates immobilised at a neomembrane inter-
face and proteins in solution. Surface plasmon resonance (SPR)
was used to follow binding events at a synthetic membrane
using a BIAcore 2000 instrument. A recent review describes
many advantages of using SPR to measure the interaction
between saccharide ligands presented in a membrane environ-
ment and protein analytes.13 The experiments described below
focused on neoglycolipid 30 (see Scheme 2 later). Experiments
with the other glycolipids prepared will be described elsewhere.
operativity and
a reduced hyperbolic plot for negative
cooperativity. Such plots compare with the standard hyperbolic
plot to Vmax expected from the Michaelis–Menten equation.
Whilst the concept of cooperativity has been applied to a
number of specific molecular interactions in chemistry it is dif-
ficult to quantitate for polyvalent systems unless a quantitative
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 2 8 – 9 3 8
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
928

View Article Online
Scheme 1 Reagents and conditions: i, MeONa, MeOH; ii, NaOH, Bu4NHSO4, NaI; iii, Na–NH3.
Scheme 2 Reagents and conditions: i, NaOAc, Ac2O; ii, BF3OEt2; iii, NaOMe, MeOH.
with
a monochloro end-group were converted into the
Results and discussion
corresponding benzyl thiol derivatives 3 and 4 respectively.15
Further reaction with C10, C12 and C18 1-bromoalkanes in the
presence alkali, sodium iodide and a phase transfer catalyst
gave the corresponding alkylated compounds 5–7 and 8–10
respectively.16 Debenzylation with sodium and liquid ammonia
liberated the thiol group for condensation with octaacetyl-
lactose 18 prepared from lactose 17 by a standard procedure
(Scheme 2).17 The condensation reaction was promoted by
boron trifluoride etherate in dichloromethane to yield the
protected glycolipids 19–24.18 Treatment of each of the per-
acetylated compounds with sodium methoxide/methanol gave
the neoglycolipids 25–30 which were purified separately by flash
chromatography on silica (Scheme 2).
A series of novel thiolactosyl glycolipids were synthesised
with PEG linkers between the disaccharide and an alkyl chain.
Alkyl chains of various lengths were incorporated in order
to determine the effect of chain length on the behaviour of
the neoglycolipid in artificial membranes. The PEG linkers
were introduced to study the effect of PEG chain length on
recognition of the carbohydrate by lectins. It was expected that
the nature of the PEG linker would affect the way in which the
carbohydrate was presented on the surface of the membrane
and hence the interaction with lectins. Lactose was chosen
because it is readily available, it is a common core of glycolipids,
and it binds readily to lectins such as the 120 kDa agglutinin
from Ricinus communis that recognizes β-linked galactosyl
residues. It was intended to use enzymatic reactions to change
the ligand on the surface of the chip. The thiolactosyl linkage
was selected because it is more resistant to hydrolysis by glyco-
sidases than O-linked glycoside and therefore resistant to
adventitious glycosidases in the enzyme preparation to be used
to modify the chip surface. A number of thiolactosyl derivatives
were synthesised with different degrees of hydrophilic character
in the spacer group between the thiolactosyl residue and the
hydrocarbon chain. A similar series of compounds have been
synthesized with a normal O-glycosidic linkage.14
Biological results
It was planned to attach the thioglycolipid on the surface of the
SPR chip to avoid the potential difficulty of aggregation of the
glycolipid in solution which would have made the interpretation
of data (that requires an accurate concentration term for the
analyte) very difficult. It is also known that the best results with
the BIAcore system are obtained with analytes of more than 10
kDa molecular weight. Earlier work showed that glycopeptides
could be attached to a CM5 sensor chip (that has a surface of
oxidised dextran) by amide links19 and also to the surface
of a chip coated with streptavidin by forming a biotinylated-
hydrazide of the saccharides that then bound to the strept-
avidin.20 Both of these methods fix the ligands tightly at the
surface of a sensor chip preventing diffusion and therefore
Chemical synthesis
The synthesis of the linkers was carried out as shown in Scheme
1. In the first stage α,ω-bifunctional PEG dimer 1 and trimer 2
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 2 8 – 9 3 8
929

View Article Online
Table 1 Relative rates of associationa of β-galactosidase with neo-
glycolipid 30
multivalent binding. In the planned experiments, lateral diffu-
sion within the lipid layer would be possible which would make
possible the formation of lipid rafts. Such phase separation
would be encouraged by using alkyl chain lengths in the
neoglycolipid different from those of the components of the
membrane.21,22 In the experiments reported here this condition
was approached by incorporating the neoglycolipid in a
Concentration of enzyme/nM
Relative rate of association
25
33
50
75
0.022
0.019
0.018
0.017
dipalmitoylphosphatidylcholine neomembrane.
A
similar
approach was used to follow the binding of glycopeptide
antibiotics such as vancomycin to synthetic lipopeptides held
on an HPA chip surface.23
a The relative rates of association calculated from the linear fit would
normally rise with concentration. The decrease is due to the loss of
ligand through enzyme attack.
The absorption of phospholipid and phospholipid/glycolipid
mixtures onto the alkane thiol surface of an HPA sensor chip
was accomplished by flowing a suspension of lipid, homo-
genized into buffer at 40 ЊC, slowly over the chip which was held
at 35 ЊC. The surface can be divided into four separate tracks
separated by micropneumatically-controlled valves within
the BIAcore apparatus. Previous reports have described the
formation of liposomes or small unilamellar vesicles that
were then exposed to the chip surface.23 It was found that the
formation of carefully sized liposomal or vesicular structures
was unnecessary. Deposition followed a stepped curve with
periods of association followed by a sudden loss of signal as
assemblies of lipids peeled away from the surface. An uptake
corresponding to 3,000–4,000 response units (RU) over a
period of three hours was common. At the end of each phase of
deposition the flow rate was increased to 100 µl minϪ1 to remove
loosely adherent lipid. This was followed by a pulse of 20 mM
sodium hydroxide to stabilize the neomembrane formed.
Neomembranes could be used for several weeks with repeated
regenerations with sodium hydroxide to remove any bound
lectin. Moreover, when the response on a surface finally
decayed to a non-reproducible level, a new neomembrane of the
same or a different glycolipid/phospholipid could be deposited
on the thioalkyl layer of the chip surface by raising the tem-
perature to 35 ЊC and following the procedure described in the
experimental section to deposit a new neomembrane. It was
observed that 15 mol% of glycolipid in a 0.5 mM suspension of
dipalmitoylphosphatidylcholine in buffer presented a suitable
concentration of glycolipid and that a 12.5 µM solution of
RCA120 in the same buffer gave highly reproducible results,
although under carefully controlled conditions a 10 mol%
ligand concentration and a much lower concentrations of
RCA120 were used for kinetic studies. Attempts to use dis-
tearoylphosphatidylcholine or dioleoloylphosphatidylcholine
as the carrier failed to give good deposition/response behaviour
of the glycolipid analyte.
Fig.
1
Sensorgram showing the binding of Ricinus communus
agglutinin (RCA120) to a neomembrane of 15 mol% of compound 30 in
dipalmitoylphosphatidylcholine. The upper curve (A) represents the
binding to RCA120 before treatment with β-galactosidase and the lower
curve (B) shows the binding after exposure of the ligand to the enzyme
(7.7 µM) for 45 min at a flow rate of 5 µL minϪ1
.
but the enzymatic digestion with β-galactosidase from jack
bean was carried out in solution and then starting material
and the enzyme-digested asialofetuin glycopeptide were
immobilised on the chip to measure the RCA120 binding.20
An attempt was made to follow the binding of the enzyme
to the surface at very low concentrations (25–75 nM) where the
hydrolysis reaction would be slow. The sensorgrams showed
that binding did occur but it was reduced as the concentration
of enzyme increased confirming the rapid hydrolytic reaction
(Table 1). The observation that all of the curves gave a lower
baseline and a lower response to the binding of RCA120 after
exposure to enzyme confirmed the loss of ligand and made
a full kinetic analysis difficult because the dissociation part of
the curve represented loss of the two components namely the
enzyme, by reversible dissociation, and enzymatically-cleaved
ligand. However, the potential of this approach to the study
of enzyme–substrate interactions is clear, provided only that
non-reacting substrate analogues are attached to the sensor
chip.
Kinetic measurements were carried out with 10 mol%
glycolipid 30 and dilute solutions (25–150 nM) of RCA120 to
minimise bulk effects attributable to changes in the refractive
index of the solution flowing over the chip. The set of sensor-
grams obtained (Fig. 2) showed clear rates of binding and of
dissociation. The equations used to calculate the equilibrium
association and dissociation constants have been described in
detail24–26 and are based on 1:1 Langmuir binding. In brief,
glycolipid 30 binds to the RCA120 lectin and the rate of binding
for components A and B binding to give complex AB can be
described by the equation
Albumin binds well to a hydrophobic surface such as that
presented by the alkylthiol chains on an HPA sensor chip with
no neomembrane present. There was no significant uptake of
albumin onto the chip, proving that there was good coverage
of the chip surface. Non-specific binding was not a problem
and therefore the association measured was specific. Equally,
binding of RCA120 to the diphosphatidylcholine alone (the
negative control) was negligible but was in any case subtracted
from all test runs. There was no evidence of mass transfer
problems above a flow rate of 10 µl minϪ1 so all measurements
of binding were done at 10 or 20 µl minϪ1
.
A typical binding experiment is illustrated in Fig. 1. The
upper curve A shows the response obtained with a layer of 15%
neoglycolipid 30 (Scheme 2) in a dipalmitoylphosphatidyl-
choline neomembrane. Evidence for the specificity of the inter-
action was obtained by exposing the sensor chip surface to the
β-galactosidase from Bacillus circulans. Measurement of the
binding of RCA120 both before and after exposure to an 8.6 µM
solution of enzyme for 45 min showed that there was a 57%
loss in binding after exposure to enzyme (curve B, Fig. 1) con-
firming that the RCA120 was binding to the β-galactosyl unit
of the thiolactosyl glycolipid. A similar experiment with
asialofetuin peptides on a CM5 sensor chip has been reported
where ka and kd are respectively the rate constants for associ-
ation and dissociation. In the BIAcore apparatus the glycolipid
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 2 8 – 9 3 8
930

View Article Online
where r0 = kaCRmax and
(6)
The BIAcore software evaluates the initial rate r0 and the
observed rate constant for association kobs by numerical fitting
of the Rt versus t plot, against eqn. (5). The true rate constant
for association, ka, is then obtained from a plot of kobs versus C,
the analyte (lectin in the present case) concentration (eqn. (6)).
The dissociation constant, kd, is given by the intercept on the
kobs axis by extrapolation to C = 0. In practice, the accuracies of
values of kd obtained in this way are not very good, particularly
for low values. Accordingly, as in the present work, kd is best
evaluated from the dissociation phase of the sensorgram after
the analyte solution has been replaced by the solution (buffer)
minus the analyte. Under these conditions, the dissociation
curve is given by
Fig. 2 Sensorgram showing the binding of RCA120 agglutinin to a
neomembrane of 10 mol% of compound 30 in
dipalmitoylphosphatidylcholine with a range of concentrations of
RCA120. NB The lower two concentration curves are nearly coincident.
which on integration gives
B is immobilised on the surface of the sensor chip and so the
concentration of the complex AB is identical to that of
bound lectin A. The concentration of bound lectin gives a pro-
portional response Rt.at time t. If [B]0 is the total concentration
of ligand then
(7)
where R0 is the sensor response at the beginning of the dissoci-
ation phase and Rt is the response after time t. This equation
can be rewritten as
[B] = [B]0 Ϫ [AB]
The maximum response (at saturation) Rmax is proportional
to the total ligand concentration and (Rmax Ϫ Rt) is pro-
portional to the free ligand concentration. The flow of lectin is
constant during the association phase and therefore the con-
centration of free analyte is equal to the constant concentration
of analyte passing over the sensor chip. If the concentrations of
complex and free glycolipid are expressed in terms of sensor
response the equation can be rewritten as
From this equation, the BIAcore software evaluates kd by
numerical fitting of the dissociation curve. With the values of ka
and kd to hand, the equilibrium constant K is directly evaluated.
All of these calculations assume a 1:1 interaction between
lectin and glycolipid. While there is a possibility that RCA120
agglutinin might bind to two adjacent glycolipid molecules no
evidence for this could be found. Thus plots of
(1)
where C is the concentration of injected lectin. Eqn. (1) can be
rearranged to give eqn. (2):
versus R and
(2)
This can be integrated to give eqn. (3):
versus t (eqn. (7)) corresponding to association and dissociation
phases respectively were linear.
(3)
The observed equilibrium constant for association, KA, is the
mean of values obtained from ten data sets and has a value
of 1.14 ( 0.12) × 107 MϪ1. This value is almost an order of
magnitude lower than that of that for the binding of RCA120 to
asialofetuin23 (KA=1.62 × 108 MϪ1) but almost four orders of
magnitude higher than the binding constant for free lactose to
Eqn. (3) can be rewritten as:
(4)
= 1 and Rt= 0. Thus from eqn. (2), the
ϩ
kd)t
At t = 0, eϪ(k C
initial slope is
a
RCA120 where KA=3.8 × 103 MϪ1 1
.
Inhibition experiments were carried out by adding lactose
(0–100 mM) to the RCA120 solution before it was passed over
the ligand on the chip surface. Maximum binding was observed
at zero concentration of free lactose. Binding was completely
inhibited at 80 mM lactose (Fig. 3). That such a high concen-
tration of lactose (80 mM) was required to inhibit the binding
of RCA120 to the thiolactosyl glycolipid 30 confirmed the tight
binding of the RCA120 to the thiolactosyl glycolipid observed in
the kinetic experiments.
Also, from eqn. (2), Ϫ(kaC ϩ kd is the observed rate constant
for the association step. Eqn. (4) can be rewritten as eqn. (5)
(5)
The experiments described above provide a further example
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 2 8 – 9 3 8
931

View Article Online
added dropwise. The reaction mixture was heated under reflux
overnight. The reaction was followed by TLC (CH2Cl2–MeOH,
9:1 v/v, Rf 0.71). After 19 h, the reaction mixture was poured
into a saturated aqueous solution of NaCl (500 cm3) and
product was extracted with CH2Cl2 (3 × 1 dm3). The organic
phase was dried (MgSO4), filtered and evaporated under
reduced pressure. The crude oil was purified by chromato-
graphy on silica (CH2Cl2
CH2Cl2–MeOH, 15:1 v/v) to give
the sulfide 3 (75.36 g, 88%) as a clear oil; δH(300 MHz; CDCl3)
2.47 (1 H, t, J 6.2, OH), 2.64 (2 H, t, J 6.6, 2-CH2), 3.50–3.56
(2 H, m, 4-CH2), 3.60 (2 H, t, J 6.6, 3-CH2), 3.69–3.73 (2 H,
m, 5-CH2), 3.77 (2 H, s, 1-CH2), 7.23–7.37 (5 H, m, Ph);
δC(75 MHz; CDCl3) 30.7 (2-C), 36.5 (1-C), 61.6 (5-C), 70.1
(4-C), 71.9 (3-C), 127.0 (9-C), 128.4, 128.8 (7, 7Ј, 8, 8Ј-C), 138.1
(6-C); m/z(CI) 230.1215 (M ϩ NH4ϩ. C11H20NO2S requires
230.1212), 151 (100%), 91 (17.2) and 35 (6.0).
The same synthetic pathway was used to prepare the two
series of compounds described below and therefore the detailed
synthetic method is only described for the first member of each
group of compounds. A complete NMR spectrum and mass
spectral analysis is given for each compound of the two related
series (i.e. those derived from compound 3 (series A) and those
derived from compound 4 (series B).
Fig. 3 Sensorgram showing the inhibition of binding of RCA120 to
neoglycolipid 30. The curves show binding at various concentrations of
lactose.
of the greatly increased strength of interaction between a
carbohydrate-binding protein (lectin) and its carbohydrate
receptor on a membrane surface compared with the corre-
sponding interaction in solution. As discussed earlier, other
factors also are probably important, included amongst which
are thermodynamic factors associated with entropy changes
and in torsional entropy changes in particular,27 and differences
in water activity between the bulk solution and the interface
which would affect the solvation of the interacting species.
By determination of KA at 5 ЊC temperature intervals
between 5 ЊC and 30 ЊC the standard enthalpy of association,
∆H, was determined from the van’t Hoff equation:
Series A
[2-(2-Decyloxyethoxy)ethylsulfanylmethyl]benzene
5.
Aqueous NaOH [37.7 g, 943 mmol, in 38 g H2O (50% w/w)],
tetrabutylammonium hydrogensulfate (1.28 g, 3.78 mmol) and
sodium iodide (0.71 g, 4.72 mmol) were added to a mixture
of 2-(2-benzylsulfanylethoxy)ethanol 3 (10 g, 47.15 mmol) and
1-bromodecane (12.7 cm3, 61.29 mmol). The two-phase mixture
was stirred vigorously overnight at 90 ЊC under an atmosphere
of nitrogen. The reaction was followed by TLC (hexane–
EtOAc, 3:7 v/v, Rf 0.69). After stirring overnight, the reaction
mixture was poured into water (500 cm3) and the product was
extracted with CH2Cl2 (3 × 500 cm3). The organic extracts were
combined, dried (MgSO4), filtered and evaporated under
reduced pressure to give a crude oil which was purified by silica
chromatography (hexane–EtOAc, 17:1 v/v) to give compound
5 (15.08 g, 91%) as clear oil; δH(300 MHz; CDCl3) 0.82 (3 H, t,
J 6.7, CH3), 1.20–1.35 (14 H, m, 8, 9, 10, 11, 12, 13, 14-CH2),
1.46–1.57 (2 H, m, 7-CH2), 2.56 (2 H, t, J 6.8, 2-CH2), 3.39 (2 H,
t, J 6.8, 6-CH2), 3.47–3.58 (6 H, m, 3, 4, 5-CH2), 3.70 (2 H,
s, 1-CH2), 7.13–7.27(5 H, m, Ph); δC(75 MHz; CDCl3) 14.1
(15-C), 22.6 (14-C), 26.0 (8-C), 29.3, 29.4, 29.5, 29.5, 29.6 (7, 9,
10, 11, 12-C), 31.8 (13-C), 30.4 (2-C), 36.5 (1-C), 70.0, 70.3,
The gradient of a plot of ln K versus (1/T ) based on fourteen
observations taken at 5 ЊC intervals was used to calculate ∆H.
The corresponding entropy of association was then determined
from the relationship:∆G = ∆H Ϫ T ∆S. At 290 K, ∆G = Ϫ39.8
kJ molϪ1, ∆H = Ϫ33.6 kJ molϪ1 and ∆S = 21 J molϪ1 KϪ1. These
figures are similar to those found by other workers for lectin to
carbohydrate binding using titration microcalorimetry.28,29
70.8, 71.5 (3, 4, 5, 6-C), 126.8 (19-C), 128.4, 128.8 (17, 17Ј,
ϩ
.
18, 18Ј-C), 138.4 (16-C); m/z(CI) 370.2780 (M ϩ NH4
Experimental
C21H40NO2S requires 370.2780), 280 (9%), 220 (13.1), 151
(100.0), 108 (12.3) and 91 (8.5%).
NMR spectra were obtained using a Bruker 400 MHz spectro-
meter. Chemical shifts are given as parts per million (δ).
Coupling constants are quoted in Hz. MS spectra were deter-
mined using a Waters Micromass AutoSpec mass spectrometer.
Optical rotations were measured using an Optical Activity Ltd.
Model AA1000 polarimeter at 589 nM with a path length of
2 dm. Concentrations (c) are quoted in g 1 cmϪ1. TLC analysis
was carried out on aluminium backed plates (Merck Kieselgel
60 F254). Spots were visualised by spraying with with 5%
H2SO4 and heating. Column chromatography was carried out
using Kieselgel 60F254 (Merck). RCA120 and dipalmitoylphos-
phatidylcholine were obtained from Sigma. Binding studies
were monitored, in real time, by surface plasmon resonance
using an HPA chip in a BIAcore 2000 automatic instrument.
2-(2-Decyloxyethoxy)ethanethiol 11. Liquid ammonia (150
cm3) was condensed into a flask at Ϫ78 ЊC. Small pieces
of cleaned and dried sodium (1.2 g, 52.2 mmol, were added to
obtain a permanent blue colouration. The mixture was stirred
at Ϫ78 ЊC for 10 min. A solution of [2-(2-decyloxyethoxy)-
ethylsulfanylmethyl]benzene 5 (3 g, 8.52 mmol) in anhydrous
THF (25 cm3) was added dropwise and the reaction mixture
was stirred at Ϫ78 ЊC for 1 h. The temperature was slowly
increased to Ϫ30 ЊC (1.5 h) and the reaction refluxed for 30 min.
The reaction mixture was cooled to Ϫ78 ЊC and quenched with
wet THF (75% THF in H2O, 25 cm3) until the blue colouration
disappeared. The reaction flask was warmed to room temper-
ature under nitrogen to facilitate the evaporation of ammonia.
The resulting white suspension was slowly diluted with water
(200 cm3) and the product extracted with CH2Cl2 (3 × 150 cm3).
The combined organic phase was dried (MgSO4), filtered and
evaporated under reduced pressure to give a crude yellow oil
which was purified by chromatography on silica (CH2Cl2) to
2-(2-Benzylsulfanylethoxy)ethanol 3
Benzyl mercaptan (31.4 cm3, 267.8 mmol) under an atmosphere
of nitrogen was added to a mixture of sodium methoxide
(24 g, 446.3 mmol) in anhydrous MeOH (500 cm3). To this
mixture, 2-(2-chloroethoxy)ethanol 1 (50 g, 403 mmol) was
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 2 8 – 9 3 8
932

View Article Online
give compound 11 (1.94 g, 87%) as a clear oil; TLC (CH2Cl2,
Rf 0.50); δH(300 MHz; CDCl3) 0.85 (3 H, t, J 6.0, 7.0, CH3),
1.15–1.36 (14 H, m, 7, 8, 9, 10, 11, 12, 13-CH2), 1.50–1.61
(3 H, m, SH (t, J 8.3), 6-CH2), 2.67 (2 H, dt, J 6.4, 6.6, 1-CH2),
3.43 (2 H, t, J 6.8, 5-CH2), 3.52–3.63 (6 H, m, 2, 3, 4-CH2);
δC(75 MHz; CDCl3) 14.1 (14-C), 22.6 (13-C), 24.2 (1-C), 26.0
(7-C), 29.3, 29.4, 29.5, 29.6, 29.6 (6, 8, 9, 10, 11-C), 31.8 (12-C),
69.9, 70.2, 72.8 (2, 3, 4-C), 71.5 (5-C); m/z(TOF MS ESϩ)
285.1862 (M ϩ Na. C14H30O2NaS requires 285.1864).
(CH3 × 3), 20.7 (CH3), 20.7 (CH3), 20.8 (CH3), 22.6 (19-C),
26.0 (13-C), 29.3, 29.4, 29.6 (× 4) (7, 12, 14, 15, 16, 17-C),
31.8 (18-C), 60.7 (6Ј-C), 62.1 (6-C), 66.5 (4Ј-C), 69.0 (2Ј-C),
70.4 (2-C), 70.6 (5Ј-C), 70.9 (3Ј-C), 69.9, 70.3, 71.1, 71.6 (8, 9,
10, 11-C), 73.7 (3-C), 76.2 (4-C), 76.7 (5-C), 83.5 (1-C), 101.1
(1Ј-C), 169.0 (C᎐O), 169.6 (C᎐O), 169.6 (C᎐O), 170.0 (C᎐O),
᎐
᎐
᎐
᎐
170.1 (C᎐O), 170.3 (C᎐O × 2); m/z(TOF MS ESϩ) 903.3671
᎐
᎐
(M ϩ Na. C40H64O19NaS requires 903.3660).
1-{2-[2-(ꢀ-D-Galactopyranosyl-(1 4)-1-thio-ꢀ-D-glucopyra-
nosyl)ethoxy]ethoxy}decane 25. Anhydrous MeOH (40 cm3)
was added to a flask containing 1-{2-[2-(2,3,4,6-tetra-O-acetyl-
β--galactopyranosyl-(1 4)-2,3,6-tri-O-acetyl-1-thio-β--
glucopyranosyl)ethoxy]ethoxy}decane 19 (3.42 g, 3.88 mmol)
and sodium methoxide (21 mg, 0.39 mmol) at room temper-
ature under an atmosphere of nitrogen. The reaction mixture
was stirred at room temperature and reaction was followed by
TLC (CH2Cl2–MeOH, 9:1 v/v, Rf 0.06). A precipitate started
to form after about 1 h. The reaction mixture was left to stir
overnight at room temperature. After 15 h it was diluted with
CH2Cl2 (30 cm3) and MeOH (20 cm3) to dissolve the precipitate.
The solution was neutralised with Amberlyst 15 (pH 7.0,
pH paper), filtered and evaporated under reduced pressure to
give compound 25 (2.19 g, 96%) as a white crystalline solid;
mp 160–163 ЊC, Ϫ16.8 (c 0.44 in MeOH); δH(300 MHz;
MeOH-d 4) 0.80 (3 H, m, CH3), 1.13–1.32 (14 H, m, 13, 14, 15,
16, 17, 18, 19-CH2), 1.40–1.53 (2 H, m, 12-CH2), 2.66–2.77
(1 H, m, 7-CH2a), 2.78–2.91 (1 H, m, 7-CH2b), 3.17 (1 H, dd,
J 8.1, 8.7, 2-H), 3.31–3.87 (19 H, m, 2Ј, 3, 3Ј, 4, 4Ј, 5, 5Ј, 6a, 6aЈ,
6b, 6bЈ-H, 8, 9, 10, 11-CH2), 4.26 (1 H, d, J 7.4, 1Ј-H), 4.35
(1 H, d, J 9.8, 1-H); δC(75 MHz; MeOH-d 4) 14.5 (CH3), 23.7
(19-C), 26.9 (13-C), 30.1, 30.2, 30.4, 30.4, 30.5, 30.5 (7, 12, 14,
15, 16, 17-C), 33.0 (18-C), 62.0 (6Ј-C), 62.4 (6-C), 71.1, 71.2,
72.2, 72.4 (8, 9, 10, 11-C), 70.3, 72.5, 74.1, 74.8, 77.1, 77.8, 80.4,
80.5 (2, 3, 4, 5, 2Ј, 3Ј, 4Ј, 5Ј-C), 87.0 (1-C), 105.0 (1Ј-C); m/z-
(TOF MS ESϩ) 609.2926 (M ϩ Na. C26H50O12NaS requires
609.2921).
Acetyl 2,3,4,6-tetra-O-acetyl-ꢀ-D-galactopyranosyl-(1 4)-
2,3,6-tri-O-acetyl-ꢀ-D-glucopyranoside 18. A mixture of -
lactose 17 (50 g, 146 mmol), anhydrous sodium acetate (48 g,
584 mmol) and acetic anhydride (124 cm3, 1.31 mol) was heated
(95–100 ЊC) on an oil-bath under an atmosphere of nitrogen
until a clear solution was obtained. The mixture was then
boiled under reflux for 2 h. The reaction was followed by TLC
(toluene–EtOAc, 1:1 v/v, Rf 0.47). The reaction mixture was
poured into crushed ice (1 L) and stirred for 1 h. The precipitate
formed was filtered, washed with ice-cold water and dried under
reduced pressure over P2O5 The white/off-brown solid was re-
crystallized from EtOH to give the peracetate 18 (78.3 g, 79%)
as a white crystalline solid; mp 90–92 ЊC (from EtOH) (lit.30
91–93 ЊC, lit.,31 90–91 ЊC) [α]2D3; Ϫ4.13 (c 1 in CHCl3) (lit.,32
[α]2D2 Ϫ4.4 (c 1 in CHCl3)); δH(300 MHz; CDCl3) 1.92 (3 H, s,
CH3), 2.00 (3 H, s, CH3), 2.01 (9 H, s, CH3), 2.02 (3 H, s, CH3),
2.03 (3 H, s, CH3), 2.06 (3 H, s, CH3), 2.08 (3 H, s, CH3), 2.12
(3 H, s, CH3), 3.74 (1 H, m, 5-H), 3.80 (1 H, m, 4-H), 3.85 (1H,
m, 5Ј-H), 3.98–4.15 (3 H, m, 6a-H, 6aЈ-H, 6bЈ-H), 4.42 (1 H, dd,
J 1.7, 12.1, 6b-H), 4.44 (1 H, d, J 7.9, 1Ј-H), 4.92 (1 H, dd, J 3.4,
10.5, 3Ј-H), 5.01 (1 H, dd, J 8.3, 9.4, 2-H), 5.07 (1 H, dd, J 7.9,
10.5, 2Ј-H), 5.21 (1 H, dd, J 8.7, 9.4, 3-H), 5.33 (1 H, dd, J 0.9,
3.4, 4Ј-H), 5.64 (1 H, d, J 8.3, 1-H); δC(75 MHz; CDCl3) 20.4
(CH3), 20.5 (CH3 × 2), 20.6 (CH3 × 2), 20.7 (CH3 × 2), 20.8
(CH3), 60.8 (6Ј-C), 61.6 (6-C), 66.5 (4Ј-C), 68.9 (2Ј-C), 70.4
(2-C), 70.6 (3Ј-C), 70.9 (3-C), 73.4 (5-C), 75.6 (4-C), 91.4 (1-C),
100.9 (1Ј-C), 168.8 (C᎐O), 169.0 (C᎐O), 169.5 (C᎐O), 169.6
᎐
᎐
᎐
(C᎐O), 170.0 (C᎐O), 170.1 (C᎐O), 170.3 (C᎐O), 170.3 (C᎐O);
᎐
᎐
᎐
᎐
᎐
m/z(FAB, NBA) 701.1909 (M ϩ Naϩ. C28H38Na O19 requires
701.1905), 619 (78%), 559 (12.5), 331 (100), 289 (11.3), 229 (6.9)
and 137 (42.5).
[2-(2-Dodecyloxyethoxy)ethylsulfanylmethyl]benzene 6. An
aqueous solution containing NaOH [37.7 g, 943 mmol, in 38 g
H2O (50% w/w)], tetrabutylammonium hydrogensulfate (1.28 g,
3.78 mmol) and sodium iodide (0.71 g, 4.72 mmol) was added
to a mixture of. 2-(2-benzylsulfanylethoxy)ethanol 2 (10 g,
47.15 mmol) and 1-bromododecane (14.7 cm3, 61.29 mmol),
The reaction conditions and isolation of the crude oil were the
same as described for compound 5. The reaction was followed
by TLC (hexane–EtOAc, 10:1 v/v, Rf 0.33). The oil was purified
by chromatography on silica (hexane–EtOAc, 12:1 v/v) to give
compound 6 (16.17 g, 90%) as clear oil; δH(300 MHz; CDCl3)
0.82 (3 H, t, J 6.7, CH3), 1.15–1.35 (18 H, m, 8, 9, 10, 11, 12,
13, 14, 15, 16-CH2), 1.45–1.54 (2 H, m, 7-CH2), 2.56 (2 H, t,
J 6.8, 2-CH2), 3.39 (2 H, t, J 6.8, 6-CH2), 3.47–3.58 (6 H,
m, 3, 4, 5-CH2), 3.70 (2 H, s, 1-CH2), 7.13–7.22 (5 H, m, Ph);
δC(75 MHz; CDCl3) 14.1 (17-C), 22.6 (16-C), 26.0 (8-C),
29.3, 29.5, 29.6, 29.6 (× 4) (7, 9, 10, 11, 12, 13, 14-C), 30.4
(2-C), 31.8 (15-C), 36.5 (1-C), 70.0, 70.3, 70.8, 71.5 (3, 4, 5,
6-C), 126.8 (21-C), 128.4, 128.8 (19, 19Ј, 20, 20Ј-C), 138.4
(18-C); m/z(CI) 398.3096 (M ϩ NH4ϩ. C23H44NO2S requires
398.3093), 308 (5%), 248 (6.5), 151 (43.2), 108 (8.5), 91 (8.5) and
44 (100.0).
1-{2-[2-(2,3,4,6-Tetra-O-acetyl-ꢀ-D-galactopyranosyl-(1 4)-
2,3,6-tri-O-acetyl-1-thio-ꢀ-D-glucopyranosyl)ethoxy]ethoxy}-
decane 19. BF3OEt2 (1.0 cm3, 8.13 mmol) was added dropwise
over a period of 15 min to a solution of acetyl 2,3,4,6-tetra-
O-acetyl-β--galactopyranosyl-(1 4)-2,3,6-tri-O-acetyl-β--
glucopyranoside 18 (3.67 g, 5.42 mmol) and [2-(2-decyloxy)-
ethoxy]ethanethiol 11 (1.85 g, 7.05 mmol) in anhydrous CH2Cl2
(40 cm3) at room temperature under an atmosphere of nitrogen.
The reaction was followed by TLC (hexane–EtOAc, 1:1 v/v,
Rf 0.36). After 3 h, the reaction mixture was diluted with
CH2Cl2(200 cm3) and washed with 1 M HCl(aq) (3 × 140 cm3)
and brine (3 × 140 cm3). The organic layer was dried (MgSO4),
filtered and evaporated under reduced pressure. The residue was
purified by chromatography on silica (hexane–EtOAc, 1:1 v/v)
to give compound 19 (3.48 g, 73%) as a white foamy solid/
syrup; [α]2D6 Ϫ17.9 (c 0.44 in CHCl3), δH(300 MHz; CDCl3) 0.85
(3 H, t, J 6.4, 7.0, CH3), 1.17–1.36 (14 H, m, 13, 14, 15, 16, 17,
18, 19-CH2), 1.50–1.61 (2 H, m, 12-CH2), 1.93 (3 H, s, CH3),
2.01 (3 H, s, CH3), 2.02 (3 H, s, CH3), 2.02 (3 H, s, CH3), 2.03
(3 H, s, CH3), 2.09 (3 H, s, CH3), 2.12 (3 H, s, CH3), 2.66–2.78 (1
H, m, 7-CH2a), 2.83–2.95 (1 H, m, 7-CH2b), 3.41 (2 H, t, J 6.8,
11-CH2), 3.50–3.70 (7 H, m, 5-H, 8, 9, 10-CH2), 3.74 (1 H, dd,
J 9.2, 9.4, 4-H), 3.84 (1 H, b dd, J 6.8, 6.8, 5Ј-H), 4.00–4.15
(3 H, m, 6a-H, 6aЈ-H, 6bЈ-H), 4.40–4.48 (2 H, m, 1Ј-H (d, J 7.7),
6b-H), 4.53 (1 H, d, J 10.2, 1-H), 4.88 (1 H, dd, J 9.4, 10.2,
2-H), 4.92 (1 H, dd, J 3.4, 10.6, 3Ј-H), 5.07 (1 H, dd, J 7.9, 10.4,
2Ј-H), 5.17 (1 H, dd, J 9.0, 9.2, 3-H), 5.35 (1 H, dd, J 0.8, 3.4,
4Ј-H) δC(75 MHz; CDCl3) 14.1 (20-C), 20.5 (CH3), 20.6
2-(2-Dodecyloxyethoxy)ethanethiol 12. Liquid ammonia
(150 cm3) was condensed into a flask at Ϫ78 ЊC. Small pieces
of cleaned and dried sodium (1.2 g, 52.2 mmol) were added
until a permanent blue colouration was obtained. The mixture
was stirred at Ϫ78 ЊC for 10 min. A solution of [2-(2-dodecyloxy-
ethoxy)ethylsulfanylmethyl]benzene 6 (3 g, 7.89 mmol) in
anhydrous THF (25 cm3) was added dropwise and the reaction
mixture was stirred at Ϫ78 ЊC for 1 h. The reaction conditions
and purification were the same as described for compound 11
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 2 8 – 9 3 8
933

View Article Online
and gave a crude yellow oil which was purified by chromato-
graphy on silica (CH2Cl2) to give compound 12 (1.72 g, 75%) as
a clear oil; TLC (CH2Cl2, Rf 0.52); δH(300 MHz; CDCl3) 0.85
(3 H, t, J 6.0, 7.0, CH3), 1.15–1.38 (18 H, m, 7, 8, 9, 10, 11, 12,
13, 14, 15-CH2), 1.50–1.61 (3 H, m, SH (t, J 8.3), 6-CH2), 2.67
(2 H, dt, J 6.4, 6.6, 1-CH2), 3.43 (2 H, t, J 6.8, 5-CH2), 3.52–3.63
(6 H, m, 2, 3, 4-CH2); δC(75 MHz; CDCl3) 14.1 (16-C), 22.6
(15-C), 24.2 (1-C), 26.0 (7-C), 29.3, 29.4, 29.6 (× 5) (6, 8, 9, 10,
11, 12, 13-C), 31.8 (14-C), 69.9, 70.2, 72.8 (2, 3, 4-C), 71.5 (5-C);
m/z(TOF MS ESϩ) 313.2182 (M ϩ Na. C16H34O2NaS requires
313.2177).
[2-(2-Octadecyloxyethoxy)ethylsulfanylmethyl]benzene 7. An
aqueous solution containing NaOH [37.7 g, 943 mmol, in
38 g H2O (50% w/w)], tetrabutylammonium hydrogensulfate
(1.28 g, 3.78 mmol) and sodium iodide (0.71 g, 4.72 mmol)
was added to a mixture of 2-(2-benzylsulfanylethoxy)ethanol 3
(10 g, 47.15 mmol) and 1-bromooctadecane (20.9 cm3, 61.29
mmol). The reaction conditions and purification were the same
as those described for compound 6. Compound 7 (18.79 g,
86%) was obtained as a clear oil; δH(300 MHz; CDCl3) 0.82
(3 H, t, J 6.7, CH3), 1.15–1.35 (30 H, m, 8, 9, 10, 11, 12, 13, 14,
15, 16, 17, 18, 19, 20, 21, 22-CH2), 1.45–1.54 (2 H, m, 7-CH2),
2.56 (2 H, t, J 6.8, 2-CH2), 3.39 (2 H, t, J 6.8, 6-CH2), 3.47–3.58
(6 H, m, 3, 4, 5-CH2), 3.70 (2 H, s, 1-CH2), 7.13–7.22 (5 H, m,
Ph); δC(75 MHz; CDCl3) 14.1 (23-C), 22.6 (22-C), 26.0 (8-C),
29.3, 29.5, 29.6, 29.7 (× 9) (7, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18,
19, 20-C), 30.4 (2-C), 31.8 (21-C), 36.5 (1-C), 70.0, 70.3, 70.8,
71.5 (3, 4, 5, 6-C), 126.8 (27-C), 128.4, 128.8 (25, 25Ј, 26, 26Ј-C),
138.4 (24-C); m/z(CI) 482.4034 (M ϩ NH4ϩ. C29H56NO2S
requires 482.4032), 342 (38%), 217 (7.3), 151 (100.0), 108 (10.5)
and 91 (10.0).
1-{2-[2-(2,3,4,6-Tetra-O-acetyl-ꢀ-D-galactopyranosyl-(1 4)-
2,3,6-tri-O-acetyl-1-thio-ꢀ-D-glucopyranosyl)ethoxy]ethoxy}-
dodecane 20. BF3OEt2 (0.90 cm3, 7.24 mmol) was added drop-
wise over a period of 15 min to a solution of acetyl 2,3,4,6-
tetra-O-acetyl-β--galactopyranosyl-(1 4)-2,3,6-tri-O-acetyl-
β--glucopyranoside 18 (3.27 g, 4.82 mmol) and [2-(2-dodecyl-
oxy)ethoxy]ethanethiol 12 (1.68 g, 5.79 mmol) in anhydrous
CH2Cl2 (40 cm3) at room temperature under an atmosphere of
nitrogen. The reaction conditions and purification were the
same as described for compound 19. Compound 20 (3.48 g,
73%) was obtained as a white foamy solid/syrup; [α]2D6 Ϫ17.9
(c 0.44 in CHCl3); δH(300 MHz; CDCl3) 0.85 (3 H, t, J 6.4,
7.0, CH3), 1.15–1.346 (18 H, m, 13, 14, 15, 16, 17, 18, 19. 20,
21-CH2), 1.50–1.61 (2 H, m, 12-CH2), 1.93 (3 H, s, CH3), 2.01
(3 H, s, CH3), 2.02 (3 H, s, CH3), 2.02 (3 H, s, CH3), 2.03 (3 H,
s, CH3), 2.09 (3 H, s, CH3), 2.12 (3 H, s, CH3), 2.66–2.78
(1 H, m, 7-CH2a), 2.83–2.95 (1 H, m, 7-CH2b), 3.41 (2 H, t, J 6.8,
11-CH2), 3.50–3.70 (7 H, m, 5-H, 8, 9, 10-CH2), 3.74 (1 H, dd,
J 9.2, 9.4, 4-H), 3.84 (1 H, b dd, J 6.8, 6.8, 5Ј-H), 4.00–4.15
(3 H, m, 6a-H, 6aЈ-H, 6bЈ-H), 4.40–4.48 (2 H, m, 1Ј-H (d, J 7.7),
6b-H), 4.53 (1 H, d, J 10.2, 1-H), 4.88 (1 H, dd, J 9.4, 10.2,
2-H), 4.92 (1 H, dd, J 3.4, 10.6, 3Ј-H), 5.07 (1 H, dd, J 7.9, 10.4,
2Ј-H), 5.17 (1 H, dd, J 9.0, 9.2, 3-H), 5.35 (1 H, dd, J 0.8, 3.4,
4Ј-H) δC(75 MHz; CDCl3) 14.1 (22-C), 20.5 (CH3), 20.6 (CH3
× 3), 20.7 (CH3), 20.7 (CH3), 20.8 (CH3), 22.6 (21-C), 26.0
(13-C), 29.3, 29.4, 29.6 (× 4) (7, 12, 14, 15, 16, 17, 18, 19-C),
31.8 (20-C), 60.7 (6Ј-C), 62.1 (6-C), 66.5 (4Ј-C), 69.0 (2Ј-C), 70.4
(2-C), 70.6 (5Ј-C), 70.9 (3Ј-C), 69.9, 70.3, 71.1, 71.6 (8, 9, 10,
11-C), 73.7 (3-C), 76.2 (4-C), 76.7 (5-C), 83.5 (1-C), 101.1
[2-(2-Octadecyloxy)ethoxy]ethanethiol 13. Liquid ammonia
(150 cm3) was condensed into a flask at Ϫ78 ЊC. Small pieces of
cleaned dried sodium (1.2 g, 52.2 mmol) were added until a
permanent blue colouration was obtained. The mixture was
stirred at Ϫ78 ЊC for 10 min. A solution of [2-(2-octadecyloxy-
ethoxy)ethylsulfanylmethyl]benzene 7 (3 g, 6.46 mmol) in
anhydrous THF (25 cm3) was added dropwise and the reaction
mixture was stirred at Ϫ78 ЊC for 1 h. The reaction conditions
and purification were the same as those described for com-
pound 11. Compound, 13, (1.80 g, 74%) was obtained as a
white waxy solid; TLC (CH2Cl2, Rf 0.58); δH(300 MHz; CDCl3)
0.85 (3 H, t, J 6.0, 7.0, CH3), 1.15–1.38 (30 H, m, 7, 8, 9, 10, 11,
12, 13, 14, 15, 16, 17, 18, 19, 20, 21-CH2), 1.50–1.61 (3 H, m,
SH (t, J 8.3), 6-CH2), 2.67 (2 H, dt, J 6.4, 6.6, 1-CH2), 3.43 (2 H,
t, J 6.8, 5-CH2), 3.52–3.63 (6 H, m, 2, 3, 4-CH2); δC(75 MHz;
CDCl3) 14.1 (22-C), 22.6 (21-C), 24.2 (1-C), 26.0 (7-C), 29.3,
29.5, 29.6, 29.7 (× 10) (6, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18,
19-C), 31.8 (20-C), 69.9, 70.2, 72.8 (2, 3, 4-C), 71.5 (5-C);
m/z(TOF MS ESϩ) 397.3118 (M ϩ Na. C22H46O2NaS requires
397.3116).
(1Ј-C), 169.0 (C᎐O), 169.6 (C᎐O), 1696 (C᎐O), 170.0 (C᎐O),
᎐
᎐
᎐
᎐
170.1 (C᎐O), 170.3 (C᎐O × 2); m/z(TOF MS ESϩ) 931.3981
(M ϩ Na. C42H68O19NaS requires 931.3973).
1-{2-[2-(2,3,4,6-Tetra-O-acetyl-ꢀ-D-galactopyranosyl-(1 4)-
2,3,6-tri-O-acetyl-1-thio-ꢀ-D-glucopyranosyl)ethoxy]ethoxy}-
octadecane 21. BF3OEt2 (0.70 cm3, 5.68 mmol) was added
dropwise over a period of 15 min to a solution of acetyl 2,3,4,6-
tetra-O-acetyl-β--galactopyranosyl-(1 4)-2,3,6-tri-O-acetyl-
β--glucopyranoside 18 (2.57 g, 3.78 mmol) and [2-(2-octa-
decyloxy)ethoxy]ethanethiol 13 (1.7 g, 4.54 mmol) in anhydrous
CH2Cl2 (40 cm3) at room temperature under an atmosphere of
nitrogen. The reaction was followed by TLC (hexane–EtOAc,
1:1 v/v, Rf 0.40). Reaction conditions and purification were the
same as described for compound 19. Compound 21 (2.95 g,
78%) was obtained as a white foamy solid; [α]2D6 Ϫ17.9 (c 0.44 in
CHCl3); δH(300 MHz; CDCl3) 0.85 (3 H, t, J 6.4, 7.0, CH3),
1.15–1.34 (30 H, m, 13, 14, 15, 16, 17, 18, 19. 20, 21, 22, 23, 24,
25, 26, 27-CH2), 1.50–1.61 (2 H, m, 12-CH2), 1.93 (3 H, s, CH3),
2.01 (3 H, s, CH3), 2.02 (3 H, s, CH3), 2.02 (3 H, s, CH3), 2.03
(3 H, s, CH3), 2.09 (3 H, s, CH3), 2.12 (3 H, s, CH3), 2.66–2.78
(1 H, m, 7-CH2a), 2.83–2.95 (1 H, m, 7-CH2b), 3.41 (2 H, t, J 6.8,
11-CH2), 3.50–3.70 (7 H, m, 5-H, 8, 9, 10-CH2), 3.74 (1 H, dd,
J 9.2, 9.4, 4-H), 3.84 (1 H, b dd, J 6.8, 6.8, 5Ј-H), 4.00–4.15
(3 H, m, 6a-H, 6aЈ-H, 6bЈ-H), 4.40–4.48 (2 H, m, 1Ј-H (d, J 7.7),
6b-H), 4.53 (1 H, d, J 10.2, 1-H), 4.88 (1 H, dd, J 9.4, 10.2,
2-H), 4.92 (1 H, dd, J 3.4, 10.6, 3Ј-H), 5.07 (1 H, dd, J 7.9, 10.4,
2Ј-H), 5.17 (1 H, dd, J 9.0, 9.2, 3-H), 5.35 (1 H, dd, J 0.8, 3.4,
4Ј-H) δC(75 MHz; CDCl3) 14.1 (28-C), 20.5 (CH3), 20.6 (CH3
× 3), 20.7 (CH3), 20.7 (CH3), 20.8 (CH3), 22.6 (27-C), 26.0
(13-C), 29.3, 29.5, 29.6, 29.7 (× 11) (7, 12, 14, 15, 16, 17, 18, 19,
20, 21, 22, 23, 24, 25-C), 31.8 (26-C), 60.7 (6Ј-C), 62.1 (6-C),
᎐
᎐
1-{2-[2-(ꢀ-D-Galactopyranosyl-(1 4)-1-thio-ꢀ-D-glucopyra-
nosyl)ethoxy]ethoxy}dodecane 26. Anhydrous MeOH (50 cm3)
was added to a flask containing 1-{2-[2-(2,3,4,6-tetra-O-acetyl-
β--galactopyranosyl-(1 4)-2,3,6-tri-O-acetyl-1-thio-β--
glucopyranosyl)ethoxy]ethoxy}dodecane 20 (3.1 g, 3.41 mmol)
and sodium methoxide (19 mg, 0.34 mmol) at room temper-
ature under an atmosphere of nitrogen. The reaction mixture
was stirred at room temperature for 18 h. The reaction was
followed by TLC (CH2Cl2–MeOH, 9:1 v/v, Rf 0.10). The reac-
tion conditions and purification were the same as those
described for compound 25. Compound 26 (2.08 g, 99%)
was obtained as a white amorphous solid; mp 158–160 ЊC;
[α]2D6 Ϫ19.0 (c 0.42 in pyridine); δH(300 MHz; MeOH-d 4-CDCl3,
2:1 v/v) 0.81 (3 H, t, J 6.4, 7.0, CH3), 1.13–1.32 (18 H, m, 13, 14,
15, 16, 17, 18, 19, 20, 21-CH2), 1.40–1.53 (2 H, m, 12-CH2),
2.66–2.77 (1 H, m, 7-CH2a), 2.78–2.91 (1 H, m, 7-CH2b), 3.17
(1 H, dd, J 8.1, 8.7, 2-H), 3.31–3.87 (19 H, m, 2Ј, 3, 3Ј, 4, 4Ј, 5,
5Ј, 6a, 6aЈ, 6b, 6bЈ-H, 8, 9, 10, 11-CH2), 4.26 (1 H, d, J 7.4,
1Ј-H), 4.35 (1 H, d, J 9.8, 1-H); δC(75 MHz; MeOH-d 4-CDCl3,
2:1 v/v) 14.5 (22-C), 23.7 (21-C), 26.9 (13-C), 30.0, 30.1, 30.3,
30.3, 30.4 (× 4) (7, 12, 14, 15, 16, 17, 18, 19-C), 33.0 (20-C), 62.0
(6Ј-C), 62.4 (6-C), 71.1, 71.2, 72.2, 72.4 (8, 9, 10, 11-C), 70.3,
72.5, 74.1, 74.8, 77.1, 77.8, 80.4, 80.5 (2, 3, 4, 5, 2Ј, 3Ј, 4Ј, 5Ј-C),
87.0 (1-C), 105.0 (1Ј-C); m/z(TOF MS ESϩ) 637.3220 (M ϩ
Na. C28H54O12NaS requires 637.3234).
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 2 8 – 9 3 8
934

View Article Online
66.5 (4Ј-C), 69.0 (2Ј-C), 70.4 (2-C), 70.6 (5Ј-C), 70.9 (3Ј-C),
J 6.8, 2-CH2), 3.36 (2 H, t, J 6.8, 8-CH2), 3.47–3.58 (10 H, m, 3,
69.9, 70.3, 71.1, 71.6 (8, 9, 10, 11-C), 73.7 (3-C), 76.2 (4-C), 76.7
4, 5, 6, 7-CH2), 3.69 (2 H, s, 1-CH2), 7.13–7.27 (5 H, m, Ph);
δC(75 MHz; CDCl3) 14.0 (17-C), 22.6 (16-C), 26.0 (10-C), 29.3,
29.4, 29.5, 29.6 (× 2) (9, 11, 12,13, 14-C), 30.5 (2-C), 31.8
(15-C), 36.5 (1-C), 70.0, 70.2, 70.5, 70.6, 70.8, 71.5 (3, 4, 5, 6, 7,
8-C), 126.9 (21-C), 128.4, 128.9 (19, 19Ј, 20, 20Ј-C), 138.3
(18-C); m/z(CI) 414.3046 (M ϩ NH4ϩ. C23H44NO3S requires
414.3042), 386 (13.0%), 326 (100.0), 292 (20.0), 264 (64.5), 236
(35.0), 151 (50.5), 108 (34.0) and 91 (7.0).
(5-C), 83.5 (1-C), 101.1 (1Ј-C), 169.0 (C᎐O), 169.6 (C᎐O), 169.6
᎐
᎐
(C᎐O), 170.0 (C᎐O), 170.1 (C᎐O), 170.3 (C᎐O × 2); m/z(TOF
᎐
᎐
᎐
᎐
MS ESϩ) 1015.4904 (M ϩ Na. C48H80O19NaS requires
1015.4912).
1-{2-[2-(ꢀ-D-Galactopyranosyl-(1 4)-1-thio-ꢀ-D-glucopyra-
nosyl)ethoxy]ethoxy}octadecane 27. Anhydrous MeOH (60
cm3) was added with stirring to a mixture of 1-{2-[2-(2,3,4,6-
tetra-O-acetyl-β--galactopyranosyl-(1 4)-2,3,6-tri-O-acetyl-
1-thio-β--glucopyranosyl)ethoxy]ethoxy}octadecane 21 (2.9 g,
2.92 mmol) and sodium methoxide (16 mg, 0.29 mmol) at room
temperature under an atmosphere of nitrogen. Reaction con-
ditions and purification was the same as that described for
compound 25. The reaction was followed by TLC (CH2Cl2–
MeOH, 9:1 v/v, Rf 0.14). Compound 27 (1.91 g, 94%) was
obtained as a white amorphous solid; mp 162–165 ЊC; [α]2D6
Ϫ16.0 (c 0.42 in pyridine); δH(300 MHz; MeOH-d 4-CDCl3,
2:1 v/v) 0.81 (3 H, t, J 6.4, 7.0, CH3), 1.13–1.32 (30 H, m, 13, 14,
15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27-CH2), 1.40–1.53
(2 H, m, 12-CH2), 2.66–2.77 (1 H, m, 7-CH2a), 2.78–2.91 (1 H,
m, 7-CH2b), 3.17 (1 H, dd, J 8.1, 8.7, 2-H), 3.31–3.87 (19 H, m,
2Ј, 3, 3Ј, 4, 4Ј, 5, 5Ј, 6a, 6aЈ, 6b, 6bЈ-H, 8, 9, 10, 11-CH2), 4.26
(1 H, d, J 7.4, 1Ј-H), 4.35 (1 H, d, J 9.8, 1-H); δC(75 MHz;
MeOH-d 4-CDCl3, 2:1 v/v) 14.5 (28-C), 23.7 (27-C), 26.9 (13-C),
30.0, 30.1, 30.3, 30.3, 30.4 (× 4) (7, 12, 14, 15, 16, 17, 18, 19, 20,
21, 22, 23, 24, 25-C), 33.0 (26-C), 62.0 (6Ј-C), 62.4 (6-C), 71.1,
71.2, 72.2, 72.4 (8, 9, 10, 11-C), 70.3, 72.5, 74.1, 74.8, 77.1, 77.8,
80.4, 80.5 (2, 3, 4, 5, 2Ј, 3Ј, 4Ј, 5Ј-C), 87.0 (1-C), 105.0 (1Ј-C);
m/z(TOF MS ESϩ) 721.4164 (M ϩ Na. C34H66O12NaS requires
721.4173).
2-[2-(2-Decyloxyethoxy)ethoxy]ethanethiol 14
Liquid ammonia (150 cm3) was condensed into a flask at
Ϫ78 ЊC. Small pieces of cleaned and dried sodium (1.2 g,
52.2 mmol) were added until a permanent blue colouration was
obtained. The mixture was stirred at Ϫ78 ЊC for 10 min.
A solution of {2-[2-(2-decyloxyethoxy)ethoxy]ethylsulfanyl-
methyl}benzene 8 (3 g, 7.57 mmol) in anhydrous THF (25 cm3)
was added via a syringe and the reaction mixture was stirred at
Ϫ78 ЊC for 1 h. The reaction conditions and purification were
the same as those described for compound 11. The crude yellow
oil was purified by chromatography on silica (CH2Cl2–EtOAc,
30:1 v/v) to give compound 14 (1.71 g, 74%) as a clear oil; TLC
(CH2Cl2–EtOAc, 30:1 v/v, Rf 0.35); δH(300 MHz; CDCl3) 0.85
(3 H, t, J 6.4, 7.0, CH3), 1.15–1.38 (14 H, m, 9, 10, 11, 12, 13, 14,
15-CH2), 1.50–1.61 (3 H, m, SH(t, J 8.3), 8-CH2), 2.67 (2 H, dt,
J 6.4, 6.6, 1-CH2), 3.43 (2 H, t, J 6.8, 7-CH2), 3.50–3.65 (10 H,
m, 2, 3, 4, 5, 6-CH2); δC(75 MHz; CDCl3) 14.1 (16-C), 22.6
(15-C), 24.2 (1-C), 26.0 (9-C), 29.3, 29.5, 29.5, 29.6 (× 2) (8,
10, 11, 12, 13-C), 31.8 (14-C), 70.0, 70.2, 70.5, 70.6, 71.5, 72.8
(2, 3, 4, 5, 6, 7-C); m/z(TOF MS ESϩ) 329.2132 (M ϩ Na.
C16H34O3NaS requires 329.2126).
1-{2-[2-(2-[2,3,4,6-Tetra-O-acetyl-ꢀ-D-galactopyranosyl-
(1 4)-2,3,6-tri-O-acetyl-1-thio-ꢀ-D-glucopyranosyl]ethoxy)-
ethoxy]ethoxy}decane 22. BF3OEt2 (1.0 cm3, 8.08 mmol) was
added dropwise over a period of 15 min to a solution of acetyl
2,3,4,6-tetra-O-acetyl-β--galactopyranosyl-(1 4)-2,3,6-tri-
O-acetyl-β--glucopyranoside 18 (3.66 g, 5.39 mmol) and 2-[2-
(2-decyloxyethoxy)ethoxy]ethanethiol 14 (1.7 g, 5.56 mmol) in
anhydrous CH2Cl2 (40 cm3) at room temperature under an
atmosphere of nitrogen. The reaction was followed by TLC
(CH2Cl2-acetone, 10:1 v/v, Rf 0.18). The reaction conditions
and purification were the same as used for compound 19. The
residue was deacetylated without further purification.
Series B
2-[2-(2-Benzylsulfanylethoxy)ethoxy]ethanol
4.
2-[2-(2-
Chloroethoxy)ethoxy]ethanol 2 (50 g, 298 mmol) was added
dropwise under nitrogen to a mixture of benzyl mercaptan
(31.4 cm3, 362.8 mmol) and sodium methoxide (24 g,
268 mmol) in anhydrous MeOH (500 cm3). The reaction
mixture was boiled under reflux overnight. The reaction was
followed by TLC (CH2Cl2–MeOH, 15:1 v/v, Rf 0.62). After
20 h, the reaction mixture was poured into a saturated aqueous
solution of NaCl (500 cm3) and the product was extracted with
CH2Cl2 (3 × 1 dm3). The organic phase was dried (MgSO4),
filtered and evaporated under reduced pressure. The crude oil
was purified by chromatography on silica (CH2Cl2
CH2Cl2–
1-{2-[2-(2-[ꢀ-D-Galactopyranosyl-(1 4)-1-thio-ꢀ-D-gluco-
pyranosyl]ethoxy)ethoxy]ethoxy}decane 28. Anhydrous MeOH
(60 cm3) was added to a mixture of 1-{2-[2-(2-[2,3,4,6-tetra-O-
acetyl-β--galactopyranosyl-(1 4)-2,3,6-tri-O-acetyl-1-thio-
β--glucopyranosyl]ethoxy)ethoxy]ethoxy}decane 22 (4.51 g,
6.65 mmol) and sodium methoxide (36 mg, 0.67 mmol) at room
temperature under an atmosphere of nitrogen. The reaction
mixture was stirred at room temperature and the progress of
the reaction was followed by TLC (CH2Cl2–MeOH, 4:1 v/v,
Rf 0.37). A precipitate started to form after about 1.5 h. The
reaction was left to stir overnight at room temperature. After
20 h the reaction mixture was diluted with CH2Cl2 (100 cm3)
and MeOH (150 cm3) to dissolve the precipitate. The solution
was neutralised with Amberlyst 15, filtered and evaporated
under reduced pressure to give a crude white solid which was
purified by chromatography on silica (CH2Cl2–MeOH, 4:1 v/v)
to give compound 28 (1.85 g, 63% over two steps) as a white
crystalline solid; mp 155–157 ЊC; [α]2D6 Ϫ17.5 (c 0.41 in MeOH);
δH(300 MHz; MeOH-d 4) 0.80 (3 H, m, CH3), 1.12–1.35 (14 H,
m, 15, 16, 17, 18, 19, 20, 21-CH2), 1.40–1.53 (2 H, m, 14-CH2),
2.66–2.7 (1 H, m, 7-CH2a), 2.78–2.91 (1 H, m, 7-CH2b), 3.17
(1 H, dd, J 8.1, 8.7, 2-H), 3.26–3.84 (23 H, m, 2Ј, 3, 3Ј, 4, 4Ј, 5,
5Ј, 6a, 6aЈ, 6b, 6bЈ-H, 8, 9, 10, 11, 12, 13-CH2), 4.25 (1 H, d,
J 7.3, 1Ј-H), 4.34 (1 H, d, J 9.8, 1-H); δC(75 MHz; MeOH-d 4)
MeOH, 15:1 v/v) to give compound 4 (68.68 g, 90%) as a clear
oil; δH(300 MHz; CDCl3) 2.51–2.59 (3 H, m, 2-CH2,-OH ),
3.50–3.61 (8 H, m, 3, 4, 5, 6-CH2), 3.62–3.68 (2 H, m, 7-CH2),
3.70 (2 H, s, 1-CH2), 7.23–7.37 (5 H, m, Ph); δC(75 MHz;
CDCl3) 30.5 (2-C), 36.5 (1-C), 61.6 (7-C), 70.2, 70.2, 70.6 (3, 4,
5-C), 72.4 (6-CH2), 127.0 (11-C), 128.4, 128.8 (9, 9Ј, 10, 10Ј-C),
138.2 (8-C); m/z 274.1478 (M ϩ NH4ϩ. C13H24NO3S requires
274.1477), 184 (34.0%), 168 (41.5), 152 (93.0), 124 (73.5) and 91
(89.0).
{2-[2-(2-Decyloxyethoxy)ethoxy]ethylsulfanylmethyl}benzene
8. Aqueous NaOH [31.2 g, 781 mmol, in 32 g H2O (50%
w/w)],tetrabutylammonium hydrogensulfate (1.06 g, 3.12
mmol) and sodium iodide (0.59 g, 3.91 mmol) were added
to a mixture of 2-[2-(2-benzylsulfanylethoxy)ethoxy]ethanol 4
(10 g, 39.05 mmol) and 1-bromodecane (10.6 cm3, 50.76 mmol).
Reaction conditions and purification were the same as
described for compound 5. The reaction was followed by TLC
(hexane–EtOAc, 5:1 v/v, Rf 0.35). The crude oil obtained was
purified by chromatography on silica (hexane–EtOAc, 5:1 v/v)
to give compound 8 (12.03 g, 78%) as clear oil; δH(300 MHz;
CDCl3) 0.80 (3 H, t, J 6.8, CH3), 1.10–1.32 (14 H, m, 10, 11, 12,
13, 14, 15, 16-CH2), 1.46–1.55 (2 H, m, 9-CH2), 2.56 (2 H, t,
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 2 8 – 9 3 8
935

View Article Online
14.5 (22-C), 23.7 (21-C), 27.2 (15-C), 30.2, 30.5, 30.6, 30.7 (× 2),
30.8 (7, 14, 16, 17, 18, 19-C), 33.0 (20-C), 62.0 (6Ј-C), 62.4
(6-C), 71.1, 71.3, 72.5 (× 2), 72.3, 72.4 (8, 9, 10, 11, 12, 13-C),
70.3, 72.5, 74.1, 74.8, 77.1, 77.8, 80.4, 80.5 (2, 3, 4, 5, 2Ј, 3Ј, 4Ј,
5Ј-C), 87.0 (1-C), 105.0 (1Ј-C); m/z(TOF MS ESϩ) 653.3188
(M ϩ Na. C28H54O13NaS requires 653.3183).
nitrogen. Reaction conditions and purification were the same
as described for compound 28. Compound 29 (1.12 g, 67% over
two steps) was obtained as a white amorphous solid; mp 147–
150 ЊC; [α]2D6 Ϫ19.1 (c 0.45 in pyridine); δH(300 MHz; MeOH-d 4)
0.85 (3 H, m, CH3), 1.20–1.45 (18 H, m, 15, 16, 17, 18, 19, 20,
21, 22, 23-CH2), 1.40–1.60 (2 H, m, 14-CH2), 2.66–2.77 (1 H, m,
7-CH2a), 2.78–2.91 (1 H, m, 7-CH2b), 3.17 (1 H, dd, J 8.1, 8.7,
2-H), 3.26–3.84 (23 H, m, 2Ј, 3, 3Ј, 4, 4Ј, 5, 5Ј, 6a, 6aЈ, 6b, 6bЈ-H,
8, 9, 10, 11, 12, 13-CH2), 4.25 (1 H, d, J 7.3, 1Ј-H), 4.34 (1 H, d,
J 9.8, 1-H); δC(75 MHz; MeOH-d 4) 14.5 (24-C), 23.7 (23-C),
27.2 (15-C), 29.9, 30.0, 30.1, 30.2, 30.3 (× 4) (7, 14, 16, 17, 18,
19, 20, 21-C), 33.0 (22-C), 62.0 (6Ј-C), 62.4 (6-C), 71.1, 71.3,
72.5 (× 2), 72.3, 72.4 (8, 9, 10, 11, 12, 13-C), 70.3, 72.5, 74.1,
74.8, 77.1, 77.8, 80.4, 80.5 (2, 3, 4, 5, 2Ј, 3Ј, 4Ј, 5Ј-C), 87.0
(1-C), 105.0 (1Ј-C); m/z(TOF MS ESϩ) 681.3494 (M ϩ Na.
C30H58O13NaS requires 681.3496).
{2-[2-(2-Dodecyloxyethoxy)ethoxy]ethylsulfanylmethyl}-
benzene 9. Aqueous NaOH [31.2 g, 781 mmol, in 32 g H2O
(50% w/w)], tetrabutylammonium hydrogensulfate (1.06 g, 3.12
mmol) and sodium iodide (0.59 g, 3.91 mmol) were added to a
mixture of 2-[2-(2-benzylsulfanylethoxy)ethoxy]ethanol 4 (10 g,
39.05 mmol) and 1-bromododecane (12.7 cm3, 50.76 mmol).
The reaction conditions and purification were the same as
described for compound 5. The reaction was followed by TLC
(hexane–EtOAc, 9:2 v/v, Rf 0.41). The crude oil was purified by
chromatography on silica (hexane–EtOAc, 6:1 v/v) to give
compound 9 (13.15 g, 80%) as a clear oil; δH(300 MHz; CDCl3)
0.80 (3 H, t, J 6.8, CH3), 1.10–1.32 (18 H, m, 10, 11, 12, 13, 14,
15, 16, 17, 18-CH2), 1.46–1.55 (2 H, m, 9-CH2), 2.56 (2 H, t,
J 6.8, 2-CH2), 3.36 (2 H, t, J 6.8, 8-CH2), 3.47–3.58 (10 H, m, 3,
4, 5, 6, 7-CH2), 3.69 (2 H, s, 1-CH2), 7.13–7.27 (5 H, m, Ph);
δC(75 MHz; CDCl3) 14.0 (19-C), 22.6 (18-C), 26.0 (10-C), 29.3,
29.4, 29.5 (× 4), 29.6 (9, 11, 12, 13, 14, 15, 16-C), 30.5 (2-C),
31.8 (17-C), 36.5 (1-C), 70.0, 70.2, 70.5, 70.6, 70.8, 71.5 (3, 4, 5,
6, 7, 8-C), 126.9 (21-C), 128.4, 128.9 (19, 19Ј, 20, 20Ј-C), 138.3
(18-C); m/z(CI) 442.3354 (M ϩ NH4ϩ. C25H48NO3S requires
442.3355), 320 (6.5), 292 (26.0), 184 (6.5), 151 (100.0), 108
(19.0) and 91 (12.5).
{2-[2-(2-Octadecyloxyethoxy)ethoxy]ethylsulfanylmethyl}-
benzene 10. Aqueous NaOH [31.2 g, 781 mmol, in 32 g H2O
(50% w/w)], tetrabutylammonium hydrogensulfate (1.06 g,
3.12 mmol) and sodium iodide (0.59 g, 3.91 mmol) were added
to a mixture of 2-[2-(2-benzylsulfanylethoxy)ethoxy]ethanol 4
(10 g, 39.05 mmol) and 1-bromooctadecane (17.3 cm3, 50.76
mmol). Reaction conditions and purification were the same
as described for compound 5. The reaction was followed by
TLC (hexane–EtOAc, 9:2 v/v, Rf 0.47). The crude oil obtained
was purified by chromatography on silica (hexane–EtOAc,
6:1 v/v) to give compound 10 (14.2 g, 72%) as a white waxy
solid; δH(300 MHz; CDCl3) 0.81 (3 H, t, J 6.8, CH3), 1.10–1.32
(30 H, m, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23,
24-CH2), 1.46–1.55 (2 H, m, 9-CH2), 2.56 (2 H, t, J 6.8, 2-CH2),
3.36 (2 H, t, J 6.8, 8-CH2), 3.47–3.58 (10 H, m, 3, 4, 5, 6, 7-
CH2), 3.69 (2 H, s, 1-CH2), 7.13–7.27 (5 H, m, Ph); δC(75 MHz;
CDCl3) 14.0 (25-C), 22.6 (24-C), 26.0 (10-C), 29.3, 29.4, 29.6,
29.7 (× 10) (9, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22-C),
30.5 (2-C), 31.8 (23-C), 36.5 (1-C), 70.0, 70.2, 70.5, 70.6, 70.8,
2-[2-(2-Dodecyloxyethoxy)ethoxy]ethanethiol 15. Liquid
ammonia (150 cm3) was condensed into a flask at Ϫ78 ЊC
under an atmosphere of nitrogen. Small pieces of clean dried
sodium (1.2 g, 52.2 mmol) were added until a permanent blue
colouration was obtained. The mixture was stirred at Ϫ78 ЊC
for 10 min. A solution of {2-[2-(2-dodecyloxyethoxy)ethoxy]-
ethylsulfanylmethyl}benzene 9 (3 g, 7.07 mmol) in anhydrous
THF (25 cm3) was added via a syringe and the reaction mixture
was stirred at Ϫ78 ЊC for 1 h. Reaction conditions and purifi-
cation were the same as those described for compound 11. The
yellow oil was purified by chromatography on silica (CH2Cl2–
EtOAc, 30:1 v/v) to give compound 15 (1.03 g, 44%) as a clear
oil; TLC (CH2Cl2–EtOAc, 30:1 v/v, Rf 0.44); δH(300 MHz;
CDCl3) 0.85 (3 H, t, J 6.4, 7.0, CH3), 1.15–1.38 (18 H, m, 9, 10,
11, 12, 13, 14, 15, 16, 17-CH2), 1.50–1.61 (3 H, m, SH(t, J 8.3),
8-CH2), 2.67 (2 H, dt, J 6.4, 6.6, 1-CH2), 3.43 (2 H, t, J 6.8,
7-CH2), 3.50–3.65 (10 H, m, 2, 3, 4, 5, 6-CH2); δC(75 MHz;
CDCl3) 14.1 (18-C), 22.6 (17-C), 24.2 (1-C), 26.0 (9-C), 29.3,
29.5, 29.6 (× 4), 29.7 (8, 10, 11, 12, 13, 14, 15-C), 31.8 (16-C),
70.0, 70.2, 70.5, 70.6, 71.5, 72.8 (2, 3, 4, 5, 6, 7-C); m/z(TOF MS
ESϩ) 357.2432 (M ϩ Na. C18H38O3NaS requires 357.2439).
71.5 (3, 4, 5, 6, 7, 8-C), 126.9 (29-C), 128.4, 128.9 (27, 27Ј,
ϩ
.
28, 28Ј-C), 138.3 (26-C); m/z(CI) 526.4298 (M ϩ NH4
C25H48NO3S requires 526.4294), 436 (7.0%), 404 (8.0), 376
(19.0), 332 (7.5), 268 (10.2), 151 (100.0), 108 (44.0) and 91
(27.0).
2-[2-(2-Octadecyloxyethoxy)ethoxy]ethanethiol 16. Liquid
ammonia (150 cm3) was condensed into a flask at Ϫ78 ЊC under
nitrogen. Small pieces of clean dried sodium (1.2 g, 52.2 mmol)
were added until a permanent blue colouration was obtained.
The mixture was stirred at Ϫ78 ЊC for 10 min. A solution of {2-
[2-(2-octadecyloxyethoxy)ethoxy]ethylsulfanylmethyl}benzene
10 (3 g, 5.90 mmol) in anhydrous THF (25 cm3) was added via
a syringe and the reaction mixture was stirred at Ϫ78 ЊC for 1 h.
Reaction conditions and purification were the same as
described for compound 15. Compound 16 (1.57 g, 67%) was
obtained as a white waxy solid; TLC (CH2Cl2–EtOAc, 30:1 v/v,
Rf 47); δH(300 MHz; CDCl3) 0.85 (3 H, t, J 6.4, 7.0, CH3), 1.20–
1.38 (30 H, m, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,20, 21, 22,
23-CH2), 1.50–1.61 (3 H, m, SH(t, J 8.3), 8-CH2), 2.67 (2 H, dt,
J 6.4, 6.6, 1-CH2), 3.43 (2 H, t, J 6.8, 7-CH2), 3.50–3.65 (10 H,
m, 2, 3, 4, 5, 6-CH2); δC(75 MHz; CDCl3) 14.1 (24-C), 22.6
(23-C), 24.2 (1-C), 26.0 (9-C), 29.4, 29.5, 29.6, 29.7 (× 10) (8, 10,
11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21-C), 31.8 (22-C), 70.0,
70.2, 70.5, 70.6, 71.5, 72.8 (2, 3, 4, 5, 6, 7-C); m/z(TOF MS
ESϩ) 441.3376 (M ϩ Na. C24H50O3NaS requires 441.3378).
1-{2-[2-(2-[2,3,4,6-Tetra-O-acetyl-ꢀ-D-galactopyranosyl-
(1 4)-2,3,6-tri-O-acetyl-1-thio-ꢀ-D-glucopyranosyl]ethoxy)-
ethoxy]ethoxy}dodecane 23. BF3OEt2 (0.46 cm3, 3.78 mmol)
was added dropwise over a period of 15 min to a solution
of acetyl 2,3,4,6-tetra-O-acetyl-β--galactopyranosyl-(1 4)-
2,3,6-tri-O-acetyl-β--glucopyranoside 8 (1.71 g, 2.52 mmol)
and 2-[2-(2-dodecyloxyethoxy)ethoxy]ethanethiol 15 (1.01 g,
3.02 mmol) in anhydrous CH2Cl2 (30 cm3) at room temperature
under an atmosphere of nitrogen. Reaction conditions and
purification were the same as described for compound 22.
1-{2-[2-(2-[ꢀ-D-Galactopyranosyl-(1 4)-1-thio-ꢀ-D-gluco-
1-{2-[2-(2-[2,3,4,6-Tetra-O-acetyl-ꢀ-D-galactopyranosyl-
(1 4)-2,3,6-tri-O-acetyl-1-thio-ꢀ-D-glucopyranosyl]ethoxy)-
ethoxy]ethoxy}octadecane 24. BF3OEt2 (0.46 cm3, 3.78 mmol)
was added dropwise over a period of 15 min to a solution
of acetyl 2,3,4,6-tetra-O-acetyl-β--galactopyranosyl-(1 4)-
2,3,6-tri-O-acetyl-β--glucopyranoside 18 (2.21 g, 3.26 mmol)
and 2-[2-(2-octadecyloxyethoxy)ethoxy]ethanethiol 15 (1.5 g,
pyranosyl]ethoxy)ethoxy]ethoxy}dodecane
29.
Anhydrous
MeOH (60 cm3) was added to a mixture of crude 1-{2-[2-(2-
[2,3,4,6-tetra-O-acetyl-β--galactopyranosyl-(1 4)-2,3,6-tri-
O-acetyl-1-thio-β--glucopyranosyl]ethoxy)ethoxy]ethoxy}do-
decane 23 (2.61 g, 3.85 mmol) and sodium methoxide (21 mg,
0.39 mmol) at room temperature under an atmosphere of
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 2 8 – 9 3 8
936

View Article Online
3.59 mmol) in anhydrous CH2Cl2 (30 cm3) at room temperature
under an atmosphere of nitrogen. The reaction was followed by
TLC (CH2Cl2–acetone, 10:1 v/v, Rf 0.47). Reaction conditions
and purification were the same as those described for com-
pound 22.
Tests for non-specific binding, a negative control, and mass
transfer effects. To test for non-specific binding channel 1 (with
phospholipid ligand only) was exposed to the analyte RCA120
(12.5 µM, the optimum concentration) at 5 µl minϪ1 for 5 min.
Channel 4 (with the highest concentration of thioglycolipid)
was exposed to a solution of albumin (25 µM) at 5 µl minϪ1 for
6 min as a negative control.
1-{2-[2-(2-[ꢀ-D-Galactopyranosyl-(1 4)-1-thio-ꢀ-D-gluco-
pyranosyl]ethoxy)ethoxy]ethoxy}octadecane 30. Anhydrous
MeOH (70 cm3) was added to a mixture of crude 1-{2-[2-(2-
[2,3,4,6-tetra-O-acetyl-β--galactopyranosyl-(1 4)-2,3,6-tri-
O-acetyl-1-thio-β--glucopyranosyl]ethoxy)ethoxy]ethoxy}o-
ctadecane 24 (3.47 g, 5.12 mmol) and sodium methoxide (28
mg, 0.51 mmol) with stirring under an atmosphere of nitrogen.
Reaction conditions and purification was the same as described
for compound 28. The reaction was followed by TLC (CH2Cl2–
MeOH, 4:1 v/v, Rf 0.40). Compound 30 (1.47 g, 61% over two
steps) was obtained as a white amorphous solid; mp 141–144
ЊC; [α]2D6 Ϫ16.9 (c 0.46 in pyridine); δH(300 MHz; MeOH-d 4)
0.85 (3 H, m, CH3), 1.20–1.45 (30 H, m, 15, 16, 17, 18, 19, 20,
21, 22, 23, 24, 25, 26, 27, 28, 29-CH2), 1.40–1.60 (2 H, m,
14-CH2), 2.66–2.77 (1 H, m, 7-CH2a), 2.78–2.91 (1 H, m, 7-
CH2b), 3.17 (1 H, dd, J 8.1, 8.7, 2-H), 3.26–3.84 (23 H, m, 2Ј, 3,
3Ј, 4, 4Ј, 5, 5Ј, 6a, 6aЈ, 6b, 6bЈ-H, 8, 9, 10, 11, 12, 13-CH2), 4.25
(1 H, d, J 7.3, 1Ј-H), 4.34 (1 H, d, J 9.8, 1-H); δC(75 MHz;
MeOH-d 4) 14.5 (30-C), 23.7 (29-C), 27.2 (15-C), 29.8, 29.9,
30.0, 30.0, 30.1, 30.2 (× 9) (7, 14, 16, 17, 18, 19, 20, 21, 22, 23,
24, 25, 26, 27-C), 33.0 (28-C), 62.0 (6Ј-C), 62.4 (6-C), 71.1, 71.3,
72.5 (× 2), 72.3, 72.4 (8, 9, 10, 11, 12, 13-C), 70.3, 72.5, 74.1,
74.8, 77.1, 77.8, 80.4, 80.5 (2, 3, 4, 5, 2Ј, 3Ј, 4Ј, 5Ј-C), 87.0
(1-C), 105.0 (1Ј-C); m/z(TOF MS ESϩ) 765.4431 (M ϩ Na.
C36H70O13NaS requires 765.4435).
Binding experiments of analyte RCA120to 15 mol% thio-
glycolipid ligand were performed at a range of flow rates
from 5 to 50 µl minϪ1 to check whether mass transfer was a
problem.
Enzymatic modification of the galactopyranoside ligands.
Channel 4 was exposed to a solution of RCA120 (12.5 µM) for
5 min to establish a positive response level before exposure to
β-galactosidase in buffer (1 mg mlϪ1) (Bacillus circulans) at a
flow rate of 5 µl minϪ1 for 45 min. A further injection of RCA120
for 5 min showed that the binding was reduced by 57%.
Kinetic measurements (binding constants and inhibition by free
lactose). The concentration of RCA120 analyte was reduced to
25, 75, 100, 125 and 150 nM to minimize bulk refractive index
changes and possible mass transport effects. Each solution was
injected into both channels and channel 1 (the negative control
of phospholipid) was subtracted from channel 2 (10 mol%
thioglycolipid) to remove any bulk refractive index changes or
possible non-specific interactions. The main thioglycolipid used
in these studies was compound 30.
Inhibition of binding of the analyte RCA120 was achieved by
adding free lactose to the analyte before passing the solution
over the thioglycolipid ligand. Solutions of 100 nm RCA120
contained 0, 25, 50, 80 and 100 µM lactose.
BIAcore experiments
Acknowledgements
Preparation of the sensor chip. An HPA sensor chip was
washed overnight at 25 ЊC with degassed, ultrafiltered 20 mM
HEPES buffer pH 7.0 (containing 90 mM Na Cl) at a flow rate
of 2 µl minϪ1. The chip was then washed with 40 mM octyl
glucoside dissolved in buffer for 7 min at a flow rate of 5 µl
minϪ1 immediately prior to deposition of the selected lipids.
Dipalmitoyl phosphatidylcholine dissolved in chloroform-
:methanol 1:1 v/v was mixed with varying amounts of the
selected thiolactosyl lipid in the same solvent to give solutions
of 10, 15 and 25 mol% of thioglycolipid. A control sample of
pure dipalmitoylphosphatidylcholine was used. The solutions
were evaporated under reduced pressure, dried for several hours
over P2O5 in a vacuum desiccator, and then suspended in
HEPES buffer by warming to 30 ЊC and vigorously mixed
(vortex mixer) to give a final concentration of phospholipid of
0.5 mM. Each of the four channels of the sensor chip was
exposed for 3 h to a different concentration of thioglycolipid
from 0 to 25 mol% at a flow rate of 2 µl minϪ1 and 35 ЊC, to
allow the lipids to impregnate the existing thioalkyl layer
attached to the gold surface by the manufacturers.
We thank Dr Jose A. Ribeiro Martins and Robert Jenkins
for advice on synthetic techniques and the BBSRC and the
Erasmus Organisation for financial support.
References
1 C. W. M. Grant and M. W. Peters, Biochim. Biophys. Acta, 1984, 779,
403–422.
2 A. Varki, Glycobiology, 1993, 3, 97–130.
3 N. Gilboa-Garber, D. Avichezer and N. C. Garber, Bacterial
lectins (ch. 21), H. D. Ward, Parasite lectins (ch. 22), H. Rudiger,
Plant lectins (ch. 23), J-P. Zanetta, Animal lectins (ch. 24), in
Glycosciences: Status and Perspectives, ed. H.-J. Gabius and
S. Gabius, Chapman and Hall, Weinheim, 1997, pp. 369-438.
4 H. Lis and N. Sharon, Chem. Rev., 1998, 98, 637–674.
5 M. Mammen, S.-K. Choi and G. M. Whitesides, Angew. Chem. Int.
Ed. Engl., 1998, 37, 2754–2794.
6 Y. C. Lee and R. T. Lee, Acc. Chem. Res., 1995, 28, 321–327.
7 G. B. Sigal, M. Mrksich and G. M. Whitesides, J. Am. Chem. Soc.,
1998, 120, 3464–3473.
8 P. Harder, M. Grunze, R. Dahint, G. M. Whitesides and
P. E. Laibinis, J. Phys. Chem. B., 1998, 102, 426–436.
9 N. Stromberg, P. G. Nyholm, I. Pascher and S. Normark, Proc. Natl.
Acad. Sci. USA, 1991, 88, 9340–9344.
The chip surface was washed with a fast flow of buffer (100 µl
minϪ1 for 5 min) to remove loosely adherent lipid, followed by
two washes with 20 mM NaOH (5 µl minϪ1 for 5 min each) and
then with buffer to stabilize the phospholipid monolayer.
10 T. Sato, T. Serizawa and Y. Okahata, Biochim. Biophys Acta-
Membranes, 1996, 1285, 14–20.
11 A. Kiarash, B. Boyd and C. A. Lingwood, J. Biol. Chem., 1994, 269,
11138–11146.
12 R. J. Stewart and J. M. Boggs, Biochemistry, 1993, 32, 5605–5614.
13 C. R. Mackenzie and T. Hirama, Methods Enzymol., 2000, 312, 205–
216.
14 F. Wilhelm, S. K. Chaterjee, B. Rattay, P. Nahn, R. Benecke and
J. Ortwein, Leibigs Ann. Chim., 1995, 9, 1673–1679.
15 N. Boden, R. J. Bushby, S. Clarkson, S. D. Evans, P. F. Knowles and
A. Marsh, Tetrahedron, 1997, 53, 10939–10952.
16 R. A. Bartsch, C. V. Cason and B. P. Czech, J. Org. Chem., 1989, 54,
857–860.
Determination of the optimum glycolipid and analyte concen-
trations. A solution of Ricinus communis agglutinin (RCA120
)
(12.5 µM) was passed over each ligand channel (5 µl minϪ1 for
5 min), followed by buffer to establish association and dissoci-
ation rates. All binding studies were measured at 25 ЊC and a
solution of 50 mM NaOH (2 × 5 µl minϪ for 3 min) followed by
buffer was used to regenerate the ligand surface.
Channel 3 (with 15 mol% thioglycolipid) gave the best
response and was exposed to varying concentrations of RCA120
(5, 12.5, 25, 50 µM) at a flow rate of 5 µl minϪ1 for 5 min.
17 T. W. Greene and P. G. M. Wuts, in Protective Groups in Organic
Synthesis, Wiley, New York, 2nd edn., 1991.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 2 8 – 9 3 8
937

View Article Online
18 M. Elofsson, S. Roy, B. Walse and J. Kihlberg, Tetrahedron Lett.,
1991, 32, 7613.
19 Y. Shinohara, H. Sota and Y. Hasegawa, in Techniques in
Glycobiology, ed. R. R. Townsend and A. T. Hotchkiss Jr., Marcel
Dekker Inc, New York, 1997, pp. 209–228.
20 Y. Shinohara, F. Kim, M. Shimizu, M. Goto, M. Tosu and
Y. Hasegawa, Eur. J. Biochem., 1994, 223, 189–194.
21 M. Hashizume, T. Sato and Y. Okahata, Chem. Lett., 1998, 399–400.
22 D. L. Lu, D. Singh, M. R. Morrow and C. W. M. Grant,
Biochemistry, 1993, 32, 290–297.
25 R. Karlsson and A. Falt, J. Immunol. Methods, 1997, 200,
121–133.
26 D. J. O. Shannessy, M. Brigham-Burke, K. K. Soneson, P. Hensley
and I. Brooks, Anal. Biochem., 1993, 212, 457–468.
27 M. Mammen, E. I. Shakhnovich and G. M. Whitesides, J. Org.
Chem., 1998, 63, 3168–3175.
28 M. C. Chervenak and E. J. Toone, J. Am. Chem. Soc., 1994, 116,
10533–10539.
29 F. P. Schwarz, K. D. Puri, R. G. Bhat and A. Surolia, J. Biol. Chem.,
1993, 268, 7668–7677.
23 M. Cooper, D. H. Williams and Y. R. Cho, Chem Commun., 1997,
1625–1626.
24 R. Karlsson, A. Michaelsson and L. Mattsson, J. Immunol.
Methods, 1991, 145, 229–240.
30 A. Isbell, F. A. Smith, E. C. Creitz, H. L. Frush, J. D. Moyer and
J. E. Stewart, J. Res. Natl. Bur. Stand (US), 1957, 57, 41–43.
31 J. Everett, J. Agric. Food Chem., 1966, 14, 47–51.
32 J. D. McCarter, J. Am. Chem. Soc., 1997, 119, 5792–5797.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 9 2 8 – 9 3 8
938