Synthesis and interfacial behavior of three homologous glycero neoglycolipids with various chain lengths

Article abstract of DOI:10.1016/S0009-3084(97)00076-5

Three neoglycolipids 1a-c derived from glycerol were synthezised and their molecular arrangements were studied at the air/water interface through Langmuir-Blodgett technique. The common structural features of these neoglycolipids are: a triethyleneglycol spacer at C-2 of glycerol, a GlcNAc hydrophilic head group at the end of the spacer, two saturated aliphatic chains at C-1 and C-3 of glycerol, linked by ether bonds. Compounds 1a-c differ only by the length of their lipid moieties. By increasing the hydrocarbon chain length from C₁₁ to C₁₆, a densely packed state was reached in the monolayer. The compression isotherms displayed an expanded state during the whole compression for compounds 1a and 1c bearing two C₁₁ or one C₁₁ and one C₁₆ chains. Compound 1b bearing two C₁₆ chains displayed a quite different interfacial behavior. The transfer of these monolayers onto a solid substrate can be obtained only with a trigger molecule, a phosphatidic acid.

Full text of DOI:10.1016/S0009-3084(97)00076-5

Chemistry and Physics of Lipids
90 (1997) 63–74
Synthesis and interfacial behavior of three homologous glycero
neoglycolipids with various chain lengths
P. Boullanger a,*, M.-R. Sancho-Camborieux a, M.-N. Bouchu a,
L. Marron-Brignone b, R.M. Morelis b, P.R. Coulet b
a Laboratoire de Chimie Organique 2 (U.M.R. 5622 du C.N.R.S.), Uni6ersite´ Claude Bernard Lyon 1,
Ecole Supe´rieure de Chimie Physique Electronique de Lyon, 43 Bd du 11 No6embre 1918, F-69622 Villeurbanne, France
b Laboratoire de Ge´nie Enzymatique (U.P.R.E.S. A 5013 du C.N.R.S.), Uni6ersite´ Claude Bernard Lyon 1,
Ecole Supe´rieure de Chimie Physique Electronique de Lyon, 43 Bd du 11 No6embre 1918, F-69622 Villeurbanne, France
Received 23 June 1997; received in revised form 9 September 1997; accepted 15 September 1997
Abstract
Three neoglycolipids 1a–c derived from glycerol were synthezised and their molecular arrangements were studied
at the air/water interface through Langmuir–Blodgett technique. The common structural features of these neoglycol-
ipids are: a triethyleneglycol spacer at C-2 of glycerol, a GlcNAc hydrophilic head group at the end of the spacer, two
saturated aliphatic chains at C-1 and C-3 of glycerol, linked by ether bonds. Compounds 1a–c differ only by the
length of their lipid moieties. By increasing the hydrocarbon chain length from C11 to C16, a densely packed state was
reached in the monolayer. The compression isotherms displayed an expanded state during the whole compression for
compounds 1a and 1c bearing two C11 or one C11 and one C16 chains. Compound 1b bearing two C16 chains displayed
a quite different interfacial behavior. The transfer of these monolayers onto a solid substrate can be obtained only
with a trigger molecule, a phosphatidic acid. © 1997 Elsevier Science Ireland Ltd.
Keywords: Synthesis; Neoglycolipid; Glucosylglyceride; Monolayers
1. Introduction
This phenomenon is strongly affected by the con-
formations of both the protein and the ligand.
When the carbohydrate lies at the surface of a
cell, the membrane supramolecular organization
can induce conformational changes with regard to
the isotropic medium (Nyholm and Pascher,
1993). The binding with a specific protein is there-
fore affected, not only because of these conforma-
Molecular recognition of carbohydrates is of
crucial interest in the understanding of major
biological events (Lee and Lee, 1995; Feizi, 1993).
* Corresponding author. Tel.: +33 472431162; fax: +33
478898914; e-mail: paul.boullanger@univ-lyon1.fr
0009-3084/97/$17.00 © 1997 Elsevier Science Ireland Ltd. All rights reserved.
PII S0009-3084(97)00076-5

64
P. Boullanger et al. / Chemistry and Physics of Lipids 90 (1997) 63–74
tional changes but also because of thermodynamic
shifts. By comparison with the isotropic medium,
the molecular recognition at interfaces most often
displays favorable enthalpy and disfavorable en-
tropy changes (Carver, 1993). In the field of our
researches devoted to saccharide recognition in
organized systems (Chierici et al., 1997; Boul-
langer, 1997), we were interested in the possible
conformational changes induced by the nature of
the hydrophobic moiety of a neoglycolipid. In a
search for a system that can, not only give in-
sights in the conformational changes, but also rely
on the latter with differences in the molecular
recognition, we have chosen the air/water mono-
layer. This supramolecular organization consti-
tutes a 2-dimensional model that can afford a
quantitative measurement of the binding, while
3-dimensional models (liposomes or vesicles) only
give qualitative informations (Charych et al.,
1996).
We describe in this paper, the synthesis and
monolayer organization of model compounds 1a–
c (Fig. 1). The latter were chosen for several
reasons: (a) their structures are close to glycoglyc-
erolipids widely distributed in the vegetal king-
dom; (b) the GlcNAc hydrophilic moiety can be
recognized by a specific lectin (wheat germ agglu-
tinin); (c) the ether linkage between the glycerol
and lipid moieties is more stable than the ester
binding found in natural products; (d) the junc-
tion of the hydrophilic moiety at C-2 of glycerol is
prone to give 3-dimensional arrangements by
keeping both hydrophobic alkyl chains apart (Is-
raelachvili et al., 1980); and (e) the triethylene
glycol spacer between the carbohydrate and lipid
moities might favor an ensuing recognition by
restoring a part of the saccharide mobility.
colipid and the mixed neoglycolipid/phospholipid
solutions were prepared in a mixture of analytical
grade 2:1 CHCl3-MeOH at 1.10−3 M. Behenic
acid was dissolved in CHCl3 at 2.10−3 M. The
solutions were stored at 4°C.
Pure water (resistivity=18.2 MV cm) was ob-
tained through a milli-Q four-cartridge purifica-
tion system (Millipore). CaF2 substrates (Sorem,
France) whose dimensions were 35×9.5×2 mm,
were cleaned before use as previously described
(Morelis et al., 1993). Glass substrates, whose
dimensions were 35×8×1 mm were silanized by
using dichlorodimethylsilane (1% in CHCl3).
Melting points were determined on a Bu¨chi
apparatus and were uncorrected. TLC analyses
were performed on aluminium sheets coated with
Silica Gel 60 F 254 (Merck). Compounds were
visualized by spraying the TLC plates with dilute
20% H2SO4 in MeOH followed by charring at
150°C for a few minutes. Column chromatogra-
phy was performed on Silica Gel Geduran Si 60
(Merck). Optical rotations were recorded on a
1
Perkin Elmer 241 polarimeter in a 1 dm cell. H
and 13C NMR spectra were recorded on a Bruker
AC-200 or AM-300 spectrometer working at 200
or 300 MHz and 50 or 75.5 MHz, respectively,
with Me4Si as the internal chemical shift refer-
ence; coupling constants (J) are given in Hz. Mass
spectra were recorded on a Fisons VG-Quatro
quadripole instrument in the ion spray mode.
Elemental analyses were performed by the ‘Labo-
ratoire Central d’Analyses du CNRS’ (Vernaison,
France).
2. Material and methods
Docosanoic acid (behenic acid, purity: 99%),
1,2-bis-O-tetradecanoyl-sn-glycerol-3-phosphate,
(dimyristoyl-
rity: 99%), 1,2-bis-O-hexadecanoyl-sn-glycerol-3-
phosphate, (dipalmitoyl-
L
-a-phosphatidic acid, DMPA, pu-
L
-a-phosphatidic acid,
DPPA, purity: 98%) were purchased from Sigma
(St Quentin Fallavier, France). The pure neogly-
Fig. 1. Structures of the glycero neoglycolipids used through-
out this paper.

P. Boullanger et al. / Chemistry and Physics of Lipids 90 (1997) 63–74
65
2.1. Synthesis
2.1.3. 1,3-Bis(hexadecyloxy)propan-2-ol 4b
The same experimental and extraction proce-
dures were applied to hexadecan-1-ol (4.58 g, 18.9
mmol). The organic extract was evaporated to
dryness and the dark oily material thus obtained
was chromatographed on silica gel with 1:5 ethyl
acetate–hexane as the eluent. Compound 4b was
obtained as a white crystalline material (2.14 g,
2.1.1. 1-Chloro-3-undecyloxypropan-2-ol 3
(RS)-Epichlorohydrin 2 (1 ml, 12.8 mmol) and
undecan-1-ol (5.3 ml, 25.5 mmol) were stirred for
6 h at 120°C in the presence of conc. sulfuric acid
(70 vl), according with the procedure described in
the literature (Novak and Reznicek, 1969). After
neutralization with BaCO3, the mixture was
purified by column chromatography on silica gel
with 1:3 ethyl acetate-hexane as the eluent, to
afford compound 3 (950 mg, 28%) as an oily
1
21%). M.p. 55°C. H NMR, same chemical shifts
(90.01 ppm) and integrals as 4a, except that of
protons at 1.26 ppm (52 H). 13C NMR, same
chemical shifts (90.1 ppm) as 4a. Anal. Calcd for
C35H72O3 (Mw 540.96) C, 77.71; H, 13.42. Found:
C, 77.52; H, 13.24.
1
liquid. H NMR (CDCl3, 200 MHz) l 3.93–4.00
(m, 1H, CHOH), 3.58–3.70 (m, 2H, CH2Cl);
3.43–3.55 (m, 4H, CH2OCH2), 2.57 (s, 1H, OH),
1.55–1.62 (m, 2H, CH2), 1.27 (m, 16H, CH2), 0.89
(t, 3H, CH3). 13C NMR (CDCl3, 50 MHz) l 71.8
and 71.2 (CH2OCH2), 70.3 (CHOH), 46.0
(CH2Cl), 32.0, 29.4–29.7, 26.1, 22.7 (CH2), 14.1
(CH3). Anal. Calcd for C14H29O2Cl (Mw 264.84)
C, 63.49; H, 11.04; Cl, 13.39. Found: C, 63.35; H,
11.03, Cl, 13.46.
2.1.4. 1-Hexadecyloxy-3-undecyloxypropan-2-ol
4c
Hexadecan-1-ol (4.58 g, 18.9 mmol) was
refluxed in toluene (15 ml) in the presence of
sodium (0.35 g, 15.2 mmol) for 5 h. Compound 3
(2.03 g, 7.67 mmol) was then added and the reflux
was maintained for 5 h, followed by a period of
14 h at room temperature. The mixture was then
poured into icy water (50 ml) and extracted with
ether (3×50 ml); the organic extracts were dried
(CaCl2), evaporated to dryness and finally chro-
matographed on silica gel with 2:5 ethyl acetate–
hexane as the eluent. Racemic compound 4c (1.12
g, 31%) was obtained as white crystalline material.
2.1.2. 1,3-Bis(undecyloxy)propan-2-ol 4a
Undecan-1-ol (50.0 ml, 0.24 mol) in THF (80
ml) was added dropwise, at room temperature
under argon, to a mixture of NaH (7.5 g, 0.31
mol) in THF (100 ml) and HMPA (8ml). The
mixture was refluxed for 1 h and the temperature
was then decreased to 0°C, before addition of
epichlorohydrin 2 (9.3 ml, 0.118 mol) in THF (40
ml). The mixture was then refluxed for 18 h, then
cooled to room temperature, poured into a satu-
rated aqueous solution of NH4Cl (250 ml) and
extracted with ethyl acetate (3×200 ml). The
organic extract was evaporated and the excess of
unreacted materials was distilled off. The fraction
boiling at 190–245°C/0.6 mmHg was then recrys-
tallized in a 1:5 ethyl acetate-hexane mixture to
leave compound 4a as a white crystalline material
1
M.p. 43°C. H NMR, same chemical shifts (9
0.02 ppm) and integrals as 4a and 4b, except that
of protons at 1.26 ppm (42 H). 13C NMR, same
chemical shifts (90.1 ppm) as 4a and 4b. Anal.
Calcd for C30H62O3 (Mw 470.83) C, 76.53; H,
13.27. Found: C, 76.80; H, 13.62.
2.1.5. 1-Chloro-10-undecyloxymethyl-3,6,9,12-
tetraoxatricosane 5a
Compound 4a (0.87 g, 2.17 mmol), tri-
ethyleneglycol dichloride (1.21 g, 6.48 mmol) and
tetrabutylammonium hydrogen sulfate (0.45 g,
1.33 mmol) were vigourously stirred at 60°C in a
12 M aqueous solution of sodium hydroxide (1.5
ml) for 28 h. The mixture was then poured into
water (20 ml) and extracted with CH2Cl2 (3×50
ml); the organic extracts were dried (Na2SO4),
evaporated and purified by column chromatogra-
phy on silica gel, using 2:5 ethyl acetate–hexane
1
(11.3 g, 24%). M.p. 38°C. H NMR (CDCl3, 200
MHz) l 3.93–3.95 (m, 1 H, CHOH), 3.38–3.53
(m, 8 H, CH2OCH2), 2.66 (d, 1 H, JOH,CH 4.1,
OH), 1.54–1.60 (m, 4 H, CH2), 1.26 (m, 32 H,
CH2), 0.88 (t, 6 H, CH3). 13C NMR (CDCl3, 50
MHz) l 71.9 and 71.7 (CH2OCH2), 69.5 (CHOH),
31.9, 29.7, 29.5, 29.4, 26.1, 22.7 (CH2), 14.1
(CH3). Anal. Calcd for C25H52O3 (Mw 400.69) C,
74.94; H, 13.08. Found: C, 74.66; H, 13.04.

66
P. Boullanger et al. / Chemistry and Physics of Lipids 90 (1997) 63–74
as the eluent. Compound 5a (478 mg, 40%) was
obtained as a colorless oil. H NMR (CDCl3, 200
aqueous solution of sodium hydroxide (0.17 ml).
After concentration under vacuum, water (4 ml)
was added and the mixture was extracted with
CH2Cl2 (3×10 ml). The combined organic ex-
tracts were washed with water (2×5 ml), dried
over Na2SO4, evaporated and purified by column
chromatography, using 3:1 ethyl acetate–hexane
as the eluent. Compound 6a (635 mg, 89%) was
1
MHz) l 3.73–3.79 (m, 4 H, CH2CH2Cl), 3.60–
3.67 (m, 9 H, CHOH and OCH2), 3.41–3.54 (m, 8
H, OCH2), 1.51–1.58 (m, 4 H, CH2), 1.26 (m, 32
H, CH2), 0.88 (t, 6 H, CH3). 13C NMR (CDCl3,
50 MHz) l 71.6, 71.4, 70.6–70.9, 69.8 (OCH2),
78.4 (CH), 32.0, 29.7, 29.5, 29.4, 26.2, 22.7 (CH2),
14.1 (CH3). Anal. Calcd for C31H63O5Cl (Mw
551.30) C, 67.54; H, 11.52; Cl, 6.43. Found: C,
67.65; H, 11.39; Cl, 6.52.
1
obtained as a colorless oil. H NMR (CDCl3, 200
MHz) l 3.70–3.79 (m, 4 H, CH2CH2OH), 3.59–
3.66 (m, 9 H, CHOH and OCH2), 3.41–3.59 (m, 8
H, OCH2), 3.16 (s, 1H, OH), 1.53–1.58 (m, 4 H,
2.1.6. 1-Chloro-10-hexadecyloxymethyl-3,6,9,12-
tetraoxaoctacosane 5b
CH2), 1.26 (m, 32 H, CH2), 0.88 (t, 6 H, CH3). 13
C
NMR (CDCl3, 50 MHz) l 78.4 (CH), 72.7, 71.6,
70.8, 70.7, 70.4, 69.6 (OCH2), 61.7 (CH2OH),
31.9, 29.7, 29.5, 29.4, 26.1, 22.7 (CH2), 14.1
(CH3). Anal. Calcd for C31H64O6 (Mw 532.85) C,
69.88; H, 12.11. Found: C, 69.59; H, 12.25.
The same experimental and extraction proce-
dures were applied to compound 4b (1.17 g, 2.17
mmol). After column chromatography, using 1:4
ethyl acetate–hexane as the eluent, compound 5b
1
(491 mg, 33%) was obtained as a colorless oil. H
NMR, same chemical shifts (90.02 ppm) and
integrals as 5a, except that of protons at 1.26 ppm
(52 H). 13C NMR, same chemical shifts (90.1
ppm) as 4a. Anal. Calcd for C41H83O5Cl (Mw
691.57) C, 71.21; H, 12.10; Cl, 5.13. Found: C,
71.19; H, 12.09; Cl, 4.74.
2.1.9. 10-Hexadecyloxymethyl-3,6,9,12-
tetraoxaoctacosan-1-ol 6b
The same experimental and extraction proce-
dures were applied to compound 5b (927 mg, 1.34
mmol). After column chromatography, using 3:1
ethyl acetate–hexane as the eluent, compound 6b
1
2.1.7. 1-Chloro-10-undecyloxymethyl-3,6,9,12-
tetraoxaoctacosane 5c
(755 mg, 84%) was separated as a colorless oil. H
NMR, same chemical shifts (90.02 ppm) and
integrals as 6a, except that of protons at 1.26 ppm
(52 H). 13C NMR, same chemical shifts (90.1
ppm) as 6a. Anal. Calcd for C41H84O6, 0.5 H2O
(Mw 682.13) C, 72.19; H, 12.56. Found: C, 72.09;
H, 12.54.
The same experimental and extraction proce-
dures were applied to compound 4c (1.02 g, 2.17
mmol). After column chromatography, using 1:4
ethyl acetate–hexane as the eluent, racemic com-
pound 5c (512 mg, 38%) was obtained as a color-
1
less oil. H NMR, same chemical shifts (90.02
ppm) and integrals as 5a and 5b, except that of
protons at 1.26 ppm (42 H). 13C NMR, same
chemical shifts (90.2 ppm) as 5a and 5b. Anal.
Calcd for C36H73O5Cl (Mw 621.43) C, 69.58 H,
11.84. Found: C, 69.97; H, 12.22.
2.1.10. 10-Undecyloxymethyl-3,6,9,12-
tetraoxaoctacosan-1-ol 6c
The same experimental and extraction proce-
dures were applied to compound 5c (832 mg, 1.34
mmol). After column chromatography, using 3:1
ethyl acetate–hexane as the eluent, racemic com-
pound 6c (695 mg, 86%) was separated as a
2.1.8. 10-Undecyloxymethyl-3,6,9,12-
tetraoxatricosan-1-ol 6a
1
Compound 5a (739 mg, 1.34 mmol), dissolved
in DMF (4 ml), was treated for 22 h at 115°C
with sodium formate (186 mg, 2.74 mmol) and
tetrabutylammonium bromide (6.5 mg, 0.02
mmol). The mixture was then cooled to room
temperature and stirred for 30 min with a 12 M
colorless oil. H NMR, same chemical shifts (9
0.02 ppm) and integrals as 6a and 6b, except that
of protons at 1.26 ppm (42 H). 13C NMR, same
chemical shifts (90.1 ppm) as 6a and 6b. Anal.
Calcd for C36H74O6 (Mw 602.99) C, 71.71; H,
12.37. Found: C, 72.02; H, 12.41.

P. Boullanger et al. / Chemistry and Physics of Lipids 90 (1997) 63–74
67
2.1.11. 10-Undecyloxymethyl-3,6,9,12-tetraoxatri-
cosyl 3,4,6-tri-O-acetyl-2-allyloxycarbonylamino-
(90.1 Hz), chemical shifts (90.02 ppm) and
integrals as 7a, except that of protons at 1.26 ppm
(52 H). 13C NMR, same chemical shifts (90.1
ppm) as 7a. Anal. Calcd for C57H105NO15 (Mw
1044.47) C, 65.55; H, 10.13; N, 1.34. Found: C,
65.68; H, 10,15; N, 1.49.
2-deoxy-i- -glucopyranoside 7a
D
A solution of compound 6a (969 mg, 1.82
mmol) and 1,3,4,6-tetra-O-acetyl-2-allyloxycar-
bonylamino-2-deoxy-i-D-glucopyranose (850 mg,
1.97 mmol) in dry alcohol-free CH2Cl2 (10 ml)
was cooled to −25°C under argon, according
with the procedure described in the literature
(Boullanger et al., 1987). Trimethylsilyl trifl-
uoromethanesulfonate (330 vl, 1.82 mmol) was
then added through a syringe and the mixture was
stirred at −25°C for 14 h. After neutralization
(triethylamine, 0.5 ml, 3.6 mmol), the mixture was
evaporated to dryness and purified by two succes-
sive column chromatographies on silica gel, using
2:1 ethyl acetate–hexane and 1:1 acetone–hexane
as the eluents, respectively. Compound 7a (1.50 g,
91%) was obtained as a colorless syrup. [h]D21
−5.1° (c 2.5, CHCl3). 1H NMR (CDCl3, 300
MHz) l 5.84–5.95 (m, 2H, NH and CHꢀCH2),
5.29 and 5.09 (2m, 2H, CHꢀCH2), 5.18 (t, 1H,
J2,3ꢀJ3,4 9.6, H3), 5.04 (dd, 1H, J4,5 9.4, H4), 4.81
(d, 1H, J1,2 8.6, H1), 4.57 (d, 2H, CH2–CHꢀCH2),
4.28 (dd, 1H, J5,6 4.7, J6,6% 12.3, H6), 4.11 (dd, 1H,
J5,6% 2.2, H6%), 3.60–3.93 (m, 10H, H2, CH glyc-
erol and CH2O), 3.40–3.60 (m, 13H, H5 and
CH2O), 2.17, 2.08, 2.02, 2.01 (4s, 12 H,
OCOCH3), 1.51–1.58 (m, 4H, CH2), 1.26 (m,
32H, CH2), 0.88 (t, 6H, CH3). 13C NMR (CDCl3,
50 MHz) 170.8, 170.5, 169.5 (OCOCH3), 156.2
2.1.13. 10-Undecyloxymethyl-3,6,9,12-tetraoxao-
ctacosyl 3,4,6-tri-O-acetyl-2-allyloxycarbonyl-
amino-2-deoxy-i- -glucopyranoside 7c
D
The same experimental and extraction proce-
dures were applied to compound 6c (1.10 g, 1.82
mmol). The purification was effected by column
chromatography on silica gel with 1:2 acetone–
hexane as the eluent. Compound 7c (diastereoiso-
meric mixture, 1.60 g, 90%) was obtained as a
1
colorless syrup. [h]2D1 −6.2° (c 1.7, CHCl3). H
NMR, same coupling constants (90.1 Hz),
chemical shifts (90.02 ppm) and integrals as 7a
and 7b, except that of protons at 1.26 ppm (42 H).
13C NMR, same chemical shifts (90.2 ppm) as
7a and 7b. Anal. Calcd for C52H95NO15 (Mw
974.33) C, 64.10; H, 9.83; N, 1.44. Found: C,
63.86; H, 9.97; N, 1.60.
2.1.14. 10-Undecyloxymethyl-3,6,9,12-tetraoxatri-
cosyl 2-amino-2-deoxy-i- -glucopyranoside 8a
D
Compound 7a (615 mg, 0.68 mmol) was
refluxed for 7 h in 2.5 N aqueous sodium hydrox-
ide (20 ml). The mixture was then partly neutral-
ized (pH 8.2) by addition of 2.5 N aqueous
hydrochloric acid, filtered and exhaustively ex-
tracted with ether. After drying (Na2SO4) and
evaporation, the crude product was purified by
column chromatography on silica gel, using 1:1
ethyl acetate–methanol as the eluent, to afford
compound 8a (392 mg, 83%) as an amorphous
(NHCOO),
l
133.0
(CH=CH2),
117.2
(CHꢀCH2), 101.9 (C1), 78.3 (CH, glycerol), 73.0
(C3), 71.7 (C5), 71.6, 71.3, 70.7, 70.6, 70.0, 68.9
(OCH2), 68.8 (C4), 65.4 (CH2–CHꢀCH2), 62.2
(C6), 56.0 (C2), 31.9, 29.7, 29.6, 29.5, 29.3, 26.1,
22.7 (CH2), 14.1 (CH3). Anal. Calcd for
C47H85NO15 (Mw 904.20) C, 62.43; H, 9.48; N,
1.55. Found: C, 62.47; H, 9.49; N, 1.56.
1
solid. [h]2D1 −14° (c 2.2, MeOH). H NMR (1:2
CDCl3–CD3OD, 300 MHz) l 4.37 (d, 1H, J1,2
8.1, H1), 3.25–4.05 (m, 27H, NH2, H3-6%, OCH2),
2.60 (m, 1H, H2), 1.54 (m, 4H, CH2), 1.29 (s,
32H, CH2), 0.90 (m, 6H, CH3). 13C NMR (1:2
CDCl3–CD3OD, 75.5 MHz) d 104.5 (C1), 79.5
(CH, glycerol), 78.5 (C5), 77.4 (C3), 71.7 (C4),
72.5, 71.8, 71.6, 71.4, 71.3, 70.6, 69.6 (OCH2),
62.7 (C6), 58.2 (C2), 33.0, 30.5–30.7, 27.2, 23.7
(CH2), 14.5 (CH3). Anal. Calcd for C37H75NO10
2.1.12. 10-Hexadecyloxymethyl-3,6,9,12-tetraox-
aoctacosyl 3,4,6-tri-O-acetyl-2-allyloxycarbonyl-
amino-2-deoxy-i- -glucopyranoside 7b
D
The same experimental, extraction and purifica-
tion procedures were applied to compound 6b
(1.23 g, 1.82 mmol). Compound 7b (1.29 g, 68%)
was obtained as a colorless syrup. [h]2D1 −4.5° (c
1
1.0, CHCl3). H NMR, same coupling constants

68
P. Boullanger et al. / Chemistry and Physics of Lipids 90 (1997) 63–74
(Mw 694.01) C, 64.04; H, 10.89; N, 2.02. Found:
C, 63.91; H, 10.93; N, 1.89.
81%) was obtained as an amorphous solid. [h]D21
−30° (c 0.7, CHCl3). 1H NMR (1:1 CDCl3–
CD3OD, 300 MHz) l 4.48 (d, 1H, J1,2 8.0, H1),
3.84–3.98 (m, 2H, H6%, 1/2 OC1%H2–aglycon),
3.59–3.77 (m, 14H, H2, H6, CH–glycerol, 1/2
OC1%H2–aglycon, OCH2–ethyleneglycol), 3.40–
3.56 (m, 9H, H3, OCH2), 3.24–3.37 (m, 2H, H4,
H5), 1.98 (s, 3H, COCH3), 1.54 (m, 4H, CH2),
1.29 (bs, 32H, CH2), 0.90 (t, 6H, CH3). 13C NMR
2.1.15. 10-Hexadecyloxymethyl-3,6,9,12-tetraox-
aoctacosyl 2-amino-2-deoxy-i-D-glucopyranoside
8b
The same experimental and extraction proce-
dures were applied to compound 7b (710 mg, 0.68
mmol). After column chromatography on silica
gel, using 2:1 ethyl acetate–methanol as the elu-
ent, compound 8b (452 mg, 80%) was obtained as
an amorphous solid. 1H NMR, same coupling
constants, chemical shifts (90.02 ppm) and inte-
grals as 7a, except that of protons at 1.29 ppm (52
H). 13C NMR same chemical shifts (90.2 ppm)
as 7a. This compound was not fully characterized
but directly used for the next step.
(1:1 CDCl3–CD3OD, 75.5 MHz)
l
173.6
(COCH3), 102.8 (C1), 79.5 (CH, glycerol), 77.9
(C5), 76.2 (C3), 72.0 (C4), 72.4, 71.8, 71.5, 70.6
(OCH2), 69.7 (OC%1H2–aglycon), 62.7 (C6), 57.3
(C2), 33.0, 30.4–30.7, 27.2, 23.6 (CH2), 23.1
(COCH3), 14.5 (CH3). Anal. Calcd for
C39H77NO11 (Mw 736.05) C, 63.64; H, 10.54; N,
1.90. Found: C, 63.42; H, 10.42; N, 1.84.
2.1.16. 10-Undecyloxymethyl-3,6,9,12-tetraoxao-
2.1.18. 10-Hexadecyloxymethyl-3,6,9,12-tetraox-
ctacosyl 2-amino-2-deoxy-i-
D
-glucopyranoside 8c
aoctacosyl 2-acetamido-2-deoxy-i-D-glucopyra-
The same experimental and extraction proce-
dures were applied to compound 7c (663 mg, 0.68
mmol). After purification by column chromatog-
raphy on silica gel, using 1:1 ethyl acetate–
methanol as the eluent, compound 8c (dia-
stereoisomeric mixture) was obtained as an amor-
noside 1b
The same experimental procedure was applied
to compound 8b (309 mg, 0.37 mmol). After the
treatment, the crude product was purified by
column chromatography on silica gel, using 2:1
ethyl acetate–MeOH as the eluent. Compound 1b
(211 mg, 65%) was obtained as an amorphous
1
phous solid (416 mg, 80%). H NMR, same cou-
1
pling constants, chemical shifts (90.02 ppm) and
integrals as 8a and 8b, except that of protons at
1.29 ppm (42 H). 13C NMR same chemical shifts
(90.2 ppm) as 8a and 8b. Anal. Calcd for
C42H85NO10 (Mw 764.15) C, 66.02; H, 11.21; N,
1.83. Found: C, 65.53; H, 10.89; N, 1.85. Mass
spectrometry, (M+H)+ 764.73, (M+Na)+
786.71.
solid. [h]2D1 −23° (c 1.0, CHCl3). H NMR, same
coupling constants, chemical shifts (90.02 ppm)
and integrals as 1a, except that of protons at 1.29
ppm (52 H). 13C NMR, same chemical shifts
(90.2 ppm) as 1a. Anal. Calcd for C49H97NO11
(Mw 876.32) C, 67.16; H, 11.16; N, 1.60. Found:
C, 66.79; H, 11.41; N, 1.79.
2.1.19. 10-Undecyloxymethyl-3,6,9,12-tetraox-
2.1.17. 10-Undecyloxymethyl-3,6,9,12-tetraoxatri-
aoctacosyl 2-acetamido-2-deoxy-i-
D-gluco-
cosyl 2-acetamido-2-deoxy-i-
1a
D-glucopyranoside
pyranoside 1c
The same experimental procedure was applied
to compound 8c (283 mg, 0.37 mmol). After the
treatment, the crude product was purified by
column chromatography on silica gel using 20:7
ethyl acetate–methanol as the eluent. Compound
1c (diastereoisomeric mixture, 188 mg, 63%) was
separated as an amorphous solid. [h]2D1 −28° (c
0.6, CHCl3). H NMR, same coupling constants,
chemical shifts (90.02 ppm) and integrals as 1a
and 1b, except that of protons at 1.29 ppm (42 H).
Compound 8a (257 mg, 0.37 mmol), dissolved
in a 1:1 CHCl3–MeOH mixture (3 ml), was
treated with acetic anhydride (60 vl, 0.63 mmol)
at room temperature for 2 h. After addition of
water (100 vl), the mixture was evaporated to
dryness and the solid thus obtained was washed
with water (100 ml) and centrifuged; the superna-
tant was removed and the same operations were
repeated three times. Compound 1a (221 mg,
1

P. Boullanger et al. / Chemistry and Physics of Lipids 90 (1997) 63–74
69
13C NMR, same chemical shifts (90.2 ppm) as
3. Results
1a and 1b. Anal. Calcd for C44H87NO11 (Mw
806.18) C, 65.55; H, 10.88; N, 1.74. Found: C,
65.33; H, 10.69; N, 1.65. (M+H)+ 806.76,
(M+Na)+ 828.68.
3.1. Synthesis
Epichlorohydrin was reacted with undecanol,
either in the presence of sodium hydride, pathway
a (Jungk et al., 1983) or in the presence of sulfuric
acid, pathway b (Fauran et al., 1964). In the first
instance (pathway a), a symmetrical disubstituted
derivative 4a was obtained as the only product
(24% yield), via the intermediate epoxide A; the
same reaction pathway led to 4b (21% yield) from
hexadecanol (Fig. 2). In the second case (pathway
b), a monosubstituted chlorohydrin 3 could be
obtained in 28% yield. The epoxide A could be
formed from 3 in alkaline medium, which in the
presence of hexadecanol, allowed the formation of
4c. Starting from epichlorohydrin, it was thus
possible to prepare 1,3-diethers of glycerol either
bearing identical alkyl chains or not (Fig. 2).
Triethylene glycol was then linked at the C-2
position of glycerol. Amongst methods used, the
best results were obtained by reacting compounds
4a–c with a 3-fold excess of triethyleneglycol
dichloride under phase transfer conditions (Am-
rani et al., 1985). The adducts 5a–c were obtained
in 33–40% yield. It should be mentioned at this
point that the modest yields reported in the first
steps of the syntheses are not mainly due to the
lack of reactivity of the reagents but mostly to
difficulties in the separations of the products from
2.2. Surface pressure measurements
Langmuir monolayer experiments were per-
formed using either a computerized KSV 3000
Langmuir–Blodgett trough (Finland) or a Model
LB-105 from Atemeta (Paris, licence CEA,
patent 8319770). Both were enclosed in a filtered
air flow cabinet to avoid dust deposition.
With the KSV 3000 trough, the surface pres-
sure was measured with a platinum Wilhelmy
plate attached to a sensitive balance. A symmet-
rical compression of the monolayer was achieved
with two moving barriers made of delrin and a
continuous compression was achieved in a single
step.
With the Atemeta trough, the Wilhelmy bal-
ance was made with filter paper, hung up to a
mylar support. A discontinuous compression was
achieved by steps of 1 mN/m with a single bar-
rier in PTFE.
2.3. Monolayer deposition conditions
The transfer was performed with the KSV
3000 trough on a pure water subphase ther-
mostated at 18°C. The hydrophilic substrates
used were CaF2 and CaF2-Beh. (CaF2-Beh. is a
CaF2 substrate coated with behenic acid LB films
in order to obtain a hydrophilic surface consti-
tuted by the carboxylic head groups; Marron-
Brignone et al., 1997). The hydrophobic
substrates used were silanized glass.
When the substrates, clamped parallel to the
barriers, were immersed in the subphase, the li-
pidic solution was spread on the water surface
and the monolayer was compressed up to the
surface pressure at which deposition was done.
Deposition of the first layer was performed after
allowing the monolayer to relax for 5 min. Dur-
ing the deposition, the surface pressure was kept
constant and the dipping rate was 5 mm/min.
1
the unreacted starting materials. The TLC and H
NMR spectra of the reaction mixtures often dis-
played crude yields in the range of 70–80%. The
hydrolysis of the chlorine atom of 5a–c was then
realized via a two-step one-pot reaction (Zahalka
and Sasson, 1986), i.e. substitution to a formate
ester (HCOONa, Bu4NBr) followed by the hy-
drolysis of the latter (aq NaOH); compounds
6a–c were obtained in 84–89% yields. The glyco-
sylation was effected by the ‘allyloxycarbonyl pro-
cedure’ (Boullanger et al., 1987); thus, compounds
6a–c were reacted with 1,3,4,6-tetra-O-acetyl-2-
allyloxycarbonylamino-2-deoxy-i-D-glucopyra-
nose (Boullanger et al., 1990) in the presence of
trimethylsilyl triflate to give the expected gly-
cosides 7a–c in 70–90% yields. After a usual
deprotection (aq NaOH) and re-N-acetylation,

70
P. Boullanger et al. / Chemistry and Physics of Lipids 90 (1997) 63–74
Fig. 2. Reaction pathway to the synthesis of compounds 1a–c.
pure compounds 1a–c were obtained in 65–75%
yields.
obtained whatever the compression mode used.
1a and 1c neoglycolipids display smooth and
continuous isotherms (Fig. 3a, c). Considering the
isotherm shape, it can be assumed that the two
monolayers remain in an expanded state during
the whole compression. This is corroborated by
the extrapolated mean molecular area values of 73
and 90 A for 1a and 1c, respectively. These
molecular areas are higher than the expected val-
ues for a neoglycolipid in a condensed state
(Boggs, 1987; Curatolo, 1987).
3.2. Interfacial beha6ior
The neoglycolipid solutions in 2:1 CHCl3–
MeOH were spread on a pure water subphase at
2091°C. The solvent was allowed to evaporate
for 10 min before the monolayer was compressed.
The 1a, 1b and 1c isotherms are shown in Fig. 3.
It is to be noted that these isotherm shapes were
2
˚

P. Boullanger et al. / Chemistry and Physics of Lipids 90 (1997) 63–74
71
tween 0.5 and 45 mN/m gave perfectly overlapped
isotherms. This reveals that the monolayer keeps
its bidimensional structure at the air/water inter-
face during the compression. Moreover, when the
temperature of the subphase was varied from 5 to
35°C the plateau was displaced towards higher
pressures. This phenomenon is characteristic of a
1b Glycolipid has a very different interfacial
behaviour as revealed by its isotherm shape (Fig.
3b). Two sets of experiments were carried out in
order to characterize the successive different states
which occurred during the compression. Com-
pression–decompression cycles performed be-
phase transition typical of
a
phospholipid
isotherm. Finally, the compressibility modulus
values (calculated at 10 and 50 mN/m) and the
mean molecular areas (extrapolated from the two
parts of the isotherm) suggest that the monolayer
exists in an expanded state below the plateau (34
2
˚
mN/m, 119 A ) and in a more condensed state
2
˚
above (250 mN/m, 54 A ).
3.2.1. Monolayer deposition
3.2.1.1. Deposition of 1c neoglycolipid monolayer.
Before deposition, the monolayer stability was
studied when the monolayer was in the most
packed state, at different surface pressures. When
the monolayer was allowed to relax, a collapse
takes place whatever the surface pressure chosen.
The monolayer instability could be responsible for
the deposition type obtained at 40 mN/m onto
hydrophilic or hydrophobic substrates: the trans-
fer ratio was close to one at the first stroke and no
transfer occurred at the next one. Considering
these results, the 1c molecules cannot be used to
build up LB films.
3.2.1.2. Deposition of 1a or 1b neoglycolipid mono-
layer. The stability of 1a and 1b monolayers was
studied at the chosen surface pressures for deposi-
tion. Unlike 1c monolayer, 1a and 1b monolayers
kept their integrity at 34 and 40 mN/m, respec-
tively.
On hydrophobic (silanized glass) or hydrophilic
(CaF2) substrates, a layer was transferred either at
the downstroke or at the upstroke, respectively.
This layer peeled off at the following stroke.
Consequently no LB films can be built up onto
these substrates. A previous study (Marron-
Brignone et al., 1997) showed that on the sub-
strate surface constituted by the head groups of
behenic acid LB films, a second layer of 1a was
transferred. So, such a coated substrate was used
Fig. 3. y-A isotherms at the air/water interface (at 20°C) of 1a
(a), 1b (b), 1c (c).

72
P. Boullanger et al. / Chemistry and Physics of Lipids 90 (1997) 63–74
Table 1
Deposition of mixed monolayers onto CaF2 substrate
Transfer ratio values
1a/DMPA 1/1
1b/DPPA 2/1
1st layer
2nd layer
3rd layer
1.12
0.10
0.32
1.03
0.78
0.78
ers. These ratios were selected since they gave the
best results with regard to monolayer deposition.
Compressibility modulus values and molecular ar-
eas indicate that the added phospholipids allow
the monolayer to be in a tighter state.
The mixed monolayers were transferred onto
CaF2 substrate at 36 mN/m for 1a and 40 mN/m
for 1b. The transfer ratio values are given in Table
1. It must be noticed that the presence of DMPA
prevented the peeling-off of the first 1a layer but
no second layer was transferred when the sub-
strate was withdrawn. At the third stroke, a weak
but actual deposition occurred; nevertheless the
nature of the transferred molecules could not be
determined because 1a molecules are very difficult
to transfer whereas DMPA monolayer is easily
transferrable. Addition of DPPA in the 1b mono-
layer led to both the retention of the first layer
with the concomitant second layer deposition and,
moreover, a third layer can be transferred.
Fig. 4. Mixed monolayers y-A isotherms. (A) y-A isotherms
of 1a (a), 1a/DMPA 1/1 (b) and DMPA (c). (B) y-A isotherms
of 1b (a), 1b/DPPA 1/2 (b) and DPPA (c).
in order to transfer the 1b monolayer. Surpris-
ingly, the first layer transferred at the upstroke
was not retained, although the interacting chemi-
cal groups were the same as those for 1a.
4. Discussion
3.2.2. Mixed monolayers
It must be stressed that 1a, 1b and 1c which are
neoglycolipids with the same saccharidic moiety
have a very different molecular area at the air/wa-
ter interface. This suggests that the molecular area
does not depend exclusively on the hydrated moi-
ety but is also due to the steric hindrance of the
hydrocarbon chains. Indeed, it must be stressed
that the molecular backbone of these neoglycol-
ipids may be unfavorable for intramolecular inter-
actions to occur since the hydrocarbon chains are
not parallel. Longer are the hydrocarbon chains,
denser is the packing of the monolayer. Therefore,
the 1a monolayer remains in an expanded state
In order to improve the efficiency of the trans-
fer, the molecular packing has to be tightened up
by addition of specific molecules known for their
transferrability. DMPA molecules were used and
mixed with 1a with regard to the fit of their
hydrocarbon chain length. For the same reason,
DPPA molecules were mixed with 1b neoglycol-
ipids. Before deposition it was compulsory to
check the miscibility of the two components and
to determine the more suitable glycolipid/phos-
pholipid ratio for building up the LB films. Fig. 4
shows the y-A isotherms obtained with 1a/
DMPA (1/1) and with 1b/DPPA (2/1) monolay-

P. Boullanger et al. / Chemistry and Physics of Lipids 90 (1997) 63–74
73
whereas the 1b one reaches a condensed state.
This is attributable to more numerous interactions
between the C16 chains than between the C11 ones.
Surprisingly, the presence of one C16 chain in 1c
molecules does not improve the molecular pack-
ing since the molecular area is the highest. On the
contrary, the difference in the hydrocarbon chain
lengths seems to provok some disorganization in
the monolayer. The five carbons at the end of the
C16 chain do not participate to additional in-
tramolecular interactions; more, they increase the
steric hindrance of the neoglycolipid hydrophobic
part.
The 1a and 1b neoglycolipids cannot be trans-
ferred with efficiency; this is not surprising since
as previously seen for the 1a monolayer, entrain-
ment of water molecules during the transfer leads
to a poor adhesion between the neoglycolipid
molecules and the substrate. This strong hydra-
tion of saccharidic moieties and the results on the
transferrability have already been reported
(Marra, 1985; Luckham et al., 1993). However, it
has been shown that a firm retention of the first
1a layer was obtained onto CaF2-Beh. substrate,
this was attributed to the good adequacy between
the networks of behenic acid and of 1a mono-
layer. In the case of 1b, the different molecular
packing suggests an unsuitable fit between the
carboxylate groups of behenic acid LB films and
the saccharidic moieties.
Whatever the monolayer packing is, the neogly-
colipids cannot establish strong interactions with
hydrophilic or hydrophobic substrates, this could
be due to the nature of the molecule itself, espe-
cially to the strong hydrophilicity of the saccha-
ridic moiety. Therefore, these neoglycolipids were
mixed with phospholipids known for their trans-
ferrability such as DMPA and DPPA; they allow
the transfer of DMPC or DPPC (of which only
one layer can be transferred, Hasmonay et al.,
1979). DMPA molecules slightly improve the 1a
glycolipid deposition; the phosphatidic acid
groups interact with the substrate and thus, less
water molecules should be dragged during the
lipidic monolayer transfer. Nevertheless, hydro-
phobic chains are too short to allow the efficient
hydrophobic interactions between the transferred
layer and the floating monolayer. Therefore, a
second layer cannot be transferred with efficiency.
Several layers are transferred from the DPPA/1b
monolayer. In addition of the previously effects
seen with DMPA, the DPPA should also allow
strong hydrophobic interactions to set up between
the two first layers transferred leading to their
retention.
5. Conclusion
With the view of performing molecular recogni-
tion based on N-acetylglucosamine groups, we
used the Langmuir–Blodgett technique for ob-
taining organized glycolipid molecular assemblies.
Thus, after checking the interfacial behavior of
the three neoglycolipids, it was compulsory to
transfer the monolayer onto solid substrates with
the carbohydrate groups laid at the external sur-
face.
With the hydrophilic part constituted by the
N-acetylglucosamine linked to the ethylene glycol
arm, the densely packed state of the monolayer
was directly related to the length of the hydrocar-
bon chain. Nevertheless, this solid state was not
sufficient to allow the building-up of LB films. It
was necessary to use a trigger molecule to ensure
a good adhesion of the layer on the solid sub-
strate. With 1a or 1b mixed with DMPA or
DPPA, respectively, it was possible to obtain a
surface exhibiting N-acetylglucosamine. These re-
sults allow to consider a further molecular recog-
nition study with these glycolipidic LB films.
Acknowledgements
The research was partly supported by CNRS-
ULTIMATECH and Region Rhoˆne-Alpes grants.
The authors are indebted to B. Fenet (Universite´
Lyon I) for recording HSQC gradient TOCSY
decoupled and phased 13C NMR.
References
Amrani, Y., Lafont, D., Sinou, D., 1985. Chiral polyoxa-1,2-
diphosphines. Synthesis and applications in assymetric hy-
drogenation. J. Mol. Catal. 32, 333–341.

74
P. Boullanger et al. / Chemistry and Physics of Lipids 90 (1997) 63–74
Boullanger, P., Banoub, J., Descotes, G., 1987. N-allyloxycar-
bonyl derivatives of -glucosamine as promotors of 1,2-
principles of membrane organization. Quat. Rev. Biophys.
13, 121–200.
D
trans-glucosylation in Koenigs–Knorr reactions and in
Lewis acid catalyzed condensations. Can. J. Chem. 65,
1343–1348.
Jungk, S.J., Moore, J.A., Gandour, R.D., 1983. Efficient
synthesis of C-Pivot Lariat Ethers. 5-(Alkoxymethyl)-
1,4,7,10,13,16-hexaoxacyclooctadecanes. J. Org. Chem. 48,
1116–1120.
Lee, Y.C., Lee, R.T., 1995. Carbohydrate-protein interactions:
basis of glycobiology. Accounts Chem. Res. 28, 321–327.
Luckham, P., Wood, J., Froggatt, S., Swart, R., 1993. The
surface properties of gangliosides. I-Monolayer properties.
J. Colloid Interface Sci. 156, 164–172.
Marra, J., 1985. Controlled deposition of lipid monolayers and
bilayers onto mica and direct force measurements between
galactolipid bilayers in aqueous solutions. J. Colloid Inter-
face Sci. 107, 446–458.
Marron-Brignone, L., Morelis, R.M., Coulet, P.R., 1997. Syn-
thetic glycolipids monolayer and related Langmuir-Blod-
gett films. Importance of the amphipathic balance. J.
Colloid Interface Sci. 191, 349–556.
Morelis, R.M., Girard-Egrot, A.P., Coulet, P.R., 1993. Depen-
dence of Langmuir–Blodgett films quality on fatty acid
monolayer integrity. I. Nucleation crystal growth avoid-
ance in the monolayer through the optimized compression
procedure. Langmuir 9, 3101–3106.
Novak, J., Reznicek, J., 1969. Gas-chromatographic identifica-
tion of 3-chloro-2-hydroxypropyl ethers of aliphatic pri-
mary alcohols. J. Chromatogr. 43, 437–443.
Boullanger, P., Jouineau, M., Bouammali, B., Lafont, D.,
Descotes, G., 1990. The use of N-alkoxycarbonyl deriva-
tives of 2-amino-2-deoxy-D-glucose as donors in glycosyla-
tion reactions. Carbohydr. Res. 202, 151–164.
Boullanger, P., 1997. Amphiphilic carbohydrates as a tool for
molecular recognition in organized systems. In: Thiem, J.,
Driguez, H. (Eds.), Topics in Current Chemistry, vol. 187,
Springer, Berlin, pp. 275–312.
Boggs, J.M., 1987. Lipid intermolecular hydrogen bonding:
influence on structural organization and membrane func-
tion. Biochim. Biophys. Acta 906, 353–404.
Carver, J.P., 1993. Oligosaccharides: how can flexible
molecules act as signals?. Pure Appl. Chem. 65, 763–770.
Charych, D., Cheng, Q., Reichert, A., Kuziemko, G., Stroh,
M., Nagy, J.O., Spevak, W., Stevens, R.C., 1996. A ‘litmus
test’ for molecular recognition using artificial membranes.
Chem. Biol. 3, 113–120.
Chierici, S., Boullanger, P., Marron-Brignone, L., Morelis,
R.M., Coulet, P.R., 1997. Synthesis and interfacial be-
haviour of a gemini neoglycolipid. Chem. Phys. Lipids 87,
91–101.
Curatolo, W., 1987. The physical properties of glycolipids.
Biochim. Biophys. Acta 906, 111–136.
Nyholm, P.G., Pascher, I., 1993. Orientation of the saccharide
chains of glycolipids at the membrane surface: conforma-
tional analysis of the glucose–ceramide and the glucose–
glyceride linkages using molecular mechanics (MM3).
Biochemistry 32, 1225–1234.
Zahalka, H.A., Sasson, Y., 1986. One-pot conversion of pro-
mary alkyl chlorides into alcohols, diols and ethers via
formic ester intermediates under phase-transfer conditions.
Synthesis 763–765.
Fauran, C., Miocque, M., Gauthier, J.A., 1964. Ethers-oxydes
dissyme´triques du glyce´rol en se´rie aliphatique. Me´thodes
de pre´paration. Bull. Soc. Chim. Fr. 4, 831–835.
Feizi, T., 1993. Oligosaccharides that mediate mammalian
cell–cell adhesion. Curr. Opin. Struct. Biol. 3, 701–710.
Hasmonay, H., Caillaud, M., Dupeyrat, M., 1979. Langmuir–
Blodgett multilayers of phosphatidic acid and mixed phos-
pholipids. Biochem. Biophys. Res. Commun. 89, 338–344.
Israelachvili, J.N., Marcelja, S., Horn, R.G., 1980. Physical
.
.