Synthesis and antigenicity of BBGL-2 glycolipids of Borrelia burgdorferi, the causative agent of Lyme disease

Article abstract of DOI:10.1016/j.carres.2011.04.045

Borrelia burgdorferi is the etiological agent for Lyme disease (LD), the most common vector borne disease in the United States. There is no human vaccine against LD currently available. Our approach to a vaccine is based on its surface-exposed glycolipids. One group of these glycolipids termed BBGL-2 consists of 1,2-di-O-acyl-3-O-(α-d-galactopyranosyl)-sn-glycerol congeners having palmitic, oleic, stearic, linoleic, and myristic acids. In order to delineate the immunodominant region(s) of the BBGL-2 components, we embarked on a synthetic project to provide available structurally defined, homogeneous analogs of BBGL-2 that might help identify the best vaccine candidate. The antigenicity of the synthetic glycolipids was examined by dot-blot analysis using mice sera obtained by immunization with killed B. burgdorferi cells, with native BBGL-2 in complete Freund's adjuvant, as well as sera obtained from patients with Lyme disease. We found that the presence of two acyl groups in the glycerol moiety was essential for antigenicity. At least one of these groups must be an oleoyl moiety. Neither the anomeric configuration of the galactose nor the configuration of the glycerol at C-2 was a decisive factor. Based on these findings we designed an 'unnatural' BBGL-2 analog having the structure 3-O-(β-d-galactopyranosyl)-1,2-di-O-oleoyl-dl-glycerol which is easier and less expensive to synthesize than the other BBGL-2 congeners prepared in this study. This substance proved to be antigenic and is considered a candidate vaccine for Lyme disease.

Full text of DOI:10.1016/j.carres.2011.04.045

Carbohydrate Research 346 (2011) 1551–1563
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Synthesis and antigenicity of BBGL-2 glycolipids of Borrelia burgdorferi,
the causative agent of Lyme disease
Vince Pozsgay a, , Joanna Kubler-Kielb a, Bruce Coxon a, Adriana Marques b, John B. Robbins a,
⇑
Rachel Schneerson a
a Program on Developmental and Molecular Immunity, Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, MD 20892-2423, United States
b Laboratory of Clinical Infectious Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892-2423, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Borrelia burgdorferi is the etiological agent for Lyme disease (LD), the most common vector borne disease
in the United States. There is no human vaccine against LD currently available. Our approach to a vaccine
is based on its surface-exposed glycolipids. One group of these glycolipids termed BBGL-2 consists of 1,2-
Received 2 February 2011
Received in revised form 18 April 2011
Accepted 27 April 2011
di-O-acyl-3-O-(a-D-galactopyranosyl)-sn-glycerol congeners having palmitic, oleic, stearic, linoleic, and
Available online 3 May 2011
myristic acids. In order to delineate the immunodominant region(s) of the BBGL-2 components, we
embarked on a synthetic project to provide available structurally defined, homogeneous analogs of
BBGL-2 that might help identify the best vaccine candidate. The antigenicity of the synthetic glycolipids
was examined by dot-blot analysis using mice sera obtained by immunization with killed B. burgdorferi
cells, with native BBGL-2 in complete Freund’s adjuvant, as well as sera obtained from patients with Lyme
disease. We found that the presence of two acyl groups in the glycerol moiety was essential for antige-
nicity. At least one of these groups must be an oleoyl moiety. Neither the anomeric configuration of the
galactose nor the configuration of the glycerol at C-2 was a decisive factor. Based on these findings we
Dedicated to Professor Dr. András Lipták on
the occasion of his 75th birthday
Keywords:
Antigen
Borrelia burgdorferi
Vaccine
Diacylglycerol
Dot-blot
designed an ‘unnatural’ BBGL-2 analog having the structure 3-O-(b-D-galactopyranosyl)-1,2-di-O-
oleoyl-DL-glycerol which is easier and less expensive to synthesize than the other BBGL-2 congeners pre-
pared in this study. This substance proved to be antigenic and is considered a candidate vaccine for Lyme
Glycolipid
disease.
Published by Elsevier Ltd.
1. Introduction
effective above the age of 12 years when administered with alumi-
num hydroxide as the adjuvant4 was withdrawn from the market
Lyme disease is a multisystem infection caused by the spiro-
chete Borrelia burgdorferi that can involve the skin, nervous system,
heart, and joints.1 It is transmitted by some of the Lyxodes tick spe-
cies, and it is the most common vector borne disease in the United
States with more than 304,000 LD cases have been reported to the
Centers for Disease Control and Prevention from 1995 to 2009. The
majority of cases occur in the Middle Atlantic, Northeastern, and
North Central states, reaching 30,000 confirmed cases in 2009 to-
in early 2002 after less than 5 years of use, because of inadequate
market results. Moreover, there was also the contentious issue of
the vaccine’s hypothetical potential to induce autoimmunity be-
cause of OspA’s partial homology to the human leukocyte func-
tion-associated antigen
1 in persons with certain HLA-DR
alleles5,6 but studies have shown no increase in the development
of arthritis or other adverse effects.7
Currently, there is no vaccine for human use against LD, and
prevention of the disease is limited to protective measures to avoid
tick bites. An effective vaccine to prevent human Lyme disease
would be of great benefit for populations with high risk of acquir-
ing the infection.8,9
gether with more than 8000 probable cases in the same year.2,3
licensed vaccine (LYMErix™) containing a lipidated recombinant
surface protein of B. burgdorferi designated as -OspA, although
A
L
B. burgdorferi produces neither a lipopolysaccharide nor a
capsular polysaccharide.10,11 On the other hand, immunoreactive
glycolipids were isolated from B. burgdorferi that were shown to
Abbreviations: API-ES-MS, atmospheric pressure ionization electrospray mass
spectrometry; BBGL, Borrelia burgdorferi glyco lipid; DCC, N,N-dicyclohexylcarbo-
diimide; DMAP, 4-dimethylaminopyridine; FA, fatty acid; Gal, galactosyl; Gro,
glycerol; IL, interleukin; iNKR, invariant natural killer T cell; LD, Lyme disease; Ole,
oleoyl; Pal, palmitoyl; sn, stereochemical nomenclature; Ste, stearoyl; TLC, thin
layer chromatography.
be
a
-galactosyl diacylglycerols.12 However, neither the location
of the acyl groups nor the stereochemistry of the glycerol residue
was defined in the early studies. Our laboratory reported the
isolation and structural characterization of two groups of
⇑
Corresponding author. Tel.: +1 301 295 2266; fax: +1 301 295 1435.
E-mail addresses: pozsgayv@mail.nih.gov, vipo@helix.nih.gov (V. Pozsgay).
0008-6215/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.carres.2011.04.045

1552
V. Pozsgay et al. / Carbohydrate Research 346 (2011) 1551–1563
surface-exposed glycolipids termed BBGL-1 and BBGL-2.11 Using a
variety of chemical and spectroscopic methods, BBGL-1 was iden-
NMR sensitivity of carbonyl carbons, this method requires long
acquisition times to obtain good-quality spectra. Unfortunately,
the signal-to-noise ratio of the carbonyl region is insufficient in
most reports to allow rigorous assessment of homogeneity.
We are examining aproaches to producing BBGL components
for use in a vaccine against B. burgdorferi. Growing B. burgdorferi
to produce BBGL’s in sufficient quantities for immunization exper-
iments is difficult. In addition, isolation of the glycolipids in a
homogeneous form has not been possible, raising reproducibility
concerns. To circumvent these difficulties, we are preparing BBGL
components by using synthetic chemical methods. So far, we have
reported the synthesis of the major BBGL-1 components in their
native and bioconjugatable forms28 and prepared a semisynthetic
experimental vaccine against B. burgdorferi consisting of the
BBGL-1 glycolipids covalently linked to bovine serum albumin
through an oxime linkage.29 The aim of the present work is to
delineate the immunodominant region of BBGL-2 components by
assessing their antigenicity. It is expected that such recognition
will facilitate the design of a vaccine against LD.
Here we describe experiments aimed at synthesizing the puta-
tive BBGL-2 components 1–4. In order to evaluate the biological
importance of various structural features we also describe the syn-
thesis of the diastereoisomers 5–8 that differ from the native ones
only in the stereochemistry at C-2 of the glycerol moiety, the sat-
urated analogs 9 and 10, the mono-O-acyl derivatives 11 and 12,
as well as two additional BBGL-2 analogs (13 and 14) that contain
the galactose moiety in the unnatural, beta glycosidic linkage. (Ta-
ble 1).
tified as 6-O-acyl-b-
1,2-di-O-acyl-3-O-
ings were confirmed by other workers who also reported the isola-
tion of cholesteryl b- -galactopyranoside without an acyl group at
the galactose unit as well as 6-O-acyl-b- -glucopyranosyl-choles-
D
-galactopyranosyl-cholesterol and BBGL-2 as
a
-D
-galactopyranosyl-sn-glycerol.11 These find-
D
D
terol.13,14 In BBGL-2 the most common fatty acid is palmitic acid
(1.00 mol), followed by oleic and stearic acids (approx. 0.65 and
0.25 mol, respectively). Myristic and linoleic acids were also de-
tected in 0.15 and 0.12 mol amounts. Based on the relative propor-
tions of the palmitic and oleic acids it was proposed that the major
components of BBGL-2 are 3-O-
a-
D-galactopyranosyl-1-O-oleoyl-
2-O-palmitoyl-sn-glycerol (4) and/or 3-O-
a
-D
-galactopyranosyl-2-
O-oleoyl-1-O-palmitoyl-sn-glycerol (3) whereas the other fatty
acids detected in the BBGL-2 fraction remain unaccounted for.11
Because the individual components of the BBGL-2 complex could
not be separated, no single structure could be identified. Therefore,
the presence of galactosyl-homodiacyl-glycerol derivatives and of
those incorporating the minor fatty acids cannot be excluded. Be-
cause of the lack of genes in B. burgdorferi for the synthesis or elon-
gation of fatty acids,15 its fatty acids are incorporated from the host
or from the environment. This may explain the differences in fatty
acid composition reported by different laboratories for in vitro cul-
tivated B. burgdorferi cells.11,12 In mice and rabbits, BBGL-2 elicited
antibodies that reacted with both BBGL-1 and -2, and the sera of LD
patients had a strong IgG reaction with BBGL-2.11,16 These propen-
sities make the BBGL-2 glycolipids candidates for developing diag-
nostics and vaccines against B. burgdorferi devoid of any
immunogenic proteins such as L-OspA that might have the poten-
tial to elicit autoantibodies.5
Following the synthetic studies, we disclose salient NMR fea-
tures of several of the synthetic glycolipids, and discuss the antige-
nicity of the synthesized glycolipids.
Chemical syntheses of BBGL-2 glycolipids having one oleoyl and
one palmitoyl group on their glycerol moieties have been reported,
but the published synthetic protocols lack rigorous proof of their
2. Results and discussion
2.1. Synthetic studies
homogeneity.17–20 An approach by
a-galactosylation of the com-
mercially available diglyceride 2-O-oleoyl-1-O-palmitoyl-sn-glyc-
erol (Avanti Polar Lipids) has recently been described.19,20
Examination of the 13C NMR spectrum of the commercial material,
obtained from the same source revealed that it contains up to 25%
of an accompanying compound that is most likely 1-O-oleoyl-2-O-
palmitoyl-sn-glycerol. The basis of this assumption is that while
the 1H NMR spectrum of the commercial material is fully consis-
tent with the proposed structure, the 13C NMR spectrum exhibits,
in addition to two major carbonyl carbon signals, two additional
ones that are approximately 1/4th of the major ones, together with
several similarly low-intensity peaks. Because in our experience
separation of 3 and 4 is not possible, we are tempted to assume
that the material reported in Refs. 19 and 20 as compound 3 is a
ca. 4:1 mixture of 3 and 4. An alternative approach17,19 starts with
the commercially available 3-O-benzyl-sn-glycerol which is regio-
selectively acylated at O-1 followed by a second acylation at O-2
and removal of the O-benzyl group with trichloroborane to afford
the targeted 1-O-acyl(1)-2-O-acyl(2)-sn-glycerols17,19,21 which are
then galactosylated. In applying this approach to 1,2-diacyl-sn-
glycerol precursors, care has to be taken to avoid the well-docu-
mented acyl migration that can take place under both acidic and
basic conditions.22–27 We did not observe acyl migration from
the O-1 to the O-2 position of the glycerol moieties, but we did ob-
serve migration from O-2 to O-3 over extended periods at room
temperature, or even during silica gel column chromatography.
In order to prevent undesirable acyl migration, short reaction times
and quick chromatographic procedures have been suggested.25 It
has been proposed that the regioisomeric purity of acyl glycerols
can best be determined by integration of the carbonyl signals of
the acyl moieties in high field 13C NMR spectra.22,23 Because of
the low natural abundance (1.1%) of 13C combined with the low
We explored an approach in which the introduction of the a-D-
galactopyranosyl moiety to sn-3-O of the glycerol unit precedes the
attachment of the fatty acyl groups, thereby preventing acyl migra-
tion to that site.18 This concept is shown in Scheme 1. Precursor to
the target compounds was phenyl 1-thio-b-D-galactopyrano-
side30,31 15. In preliminary trials, compound 15 was converted into
its tetra-O-benzyl derivative, which was reacted with the commer-
cially available glycerol-sn-1,2-acetonide 19 under a variety of
published thioglycoside activation protocols (not described in the
experimental part). Unexpectedly, the products contained mostly
b-interglycosidic linkages, despite the presence of the non-partici-
pating benzyl protecting group at the O-2 position of the galactose
moiety. We hypothesized that tuning down the reactivity of the
Table 1
Synthetic BBGL-2 analogs
3-O-
3-O-
3-O-
3-O-
1-O-
1-O-
1-O-
1-O-
3-O-
3-O-
3-O-
3-O-
a
a
a
a
a
a
a
a
a
a
a
a
-
-
-
-
-
-
-
-
-
-
-
-
D
D
D
D
D
D
D
D
D
D
D
D
-Galp-1,2-di-O-Pal-sn-GRO
-Galp-1,2-di-O-Ole-sn-GRO
-Galp-2-O-Ole-1-O-Pal-sn-GRO
-Galp-1-O-Ole-2-O-Pal-sn-GRO
-Galp-2,3-di-O-Pal-sn-GRO
-Galp-2,3-di-O-Ole-sn-GRO
-Galp-2-O-Ole-3-O-Pal-sn-GRO
-Galp-3-O-Ole-2-O-Pal-sn-GRO
-Galp-1,2-di-O-Ste-sn-GRO
-Galp-2-O-Pal-1-O-Ste-sn-GRO
-Galp-1-O-Pal-sn-GRO
1
2
3
4
5
6
7
8
9
10
11
12
13
14
-Galp-1-O-Ole-sn-GRO
3-O-b-
3-O-b-
D
-Galp-1,2-di-O-Pal-sn-GRO
-Galp-1,2-di-O-Ole-sn-GRO
D

V. Pozsgay et al. / Carbohydrate Research 346 (2011) 1551–1563
1553
BzO
HO
HO
BzO
O
c
CH2OH
O
CH2OBz
O
CH2OBz
O
CH2OH
O
a
b
19
O
HO
SPh
SPh
SPh
HO
BzO
B
HO
Ref. 32
BnO
16
BnO
17
BnO
HO
15
Cl
18
BzO
CH2OBz
LevO
20
21
HO
HO
O
CH2OH
O
CH2OLev
O
CH2OH
BzO
O
O
HO
LevO
HO
g
BnO
O
O
d
O
e
f
O
O
O
HO
LevO
O
BnO
Ref. 16
O
O
O
O
O
BzO
CH2OBz
O
22
23
24
O
O
BzO
BnO
LevO
CH2OLev
O
LevO
OH
2
LevO
O
OH
3
1
25
Scheme 1. Reagents and conditions: (a) 4.9 equiv BzCl, CH2Cl2, C6H5N, DMAP (cat), rt, 24 h, 91%; (b) Cl2 in CCl4 (excess), CH2Cl2, 0 °C, 30 min, hex-1-ene (excess), 69%); (c)
1.4 equiv 19, 2,4,6-tri-tert-butylpyrimidine (22.5 g, 184 mmol), 4 Å molecular sieves, AgOTf (15 g, 58.2 mmol), ꢀ40 °C, 15 min, 61%); (d) NaOMe (excess), CH2Cl2, MeOH,
Dowex 50WX8, CH2N2, 91%); (e) H2, Pd/C, 2,4,6-tri-tert-butylpyrimidine, EtOAc, 15 min, 87%; (f) levulinic acid (6 equiv), DCC (7 equiv), 4-dimethylaminopyridine (cat), EtOAc,
94%); (g) AcOH, MeOH, reflux, 5 min, 83%.
galactosyl donor would improve this situation. After attempts with
various tri-O-acyl derivatives of the thiogalactoside triol 16,
including levulinoyl,32 pentafluoropropionyl,33 and trifluoroace-
tyl33 groups we eventually selected the benzoylated galactosyl
chloride 18 obtained by chlorinolysis of thiogalactoside 17, which
in turn was prepared by conventional benzoylation of 16.34 Com-
pound 18 was reacted with acetonide 19 under AgOTf activation
LevO
CH2OLev
O
O
LevO
OH
HO
O
LevO
O
OH
26
27
Chart 1.
to afford the desired
a-linked intermediate 20 in 61% yield. The
1,2-cis ( ) interglycosidic linkage was proven by the value of the
a
J1,2 coupling constant being 3.6 Hz (see Section 4). The correspond-
ing b-linked glycoside 21 was also isolated, in 19% yield. Its ano-
meric configuration was indicated by the doublet in its 1H NMR
spectrum at 4.71 ppm, with the J1,2 coupling constant being
7.7 Hz. Next, the benzoyl groups were cleaved from 20 by sodium
methoxide uneventfully, to afford the triol 22 in excellent yield,
followed by careful hydrogenolytic removal of the benzyl group,
using ethyl acetate as solvent and a commercial palladium-on-
charcoal catalyst, in admixture with 2,4,6-tri-tert-butylpyrimidine,
thus affording 23 in 87% yield. In our experience, extended hydrog-
enolysis often leads to a significant drop in pH. A likely reason for
this is the presence of residual palladium chloride in the catalyst,
from which hydrogen chloride may be generated upon hydrogena-
tion. Treatment of the tetraol 23 with levulinic acid and DCC in the
presence of DMAP afforded a nearly stoichiometric yield of the
fully protected galactoside 24.18 Subsequently, the isopropylidene
group was removed from the glycerol moiety by hydrolysis in ace-
tic acid to yield diol 25 in 83% yield, for incorporation of the lipid
chains. Employing identical reaction conditions, diastereomer 27
having the galactose moiety at the sn-1 position was also synthe-
sized, using the commercially available glycerol derivative 26,
the enantiomer of 19. (Chart 1).
afforded the fully protected intermediate 30, from which acetic
acid hydrolysis yielded diol 31 in high yield. (Scheme 2).
With the availability of compounds 25, 27, and 31 incorporating
both the galactose and the glycerol moieties in the required stereo-
chemistries, the stage was set for the introduction of the oleoyl,
palmitoyl, and stearoyl groups.
For the synthesis of the sn-1-O and sn-2-O hetero-di-substituted
Gro-moieties, we opted initially to adapt a published protocol18
that, as we eventually found out by 13C NMR spectroscopy, affor-
ded regioisomeric mixtures. In Ref. 18, diol 25 was treated with
one equivalent of palmitic or oleic acid in the presence of DCC,
apparently without the use of the nucleophilic catalyst DMAP that
in our experience is essential to induce O-acylation. Hexane–EtOAc
mixtures were used to follow the course of the reaction by TLC, and
to isolate the required product ‘singly acylated’ at sn-1-O.18 Using
95:5 and 99:1 v/v mixtures of CH2Cl2 and MeOH as the TLC devel-
oping solvent allowed us to observe that the ‘singly acylated’ prod-
uct is, in fact, a mixture of sn-1-O and sn-2-O mono-acylated
isomers, with the former predominating in a range of approxi-
mately 4:1–9:1. The formation of this mixture remained unnoticed
in an earlier protocol,18 and no physical data have been reported
for the ‘singly acylated’ derivatives either.18 All of our efforts to
separate the ‘singly acylated’ sn-1-O and 2-O products by TLC or
column chromatography using a range of hexane–EtOAc mixtures,
reported in the published protocol,18 proved to be futile.
The synthetic sequence toward BBGL-2 analogs 13 and 14, that
contain the galactose moiety in the unnatural, beta glycosidic link-
age started from imidate 2835 that was reacted with alcohol 19 un-
der activation by TMSOTf, to afford disaccharide 29 in which the b
interlycosidic linkage was proven by the value J1,2 7.9 Hz. A two-
step replacement of the acyl groups in 29 by levulinoyl groups
We also noted the formation of a di-O-Gro acylated product
before the disappearance of the starting diol, an observation that

1554
V. Pozsgay et al. / Carbohydrate Research 346 (2011) 1551–1563
AcO
LevO
LevO
O
a
O
CH2OAc
O
CH2OLev
O
AcO
OH
CH2OLev
O
O
CH2OAc
O
d
b,c
19
O
HO
O
O
O
OH
AcO
LevO
LevO
O
O
AcO
AcO
LevO
LevO
O
NH
CCl3
AcO
29
30
31
28
Scheme 2. Reagents and conditions: (a) 2 equiv 19, CH2Cl2, TMSOTf (cat), 0 °C, 1 h, 80%); (b) NaOMe (excess), MeOH, 24 h, Dowex 50WX8 (H+); (c) levulinic acid (8 equiv),
DCC (8 equiv), EtOAc, 93%); (d) AcOH, MeOH, reflux, 5 min, 85%.
required the termination of the reaction when some of the diols 25,
27, or 31 were still present.
Table 3
Doubly Gro-acylated protected BBGL-2 congeners
The faster-moving mono-O-acylated component was invariably
LevO
the targeted sn-1 acyl product identified by the presence of a char-
acteristic doublet of doublets in its 1H NMR spectrum at ca.
4.37 ppm, indicating that acylation took place at the primary hy-
droxyl group of the glycerol moiety, leaving HO-2 unsubstituted.
The slightly slower-migrating product was the undesired sn-2 acyl
CH2OLev
O
LevO
R2
R1
LevO
O
OR3
Compound
R1
R2
R3
isomer showing
5.1 ppm.
a diagnostic one-proton multiplet at about
37
38
39
40
41
42
43
44
45
H
H
H
H
OPal
OOle
OOle
OPal
OSte
OH
OH
OH
OH
Pal
Ole
Pal
Ole
Ste
Pal
Ole
Pal
Ole
With the homogeneous O-1 (sn) acylated derivatives in hand,
the second acyl group was introduced at O-2 of the glycerol moi-
ety, in uneventful reactions using the DCC/DMAP method.
Treatment of the diols 25, 27, and 31 with various FAs used in
excess under the agency of DCC/DMAP proceeded as expected,
and produced the fully substituted intermediates featuring two
identical fatty acids on the glycerol moiety.
H
OPal
OOle
OOle
OPal
The singly and the doubly glycerol-acylated protected BBGL-2
congeners prepared in this study are listed in Tables 2 and 3,
respectively. In the final stage of the syntheses, the levulinoyl pro-
tecting groups were conventionally removed from the galactose
moiety by hydrazine acetate in pyridine32 to afford 1–14.
The structures of regioisomers 3 and 4 are not easily differenti-
ated by 1H NMR at 500 MHz, but we have found that 1D 13C NMR
provides an excellent method for distinguishing the structures of
the compounds, and assessing their purity. The key parameter is
the 13C@O chemical shift (see Table 5), which appears to be quite
sensitive to the type and chemical environment of the acyl groups.
In this regard we note that HMBC was less suitable to detect con-
nectivities in minor isomers. This is because of the limited spectral
resolution imposed by small numbers of data points in 2D HMBC,
but also because of the restricted signal:noise ratio and dynamic
range caused by the small number of scans per FID in 2D NMR.
By contrast, 1D 13C NMR has better spectral resolution because of
the far greater number of points sampled in the FID, as compared
with those in the indirectly detected, 13C dimension of 2D HMBC.
Analysis of the 13C NMR data of pure oleoyl and palmitoyl digly-
cerides and their mixtures suggests the following rules:
2.2. Nuclear magnetic resonance studies
A detailed NMR investigation was performed on compounds 1–
4. 1H NMR spectra were assigned using 2D COSY and high-resolu-
tion, 2D TOCSY, the latter technique being used to generate sepa-
rate 1D sub-spectra for the Gal, Gro, Ole, and Pal residues.
Selected 1H NMR chemical shifts are reported in Table 4 and cou-
pling constants in Table 6. The small J4,5 value of the Gal residue re-
sulted in inefficient magnetization transfer from H-4 to H-5 and
beyond, so thatthe assignments for H-5, H-6, and H-60 were ob-
tained from the one-bond, 1H–13C correlations established by 2D
HSQC. Selected 13C NMR chemical shifts are listed in Table 5. Sub-
stituent positions in compounds 1–4 were verified by the inter-res-
idue connectivities observed by high-resolution, 2D HMBC.
(a) The 13C@O chemical shift is primarily influenced by the sub-
stituent’s position on the glycerol: the 1-sn-substituent
13C@O’s resonate at lower field than the 2-sn-substituent
13C@O’s, for example, 173.65 versus 173.34 ppm for the Pal
substituent at O-1 versus O-2.
(b) The origin of the minor influence derives from the type of
the substituent: the 13C@O’s of the palmitoyl groups reso-
nate at lower field than those of the oleoyl groups, for exam-
ple, 173.34 versus 173.29 for the Pal versus the Ole group at
sn O-2.
1
Anomeric JC-1,H-1 coupling constants of 1–4 were measured by
1H-coupled, 2D HSQC, and were found to lie in the range of
170.1–170.4 Hz, which together with the small values of J1,2 3.7–
3.8 Hz confirms the
designed.
a configuration of the glycosidic linkages, as
Table 2
Singly Gro-acylated protected BBGL-2 congeners
LevO
CH2OLev
O
LevO
1 R2
These differences have proved to be extremely useful in detect-
ing the presence of glyceride mixtures produced by acylation reac-
tions of imperfect regioselectivity, a facet that has not previously
been recognized.
R
LevO
O
OR3
Compound
R1
R2
R3
32
33
34
35
36
H
H
H
OH
OH
OH
OH
OH
H
Pal
Ole
Ste
Pal
Ole
2.3. Immunochemistry
That sera from LD patients react with carbohydrate and
lipid-containing substances extracted from B. burgdorferi was
H

V. Pozsgay et al. / Carbohydrate Research 346 (2011) 1551–1563
1555
demonstrated many years ago,36–38 but not until 2003 was it
shown that the surface of this pathogen contains two groups of
glycolipids, which were termed BBGL-1 and 2.11 BBGL-2 was re-
ported to contain several FA’s of which palmitate and oleate were
shown to be the major components, but their exact location on the
glycerol moiety is unknown.11 This knowledge may be crucial for
the preparation of a chemically well-characterized, synthetic vac-
cine. In order to determine the structural features necessary for
recognition by antibodies, 14 glycolipids have been synthesized,
(compound 10). The observed non-reactivity might be related to
the low solubility of these compounds as compared to the rest of
the glycolipids in Table 1. All of the compounds having two FA’s
of which at least one is an oleoyl residue reacted with each of
the three sera (3, 4, 7, and 8). The glycolipids having two oleates
(2, 6, and 14) were also reactive. Unexpectedly, compounds with
the Gal moiety in the unnatural b configuration exhibited the same
reactivities as did their natural counterparts: the di-palmitoyl
derivative 13 was unreactive, whereas the di-oleoyl congener
was seemingly as reactive as compound 2.
of which 12 contain the Gal residue in the native,
a anomeric con-
figuration, and 2 in the unnatural, b one. The synthetic glycolipids
were reacted with mouse sera induced by (1) formaldehyde killed
B. burgdorferi cells and (2) by native purified BBGL-2, and also with
serum of patients with late Lyme neuroborreliosis (Fig. 1). There
was no reaction of any glycolipid with control sera (not shown).
All of the glycolipids exhibited similar antigenic activities to all
three sera. No binding occurred with the glycolipids containing
only a single FA moiety (compounds 11 and 12). The lipids contain-
ing two saturated FA’s were also unreactive: no binding occurred
with the compounds containing two palmitates (1, 5, and 13),
two stearates (compound 9), or one palmitate and one stearate
These observations led us to the following conclusions:
(a) The presence in the glycerol part of two fatty acid moieties is
required for antigenicity, leading to the hypothesis that the
binding to antibodies involves two fatty acids and two clo-
sely located binding sites. Alternatively, one FA may position
the other in order to achieve the proper conformation for
binding with antibodies.
(b) Of the two fatty acids, at least one should be oleic acid.
(c) The position of the oleic acid moiety is not critical.
A
3-O-α-D-Galp-1,2-di-O-Pal-sn-GRO
3-O-α-D-Galp-1,2-di-O-Ole-sn-GRO
3-O-α-D-Galp-2-O-Ole-1-O-Pal-sn-GRO
3-O-α-D-Galp-1-O-Ole-2-O-Pal-sn-GRO
1-O-α-D-Galp-2,3-di-O-Pal-sn-GRO
1-O-α-D-Galp-2,3-di-O-Ole-sn-GRO
1-O-α-D-Galp-2-O-Ole-3-O-Pal-sn-GRO
1-O-α-D-Galp-3-O-Ole-2-O-Pal-sn-GRO
3-O-α-D-Galp-1,2-di-O-Ste-sn-GRO
3-O-α-D-Galp-2-O-Pal-1-O-Ste-sn-GRO
3-O-α-D-Galp-1-O-Pal-sn-GRO
1
2
3
4
5
6
7
8
9
10
11
12
13
14
B
3-O-α-D-Galp-1-O-Ole-sn-GRO
3-O-β-D-Galp-1,2-di-O-Pal-sn-GRO
3-O-β-D-Galp-1,2-di-O-Ole-sn-GRO
A - anti-whole bacteria
B - anti-BBGL-2
C-Lyme disease patient serum
C
Figure 1. Immunoblotting of BBGL-2 derivatives 1–14 with mice sera induced against killed B. burgdorferi cells (A), against purified BBGL-II (B), and with LD patient serum
(C). No reaction was observed with control serum (picture not shown).

1556
V. Pozsgay et al. / Carbohydrate Research 346 (2011) 1551–1563
(d) Replacement of the oleic acid by stearic acid abolishes
use. Anhydrous solvents were obtained from Aldrich. Column chro-
matography was performed on Silica Gel 60 (0.040–0.063 mm) and
thin layer chromatography was performed on glass-supported sil-
ica gel layers obtained from Analtech (Uniplate) or on HPTLC plates
from Merck. Visualization was carried out by inspection under UV
light (254 nm), by iodine adsorption, and by charring using a solu-
tion of ammonium cerium(IV) sulfate and ammonium molybdate
in sulfuric acid. 2-O-Oleoyl-1-O-palmitoyl-sn-glycerol (lot #160-
181DG-22) was obtained from Avanti Polar Lipids Inc., Alabaster,
AL. API-ES mass spectra were recorded on an Agilent Technologies
LC MDS CL spectrometer.
NMR spectra were recorded at 300 K, using a Bruker DRX-500
spectrometer equipped with a 5 mm, HCN cryoprobe. Solutions
containing ꢁ20 mg of glycolipid in CDCl3 unless indicated other-
wise (0.5 mL) were used, except for compounds 1, 5, 9, and 13
which were not soluble to this extent, and for which a saturated
solution was employed. Tetramethylsilane was used as an internal
chemical shift reference at d = 0, for 1H and 13C NMR spectra. The
NMR data were acquired and processed by means of the Bruker
Topspin program, version 1.3.
antigenicity.
(e) The anomeric configuration has no influence on antigenicity:
a
and b-galactoside derivatives can be equally antigenic.
(f) The galactose moiety may be linked to either O-1 or to O-3 of
sn-glycerol to maintain antigenicity.
Based on the above observations, we prepared an ‘unnatural’
BBGL-2 analog, using the chemistry described in Section 2.1 having
the structure 3-O-(b-D-galactopyranosyl)-1,2-di-O-oleoyl-DL-glyc-
erol 46, which is much easier and less expensive to synthesize than
the native analogs, by starting from DL-glycerol. Figure 2. shows that
the ‘unnatural’ congener is just as good an antigen as are the reactive
native analogs, and is a candidate for a vaccine against LD.
There is some similarity between our results on antigenic spec-
ificity and those published recently by others demonstrating that
iNKT cells react with BBGL-2 glycolipids in a FA structure-depen-
dent manner, and that the difference in fatty acids influences the
steric position of the sugar epitope.18,39 The highest response, as-
sessed by the level of induced IL-2 was to compound 4 having
Ole and Pal residues in the sn-1 and sn-2 positions, respectively,
whereas the reverse arrangement produced less than half of that.
In our present study, compound 4 also showed a high level of anti-
genicity. On the other hand, compound 1 produced only slightly
more than baseline IL-2 levels, as shown by others.18,39 In a similar
fashion, in our study compound 1 was not recognized by sera in-
duced to B. burgdorferi, or to purified native BBGL-2.
One-dimensional (1D) 1H NMR spectra were acquired in 32,768
data points, zero-filled to 65,536 points, with use of a spectral
width of 4.01 kHz, a 30° pulse (2.67 ls), and a recycle time of 6 s.
For integration, the free induction decay (FID) was subjected to
Gaussian multiplication, using a line-broadening of ꢀ0.5 Hz and a
truncation fraction of 0.3. However, coupling constants were mea-
sured by using a line-broadening of ꢀ1.0 to ꢀ2.0 Hz. 1D 13C NMR
spectra were acquired in 65,536 data points, zero-filled to
131,072 points, by using a 25.15 kHz spectral width, a 45° pulse
3. Conclusions
(7.7 ls), and a recycle time of 2 s. The free induction decays were
subjected either to exponential multiplication using a line-broad-
ening of 1.0 Hz, or to Gaussian multiplication with a line-broaden-
ing of ꢀ0.1 to ꢀ0.5 Hz.
We have described synthetic schemes to homogeneous BBGL-2
congeners that make up a distinctive group of major glycolipids on
the surface of B. burgdorferi, the causative organism of Lyme dis-
2D COSY NMR spectra were acquired by using a 30° 1H pulse,
and 2048 ꢂ 512 point datasets, zero-filled to 2048 ꢂ 2048 points.
The data were processed by using an unshifted sine-bell squared
window in both dimensions, a magnitude calculation in F1, and
symmetrization to remove artifacts. 2D TOCSY spectra were col-
lected in 16,384 ꢂ 512 point datasets, zero-filled to
32,768 ꢂ 2048 points. Resolution enhancement by Gaussian multi-
plication was used in F2 with a line-broadening of ꢀ1.5 Hz, and in
ease, comprising
a-galactopyranosyl-diacyl glycerols. We have
presented evidence that the diastereomeric purity of these glyco-
lipids can best be ascertained by examination of the carbonyl car-
bon resonances in their 13C NMR spectra, whereas the 1H NMR
spectra are less informative. We found that the antigenicity of
the synthetic BBGL-2 glycolipids depends on the unsaturation of
the lipid components, but is independent of the anomeric configu-
ration of the sn-3 linked hydrophilic galactose head-group. The
presence of two FAs in the glycerol part is necessary, of which at
least one has to be an oleic acid residue for antigenicity, a property
that is lost completely upon saturation.
F1, a sine-bell squared window, shifted by
p/2 rad. The spectra
were analyzed by means of extraction of 1D slices in the F2 dimen-
sion. 2D HSQC spectra were recorded in 2048 ꢂ 512 point datasets,
zero-filled to 8192 ꢂ 4096 points. A sine-bell squared window
shifted by
p/4 rad was used in both dimensions. 2D HMBC was
4. Experimental
conducted with an evolution delay of 83 ms, corresponding to opti-
2,3
mization for
J
CH
6 Hz, together with 2048 ꢂ 512 or 2048 point
4.1. General methods
datasets, zero-filled to 4096 ꢂ 4096 points. A magnitude calcula-
tion was used in F1, with sine-bell squared windows shifted by
All chemicals were of commercial grade and used without
purification. Solvents for chromatography were distilled prior to
p
/2 rad in both dimensions. 1H-coupled 2D HSQC spectra were ac-
quired in 8192 ꢂ 512 point datasets, zero-filled to 8192 ꢂ 2048
points, with a sine-bell squared window in both dimensions,
shifted by
p/2 rad. All 2D pulse sequences were field-gradient se-
lected. 1H and 13C spectral widths were 3.005 and 25.15 kHz,
respectively. The phase-sensitive, echo–anti-echo protocol was
used for 2D TOCSY, HSQC, and 1H-coupled HSQC.
3-O-α-D-Galp-1,2-di-O-Ole-sn-GRO (2)
1-O-α-D-Galp-2,3-di-O-Ole-sn-GRO (6)
The synthetic glycolipids were dissolved in CHCl3/MeOH 80:20
(v/v) at 1 mg/ml concentration. Then, 3 lL of the solution was di-
rectly pipetted onto stripes of a PVDF membrane (Millipore, Bed-
ford, MA). After 10 min, the membranes were blocked with 5%
BSA in PBS for 1 h at rt and washed three times with PBS. Next,
the membranes were incubated for 2 h at rt with the following
sera: (1) induced to heat killed B. burgdorferi cells (strain B31, ATCC
35210), diluted 1:100 in blocking buffer; (2) induced to purified
BBGL-2, diluted 1:50 in blocking buffer; (3) serum from human
3-O-β-D-Galp-1,2-di-O-Ole-DL-GRO (46)
Figure 2. Dot-blot of BBGL-2 derivatives 2, 6, and 46 with mice sera induced by
native purified BBGL-2 injected in Freund’s adjuvant.

V. Pozsgay et al. / Carbohydrate Research 346 (2011) 1551–1563
1557
Table 5
patients diagnosed with Lyme disease, diluted 1:50 in blocking
buffer; (4) control mouse and human sera. Mouse sera were pre-
pared as described in Refs. 11 and 40. After washing as above,
the membranes were incubated with goat anti-mouse or anti-hu-
man IgG conjugated to alkaline phosphatase (KPL, Gaithersburg,
MD) for 1 h at rt, then washed and developed with BCIP/NBT phos-
phate substrate (KPL, Gaithersburg, MD).
Selected 13C NMR chemical shifts (ppm) of galactopyranosyl diglyceride analogs 1–4
in CDCl3
1
2
3
4
Gal C-1
C-2
C-3
C-4
C-5
C-6
Gro C-1
C-2
C-3
Ole C-1
C-2
99.25
69.45
71.00
70.31
69.98
63.28
61.95
69.87
99.37
69.13
70.64
70.14
70.14
62.45
62.42
69.88
99.31
69.30
70.81
70.29
70.03
62.93
62.17
69.87
66.67
99.34
69.28
70.79
70.29
70.06
62.85
62.24
69.87
66.65
4.2. 3-O-a-D
-Galactopyranosyl-1,2-di-O-palmitoyl-sn-glycerol18
(1)
66.88
66.46
TLC: CH2Cl2–MeOH (95:5). For the 1H and 13C NMR spectra see
Tables 4–6. API-ES-MS: m/z calcd for [C41H78O10]NH4þ: 748.6.
Found 748.4.
—
—
173.73, 173.35 173.29
34.28, 34.11
173.63
34.28
34.30
C-9
—
—
—
130.03
129.69
14.13, 14.13
130.05
129.69
14.13
130.04
129.69
14.13
C-10
C-18
Pal C-1
C-2
4.3. 3-O-a-D-Galactopyranosyl-1,2-di-O-oleoyl-sn-glycerol (2)
173.50, 173.25
34.29, 34.10
14.13, 14.13
—
—
—
173.65
34.16
14.13
173.34
34.10
14.13
TLC: EtOAc–hexanes (1:1, 4:1). For the 1H and 13C NMR spectra
see Tables 4–6. API-ES-MS: m/z calcd for [C45H82O10]NH4þ: 800.6.
Found 800.6.
C-16
Table 6
4.4. 3-O-a-D-Galactopyranosyl-2-O-oleoyl-1-O-palmitoyl-sn-
Selected two and three bond homonuclear H-H and one-bond heteronuclear C–H
NMR coupling constants (Hz) of galactopyranosyl diglyceride analogs 1–4 in CDCl3
glycerol17–20 (3)
1
2
3
4
TLC: EtOAc–MeOH (99:1, 95:5), CH2Cl2–MeOH (95:5). For the
1H and 13C NMR spectra see Tables 4–6. API-ES-MS: m/z calcd for
[C43H80O10]NH4þ: 774.6. Found 774.4.
Gal J1,2
J2,3
J3,4
J4,5
J5,6
3.8
9.7
3.3
0.9
5.1
3.7
10.3
3.2
3.8
9.8
3.2
1.1
5.0
3.8
9.9
3.3
1.3
5.0
<0.5
Nra
Nr
4.5. 3-O-a-D-Galactopyranosyl-1-O-oleoyl-2-O-palmitoyl-sn-
glycerol18–20 (4)
0
J5,6
Nr
4.6
4.7
0
J6,6
11.6
170.1
11.8
4.2
nr
11.4
11.3
JC-1,H-1
Gro J1,1
J1,2
170.1
12.0
3.6
170.4
11.9
4.0
170.1
12.0
3.9
TLC: EtOAc–MeOH (99:1, 95:5), CH2Cl2–MeOH (95:5). For the
1H and 13C NMR spectra see Tables 4–6. API-ES-MS: m/z calcd for
[C43H80O10]NH4þ: 774.6. Found 774.6.
0
0
J1 ,2
5.7
4.6
6.2
5.4
5.9
5.0
6.0
5.0
J2,3
0
J2,3
6.3
5.8
6.1
6.0
4.6. 1-O-a-D-Galactopyranosyl 2,3-di-O-palmitoyl-sn-glycerol
(5)
0
J3,3
11.0
10.8
10.9
10.9
a
Nr = not resolved.
TLC: CH2Cl2–MeOH (95:5); 1H NMR (CDCl3): d 5.24 (m, 1H), 4.97
(d, 1H, J = 3.7 Hz), 4.29 (m, 1H, J = 4.1 Hz, J = 11.9 Hz), 4.20 (dd, 1H,
J = 6.0 Hz, J = 11.9 Hz), 4.09 (br d, 1H, J = 2.1 Hz), 3.94 (dd, 1H,
J = 6.3 Hz), J = 11.6 Hz), 3.87–3.81 (m, 4H), 3.74 (dd, 1H, J = 3.1 Hz,
J = 9.8 Hz), 3.63 (dd, 1H, J = 4.8 Hz, J = 11.1 Hz); 13C NMR (CDCl3):
d 173.55, 173.35, 99.29, 70.89, 70.28, 70.11, 69.92, 69.36, 66.89,
63.14, 62.17, 34.31, 34.10, 31.94, 29.71, 29.69, 29.67, 29.66,
29.51, 29.37, 29.30, 29.15, 29.12, 24.89, 22.70, 14.13. API-ES-MS:
m/z calcd for [C41H78O10]NH4þ: 748.6. Found 748.4.
Table 4
Selected 1H NMR chemical shifts (ppm) of galactopyranosyl diglyceride analogs 1–4
in CDCl3
1
2
3
4
4.7. 1-O-a-D-Galactopyranosyl-2,3-di-O-oleoyl-sn-glycerol (6)
Gal H-1
H-2
H-3
H-4
H-5
H-6
H-60
Gro H-1
H-10
4.954
3.829
3.758
4.098
3.824
3.962
3.851
4.379
4.117
5.254
3.856
3.641
—
4.925
3.868
3.788
4.087
3.798
3.833
3.799
4.374
4.148
5.250
3.812
3.624
2.318, 2.309
2.009, 2.008
5.342, 5.342
2.009, 2.008
0.880, 0.880
—
4.940
3.850
3.774
4.095
3.808
3.914
3.837
4.377
4.132
5.252
3.834
3.632
2.321
2.010
5.344
2.010
0.880
2.313
0.880
4.937
3.856
3.776a
4.093a
3.806
3.905
3.834
4.373
4.135
5.254
5.254
3.631
2.313
2.009
5.343
2.009
0.880
0.880
0.880
TLC: EtOAc–hexanes (1:1, 4:1); 1H NMR (CDCl3): d5.39–5.29 (m,
4H), 5.25 (m, 1H), 4.97 (d, 1H, J = 3.6 Hz), 4.30 (dd, 1H, J = 3.9 Hz,
J = 11.9 Hz), 4.19 (dd, 1H, J = 6.2 Hz, J = 11.9 Hz), 4.09 (d, 1H,
J = 3.2 Hz), 3.93 (dd, 1H, J = 5.0 Hz, J = 11.6 Hz), 3.88–3.78 (m, 4H),
3.75 (dd, 1H, J = 3.3 Hz, J = 10.3 Hz), 3.62 (dd, 1H, J = 5.0 Hz,
J = 11.1 Hz), 2.35–2.27 (m, 4H), 2.04–1.95 (m, 8H), 1.66–1.54 (m.
4H), 1.37–1.19 (m, 40H), 0.91–0.82 (m, 6H); 13C NMR (CDCl3): d
173.60, 173.34, 129.98, 129.64, 129.62, 99.37, 70.58, 70.18, 70.09,
69.95, 69.09, 66.63, 62.45, 62.41, 34.24, 34.04, 31.87, 29.73,
29.70, 29.70, 29.49, 29.28, 29.19, 29.13, 29.09, 29.06, 27.19,
27.16, 24.89, 24.82, 22.65, 14.08. API-ES-MS: m/z calcd for
[C45H82O10]NH4þ: 800.6. Found 800.4.
H-2
H-3
H-30
Ole H-2,20
H-8,80
H-9,10
H-11,110
H-18
—
—
—
—
Pal H-2,20
H-16
2.324, 2.317
0.881, 0.881
4.8. 1-O-
a-D-Galactopyranosyl-2-O-oleoyl-3-O-palmitoyl-sn-
—
glycerol (7)
a
Our assignments are at variance with those published in Ref. 19 for compound
TLC: EtOAc–MeOH (99:1, 95:5), CH2Cl2–MeOH (95:5); 1H NMR
(CDCl3): 5.38–5.31 (m, 2H), 5.24 (m, 1H), 4.30 (dd, 1H,
4: H-3 of Gal was reported in the range of 4.03–4.12 ppm, while H-4 was in the
3.79–3.94 ppm range, that is, reversed from the data shown above.
d

1558
V. Pozsgay et al. / Carbohydrate Research 346 (2011) 1551–1563
J = 4.9 Hz, J = 11.8 Hz), 4.19 (dd, 1H, J = 6.2 Hz, J = 11.8 Hz), 4.09 (d,
1H, J = 2.8 Hz), 3.93 (dd, 1H, J = 5.3 Hz, J = 11.8 Hz), 3.86–3.79 (m,
4H), 3.75 (dd, 1H, J = 2.8 Hz, J = 9.6 Hz), 3.62 (dd, 1H, J = 4.9 Hz,
J = 10.9 Hz), 2.33 (t, 1H, J = 6.9 Hz), 2.31 (t, 2H, J = 7.4 Hz), 2.05–
1.99 (m, 4H), 2.05–1.99 (m, 2H), 1.65–1.56 (m, 4H), 1.38–1.19
(m), 0.88 (t, 6H, J = 7.0 Hz); 13C NMR (CDCl3): d 173.69, 173.38,
130.02, 129.66, 99.39, 70.84, 70.28, 70.13, 69.94, 69.34, 66.83,
63.00, 62.23, 34.28, 34.10, 31.93, 31.91, 29.76, 29.71, 29.68,
29.67, 29.65, 29.54, 29.51, 29.37, 29.33, 29.32, 29.30, 29.21,
29.15, 29.09, 27.19, 27.18, 24.92, 24.88, 22.70, 22.69, 14.13. API-
ES-MS: m/z calcd for [C43H80O10]NH4þ: 774.6. Found 774.6.
64.44, 61.07, 33.38, 31.27, 29.01, 28.98, 28.95, 28.82, 28.69,
28.63, 28.48, 24.22, 21.99, 13.11. API-ES-MS: m/z calcd for
[C25H48O9]NH4þ: 510.4. Found 510.5.
4.13. 3-O-a-D-Galactopyranosyl-1-O-oleoyl-sn-glycerol (12)
TLC: EtOAc–MeOH (95:5), CH2Cl2–MeOH (95:5, 9:1); 1H NMR
(CDCl3–MeOH-d4): d 5.38–5.30 (m, 2H), 4.89 (d, 1H, J = 3.7 Hz),
4.17 (dd, 1H, J = 4.1 Hz, J = 11.3 Hz), 4.10 (dd, 1H, J = 5.8 Hz,
J = 11.3 Hz), 4.05 (m 1H), 3.95 (d, 1H, J = 3.4 Hz), 3.89–3.69 (m,
6H), 3.42 (dd, 1H, J = 7.6 Hz, J = 10.4 Hz), 2.37 (t, 2H, J = 7.4 Hz),
2.08–1.98 (m, 4H), 1.69–1.58 (m, 2H), 1.40–1.20 (m, 20H), 0.89
(t, 3H, J ꢁ6.8 Hz); 13C NMR (CDCl3–MeOH-d4): d 173.89, 129.36,
129.12, 99.07, 70.26, 69.67, 69.25, 69.10, 68.57, 67.93, 64.50,
61.13, 33.45, 31.32, 29.15, 29.11, 28.91, 28.72, 28.69, 28.61,
28.53, 28.51, 26.56, 26.54, 24.27, 22.06, 13.23. API-ES-MS: m/z
calcd for [C27H48O9]NH4þ: 536.4. Found 536.5.
4.9. 1-O-
a-D-Galactopyranosyl-3-O-oleoyl-2-O-palmitoyl-sn-
glycerol (8)
TLC: EtOAc–MeOH (99:1, 95:5), CH2Cl2–MeOH (95:5); 1H NMR
(CDCl3–MeOH-d4): d 5.30–5.23 (m, 2H), 5.15 (m, 1H), 4.85 (d, 1H,
J = 3.4 Hz), 4.27 (dd, 1H, J = 3.7 Hz, J = 12.0 Hz), 4.11 (dd, 1H,
J = 6.5 Hz, J = 12.0 Hz), 4.01 (d, 1H, J = 3.3 Hz), 3.48–3.65 (m, 5H),
3.54 (dd, 1H, J = 5.4 Hz, J = 10.8 Hz), 2.25 and 2.23 (2t, 2 ꢂ 2H,
J = 7.2 Hz for each), 1.99–1.91 (m, 4H), 1.60–1.48 (m, 4H), 1.31–
4.14. 3-O-b-D-Galactopyranosyl-1,2-di-O-palmitoyl-sn-glycerol
(13)
1.14 (m), 0.86–0.77 (m, 6H); 13C NMR (CDCl3–MeOH-d4):
d
TLC: EtOAc–MeOH (95:5), CH2Cl2–MeOH (95:5, 9:1); 1H NMR
(CDCl3–MeOH-d4): 5.28 (m, 1H), 4.43 (dd, 1H, J = 3.0 Hz,
173.68, 173.35, 129.96, 129.6, 99.37, 70.51, 70.19, 69.96, 68.99,
66.56, 62.54, 62.16, 34.24, 34.07, 31.89, 31.86, 29.72, 29.71,
29.68, 29.63, 29.51, 29.49, 29.42, 29.33, 29.29, 29.28, 29.21,
29.14, 29.06, 28.18, 27.15, 24.88, 24.84, 22.65, 22.64, 14.07. API-
ES-MS: m/z calcd for [C43H80O10]NH4þ: 774.6. Found 774.6.
d
J = 12.0 Hz), 4.25 (d, 1H, J = 7.5 Hz), 4.24 (dd, 1H, J = 6.8 Hz,
J = 12.0 Hz), 3.99 (dd, 1H, J = 5.4 Hz, J = 12.0 Hz), 3.88 (dd, 1H,
J = 1.1 Hz, J = 3.3 Hz), 3.81 (dd, 1H, J = 6.6, J = 11.5 Hz), 3.75 (dd,
1H, J = 5.4 Hz, J = 11.5 Hz), 3.74 (dd, 1H, 5.9 Hz, J = 11.0 Hz), 3.54
(dd, 1H, J = 7.5 Hz, J = 9.7 Hz), 3.89 (dd, 1H, J = 3.4 Hz, J = 9.7 Hz),
2.35–2.31 (m, 4H), 1.61 (m, 4H), 1.36–1.22 (m, 48H), 0.89 (t, 6H,
J ꢁ7.1 Hz); 13C NMR (CDCl3–MeOH-d4): 173.44, 173.11, 103.39,
74.60, 72.82, 70.47, 69.79, 68.19, 67.00, 62.17, 60.57, 33.50,
33.36, 31.21, 28.96–28.37, 24.19, 21.92, 12.98. API-ES-MS: m/z
calcd for [C41H78O10]NH4þ: 748.6. Found 748.7.
4.10. 3-O-a-D
-Galactopyranosyl-1,2-di-O-stearoyl-sn-glycerol17
(9)
TLC: EtOAc–MeOH (99:1, 95:5), CH2Cl2–MeOH (95:5); 1H NMR
(CDCl3–MeOH-d4): d 5.26 (m, 1H), 4.88 (d, 1H, J = 3.7 Hz), 4.39
(dd, 1H, J = 3.6 Hz, J = 12.0 Hz), 4.16 (dd, 1H, J = 6.4 Hz,
J = 12.0 Hz), 3.99 (d, 1H, J = 3.3 Hz), 3.85–3.75 (m), 3.73 (dd, 1H,
J = 3.3 Hz), J = 10.0 Hz), 3.63 (dd, 1H, J = 5.9 Hz, J = 10.9 Hz), 2.53
(m, 1H), 2.45 (m, 1H), 2.35–2.30 (m, 4H), 1.65–1.56 (m, 4H),
1.36–1.19 (m, 56H), 0.88 (t, 6H, J ꢁ6.8 Hz); 13C NMR (CDCl3–
MeOH-d4): d 173.82, 173.48, 99.34, 70.27, 70.03, 69.85, 69.62,
68.82, 66.15, 62.34, 61.71, 34.06, 33.89, 31.70–28.89, 25.62,
24.67, 24.65, 22.45, 13.79. API-ES-MS: m/z calcd for
[C45H86O10]NH4þ: 804.6. Found 804.7.
4.15. 3-O-b-D-Galactopyranosyl-1,2-di-O-oleoyl-sn-glycerol (14)
TLC: EtOAc–MeOH (95:5), CH2Cl2–MeOH (95:5, 9:1); 1H NMR
(CDCl3–MeOH-d4): d 5.37–5.31 (m, 4H), 5.29 (m, 1H), 4.39 (dd,
1H, J = 3.1 Hz, J = 12.2 Hz), 4.27 (d, 1H, J = 7.6 Hz), 4.21 (dd, 1H,
J = 6.6 Hz, J = 12.1 Hz), 4.02 (d, 1H, J = 2.6 Hz), 3.92 (dd, 1H,
J = 5.4 Hz, J = 11.0 Hz), 3.84 (dd, 1H, J = 4.9 Hz, J = 12.0 Hz), 3.74–
3.65 (m, 3H), 3.59 (dd, 1H, J = 2.9 Hz, J = 9.6 Hz), 3.54 (t, 1H, J
ꢁ4.6 Hz), 2.32 and 2.32 (2t, 2 ꢂ 2H, J ꢁ7.4 Hz), 2.08–1.94 (m,
4.11. 3-O-
a-
D-Galactopyranosyl-2-O-palmitoyl-1-O-stearoyl-sn-
8H), 1.65–1.54 (m, 4H), 1.37–1.19 (m, 40H), 0.88 (t, 6H, J
glycerol (10)
ꢁ6.8 Hz); 13C NMR (CDCl3–MeOH-d4): d 173.88, 173.47, 129.97,
129.64, 104.00, 74.52, 73.37, 71.11, 70.12, 68.88, 68.04, 62.87,
TLC: EtOAc–MeOH (99:1, 95:5), CH2Cl2–MeOH (95:5); 1H NMR
(CDCl3–MeOH-d4): d 5.28 (m, 1H), 4.87 (d, 1H, J = 3.6 Hz), 4.44
(dd, 1H, J = 3.2 Hz, J = 12. Hz), 4.19 (dd, 1H, J = 6.4 Hz, J = 12.0 Hz,
3.95 (d, 1H, J = 3.2 Hz), 3.86 (dd, 1H, J = 5.5 Hz, J = 10.8 Hz), 3.83–
3.71 (m, 6H), 3.65 (dd, 1H, J = 5.7 Hz, J = 10.8 Hz), 2.35 and 2.34
(2t, 2 ꢂ 2H, J ꢁ7.1 Hz), 1.68–1.57 (m, 4H), 1.40–1.20 (52H), 0.89
(t, 6H, J ꢁ7 Hz); 13C NMR (CDCl3–MeOH-d4): d 173.54, 173.19,
99.03, 70.37, 69.64, 69.58, 69.19, 68.36, 65.53, 62.09, 61.06,
33.64, 33.47, 31.32, 29.07–28.49, 24.30, 22.03, 13.16. API-ES-MS:
m/z calcd for [C43H82O10]NH4þ: 776.6. Found 776.6.
61.52, 34.24, 34.11, 31.88, 27.78–29.08, 27.20, 27.18, 24.86,
þ
24.84, 22.65, 14.09. API-ES-MS m/z calcd for [C45H82O10]NH4
:
800.6. Found 800.4.
4.16. Phenyl 3,4,6-tri-O-benzoyl-2-O-benzyl-1-thio-b-D-
galactopyranoside (17)
To a solution of compound 16 (32 g, 88.4 mmol) in a mixture of
anhydrous CH2Cl2 (300 mL) and C5H5N (160 mL) containing a cat-
alytic amount of DMAP was added benzoyl chloride (50 mL, 60.5 g,
430 mmol), dropwise at room temperature under stirring. After
24 h, the solution was treated with MeOH (70 mL). The solution
was concentrated under reduced pressure. The residue was equil-
ibrated between CHCl3 and water. The organic layer was extracted
with 4 N hydrochloric acid followed by water and saturated aq
NaHCO3. Column chromatographic purification of the product
using a hexanes–EtOAc gradient 10:1?4:1 afforded 17 (54.0 g,
91%) as a solid: 1H NMR (CDCl3): d 8.03–7.13 (m, 25H), 5.92
(ddd, 1H, J3,4 = 3.5 Hz, J4,5 = 0.9 Hz, H-4), 5.50 (dd, J2,3 = 9.5 Hz,
J3,4 = 3.5 Hz, H-3), 4.89 (d, 1H, J1,2 = 9.6 Hz, H-1), 4.84 and 4.61
4.12. 3-O-a-D-Galactopyranosyl-1-O-palmitoyl-sn-glycerol (11)
TLC: EtOAc–MeOH (95:5), CH2Cl2–MeOH (95:5, 9:1); 1H NMR
(CDCl3–MeOH-d4): 4.88 (d, 1H, J = 3.6 Hz), 4.17 (dd, 1H,
d
J = 4.6 Hz, J = 11.3 Hz), 4.10 (dd, 1H, J = 5.9 Hz, J = 11.3 Hz), 4.06
(m, 1H), 3.95 (d, 1H, J = 3.1 Hz), 3.85–3.72 (m, 6H), 3.42 (dd, 1H,
J = 7.6 Hz, J = 10.3 Hz), 2.36 (t, 2H, J = 7.5 Hz), 1.63 (m, 2H), 1.36–
1.24 (m, 24H), 0.89 (t, 3H, J ꢁ6.9 Hz); 13C NMR (CDCl3–MeOH-
d4): d 173.86, 99.01, 70.26, 69.61, 69.21, 68.97, 68.50, 67.85,

V. Pozsgay et al. / Carbohydrate Research 346 (2011) 1551–1563
1559
[2d, 2 ꢂ 1H J = 10.8 Hz each, CH2 (Bn)], 4.62 (dd, 1H, J5,6 = 7.1 Hz,
J = 12.3 Hz each, CH2 (Bn)], 4.71 (d, 1H, J1,2 = 7.7 Hz, H-1), 4.39
(m, 1H, H-20), 4.36 (dd, 1H, J = 6.5 Hz, J = 11.3 Hz), 4.10 (dd, 1H,
J = 6.5 Hz, J = 8.3 Hz), 4.08 (dd, 1H, J = 5.0 Hz, J = 10.4 Hz), 3.94
(dd, 1H, J = 7.8 Hz, J = 10.4 Hz), 3.90 (dd, 1H, J = 6.0 Hz, J = 8.2 Hz),
3.72 (dd, 1H, J = 6.0 Hz, J = 10.4 Hz), 1.45 and 1.37 (2s, 2 ꢂ 3H,
0
0
0
J6,6 = 11.5 Hz, H-6), 4.41 (dd, 1H, J5,6 = 5.7 Hz, J6,6 = 11.5 Hz), 4.27
0
(ddd, 1 J, J4,5 = 0.9 Hz, J5,6 = 7.1 Hz, J5,6 = 5.7 Hz, H-5), 4.01 (t, 1H,
J1,2 = J2,3 = 9.5 Hz); 13C NMR (CDCl3): d 166.01, 165.35, 137.36,
133.50, 133.22, 133.18, 132.81, 132.71, 129.92, 129.79, 129.64,
129.46, 129.29, 129.21, 128.99, 128.55, 128.41, 128.29, 128.27,
128.10, 127.93, 127.83, 87.51, 75.48, 75.22, 74.80, 74.77, 68.79,
62.58, 62.52. API-ES-MS: m/z calcd for [C40H34O8S]NH4þ: 692.2.
Found 692.2.
2CH3); 13C NMR (CDCl3):
d 165.89, 165.41, 165.33, 137.65,
133.32, 133.13, 133.00, 129.80, 129.62, 129.60, 129.33, 129.23,
129.18, 128.41, 128.34, 128.11, 128.09, 127.97, 127.53, 109.43,
104.18, 76.08, 74.56, 74.24, 72.62, 70.93, 70.60, 68.39, 66.56,
þ
61.97, 26.71, 25.21. API-ES-MS: m/z calcd for [C40H40O11]NH4
:
4.17. 3,4,6-Tri-O-benzoyl-2-O-benzyl-
a-
D-galactopyranosyl
714.3. Found 714.3.
chloride (18)
4.19. 3-O-(2-O-Benzyl-a-D-galactopyranosyl)-1,2-O-
To a solution of compound 17 (25.0 g, 41.6 mol) in anhydrous
CH2Cl2 (250 mL) was added a saturated solution of chlorine in
CCl4 at 0 °C. After 30 min TLC (hexanes–EtOAc, 6:1) showed com-
plete conversion to a slightly faster-moving compound. The reac-
tion mixture was treated with 1-hexene (excess) followed by
concentration. Column chromatographic purification of the residue
(hexanes–EtOAc, 5:1) afforded amorphous 18 (15.4 g, 69%): 1H
NMR (CDCl3, d): 8.04–7.19 (m, 20H), 6.29 (d, 1H, J1,2 = 3.8 Hz, H-
1), 5.99 (dd, 1H, J3,4 = 3.4 Hz, J4,5 = 1.2 Hz, H-4), 5.81 (dd, 1H,
J2,3 = 10.3 Hz, J3,4 = 3.4 Hz, H-3), 4.84 (m, 1H, H-5), 4.70 and 4.64
[2d, 2 ꢂ 1H J = 12.0 Hz each, CH2 (Bn)], 4.56 (dd, 1H, J5,6 = 6.8 Hz,
isopropylidene-sn-glycerol (22)
To a stirred solution of tribenzoate 20 (8.0 g, 11.4 mmol), anhy-
drous CH2Cl2 (50 mL) and MeOH (50 mL) was added a 25% solution
of NaOMe in MeOH (10 mL) at rt. After 12 h, the solution was care-
fully treated with Dowex 50WX8 to approx. pH 5 as seen by indi-
cator paper. The resin was removed by filtration and the filtrate
was treated with an ethereal solution of CH2N2 until the yellow
color persisted. Purification of the residue by silica gel chromatog-
raphy using EtOAc as the eluant afforded solid 22 (5.0 g, 91%): 1H
NMR (CDCl3): d 7.38–7.29 (m, 5H), 4.95 (d, 1H, J = 3.5 Hz, H-1),
4.66 [s, 2H, CH2 (Bn)] 4.32 (m, 1H), 4.08 (dd, 1H, J = 1.1 Hz,
J = 3.4 Hz, H-4), 4.05 (dd, 1H, J = 6.5 Hz, J = 8.3 Hz), 3.99 (dd, 1H,
J = 3.4 Hz, J = 9.8 Hz), 3.91 (dd, 1H, J = 5.2 Hz, J = 11.2 Hz), 3.85 (dt,
1H, J = 1.2 Hz, J = 4.3 Hz, J = 5.4 Hz), 3.81 (dd, 1H, J = 4.2 Hz,
J = 11.2 Hz, 3.76 (dd, 1H, J = 3.5 Hz, J = 9.7 Hz), 3.73 (dd, 1H,
J = 6.2 Hz, J = 8.2 Hz), 3.65 (dd, 1H, J = 4.9 Hz, J = 10.7 Hz), 3.46
(dd, 1H, J = 6.2 Hz, J = 10.7 Hz), 1.43 and 1.36 (2s, 2 ꢂ 3H, 2CH3);
13C NMR (CDCl3): d 137.88, 128.63, 128.20, 128.15, 109.63, 97.11,
76.48, 74.63, 72.67, 70.29, 69.32, 68.98, 68.75, 66.47, 63.02,
26.78, 25.39. API-ES-MS: m/z calcd for [C19H28O8]NH4þ: 402.2.
Found 402.2.
J6,6 = 11.5 Hz, H-6), 4.34 (dd, 1H, J5,6 = 6.0 Hz, J6,6 = 11.5 Hz, H-60),
4.32 (dd, 1H, J1,2 = 3.8 Hz, J2,3 = 10.3 Hz, H-2); 13C NMR (CDCl3): d
165.90, 165.26, 165.21, 136.88, 133.55, 133.24, 133.18, 129.79,
129.76, 129.66, 129.33, 129.28, 129.08, 128.58, 128.51, 128.40,
128.28, 128.17, 128.04, 93.01, 73.09, 72.81, 69.89, 69.71, 68.59,
61.84.
0
0
4.18. (3,4,6-Tri-O-benzoyl-2-O-benzyl-a-D-galactopyranosyl)-
1,2-O-isopropylidene-sn-glycerol (20) and (3,4,6-tri-O-benzoyl-
2-O-benzyl-b-D-galactopyranosyl)-1,2-O-isopropylidene-sn-
glycerol (21)
To a stirred mixture of chloride 18 (15.2 g, 25.3 mmol), 2,4,6-tri-
tert-butylpyrimidine41 (22.5 g), crushed 4 Å molecular sieves
(15 g), 1,2-O-isopropylidene-sn-glycerol 19 (4.5 mL, 4.8 g,
36.5 mmol), and anhydrous CH2Cl2 (125 mL) was added AgOTf
(15 g, 58.2 mmol) at ꢀ40 °C. After 15 min TLC (3:1 hexanes–EtOAc)
indicated the disappearance of 18 and the formation of two closely
migrating products. To the reaction mixture were added Bu4NBr
(11.5 g) and saturated aqueous NaHCO3 and the resulting mixture
was filtered through a layer of Celite. The organic layer was con-
centrated and the residue was chromatographed on silica gel using
a 20:1?2:1 hexanes–EtOAc gradient to afford 20 (10.8 g, 61%) as
an amorphous substance: 1H NMR (CDCl3, d): 8.01–7.25 (m,
20H), 5.94 (dd, 1H, J3,4 = 3.0 Hz, J4,5 = 1.2 Hz, H-4), 5.76 (dd, 1H,
J2,3 = 10.5 Hz, J3,4 = 3.0 Hz, H-3), 5.14 (d, 1H, J1,2 = 3.6 Hz, H-1),
4.71 and 4.63 [2d, 2 ꢂ 1H J = 12.3 Hz each, CH2 (Bn)], 4.58 (m,
1H, H-5), 4.53 (dd, 1H, J7.5 Hz, J = 11.1), 4.36 (m, 1H, H-2), 4.32
(dd, 1H, J = 5.2 Hz, J = 11.0 Hz), 4.16 (dd, 1H, J = 3.5 Hz, J = 10.5 Hz,
H-2), 4.09 (dd, J = 6.5 Hz, 8.4 Hz), 3.78 (dd, J = 6.5 Hz, J = 8.4 Hz),
3.74 (dd, 1H, J = 5.5 Hz, J = 10.5 Hz), 3.64 (dd, 1H, J = 5.5 Hz,
J = 10.5 Hz), 1.42 and 1.37 (2s, 2 ꢂ 3H, 2CH3); 13C NMR (CDCl3): d
166.02, 165.47, 165.46, 137.78, 133.40, 133.18, 133.02, 129.83,
129.72, 129.66, 129.53, 129.41, 128.54, 128.41, 128.25, 127.96,
127.93, 109.54, 97.85, 74.55, 73.44, 72.87, 70.05, 69.51, 69.16,
67.10, 66.85, 62.72, 26.81, 25.50. API-ES-MS: m/z calcd for
[C40H40O11]NH4þ: 714.3. Found 714.2.
4.20. 3-O-(
glycerol (23)
a-D-Galactopyranosyl)-1,2-O-isopropylidene-sn-
A mixture of compound 22 (480 mg), 2,4,6-tri-tert-butylpyrim-
idine (500 mg), 10% palladium-on-charcoal (0.5 g), EtOAc (10 mL)
was stirred under hydrogen at 200 psi. After 10 min, the mixture
was filtered and the solids were washed with MeOH. Concentra-
tion of the combined solutions under reduced pressure afforded a
semisolid which was purified by silica gel column chromatography
using 9:1 CH2Cl2–MeOH as the eluant to afford amorphous 23
(320 mg, 87%): 1H NMR (CDCl3): d 4.91 (d, 1H, J = 3.5 Hz), 4.37
(m, 1H, H-20), 4.11 (dd, 1H, J = 6.8 Hz, J = 8.5 Hz), 3.97 (br d, 1H,
J = 3.2 Hz), 3.85–3.73 (m, 7H), 3.50 (dd, 1H, J = 6.8 Hz, J = 10.5 Hz),
1.46 and 1.39 (2s, 2 ꢂ 3H, 2CH3); 13C NMR (CDCl3): d 109.45,
98.89, 74.39, 70.25, 69.92, 69.39, 68.78, 68.68, 65.85, 61.37,
26.07, 24.72. API-ES-MS: m/z calcd for [C12H22O8]NH4þ: 312.2.
Found 312.2.
4.21. 1,2-O-Isopropylidene-3-O-(2,3,4,6-tetra-O-levulinoyl-
galactopyranosyl)-sn-glycerol (24)
a-D-
To a stirred solution of tetraol 23 (320 mg, 1.09 mmol) in EtOAc
(10 mL) were added sequentially levulinic acid (760 mg,
6.55 mmol), N,N-dicyclohexylcarbodiimide (1.6 g, 7.7 mmol) and
a catalytic amount of 4-dimethylaminopyridine at rt. After 24 h,
the solids were removed by filtration and washed with EtOAc).
The solutions were combined and concentrated under reduced
pressure. The residue was purified by silica gel column chromatog-
raphy using 1:1?1:4 hexanes–EtOAc as the eluant to afford 24
(702 mg, 94%) as a syrup: 1H NMR (CDCl3): d 5.43 (dd, 1H,
Also isolated was (3,4,6-tri-O-benzoyl-2-O-benzyl-b-D-galacto-
pyranosyl)-1,2-O-isopropylidene-sn-glycerol 21 (3.35 g, 19%): 1H
NMR (CDCl3): d 8.03–7.09 (m, 20H), 5.88 (br d, 1H, J3,4 = 3.5 Hz,
H-4), 5.44 (dd, 1H, J2,3 = 10.1 Hz, J3,4 = 3.5 Hz, H-3), 4.87 and 4.69
[2d, 2 ꢂ 1H J = 11.8 Hz each, CH2 (Bn)], 4.71 (d, [2d, 2 ꢂ 1H

1560
V. Pozsgay et al. / Carbohydrate Research 346 (2011) 1551–1563
J = 1.2 Hz, J = 3.7 Hz), 5.34 (dd, 1H, J = 3.7 Hz, J = 10.6 Hz), 5.13 (dd,
1H, J = 3.7 Hz, J = 9.5 Hz), 5.12 (br s, 1H), 4.30 (dd, 1H, J = 5.8 Hz,
J = 11.6 Hz), 4.26 (t, J = 6.6 Hz), 4.11 (br s, 1H), 4.04 (br s, 1H),
4.08 (dd, 1H, J = 6.4 Hz, J = 8.3 Hz), 3.76 (dd, 1H, J = 6.4 Hz,
J = 8.3 Hz), 3.72 (dd, 1H, J = 5.2 Hz, J = 10.5 Hz), 3.56 (dd, 1H,
J = 5.2 Hz, J = 10.5 Hz), 2.82–2.41 (m, 16H), 2.19, 2.177, 2.172,
2.171 (4s, each 3H); 13C NMR (CDCl3): d 206.51, 206.36, 206.31,
206.06, 172.18, 172.08, 171.98, 171.84, 109.52, 96.63, 74.38,
69.15, 68.20, 67.97, 67.52, 66.49, 66.45, 61.84, 37.80, 37.78,
37.75, 37.70, 29.79, 29.78, 29.76, 27.92, 27.77, 27.75, 26.71,
25.47. API-ES-MS: m/z calcd for [C32H46O16]NH4þ: 704.3. Found
704.2.
61.21, 26.53, 25.09, 20.68, 20.60, 20.59, 20.51. API-ES-MS: m/z
calcd for [C20H30O12]NH4þ: 480.2. Found 480.2.
4.25. 1,2-O-Isopropylidene-3-O-(2,3,4,6-tetra-O-levulinoyl-b-
galactopyranosyl)-sn-glycerol (30)
D-
To a solution of compound 29 (3.5 g, 7.6 mmol) in anhydrous
MeOH was added NaOMe in MeOH (excess) until the pH of the
solution reached 12 as seen with an indicator paper. After 24 h,
the solution was carefully treated with Dowex 50WX8 (H+) ion ex-
change resin until the pH reached approx. 7, then the solution was
concentrated and the residue dried under vacuum to obtain a
semisolid. A solution of this material in EtOAc (50 mL) was treated
under stirring with levulinic acid (7.1 g, 61 mmol) followed by N,N-
dicyclohexylcarbodiimide (12.5 g, 61 mmol). After 3 h, MeOH
(5 mL) was added and stirring was continued for an additional
2 h. The solids were filtered and the filtrate concentrated followed
by column chromatographic purification using a hexane–EtOAc
10:1?2:1 gradient to yield 30 (4.8 g, 93%) as a syrup: 1H NMR
4.22. 3-O-(2,3,4,6-Tetra-O-levulinoyl-a-D-galactopyranosyl)-sn-
glycerol (25)
A stirred solution of 24 (1.25 g, 1.8 mmol) in a mixture of MeOH
(5 mL) and AcOH (5 mL) was heated at reflux for 5 min when TLC
(EtOAc) indicated disappearance of 24. The solution was concen-
trated and the residue purified by column chromatography using
a 1:1?1:0 EtOAc–hexanes gradient as the eluant to afford 25
(CDCl3):
d 5.36 (d, 1H, J = 3.6 Hz), 5.17 (dd, 1H, J = 8.0 Hz,
J = 10.4 Hz), 5.02 (dd, 1H, J = 3.6 Hz, J = 10.4 Hz), 4.57 (d, 1H,
J = 8.2 Hz), 4.27 (m, 1H), 4.20 (dd, 1H, J = 6.7 Hz), J = 11.4 Hz),
4.11 (dd, 1H, J = 6.7 Hz, J = 11.6 Hz), 4.04 (dd, 1H, J = 6.7 Hz,
J = 8.2 Hz), 3.91–3.87 (m, 2H), 3.83 (dd, 1H, J = 5.9 Hz, J = 8.2 Hz),
3.63 (dd, 1H, J = 6.3 Hz, J = 10.8 Hz), 2.83–2.42 (m, 16H), 2.19,
2.18, 2.17, 2.16, 1.41, 1.34 (6s, 6 ꢂ 3H); 13C NMR (CDCl3): d
207.74, 206.82, 206.64, 205.30, 172.16, 171.89, 171.88, 171.41,
109.29, 101.26, 74.24, 70.82, 70.75, 69.39, 68.85, 67.23, 66.39,
(1.0 g, 83%) as
a d 5.42 (dd, 1H,
syrup: 1H NMR (CDCl3):
J = 1.3 Hz, J = 3.8 Hz), 5.37 (dd, 1H, J = 3.8 Hz, J = 10.6 Hz), 5.14 (d,
1H, J = 3.8 Hz), 5.09 (dd, 1H, J = 3.8 Hz, J = 10.6 Hz), 4.26 (br t, 1H,
J = 6.7 Hz), 4.15–4.07 (m, 2H), 3.92 (m, 1H), 3.81 (dd, 1H,
J = 3.8 Hz, J = 10.7 Hz), 3.72 (dd, 1H, J = 4.3 Hz, J = 11.2 Hz), 3.62
(dd, 1H, J = 5.6 Hz, J = 11.2 Hz), 3.53 (dd, 1H, J = 6.5 Hz,
J = 10.6 Hz), 2.84–2.41 (m, 16H), 2.20 (s, 3H), 2.19 (s, 6H), 2.17 (s,
3H); 13C NMR (CDCl3): d 207.77, 207.20, 206.62, 206.19, 172.18,
171.88, 171.79, 171.73, 96.40, 70.50, 69.93, 68.22, 68.10, 67.30,
66.37, 63.30, 61.92, 37.63, 37.58, 37.50, 29.65, 29.63, 29.60,
29.58, 29.50, 27.94, 27.76, 27.62, 27.59, 27.43. API-ES-MS: m/z
calcd for [C29H42O16]NH4þ: 664.3. Found 664.2.
61.38, 37.85, 37.78, 37.76, 37.68, 29.76, 29.75, 29.67, 27.79,
þ
27.76, 26.67, 25.19. API-ES-MS: m/z calcd for [C32H46O16]NH4
:
704.3. Found 704.3.
4.26. 3-O-(2,3,4,6-Tetra-O-levulinoyl-b-D-galactopyranosyl)-sn-
glycerol (31)
4.23. 1-O-(2,3,4,6-Tetra-O-levulinoyl-a-D-galactopyranosyl)-sn-
glycerol (27)
Compound 30 was treated with acetic acid as described for 24
to afford 31 as 5.36 (d, 1H,
syrup: 1H NMR (CDCl3):
a
d
1H NMR (CDCl3): d 5.42 (d, 1H, J = 3.5 Hz), 5.35 (dd, 1H,
J = 3.5 Hz, J = 10.8 Hz), 5.17 (d, 1H, J = 3.6 Hz), 5.08 (dd, 1H,
J = 3.6 Hz, J = 10.8 Hz), 4.24 (m, 1H), 4.16 (dd, 1H, J = 5.2 Hz,
J = 11.3 Hz), 4.08 (dd, 1H, J = 7.6 Hz, 11.3 Hz), 3.88 (m, 1H), 3.79
(dd, 1H, J = 5.7 Hz, J = 10.4 Hz), 3.72 (d, 1H, J = 4.5 Hz), 3.62 (dd,
1H, J = 5.1 Hz, J = 10.4 Hz), 2.83–2.41 (m, 16H), 2.19 (s, 6H), 2.18
and 2.17 (2s, 3H each); 13C NMR (CDCl3): d 207.75, 207.29,
206.46, 206.03, 172.31, 171.98, 171.93, 171.77, 96.19, 70.15,
69.33, 68.37, 68.28, 67.40, 66.65, 63.33, 62.31, 37.80, 37.76,
37.74, 37.66, 29.86, 29.78, 29.76, 27.87, 27.74, 27.70. API-ES-MS:
m/z calcd for [C29H42O16]NH4þ: 664.3. Found 664.2.
J = 3.7 Hz), 5.17 (dd, 1H, J = 7.9 Hz, J = 10.5 Hz), 5.04 (dd, 1H,
J = 3.4 Hz, J = 10.5 Hz), 4.53 (d, 1H, J = 8.1 Hz), 4.21 (dd, 1H,
J = 7.2 Hz, J = 11.4 Hz), 4.11 (dd, 1H, J = 6.0 Hz, J = 11.4 Hz), 3.94–
3.90 (m, 2H), 3.86 (m, 1H), 3.77 (dd, 1H, J = 3.6 Hz, J = 10.3 Hz),
3.68 (dd, 1H, J = 4.3 Hz, J = 11.5 Hz), 3.65 (dd, 1H, J = 5.1 Hz,
J = 11.5 Hz), 2.91–2.41 (m, 16H), 2.20, 2.19, 2.18, 2.16 (4s,
4 ꢂ 3H); 13C NMR (CDCl3): d 207.71, 206.76, 206.56, 206.09,
172.18, 171.87, 171.72, 171.67, 101.49, 72.04, 70.83, 70.54, 70.34,
68.92, 67.24, 63.24, 61.54, 37.76, 37.74, 37.69, 37.57, 29.76,
29.73, 29.69, 29.61, 27.71, 27.68, 27.66. API-ES-MS: m/z calcd for
[C29H42O16]NH4þ: 664.3. Found 664.5.
4.24. 3-O-(2,3,4,6-Tetra-O-acetyl-b-
isopropylidene-sn-glycerol (29)
D-galactopyranosyl)-1,2-O-
4.27. General procedure for the preparation of singly Gro-
substituted intermediates featuring tetra-levulinoylated Gal
moieties
To a stirred solution of imidate35 28 (4.0 g, 8 mmol) and alcohol
19 (2.1 g, 16 mmol) in anhydrous CH2Cl2 was added at 0 °C TMSOTf
(20 lL). After 1 h the solution was treated with a saturated aque-
To a solution of the diol 3-O-(2,3,4,6-tetra-O-levulinoyl-a-D-
galactopyranosyl)-sn-glycerol 25 (129 mg, 0.2 mmol) in anhydrous
CH2Cl2 (3 mL) was added fatty acid (0.2 mmol) followed by DCC
(60 mg, 0.29 mmol) and a catalytic amount of DMAP at room tem-
perature. After 16 h MeOH (0.5 mL) was added followed by stirring
for 1 h. TLC (CH2Cl2–MeOH 100:4) indicated that most of the start-
ing diol had disappeared and two closely migrating mono-substi-
tuted products had formed. Some less polar material was also
formed that was the doubly-substituted product by mass spec-
trometry. The mixture was concentrated, then CH2Cl2 was added,
followed by removal of the solids by filtration. The residue was ap-
plied to a silica gel column made in CH2Cl2 which was then eluted
ous solution of NaHCO3. The usual processing followed by column
chromatographic purification afforded 29 as a syrup (3.0 g, 80%):
1H NMR (CDCl3): d 5.38 (dd, 1H, J = 1.1 Hz, J = 3.5 Hz), 5.21 (dd,
1H, J = 7.9 Hz, J = 10.4 Hz), 5.01 (dd, 1H, J = 3.5 Hz, J = 10.4 Hz),
4.58 (d, 1H, J = 7.9 Hz), 4.26 (m, 1H), 4.18 (dd, 1H, J = 6.7 Hz,
J = 11.2 Hz), 4.13 (dd, 1H, J = 6.7 Hz, J = 11.2 Hz), 4.02 (dd, 1H,
J = 6.5 Hz, J = 8.3 Hz), 3.92 (dt, 1H, J = 1.1 Hz, J = 6.8 Hz), 3.89 (dd,
1H, J = 4.4 Hz, J = 10.8 Hz), 3.81 (dd, 1H, J = 5.9 Hz, J = 8.3 Hz), 3.63
(dd, 1H, J = 6.0 Hz, J = 10.7 Hz), 2.15, 2.07, 2.05, 1.99, 1.42, 1.35
(6s, 6 ꢂ 3H); 13C NMR (CDCl3): d 170.32, 170.17, 170.08, 169.35,
109.26, 101.35, 74.19, 70.82, 70.66, 69.03, 68.66, 66.96, 66.16,

V. Pozsgay et al. / Carbohydrate Research 346 (2011) 1551–1563
1561
with CH2Cl2–MeOH 100:1 to afford the mono-substituted product
in 65–75% yield, free of the slower-migrating isomer.
J = 10.7 Hz), 3.51 (dd, 1H, J = 6.8 Hz, J = 10.7 Hz), 2.84–2.41 (m,
16H), 2.34 (t, 2H, J = 7.6 Hz), 2.188 (s, 6H), 2.181 (s, 3H), 2.17 (s,
3H), 2.04–1.98 (m, 2H), 1.65–1.59 (m, 2H), 1.36–1.24 (m), 0.89–
0.86 (m, 3H); 13C NMR (CDCl3): d 206.50, 206.42, 206.22, 205.98,
173.78, 172.20, 171.95, 171.80, 171.76, 129.96, 129.72, 96.86,
70.31, 68.82, 68.25, 68.16, 67.33, 66.60, 65.03, 61.99, 53.42,
37.76, 37.72, 37.69, 37.66, 34.08, 31.88, 29.77, 29.76, 29.74,
29.72, 29.69, 29.49, 29.29, 29.18, 29.10, 27.94, 27.74, 27.72,
27.69, 27.19, 27.16, 24.86, 22.66, 14.09. API-ES-MS: m/z calcd for
[C47H74O17]NH4þ: 928.5. Found 928.4.
4.28. General procedure for the preparation of fully substituted
intermediates having two different fatty acyl groups in the Gro
moiety
To a stirred solution of the singly Gro-substituted intermediate
in anhydrous CH2Cl2 was added fatty acid (1.3 equiv) followed by
DCC (1.5 equiv) and a catalytic amount of DMAP. After 3 h, the
solution was treated with MeOH (excess). After 1 h, the mixture
was concentrated. To the residue was added CH2Cl2 followed by re-
moval of the solids by filtration. Column chromatographic purifica-
tion of the residue with a CH2Cl2–MeOH mixture of appropriate
polarity afforded a near quantitative yield of the doubly Gro-
substituted product.
4.33. 3-O-(2,3,4,6-Tetra-O-levulinoyl-a-D-galactopyranosyl)-1-
O-stearoyl-sn-glycerol (34)
1H NMR (CDCl3): d 5.42 (d, 1H, J = 3.6 Hz), 5.36 (dd, 1H,
J = 3.6 Hz, J = 11.0 Hz), 5.16 (d, 1H, J = 3.7 Hz), 5.09 (dd, 1H,
J = 3.7 Hz, J = 11.0 Hz), 4.25 (t, 1H, J = 6.6 Hz), 4.19–4.07 (m, 5H),
3.81 (dd, 1H, J = 3.1 Hz, J = 10.6 Hz), 3.58 (d, 1H, J = 3.9 Hz), 3.51
(dd, 1H, J = 6.7 Hz, J = 10.6 Hz), 2.84–2.41 (m, 16H), 2.34 (t, 2H, J
ꢁ7.6 Hz), 2.19 (s, 6H), 2.18 and 2.17 (2s, 2 ꢂ 3H), 1.62 (m, 2H),
1.38–1.18 (m, 28H), 0.88 (t, 3H, J ꢁ6.7 Hz); 13C NMR (CDCl3): d
207.31, 206.59, 206.37, 205.93, 173.67, 172.07, 171.83, 171.69,
171.63, 96.73, 70.16, 68.65, 68.13, 68.04, 67.25, 66.48, 64.88,
61.86, 37.63, 37.60, 37.57, 37.53, 33.97, 31.77, 29.69–29.02,
27.82, 27.63, 27.61, 27.57, 24.76, 22.54, 13.98. API-ES-MS: m/z
calcd for [C47H76O17]NH4þ: 930.5. Found 930.4.
4.29. General procedure for the preparation of fully protected
intermediates having two identical fatty acyl groups in the Gro
moiety
To a stirred solution of 3-O-(2,3,4,6-tetra-O-levulinoyl-a-D-
galactopyranosyl)-sn-glycerol (25) in anhydrous CH2Cl2 was added
fatty acid (3 equiv) followed by DCC (6 equiv) and a catalytic
amount of DMAP. After 3 h, the solution was treated with MeOH
(excess). After 1 h, the mixture was concentrated and processed
as described in Section 4.28.
4.34. 1-O-(2,3,4,6-Tetra-O-levulinoyl-a-D-galactopyranosyl)-3-
4.30. General procedure for the removal of the levulinoyl
groups
O-palmitoyl-sn-glycerol (35)
1H NMR (CDCl3): d 5.42 (dd, 1H, J = 1.2 Hz, J = 3.4 Hz), 5.34 (dd,
1H, J = 3.4 Hz, J = 10.7 Hz), 5.15 (d, 1H, J = 3.8 Hz), 5.10 (dd, 1H,
J = 3.8 Hz, J = 10.7 Hz), 4.24 (t, 1H, J ꢁ6.5 Hz), 4.18–4.14 (m, 2H),
4.08 (dd, 1H, J = 7.5 Hz, J = 11.4 Hz), 4.05–4.01 (m, 1H), 3.75 (dd,
1H, J = 5.7 Hz, J = 11.0 Hz), 3.60 (dd, 1H, J = 4.4 Hz, J = 10.8 Hz), 3.46
(d, 1H, J = 5.6 Hz), 2.83–2.41 (m, 16H), 2.34 (d, 2H, J ꢁ7.6 Hz), 2.19
(s, 9H), 2.17 (s, 3H), 1.66–1.58 (m, 2H), 1.35–1.21 (m, 24H), 0.88 (t,
3H, J ꢁ7 Hz); 13C NMR (CDCl3): d 207.18, 206.98, 206.45, 206.02,
173.86, 172.23, 171.96, 171.94, 171.76, 96.39, 76.51, 69.53, 68.46,
6.31, 68.13, 67.39, 66.68, 65.12, 62.17, 37.76, 37.71, 37.69, 34.14,
31.92, 29.83, 29.75, 29.69, 29.66, 29.48, 29.35, 29.30, 29.18, 27.92,
27.75, 29.69, 24.91, 22.69, 14.12. API-ES-MS: m/z calcd for
[C45H72O17]NH4þ: 902.5. Found 902.4.
The levulinylated material is dissolved in a 1 M solution of
hydrazine hydrate in pyridine–AcOH (3:2) at rt. After 3 h, the solu-
tion is treated with 2,4-pentanedione (excess). The solution is con-
centrated under reduced pressure and the residue is purified by
silica gel column chromatography. Depending on the acyl groups
attached to the Gro moiety, the product may precipitate as a solid
material during work-up.
4.31. 3-O-(2,3,4,6-Tetra-O-levulinoyl-a-D-galactopyranosyl)-1-
O-palmitoyl-sn-glycerol (32)
1H NMR (CDCl3): d 5.42 (dd, 1H, J = 1.4 Hz, J = 3.5 Hz), 5.36 (dd,
1H, J = 3.5 Hz, J = 10.7 Hz), 5.16 (d, 1H, J = 3.7 Hz), 5.09 (dd, 1H,
J = 3.7 Hz, J = 10.7 Hz), 4.25 (ddd, 1H), 4.18 (dd, 1H, J = 4.6 Hz,
J = 11.3 Hz), 4.13 (dd, 1H, J = 6.1 Hz, J = 11.2 Hz), 4.09 (dd, 1H,
J = 7.0 Hz, J = 11.2 Hz), 3.81 (dd, 1H, J = 3.4 Hz, J = 10.9 Hz), 3.57
(d, 1H, J = 4.4 Hz), 3.51 (dd, 1H, J = 7.0 Hz, J = 10.9 Hz), 2.84–2.41
(m, 16H), 2.34 (dd, 1H, J = 7.3 Hz, J = 7.8 Hz), 2.189, 2.187, 2.181,
2.17 (4s, 4 ꢂ 3H), 1.66–1.58)m, 2H), 1.35–1.20 (m, 24H), 0.91–
0.84 (m, 3H); 13C NMR (CDCl3): d 207.35, 206.63, 206.40, 205.97,
173.71, 172.11, 171.86, 171.72, 171.67, 96.76, 70.18, 68.68, 68.15,
68.07, 67.27, 66.51, 64.91, 61.89, 37.66, 37.62, 37.59, 37.55, 33.99
31.79, 29.66, 29.63, 29.62, 29.56, 29.53, 29.49, 29.43, 29.36,
29.23, 29.17, 29.04, 27.84, 27.66, 27.64, 27.60, 24.78, 22.56,
14.00. API-ES-MS: m/z calcd for [C45H72O17]NH4þ: 902.5. Found
902.4.
4.35. 1-O-(2,3,4,6-Tetra-O-levulinoyl-
O-oleoyl-sn-glycerol (36)
a-D-galactopyranosyl)-3-
1H NMR (CDCl3): d 5.42 (d, 1H, J = 3.3 Hz), 5.37–5.30 (m, 3H),
5.14 (d, 1H, J = 3.6 Hz), 5.10 (dd, 1H, J = 3.6 Hz, J = 10.6 Hz), 3.24
(m, 1H), 4.20–4.11 (m, 3H), 4.07 (dd, 1H, J = 7.4 Hz, J = 11.3 Hz),
3.76 (dd, 1H, J = 5.5 Hz, J = 10.8 Hz), 2.83–2.41 (m, 16H), 2.32 (t,
2H, J ꢁ7.6 Hz), 2.19 (s, 9H), 2.17 (s, 3H), 2.03–1.99 (m 4H), 1.67–
1.28 (m, 2H), 1.39–1.20 (m, 20H), 0.89–0.86 (m, 3H); 13C NMR
(CDCl3): d 207.13, 206.94, 206.42, 205.93, 173.74, 172.16, 171.90,
171.87, 171.69, 129.91, 129.67, 96.31, 69.44, 68.36, 68.22, 58.06,
67.32, 66.59, 65.05, 62/08, 37.69, 37.68, 37.64, 37.60, 34.04,
35.00, 31.81, 29.74, 29.68, 29.64, 29.43, 29.23, 29.18, 29.12,
29.06, 29.05, 27.85, 27.69, 27.67, 27.62, 27.13, 27.10, 24.82,
22.59, 14.04. API-ES-MS: m/z calcd for [C47H74O17]NH4þ: 928.5.
Found 928.4.
4.32. 3-O-(2,3,4,6-Tetra-O-levulinoyl-a-D-galactopyranosyl)-1-
O-oleoyl-sn-glycerol (33)
1H NMR (CDCl3): d 5.42 (d, 1H, J = 3.4 Hz), 5.36 (dd, 1H,
J = 3.6 Hz, J = 10.4 Hz), 5.36–5.32 (m, 2H), 5.16 (d, 1H, J = 3.7 Hz),
5.09 (dd, 1H, J = 3.7 Hz, J = 10.7 Hz), 4.24 (ddd,1H), 4.18 (dd, 1H,
J = 4.4 Hz, J = 11.3 Hz), 4.13 (dd, 1H, J = 5.8 Hz, J = 10.7 Hz), 4.09
(dd, 1H, J = 7.2 Hz), J = 11.3 Hz), 3.81 (dd, 1H, J = 3.4 Hz,
4.36. 3-O-(2,3,4,6-Tetra-O-levulinoyl-
di-O-palmitoyl-sn-glycerol (37)
a-D-galactopyranosyl)-1,2-
1H NMR (CDCl3): d 5.43 (dd, 1H, J = 1.2 Hz, J = 3.4 Hz), 5.32 (dd,
1H, J = 3.4 Hz, J = 10.8 Hz), 5.22 (m, 1H), 5.12 (dd, 1H, J = 3.7 Hz,

1562
V. Pozsgay et al. / Carbohydrate Research 346 (2011) 1551–1563
J = 10.8 Hz), 5.06 (d, 1H, J = 3.7 Hz), 4.37 (dd, 1H, 3.9 Hz,
J = 11.8 Hz), 4.20 (ddd, 1H), 4.18–4.06 (m, 3H), 3.83 (dd, 1H,
J = 4.5 Hz, J = 11.2 Hz), 3.65 (dd, 1H, J = 5.0 Hz, J = 11.0 Hz), 2.83–
2.42 (m, 16H), 2.33 and 2.30 (2t, 2 ꢂ 2H, J ꢁ7.8 Hz), 1.66–1.56
(m, 4H), 1.36–1.20 (m, 44H), 0.92–0.85 (m, 6H); 13C NMR (CDCl3):
d 206.40, 206.35, 206.30, 206.00, 173.25, 172.97, 172.09, 172.08,
171.90, 171.79, 96.58, 69.73, 67.99, 67.81, 67.44, 66.63, 66.47,
62.19, 61.65, 37.73, 37.68, 37.66, 34.20, 34.01, 31.88, 29.69,
29.66, 29.61, 29.49, 29.47, 29.31, 29.29, 29.27, 29.11, 29.10,
27.81, 27.74, 27.71, 27.65, 24.89, 24.84, 22.64, 14.07. API-ES-MS:
m/z calcd for [C61H102O18]NH4þ: 1140.7. Found 1160.6.
4.40. 3-O-(2,3,4,6-Tetra-O-levulinoyl-
di-O-stearoyl-sn-glycerol (41)
a-D-galactopyranosyl)-1,2-
1H NMR (CDCl3): d 5.43 (d, 1H, J = 3.4 Hz), 5.32 (dd, 1H,
J = 3.4 Hz, J = 10.7 Hz), 5.22 (m, 1H), 5.12 (dd, 1H, J = 3.6 Hz,
J = 10.7 Hz), 5.06 (d, 1H, J = 3.6 Hz), 4.37 (dd, 1H, J = 3.9 Hz,
J = 12.0 Hz), 4.22–4.06 (m, 4H), 3.83 (d, 1H, J = 4.6 Hz, J = 11.2 Hz),
3.66 (dd, 1H, J = 5.1 Hz, J = 11.2 Hz), 2.86–2.41 (m, 16H), 2.33 and
2.30 (2t, 2 ꢂ 2H, J ꢁ7.5 Hz), 2.18 and 2.17 (2s, 2 ꢂ 3H), 2.16 (s,
6H), 1.66–1.56 (m, 4H), 1.59–1.21 (m, 56H), 0.88 (t, 6H,
J
ꢁ7.0 Hz); 13C NMR (CDCl3): d 206.32, 206.10, 205.87, 173.18,
172.91, 172.03, 172.01, 171.85, 171.74, 96.52, 69.68, 67.94, 67.76,
67.39, 66.58, 66.42, 62.14, 61.60, 37.67, 37.62, 37.61, 34.14,
4.37. 3-O-(2,3,4,6-Tetra-O-levulinoyl-a-D-galactopyranosyl)-1,2-
di-O-oleoyl-sn-glycerol (38)
33.96, 31.82, 29.74–29.04, 27.76, 27.69, 27.66, 27.60, 24.84,
þ
24.78, 22.59, 14.02. API-ES-MS: m/z calcd for [C65H110O18]NH4
:
1H NMR (CDCl3): d 5.43 (d, 1H, J = 3.7 Hz), 5.38–5.30 (m 5H),
5.22 (m, 1H), 5.13 (dd, 1H, J = 3.7 Hz, J = 10.8 Hz), 5.06 (d, 1H,
J = 3.7 Hz), 4.37 (dd, 1H, J = 4.0 Hz, J = 11.8 Hz), 4.22–4.06 (m, 4H),
3.83 (dd, 1H, J = 4.5 Hz, J = 11.2 Hz), 3.65 (dd, 1H, J = 5.0 Hz,
J = 11.2 Hz), 2.83–2.42 (m, 16H), 2.33 and 2.30 (2t, 2 ꢂ 3H, J
ꢁ7.7 Hz), 2.18 (s, 3H), 2.177 (s, 3H), 2.170 (s, 6H), 2.05–1.97 (m,
1196.8. Found 1197.0.
4.41. 1-O-(2,3,4,6-Tetra-O-levulinoyl-
di-O-palmitoyl-sn-glycerol (42)
a-D-galactopyranosyl)-2,3-
1H NMR (CDCl3): d 5.42 (dd, 1H, J = 1.3 Hz, J = 3.5 Hz), 5.30 (dd,
1H, J = 3.5 Hz, J = 10.7 Hz), 5.20 (m, 1H), 5.12 (dd, 1H, J = 3.7 Hz,
J = 10.7 Hz), 5.08 (d, 1H, J = 3.7 Hz), 4.32 (dd, 1H, J = 3.7 Hz,
J = 12.5 Hz), 4.21–4.17 (m, 2H), 4.11 (s, 1H), 4.09 (m, 1H), 3.82
(dd, 1H, J = 5.4 Hz, J = 11.1 Hz), 3.63 (dd, 1H, J = 5.4 Hz,
J = 11.1 Hz), 2.83–2.42 (m, 16H), 2.31 and 2.30 (2t, 2 ꢂ 2H, J
ꢁ7.5 Hz), 2.18, 2.176, 2.174, 2.171 (4s, 3 ꢂ 3H), 1.65–1.56 (m,
4H), 1.34–1.21 (m, 48H), 0.89–0.86 (m, 6H); 13C NMR (CDCl3): d
206.37, 206.26, 206.20, 206.19, 173.21, 172.78, 172.07, 172.03,
171.85, 171.71, 96.61, 69.67, 67.97, 67.75, 67.31, 66.55, 66.28,
62.18, 61.63, 37.70, 37.68, 37.63, 37.61, 34.10, 33.97, 31.83,
29.66, 29.64, 29.61, 29.58, 29.57, 29.43, 29.42, 29.27, 29.23,
29.22, 29.06, 29.04, 27.77, 27.69, 27.66, 27.63, 24.81, 24.78,
22.60, 14.03. API-ES-MS: m/z calcd for [C61H102O18]NH4þ: 1140.7.
Found 1140.6.
8H), 1.66–1.56 (m, 4H), 1.38–1.20 (m, 42H), 0.88 (t, 6H,
J
ꢁ7.0 Hz); 13C NMR (CDCl3): d 206.35, 206.12, 205.90, 173.17,
172.89, 172.04, 172.03, 171.87, 171.75, 129.90, 129.63, 96.55,
69.70, 67.95, 67.77, 67.40, 66.58, 66.44, 62.15, 61.61, 37.68,
37.64, 37.62, 34.13, 33.94, 31.81, 29.67, 29.66, 29.65, 29.43,
29.22, 29.15, 29.13, 29.09, 29.06, 29.03, 29.02, 27.77, 27.70,
27.67, 27.62, 27.13, 27.11, 27.10, 24.82, 24.77, 22.59, 14.03. API-
ES-MS: m/z calcd for [C65H106O18]NH4þ: 1192.8. Found 1192.6.
4.38. 3-O-(2,3,4,6-Tetra-O-levulinoyl-a-D-galactopyranosyl)-2-
O-oleoyl-1-O-palmitoyl-sn-glycerol (39)
1H NMR (CDCl3): d 5.43 (d, 1H, J = 1.2 Hz, J = 3.5 Hz), 5.38–5.30
(m, 3H), 5.22 (m, 1H), 5.12 (dd, 1H, J = 3.7 Hz, J = 10.7 Hz), 4.37
(dd, 1H, J = 4.0 Hz, 12.0 Hz), 4.22–4.06 (m, 4H), 3.83 (dd, 1H,
J = 4.6 Hz, J = 11.1 Hz), 3.65 (4.9 Hz, J = 11.3 Hz), 2.82–2.42 (m,
16H), 2.33 and 2.30 (2t, s ꢂ 2H, J = 7.5 Hz for each), 2.05–1.96 (m,
4H), 1.65–1.55 (m, 4H), 1.37–1.19 (m), 0.88 (t, 6H, J = 6.7 Hz); 13C
NMR (CDCl3): d 206.41, 206.18, 206.96, 173.25, 172.94, 172.09,
172.08, 171.91, 171.80, 129.94, 129.66, 96.58, 69.73, 67.99, 67.81,
67.44, 66.61, 66.47, 62.17, 61.65, 37.72, 37.67, 37.66, 34.17,
34.00, 31.87, 31.85, 29.71, 29.70, 29.69, 29.68, 29.65, 29.62,
29.60, 29.47, 29.46, 29.31, 29.26, 29.19, 29.12, 29.10, 29.06,
27.80, 27.73, 27.70, 27.65, 27.17, 27.15, 24.86, 24.83, 22.64,
22.63, 14.06. API-ES-MS: m/z calcd for [C63H104O18]NH4þ: 1166.8.
Found 1166.6.
4.42. 1-O-(2,3,4,6-Tetra-O-levulinoyl-
di-O-oleoyl-sn-glycerol (43)
a-D-galactopyranosyl)-2,3-
1H NMR (CDCl3): d 5.42 (dd, 1H, J = 1.1 Hz, J = 3.4 Hz), 5.37–5.31
(m, 4H), 5.30 (dd, 1H, J = 3.4 Hz, J = 10.7 Hz), 5.20 (m, 1H), 5.12 (dd,
1H, J = 3.8 Hz, J = 10.7 Hz), 5.08 (d, 1H, J = 3.8 Hz), 4.32 (dd, 1H,
J = 3.8 Hz, J = 12.0 Hz), 4.1–4.17 (m, 2H), 4.09 (m, 1H), 3.82 (dd,
1H, 5.3 Hz, J = 11.2 Hz), 3.63 (dd, 1H, J = 3.63 Hz, J = 11.2 Hz),
2.82–2.42 (m, 16H), 2.31 and 2.30 (2t, 2 ꢂ 2H, J ꢁ7.5 Hz), 2.183,
2/176, 2.173, 2.170 (4s, 3H each), 2.03–1.99 (m, 4H), 1.36–1.22
(m, 40H), 0.89–0.86 (m, 6H); 13C NMR (CDCl3): d 206.48, 206.25,
206.20, 206.17, 173.21, 172.77, 172.09, 172.05, 171.87, 172.77,
172.09, 172.05, 171.88, 171.74, 129.92, 129.64, 96.63, 69.70,
67.99, 7.77, 67.33, 66.57, 66.30, 62.20, 61.64, 37.72, 37.70, 37.66,
37.63, 34.09, 33.96, 31.83, 29.69, 29.66, 29.45, 29.24, 29.15,
29.09, 29.07, 29.05, 29.02, 27.79, 27.70, 27.68, 27.65, 27.15,
27.12, 27.11, 24.80, 24.78, 22.60, 14.04. API-ES-MS: m/z calcd for
[C65H106O18]NH4þ: 1192.8. Found 1192.6.
4.39. 3-O-(2,3,4,6-Tetra-O-levulinoyl-a-D-galactopyranosyl)-1-
O-oleoyl-2-O-palmitoyl-sn-glycerol (40)
1H NMR (CDCl3): d 5.43 (d, 1H, J = 3.7 Hz), 5.36–5.29 (m, 3H),
5.22 (m, 1H), 5.12 (dd, 1H, J = 3.9 Hz, J = 10.6 Hz), 5.06 (d, 1H,
J = 3.5 Hz), 4.36 (dd, 1H, J = 4.1 Hz, J = 12.3 Hz), 4.21–4.06 (m, 4H),
3.83 (dd, 1H, J = 4.5 Hz, J = 11.3 Hz), 3.65 (dd, 1H, J = 5.0 Hz,
J = 11.0 Hz), 2.82–2.42 (m, 16H), 2.18 and 2.176 (2s, 2 ꢂ 3H),
2.172 (s, 3H), 2.33 and 2.30 (2t, 2 ꢂ 2H, J = 7.5 Hz), 2.01 (m, 4H),
1.62 (m, 4H), 1.37–1.21 (m), 0.88 (t, 6H, J = 7.3 Hz); 13C NMR
(CDCl3): d 206.45, 206.23, 206.22, 205.99, 173.29, 173.03, 172.15,
172.14, 171.97, 171.86, 130.01, 129.73, 96.64, 69.79, 68.05, 67.87,
67.50, 66.70, 66.53, 62.26, 61.71, 37.79, 36.74, 37.73, 34.25,
34.05, 31.93, 31.91, 31.79, 29.77, 29.75, 29.72, 29.71, 29.68,
29.55, 29.54, 29.46, 29.38, 29.35, 29.34, 29.33, 29.24, 29.17,
4.43. 1-O-(2,3,4,6-Tetra-O-levulinoyl-
O-oleoyl-3-O-palmitoyl-sn-glycerol (44)
a-D-galactopyranosyl)-2-
1H NMR (CDCl3): d 5.43 (d, 1H, J = 3.4 Hz), 5.37–5.27 (m, 3H),
5.20 (m, 1H), 5.12 (dd, 1H, J = 3.7 Hz, J = 10.7 Hz), 5.08 (d, 1H,
J = 3.7 Hz), 4.32 (dd, 1H, J = 3.7 Hz, J = 11.9 Hz), 4.22–4.16 (m, 2H),
4.12–4.08 (m, 2H), 3.82 (dd, 1H, J = 5.3 Hz, J = 11.1 Hz), 3.63 (dd,
1H, J = 5.3 Hz, J = 11.1 Hz), 2.82–2.42 (m, 16H), 2.31 and 2.30 (2t,
2H, J ꢁ7.6 Hz), 2.180 (s, 3H), 2.175 (s, 6H), 2.172 (s, 3H), 2.06–
1.98 (m, 4H), 1.66–1.56 (m, 4H), 1.38–1.20 (m, 44H), 0.92–086
(m, 3H); 13C NMR (CDCl3): d 206.44, 206.26, 206.18, 205.99,
29.14, 27.86, 27.80, 27.76, 27.71, 27.23, 27.20, 24.95, 24.86,
þ
22.70, 22.69, 14.13. API-ES-MS: m/z calcd for [C63H104O18]NH4
:
1166.8. Found 1166.6.

V. Pozsgay et al. / Carbohydrate Research 346 (2011) 1551–1563
1563
173.33, 172.86, 172.18, 172.14, 171.96, 171.83, 130.01, 129.72,
References
96.73, 69.78, 68.07, 67.86, 67.41, 66.65, 66.38, 62.27, 61.73,
37.81, 37.79, 37.74, 37.72, 34.18, 34.06, 31.93, 31.91, 29.78,
29.76, 29.71, 29.69, 29.67, 29.54, 29.52, 29.37, 29.34, 29.33,
1. Burgdorfer, W.; Barbour, A. G.; Hayes, S. F.; Benach, J. L.; Grunwaldt, E.; Davis, J.
P. Science 1982, 216, 1317–1319.
2. http://www.cdc.gov.
3. Bacon, R. M.; Kugeler, K. J.; Mead, P. S. MMWR 2008, 57, 1–9.
4. Steere, A. C.; Sikand, V. K.; Meurice, F.; Parenti, D. L.; Fikrig, E.; Schoen, R. T.;
Nowakowski, J.; Schmid, C. H.; Laukamp, S.; Buscarino, C.; Krause, D. S.Lyme Dis
Vaccine Study Grp N. Eng. J. Med. 1998, 339, 209–215.
29.24, 29.18, 29.11, 27.87, 27.78, 27.76, 27.73, 27.24, 27.21,
þ
24.89, 22.70, 14.13. API-ES-MS: m/z calcd for [C63H104O18]NH4
:
1166.8. Found 1166.6.
5. Gross, D. M.; Forsthuber, T.; Tary-Lehmann, M.; Etling, C.; Ito, K.; Nagy, Z. A.;
Field, J. A.; Steere, A. C.; Huber, B. T. Science 1998, 281, 703–706.
6. Nigrovic, L. E.; Thompson, K. M. Epidemiol. Infect. 2007, 135, 1–8.
7. Ball, R.; Shadomy, S. V.; Meyer, A.; Huber, B. T.; Leffell, M. S.; Zachary, A.;
Belotto, M.; Hilton, E.; Bryant-Genevier, M.; Schriefer, M. E.; Miller, F. W.;
Braun, M. M. Arthritis Rheum. 2009, 60, 1179–1186.
8. Shen, A. K.; Mead, P. S.; Beard, C. Clin. Infect. Dis. 2011, 52, S247–S252.
9. Plotkin, S. A. Clin. Infect. Dis. 2011, 52, S271–S275.
10. Takayama, K.; Rothenberg, R. J.; Barbour, A. G. Infect. Immun. 1987, 55, 2311–
2313.
11. Ben-Menachem, G.; Kubler-Kielb, J.; Coxon, B.; Yergey, A.; Schneerson, R. Proc.
Natl. Acad. Sci. U.S.A. 2003, 100, 7913–7918.
12. Hossain, H.; Wellensiek, H. J.; Geyer, R.; Lochnit, G. Biochimie 2001, 83, 683–
692.
13. Schroder, N. W. J.; Schombel, U.; Heine, H.; Gobel, U. B.; Zähringer, U.;
Schumann, R. R. J. Biol. Chem. 2003, 278, 33645–33653.
14. Stubs, G.; Fingerle, V.; Wilske, B.; Gobel, U. B.; Zähringer, U.; Schumann, R. R.;
Schroder, N. W. J. J. Biol. Chem. 2009, 284, 13326–13334.
4.44. 1-O-(2,3,4,6-Tetra-O-levulinoyl-
O-oleoyl-2-O-palmitoyl-sn-glycerol (45)
a-D-galactopyranosyl)-3-
1H NMR (CDCl3): d 5.42 (d, 1H, J = 3.4 Hz), 5.38–5.32 (m, 2H),
5.30 (dd, 1H, J = 3.4 Hz, J = 10.6 Hz), 5.20 (m, 1H), 5.12 (dd, 1H,
J = 3.6 Hz, J = 10.6 Hz), 5.08 (dd, 1H, J = 3.6 Hz), 4.32 (dd, 1H,
J = 3.6 Hz, J = 11.8 Hz), 4.22–4.17 (m, 2H), 4.12–4.08 (m, 2H), 3.82
(dd, 1H, J = 5.3 Hz, J = 11.1 Hz), 3.63 (dd, 1H, J = 5.3 Hz,
J = 11.1 Hz), 2.83–2.42 (m, 16H), 2.307 and 2.305 (2t, 2 ꢂ 2H, J
ꢁ7.4 Hz each), 2.18 (s, 3H), 2.175 (s, 6H), 2.171 (s, 3H), 2.06–1.97
(m, 4H), 1.67–1.56 (m, 4H), 1.42–1.19 (m, 44H), 0.94–0.85 (m,
6H); 13C NMR (CDCl3): d 206.44, 206.26, 206.18, 205.99, 173.29,
172.89, 172.18, 172.14, 171.96, 171.82, 130.01, 129.73, 96.72,
69.77, 68.07, 67.86, 67.41, 66.65, 66.39, 62.29, 61.73, 37.81,
37.79, 37.74, 37.72, 34.20, 34.05, 31.93, 31.91, 29.78, 29.75,
29.72, 29.70, 29.67, 29.54, 29.37, 29.33, 29.23, 29.16, 29.14,
27.87, 27.78, 27.76, 27.73, 27.23, 27.20, 24.91, 24.87, 22.70,
22.69, 14.13. API-ES-MS: m/z calcd for [C63H104O18]NH4þ: 1166.8.
Found 1166.6.
15. Fraser, C. M.; Casjens, S.; Huang, W. M.; Sutton, G. G.; Clayton, R.; Lathigra, R.;
White, O.; Ketchum, K. A.; Dodson, R.; Hickey, E. K.; Gwinn, M.; Dougherty, B.;
Tomb, J. F.; Fleischmann, R. D.; Richardson, D.; Peterson, J.; Kerlavage, A. R.;
Quackenbush, J.; Salzberg, S.; Hanson, M.; van Vugt, R.; Palmer, N.; Adams, M.
D.; Gocayne, J.; Weidman, J.; Utterback, T.; Watthey, L.; McDonald, L.; Artiach,
P.; Bowman, C.; Garland, S.; Fujii, C.; Cotton, M. D.; Horst, K.; Roberts, K.; Hatch,
B.; Smith, H. O.; Venter, J. C. Nature 1997, 390, 580–586.
16. Jones, K. L.; Seward, R. J.; Ben-Menachem, G.; Glickstein, L. J.; Costello, C. E.;
Steere, A. C. Clin. Immunol. 2009, 132, 93–102.
17. Wu, D.; Xing, G. W.; Poles, M. A.; Horowitz, A.; Kinjo, Y.; Sullivan, B.; Bodmer-
Narkevitch, V.; Plettenburg, O.; Kronenberg, M.; Tsuji, M.; Ho, D. D.; Wong, C. H.
Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 1351–1356.
18. Kinjo, Y.; Tupin, E.; Wu, D.; Fujio, M.; Garcia-Navarro, R.; Benhnia, M. R. E. I.;
Zajonc, D. M.; Ben-Menachem, G.; Ainge, G. D.; Painter, G. F.; Khurana, A.;
Hoebe, K.; Behar, S. M.; Beutler, B.; Wilson, I. A.; Tsuji, M.; Sellati, T. J.; Wong, C.
H.; Kronenberg, M. Nat. Immunol. 2006, 7, 978–986.
19. Du, W.; Kulkarni, S. S.; Gervay-Hague, J. Chem. Commun. 2007, 2336–2338.
20. Schombs, M.; Park, F. E.; Du, W. J.; Kulkarni, S. S.; Gervay-Hague, J. J. Org. Chem.
2010, 75, 4891–4898.
21. Xia, J.; Hui, Y. Z. Tetrahedron: Asymmetry 1997, 8, 3019–3021.
22. Fraser, B. H.; Perlmutter, P.; Wijesundera, C. J. Am. Oil Chem. Soc. 2007, 84, 11–
21.
23. Andrews, P. C.; Fraser, B. H.; Junk, P. C.; Massi, M.; Perlmutter, P.; Thienthong,
N.; Wijesundera, C. Tetrahedron 2008, 64, 9197–9202.
24. Fureby, A. M.; Virto, C.; Adlercreutz, P.; Mattiasson, B. Biocatal. Biotransform.
1996, 14, 89–111.
25. Greimel, P.; Ito, Y. Tetrahedron Lett. 2008, 49, 3562–3566.
26. Martin, S. F.; Josey, J. A.; Wong, Y. L.; Dean, D. W. J. Org. Chem. 1994, 59, 4805–
4820.
4.45. 3-O-a-D-Galactopyranosyl)-1,2-di-O-oleoyl-DL-glycerol (46)
1H NMR (CDCl3): d 5.39–5.28 (m, 5H), 4.40 (dd, 1H, 3.6 Hz,
J = 12.5 Hz), 4.28 (d, 1H, J = 7.6 Hz), 4.21 (dd, 1H, J = 6.7 Hz,
J = 12.0 Hz, 4.14 (dd, 1H, J = 6.0 Hz, J = 12.0 Hz), 4.02–3.84 (m,
4H), 3.77 (dd, 1H, J = 6.4 Hz, J = 11.5 Hz), 3.75 (dd, 1H, J = 6.4 Hz,
J = 11.5 Hz), 3.66 (t, 1H, J = 8.4 Hz), 3.61–3.57 (m, 1H), 3.55 (m,
2H), 2.35–2.29 (m, 4H), 2.06–1.96 (m, 8H), 1.67–1.54 (m, 6H),
1.38–1.19 (m, 38H), 0.88 (t, 6H, J ꢁ7.2 Hz); 13C NMR (CDCl3): d
173.76, 173.70, 173.52, 173.51, 130.06, 129.72, 104.06, 103.65,
74.53, 74.40, 73.46, 73.36, 71.76, 71.67, 70.19, 70.14, 69.58,
68.49, 68.26, 63.02, 62.96, 62.70, 62.44, 34.33, 34.31, 34.14,
34.13, 31.92, 29.79, 29.74, 29.55, 29.35, 29.34, 29.21, 29.15,
29.12, 29.08, 27.25, 27.20, 24.91, 24.86, 22.70, 14.13. API-ES-MS:
m/z calcd for [C45H82O10]NH4þ: 800.6. Found 800.5.
27. Rosseto, R.; Bibak, N.; Hajdu, J. Org. Biomol. Chem. 2006, 4, 2358–2360.
28. Pozsgay, V.; Kubler-Kielb, J.; Coxon, B.; Ekborg, G. Tetrahedron 2005, 61, 10470–
10481.
5. Note added after the review process
29. Pozsgay, V.; Kubler-Kielb, J. Carbohydr. Res. 2007, 342, 621–626.
30. Khiar, N.; Martin-Lomas, M. J. Org. Chem. 1995, 60, 7017–7021.
31. Osswald, M.; Lang, U.; Friedrich-Bochnitschek, S.; Pfrengle, W.; Kunz, H. Z.
Naturforsch., B 2003, 58, 764–774.
32. van Boom, J. H.; Burgers, P. M. J. Tetrahedron Lett. 1976, 4875–4878.
33. Takatani, M.; Matsuo, I.; Ito, Y. Carbohydr. Res. 2003, 338, 1073–1081.
34. Pozsgay, V. Tetrahedron: Asymmetry 2000, 11, 151–172.
35. Eisele, T.; Windmüller, R.; Schmidt, R. R. Carbohydr. Res. 1998, 306, 81–91.
36. Cinco, M.; Banfi, E.; Balanzin, D.; Godeas, C.; Panfili, E. FEMS Microbiol. Immunol.
1991, 3, 33–38.
An anonymous referee suggested the use of Lemieux’s halide-
ion catalysis protocol. Accordingly, tetra-O-benzyl-D-galactopyran-
osyl bromide, obtained in situ by the action of bromine on the cor-
responding ethylthio-galactoside, was treated with 1,2-O-
isopropylidene-sn-glycerol (6 molar equiv) in the presence of
Bu4NBr and Hünig’s base for 4 days at rt to afford the expected
-galactopyanosyl-glycerol derivative together with an insepara-
ble minor product (<5%, NMR) in a combined yield of 78% yield
(for two steps).
a
37. Eiffert, H.; Lotter, H.; Jareckikhan, K.; Thomssen, R. Med. Microbiol. Immunol.
1991, 180, 229–237.
38. Wheeler, C. M.; Monco, J. C. G.; Benach, J. L.; Golightly, M. G.; Habicht, G. S.;
Steere, A. C. J. Infect. Dis. 1993, 167, 665–674.
39. Wang, J.; Li, Y. L.; Kinjo, Y.; Mac, T. T.; Gibson, D.; Painter, G. F.; Kronenberg, M.;
Zajonc, D. M. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 1535–1540.
40. Chu, C.; Liu, B.; Watson, D.; Szu, S.; Bryla, D.; Shiloach, J.; Schneerson, R.;
Robbins, J. B. Infect. Immun. 1991, 59, 4450–4458.
Acknowledgments
This work was supported by the intramural programs of the Eu-
nice Kennedy Shriver National Institute of Child Health and Human
Development, and the National Institute of Allergy and Infectious
Diseases, National Institutes of Health, Bethesda, MD.
41. Crich, D.; Smith, M.; Yao, Q. J.; Picione, J. Synthesis-Stuttgart 2001, 323–326.