Chemical synthesis of a glycolipid library by a solid-phase strategy allows elucidation of the structural specificity of immunostimulation by rhamnolipids

Article abstract of DOI:10.1002/chem.200600482

The first synthesis of a glycolipid library by hydrophobically assisted switching phase (HASP) synthesis is described. HASP synthesis enables flexible switching between solution-phase steps and solid-supported reactions conducted with molecules attached to a hydrophobic silica support. A library of glycolipids derived from the lead compound 1 - a strongly immunostimulatory rhamnolipid - with variations in the carbohydrate part, the lipid components, and the stereochemistry of the 3-hydroxy fatty acids was designed and synthesized. The enantiose lective synthesis of the 3-hydroxy fatty acid building blocks was achieved by employing asymmetric hydrogenation of 3-oxo fatty acids. Glycolipids were prepared by this approach without any intermediary isolation steps, mostly in excellent yields. Final deprotection to the carboxylic acids was accomplished by enzymatic ester cleavage. All prepared rhamnolipids were tested for their immunostimulatory properties against human monocyte cells by assaying the secretion of the cytokine tumor necrosis factor α (TNFα) into the medium. The observed structure-activity relationships of rhamnolipids indicate a specific, recognition-based mode of action, with small structural variations in the rhamnolipids resulting in strong effects on the immunostimulatory activities of the rhamnolipids at low micromolar concentrations.

Full text of DOI:10.1002/chem.200600482

DOI: 10.1002/chem.200600482
Chemical Synthesis of a Glycolipid Library by a Solid-Phase Strategy Allows
Elucidation of the Structural Specificity of Immunostimulation by
Rhamnolipids**
Jörg Bauer,[a, b] Klaus Brandenburg,[c] Ulrich Zähringer,[c] and Jörg Rademann*[a, d]
Abstract: The first synthesis of a glyco-
lipid library by hydrophobically assist-
ed switching phase (HASP) synthesis is
lective synthesis of the 3-hydroxy fatty
acid building blocks was achieved by
employing asymmetric hydrogenation
of 3-oxo fatty acids. Glycolipids were
prepared by this approach without any
intermediary isolation steps, mostly in
excellent yields. Final deprotection to
the carboxylic acids was accomplished
by enzymatic ester cleavage. All pre-
pared rhamnolipids were tested for
their immunostimulatory properties
against human monocyte cells by assay-
ing the secretion of the cytokine tumor
necrosis factor a (TNFa) into the
medium. The observed structure–activi-
ty relationships of rhamnolipids indi-
cate a specific, recognition-based mode
of action, with small structural varia-
tions in the rhamnolipids resulting in
strong effects on the immunostimulato-
ry activities of the rhamnolipids at low
micromolar concentrations.
described. HASP synthesis enables
A
flexible switching between solution-
phase steps and solid-supported reac-
tions conducted with molecules attach-
ed to a hydrophobic silica support. A
library of glycolipids derived from the
lead compound 1—a strongly immu-
nostimulatory rhamnolipid—with varia-
tions in the carbohydrate part, the lipid
components, and the stereochemistry
of the 3-hydroxy fatty acids was de-
signed and synthesized. The enantiose-
Keywords: combinatorial chemis-
try · glycolipids · innate immunity ·
rhamnolipids · solid-phase synthesis
Introduction
Glycolipids are structurally heterogeneous membrane com-
ponents found in all species from bacteria to man. Bacterial
glycolipids, thanks to their enormous structural diversity,
serve as efficient agonists (pathogen-associated molecular
patterns, PAMPs)[1] for the receptors of the innate immune
system (pathogen recognition receptors, PRRs).[2] The im-
munostimulatory effects of glycolipids can lead to septic
shock[3] and are mediated in a very specific way through the
toll-like receptors TLR-2 and TLR-4.[4] Strong immunosti-
mulatory properties were recently discovered in the rham-
nolipid RL-2,214,14 (1), an l-rhamnose-containing glycolipid,
triggering the excretion of cytokines[5a] and antibacterial
peptides[5b] in human cells. The activation was independent
of the known glycolipid receptors TLR-2 and TLR-4,[5a] sug-
gesting a relevant alternative antibacterial strategy of the
innate immune system, such as the activation of natural
killer T cells by a-galactosylceramide.[2a]
[a] Dr. J. Bauer, Prof. Dr. J. Rademann
Medicinal Chemistry Division
Leibniz Institute for Molecular Pharmacology (FMP)
Robert-Rössle-Str. 10, 13125 Berlin (Germany)
Fax : (+49)30-94062901
E-mail: rademann@fmp-berlin.de
[b] Dr. J. Bauer
Institute for Organic Chemistry
Eberhard-Karls-University Tübingen
Auf der Morgenstelle 18, 72076 Tübingen (Germany)
[c] Prof. Dr. K. Brandenburg, Prof. Dr. U. Zähringer
Research Center Borstel
Leibniz-Center for Medicine and Biosciences
Parkallee 1–40, 23845 Borstel (Germany)
[d] Prof. Dr. J. Rademann
Free University Berlin, Institute for Chemistry and Biochemistry
Takusstr. 3, 14195 Berlin (Germany)
[**] Essential parts of this work were presented at Eurocombi3 in Win-
chester (UK), July 18–21, 2005 as an oral contribution by J.B., who
received a “Young Investigators Presentation Award”.
For investigation of the biological structure–activity rela-
tionships of rhamnolipids, a synthetic rhamnolipid library
featuring systematic variation of the lipid and of the carbo-
hydrate parts was an essential tool. Carbohydrate-based li-
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
7116
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7116 – 7124

FULL PAPER
braries are difficult to obtain, however, and suitable meth-
ods to achieve them require further development.[6] The syn-
thesis of glycolipids is even more demanding, due to the am-
phiphilic character hampering all synthesis and isolation
steps.[7] No example of a solid-supported library synthesis of
rhamnolipids (or any other glycolipids) has been reported so
2,212,12-COOH, 30), is based on two b-(R)-hydroxymyristic
acid components [(14:0(3-OH)] connected through an ester
N
linkage.[5a,10] We designed a rhamnolipid library representing
variations in the amphiphilic character of such molecules by
introducing specific modifications both in the fatty acid part
and in the sugar component. This was achieved particularly
through variation in the number, lengths, and stereochemis-
try of the b-hydroxylated carboxylic acids and in the
number of sugars, and also through reduction of the free
carboxylate to an uncharged alcohol moiety.
far, and only a single report has described the preparation
A
of a rhamnolipid in standard solution synthesis.[8]
Therefore, we embarked on the development of a synthet-
ic strategy that would be broadly applicable for the prepara-
tion of glycolipid libraries. Recently, we found that oligosac-
charides can be synthesized efficiently in a solid-supported
manner by employing a specifically designed dilipid anchor-
ing molecule for quantitative and reversible retention on a
hydrophobic solid phase.[9] These initial findings inspired the
idea that hydrophobically assisted switching phase (HASP)
synthesis might be an ideal basis for the preparation of gly-
colipid libraries, which we have investigated in this contribu-
tion.
Achiral b-ketoesters 3n of different lengths were generat-
ed by C-acylation of Meldrum’s acid (2, Scheme 1).[11]
Asymmetric reduction of the ketones 3 in the presence of a
chiral ruthenium 2,2’-bis(diphenylphosphino)-1,1’-binaphthyl
(BINAP) catalyst yielded the enantiomerically pure b-hy-
droxy esters (R)-4n and (S)-4n.[12] These secondary alcohols
were obtained in high yields and with optical purities, as de-
termined by NMR spectroscopy and chiral GC separation of
the corresponding Mosher’s ester derivatives 5n
(Figure 2).[13] Subsequent saponification and silylation fur-
nished the corresponding 3-O-triethylsilyl-substituted car-
boxylic acids 6n and 7n as optically pure key building blocks
(Scheme 1). For the synthesis of rhamnolipid alcohols, the
terminally protected 1,3-diol building block 9 was required.
This could be accessed via the 3-(2-methoxyethoxy)methyl-
protected acetate 8.
Results and Discussion
The lead rhamnolipid RL-2,214,14-COOH 1 (Figure 1), isolat-
ed from Burkholderia pseudomallei and Pseudomonas aeru-
ginosa (together with the shorter-fatty-acid compound RL-
After the esterification of 6i and 4j, C18 reversed-phase
silica support (RP-18) was added to the reaction mixture,
the solvents were evaporated, and all monolipid starting ma-
terials were easily and quantitatively removed by filtration
with MeOH/water (Scheme 2). Subsequently, the silylated
esters were deprotected with dilute trifluoroacetic acid
(TFA) while attached to the solid support. After washing
and filtration, MgSO4 was added and the diastereomerically
pure 3-hydroxy dilipid compounds 12i,j were desorbed quan-
titatively with CH2Cl2. The pure products obtained from the
previous solid-phase step were glycosylated with an excess
(1.3 equiv) of rhamnose donor 13 in CH2Cl2.[9] The progress
of the reaction was monitored by TLC and MALDI-MS.
Workup of 14i,j (and 17i,j and 20i,j) was performed by a phase
switch from solution phase back to the RP-18 solid support.
Removal of the temporary phenoxyacetate (POAc) protect-
ing group at the 2-position of rhamnose was effected on the
solid support, as was the final removal of the butane-2,3-
dione (BDA) protective group, yielding 16i,j (and 19i,j and
22i,j). For the construction of higher glycosylated rhamnoli-
pid methyl esters, compounds 15i,j (18i,j) were glycosylated in
a second (third) HASP reaction cycle, yielding pure prod-
Abstract in German: Die erste Synthese einer Glycolipidbi-
bliothek unter Verwendung der hydrophob unterstützten pha-
senwechselsynthese—hydrophobically
assisted
switching
phase (HASP) synthesis—wird beschrieben. Das HASP-Syn-
thesekonzept ermöglicht den flexiblen Wechsel zwischen Re-
aktionsschritten in Lösung und an der Festphase, bei denen
die Moleküle an einem hydrophoben Kieselgelträger immobi-
lisiert sind. Abgeleitet von der Leitverbindung 1, einem stark
immunstimulierendem Rhamnolipid, wurde eine Glycolipid-
bibliothek entworfen und synthetisiert, die Variationen im
Kohlenhydratteil, den Lipidmuster und der Konfiguration der
3-Hydroxyfettsäuren enthielt. Die enantioselektive Synthese
von 3-Hydroxyfettsäuren wurde durch die asymmetrische Hy-
drierung von 3-Oxofettsäuren realisiert. Mit diesem Ansatz
wurden Glycolipide in exzellenten Ausbeuten hergestellt,
ohne dass eine einzige Isolierung von Zwischenprodukten
notwendig war. Die abschließende Entschützung zu den Car-
bonsäuren wurde durch enzymatische Esterspaltung erreicht.
Alle hergestellten Rhamnolipide wurden auf ihre immunsti-
mulierende Wirkung gegenüber menschlichen Monozyten ge-
testet, indem die Ausschüttung des Cytokines Tumornekrose-
faktor a (TNFa) in das Zellmedium gemessen wurde. Die
beobachteten Struktur–Aktivitätsbeziehungen von Rhamnoli-
piden deuten auf einen spezifischen, Erkennungs-basierten
Wirkmechanismus hin. Geringfügige strukturelle Variationen
der Rhamnolipide führten zu einem starken Effekt auf die
immunstimulierende Aktivität der Rhamnolipide bei Konzen-
trationen im unteren mikromolaren Bereich.
ucts without the need for
(Figure 3).
a single purification step
Cleavage of the methyl ester group was most successful
under enzymatic conditions. Results of a primary screen of a
set of commercially available lipases (Table 1) and a secon-
dary screen for the optimal solvent conditions (data not
shown) led us to use the solid-supported lipase of Candida
antarctica, which furnished the rhamnolipid acids in good to
excellent yields and with exceptionally broad substrate toler-
Chem. Eur. J. 2006, 12, 7116 – 7124
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7117

J. Rademann et al.
Figure 1. Rhamnolipid library obtained by HASP synthesis.
ance (Table 2). All the (R,R’)-configured rhamnolipid
methyl esters 16i,j, 19i,j, and 22i,j could be hydrolyzed to the
corresponding diastereomeric RL-acids, however, the (3S)-
configured fatty acid methyl esters resisted successful sub-
strate recognition by the enzyme. Therefore, lipid esters
amenable to chemical deprotection were synthesized and
employed for HASP construction of the corresponding RL
esters. Because the trichloroethyl ester 10 underwent trans-
esterification upon treatment with MeNH2 in MeOH, the
benzyl ester 11 was employed to access (R,S’)-configured
rhamnolipid acids (Scheme 3). Excellent yields of both
(R,S’)-RL acids, formed by employing the benzyl ester 11,
and rhamnolipid alcohols, synthesized from the acetate 10,
were accomplished through analogous HASP cycles
(Table 2).
To assay the immunostimulation produced by the individ-
ual rhamnolipids (Figure 1), human mononuclear cells were
treated with a rhamnolipid preparation and the amount of
the cytokine tumor necrosis factor (TNFa) secreted into the
medium was determined by conducting sandwich-ELISA[5a]
(Figure 4).
7118
www.chemeurj.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7116 – 7124

Synthesis of a Glycolipid Library by a Solid-Phase Strategy
FULL PAPER
Figure 2. 1H NMR (400 MHz, CDCl3, OMe area, d=3.7–3.5 ppm) of
MTPA derivative 514. The MTPA ester obtained from racemic methyl 3-
hydroxymyristate (rac-414) shows signals for both the (R,S) and the (R,R)
diastereomers (dotted line). The corresponding Mosher’s ester derivative
(R,R)-514 obtained by asymmetric (R)-BINAP-RuCl2-catalyzed hydroge-
nation shows only one set of signals, allowing the determination of opti-
cal purities by integration (98.3% dr in (R,R)-514, corresponding to
98.3% er in (R)-414).
Scheme 1. Synthesis of enantiopure b-hydroxy lipid building blocks from
Meldrum’s acid (2). a) CH2Cl2, pyridine, 1 h (08C), 77.9%; b) MeOH,
3 h, reflux; c) 0.1 mol% (R)-BINAP-RuII, MeOH, 5 bar H2, 5 h, 1008C,
95.6%, optical purity 98.3%er determined by 1H NMR as %dr (diaste-
A
tical to (c), but with (S)-BINAP; e) LiOH, MeOH/H2O (4:1), 12 h,
97.6%; f) triethylsilyl chloride, pyridine, 2 h, 608C, 85.6%; g) 2-methox-
yethoxymethyl chloride, DIPEA, CH2Cl2, 20 h, 93%; h) LiAlH4, THF,
2 h, 08C, 99%; i) Ac2O, cat. DMAP, pyridine, 12 h, 92%; j) dry ZnBr2,
CH2Cl2, 16 h, 87%; k) trichloroethanol, EDC, cat. DMAP, CH2Cl2, 12 h;
l) TFA, 10 min, 62% (2 steps); m) benzyl alcohol, EDC, cat. DMAP,
CH2Cl2, 12 h, 81%; n) TFA, CH2Cl2, 10 min, 98%. Yields and purities
given refer to the corresponding myristic acid derivatives (R=C11).
Figure 3. ESI-MS of crude BDA-protected rhamnolipid RL-3,214,14
-
COOMe 2114,14 after three glycosylation/deprotection cycles, demonstrat-
ing the high purity of glycolipids obtained by HASP without the need for
intermediary purification steps.
Comparison of the library of rhamnolipids indicated a
strong influence of glycolipid structure on the proinflamma-
tory activity in human mononuclear cells, as determined by
TNFa secretion. Lead structure 1 and rhamnolipids 23, 26,
and 29 were active in human macrophages in a dose-de-
pendent manner, expressing their half-maximal activities at
observed if the negative charge was removed (31, 32), or if a
single enantiomeric center in the fatty acid was changed (33,
34).
a low-mm concentration range of the stimulus (1 mgmLꢀ1
=
1.3 mm for compound 1). The number of rhamnoses and the
length of the second fatty acid were not critical for bioactivi-
ty, but the biological activity dropped significantly as the
lengths of both fatty acids were either increased (27) or de-
creased (30). Dramatic attenuation in bioactivity was also
Conclusion
We have demonstrated the preparation of a novel glycolipid
library by using the HASP synthesis concept. Our strategy
allowed flexible combinations of solid-phase reactions with
Chem. Eur. J. 2006, 12, 7116 – 7124
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7119

J. Rademann et al.
couplings
solution-phase
through quantitative and re-
versible phase switching in a
commercial parallel synthesiz-
er. This work should facilitate
access to libraries and enable
structural and biological stud-
ies of hydrophobic molecules
that have not yet been attaina-
ble so far.
The biological results strong-
ly suggest a specific—that is,
recognition-based—mode
of
monocyte activation by rham-
nolipids. The library is current-
ly the subject of further inves-
tigations, including the bio-
physical profiling of the syn-
thesized glycolipids, the struc-
tural characterization of active
aggregates, and the tracing of
rhamnolipids within target
cells.
Scheme 2. HASP synthesis of rhamnolipid methyl esters on RP-18 silica as polymer support. a) EDC, cat.
DMAP, CH2Cl2, 12 h, addition of RP-support, evaporation, washing steps (MeOH/H2O 80:20), 97.5%. b) 5%
TFA, MeOH/H2O (80:20), washing steps (MeOH/H2O 80:20), MgSO4, desorption (CH2Cl2), 98%. c) 1.3 equiv
donor 13, 0.05 equiv of TMSOTf, CH2Cl2, 30 min, neutralization (DIPEA), evaporation, washing steps
(MeOH/H2O 80:20), 94%. d) 40 equiv aqueous MeNH2, MeOH/THF (1:1), 1 h, washing steps (MeOH/H2O
80:20), 93%. e) TFA/H2O (90:10), 3×10 min, washing steps (MeOH/H2O 80:20), MgSO4, desorption
(CH2Cl2), 98%. Yields given refer to the respective myristic acid derivatives (n=1, R=C11). For the purpose
of fully quantifying all HASP steps, the products of solid-supported steps were desorbed, characterized, and
readsorbed.
Experimental Section
General remarks: Hydrophobically
assisted synthesis was conducted with
bulk reversed-phase Grom-Sil ODS-
4 HE silica gel (50 m, 120 Å). All
chemicals were purchased from
Sigma, Aldrich, or Fluka and were
used without further purification. Reactions were carried out in commer-
cially available dry solvents unless otherwise stated. Column chromatog-
raphy was performed by using Kieselgel 60 silica gel (Merck, 0.04–
0.063 mm). TLC analysis was carried out by using Merck silica gel 60 F254
plates and Merck RP-18 F254 plates, detection of compounds was ach-
ieved by measuring UV absorption (254 nm) and by charring after spray-
ing with 5% H2SO4 or with ammonium molybdate and ceric ammonium
sulfate in 10% aqueous H2SO4. 1H and 13C NMR spectra were recorded
by using a Bruker AMX2 600 MHz, a Bruker Avance 400 MHz, or a Bru-
ker AC 250 MHz spectrometer. 1H and 13C chemical shifts are given in
ppm relative to the solvent signal as internal standard. Assignments were
based on homonuclear decoupling experiments and homo- and heteronu-
clear correlation. Mass spectra were measured by using a Bruker Auto-
flex MALDI-TOF instrument with a-cyano-4-hydroxycinnamic acid
(CHCA) as matrix. HRESI-MS was performed by using a 4.7 Tesla
APEX II spectrometer (Bruker Daltonics, Bremen). Optical rotations
were measured by using a Perkin–Elmer 341 polarimeter.
Table 1. Screening of commercially available hydrolases for their activity
against RL-2,214,14-COOMe 1914,14. Data show the results for the OMe-hy-
drolysis of 3 mmol (2.3 mg) RL-2,214,14-COOMe 1914,14 on employment of
3 mg enzyme in ACN/NaiPO4 buffer (8 mL, pH 7.0, 1:3 v/v) or n-hexane/
NaiPO4 buffer (5 mL, 3:1 v/v) at RT for 72 h. Hydrolytic activity was
evaluated by performing MALDI-MS of the reaction mixture: (-o-)=no
turnover, (+)=turnover<1/3, (++)=1/3<turnover<2/3; (+++)=
turnover>2/3.
Name/source of the hydrolase
ACN/buffer n-hexane/buffer
Alcaligines sp.
Aspergillus niger
Candida antarctica, fract. A
Candida antarctica, fract. B
Candida antarctica
(immobilized)
ASL
ANL
CAL-A
CAL-B
CAL-B
-o-
-o-
-o-
+
+++
-o-
-o-
+
+
-o-
Methyl (R)-3-hydroxymyristate (414): Methyl 3-oxomyristate (314, 7.50 g,
29.25 mmol) was dissolved in dry MeOH (30 mL) under argon and
placed in a hydrogenation autoclave. The freshly prepared (R)-BINAP-
RuCl2 hydrogenation catalyst (29.3 mmol, 0.1 mol%) was added, the au-
toclave was sealed and flushed with hydrogen, and a H2 pressure of 5 bar
was provided. After the reaction temperature of 1008C had been reached
(by heating in a silicon oil bath), the reaction mixture was stirred for 5 h.
The crude product was concentrated and purified by column chromatog-
raphy (hexane/ethyl acetate 5:1) to yield 414 as a white solid (7.22 g,
95.6%).
Candida eugosa
Bovine pancreas chymotrypsine CHYM
Horse liver
Rhizomucor miehei
Rhizomucor miehei (Lipozym) RML
Pig liver (E-1)
Pig liver (E-2)
Pig liver
CRL
-o-
-o-
-o-
-o-
-o-
-o-
-o-
-o-
-o-
-o-
+
+++
-o-
-o-
++
+
HLE
RML
PLE-E1
PLE-E2
PLE
PPL
RAL
TLL
-o-
-o-
-o-
-o-
++
+
Porcine pancreas
Rhizopus arrhizus
Thermomyces lanuginosa
Enantiomeric purity: 98.3% er (enantiomeric ratio, determined via the
(R,R)-a-methoxy-a-(trifluoromethyl)phenylacetic acid (MTPA)-ester de-
rivative 514); [a]2D0 =ꢀ13.98 (c=1.0 in CHCl3); Rf =0.2 (hexane/ethyl ace-
7120
www.chemeurj.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7116 – 7124

Synthesis of a Glycolipid Library by a Solid-Phase Strategy
FULL PAPER
Table 2. Summary of building blocks used and yields [%] of each HASP step in the synthesis of the focused rhamnolipid library.
Building
blocks
Desilylated
lipid ester[a]
Glycosyl
ation[e]
POAc
removal
BDA
removal
Removal of terminal
OMe/OBn/OAc
Final RL ([mg])
64–414
87.5
88.1
84.9
83.5
91.6
95.1
86.7
93.8
94.8
98.3
98.7
94.8
87.2
97.1
76.9
97.6
92.7
94.0
44.6[h]
68.4[h]
73.0[h]
85.6[h]
31.4[h]
46.6[h]
57.0[h]
64.8[h]
83.4[h]
70.2[i]
72.1[i]
75.4[j]
81.0[j]
1,24,14-COOH, 23
ACHTREUNG
614–614–414
614–614–414
614–414
618–418
614–44
614–414
612–412
614–414
614–9
95.7[b]
1,314,14,14-COOH, 24
2,314,14,14-COOH, 25
3,214,14-COOH, 26
2,218,18-COOH, 27
1,214,4-COOH, 28
1,214,14-COOH, 29
2,212,12-COOH, 30
2,214,14-COOH, 1
2,214,14-CH2OH, 31
1,214,14-CH2OH, 32
1,2(R)14,(S)14-COOH, 33
2,2(R)14,(S)14-COOH, 34
ACHTREUNG
[c]
96.7
71.5
ACHTREUNG
[c]
[f]
70.7[g]
93.5[g]
95.2
ACHTREUNG
[f]
99.2
ACHTREUNG
73.6
70.5
86.1
94.9
ACHTREUNG
98.2[d]
96.9
81.7
ACHTREUNG
93.5
ACHTREUNG
[c]
92.5
98.7
ACHTREUNG
[c]
[f]
92.5[g]
96.3
ACHTREUNG
614–9
614–11
614–11
86.7
89.1
ACHTREUNG
99.5
85.0
98.8
ACHTREUNG
[c]
[f]
93.4[g]
ACHTREUNG
[a] Combined yield for esterification and TES-deprotection. [b] Combined yield for the route of two esterifications and two TES-desilylations. [c] As for
the corresponding monoglycosylated lipid. [d] Isolated yield for the desilylation step. [e] Yields refer to the last glycosylation step of the final RL com-
pound. [f] Not isolated for quantification. [g] Combined yields for POAc and BDA removal. [h] Yields of the enzymatic OMe ester hydrolysis after chro-
matographic purification. [i] Yields of the OAc ester hydrolysis after chromatographic purification. [j] Yields of the hydrogenolytic OBn ester deprotec-
tion after chromatographic purification.
Scheme 3. General synthesis of RL acids and alcohols from RL esters.
a) Terminal (R)-configured RL methyl ester, lipase from Candida antarc-
tica, Na2HPO4 buffer, pH 7.0, 20% DMSO, 408C, 72 h. b) Pd/C, H2,
THF/HOAc, 5 h. c) Cat. LiOH, pH 8.5, THF/MeOH/H2O.
1
Figure 4. TNFa responses of human mononuclear cells (MNCs) after
stimulation with various synthetic rhamnolipids.
tate 5:1); H NMR (400 MHz, CDCl3): d=3.98 (m, 1H; CH-OH), 3.69 (s,
3H; OMe-CH3), 2.87 (br, 1H; OH), 2.52–2.36 (m, 2H; CH-CH2-CO),
1.52–1.24 (m, 20H; CH2), 0.86 ppm (t, 3H; CH3); 13C NMR (100 MHz,
CDCl3): d=173.5, 68.0, 51.7, 41.1, 36.5, 31.9, 29.6, 29.5, 29.4, 29.3, 25.5,
22.7, 14.1 ppm.
for C20H42O3SiNa [M+Na]+: 381.27954; found: 381.27959, Dm/z=
0.13 ppm.
(R)-3-O-Triethylsilyl myristic acid (614): A solution of (R)-3-hydroxymyr-
istic acid (2.5 g, 10.23 mmol) in pyridine (60 mL) was heated to 608C,
and triethylsilyl chloride (1.80 mL, 10.74 mmol) was added dropwise
from a dropping funnel. The reaction mixture was allowed to stir for 2 h
and was then concentrated, and the residue was dissolved in CHCl3
(200 mL), washed with an aqueous solution of NaHCO3 (3×200 mL,
5%), dried with sodium sulfate, and concentrated. The crude product
was concentrated and purified by column chromatography (hexane/ethyl
acetate 3:1, 1% triethylamine (TEA)) to yield 614 as a white and waxy
solid (3.14 g, 85.6%).
[a]2D0 =ꢀ3.18 (c=1.0 in CHCl3); Rf =0.4 (hexane/ethyl acetate 3:1);
1H NMR (400 MHz, CDCl3): d=4.07 (m, 1H; CH-OH), 2.59–2.44 (m,
2H; a-CH2), 1.54–1.24 (m, 20H; CH2), 0.98–0.92 (m, 12H; CH3, TES-
CH3), 0.61 ppm (q, 6H; TES-CH2); 13C NMR (100 MHz, CDCl3): d=
175.4, 69.4, 41.8, 37.3, 31.9, 29.6, 29.6, 29.5, 29.5, 29.3, 25.3, 22.7, 14.1, 6.8,
4.8 ppm; HRMS (FT-ICR-MS, ICR=ion cyclotron resonance): m/z calcd
Benzyl (S)-3-hydroxymyristate (11): Benzyl alcohol (435 mL, 4.2 mmol)
was added to a solution of (S)-3-O-TES-myristic acid (TES=triethylsilyl)
714 (502 mg, 1.40 mmol) and 1-(3-dimethylaminopropyl)-3-ethylcarbodii-
mide (EDC) (805 mg, 4.20 mmol) in dry CH2Cl2 (5 mL). A catalytic
amount of 4-dimethylaminopyridine (DMAP) (10 mg) was added and the
reaction mixture was allowed to stir for 12 h at RT. After removal of the
solvents by concentration, the product was obtained by column chroma-
tography (hexane/ethyl acetate 30:1) as a clear and viscous oil (511 mg,
81%).
Rf =0.3 (hexane/ethyl acetate 30:1); [a]2D0 =+10.38 (c=1.0 in CHCl3);
1H NMR (400 MHz, CDCl3): d=7.36–7.31 (m, 5H; Ar-H), 5.11 (s, 2H;
Bn-CH2), 4.16 (m, 1H; CH-OSi), 2.57–2.42 (m, 2H; a-CH2), 1.53–1.25
(m, 20H; CH2), 0.96–0.87 (m, 12H; CH3, TES-CH3), 0.59 ppm (q, 6H;
TES-CH2); 13C NMR (100 MHz, CDCl3): d=171.6, 135.9, 128.5, 128.2,
128.1, 69.4, 66.1, 42.8, 37.7, 31.9, 29.6, 29.5, 29.3, 25.1, 22.7, 14.1, 6.8,
Chem. Eur. J. 2006, 12, 7116 – 7124
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7121

J. Rademann et al.
4.9 ppm; HRMS (FT-ICR-MS): m/z calcd for C27H48O3SiNa [M+Na]+:
471.32649; found: 471.32456, Dm/z=0.15 ppm.
CH2Cl2 (10 mL) was added and the suspension was shaken for 3 min and
filtered twice more. The filtrates containing the desorbed compound
were collected and concentrated to give
a clear and colorless oil
Trifluoroacetic acid (2.5 mL) was added to a solution of benzyl (S)-3-O-
TES-hydroxymyristate (505 mg, 1.05 mmol), prepared as described
above, in dry CH2Cl2 (50 mL). The mixture was stirred for 10 min and
then washed with water (3×50 mL), and the combined organic layers
were collected, dried with sodium sulfate, and concentrated. Purification
by column chromatography (hexane/ethyl acetate 8:1) yielded 11 as a
colorless wax (357 mg, 98%).
Rf =0.25 (hexane/ethyl acetate 8:1); [a]2D0 =+12.98 (c=1.0 in CHCl3);
1H NMR (400 MHz, CDCl3): d=7.36–7.32 (m, 5H; Ar-H), 5.15 (s, 2H;
Bn-CH2), 4.01 (m, 1H; CH-OH), 2.77 (br, 1H; OH), 2.55–2.45 (m, 2H;
a-CH2), 1.49–1.26 (m, 20H; CH2), 0.87 ppm (t, 3H; CH3); 13C NMR
(100 MHz, CDCl3): d=172.9, 135.6, 128.6, 128.4, 128.2, 68.0, 66.5, 41.3,
36.5, 31.9, 29.6, 29.6, 29.5, 29.5, 29.4, 29.3, 25.4, 22.7, 14.1 ppm; HRMS
(FT-ICR-MS): m/z calcd for C21H34O3Na [M+Na]+: 357.24002; found:
357.23980, Dm/z=0.6 ppm.
(233.6 mg, 97.5%). The RP-silica-bound TES-dilipids were desorbed for
analytical purposes only and usually remained immobilized for further
solid-supported chemical transformation.
1H NMR (400 MHz, CDCl3): d=5.18 (m, 1H; CH-O), 4.07 (m, 1H; CH-
OSi), 3.65, (s, 3H; OMe), 2.62–2.40 (m, 4H; a-CH2, a’-CH2), 1.58–1.24
(m, 40H; CH2), 0.93 (t, 9H; CH3, TES-CH3), 0.86 (t, 6H; CH3), 0.58 ppm
(q, 6H; TES-CH2); 13C NMR (100 MHz, CDCl3): d=171.0, 170.8, 70.5,
69.1, 51.7, 42.9, 38.9, 37.5, 33.9, 31.9, 29.7, 29.6, 29.6, 29.5, 29.5, 29.4, 29.3,
25.1, 22.7, 14.1, 6.8, 4.9 ppm.
HASP synthesis of methyl (R)-3-O-[(R)-3’-hydroxymyristyl]myristate
(1214,14): TES-dilipid ester prepared as described above (239.6 mg,
0.40 mmol), still immobilized on RP-silica from the previous step, was
suspended in methanol/water (3.8 mL, 80/20 v/v). Trifluoroacetic acid
(200 mL) was added and the suspension was shaken vigorously for
20 min. After completion of the reaction, the workup by HASP filtration
and HASP release of products was performed as described above to
yield 1214,14 as a colorless, clear oil (190.4 mg, 98.2%).
1H NMR (400 MHz, CDCl3): d=5.25 (m, 1H; CH-O-C=O), 3.98 (m,
1H; CH-OH), 3.66, (s, 3H; OMe), 2.81 (br, 1H; OH), 2.57–2.34 (m, 4H;
a-CH2, a’-CH2), 1.61–1.24 (m, 40H; CH2), 0.87 ppm (t, 6H; CH3);
13C NMR (100 MHz, CDCl3): d=172.5, 171.1, 70.9, 68.3, 51.9, 41.8, 39.0,
36.5, 34.0, 31.9, 29.6, 29.6, 29.6, 29.5, 29.5, 29.4, 29.3, 29.3, 25.5, 25.1, 22.7,
14.1 ppm; HRMS (FT-ICR-MS): m/z calcd for C29H56O5Na [M+Na]+:
507.40200; found: 507.40241, Dm/z=0.8 ppm.
(R)-3-MEM-O-myristyl acetate (MEM=methoxyethoxymethyl) (8): (R)-
3-MEM-O-myristyl alcohol (2.25 g, 7.06 mmol) was dissolved in pyridine
(50 mL), and acetic anhydride (6.67 mL, 70.6 mmol) was added slowly.
The reaction mixture was stirred with a catalytic amount of DMAP
(15 mg) for 12 h. After completion of the reaction, the mixture was con-
centrated and the residue was dissolved in CHCl3 (100 mL), washed with
an aqueous solution of KHSO4 (3×100 mL, 5%), concentrated, and puri-
fied by column chromatography (hexane/ethyl acetate 4:1) to give 8 as a
clear oil (2.34 g, 92%).
[a]2D0 =ꢀ16.18 (c=1.0 in CHCl3); Rf =0.2 (hexane/ethyl acetate 4:1);
1H NMR (400 MHz, CDCl3): d=4.72 (2d, 2H; O-CH2-O (AB-system)),
4.12 (m, 2H; CH2-OAc), 3.70–3.66 (m, 3H; CH-O-MEM, O-CH2-CH2),
3.53 (m, 2H; O-CH2-CH2), 3.37 (s, 3H; OCH3), 2.02 (s, 3H; OAc), 1.79
(m, 2H; CH2-CH2-OAc), 1.54–1.24 (m, 20H; CH2), 0.86 ppm (t, 3H;
CH3); 13C NMR (100 MHz, CDCl3): d=171.1, 94.4, 74.5, 71.7, 67.1, 61.4,
59.0, 34.4, 33.2, 31.9, 29.7, 29.6, 29.6, 29.6, 29.3, 25.0, 22.7, 21.0, 14.1 ppm;
HRMS (FT-ICR-MS): m/z calcd for C20H40O5Na: 383.27680 [M+Na]+;
found: 383.27654, Dm/z=0.7 ppm.
HASP synthesis of 1-O-{methyl (R)-3-O-[(R)-3’-O-myristyl]myristate}-2-
O-phenoxyacetyl-[3,4-O-(2,3-dimethoxybutane-2,3-diyl)]-(1!2)-a-l-
rhamnopyranoside (1414,14): Dilipid ester methyl (R)-3-O-[(R)-3’-hydroxy-
myristyl]myristate 1214,14 (104.2 mg, 0.215 mmol) and donor 3,4-O-(2,3-di-
methoxybutane-2,3-diyl)-2-O-phenoxyacetyl-a-l-rhamnopyranosyl
tri-
chloroacetimidate (13, 155.6 mg, 0.28 mmol) were dissolved in dry
CH2Cl2 (3 mL) under argon.[9] A freshly prepared trimethylsilyl trifluoro-
methanesulfonate (TMSOTf) solution (0.05 equiv) in dry CH2Cl2
(215 mL, 0.05m) was added and the reaction mixture was shaken for
30 min. After completion, the solution was neutralized with N,N-diisopro-
pylethylamine (DIPEA) (10 mL), and HASP filtration and HASP release
of products were conducted as described above to yield 1414,14 as a color-
less, clear oil (177.3 mg, 93.8%).
(R)-3-Hydroxymyristyl acetate (9): Powdered anhydrous zinc bromide
(7.18 g, 31.9 mmol) was added under argon to a solution of (R)-3-MEM-
O-myristyl acetate 8 (2.30 g, 6.38 mmol) in dry CH2Cl2 (50 mL). The sus-
pension was agitated for 16 h at RT. After the reaction was complete,
CHCl3 (100 mL) was added, the mixture was washed with water (3×
200 mL), and the combined organic layers were dried with sodium sulfate
and concentrated. The crude product was purified by column chromatog-
raphy (hexane/ethyl acetate 2:1) to give 9 as a colorless, waxy solid
(1.51 g, 87%).
[a]2D0 =ꢀ2.78 (c=1.0 in CHCl3); Rf =0.3 (hexane/ethyl acetate 2:1);
1H NMR (400 MHz, CDCl3): d=4.25–4.05 (m, 2H; CH2-OAc), 3.58 (m,
1H; CH-OH), 2.54 (br, 1H; OH), 1.97 (s, 3H; OAc), 1.77–1.54 (m, 2H;
CH2-CH2-OAc), 1.37–1.18 (m, 20H; CH2), 0.80 ppm (t, 3H; CH3);
13C NMR (100 MHz, CDCl3): d=171.3, 68.4, 67.1, 37.4, 36.1, 31.8, (29.5),
(29.4), 29.2, 25.5, 22.5, 20.8, 13.9 ppm; HRMS (FT-ICR-MS): m/z calcd
for C16H32O3Na [M+Na]+: 295.22437; found: 295.22418, Dm/z=0.6 ppm.
1H NMR (400 MHz, CDCl3): d=7.29–6.91 (m, 5H; ArH), 5.14 (m, 1H;
CH-O-C=O), 5.09 (d, 3J1,2 <1 Hz, 3J2,3 =2.8 Hz, 1H; H-2), 4.80 (d, J1,2
<
3
1 Hz, 1H; H-1), 4.65 (s, 2H; POAc-CH2), 3.97 (dd, 3J2,3 =2.8 Hz, J3,4
=
3
10.1 Hz, 1H; H-3), 3.96 (m, 1H; CH-O-Rha), 3.80 (m, 1H; H-5), 3.60, (s,
3H; OMe), 3.54 (t, 3J4,5 =10.1 Hz, 1H; H-4), 3.17 (s, 3H; BDA-OMe),
3.15 (s, 3H; BDA-OMe), 2.60–2.34 (m, 4H; a-CH2, a’-CH2), 1.54–1.18
(m, 49H; CH2, H-6, BDA-Me), 0.81 ppm (t, 6H; CH3); 13C NMR
(100 MHz, CDCl3): d=170.8, 170.4, 168.5, 157.9, 129.5, 121.6, 114.8,
100.0, 99.7, 97.2, 75.4, 71.7, 70.8, 68.6, 67.0, 66.1, 65.2, 51.7, 48.0, 47.6,
40.3, 38.9, 33.9, 31.9, 31.8, 29.6, 29.6, 29.6, 29.5, 29.5, 29.5, 29.3, 29.3, 25.1,
24.8, 22.7, 17.7, 17.6, 16.5, 14.1 ppm; HRMS (FT-ICR-MS): m/z calcd for
C49H82O13Na [M+Na]+: 901.56476; found: 901.46425, Dm/z=0.57 ppm.
HASP synthesis of methyl (R)-3-O-[(R)-3’-O-TES-myristyl]myristate:
General procedure for parallel HASP synthesis: EDC (230 mg,
1.2 mmol) and a catalytic amount of DMAP (15 mg) were added to a sol-
ution of methyl (R)-3-hydroxymyristate (414, 103.4 mg, 0.4 mmol) and
(R)-3-O-triethylsilyl myristic acid (614, 286.9 mg, 0.8 mmol) in dry CH2Cl2
(3 mL). The reaction mixture was shaken for 12 h at RT on a standard
Büchi Syncore workstation for parallel synthesis. After completion of the
reaction step, reversed-phase silica gel (1.5 g, Grom-Sil ODS-4 HE, 50 m,
120 Å) was added to the reaction vessel. To adsorb the compounds, sol-
vents were removed by evaporation. The dry reversed-phase silica gel on
which the reaction mixture was adsorbed was then washed three times
with methanol/water (80:20, 8 mL) under vigorous shaking for 3 min.
After each washing step, the solvents were filtered on the Syncore work-
station under a gentle stream of compressed air. After the third washing/
filtration step, anhydrous MgSO4 (equivalent volume to RP-silica) and
distilled CH2Cl2 (10 mL) were added, and the suspension was vigorously
shaken for 20 min and then filtered off. For complete product desorption,
HASP synthesis of 1-O-{methyl (R)-3-O-[(R)-3’-O-myristyl]myristate}-
3,4-O-(2,3-dimethoxybutane-2,3-diyl)-2-O-{a-l-rhamnopyranosyl-[3’,4’-O-
(2,3-dimethoxybutane-2,3-diyl)]-2’-O-phenoxyacetyl}-(1!2)-a-l-rhamno-
pyranoside (1714,14): The second glycosylation, at the 2-hydroxy position
of
rhamnolipid
BDA-HO-RL-1,214-COOMe
1514,14
(160.0 mg,
0.215 mmol), was performed analogously to the glycosylation step descri-
bed above, but with two equivalents of donor 13 (239.2 mg, 0.43 mmol)
to yield 1714,14 as a highly viscous oil (240.45 mg, 98.3%).
1H NMR (400 MHz, CDCl3): d=7.35–6.78 (m, 5H; ArH), 5.43 (dd, J1,2
<
<
3
3
3
1 Hz, J2,3 =2.8 Hz, 1H; H-2A), 5.18 (m, 1H; CH-O-C=O), 5.10 (d, J1,2
1 Hz, 1H; H-1A), 4.77 (d, 3J1,2 <1 Hz, 1H; H-1B), 4.67 (s, 2H; POAc-
CH2), 3.96 (dd, 3J2,3 =2.8 Hz, 3J3,4 =10.1 Hz, 1H; H-3A), 4.03–3.59 (m,
10H; H-2B, H-3B, H-4B, H-5B, H-4A, H-5A, CH-O-Rha, OMe), 3.26–3.16
(4×s, 12H; BDA-OMe), 2.63–2.35 (m, 4H; a-CH2, a’-CH2), 1.58–1.18
(m, 58H; CH2, H-6A, H-6B, BDA-Me), 0.85 ppm (t, 6H; CH3); 13C NMR
(100 MHz, CDCl3): d=170.8, 170.4, 167.7, 158.0, 129.4, 121.5, 114.7,
7122
www.chemeurj.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7116 – 7124

Synthesis of a Glycolipid Library by a Solid-Phase Strategy
FULL PAPER
100.0, 99.8, 99.7, 99.4, 99.0, 98.8, 75.6, 74.7, 71.0, 70.7, 68.7, 68.5, 68.4,
67.3, 67.1, 66.4, 65.2, 51.8, 47.9, 47.8, 47.6, 47.4, 40.4, 38.9, 33.9, 33.7, 31.9,
29.7, 29.6, 29.6, 29.5, 29.3, 25.1, 24.9, 22.7, 17.8, 17.8, 17.7, 17.7, 17.6, 17.5,
14.1 ppm; ESI-MS: m/z: 1161.3 [M+Na]+.
concentrated under reduced pressure, and the residue was purified by
column chromatography (60 mL SiO2, CHCl3/MeOH/HOAc 40:1:1). Sol-
vents were removed and the pure rhamnolipid was obtained after three
lyophilization steps from tert-butanol/water (4:1 v/v) as a colorless, fluffy
solid (41.3 mg, 75.4%).
HASP synthesis of 1-O-{methyl (R)-3-O-[(R)-3’-O-myristyl]myristate}-
3,4-O-(2,3-dimethoxybutane-2,3-diyl)-2-O-{a-l-rhamnopyranosyl-[3’,4’-O-
(2,3-dimethoxybutane-2,3-diyl)]}-(1!2)-a-l-rhamnopyranoside (1814,14):
Fully protected rhamnolipid BDA-POAc-RL-2,214-COOMe 1714,14
(224.0 mg, 0.197 mmol), still immobilized on the RP-support from the
previous HASP step, was suspended in methanol/THF (4 mL, 1:1 v/v).
Methylamine (684 mL, 40 equiv of a 40% (11.5m) aqueous solution) was
added and the suspension was shaken for 1 h. After completion of the re-
action, water was added (12 mL) and the reaction mixture was filtered
off. Subsequent HASP washing steps and HASP release of products were
performed as described above to yield 1814,14 as a colorless, highly viscous
oil (182.8 mg, 92.5%).
1H NMR (600 MHz, CD3COOD): d=5.27 (m, 1H; CH3-O), 4.98 (d,
3
3J1,2 =1.0 Hz, 1H; H-1), 4.09 (m, 1H; CH3’-O-Rha), 3.96 (dd, J1,2
=
1.0 Hz, 3J2,3 =3.1 Hz, 1H; H-2), 3.83 (dd, 3J2,3 =3.1 Hz, 3J3,4 =9.3 Hz, 1H;
H-3), 3.79 (m, 1H; H-5), 3.56 (t, 3J4,5 =9.3 Hz, 1H; H-4), 2.68–2.54 (m,
4H; a-CH2, a’-CH2), 1.69–1.26 (m, 43H; CH2, H-6), 0.89 ppm (t, 6H;
CH3); 13C NMR (150 MHz, CD3COOD): d=176.8, 172.5, 100.1, 75.9,
73.9, 72.5, 72.4, 72.2, 69.8, 41.3, 39.5, 34.8, 34.4, 32.9, 30.7, 30.6, 30.6, 30.5,
30.4, 30.3, 26.1, 25.9, 23.6, 27.7, 14.4 ppm; MALDI-TOF-MS: m/z: 639.9
[M+Na]+, 661.6 [MꢀH+2Na]+; HRMS (FT-ICR-MS): m/z calcd for
C34H64O9Na [M+Na]+: 639.44425; found: 639.44499, Dm/z=1.20 ppm.
Rhamnolipid alcohol RL-2,214,14-CH2OH (31): A mixture of THF/H2O/
MeOH (8 mL, 70:20:10 v/v) was prepared and adjusted to pH 8.5 with a
solution of lithium hydroxide (0.1m). The rhamnolipid alcohol acetate
RL-2,2-CH2OAc (0.051 mmol, 40.2 mg), prepared through analogous
HASP steps starting from lipid acetate 9, was added and the reaction
mixture was allowed to stir at RT for 2 h. After completion of the reac-
tion, the mixture was neutralized with a solution of dilute aqueous
KHSO4, the solution was concentrated under reduced pressure, and the
crude product was purified by column chromatography (60 mL SiO2,
CHCl3/MeOH/HOAc 25:2:1). Solvents were removed and the pure
rhamnolipid was obtained after three lyophilization steps from tert-buta-
nol/water (4:1 v/v) as a colorless, fluffy solid (26.7 mg, 70.2%).
1H NMR (400 MHz, CDCl3): d=5.19 (m, 1H; CH-O-C=O), 5.09 (d,
3
3J1,2 <1 Hz, 1H; H-1A), 4.77 (d, 3J1,2 <1 Hz, 1H; H-1B), 4.07 (dd, J1,2
<
1 Hz, 3J2,3 =2.8 Hz, 1H; H-2A), 3.99–3.96 (m, 2H; CH-O-Rha, H-3A),
3.87–3.83 (m, 3H; H-5A, H-2B, H-3B), 3.77–3.64, (m, 5H; H-5B, H-4A,
OMe), 3.56 (t, 3J4,5 =10.1 Hz, 1H; H-4B), 3.28–3.17 (4×s, 12H; BDA-
OMe), 2.64–2.36 (m, 4H; a-CH2, a’-CH2), 2.08 (br, 1H; OH), 1.59–1.17
(m, 58H; CH2, H-6A, H-6B, BDA-Me), 0.86 ppm (t, 6H; CH3); 13C NMR
(100 MHz, CDCl3): d=170.8, 170.4, 101.0, 100.0, 99.7, 99.5, 99.4, 99.0,
75.2, 74.6, 70.6, 69.6, 68.6, 68.4, 68.3, 67.9, 67.1, 66.8, 51.7, 47.8, 47.7, 47.6,
47.4, 40.4, 38.9, 33.9, 33.6, 31.9, 29.6, 29.6, 29.6, 29.5, 29.5, 29.3, 25.1, 24.8,
22.6, 17.8, 17.7, 17.6, 17.6, 16.5, 16.6, 14.1 ppm; ESI-MS: m/z: 1027.7
[M+Na]+.
1H NMR (600 MHz, CD3COOD): d=5.04 (d, 3J1,2 <1 Hz, 1H; H-1A),
5.01 (d, 3J1,2 <1 Hz, 1H; H-1B), 4.30–4.16 (m, 2H; CH2-OH), 4.14–4.11
(m, 2H; CH3’-O-Rha, H-2B), 3.89–3.65 (m, 6H; CH-CH2OH, H-2A, H-3A,
H-3B, H-5A, H-5B), 3.57 (t, 3J4,5 =9.3 Hz, 1H; H-4A), 3.51 (t, 3J4,5 =9.3 Hz,
1H; H-4B), 2.59–2.52 (m, 4H; a-CH2, a’-CH2), 1.89–1.69 (m, 2H; CH2-
CH2OH), 1.61–1.24 (m, 46H; CH2, H-6A, H-6B), 0.86 ppm (t, 6H; CH3);
13C NMR (150 MHz, CD3COOD, selected signals): d=102.7, 97.6, 79.6,
74.9, 73.5, 73.3, 71.6, 71.1, 69.3, 69.3, 69.1, 69.0, 65.5, 52.0, 41.0 ppm;
MALDI-TOF-MS: m/z: 771.4 [M+Na]+; HRMS (FT-ICR-MS): m/z
calcd for C40H76O12Na [M+Na]+: 771.52290; found: 771.52263, Dm/z=
0.35 ppm.
HASP synthesis of 1-O-{methyl-(R)-3-O-[(R)-3’-O-myristyl]myristate}-
2-O-a-l-rhamnopyranosyl-(1!2)-a-l-rhamnopyranoside (1914,14): BDA-
acetal protected rhamnolipid methyl ester BDA-RL-2,214-COOMe 1814,14
(144.6 mg, 0.144 mmol), still immobilized on the RP-support from the
previous HASP step, was suspended in trifluoroacetic acid/water (4 mL,
90:10 v/v). The suspension was agitated for 10 min, water (5 mL) was
added, and the reaction mixture was filtered off. For completion of the
reaction, the BDA removal was performed three times as described. Sub-
sequent HASP washing steps and HASP release of products were per-
formed as described above to yield 1914,14 as a colorless, highly viscous oil
(109.6 mg, 98.1%).
MALDI-TOF-MS: m/z: 799.1 [M+Na]+.
Stimulation of human mononuclear cells (MNCs): MNCs were isolated
from heparinized (20 IEmLꢀ1) blood taken from healthy donors and
processed directly by mixing with an equal volume of Hank’s balanced
solution and centrifugation in a Ficoll density gradient for 40 min (218C,
500 g). The interphase layer of mononuclear cells was collected and
washed twice in Hank’s medium and once in RPMI 1640 containing l-
glutamine (2 mm), penicillin (100 UmLꢀ1), and streptomycin
(100 mgmLꢀ1). The cell number was equilibrated at 5×106 NmLꢀ1. For
stimulation, 200 mL of MNCs per well (5×106 cellsmLꢀ1) were transfer-
red into 96-well culture plates. The RL-stimuli were serially diluted in
HEPES buffer (pH 7.2), added to the cultures at 20 mL per well, and the
cultures were incubated for 4 h at 378C under 5% CO2. Cell-free super-
natants were collected after centrifugation of the culture plates for
10 min at 400 g and then stored at ꢀ208C until determination of the cyto-
kine content. Immunological determination of TNFa in the cell superna-
tant was performed in a sandwich-ELISA as described previously:[14] 96-
well plates (Greiner, Solingen, Germany) were coated with a monoclonal
(mouse) anti-human TNFa antibody (clone 16 from Intex AG, Switzer-
land). Cell-culture supernatants and the standard (recombinant human
TNFa, Intex) were diluted with buffer. After exposure to appropriately
diluted test samples and serial dilutions of standard rTNFa, the plates
were exposed to peroxidase-conjugated (sheep) anti-mouse IgG antibody.
Subsequently, the color reaction was started by addition of tetramethyl-
benzidine/H2O2 in alcoholic solution and stopped after 5 to 15 min by ad-
dition of 1n sulfuric acid. In the color reaction, the substrate was cleaved
enzymatically, the product was measured photometrically by using an
ELISA reader at a wavelength of 450 nm, and the values were correlated
to the standard. TNFa was determined in duplicate at three different di-
lutions and the values were averaged.
Rhamnolipid RL-2,214,14-COOH (1): The fully deprotected rhamnolipid
methyl ester RL-2,214,14-COOMe 1914,14 (104.6 mg, 0.135 mmol) and lipase
from Candida antarctica (270 mg, 2 gmmolꢀ1; CAL-B, Fluka, >2 Umgꢀ1
)
were suspended in DMSO/phosphate buffer (5 mL, 1:4 v/v, phosphate
buffer prepared as a 0.1m solution from Na2HPO4 and NaH2PO4 and ad-
justed to pH 7). Hydrolysis was conducted for 72 h at 408C by shaking
the reaction vessel at 300 rpm. After completion of hydrolysis, the mix-
ture was concentrated and purified by column chromatography (60 mL
SiO2, CHCl3/MeOH/HOAc 30:3:1). Solvents were removed and the pure
rhamnolipid 1 was obtained after three lyophilization steps from tert-bu-
tanol/water (4:1 v/v)as a colorless, fluffy solid (85.7 mg, 83.4%).
1H NMR (600 MHz, CD3COOD): d=5.32 (m, 1H; CH3-O), 5.04 (d,
3J1,2 <1 Hz, 1H; H-1A), 5.00 (d, 3J1,2 <1 Hz, 1H; H-1B), 4.13 (m, 2H;
CH3’-O-Rha, H-2B), 3.91–3.86 (m, 3H; H-3A, H-3B, H-2A), 3.81 (m, 1H;
H-5A), 3.75 (m, 1H; H-5B), 3.57 (t, 3J4,5 =9.3 Hz, 1H; H-4A), 3.50 (t,
3J4,5 =9.3 Hz, 1H; H-4B), 2.64–2.50 (m, 4H; a-CH2, a’-CH2), 1.65–1.24
(m, 46H; CH2, H-6A, H-6B), 0.87 ppm (t, 6H; CH3); 13C NMR (150 MHz,
CD3COOD): d=104.6, 99.5, 81.8, 76.4, 75.8, 75.4, 73.8, 73.7, 73.5, 73.2,
71.4, 71.1, 42.5, 41.6, 36.5, 35.4, 34.4, 32.2, 32.1, 32.1, 32.0, 32.0, 31.9, 31.8,
27.5, 27.2, 25.1, 19.4, 19.3, 16.0 ppm; MALDI-TOF-MS: m/z: 785.4
[M+Na]+, 807.5 [MꢀH+2Na]+; HRMS (FT-ICR-MS): m/z calcd for
C40H74O13Na [M+Na]+: 785.50216; found: 785.50155, Dm/z=1.20 ppm.
Rhamnolipid RL-1,2(R)14,(S)14-COOH (33): The (3’R,3S)-configured rham-
nolipid benzyl ester RL-1,2(R)14,(S)14-COOBn 11 (0.089 mmol, 62.8 mg) was
dissolved in THF (20 mL). Acetic acid (1 mL) was added. After addition
of palladium-charcoal (10% Pd/C, 30 mg), the reaction mixture was vigo-
rously shaken at RT in a standard hydrogenation apparatus at 4 bar H2
for 5 h. The mixture was filtered through a Celite pad, the filtrate was
Chem. Eur. J. 2006, 12, 7116 – 7124
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7123

J. Rademann et al.
[6] P. M. St. Hilaire, M. Meldal, Angew. Chem. 2000, 112, 1210–1228;
Angew. Chem. Int. Ed. 2000, 39, 1163–1179.
Acknowledgements
[7] a) K. Fukase, A. Ueno, Y. Fukase, M. Oikawa, Y. Suda, S. Kusumo-
to, Bull. Chem. Soc. Jpn. 2003, 76, 485–500; b) K. Fukase, Y.
Fukase, M. Oikawa, W. C. Liu, Y. Suda, S. Kusumoto, Tetrahedron
1998, 54, 4033–4050.
The authors thank the Deutsche Forschungsgemeinschaft (DFG) for fi-
nancial support (Ra 895–3) of the presented work and the DFG graduate
college “Chemistry in Interphases” for a stipend to J.B.
[8] H. I. Duynstee, M. J. Van Vliet, G. A. van der Marel, J. H. van
Boom, Eur. J. Org. Chem. 1998, 303–307.
[9] J. Bauer, J. Rademann, J. Am. Chem. Soc. 2005, 127, 7296–7297.
[10] S. Häussler, M. Nimtz, T. Domke, V. Wray, I. Steinmetz, Infect.
Immun. 1998, 66, 1588–1593.
[1] a) R. Medzhitov, C. A. Janeway, Jr., Immunol. Rev. 2000, 173, 89–
97; b) R. Medzhitov, C. A. Janeway, Jr., N. Engl. J. Med. 2000 343,
338–344.
[2] a) L. Van Kaer, Nat. Rev. Immunol. 2005, 5, 31–42; b) P. Pizzo, A.
Viola, Curr. Opin. Immunol. 2003, 15, 255–260; c) G. A. Cabral,
Life Sci. 2005, 77, 1699–1710.
[11] Y. Oikawa, K. Sugano, O. Yonemitsu, J. Org. Chem. 1978, 43, 2087–
2088.
[12] R. Noyori, T. Ohkuma, M. Kitamura, H. Takaya, N. Sayo, H. Kumo-
bayashi, S. Akutagawa, J. Am. Chem. Soc. 1987, 109, 5856–5858.
[13] J. A. Dale, H. S. Mosher, J. Am. Chem. Soc. 1973, 95, 512–519.
[14] J. Andrä, P. Garidel, A. Majerle, R. Jerala, R. Ridge, E. Paus, T.
Novitsky, M. H. J. Koch, K. Brandenburg, Eur. J. Biochem. 2004,
271, 2037–2046.
[3] a) D. C. Angus, W. T. Linde-Zwirble, J. Lidicker, G. Clermont, J.
Carcillo, M. R. Pinsky, Crit. Care Med. 2001, 29, 1303–1310; b) H. F.
Geerdes, D. Ziegler, H. Lode, M. Hund, A. Loehr, W. Fangmann, J.
Wagner, Clin. Infect. Dis. 1992, 15, 991–1002.
[4] S. Akira, K. Takeda, Nat. Rev. Immunol. 2004, 4, 499–511.
[5] a) J. Andrä, J. Rademann, J. Howe, M. H. J. Koch, H. Heine, U.
Zähringer, K. Brandenburg, Biol. Chem. 2006, 387, 301–310; b) L.
Schwichtenberg, Doctoral Thesis, Christian-Albrechts University
Kiel, 2003, http://e-diss.uni-kiel.de/diss 809/.
Received: April 5, 2006
Published online: August 17, 2006
7124
www.chemeurj.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7116 – 7124