1
Biophysical characterization of synthetic rhamnolipids
Jörg Howe1, Jörg Bauer2, Jörg Andrä1, Andra B. Schromm1, Martin Ernst1, Manfred Rössle3,
Ulrich Zähringer1, Jörg Rademann3, and Klaus Brandenburg1*
1Forschungszentrum Borstel, Leibniz-Zentrum für Medizin und Biowissenschaften, Parkallee
1-40, D-23845 Borstel, Germany
2Leibniz-Institut für Molekulare Pharmakologie, Robert-Rössle-Str. 10, D-13125 Berlin,
Germany and Institut für Chemie und Biochemie, Freie Universität Berlin, Takustr. 3, D-
14195 Berlin, Germany
3European Molecular Biology Laboratory, Outstation Hamburg, EMBL c/o DESY, Notkestr.
85, D-22603 Hamburg, Germany
Running title: Biophysics of synthetic rhamnolipids
Key words: Rhamnolipids, endotoxins, organic synthesis, cytokine induction, X-ray
diffraction
*Corresponding author
Forschungszentrum Borstel, Leibniz-Zentrum für Medizin und Biowissenschaften, Parkallee
10, D-23845 Borstel, Germany
Tel: +49 4537-188 235, Fax: +49 4537-188 632, E-mail: [email protected]
Rhamnosyn_revidiert_.doc
2
Summary
Synthetic rhamnolipids, derived from a natural diacylated glycolipid, RL-2,214, produced by
Burkholderia (Pseudomonas) plantarii, were analysed biophysically. Changes in the chemical
structures comprised variations in the length, the stereochemistry and numbers of the lipid
chains, numbers of rhamnoses, and the occurence of charged or neutral groups. As relevant
biophysical parameters the gel (β) to liquid crystalline (α) phase behaviour of the acyl chains
of the rhamnoses, their three-dimensional supramolecular aggregate structure, and the ability
of the compounds to intercalate into phospholipid liposomes in the absence and presence of
lipopolysaccharide-binding protein (LBP) were monitored. Their biological activities were
examined as the ability to induce cytokines in human mononuclear cells (MNC) and to induce
chemiluminescence in monocytes. Depending on the particular chemical structures, the
physicochemical parameters as well as the biological test systems show large variations. This
relates to the acyl chain fluidity, aggregate structure, and intercalation ability, as well as the
bioactivity. Most importantly, the data extend our conformational concept of endotoxicity,
based on the intercalation of naturally-originating amphiphilic virulence factors into
membranes from immune cells. This ‘endotoxin conformation’, produced by amphiphilic
molecules with hydrophilic charged backbone and apolar hydrophobic moiety and adopting
inverted cubic aggegate structures, causes a strong mechanical stress in target immune cells
on integral proteins eventually leading to cell activation. Furthermore, biologically inactive
rhamnolipids with lamellar aggregate structures antagonize the endotoxin-induced activity in
a way similar to lipid A-derived antagonists.
Rhamnosyn_revidiert_.doc
3
Beside the well-known bacterial cell wall components such as lipopolysaccharide (LPS,
endotoxin), peptidoglycan (PG) and lipopeptides, which enable the immune system to
recognize and identify the invading pathogen as “non-self”, also virulence factors secreted as
exotoxins by pore-forming toxins may lead to severe infections in mammals [1]. Furthermore,
it was found in Burkholderia pseudomalei cells that also higher concentrations of
rhamnolipids were present in the biomass [2]. These heat-stable extracellular toxins,
characterized as rhamnolipids, exhibited considerable cytotoxic and hemolytic activity as well
as significant antimicrobial activity [2,3]. Their surfactant activity may play a major role in
the degradation of hydrophobic compounds used in the application of the microorganisms in
bioremediation and biotransformation [4]. Recently, we have characterized a rhamnolipid, 2-
O-α-L-rhamnopyranosyl-α-L-rhamnopyranosyl-(R)-3-hydroxytetra-decanoyl-(R)-3-hydroxy-
tetradecanoate (RL-2,214) from Burkholderia (Pseudomonas) plantarii biophysically,
including the characterization of the acyl chain melting of the hydrocarbon chains, the type of
structure of rhamnolipid aggregates, their incorporation into target cell membranes, their
critical micellar concentration, their ability to induce biological activity such as cytokines in
human mononuclear cells and the involvement of cell surface receptors in this process, their
ability to activate the Limulus cascade, and their antibacterial activity [5]. The results showed
a structure-activity relationship similar to that of bacterial LPS on the one hand but with a
number of distinct characteristics on the other. To get more general informations on the
dependence of particular functional groups in the structure-activity relationship, we have
synthesized a variety of rhamnolipids based on the structure of RL-2,214. These variations
comprise number of rhamnoses in the polar moiety (1 to 3), number of acyl chains (2 or 3),
stereochemistry (3(R) and 3(S)-OH) and length of the acyl chains (C:4 to C:18), and head
group charge (negatively charged carboxylate or neutral hydroxy group), see Fig. 1. We have
found that the different physicochemical characteristics of the compounds sensitively depend
on the particular chemical structure, and also determine their ability to cause activity in
Rhamnosyn_revidiert_.doc
4
immune cells. In this way, suitable designed rhamnolipids fulfill the criteria to act in a similar
way as bacterial endotoxins (lipopolysaccharides, LPS) and thus constitute an ‘endotoxic
conformation’.
Results
Cytokine-inducing activity
The ability of the rhamnolipids to induce cytokine production (TNFα) in human mononuclear
cells (MNC) was tested in comparison with rough mutant lipopolysaccharide (LPS Ra) (Fig.
2). Clearly, LPS Ra causes cytokine induction down to at least 1 ng/ml, whereas the activity
of all rhamnolipids presented in Fig. 2 is more than two orders of magnitude lower, but
significantly present down to 100 ng/ml. In detail, RL 9 corresponding to the natural structure
and that with one additional rhamnose residue (RL4) are the strongest activators.
Rhamnolipids RL2, 3, 5, 10 and 11 not listed in the figure are completely inactive, i.e., the
triacylated compounds (RL2 and 3), that with longer acyl chains (RL5), and the non-charged
compounds (RL10 and 11) do not induce any TNFα even up to 10 µg/ml.
It should be noted that after dilution of the rhamnolipids in RPMI nutrient as well as in
destilled water rather than in buffer all biological activities vanished.
Induction of chemiluminescence in monocytes
The ability of the rhamnolipids to induce reactive oxygen intermediates in monocytes was
tested in comparison to that of LPS Ra. From the time kinetics of the chemiluminescence of
the samples at a concentration of 0.1 µg/ml, the average chemiluminescence rates after 45
minutes were o9btained at three concentrations of 0,1, 1, and 3 µg/ml, respectively (Fig. 3).
Clearly, LPS induces the strongest reaction. The natural compound RLnat and compounds
Rhamnosyn_revidiert_.doc
5
RL4 and 9, but also RL7 exhibit significant activity at all concentrations. In particular, except
for RL10, which exhibits a small signal at 3 µg/ml, the rhamnolipids inactive in the cytokine
test are also inactive in this assay.
Fourier transform infrared spectroscopy (FTIR)
For the identification of particular functional groups, infrared spectra of rhamnolipids at 95 %
water concentration were analysed. Vibrational bands around 1410 and 1585 cm-1 for the
rhamnolipids with carboxyl groups were indicative that the hydrated molecules are negatively
charged (data not shown). Furthermore, their spectra exhibit vibrational bands typical for the
hydrophobic region (symmetrical and antisymmetrical) νs and νas stretching vibration of -
CH2- groups around 2920 and 2850 cm-1, respectively. These are sensitive markers of acyl
chain order. In Figure 4, the peak positions of the symmetric stretching vibrational band
νs(CH2), are plotted versus temperature for all investigated rhamnolipids. Clearly, the
temperature course for the natural and synthetic compound RL-2,214 (RLnat and RL9) are
nearly identical, with a phase transition temperature Tc around – 10 °C and a further gradual
increase of the wavenumber at higher temperatures indicating a high fluidity of the acyl
chains (Fig. 4C). RL7 with one rhamnose does not strongly change the picture, whereas RL4
with 3 rhamnoses exhibits a high fluidity at all temperatures (Fig. 4B). The changed
stereospecifity (RL13) is connected with a decrease of Tc to below – 15 °C, but no change of
the fluidity at higher temperatures. The Tc of RL3 with one more acyl chain is increased to –5
°C (Fig. 4B,C), whereas the change of the length of the acyl chains leads to a significant
increase in fluidity for the shorter chain RL8 and a decrease in fluidity and increase in Tc for
the longer chain RL5 (Fig. 4A-C). For compound RL12 with one rhamnose and a changed
stereospecific configuration Tc increases to higher temperatures between –3 and 0 °C (Fig.
4B), and the addition of a third rhamnose (RL2) leads to a Tc of even 15 °C (Fig. 4A).
Rhamnosyn_revidiert_.doc
6
The shortening of one of the acyl chains (RL1 and 6) in each case leads to an increase in
fluidity at all temperatures (Fig. 4A). The conversion of the charged compounds RL7 and 9
into a neutral form (RL11 and 10) leads to a sharpening of the phase transition, but not much
to a change of fluidity at higher temperatures.
X-ray diffraction
Small-angle X-ray diffraction was applied for the elucidation of the aggregate structures of
selected rhamnolipids 3, 4, 7, 10, and 13, differing in the ability to induce cytokines, at high
water contents and at different temperatures.
In Fig. 5, small-angle X-ray diffraction patterns are presented for biologically inactive RL3
and active 7 in the temperature range 5 to 60 °C. The patterns of RL3 (Fig. 5A) are nearly
identical at all temperatures and are indicative for a multilamellar structure deduced from the
periodicity at 3.77 nm and the second order reflection at 1.88 nm (Fig. 5B, bottom). The
patterns for biologically active RL7 (Fig. 5B) are much more complex and exhibit non-
lamellar characteristics, i.e., reflections are observed which are not located at equidistant
ratios. Furthermore, the spacing ratios of the reflection at lowest scattering vector, 8.93 nm at
40 °C, is untypical for a lamellar but would rather correspond to a reflection typical for non-
lamellar cubic structures (Fig. 5B, bottom). However, due to the low number of observable
reflections no assignment to a particular cubic phase is possible.
Similarly, compounds RL4 and 10 were analysed. Inactive RL10 has a mainly lamellar
characteristics with a main reflection at 4.78 (5 °C) and 4.57 nm (20 °C) which decreases to
3.47 nm (40 °C) and 3.41 nm (60 °C) (data not shown). In contrast, for active RL4 the
ocurrence of a reflection located at 1/√2 of the periodicity typical for a cubic structure is
observed (not shown).
Fluorescence resonance energy transfer (FRET)
Rhamnosyn_revidiert_.doc
7
FRET was applied to test the ability of selected rhamnolipids to intercalate into liposomal
membranes made from negatively charged phosphatidylserine (PS, Fig. 6) in the absence and
presence of lipopolysaccharide-binding protein (LBP). The results of Fig. 6 show a change of
the FRET signal after addition at 50 s for all investigated samples RL7,9,10 and 13 indicating
their incorporation into the target liposomes. The addition of LBP after 100 s leads to another
increase of the FRET signal already for the pure liposomes corresponding to an intercalation
of LBP into PS-liposomes. A stronger increase is observed for the rhamnolipids, indicative for
an LBP-mediated intercalation into the liposomes, with the smallest increase for uncharged
RL10, and higher increase in particular for RL7. All measurements were also performed by
using liposomes corresponding to the composition of the macrophage membrane, as described
in the previous report [5], with phosphatidylcholine as the main component. It turned out that
the rhamnolipids intercalated in a similar way into the liposomes, although with a smaller
amplitude (data not shown).
Involvement of cell surface receptors CD14, TLR2, and TLR4
To investigate a possible involvement of the cell surface receptors CD14, TLR2, and TLR4 in
the recognition of the rhamnolipids, its activity in a Chinese hamster ovary (CHO) cell
reporter system was compared with the activity of bacterial LPS, a bacterial lipopeptide
Pam3CSK4, and interleukin-1 (IL-1). The CHO cells are natural TLR2 knockouts and express
human CD25 surface antigen upon induction of NF-κB translocation. Experiments with
CHO/CD14 alone (expressing only endogenousTLR4) show that they react with LPS, but not
with Pam3CSK4 and RL1,4,7, and 9 (data not shown). The reactivity to cells CHO/CD14-
huTLR2 additionally transfected with human TLR2 showed the expected response to
Pam3CSK4, but again there was no significant increase of the TLR2 signal for all investigated
rhamnolipids (not shown).
Rhamnosyn_revidiert_.doc
8
MaxiK channel activity
The ability of the K+- channel (MaxiK) blocker paxilline, which effectively inhibits the
endotoxin-induced cytokine production in macrophages, to influence the TNFα production of
these cells was also tested for RL4, 9, and 13 as compared to control LPS. It was found that
paxilline is able to reduce the rhamnolipid-induced cell activation significantly at
concentrations of 10 and 20 µM at which the LPS-induced activation of cells was blocked
completely (Fig. 7). Of course, it has again to be considered that the activation of
macrophages takes place at a 1000-fold lower concentration of LPS as compared to the
rhamnolipids.
Antagonistic action of inactive rhamnolipids
The rhamnolipids RL2, 3, 5, and 10, which were inactive in the cytokine assay (see above),
were tested with respect to their ability to block the LPS-induced cytokine production in
human mononuclear cells, i.e., to act antagonistically. For this, LPS Re from S.minneota R595
was added to the cells at a concentration of 1 µg/ml and 1 ng/ml, and rhamnolipids were
added at different concentrations with final ratios of [rhamnolipid]:[LPS] 100:1 to 1:1 per
weight. In Fig. 8, the results are given for the RL-2:LPS system. Clearly, at both stock
concentrations the addition of RL-2 to LPS in excess leads to an inhibition of the LPS-
induced TNFα-production, i.e., an antagonistic action takes place. This was found to be
similarly true for the other rhamnolipids RL-3, RL-5, and RL-10 (data not shown).
Discussion
Natural rhamnolipid exotoxins have been described to be antimicrobial, but also cytotoxic to
human cells, in a way of a detergent-like action on target cells [2,4,6]. Also, they have been
Rhamnosyn_revidiert_.doc
9
shown to stimulate the uptake of hydrophobic compounds by bacteria [4]. Recently, the
natural rhamnolipid RL-2,214 from Burkholderia plantarii was characterized
physicochemically and with respect to its ability to act as a virulence factor [5]. It has been
found that this rhamnolipid exhibits a variety of endotoxin-related physicochemical
characteristics such as a cubic-inverted aggregate structure, a tendency to intercalate into
target cell membranes, and a suppression of its cytokine induction in mononuclear cells by
polymyxin B due to neutralization of the negative head group charges. A detergent-like
action, however, could not be confirmed, and RL-2,214 also did not show antimicrobial
activity.
For a more general understanding of the underlying mechanisms, we have synthesized a
variety of rhamnolipids differing in the acylation pattern, the number of monosaccharide
residues, and the charge (Fig. 1, Tab. 1). The biological data show clearly, that the active
compounds in two independent test systems (Fig. 2,3) adopt non-lamellar aggregate
structures as exemplarily demonstrated for RL7 (Fig. 5B), whereas the inactive compounds
adopt multilamellar aggregate structures (Fig. 5A). Furthermore, an intercalation into
negatively charged PS liposomes takes place which is enhanced by the action of LBP (Fig. 6).
In contrast, there is apparently no dependence of biological activity on the phase state or the
acyl chain fluidity of the rhamnolipids. For example, inactive RL5 has a relatively high Tc,
whereas for active RL9 it is very low (Fig. 4C). The phase transition behaviour is governed by
rules which also determine the phase transition behaviour of phospholipids [7]. Thus, an
addition of a third acyl chain (RL2) to a diacylated compound (RL7) leads to a drastic
increase of the hydrophobic bulk and, with that, to a considerable increase in Tc from –10 °C
to 15 °C (Fig. 4A,B). Interestingly, the removal of the charge from RL9 leading to RL10 has
– in contrast to the drastic change in bioactivity – nearly no impact on the phase behaviour
(Fig. 4C). Furthermore, the addition of more rhamnoses leads to a fluidization of the acyl
chains (compare mono-rhamnose RL7 with tri-rhamnose RL4, Fig. 4B). This is apparently
Rhamnosyn_revidiert_.doc
10
due to the fact that a higher saccharide bulk between neighbouring molecules leads to a looser
packing resulting in more space available for the acyl chains and thus, an increase in fluidity.
Noteworthy is the observation that the rhamnolipids became inactive in pure water (aqua dest)
as well as in RPMI nutrient. The former solvent is of course unphysiological, and we have
observed that also endotoxins loose much activity (one order of magnitude) when diluted in
destilled water (unpublished results). RPMI, in contrast, is such a complex mixture of
different compounds that the reason for the inactivation of the rhamnolipids remains unclear.
These results, on the other hand, confirm the necessity to use physiological buffers in
biophysical experiments.
Regarding the influence of cell activation on the presence of particular receptors, it can be
stated that neither TLR2 nor TLR4 which are important cell membrane receptors for
lipopeptide and lipopolysaccharide structures, respectively [8-10], are recognition structures
for the rhamnolipids (data not shown). In contrast, by adding the specific MaxiK channel
blocker paxilline to macrophages, not only the cell activation by LPS, but also by the active
rhamnolipids is inhibited or at least reduced (Fig. 7). This means that the MaxiK is involved
in the process of signal transduction into the cell interior. It seems obvious that similar as
observed for LPS not only one single receptor but a complete receptor cluster are involved in
cell signalling [11]. Thus, Triantafilou et al. [12] have found that different LPS and LPS part
structures trigger the recruitment of different receptors within microdomains, and the
composition of each receptor cluster seem to determine whether an immune response will be
induced or inhibited.
Independent which membrane proteins are reponsible for cell signalling, apparently one
general principle governs the process of cell activation by amphiphilic compounds: One main
prerequisite for the induction of bioactivity is the adoption of a non-lamellar, preferentially
cubic structure, and another prerequisite is the incorporation into target cell membranes either
by itself or mediated by LBP. In the membrane, the amphiphilic compounds are present in
Rhamnosyn_revidiert_.doc
11
domains, and cause a conformational change of signalling proteins. For LPS, these processes
could be verified as published previously [13]. In particular, the role of LBP and the
intercalation of LPS aggregates into target membranes could be proven [14,15]. The
interpretation that rhamnolipids may assume domains in target cell membranes is supported
by a recent study of Sanchez et al [16], who observed domain formation of a natural
dirhamnolipid in a phosphatidylethanolamine matrix.
Together with the present work for synthetic rhamnolipids we think that the above presented
signalling pathway holds true beside for LPS also for a glycolipid from Mycoplasma
fermentans [16], for synthetic phospholipid-like molecules [17,18] various monophosphoryl
lipid A analogues, and for bacterial lipopeptides (unpublished).
Interesting is the observation of an antagonistic action of those rhamnolipids, which are
themselves agonistically inactive (see Fig. 8). Compounds on the basis of lipid A part
structures such as tetraacyl lipid A - synthetic compound ‘406’ – need, beside a multilamellar
aggregate structure also a negative head group charge to be antagonistic [19]. For the
rhamnolipids, in contrast, this is not necessary, since also RL10 as uncharged compound was
antagonistic similar as the other compounds RL2, 3, and 5. As stated earlier, the necessity for
the lipid A-like compounds to have a negative charge results from the fact that they are not
able to intercalate into target cell membranes by themselves, a transport protein such as LBP
is needed [20]. For the rhamnolipids, this is apparently not necessary since they can
intercalate by themselves into target cell membranes (Fig. 6) independently of the presence of
a negative net charge.
Experimental Procedures
Synthesis
Rhamnosyn_revidiert_.doc
12
The efficient parallel synthesis of rhamnolipids was accomplished by means of the previously
developed concept of Hydrophobically Assisted Switching Phase (HASP) synthesis [21].
Rhamnolipids bearing a di-/trilipid moiety allowed to easily adapt the HASP-method of
flexible switching between solution phase steps and solid-supported reactions on a reversed
phase hydrophobic silica support (bulk RP18) to the assembly of a rhamnolipid library. The
construction of rhamnolipids based on the lead structure RL-2,214 started from
enantiomerically pure β-hydroxycarboxylic esters and -acids of various lengths C4-C18
obtained by enantioselective catalytic Noyori hydrogenation. Both, the esterification of β-
hydroxycarboxylic acids and –esters, removal of protecting groups and subsequent iterative
glycosylation cycles with the tailormade rhamnose donor 3,4-O-(2,3-dimethoxybutane-2,3-
diyl)-2-O-phenoxyacetyl-α-L-rhamnopyranosyl-trichloroacetimidate [22] were performed as
high-yielding parallel HASP-steps (Fig. 9).
Whereas the final hydrolysis of (3R,3’R) configured rhamnolipid methyl esters was performed
with solid-supported lipase from Candida antarctica which furnished the (3R,3’R)-
rhamnolipid acids in good to excellent yields, (3R,3’S) configured rhamnolipid acids were not
recognized by the enzyme and had to be accessed via a debenzylation-route of the
correspondent (S)-3-HO-benzyl esters. Rhamnolipid alcohols required the use of terminally
reversed esters which were accessed via a (R)-3-(2-methoxy ethoxymethyl) protected lipid
actetate.
Lead compound of the synthesis was RL-2,214 (RL9) corresponding to the structure of the
natural rhamnolipid. Removal of one rhamnose leads to RL-1,214 (RL7), addition of one
rhamnose to RL-3,214 (RL4), change of the stereoisomeric configuration to RL-2,2(R)14,(S)14
(RL13), addition of one acyl chains to RL-2,314 (compound 3), and change of the length of the
acyl chains to RL-2,212 (RL8) and to RL-2,218 (RL5). The compound with one rhamnose
(RL7) was additionally modified by changing the stereospecific configuration RL-1,2(R)14,(S)14
(RL12), by adding one additional acyl chain RL-1,314 (RL2), or by shortening the respective Rhamnosyn_revidiert_.doc
13
acyl chains (RL 1 and 6). Finally, the carboxylates of compounds with one and two
rhamnoses (RL7 or 9) were changed into a non-charged hydroxy group ( RL11 and 10).
The chemical structures and the nomenclature used for the single compounds are listed in Fig.
1 and Tab. 1.
Other Reagents
Lipopolysaccharides from the rough mutant Re or Ra from Salmonella minnesota (R595 or
R60), respectively, were extracted by the phenol/chloroform/petrol ether method [23] from
bacteria grown at 37 °C, purified, and lyophilized. The lipopeptide palmitoyl3CSK4 was from
EMC Microcollections (Tübingen, Germany). Lipopolysaccharide-binding protein (LBP) was
a kind gift of Russ L. Dedrick (XOMA Co, Berkeley, Ca, USA) and it was stored at -70°C as
a 1 mg/mL stock solution in 10 mM Hepes, pH 7.5, 150 mM NaCl, 0.002% (v/v) Tween 80,
0.1% F68. Bovine brain 3-sn-phosphatidylserine (PS), egg 3-sn-phosphatidylcholine (PC), 3-
sn-phosphatidylethanolamine (PE), and sphingomyelin from bovine brain were from Sigma.
Lipid sample preparation
All lipid samples were prepared as aqueous suspensions in 20 mM Hepes, pH 7. For this, the
lipids were suspended directly in buffer and were temperature-cycled 3 times between 5 and
70 °C and then stored for at least 12 h before measurement. To guarantee physiological
conditions, the water content of the samples was usually around 95 %.
For preparations of liposomes from phosphatidylserine or from a mixture corresponding to the
phospholipid composition of the macrophage membrane (phosphatidylcholine,
phosphatidylserine, phosphatidylethanolamine, and sphingomyelin in a molar ratio of
1:0.4:0.7:0.5), the lipids were solubilized in chloroform, the solvent was evaporated under a
stream of nitrogen, and the lipids were resuspended in the appropriate volume of buffer and
Rhamnosyn_revidiert_.doc
14
treated as described above (temperature-cycling). The resulting liposomes are large and
multilamellar as detected in some electron microscopic experiments (kindly performed by H.
Kühl, Div. of Pathology, Forschungszentrum Borstel).
FTIR spectroscopy
The infrared spectroscopic measurements were performed on an IFS-55 spectrometer (Bruker,
Karlsruhe, Germany). For phase transition measurements, the lipid samples were placed
between CaF2 windows with a 12.5 µm Teflon spacer. Temperature scans were performed
automatically between -20 and 70 °C with a heating rate of 0.6 °C/min. Every 3 °C, 50
interferograms were accumulated, apodized, Fourier-transformed, and converted to
absorbance spectra.
X-ray diffraction
X-ray diffraction measurements were performed at the European Molecular Biology
Laboratory (EMBL) outstation at the Hamburg synchrotron radiation facility HASYLAB
using the SAXS camera X33 [24]. Diffraction patterns in the range of the scattering vector 0.1
< s < 4.5 nm-1 (s = 2 sin θ/λ, 2θ scattering angle and λ the wavelength = 0.15 nm) were
recorded at 40 °C with exposure times of 1 min using an image plate detector with online
readout (MAR345, MarResearch, Norderstedt/Germany). The s-axis was calibrated with Ag-
Behenate which has a periodicity of 58.4 nm. The diffraction patterns were evaluated as
described previously [25] assigning the spacing ratios of the main scattering maxima to
defined three-dimensional structures. The lamellar and cubic structures are most relevant here.
They are characterized by the following features:
(1) Lamellar: The reflections are grouped in equidistant ratios, i.e., 1, 1/2, 1/3, 1/4, etc. of the
lamellar repeat distance dl
Rhamnosyn_revidiert_.doc
15
(2) Cubic: The different space groups of these non-lamellar three-dimensional structures
differ in the ratio of their spacings. The relation between reciprocal spacing shkl = 1/dhkl and
lattice constant a is
shkl = [(h2 + k2 + l2) / a ]1/2
(hkl = Miller indices of the corresponding set of plane).
Fluorescence resonance energy transfer spectroscopy (FRET)
Intercalation of the rhamnolipids into liposomes made from phosphatidylserine (PS) alone or
mediated by lipopolysaccharide-binding protein (LBP), was determined by FRET
spectroscopy applied as a probe dilution assay [20]. To the liposomes, which were labelled
with the donor dye NBD-phosphatidylethanolamine (NBD-PE) and acceptor dye Rhodamine-
PE, first the lipids and then LBP, or vice versa, were added, all at a final concentration of 1
µM. Intercalation was monitored as the increase of the ratio of the donor intensity ID at 531
nm to that of the acceptor intensity IA at 593 nm (FRET signal) in dependence on time.
Stimulation of mononuclear cells (MNC)
MNC were isolated from heparinized (20 IE/ml) 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 (21 °C, 500 g). The interphase layer of
mononuclear cells was collected and washed twice in Hank’s medium and once in RPMI
1640 containing 2 mM L-glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin. The
cell number was equilibrated at 5≅106 N/mL. For stimulation, 200 µl/well MNC (1≅106 cells)
were transferred into 96-well culture plates. The stimuli were serially diluted in RPMI-1640
and added to the cultures at 20 µl per well. The cultures were incubated for 4 h at 37 °C under
5% CO2. Cell-free supernatants were collected after centrifugation of the culture plates for 10
min at 400⋅g and stored at –20 °C until determination of the cytokine content.
Rhamnosyn_revidiert_.doc
16
Stimulation of macrophages
Monocytes were isolated from peripheral blood taken from healthy donors by the Hypaque-
Ficoll density gradient method. To differentiate the monocytes from the macrophages, cells
were cultivated in Teflon bags in the presence of 2 ng/ml M-CSF in RPMI 1640 medium
(endotoxin < 0.01 EU/ml in Limulus test; Biochrom, Berlin, Germany) containing 2 mM L-
glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin, and 4 % heat-inactivated human
serum type AB at 37 °C and 6 % CO2. On day 6 the cells were washed with PBS, detached by
trypsin-EDTA treatment and seeded at 1 ≅ 105/ml in complete medium in 96-well tissue
culture plates (NUNC, Wiesbaden, Germany). After stimulation of the cells with the
rhamnolipidsfor 4 h, cell-free supernatant of duplicate samples were collected, pooled and
stored at –20 °C until determination of cytokine content.
Determination of TNFα concentration
Immunological determination of TNFα in the cell supernatant was performed in a sandwich-
ELISA as described before [5]. 96-well plates (Greiner, Solingen, Germany) were coated with
a monoclonal (mouse) anti-human TNFα antibody (clone 16 from Intex AG, Switzerland).
Cell culture supernatants and the standard (recombinant human TNFα, Intex) were diluted
with buffer. After exposure to appropriately diluted test samples and serial dilutions of
standard rTNFα, the plates were exposed to peroxidase conjugated (sheep) anti-TNFα IgG
antibody. Subsequently, the color reaction was started by addition of
tetramethylbenzidine/H2O2 in alcoholic solution and stopped after 5 to 15 min by addition of
1N sulfuric acid. In the color reaction, the substrate is cleaved enzymatically, and the product
was measured photometrically on an ELISA reader at a wavelength of 450 nm and the values
Rhamnosyn_revidiert_.doc
17
were related to the standard. TNFα was determined in duplicate at two different dilutions and
the values were averaged.
To study the influence of the K+-channel (MaxiK) on cytokine induction, the specific channel
blocker paxilline was added at a concentration of 10 and 20 µM 10 min before stimulation by
the rhamnolipids to the mononuclear cells, and incubated at 37 °C.
Chemiluminescence of isolated monocytes
Peripheral blood monocytes were isolated from MNC by counterflow centrifugation
(elutriation) using the JE-6B-elutriator system (Beckman Instruments Inc., Palo Alto, CA,
USA) as described earlier [26]. Monocytes (200.000) were suspended in a modified RPMI-
medium (RPMI-1640-medium without phenol red and sodium bicarbonate but containing 20
mmol /l HEPES [Biochrom, Berlin, Germany]) and the monocytes in a final volume of
200 µl per well were placed in a 96 flat bottom white wells plate (Microlite TCT Flat Bottom
Plate, Dynex Technologies, Inc. Chantilly, VA, USA) . Then the plate was incubated at 37°
for at least 60 min before chemiluminescence measurement. Thereafter the plate was put into
the microplate luminometer MicroLumatPlus (LB 96V, Berthold Technologies, Bad Wildbad,
Germany) and 10 minutes prior to the chemiluminescence measurements luminol (5-amino-
2,3-dihydro-1,4-phthalazinedione, Sigma, Taufkirchen, Germany) was added (10 µl per well
of a 2 mg/ml solution) as the chemiluminescence mediating compound. Then after addition of
2 µl of medium (unstimulated control), of LPS, of natural rhamnolipid compound or
synthetic rhamnolipids, the chemiluminescence of the wells was recorded for 45 minutes
whereby the plate was always kept at 37°C in the luminometer. The data are shown as photon
count rates in relative light units per second (RLU/sec) or as mean RLU per 45 minutes (cf.
Figs. 3).
Rhamnosyn_revidiert_.doc
18
Antagonistic action of inactive rhamnolipids
The rhamnolipid samples which were found not to induce any cytokines in human
mononuclear cells were investigated with respect to their ability to block the LPS-induced
TNFα-production in the mononuclear cells. For this, LPS Re from S. minnesota R595 was
prepared at two concentrations 1 µg/ml and 1 ng/ml, and the rhamnolipids were added up to
an exess of 100:1 excess (w/w ).
Activation of CHO reporter cells
The CHO/CD25 reporter cell line, clone 3E10, is a stably transfected CD14-positive CHO
(Chinese hamster ovary) cell line that expresses inducible membrane CD25 (Tac antigen)
under transcriptional control of the human E-selectin promoter pELAM.Tac [27]. It reacts
sensitively to the activation of nuclear factor NF-κB. A TLR2-expressing cell line was
generated by stable transfection of clone 3E10 with human TLR2 (3E10-TLR2).
Acknowledgements
We thank K. Stephan, S. Groth, G. von Busse, and C. Hamann for performing the cytokine
induction assay, the paxilline assay, the infrared, and FRET measurements, respectively. We
thank the Deutsche Forschungsgemeinschaft (SFB 470, project B4 U.Z. and Ra895-3/1 to J.R.
) for financial support, and a stipend from the graduate college ‘Chemie in Interphasen’ to
J.B.
Rhamnosyn_revidiert_.doc
19
References
[1] Cosson P, Zulianello L, Join-Lambert O, Faurisson F, Gebbie L, Benghezal M, Van Delden C, Curty LK & Kohler T (2002) Pseudomonas aeruginosa virulence analyzed in a Dictyostelium discoideum host system. J Bacteriol 184, 3027-3033.
[2] Häussler S, Rohde M, von Neuhoff N, Nimtz M & Steinmetz I (2003) Structural and functional cellular changes induced by Burkholderia pseudomallei rhamnolipid. Infect Immun 71, 2970-2975.
[3] Häussler S, Nimtz M, Domke T, Wray V & Steinmetz I (1998) Purification and characterization of a cytotoxic exolipid of Burkholderia pseudomallei. Infect Immun 66, 1588-1593.
[4] Noordman WH & Janssen DB (2002) Rhamnolipid stimulates uptake of hydrophobic compounds by Pseudomonas aeruginosa. Appl Environ Microbiol 68, 4502-4508.
[5] Andrä J, Rademann J, Howe J, Koch MH, Heine H, Zähringer U & Brandenburg K (2006) Endotoxin-like properties of a rhamnolipid exotoxin from Burkholderia (Pseudomonas) plantarii: immune cell stimulation and biophysical characterization
Biol Chem 387, 301-310.
[6] Haba E, Pinazo A, Jauregui O, Espuny MJ, Infante MR & Manresa A (2003) Physicochemical characterization and antimicrobial properties of rhamnolipids produced by Pseudomonas aeruginosa 47T2 NCBIM 40044. Biotechnol Bioeng 81, 316-322.
[7] Mantsch HH & McElhaney RN (1991) Phospholipid phase transitions in model and biological membranes as studied by infrared spectroscopy. Chem Phys Lipids 57, 213-226.
[8] Beutler B (2000) Endotoxin, toll-like receptor 4, and the afferent limb of innate immunity. Curr Opin Microbiol 3, 23-28.
[9] Kopp EB & Medzhitov R (1999) The Toll-receptor family and control of innate immunity. Curr Opin Immunol 11, 13-18.
[10] Krutzik SR, Sieling PA & Modlin RL (2001) The role of Toll-like receptors in host defense against microbial infection. Curr Opin Immunol 13, 104-108.
[11] Triantafilou K, Triantafilou M & Dedrick RL (2001) A CD14-independent LPS receptor cluster. Nature Immunol 2, 338-345.
[12] Triantafilou M & Triantafilou K (2002) Lipopolysaccharide recognition: CD14, TLRs and the LPS- activation cluster. Trends Immunol 23, 301-304.
[13] Brandenburg K & Wiese A (2004) Endotoxins: relationships between structure, function, and activity. Curr Top Med Chem 4, 1127-1146.
[14] Gutsmann T, Mueller M, Carroll SF, MacKenzie RC, Wiese A & Seydel U (2001) Dual role of lipopolysaccharide (LPS)-binding protein in neutralization of LPS and
Rhamnosyn_revidiert_.doc
20
enhancement of LPS-induced activation of mononuclear cells. Infect Immun 69, 6942-6950.
[15] Gutsmann T, Haberer N, Carroll SF, Seydel U & Wiese A (2001) Interaction between lipopolysaccharide (LPS), LPS-binding protein (LBP), and planar membranes. Biol Chem 382, 425-434.
[16] Sanchez M, Teruel JA, Espuny MJ, Marques A, Aranda FJ, Manresa A, & Ortiz A (2006)Modulation of the physical properties of dielaidoylphosphatidylethanolamine membarnes by a dirhamnolipid biosurfactant produced by Pseudomonas aeruginosa. Chem. Phys. Lipids 142, 118-127.
[17] Brandenburg K, Wagner F, Müller M, Heine H, Andrä J, Koch MHJ, Zähringer U & Seydel U (2003) Physicochemical characterization and biological activity of a glycoglycerolipid from Mycoplasma fermentans. Eur J Biochem 270, 3271-3279.
[18] Seydel U, Hawkins L, Schromm AB, Heine H, Scheel O, Koch MH & Brandenburg K (2003) The generalized endotoxic principle. Eur J Immunol 33, 1586-1592.
[19] Brandenburg K, Hawkins L, Garidel P, Andrä J, Müller M, Heine H, Koch MHJ & Seydel U (2004) Structural Polymorphism and Endotoxic Activity of Synthetic Phospholipid-like Amphiphiles. Biochemistry 43, 4039-4046.
[20] Schromm AB, Brandenburg K, Loppnow H, Moran AP, Koch MHJ, Rietschel ETh & Seydel U (2000) Biological activities of lipopolysaccharides are determined by the shape of their lipid A portion. Eur J Biochem 267, 2008-2013.
[21] Gutsmann T, Schromm AB, Koch MHJ, Kusumoto S, Fukase K, Oikawa M, Seydel U & Brandenburg K (2000) Lipopolysaccharide-binding protein-mediated interaction of lipid A from different origin with phospholipid membranes. Phys Chem Chem Phys 2, 4521-4528.
[22] Bauer J, Brandenburg K, Zähringer U/ & Rademann J (2006) Chemical synthesis of a glycolipid library by a solid phase strategy allows to elucidate the structural specificity of immune stimulation by rhamnolipids. Chem Eur J. (in press).
[23] Bauer J & Rademann J (2005) Hydrophobically assisted switching phase synthesis: The flexible combination of solid-phase and solution-phase reactions employed for oligosaccharide preparation. J Am Chem Soc 127, 7296-7297.
[24] Galanos C, Lüderitz O & Westphal O (1969) A new method for the extraction of R lipopolysaccharide, Eur. J. Biochem. 9, 245-249.
[25] Koch MHJ (1988) Instruments and methods for small-angle scattering with synchrotron radiation. Makromol Chem Macromol Symp 15, 79-90.
[26] Brandenburg K, Richter W, Koch MHJ, Meyer HW & Seydel U (1998) Characterization of the nonlamellar cubic and HII structures of lipid A from Salmonella enterica serovar Minnesota by X-ray diffraction and freeze-fracture electron microscopy. Chem Phys Lipids 91, 53-69.
Rhamnosyn_revidiert_.doc
21
[27] Grage-Griebenow E, Lorenzen D, Fetting R, Flad HD & Ernst M (1993) Phenotypical and functional characterization of Fc gamma receptor I (CD64)-negative monocytes, a minor human monocyte subpopulation with high accessory and antiviral activity.
Eur J Immunol 23, 3126-3135. [28] Schromm AB, Lien E, Henneke P, Chow JC, Yoshimura A, Heine H, Latz E, Monks
BG, Schwartz DA, Miyake K & Golenbock DT (2001) Molecular genetic analysis of an endotoxin nonresponder mutant cell Line: A point mutation in a conserved region of MD-2 abolishes endotoxin-induced signaling. J Exp Med 194, 79-88.
Figure legends:
Fig. 1: Chemical structures of synthetic rhamnolipid structures.
Fig. 2: Production of TNFα by human mononuclear cells induced by various rhamnolipids in
comparison with LPS from Salmonella minnesota R60. Rhamnolipids RL2,3,5, 10,
and 11 not shown here are completely inactive at all measured concentrations. The
error bars result from the determination of TNFα in triplicate.
Fig. 3a: Kinetics of chemiluminescence in monocytes stimulated with 0.1 µg/ml LPS Re and
various natural (RLnat) and synthetic rhamnolipids.
Fig. 3b: Average chemiluminescent intensity after 45 min for LPS Re and various natural
(RLnat) and synthetic rhamnolipids.
Fig. 4: Gel to liquid crystalline phase behaviour of natural and synthetic rhamnolipids
presented as peak position of the symmetric vibrational band of the methylene groups
versus temperature. In the gel phase of the acyl chains, the peak position is located at
2849 to 2850 cm-1, in the liquid crystalline at 2852.5 to 2853.5 cm-1.
Fig. 5: Synchrotron X-ray diffraction patterns of rhamnolipids RL3 (A) and RL7 (B) in the
temperature range 5-60 °C (top) and at 40 °C (bottom). The scattering vector s = 2 sin
θ/ λ is plotted versus the logarithm of the scattering intensity log I.
Fig. 6: Fluorescence resonance energy transfer spectroscopic (FRET) measurements with
liposomes from phosphatidylserine as FRET signal ID/IA versus time. The
Rhamnosyn_revidiert_.doc
22
rhamnolipids were added at 50 s, and LBP at 100 s to the liposomes. The final
concentrations of liposomes and rhamnolipids were 1 µM, and of LBP 0.1 µM.
Fig. 7: TNFα production of human macrophages induced by rhamnolipids RL4, 9, and 13 at
concentrations 1 and 10 µg/ml and in the presence of the specific MaxiK channel
blocker paxilline at 10 and 20 µM.
Fig. 8: TNFα production of human mononuclear cells induced by two concentrations of LPS
Re (1 µg/ml and 1 ng/ml) in the presence of various concentrations of inactive
rhamnolipid RL2 (antagonistic activity).
Fig. 9: Schematic representation of the chemical synthesis of the rhamnolipids presented in
Fig. 1.
Rhamnosyn_revidiert_.doc
23
RL No. M Compound [g/mol]
1 476,60 R-4-14 2 843,22 R-14-14-14 3 989,36 R-R-14-14-14 4 909,15 R-R-R-14-14 5 875,22 R-R-18-18 6 476,60 R-14-4 7 616,87 R-14-14 8 706,90 R-R-12-12 9 763,01 R-R-14-14
10 749,0 R-R-14-14-OH 11 602,90 R-14-14-OH
12 616,9 R-14-(S)14 13 763,01 R-R-14-(S)14
Table 1: Schematic listing of the rhamnolipids investigated.
Rhamnosyn_revidiert_.doc
24
Fig. 1
Rhamnosyn_revidiert_.doc
25
R60 RL1 RL4 RL6 RL7 RL8 RL9 RL12 RL130
200
400
600
800
1000
1200
1400
1600
LPS R60
100 ng/ml 10 ng/ml 1 ng/ml
Fig.2
Con
cent
ratio
n TN
Fα (p
g/m
l) Rhamnolipids
10 µg/ml 1 µg/ml 100 ng/ml
Rhamnosyn_revidiert_.doc
26
Fig. 3
Rhamnosyn_revidiert_.doc
27
-20 -10 0 10 20 30 40 502849
2850
2851
2852
2853
2854C
RL nat RL 9 RL 5 RL 3 RL 10
Temperature (°C)
-10 0 10 20 302850
2851
2852
2853
2854
RL 3 RL 8 RL 4 RL 7 RL 13 RL 12W
aven
umbe
r (cm
-1) -10 0 10 20 30
2850
2851
2852
2853
2854
B
A
RL 6RL 11 RL 1RL 8RL 2
Fig. 4
Rhamnosyn_revidiert_.doc
28
0,1 0,2 0,3 0,4 0,5 0,6 0,7
0,1 0,2 0,3 0,4 0,5 0,6 0,7
5°C 20°C 40°C 60°C
log
I
40 °C
1.88 nm
3.77 nm
log
I
s (nm-1)
Fig. 5A
Rhamnosyn_revidiert_.doc
29
0,1 0,2 0,3 0,4 0,5
0,1 0,2 0,3 0,4 0,5 0,6 0,7
5°C 20°C 40°C 60°C
log
I
4.86 nm
40 °C
2.98 nm
4.46 nm8.93 nm
log
I
s (nm-1)
Fig. 5B
Rhamnosyn_revidiert_.doc
30
0 50 100 150 200 250 300
0,8
1,0
1,2
1,4
1,6
1,8
2,0
2,2RL13RL7RL9
RL10
Buffer/LBP
Buffer/Buffer
+LBP+ Buffer/RL's
FRET
sig
nal (
I D/I A)
Time (s)
Fig. 6
Rhamnosyn_revidiert_.doc
31
RL4 10
µg/m
l
RL4 1µ
g/mL
RL9 10
µg/m
L
RL9 1µ
g/mL
RL13 1
0µg/m
L
RL13 1
µg/m
L
0
100
200
300
400
500
600
700
800
900
1000
[Paxilline]
Con
ccen
tratio
n of
TN
Fα (p
g/m
L)
Samples
0 µM 10 µM 20 µM
Fig. 7
Rhamnosyn_revidiert_.doc
32
0
400
800
1200
1600
2000
0:1100:1
50:120:1
5:11:11:00:1100:1
1:0 50:1 20:1
5:11:1
1 ng/mlLPS Re
1 µg/ml LPS Re
TN
Fα p
rodu
ctio
n of
m
onon
ucle
ar c
ells
(pg/
ml)
[RL2]:[LPS Re] (weight ratio)
Fig. 8
Rhamnosyn_revidiert_.doc
33
Fig. 9
Rhamnosyn_revidiert_.doc