Glycobiology vol. 15 no. 1 pp. 101–108, 2005
doi:10.1093/glycob/cwh142
Advance Access publication on September 8, 2004
Sinorhizobium meliloti strain 1021 produces a low-molecular-mass capsular
polysaccharide that is a homopolymer of 3-deoxy-D-manno-oct-2-ulosonic
acid harboring a phospholipid anchor
N. Fraysse1,2, B. Lindner3, Z. Kaczynski4, L. Sharypova2,
O. Holst4, K. Niehaus2, and V. Poinsot5
2
Department of Genetics, Faculty of Biology, University of Bielefeld,
D-33615 Bielefeld, Germany; 3Division of Biophysics, Research Center
Borstel, Leibniz Center for Medicine and Biosciences,
D-23845 Borstel, Germany; 4Division of Structural Biochemistry,
Research Center Borstel, Leibniz Center for Medicine and Biosciences,
D-23845 Borstel, Germany; and 5Laboratoire des IMRCP, Universite
Paul Sabatier, F-31062 Toulouse, France
Received on July 20, 2004; revised on August 19, 2004;
accepted on August 23, 2004
Sinorhizobium meliloti strain 1021 possesses the particularity
to synthesize biologically inefficient capsular polysaccharides
(KPS). It has been assumed that this class of compounds is not
produced in high-molecular-mass (HMM) forms, even if
many genetic analyses show the existence of expression of
genes involved in the biosynthesis of capsular polysaccharides.
The expression of these genes that are involved in the export
of a KPS throughout the membrane and in the attachment of
a lipid moiety has never been related to a structurally characterized surface polysaccharide. It is now reported that
S. meliloti strain 1021 produces low-molecular-mass polysaccharides (4–4.5 kDa) that are exclusively composed of
b-(2!7)-linked 3-deoxy-D-manno-oct-2-ulopyranosonic acid
(Kdo) residues. These compounds are considered precursor
molecules of HMM KPS, whose biosynthesis is arrested in
the case of S. meliloti strain 1021. For the first time, the
phospholipid anchor of a rhizobial KPS has been found, and
its structure could be partially identified—namely, a phosphoglycerol moiety bearing a hydroxy-octacosanoic acid. When
compared to other rhizobial KPS (composed of dimeric
hexose-Kdo-like sugar repeating units), the Kdo homopolymer described here may explain why a complementation of
S. meliloti strain 1021 Exo B mutant with an effective rkpZ
gene restoring an active higher KPS size does not completely
lead to the fully effective nitrogen fixing phenotype.
Key words: capsular polysaccharides/Fourier transform ion
cyclotron resonance/phospholipic anchor/QTOF MS/
Sinorhizobium meliloti
Introduction
The Gram-negative bacterium Sinorhizobium meliloti
belonging to the Rhizobia family is able to infect leguminous
1
To whom correspondence should be addressed; e-mail:
[email protected]
plants resulting in a nitrogen-fixing symbiosis. Together
with its host, alfalfa, these bacteria constitute one of the
most studied symbiotic partners. S. meliloti produces different classes of mucoid polysaccharides that play a key role
in its interactions with the host, leading to a successful
infection (for review see Carlson et al., 1999; Fraysse et al.,
2003; Noel and Duelli, 2000; Spaink, 2002). They consist of
(1) exopolysaccharides (EPSs) that are excreted in the external environment and play a role in the suppression of the
active plant defense response; (2) the cyclic glucans, whose
precise role remains unclear but could mediate the root
attachment and the hypoosmotic adaptation of the bacterial cells; (3) the lipopolysaccharides (LPSs), whose active
role is evident in the later stages of infection, allowing for
example the penetration of the infection thread into the
cortical cells; and (4) located directly around the membrane,
the capsular polysaccharides (CPSs), whose role remains to
be clarified.
This work focuses on the CPS (or K-antigen polysaccharides, or KPS) of S. meliloti strain 1021. They have initially
been structurally characterized in S. meliloti and S. fredii
(Reuhs et al., 1993) and are analogous to group II KPSs
found in Escherichia coli, so that the K-antigen appellation,
or KPS, is mostly used for this class of polysaccharides. To
date, all K-antigens found in Rhizobia exhibit a dimeric
repeating unit of one hexose (that can be diversely substituted) and a 3-deoxy-D-manno-oct-2-ulopyranosonic acid
(Kdo) or a related sugar (Fraysse et al., 2003). The biological importance of rhizobial CPS (or KPS) has been investigated in detail in the symbiotic couple S. meliloti/alfalfa
(Pellock et al., 2000).
The nature and the size of KPS are variable within the
S. meliloti genus from strain to strain (Rheus et al., 1998).
Rm 41 exhibits two types of KPS, a low- (LMM) and a
high-molecular-mass (HMM) form. In the corresponding
exoB mutant, AK631, the HMM KPS can replace the deficiency of EPS and lead to a time-delayed normal nitrogenfixing phenotype. A mutation of the rkpZ gene involved in
the biosynthesis of KPS (more precisely on its molecular
size) results in a non-nitrogen fixating (Fix) phenotype
(Putnoky et al., 1990; Williams et al., 1990). This gene
enables the K-antigen to increase to an average molecular
mass of 20–26 kDa, which is necessary to render it symbiotically efficient. S. meliloti strain 1021 does not possess such
a gene, and an exoB mutation, resulting in a lack of EPS,
leads to a Fix phenotype. Indeed, when only a LMM KPS
is synthesized, the EPS deficiency is not overcome.
The nitrogen-fixation phenotype is restored by a complementation with rkpZ (Reuhs et al., 1995). Nevertheless,
Glycobiology vol. 15 no. 1 # Oxford University Press 2005; all rights reserved.
101
N. Fraysse et al.
this restoration is not complete and only the introduction of
the complete pSymB megaplasmid results in a total restoration of a nitrogen-fixing phenotype. This means that the
molecular size of KPS is a determinant of biological activity, but not the only one. Some other structural features are
still probably harbored or lacking in the KPS of the RkpZcomplemented S. meliloti 1021 but also in the 1021 wild
type, and this is responsible for a noncompletely restored
nitrogen-fixing phenotype.
The goal of this investigation was to characterize the
structure of the KPS produced by S. meliloti 1021, which,
unlike to those of S. meliloti strain 41 and several related
strains (Reuhs et al., 1998), has not been determined. For
the first time, it is reported that S. meliloti strain 1021 is able
to synthesize a KPS that is unique with regard to its structure (purely constituted of Kdo) and to the presence of a
phospholipid anchor. The presence in any Rhizobium strain
of this phospholipid anchor, even if it was already demonstrated for other Gram-negative bacteria, such as E. coli
(Fischer et al., 1982; Gotschlich et al., 1981; Schmidt and
Jann, 1984), has been discussed based on genetic
approaches but without being structurally demonstrated.
Results
PAGE analysis of crude polysaccharide extracts of
S. meliloti 1021
The crude extract containing the polysaccharides of
S. meliloti 1021 was first analyzed by deoxycholic acid
(DOC) polyacrylamide gel electrophoresis (PAGE) (see
Figure 1). Using silver nitrate staining, it showed the usual
molecular pattern for this strain with a major band corresponding to the rough LPS (rLPSII), and three other
higher-molecular-mass LPSs that were assumed to be
O-antigen containing (sLPSI, II, and III). Small amounts
of another lower-molecular-mass rough LPS (rLPSI) were
identified, probably corresponding to a form that lacks
the core region completely or parts thereof. When alcian
Fig. 1. DOC PAGE analysis of the Sinorhizobium meliloti 1021
polysaccharides. (A) Silver staining specific for the presence of LPS.
Five different LPSs are present. Two rough LPS (LPS I and II); the LPS I
is probably a form of LPS II that misses the outer core moiety. Three
other LPSs at higher molecular weight can be assumed to be O-antigenic
LPS. (B) Alcian blue-silver staining revealing all acidic polysaccharides.
The supplemental presence of KPS can be observed harboring a typical
veil for this compound.
102
blue silver staining, reporting the presence of acidic
polysaccharides, was used, a large gray-colored smear was
present over the rough LPS zone and further down. This
behavior is typical for the presence of a CPS, even if a ladder
pattern is not observed.
Sugar analyses
The determination of uronic acids and Kdo revealed contents of 15% (2%) and 20% (2%) of the extract dry mass,
respectively. Assays were performed in weak acidic conditions (0.1% acetic acid, 1 h), thus the Kdo residue in the main
chain is not released. The hexose content was 45% (2%).
MS analyses of the crude polysaccharide extract
Matrix-assisted laser desorption/ionization time-of-flight
(MALDI-TOF) mass spectrometry (MS) of the crude polysaccharide extract revealed a complex mixture of a series of
poorly resolved ion peaks with mass differences of 220 amu,
indicating oligosaccharides carrying different numbers of
Kdo (data not shown). Furthermore, the spectrum is
made up of ion peaks with slightly higher abundance
around m/z 2030, 1463, and 3979. The first two ions are
consistent with the lipid A structures described by Kanipes
et al. (2003), and the third may be a complete rLPS II
carrying the complete core oligosaccharide. However, due
to presence of numerous salt adducts, which could not be
reduced significantly by cation exchanger, the mass spectra
did not allow a detailed interpretation.
High-resolution Fourier transform ion cyclotron resonance (FT-ICR) MS using electrospray ionization (ESI)
was applied. The negative-ion mass spectrum of the crude
extract revealed a complex mixture of ions, each one present
in multiple charge states. To facilitate the interpretation, the
mass spectrum was charge-deconvoluted, and mass numbers given refer to the monoisotopic peaks of the neutral
molecule. Consistent with molecular species characterized
by MALDI-TOF MS, a series of compounds in the range of
2800 to 5100 amu with mass differences of 220.06 amu (one
Kdo residue) were observed. However, no ions corresponding to rLPS II and the lipid A moiety could be identified.
This might be due to the lower solubility of the highly
hydrophobic LPS in the electrospray solution, leading to a
reduced ionization efficiency. As is demonstrated by the
enlargement of the most abundant group of ions around
3926 amu (see Figure 2), each group of ions consists of two
molecular species, differing by 28 amu. The other ions
within in each group originate from multiple cationization
with a combination of sodium and potassium adducts. The
mass difference of 28 amu might point to heterogeneity with
respect to the chain length of a fatty acid residue and thus
may indicate that the Kdo polysaccharide was covalently
linked to a lipid anchor.
Capillary skimmer dissociation (CSD) was used to further
elucidate this hypothesis (see Figure 3). The negative-ion
mass spectrum and the enlargements given in Figures 3B
and 3C show different series of polymeric fragment ions.
The most abundant series (B fragments) consists of Kdo
homooligomeres (m/z 439.11, 659.17, 879.23, 1099.29,
1319.36, 1559.42) with two to seven Kdo residues accompanied by ions with mass differences of 18 and 44 amu.
S. meliloti and KPS
Fig. 2. ESI FT-ICR MS of the polysaccharides produced by S. meliloti 1021. The ionization parameters have been optimized to allow the apparition of
higher multicharged masses. (A) Primary mass spectrum. Series of four times multicharged ions are detected. (B) Corresponding deconvoluted mass
spectrum that presents the equivalent molecular species and not the monocharged ions. Two series of Kdo polymers are showed here, where starting
points are the 622.49 amu phospholipid (PL) or its related 594.46 amu compound. (C) Zoom focused on the 15-mere version (15 Kdo) of this Kdo
polysaccharide. M1 is the version whose starting point is the 594.46 amu phospholipid, M2 is the one with 622.49 amu. The low-energy ionization levels
used here appear as equivalent Na and K adducts, making the mass spectra more complicated but still interpretable.
They are induced by successive loss of H2O and/or decarboxylation of Kdo, respectively (Schwudke et al., 2003).
Two further fragment ion series differing by the number
of Kdo residues are marked by asterics. These series exhibited a mass differences of 28 amu and did not show the
satellite ions due to decarboxylation of the B fragments.
Thus, these series are interpreted as Y fragments carrying
the lipid anchor. The smallest fragment ions of this series at
m/z 622.48 and 594.46 should represent the pure lipid
anchor. Further evidence to support this interpretation is
the fact that infrared multiphoton dissociation (IRMPD)
tandem MS (MS/MS) of the Y fragment ions resulted in the
loss of at least one Kdo residues, except for the smallest ions
at m/z 622.48 and 594.46. With respect to these two compounds, the MS analysis has been acquired with enough
mass accuracy to measure the ions’ exact masses. In the case
of the ion at 622.481 amu, the deduced crude formula is
[C33O7NH69P]; the formula is [C31O7NH65P] for the ion
at 594.465. Calculation of the monoisotopic peak of theses
lipid anchors linked to a homopolymer of 15 Kdo (e.g.,
M2 ¼ 622.48 þ Hþ þ 15 Kdo ¼ 3924.362 amu) were in
agreement with the measured masses of the most abundant
molecule shown in Figure 2B.
TLC and MS/MS analyses of the phospholipid anchor
To clarify the structure of the phospholipid anchor, 10 mg
of the crude extracts were submitted to mild acid hydrolysis
in 1% acetic acid for 1 h and were then centrifuged at
5000 g. The precipitate was submitted to semi-preparative
thin-layer chromatography (TLC, see Figure 4). The two
differently stained thin-layer chromatograms revealed the
presence of lipid A (blue stained by the anthrone) and of a
compound with Rf 0.5 that did not appear to bear a sugar
moiety because it could only be stained with sulfuric acid.
After excision of this compound and extraction with
MeOH/CHCl3 (1:1), it was submitted to ESI–quadrupole
time of flight (QTOF) MS analysis. The mass spectrum
clearly identified the same phospholipid anchor
103
N. Fraysse et al.
Fig. 3. CSD ESI FT-ICR MS of the polysaccharides produced by S. meliloti 1021. The ionization parameters have been optimized to produce in-source
collision dissociation. Three series of ions are observed with starting points at 439.11 amu, corresponding to two Kdo, and the phospholipids
594.46 amu (formula C39O14NH78P) and 622.49 amu (C41O14NH82P). Notice the mass spaces of 220.06 amu between the different ions, corresponding
to a Kdo. The same fragmentations of each ion are systematically observed: successive deshydratations (18.011 amu) and/or decarboxylation
(43.990 amu), typic for acidic sugars.
(622.480 amu) that was observed in CSD FT-ICR MS of the
crude extract. Low-energy collision-induced dissociation
MS/MS of this ion (see Figure 5a) confirmed the phospholipid character by the fragment ion at m/z 152.997 corresponding to a glycerolphosphate (usually observed in
negative ESI MS/MS of phospholipids, Pulfer and Murphy,
2003) and the loss of glycerol (–74 and 92 amu) from the
parent ion. The fragment ion at m/z 439.420 could be a
cleaved ester-linked fatty acid.
NMR spectroscopy of the crude polysaccharides extracts
The 1H nuclear magnetic resonance (NMR) spectrum of the
crude extract (see Table I) contained two signals characteristic for Kdo methylene protons at d 1.774 (H3ax) and d
2.407 (H3eq), one anomeric proton of an aldose at d 4.895,
as well as broad signals observed between d 0.7 and d 1.5,
which are characteristic for acyl resonances of the LPS. The
difference d 4 0.5 between the chemical shifts of H3ax and
H3eq indicates that Kdo is in the b-pyranosyl configuration
(Unger, 1981). All 1H and 13C chemical shifts of the Kdo
104
homopolysaccharide (Table I) were established from correlation spectroscopy (COSY), total COSY, and heteronuclear multiple-quantum coherence spectra. The low-field
shifted signal of C7 of Kdo (d 71.68) demonstrated its
substitution at O7. The anomeric proton at d 4.895
belonged to the previously described homopolysaccharide
(Breedveld and Miller, 1994) ! 2)-b-D-Glcp-(1 ! (Table I).
The substitution at O7 of b-Kdo and at O2 of b-D-Glcp
were also confirmed by methylation analysis, in which 1,2,
6,7-tetra-O-acetyl-4,5,8-tri-O-methyl-octitol and 1,2,5-triO-acetyl-3,4,6-tri-O-methyl-glucitol were identified.
ESI-MS analysis of the crude polysaccharides extract
from S. meliloti strain 1021 cultivated in Vincent
minimal liquid medium
ESI-QTOF MS analysis was also performed on the crude
polysaccharides extract obtained from S. meliloti strain
1021 cultivated in Vincent minimal liquid medium. The
same Kdo homopolymer was identified (see Figure 6), possessing similar average molecular mass and polydispersity.
S. meliloti and KPS
The exact masses corresponded to the same phospholipid
anchor (ion at 622.480 amu) substituted with a certain
number of Kdo.
Discussion
In this investigation, a homopolymer composed of Kdo has
been demonstrated to be produced by S. meliloti strain 1021.
Fig. 4. Semi-preparative TLC analysis of mild acid hydrolysates from
S. meliloti 1021 crude extracts. At Rf 0.55 is observed the presence of a
nonosodic compound that is not blue colored with anthrone. This
compound has been extracted from the silica and further analyzed in
MS analysis, which showed that it correspond to the phospholipid
anchor of the Kdo polysaccharide.
The presence of Kdo in the crude extracts of S. meliloti
strain 1021 did not originate exclusively from the LPS.
Colorimetric assays showed a high content of Kdo. The
conditions used in mild acid hydrolysis were too weak to
release the main chain Kdo residue. The release of the
branching Kdo residue does not yield high quantities. Consequently, most of the Kdo in the crude polysaccharide
extract was assigned to belong to a CPS that is observed
in DOC PAGE. NMR spectroscopy and MS analyses confirmed this.
The average molecular mass of this Kdo polysaccharide
was assumed to be very close to that of the rough LPS,
the presence of which made clear observation of the
polysaccharide in DOC PAGE more complicated.
ESI-MS experiments confirmed the average molecular
mass of ~ 4–4.5 kDa.
The homopolymeric nature of this Kdo polysaccharide
was characterized on the basis of the exact masses determined by ESI FT-ICR MS and by CSD. NMR analyses
showed that the Kdo was in b-pyranosyl configuration as
it was demonstrated for S. meliloti strain AK631 (Reuhs
et al., 1993). For the first time, a lipid anchor was identified in a K-antigen polysaccharide of S. meliloti based
on the identification of two ions that corresponded to
phospholipid moieties. These two compounds possessed
a mass difference of 28.032 amu that corresponded to a
CH2-CH2 moiety. For these compounds, we propose the
formulas [C33O7NH69P] for the one with 622.480 amu
and [C31O7NH65P] for that with 694.474 amu. The heterogeneity at the level of the phospholipid anchors gave
rise to the two different series of polysaccharides. Thus
the K-antigen polysaccharides should be anchored in the
outer membrane by a phospholipid moiety. The detailed
structures of the phospholipid anchor species have not
been unequivocally identified, but the MS/MS experiments suggested the presence of a phosphoglycerol
Fig. 5. Exact mass MS and MS/MS analysis of the phospholipid anchor. (a) This residue (622 amu) fragments to give 152.99 amu typical for the
phosphoglycerol moiety. The fragment ion observed at 439.45 could reveal the presence of a hydroxy-octacosanoic acid typical for Rhizobium. (b) The
crude formula C33O7H69NP is given by exact mass 622.481, the isotopic abundancy ratio between 623.481 and 622.481: 33.5 þ 1%, and the parity
typical for the presence of one, three, or five nitrogens.
105
N. Fraysse et al.
Table I. 1H and 13C NMR chemical shifts for two homopolysaccharides isolated from S. meliloti 1021
Sugar residue
Chemical shifts 1H and
C (ppm)
H1
H2
H3
C1
C2
C3
H3eq
ax
! 7Þ-b-D-Kdop
1.774
nd
! 2Þ-b-D-Glcp
13
4.895
102.4
nd
3.585
82.78
35.22
3.780
76.17
eq
2.407
H4
H5
H6a
C4
C5
C6
3.704
68.34
3.476
69.49
Fig. 6. ESI FT-ICR MS of the polysaccharides produced by S. meliloti
1021 in liquid Vincent minimal medium. Here are represented the
molecular monocharged ions. The same masses are detected as the crude
extract produced in TY plate culture conditions, confirming that the
production of this Kdo polysaccharide does not depend on the culture
conditions.
moiety and a C28:OH fatty acid that is typical for
rhizobia species.
The presence of a lipid anchor has been discussed for
some time in investigations dealing with rhizobial KPS
structures. The rhizobial K-antigen polysaccharides were
first described to be analogous to group II KPSs found in
E. coli (Reuhs et al., 1993). This similarity is due not only to
high contents of Kdo but also to the presence of a phospholipid anchor. The phospholipid anchors of rhizobial
KPS have not been fully characterized. Their purification
and characterization is difficult for many reasons, for example, fragility of the structure in acid hydrolysis or the strong
physicochemical affinity with amphiphilic LPS that hampers the isolation. Also, direct MS analysis is difficult to
pursue. However, cyclotron resonance or high-resolution
QTOF mass spectrometers allowed the accurate determination of the complex polysaccharide structures in the mixture. Another difficulty in the structural characterization
was the presence of a phospholipid anchor that renders the
KPS amphiphilic. Reuhs et al. (1998) reported the purification and the structural characterization of rhizobial KPS.
They all consisted of simple dimeric repeating units, HexKdx sugars, without a lipid anchor. It is possible that the
observed nonpolymyxin bound KPS resulted from hydrolyzed saccharidic products whose anchor has been lost. This
highly water-soluble hydrolysate by-product from KPS
yielded much better immediate mass spectra than the entire
106
3.959
65.48
3.504
77.02
H6b
3.359
H8a
C7
C8
4.399
72.91
3.760
H7
71.68
3.607
H8b
3.979
62.52
3.940
61.36
KPS, which required a complex solvent system to obtain
good MS data. The difficulties to detect and structurally
characterize the lipid anchor might have also been due to
noncovalent interactions with LPS.
The amphiphilic properties and the noncovalent interactions of KPS with the LPS suggest that this polysaccharide
is bound to the outer membrane and does not represent an
intracellular compound or a loosely attached CPS.
It has been previously shown that the absence of the rkpZ
gene in S. meliloti strain AK631 led to a change in the
molecular mass of the KPS, that is, to a decrease from
~ 22 kDa to ~ 6 kDa. The high size range mass was also
expected for KPS of S. meliloti strain 1021 but was not
identified, even if the presence of three other RkpZ homologs in the S. meliloti 1021 genome has been recently
demonstrated (Sharypova et al., unpublished data). However, whether these were expressed remains to be determined. The KPS contained only Kdo, which could explain
the fact that even when a RkpZ complementation was
carried out, the usual phenotype made up of a hexoseKdo disaccharide repeating unit was not completely
restored, leading to a reduced nitrogen-fixation activity.
On the other hand, genetic studies on S. meliloti suggested
that certain genes involved in KPS biosynthesis correspond
to genes involved in the constitution of a KPS lipid anchor
(Kiss et al., 1997, 2001; Petrovics et al., 1993). RkpG (sharing similarity with acyltransferases), RkpH (homologous to
short-chain alcohol deydrogenase) proteins, and more generally fatty acid synthase genes from the fix23 region, are
required for KPS biosynthesis and transport. Reuhs et al.
(1995) proposed a model for KPS expression in rhizobia: A
short oligosaccharide (8–15 units) is synthesized, and such
subunits are polymerized to HMM KPS, which is finally
exported. The membrane-attached LMM oligosaccharides
that were observed in our work may represent such KPS
subunits. In the present case of S. meliloti 1021, some genes
are missing (like RkpZ ) and thus the complete HMM KPS
could no longer be polymerized and transported through
the membrane.
The fact that this LMM Kdo-rich polysaccharide was as
well produced under conditions that favor the expression
of nod genes suggests that expression of the genes involved
in its biosynthesis are probably not dependent on environmental conditions. The structural characterization of LMM
Kdo-rich polysaccharides (that can be assumed as LMM
CPSs) from S.meliloti 1021 shows that the presence of a
S. meliloti and KPS
KPS phospholipid anchor in other Rhizobium species is
possible. Consequently, using similar techniques to those
described in this article, the characterized CPSs from other
species should be reexamined.
Material and methods
Bacterial strains and cultures
When cultured on tryptone-yeast extract (TY) plates,
S. meliloti strain 1021 was obtained from the strain collection of the Department of Genetics, University of Bielefeld,
Germany. Bacteria were grown on 2000 TY plates at 28 C.
S. meliloti strain 1021 was obtained from the collection of
INRA, Auzeville, France, and cultured in liquid Vincent
minimal medium complemented with the nod gene inducer
luteoline (Fraysse et al., 2003). The cultures were grown at
36 C. Growth was stopped by the addition of sodium azide
when the optical density observed at 600 nm reached 1.5.
Extraction and purification of KPS
Bacteria were harvested by washing each plate with 0.9%
NaCl (~2 ml) and centrifugation at 10,000 g, 4 C, 20 min.
The bacteria were recovered from liquid medium using the
same centrifugation protocol. Both bacterial masses were
extracted as described. Pellets were resuspended twice in
0.9% NaCl (500 ml) and centrifuged as described. The pellets were extracted following the hot-phenol water extraction protocol (Westphal and Jann, 1965). The water phase
(~700 ml) was reduced by evaporation to 200 ml, dialyzed
three times against water (12,000 kDa molecular weight cutoff ) to eliminate traces of phenol, and then digested by
RNase, DNase, and Proteinase K (Carlson et al., 1978). Samples were dialyzed and then freeze-dried. This material was
dissolved in 20 ml water and ultracentrifuged at 100,000 g
for 18 h. Pellets were recuperated and freeze-dried.
PAGE of polysaccharides
DOC PAGE was performed as previously described (Reuhs
et al., 1998) and stained with either alcian blue-silver (Corzo
et al., 1991) or silver nitrate (Tsai and Frasch, 1982).
Semi-preparative TLC analysis and microextraction
TLC analysis was performed on 0.5 mm silica gel 60 plates
with CHCl3:MeOH (2:1, v/v) as eluent. For staining, 15%
sulfuric acid in ethanol was used. For the identification of
sugars, 15% sulfuric acid in ethanol containing 0.5%
anthrone was applied. Bands of interest were scrapped off
and the silica (about 1–5 mg) was eluted with 10 ml
CHCl3:MeOH (2:1, v/v). The solution was evaporated to
dryness and the residue dissolved in 500 ml CHCl3:MeOH
(2:1, v/v) for direct ESI-MS analysis.
ESI FT-ICR MS
ESI FT-ICR MS was performed in the negative ion mode
using an APEX II Instrument (Bruker Daltonics, Billerica,
MA) equipped with a 7 Tesla actively shielded superconducting magnet and an Apollo ion source. Mass spectra
were acquired using standard experimental sequences as
provided by the manufacturer. Samples were dissolved at
a concentration of ~ 10 ng/ml in a 50:50:0.001 (v/v/v) mixture
of 2-propanol, water, and triethylamine and sprayed at a
flow rate of 2 ml/min. Capillary entrance voltage was set to
3.8 kV and dry gas temperature to 150 C.
The spectra, which showed several charge states for each
component, were charge-deconvoluted. The mass numbers
given refer to the monoisotopic molecular masses.
CSD was induced by increasing the capillary exit voltage
from 100 V to 350 V.
ESI QTOF MS
ESI QTOF MS was performed in the negative and positive
ion mode using a QTOF Ultima Instrument (Waters,
Milford, CT). Samples were dissolved at a concentration of
about 10 ng/ml in a 50:50:0.001 (v/v/v) mixture of 2propanol, water, and triethylamine and sprayed at a flow
rate of 10 ml/min. Capillary entrance voltage was set to 3.0 kV,
and dry gas temperature to 120 C, Cone: 100 V, Rf lens: 80 V,
MS profile [150(10%), 900(80%), ramp 10%]. The collisioninduced dissociation MS/MS experiment was performed
selecting the precursor ion with the first quadrupole. Argon
was used as the collision gas at a pressure of 3.5 105 mbar.
MALDI-TOF MS
Crude extracts were analyzed on a Bruker Reflex II
(Bruker-Daltonics) in the negative-ion mode in the linear
and in the reflector TOF. Configuration using gentisic acid
matrix. For sample preparation, 1 ml saturated gentisic acid
solution was mixed with 1 ml of ~ 50 mg ml1 crude extract
lyophilisate. One microliter of this mixture was dropped
onto the MALDI target.
NMR spectroscopy
NMR spectra were obtained with a Bruker DRX Avance
600 MHz spectrometer using standard Bruker software. All
recordings were made at 305 K on crude polysaccharide
extracts in D2O (8 mg ml1) after three exchanges with
D2O. Chemical shifts were reported relative to internal
acetone (dH 2.225; dC 31.45). Two-dimensional 1H-1H
COSY was recorded with data sets (t1 t2) of 512 2048
points. Two-dimensional total COSY was performed in a
phase-sensitive manner using a data set of 512 2048 points
and a mixing time of 100 ms. Heteronuclear 2D 1H-13C
correlation was recorded in the 1H detection mode via multiple-quantum coherence by use of data sets of 2048 256.
For homo- and heterocorrelation experiments, 64 and 128
scans were acquired for each t1 value, respectively.
Methylation analysis
The methylation analysis was carried out using method
of Ciucanu and Kerek (1984). The chloroform extract
was hydrolyzed, reduced, and acetylated. The partially
methylated alditol acetates were analyzed by gas
chromatography–MS.
Acknowledgments
This work was supported by a grant from the network
‘‘Oligosaccharide signalling in plants’’ of the European
107
N. Fraysse et al.
community (grant no. HPRN-CT-2002-00251). The
authors are grateful to Dr. Anke Becker for helpful
discussions.
Abbreviations
COSY, correlation spectroscopy; CPS, capsular polysaccharide; CSD, capillary skimmer dissociation; DOC,
deoxycholic acid; EPS, exopolysaccharide; ESI, electrospray ionization; FT-ICR, Fourier transform ion cyclotron
resonance; HMM, high molecular mass; KPS, K-antigen
capsular polysaccharide; LMM, low molecular mass; LPS,
lipopolysaccharide; MALDI-TOF, matrix-assisted laser
desorption ionization time of flight; MS, mass spectrometry; MS/MS, tandem mass spectrometry; NMR, nuclear
magnetic resonance; PAGE, polyacrylamide gel electrophoresis; QTOF, quadrupole time-of-flight; TLC, thinlayer chromatography; TY, tryptone-yeast extract.
References
Breedveld, M.G. and Miller, K.J. (1994) Cyclic b-glucan of members of
the family Rhizobiaceae. Microbiol. Rev., 58, 145–161.
Carlson, R.W., Sanders, R.E., Napoli, C., and Albersheim, P. (1978) Host
symbiont interaction III. Purification and partial characterisation of
Rhizobium lipopolysaccharides. Plant Physiol., 62, 912–917.
Carlson, R.W., Reuhs, B.L., Forsberg, L.S., and Kannenberg E.L. (1999)
Rhizobial cell surface carbohydrates: their structure, biosynthesis and
functions. In Goldberg, J.B. (Ed.), Genetics of bacterial polysaccharides. CRC Press, Boca Raton, Fl.
Ciucanu, I. and Kerek, F. (1984) A simple and rapid method for the
permethylation of carbohydrates. Carbohydr. Res., 131, 209–217.
Corzo, J., Perez-Galdona, R., Leon-Barrios, M., and GutierrezNavarro, A.M. (1991) Alcian blue fixation allows silver staining of
the isolated polysaccharide component of bacterial lipopolysaccharides in polyacrylamide gels. Electrophoresis, 12(6), 439–441.
Fischer, W., Schmidt, M.A., Jan, B., and Jan, K. (1982) Structure of the
Escherichia coli K2 capsular antigen. Stereochemical configuration of
the glycerophosphate and distribution of the galactopyranosyl and
galactofuranosyl residues. Biochemistry, 21(6), 1279–1281.
Fraysse, N., Couderc, F., and Poinsot, V. (2003). Surface polysaccharide
involvement in establishing the rhizobium-legume symbiosis. Eur. J.
Biochem., 270, 1365–1380.
Gotschlich, E.C., Fraser, B.A., Nishimura, O., Robbins, J.B., and
Liu, T.Y. (1981) Lipid on capsular polysaccharides of gram negative
bacteria. J. Biol. Chem., 256, 8915–8921.
Kanipes, M.I., Kalb, S.R., Cotter, R.J., Hozbor, D.F., Lagares, A., and
Raetz, C.R. (2003). Relaxed sugar donor selectivity of a Sinorhizobium
meliloti ortholog of the Rhizobium leguminosarum mannosyl transferase LpcC. Role of the lipopolysaccharide core in symbiosis of
Rhizobiaceae with plants. J. Biol. Chem., 278, 16365–16371.
Kiss, E., Reuhs, B.L., Kim, J.S., Kereszt, A., Petrovics, G., Putnoky, P.,
Dusha, I., Carlson, R.W., and Kondorosi, A. (1997) The rkpGHI
and -J genes are involved in capsular polysaccharide production by
Rhizobium meliloti. J. Bacteriol., 179(7), 2132–2140.
Kiss, E., Kereszt, A., Barta, F., Stephens, S., Reuhs, B.L., Kondorosi, A.,
and Putnoky, P. (2001) The rkp-3 gene region of Sinorhizobium meliloti
Rm41 contains strain-specific genes that determine K antigen
structure. Mol. Plant. Microbe Interact., 14(12), 1395–1403.
108
Noel, K.D. and Duelli, D.M. (2000) Rhizobium polysaccharide and its role
in symbiosis. In Triplett, E.W. (Ed.), Prokariotic nitrogen fixation: a
model system for analyse of biological process. Horizon Scientific Press,
Wymondham, UK, pp 415–431.
Pellock, B.J., Cheng, H.P., and Walker, G.C. (2000) Alfalfa root nodule
invasion efficiency is dependent on Sinorhizobium meliloti polysaccharides. J. Bacteriol., 182, 4310–4318.
Petrovics, G., Putnoky, P., Reuhs, B.L., Kim, J., Thorp, T.A., Noel, K.D.,
Carlson, R.W., and Kondorosi, A. (1993) The presence of a novel type
of surface polysaccharide in Rhizobium meliloti requires a new fatty
acid synthase like gene cluster involved in symbiotic nodule development. Mol. Microbiol., 8(6), 1083–1094.
Pulfer, M. and Murphy, R.C. (2003) Electrospray mass spectrometry of
phospholipids. Mass Spectrom. Rev., 22, 332–364.
Putnoky, P., Petrovics, G., Kereszt, A., Grosskopf, E., Ha, D.T.C.,
Banfalvi, Z., and Kondorosi, A. (1990) Rhizobium meliloti
lipopolysaccharide and exopolysaccharide can have the same
function in the plant-bacterium interaction. J. Bacteriol., 172,
5450–5458.
Reuhs, B.L., Carlson, R.W., and Kim, J.S. (1993) Rhizobium fredii and
Rhizobium meliloti produce 3-deoxy-D-manno-2-octulosonic acidcontaining polysaccharides that are structurally analoguous to group
II K antigens (capsular polysaccharides) found in Escherichia coli.
J. Bact., 175(11), 3570–3580.
Reuhs, B.L., Kim, J.S., Badgett, A., and Carlson R.W. (1994) Production
of cell-associated polysaccharides of Rhizobium fredii USDA205 is
modulated by apigenin and host root extract. Mol. Plant. Microbe
Interact., 7, 240–247.
Reuhs, B.L., Williams, M.N.V., Kim, J.S., Carlson, R.W., and Cote, F.
(1995) Suppression of the Fix 2 phenotype of Rhizobium meliloti
exoB mutants by lpsZ is correlated to a modified expression of the
K-polysaccharide. J. Bacteriol., 177, 4289–4296.
Reuhs, B.L., Geller, D.P., Kim J.S., Fox J.E., Kolli V.S., and
Pueppke S.G. (1998) Sinorhizobium fredii and Sinorhizobium meliloti
produce structurally conserved lipopolysaccharides and strain-specific
K antigens. Appl. Environ. Microbiol., 64, 4930–4938.
Reuhs, B.L., Stephens, S.B., Geller, D.P., Kim, J.S., Glenn, J.,
Przytycki, J., and Ojanen-Reuhs, T. (1999) Epitope identification for
a panel of anti-Sinorizobium meliloti monoclonal antibodies and
application to the analysis of K-antigen and LPS from bacteroids.
Appl. Env. Microbiol., 65, 5186–5191.
Schmidt, M.A. and Jann, K. (1984) Phospholipid substitution of capsular
(K) antigens from Escherichia coli causing extra-intestinal infections
FEMS Microbiol. Lett., 14, 69–74.
Schwudke, D., Linscheid, M., Strauch, E., Appel, B., Z€ahringer, U.,
Moll, H., M€
uller, M., Brecker, L., Gronow, S., and Lindner, B. (2003)
The obligate predatory Bdellovibrio bacteriovorus possesses a neutral
lipid A containing alpha-D-mannoses that replace phosphate
residues: similarities and differences between the lipid As and the
lipopolysaccharides of the wild type strain B. bacteriovorus
HD100 and its host-independent derivative HI100. J. Biol. Chem.,
278, 27502–27512.
Spaink, H.P. (2000) Root nodulation and infection factors produced by
rhizobial bacteria. Annu. Rev. Microbiol., 54, 257–288.
Tsai, C.M. and Frasch, C.E. (1982) A sensitive silver stain for detecting
LPS in polyacrylamide gels. Anal. Biochem., 119, 115–119.
Unger, F.M. (1981) The chemistry and biological significance of 3-deoxyD-manno-2-octulosonic acid (KDO). Adv. Carbohydr. Chem. Biochem.,
38, 323–388.
Westphal, O. and Jann, K. (1965) Bacterial lipopolysaccharides. Methods
Carbohydr. Chem., 5, 83–91.
Williams, M.N.V., Hollingsworth, R.I., Klein, S., and Signer, E.R. (1990)
The symbiotic defect of Rhizobium meliloti exopolysaccharide mutants
is suppressed by lpsZþ, a gene involved in lipopolysaccharide
synthesis. J. Bacteriol., 172, 2622–2632.