Báo cáo khoa học: Identification and structural characterization of a sialylated lacto-N-neotetraose structure in the lipopolysaccharide of Haemophilus influenzae pptx
Tài liệu hạn chế xem trước, để xem đầy đủ mời bạn chọn Tải xuống
1
/ 11 trang
THÔNG TIN TÀI LIỆU
Thông tin cơ bản
Định dạng
Số trang
11
Dung lượng
511,31 KB
Nội dung
Identificationandstructuralcharacterizationofa sialylated
lacto-N-neotetraose structureinthe lipopolysaccharide
of
Haemophilus influenzae
Andrew D. Cox
1
, Derek W. Hood
2
, Adele Martin
1
, Katherine M. Makepeace
2
, Mary E. Deadman
2
,
Jianjun Li
1
, Jean-Robert Brisson
1
, E. Richard Moxon
2
and James C. Richards
1
1
Institute for Biological Sciences, National Research Council, Ottawa, Canada;
2
University of Oxford Department of Paediatrics,
Weatherall Institute for Molecular Medicine, John Radcliffe Hospital, Oxford, UK
A sialylatedlacto-N-neotetraose (Sial-lNnT) structural unit
was identified and structurally characterized inthe lipo-
polysaccharide (LPS) from the genome-sequenced strain
Road (RM118) ofthe human pathogen Haemophilus influ-
enzae grown inthe presence of sialic acid. A combination of
molecular genetics, MS and NMR spectroscopy techniques
showed that this structural unit extended from the proximal
heptose residue ofthe inner core region ofthe LPS molecule.
The structureofthe Sial-lNnT unit was identical to that
found in meningococcal LPS, but glycoforms containing
truncations ofthe Sial-lNnT unit, comprising fewer residues
than the complete oligosaccharide component, were not
detected. The finding ofsialylated glycoforms that were
either fully extended or absent suggests a novel biosynthetic
feature for adding the terminal tetrasaccharide unit of the
Sial-lNnT to the glycose acceptor at the proximal inner core
heptose.
Keywords: Haemophilus influenzae; LPS; mass spectrometry;
NMR; sialic acid.
Haemophilus influenzae remains an important cause of
disease worldwide. Encapsulated strains can cause invasive,
bacteraemic infections such as septicemia and strains
lacking a capsule, so-called nontypeable strains (NTHi),
are a common cause of otitis media and acute lower
respiratory tract infections [1]. Lipopolysaccharide (LPS) is
critical to the integrity and functioning ofthe cell wall of
H. influenzae, and as a surface component is a target for
host immune responses. Thestructureof H. influenzae LPS
has been well established from several studies [2–8].
H. influenzae LPS is composed ofa membrane-anchoring
lipid A moiety linked to a heterogeneous core oligosaccha-
ride via a phosphorylated 2-keto-3-deoxyoctulosonic acid
(Kdo) residue. The core oligosaccharide of H. influenzae
can be divided into inner and outer regions. The inner core
consists of a
L
-glycero-
D
-manno-heptose (Hep) trisaccharide
unit wherein the second heptose residue (Hep II) is substi-
tuted at the 6-position by a phosphoethanolamine (PEtn)
residue. Each ofthe Hep residues can provide a point for the
addition of hexose (Hex) residues each of which can be
further extended into oligosaccharide chains ofthe outer
core. The substitution pattern and degree of extension from
each Hep residue varies between and within strains. The
presence of nonglycose residues, including phosphate-
containing substituents and ester linked acetyl groups and
glycine molecules also contribute to thestructural variability
of these molecules [9]. The LPS of H. influenzae lacks the
O-specific side chain characteristic of enteric bacteria.
Evidence from recent structural studies has confirmed the
presence and defined the position of sialic acid inthe LPS of
H. influenzae [7,10–12]. Sialylation of LPS is a commonly
observed structural modification in Neisseria spp. and is
frequently found as thesialylated lacto-N-neotetraose
structure (Sial-lNnT) [13,14]. Indeed in neisserial species
sialylation of LPS renders the bacteria resistant to comple-
ment-mediated killing by normal human serum [15]. In the
gonococcus, LPS sialylation can occur only if an exogenous
supply of sialic acid is available [16]. In contrast, sialylated
meningococcal LPS glycoforms can be synthesized endo-
genously [17]. Sialylation of H. influenzae LPS appears to
depend upon the presence of an exogenous source of sialic
acid or its activated form CMP-sialic acid [11,18]. Sialyla-
tion of H. influenzae LPS was first indicated by alterations
in the reactivity of LPS with a mAb 3F11 that is specific for
the terminal lactosamine disaccharide ofthe lNnT group
following treatment with neuraminidase [19]. Consistent
with these findings, MS studies pointed to the presence of a
Sial-lNnT group [20]. However, the low amounts of
sialylated glycoforms present precluded definitive structural
characterization ofthesialylated moiety. Ina recent
structural study utilizing an exogenous source of sialic acid
in the growth medium we identified and localized another
sialylated species, namely sialyl-lactose (Sial-Lac) in
H. influenzae LPS [10,11].
The genetic control ofthe biosynthesis of H. influenzae
LPS inner and outer core oligosaccharides has been
investigated extensively. Most ofthe genes responsible for
Correspondence to A. D. Cox, Institute for Biological Sciences,
National Research Council, Ottawa, ON, Canada. K1A 0R6.
Fax: +44 613 952 9092, Tel.: +44 613 991 6172,
E-mail: Andrew.Cox@nrc.ca
Abbreviations: NTHi, nontypeable strains ofHaemophilus influenzae;
LPS, lipopolysaccharide; Kdo, 2-keto-3-deoxyoctulosonic acid;
Hep,
L
-glycero-
D
-manno-heptose; PEtn, phosphoethanolamine;
Hex, hexose; Sial-lNnT, sialylated lacto-N-neotetraose; Sial-Lac,
sialyl-lactose; BHI, brain heart infusion; ES-MS, electrospray;
PCho, phosphocholine.
(Received 21 March 2002, revised 20 June 2002, accepted 2 July 2002)
Eur. J. Biochem. 269, 4009–4019 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03090.x
the biosynthesis ofthe glycose residues inthe globotetraose-
containing Road (RM118) LPS glycoforms have been
recently reported [21]. Moreover, it is also known that
H. influenzae has the ability to modify its LPS structure by a
genetic mechanism known as phase variation [22]. We
recently identified one such phase variable gene lic3A as
encoding an a-2,3-sialyltransferase, which was found to be
responsible for the sialylation ofa lactose group inthe LPS
of H. influenzae strain RM118 [11]. In that study, we
observed that following insertional deletion ofthe lic3A
gene, LPS derived from the mutated strain still produced
sialylated glycoforms, suggesting the presence of alternate
sialylated structures. In this study we describe the identifi-
cation andstructural analysis ofa Sial-lNnT unit in
H. influenzae LPS.
MATERIALS AND METHODS
Bacterial strains and culture conditions
The H. influenzae strain RM118 (Rd) is derived from the
same source as the strain for which the complete genome
has been sequenced [23]. Isogenic strains RM118 lgtC and
the double mutant RM118 lgtC lic3A have been described
previously [11], as have strains RM118 lic2A, RM118 lpsA
and RM118 lgtF [21]. H. influenzae strains were grown
at 37 °C on brain heart infusion (BHI) agar (1% w/v)
supplemented with Levinthals reagent (10% v/v) and
when appropriate, Neu5Ac (25 lgÆmL
)1
), CMP-Neu5Ac
(50 lgÆmL
)1
), kanamycin (10 lgÆmL
)1
) or tetracycline
(4 lgÆmL
)1
).
Structural analysis and purification of LPS
LPS was prepared for structural analysis from cells
harvested after growth on batches of 20 BHI plates,
supplemented with Neu5Ac. LPS was extracted by the hot
phenol/water method [24] and O-deacylated as described
previously [3]. MS analyses were carried out as described
previously [8,25,26]. Electrospray (ES) MS was measured in
the negative ion mode on a VG Quattro triple quadrupole
mass spectrometer (Fisons Instruments) with an electro-
spray ion source.
Capillary electrophoresis (CE)-MS analysis was per-
formed on a crystal Model 310 CE instrument (ATI
Unicam, Boston, MA, USA) coupled to an API 3000 mass
spectrometer (MDS/Sciex, Concord, Canada) via a micro-
ionspray interface. A sheath solution (isopropanol/metha-
nol, 2 : 1, v/v) was delivered at a flow rate of 1 lLÆmin
)1
to
a low dead volume tee (250 lm internal diameter; Chro-
matographic Specialities, Brockville, Canada). All aqueous
solutions were filtered through a 0.45-lm filter (Millipore)
before use. An electrospray stainless steel needle (27 gauge)
was butted against the low dead volume tee and enabled the
delivery ofthe sheath solution to the end ofthe capillary
column. Separation was obtained on 90 cm length of bare
fused-silica capillary (192 lmo.d.· 50 lm internal diam-
eter; Polymicro Technologies, Phoenix, AZ, USA) using
30 m
M
morpholine in deionized water (negative ion mode),
pH 9.0, containing 5% methanol and 15 m
M
ammonium
acetate/ammonium hydroxide in deionized water (positive
ion mode), pH 9.0, containing 5% methanol. A voltage of
25 kV was typically applied at the injection. The outlet of
the capillary was tapered to 15 lm internal diameter
using a laser puller (Sutter Instruments, Novato, CA, USA).
Mass spectra were acquired with dwell times of 3.0 ms per
step of 1 m/z unit in full-mass scan mode. Inthe CE-ESMS,
30 nL sample was typically injected by using 300 mbar for a
duration of 0.1 min. For CE-ESMS/MS experiments
60 nL sample was introduced using 300 mbar for
0.2min.TheMS/MSdatawereacquiredwithdwelltimes
of 2.0 ms per step of 1 m/z unit. Fragment ions formed by
collision activation of selected precursor ions with nitrogen
in the RF-only quadrupole collision cell, were mass
analysed by scanning the third quadrupole.
NMRexperimentswereperformedonVarianINOVA
600 and 500 NMR spectrometers using a 5-mm triple
resonance probe with Z gradient as described previously
[25]. Measurements were made at 25 °C at concentrations of
2mgÆmL
)1
in D
2
O, subsequent to several lyophilizations
with D
2
O. For the proton chemical shift reference, the
HDO resonance was set at 4.78 p.p.m. at 25 °Crelativeto
the methyl resonance of external acetone at 2.225 p.p.m. All
of the NMR data was acquired using Varian sequences
provided with the VNMR 6.1B software. The same
program was used for processing.
The proton spectrum was acquired with a sweep width of
10 p.p.m., 1024 transients, water presaturation during the
relaxation delay of 1.5 s and acquisition time of 1.7 s. It was
processed with a low-pass digital filter for solvent suppres-
sion and zero filling for a final resolution of 0.37 Hz per
point. Homonuclear two-dimensional experiments, COSY
(16 h), TOCSY (7 h) and NOESY (7 h) were acquired using
the following parameters: water presaturation during the
relaxation delay of 1.0 s, spectral width of 7 p.p.m. (COSY;
10 p.p.m.), acquisition time in t
2
of 0.15 s (COSY; 0.20 s)
and 200 increments (COSY; 128) with 52 (TOCSY), 40
(NOESY) or 360 (COSY) scans per increment. The sign
discrimination in F
1
was achieved by the States method. The
mixing time for the NOESY and TOCSY experiments was
0.4 and 0.08 s, respectively. The phase-sensitive spectra were
processedwithforwardlinearpredictionint
1
, unshifted
Gaussian window functions, and zero filling to 1024*1024
complex points for a resolution of 2.9 Hz per point in both
dimensions. The one-dimensional NOESY experiment was
acquired with a sweep width of 16 p.p.m., water presatura-
tion during the relaxation delay of 1 s, acquisition time of 1 s,
a mixing time of 800 ms, selective pulse with 100 Hz
bandwidth, and 2048 transients for a duration of 2 h. Using
similar acquisition parameters, the one-dimensional NO-
ESY-TOCSY experiment [27] was acquired with a NOE
mixing time of 800 ms, spin lock time of 80 ms, and 12288
transients for a duration of 10 h. The one-dimensional
TOCSY experiment was acquired with a sweep width of
16 p.p.m., water presaturation during the relaxation delay of
1 s, acquisition time of 1.7 s, mixing times of 80 ms, selective
pulse with 30 Hz bandwidth, and 5120 transients for a
duration of 4 h. Using similar acquisition parameters, the
one-dimensional TOCSY-NOESY experiment [27] was
acquired with a NOE mixing time of 800 ms, spin lock time
of 80 ms, and 20480 transients for a duration of 21 h.
Electrophoretic analysis of LPS
LPSwaspreparedandanalysedbytricine-SDS/PAGEas
described previously [10].
4010 A. D. Cox et al. (Eur. J. Biochem. 269) Ó FEBS 2002
Analytical methods and methylation analysis
Sugars were determined as their alditol acetates and
partially methylated alditol acetate derivatives by GLC-
MS as described previously [8].
RESULTS
In a previous study on LPS from H. influenzae strain
RM118 lgtC we had identified asialylated lactosyl
(Sial-Lac) extension from the distal heptose residue
(Hep III) ofthe inner core [11]. Interestingly, following
mutation ofthe sialyltransferase gene (lic3A) responsible for
addition of Neu5Ac to the lactose moiety in this strain,
creating the double mutant (lgtC lic3A), sialylated glyco-
forms were still observed. In SDS/PAGE analysis sialylated
species of strain RM118 lgtC lic3A LPS were found as the
LPS constituents with the lowest mobility, identified by
virtue of an alteration in their migration pattern following
treatment with neuraminidase, an enzyme that specifi-
cally cleaves sialic acid residues (Fig. 1; lanes 1 and 2).
H. influenzae LPS typically migrates as a complex series of
bands in SDS/PAGE, each band corresponding to gain or
loss ofa glycose moiety due to the phase variable nature of
H. influenzae LPS synthesis [3,21,22]. In LPS from RM118
lgtC lic3A it was interesting to note that the band
corresponding to thesialylated LPS glycoform separated
from those of fast migrating nonsialylated glycoforms by
several glycose units. ES-MS analysis ofthe O-deacylated
LPS from the RM118 lgtC lic3A double-mutant strain
revealed ions with an expected composition consistent with
the presence ofa sialic acid residue (Fig. 2; Table 1). It
has previously been established that sialic acid residues are
not removed by the hydrazinolysis treatment used to
O-deacylate LPS [11]. The molecular mass ofthe major
sialylated glycoforms inthe lgtC lic3A double mutant
(3381.6 and 3546.0 Da) indicated the presence of an
additional 527 atomic mass units, corresponding to an
N-acetyl-hexosamine residue and two hexose residues over
the Sial-Lac glycoform inthe O-deacylated LPS from the
Fig. 1. SDS/PAGE analysis of LPS from strains RM118 lgtC lic3A,
RM118 lgtC, RM118 (wt), RM118 lic2A,RM118lpsA and RM118
lgtF before and after treatment with neuraminidase (as indicated by – and
+, respectively). Thesialylated tetrasaccharide-containing glycoforms
are indicated by an asterisk. All strains were grown on media con-
taining Neu5Ac except for the wild-type (wt) strain RM118.
Fig. 2. Triply charged region ofthe ES-MS spectrum from the O-deacylated LPS obtained from strain RM118 lgtC lic3A grown on BHI plates
supplemented with Neu5Ac. Thestructureofthe inner core glycoforms is shown and ions arising from the inner core and from the addition of the
sialylated tetrasaccharide are indicated.
Ó FEBS 2002 Sialylated LPS inHaemophilusinfluenzae (Eur. J. Biochem. 269) 4011
lgtC single mutant (Table 1) [11]. The two major sialylated
glycoforms observed differed by 165 atomic mass units,
consistent with the gain or loss ofa phosphocholine (PCho)
residue. The PCho group is appended to the glucose residue
at the proximal heptose residue (Hep I) inthe parent strain
[5] and lgtC mutant [21]. CE-MS analysis was carried out on
Table 1. Negative ion ES-MS data and proposed compositions of O-deacylated LPS from H. infl uenza e strains RM118, RM118 lgtC,RM118
lgtC lic3A, RM118 lic2A and RM118 lpsA when grown on media containing sialic acid. Average mass units were used for calculation of molecular
weight based on proposed composition as follows: lipid A, 953.00; Hex, 162.15; HexNAc, 203.19; Hep, 192.17; Kdo-P, 300.16; PEtn, 123.05; PCho,
165.05; Sial, 291.05.
Strain [M-3H]
3–
[M-2H]
2–
Observed
molecular ion
Calculated
molecular ion
Relative
intensity
Proposed
composition
RM118 lgtC 853.2 1280.3 2562.6 2562.2 0.50 3Hex, 3Hep, 2PEtn, Kdo-P, Lipid A
867.2 1300.9 2604.2 2604.2 0.85 PCho, 3Hex, 3Hep, PEtn, Kdo-P, Lipid A
908.3 1362.6 2727.5 2727.3 1.00 PCho, 3Hex, 3Hep, 2PEtn, Kdo-P, Lipid A
964.5 1446.6 2895.8 2895.3 0.35 Sial, PCho, 3Hex, 3Hep, PEtn, Kdo-P,
Lipid A
1005.2 1508.3 3018.6 3018.3 0.70 Sial, PCho, 3Hex, 3Hep, 2PEtn, Kdo-P,
Lipid A
1126.0 – 3381.0 3380.8 0.25 (Sial, 2Hex, HexNAc), 3Hex, 3Hep,
2PEtn, Kdo-P, Lipid A
1181.4 – 3547.2 3545.9 0.20 (Sial, 2Hex, HexNAc), PCho, 3Hex,
3Hep, 2PEtn, Kdo-P, Lipid A
RM118 lgtC 867.3 1301.2 2604.7 2604.2 1.00 PCho, 3Hex, 3Hep, PEtn, Kdo-P, Lipid A
lic3A 908.3 1362.6 2727.5 2727.3 0.90 PCho, 3Hex, 3Hep, 2PEtn, Kdo-P, Lipid A
1126.2 – 3381.6 3380.8 0.20 (Sial, 2Hex, HexNAc), 3Hex, 3Hep,
2PEtn, Kdo-P, Lipid A
1181.0 – 3546.0 3545.9 0.20 (Sial, 2Hex, HexNAc), PCho, 3Hex,
3Hep, 2PEtn, Kdo-P, Lipid A
RM118 812.2 – 2439.6 2439.2 0.15 3Hex, 3Hep, PEtn, Kdo-P, Lipid A
866.9 – 2603.7 2604.2 0.50 PCho, 3Hex, 3Hep, PEtn, Kdo-P, Lipid A
920.9 – 2765.7 2766.3 0.80 PCho, 4Hex, 3Hep, PEtn, Kdo-P, Lipid A
962.0 – 2889.0 2889.3 0.45 PCho, 4Hex, 3Hep, 2PEtn, Kdo-P, Lipid A
964.0 – 2895.0 2895.3 1.00 Sial, PCho, 3Hex, 3Hep, PEtn, Kdo-P,
Lipid A
989.0 – 2970.0 2969.5 0.60 PCho, 4Hex, HexNAc, 3Hep, PEtn,
Kdo-P, Lipid A
1005.0 – 3018.0 3018.3 0.70 Sial, PCho, 3Hex, 3Hep, PEtn, Kdo-P,
Lipid A
1030.0 – 3093.0 3092.6 0.25 PCho, 4Hex, HexNAc, 3Hep, 2PEtn,
Kdo-P, Lipid A
1179.7 – 3542.1 3542.9 0.20 (Sial, 2Hex, HexNAc), 4Hex, 3Hep,
2PEtn, Kdo-P, Lipid A
1235.0 – 3708.0 3708.0 0.30 (Sial, 2Hex, HexNAc), PCho, 4Hex,
3Hep, 2PEtn, Kdo-P, Lipid A
RM118 lic2A 758.0 – 2277.0 2277.4 0.40 2Hex, 3Hep, PEtn, Kdo-P, Lipid A
812.9 – 2441.7 2442.4 0.85 PCho, 2Hex, 3Hep, PEtn, Kdo-P, Lipid A
853.9 – 2564.7 2565.5 0.60 PCho, 2Hex, 3Hep, 2PEtn, Kdo-P, Lipid A
1030.9 – 3095.7 3095.6 0.30 (Sial, 2Hex, HexNAc), 2Hex, 3Hep, PEtn,
Kdo-P, Lipid A
1071.9 – 3218.7 3218.7 0.90 (Sial, 2Hex, HexNAc), 2Hex, 3Hep,
2PEtn, Kdo-P, Lipid A
1085.8 – 3260.4 3260.7 0.35 (Sial, 2Hex, HexNAc), PCho, 2Hex,
3Hep, PEtn, Kdo-P, Lipid A
1126.8 – 3383.4 3383.7 0.75 (Sial, 2Hex, HexNAc), PCho, 2Hex,
3Hep, 2PEtn, Kdo-P, Lipid A
RM118 lpsA 703.6 1055.9 2114.1 2115.2 0.45 Hex, 3Hep, PEtn, Kdo-P, Lipid A
744.9 1117.7 2237.3 2238.2 0.25 Hex, 3Hep, 2PEtn, Kdo-P, Lipid A
758.7 1138.8 2279.3 2280.2 1.00 PCho, Hex, 3Hep, PEtn, Kdo-P, Lipid A
799.7 1200.4 2402.8 2403.2 0.60 PCho, Hex, 3Hep, 2PEtn, Kdo-P, Lipid A
1017.6 1526.5 3056.1 3056.4 0.20 (Sial, 2Hex, HexNAc), Hex, 3Hep, 2PEtn,
Kdo-P, Lipid A
1072.6 1609.3 3221.1 3221.5 0.40 (Sial, 2Hex, HexNAc), PCho, Hex, 3Hep,
2PEtn, Kdo-P, Lipid A
4012 A. D. Cox et al. (Eur. J. Biochem. 269) Ó FEBS 2002
the O-deacylated material from the RM118 lgtC lic3A
double mutant in order to obtain further information on the
nature ofthesialylated glycoforms. Comparison ofthe total
ion electropherogram in negative ion mode (Fig. 3A) and
selective ion scanning for m/z 290, specific for sialic acid
(Fig. 3B), suggested the presence of two major populations
of sialylated glycoforms consistent with the ES-MS spec-
trum (Fig. 2). MS/MS analysis in positive ion mode of the
doubly charged ion at m/z 1691.6 (Fig. 3C) that corre-
sponds to the 3381.6 atomic mass unit glycoform produced
abundant fragment ions at m/z 657 (Neu5Ac-Hex-Hex-
NAc); m/z 528 (Hex-HexNAc-Hex); m/z 366 (Hex-Hex-
NAc); m/z 292 (Neu5Ac) and; m/z 204 (HexNAc). This
series of ions indicated the presence ofa Neu5Ac-Hex-
HexNAc-Hex oligosaccharide unit. In order to determine
the location of this tetrasaccharide unit, the wild-type strain
RM118 anda series of mutants thereof (lgtC, lic2A, lpsA)
that corresponded to sequential truncations ofthe globoside
trisaccharide extending from Hep III were examined, as
well as a lgtF mutant in which there is no chain extension
from Hep I [21]. In SDS/PAGE analysis ofthe LPS
obtained from the lgtC, lic2A and lpsA mutants before and
after neuraminidase treatment, the enzyme sensitive sialy-
lated glycoforms showed an increased mobility. Corre-
spondingly, the mobility ofthe major nonsialylated
glycoforms increased as the length of chain from Hep III
was reduced by consecutive glycose truncations (Fig. 1;
lanes 3–4, 7–10), confirming that thesialylated unit was not
attached via the distal heptose residue (Hep III) ofthe inner
core LPS. This was most notable for the lpsA mutant, as the
gene product of lpsA is responsible for initiation of chain
extension from Hep III [21], andin this mutant background
the LPS still elaborates the slower migrating sialylated
glycoform. However, the lgtF mutant LPS (Fig. 1; lanes 11–
12) did not contain slower migrating sialylated glycoforms,
indicating thesialylated unit to be located as an extension
from the Hep I residue ofthe inner core. LgtF has been
shown to be the glycosyltransferase required for initiation of
chain extension from Hep I [21].
As expected ES-MS analysis ofthe O-deacylated LPS of
each mutant strain (Table 1) revealed a reduction in mass of
the sialylated species by a value corresponding to the
removed residue. Thus the difference in mass between the
O-deacylated LPS of RM118 lic2A (3383.4 Da) and
RM118 lpsA (3221.1 Da) corresponds to the glucose residue
(162.3 atomic mass units) that is normally attached by the
lpsA gene product. Inthe spectra of O-deacylated LPS from
the parent strain RM118 and RM118 lgtC, Sial-Lac
glycoforms were observed in addition to the higher mass
sialylated glycoforms. As expected Sial-Lac-containing
glycoforms were not observed inthe lic2A and lpsA
mutants, as the required lactose acceptor is not present. It
is noteworthy that di-sialylated species were not observed
even inthe lgtC mutant, the most appropriate background
for Sial-Lac production [11]. In each ofthe mutant strains
the higher mass sialylated glycoforms had a molecular
weight consistent with the addition ofa complete tetrasac-
charide unit comprising Neu5Ac-Hex-HexNAc-Hex (819
atomic mass units). Closer examination ofthe ES-MS
revealed no evidence for any glycoforms corresponding to
partial addition of this sialylated unit, suggesting that this
tetrasaccharide is added as a complete unit or not at all
during LPS biosynthesis.
The structureofthesialylated glycoforms was determined
by NMR spectroscopy. To simplify the analysis the simplest
structure still containing thesialylated tetrasaccharide was
chosen for NMR studies, namely LPS from the lpsA
mutant. The one-dimensional
1
H-NMR spectrum of the
O-deacylated LPS was well resolved (Fig. 4). Characteristic
signals were observed inthe anomeric region for the H-1
1
H-resonances of a-configured residues at 5.76 p.p.m.
(Hep II), 5.42 p.p.m. (GlcN of O-deacylated lipid A), and
5.16 p.p.m. (Hep I and Hep III). Additionally, signals were
observed inthe anomeric region for the H-1
1
H-resonances
of b-configured residues, including the second GlcN residue
of O-deacylated lipid A (4.63 p.p.m.) andthe glucose
residue (Glc I) at Hep I (4.54 p.p.m.) consistent with
assignments previously obtained for O-deacylated LPS
from the lpsA mutant that had been grown on media
not containing Neu5Ac [21]. Closer examination of the
b-anomeric region revealed two minor resolved signals at
4.74 and 4.45 p.p.m. due to the anomeric protons from
Fig. 3. CE-ES mass spectrum of O-deacylated LPS from strain RM118
lgtC lic3A grown on BHI plates supplemented with Neu5Ac. (A) Solid
line indicates the total ion electropherogram (TIE), the dotted line
indicates single ion monitoring (SIM) for m/z 290
–
, obtained in neg-
ative ion mode with a high orifice voltage (120 V), and (B) a MS/MS
experiment in positive ion mode on m/z 1691.6
2+
that corresponds to
the Sial-lNnT-containing glycoform without the PCho moiety.
Ó FEBS 2002 Sialylated LPS inHaemophilusinfluenzae (Eur. J. Biochem. 269) 4013
residues inthe tetrasaccharide unit ofthesialylated glyco-
form. The signal at 4.74 p.p.m. produced a spin-system in
a TOCSY experiment that could be assigned to N-acetyl-
b-
D
-glucosamine (b-
D
-GlcNAc) by comparison to the
published data (Fig. 5) [13,14]. Similarly (data not shown)
the signal at 4.45 p.p.m. could be assigned to a b-
D
-
galactose spin-system by comparison to the published data
[13,14]. Inthe high-field region ofthe spectrum character-
istic signals were observed for the axial and equatorial H-3
1
H-resonances ofthe sialic acid residue at 1.81 and
2.76 p.p.m., respectively (Fig. 4). Integration ofthe H-3
1
H-resonances ofthe sialic acid residue indicated that the
sialylated glycoform was present at levels of 10% of the
total LPS glycoform population. Higher percentages were
anticipated from the MS analyses, but perhaps the levels of
sialylated species were over-estimated inthe MS studies due
to the presence ofthe sialic acid residue in these glycoforms
that would be more readily ionized inthe negative ion mode.
The
1
H NMR spectrum ofthe inner core LPS structure
was assigned using COSY and TOCSY experiments
(Table 2). The ring sizes and relative stereochemistries of
the component monosaccharides were established from the
1
H chemical shifts andthe magnitude ofthe coupling
constants. The sequence of glycosyl residues ofthe inner
core was determined from interresidue
1
H–
1
HNOEmeas-
urements between anomeric and aglyconic protons on
adjacent glycosyl residues. The NMR data was almost
identical to that found previously [21] for the O-deacylated
LPS ofthe RM118 lpsA mutant that had been grown
on media not containing Neu5Ac. Under these growth
conditions, the RM118 lpsA mutant does not elaborate
detectable amounts of LPS containing sialylated glycoforms
[21]. Thestructureofthe oligosaccharide chain of the
sialylated glycoforms was determined from a series of
selective excitation NMR experiments [27]. Initially the
axial resonance ofthe sialic acid residue at 1.81 p.p.m.
was selectively irradiated ina one-dimensional NOESY
Fig. 4. Anomeric region of the
1
H-NMR spectrum ofthe O-deacylated LPS obtained from strain RM118 lpsA grown on BHI plates supplemented with
Neu5Ac. Inset, complete
1
H-NMR spectrum. The spectrum was recorded in D
2
O at pH 7.0 and 295 K.
Fig. 5. Ring
1
H region ofthe TOCSY spectrum ofthe O-deacylated
LPS from strain RM118 lpsA grown on BHI plates supplemented with
Neu5Ac corresponding to the spin system from the residue at 4.74 p.p.m.
The spectrum was recorded in D
2
O at pH 7.0 and 295 K.
4014 A. D. Cox et al. (Eur. J. Biochem. 269) Ó FEBS 2002
experiment, revealing signals that by comparison to pub-
lished data [13], could be assigned at 2.76 (H-3eq), 3.68
(H-4), 3.86 (H-5) and at 4.05 p.p.m. via an interresidue
NOE. The latter signal was then irradiated ina one-
dimensional TOCSY step following the one-dimensional
NOESY step which revealed a signal at 4.56 p.p.m. that
subsequently was found to correspond to the anomeric
resonance ofthe b-
D
-galactose (Gal II) residue substituted
by sialic acid (Fig. 6A). Selective irradiation ofthe anomeric
resonance ofthe b-
D
-GlcNAc residue at 4.74 p.p.m. in a
one-dimensional TOCSY experiment revealed the ring
1
H-signals at 3.81 (H-2), 3.73 (H-3 and H-4), and
3.59 p.p.m. (H-5). Ina subsequent one-dimensional
NOESY step irradiation ofthe H-3/H-4 signals at
3.73 p.p.m. produced signals at 4.74 and 4.56 p.p.m.
corresponding to the anomeric
1
H-resonances of the
b-
D
-GlcNAc andthe b-
D
-galactose (Gal II) residue substi-
tuting the b-
D
-GlcNAc, respectively (Fig. 6B). By compar-
ison to the published data [13,14,28] it was clear that the
chemical shifts ofthe b-
D
-GlcNAc residue were consistent
with substitution at the 4-position. Thus, these experiments
established the sequence ofthe terminal trisaccharide as
a-Neu5Ac-(2–3)-b-
D
-Gal-(1–4)-b-
D
-GlcNAc. Examination
of the two-dimensional TOCSY and NOESY spectra
revealed characteristic signals for a 3-linked b-
D
-galactose
(Gal I) residue at 3.60 (H-2), 3.73 (H-3), and 4.17 p.p.m.
(H-4) inthe spin-system arising from the anomeric
resonance at 4.45 p.p.m. [14]. This inference was confirmed
by a TOCSY/NOESY selective excitation experiment,
where following selective irradiation ofthe anomeric proton
at 4.45 p.p.m. inthe TOCSY step, subsequent irradiation of
the H-3 proton at 3.73 p.p.m. inthe NOESY step revealed
the anomeric proton at 4.74 p.p.m. ofthe b-
D
-GlcNAc
residue, thus confirming this linkage (data not shown)
and establishing thestructureofthe tetrasaccharide to
be, a-Neu5Ac-(2–3)-b-
D
-Gal-(1–4)-b-
D
-GlcNAc-(1–3)-b-
D
-
Gal. Due to the low intensity ofthesialylated glycoforms
and a high degree of overlap inthe b-anomeric region of the
spectrum, NMR methods were not successful in confirming
the linkage position ofthe glucose residue at Hep I, to
which the terminal tetrasaccharide unit was expected to be
the attached. This is the only glucose residue present in the
LPS ofthe RM118 lpsA mutant. Thus a methylation
analysis was performed in order to determine its linkage
position. Methylation analysis was carried out on the
dephosphorylated O-deacylated LPS as it is known that
phosphorylated residues are not readily identified following
methylation analysis anda PCho residue substitutes the
glucose residue inthe majority of glycoforms. GLC-MS
analysis ofthe partially methylated alditol acetates identified
the expected substitution patterns for the inner core LPS,
with t-Glc, t-Hep, 2-Hep and 3,4-Hep residues identified in
approximately equimolar amounts. Less intense signals
were also identified for 3-Gal and 4-Glc establishing the
glucose residue to be 4-substituted and confirming the
3-linkages for the galactose residues ofthe Sial-lNnT unit
(Fig. 7). This confirms that a terminal Sial-lNnT unit is
linked to the Hep I residue ofthe LPS inner core in the
sialylated glycoform. Thestructureofthesialylated glyco-
forms ofthe O-deacylated LPS of H. influenzae strain
RM118 and mutants causing sequential truncations from
Hep III are illustrated in Fig. 8.
DISCUSSION
Structural analysis of H. influenzae strain RM118 lpsA LPS
has revealed glycoforms containing a Sial-lNnT unit when
the organism is grown on solid medium containing Neu5Ac.
This structure was found to extend from the proximal
heptose residue (Hep I) and is expressed in LPS glycoforms
from the wild-type RM118 strain which also expresses a
globoside unit (a-
D
-Galp-(1–4)-b-
D
-Galp-(1–4)-b-
D
-Glcp-
(1–) from the distal heptose residue (Hep III). Sial-lNnT-
containing glycoforms were also characterized in LPS from
RM118 mutant strains with sequential truncations in the
globoside unit (lgtC, lic2A,andlpsA). Sial-lNnT-containing
glycoforms were not observed however, in LPS from the
RM118 lgtF mutant strain. LgtF is the glucosyltransferase
Table 2.
1
H NMR assignments of O-deacylated LPS from H. influenzae strain RM118 lpsA grown on media containing sialic acid. Assignments at
295 K, relative to HOD 4.78 p.p.m. ND, Not determined.
H-1 H-2 H-3 H-4 H-5 NOEs
Lipid A-OH
a-GlcN 5.42 3.94 ND ND ND –
b-GlcN 4.63 3.82 3.70 ND ND ND
Inner core
Hep I 5.16 4.14 4.07 ND ND 4.31 Kdo H-5
Hep II 5.76 4.27 ND ND ND 4.07 Hep I H-3
Hep III 5.16 4.06 ND ND ND 4.27 Hep II H-2
5.76 Hep II H-1
t-b-Glc (Glc I) 4.54 3.47 3.55 ND ND 4.25 Hep I H-4
4.05 Hep I H-6
Sial-lNnT unit
-4)-b-Glc-(1-(Glc I) ND ND ND ND ND ND
-3)-b-Gal-(1-(Gal I) 4.45 3.60 3.73 4.17 ND ND
-4)-b-GlcNAc-(1- 4.74 3.81 3.73 3.74 3.59 3.73 Gal-I H-3
-3)-b-Gal-(1-(Gal II) 4.56 ND 4.05 ND ND 3.74 GlcNAc H-4
t-a-Neu5Ac 2.76
1.81
3.68 3.86 4.05 Gal-II H-3
Ó FEBS 2002 Sialylated LPS inHaemophilusinfluenzae (Eur. J. Biochem. 269) 4015
responsible for the initiation of oligosaccharide extension
from Hep I. Inthe absence ofthe glucose acceptor the
sialylated tetrasaccharide unit cannot be attached. In strain
RM118, andthe RM118 lgtC mutant strain, we have now
identified two different sialylated species, namely Sial-Lac
and sialyl-lacto-N-neotetraose (Sial-lNnT). Interestingly,
glycoforms containing both sialylated oligosaccharides were
not observed. It is possible that the expression of one
sialylated structure sterically, or in some other way, hinders
the attachment ofthe second sialylated group. Similarly, the
high molecular weight sialylated glycoform was not
observed inthe wild-type RM118 strain when there was
full extension ofthe globotetraose oligosaccharide (b-
D
-
GalpNAc-(1–4)-a-
D
-Galp-(1–4)-b-
D
-Galp-(1–4)-b-
D
-Glcp-
(1–) from Hep III, again perhaps indicating that steric
interference ofthe two chains precludes coincident expres-
sion of both. It is noteworthy that the majority of Sial-
lNnT-containing glycoforms contain two PEtn residues. In
the inner core of H. influenzae LPS, one PEtn residue is
stoichiometrically present at the 6-position ofthe Hep II
residue, anda second PEtn residue is sometimes attached to
the Kdo-P moiety. The percentage of Sial-lNnT-containing
glycoforms that elaborate two PEtn residues is considerably
higher than the percentage of other glycoforms that elabo-
rate two PEtn residues (Fig. 2), although the significance of
this is unclear. Apart from the presumed steric incompat-
ability ofthe globotetraose-containing RM118 glycoform
with elaboration ofthesialylated tetrasaccharide, all Sial-
lNnT units are found in LPS molecules that have only the
fully extended oligosaccharide at Hep III. Due to the phase-
variable expression ofthe glycosyltransferases LgtC and
Lic2A, a variety of extensions from Hep III are typically
observed inthe nonsialylated glycoforms [5,21]. Taken
together, the correlation of fully extended structures from
Hep III andthe proportion of two-PEtn-containing struc-
tures with the expression ofthe Sial-lNnT unit, would
suggest that the addition of this tetrasaccharide unit utilizes
a preferred LPS core structure or is perhaps a late event in
the LPS biosynthesis pathway. Glycoforms in which the
PCho group on Glc I is either absent or present were also
detected in each ofthe strains investigated. The Sial-lNnT
side-chain is identical to that previously observed in
meningococcal and gonococcal LPS where it is found in
an analogous molecular environment. It is of considerable
interest that this sialylatedstructure has been found to
increase the serum resistance of those organisms bearing this
group [15]. This phenomenon is of particular importance in
strains that lack a capsular structure, which in itself provides
serum resistance. Recent data from our laboratory indicate
that the ability of acapsular strains of H. influenzae to
elaborate sialylated glycoforms can confer serum
resistance in model systems. Inthe study of Hood et al.
[11], a RM118 lgtC lic3A mutant strain that cannot
synthesize Sial-Lac, but as detailed inthe present study
can express Sial-lNnT-containing glycoforms, was resistant
to the bactericidal effect of human sera compared with the
isogenic RM118 lgtC siaB strain, a strain that is unable to
incorporate any sialic acid residues. It is likely that this
Fig. 6. (A) One-dimensional NOESY-TOCSY spectrum using selective excitation of H-3
ax
1
H-resonance ofthe Neu5Ac residue inthe NOESY step
and of H-3
1
H-resonance ofthe Gal II residue inthe TOCSY step and (B) one-dimensional TOCSY-NOESY spectrum using selective excitation of H-1
1
H-resonance ofthe GlcNAc residue inthe TOCSY step andof H-3/H-4
1
H-resonance ofthe GlcNAc residue inthe NOESY step. (A) The assignments
of the
1
H-resonances ofthe Gal II residue are indicated. (B) The assignments of the
1
H-resonances ofthe GlcNAc and Gal II residues are indicated.
The spectrum was recorded in D
2
O at pH 7.0 and 295 K.
4016 A. D. Cox et al. (Eur. J. Biochem. 269) Ó FEBS 2002
residual serum resistance ofthe RM118 lgtC lic3A
strain was due to the Sial-lNnT structure. H. influenzae
requires an exogenous source of sialic acid and thus
the ability to sequester sialic acid from the host and to
then utilize this moiety to modify LPS molecules is likely to
play an important role inthe virulence of H. influenzae
strains.
Previous studies from our laboratory had identified the
sialyltransferase gene (lic3A) responsible for sialylation of
the lactose moiety [11]. As shown in this study the lic3A
mutant can elaborate Sial-lNnT-containing glycoforms
pointing to the presence of another sialyltransferase gene
in the RM118 genome that specifically adds sialic acid to the
b-
D
-Galp-(1–4)-b-
D
-GlcNAcp-(1–3)-b-
D
-Galp-(1-trisaccha-
ride. Several candidates are currently under investigation for
this role and include HI0871 that is a homologue to the
sialyltransferase gene lst of H. ducreyi [29] and HI1699 that
is a homologue to the sialyltransferase gene lst of N. men-
ingitidis [30]. This latter gene is part ofa seven-gene locus,
lsgA-G that when expressed in Escherichia coli results in the
production of chimeric LPS that reacts with MAb 3F11
(specific for the terminal lactosamine disaccharide of lNnT)
[31].
The results ofthe present study would suggest that the
Sial-lNnT group is added to the LPS as a complete unit via
a mechanism involving block addition ofthe terminal
tetrasaccharide unit to the Glc I acceptor. This behaviour
differs from the expression ofthe Sial-lNnT group of
neisserial LPS, which involves sequential addition of
monomeric units, and invariably a series of truncated
Sial-lNnT structures is observed which indeed accounts for
some ofthe different meningococcal LPS immunotypes.
The presence of truncated Sial-lNnT structures is com-
monly encountered due to the phase variable glyco-
syltransferases ofthe lgt operon that synthesize the lNnT
structure in meningococcal LPS [32]. Truncated structures
due to incomplete biosynthesis ofthe Sial-lNnT unit were
not observed in RM118 H. influenzae strains in this study.
TheSDS/PAGEprofilesandMSdataindicatedthe
presence ofthe Sial-lNnT structure as a complete unit
and no evidence for loss of any residues in this structure
was observed. This profile has also been observed in other
H. influenzae strains (unpublished data). This suggests an
interesting biosynthetic scheme for the attachment of this
structure in H. influenzae. One explanation is that biosyn-
thesis is more akin to the biosynthesis of an O-antigen,
where each repeating unit ofthe O-antigen is often
transferred as a complete unit to the growing LPS molecule.
An alternative explanation could be that cooperative
glycosylation reactions ofa series of glycosyltransferases
are required for synthesis, although such cooperativity has
not been described before in H. influenzae LPS synthesis.
Studies are underway in our laboratory to elucidate the
genetic mechanisms involved.
This study is the first to provide definitive structural
evidence for the presence ofasialylated lacto-N-neotetra-
ose moiety in H. influenzae LPS, astructure which is
potentially of great importance to the virulence of this
organism.
Fig. 7. (A) GC-MS trace of partially methylated alditol acetates derived
from O-deacylated, dephosphorylated LPS from strain RM118 lpsA
grown on BHI plates supplemented with Neu5Ac and (B) fragmentation
pattern ofa 4-linked hexose residue.
Fig. 8. Structural model ofthesialylated tetrasaccharide-containing glycoforms of O-deacylated LPS of H. in fluenza e strain RM118. Mutant strains
with truncated oligosaccharide extensions from Hep III residue are as indicated.
Ó FEBS 2002 Sialylated LPS inHaemophilusinfluenzae (Eur. J. Biochem. 269) 4017
ACKNOWLEDGEMENTS
We thank D. Krajcarski for ES-MS and S. Larocque for assistance with
NMR experiments.
REFERENCES
1. Foxwell, A.R., Kyd, J.M. & Cripps, A.W. (1998) Nontypable
Haemophilus influenzae: Pathogenesis and prevention. Microbiol.
Mol. Biol. Rev. 62, 294–308.
2. Phillips, N.J., Apicella, M.A., Griffiss, J.M. & Gibson, B.W.
(1992) Structuralcharacterizationofthe cell surface lipooligo-
saccharides from a non-typable strain ofHaemophilus influenzae.
Biochemistry 31, 4515–4526.
3. Masoud, H., Moxon, E.R., Martin, A., Krajcarski, D. &
Richards, J.C. (1997) Structureofthe variable and conserved
lipopolysaccharide oligosaccharide epitopes expressed by Haemo-
philus influenzae serotype b strain Eagan. Biochemistry 36, 2091–
2103.
4. Rahman, M.M., Gu, X X., Tsai, C M., Kolli, V.S.K. & Carlson,
R.W. (1999) Thestructural heterogeneity ofthe lipooligo-
saccharide (LOS) expressed by pathogenic non-typeable Haemo-
philus influenzae strain NTHi 9274. Glycobiology 9, 1371–1380.
5. Risberg, A., Masoud, H., Martin, A., Richards, J.C., Moxon,
E.R. & Schweda, E.K.H. (1999) Structural analysis ofthe lipo-
polysaccharide epitopes expressed by a capsular deficient strain of
Haemophilus influenzae Rd. Eur. J. Biochem. 261, 171–180.
6. Schweda, E.K.H., Hegedus, O.E., Borrelli, S., Lindberg, A.A.,
Weiser, J.W. & Moxon, E.R. (1993) Structural studies of the
saccharide portion of cell envelope lipopolysaccharide from
Haemophilus influenzae strain AH1-3 (lic3+). Carbohydr. Res.
246, 319–330.
7. Mansson, M., Bauer, S.H.J., Hood, D.W., Richards, J.C.,
Moxon, E.R. & Schweda, E.K.H. (2001) A new structural type for
Haemophilus influenzae lipopolysaccharide. Eur. J. Biochem. 268,
2148–2159.
8.Cox,A.D.,Masoud,H.,Thibault,P.,Brisson,J R.,vander
Zwan, M., Perry, M.B. & Richards, J.C. (2001) Structural analysis
of thelipopolysaccharide from the non-typable Haemophilus
influenzae strain SB 33. Eur. J. Biochem. 268, 5278–5286.
9. Li, J., Bauer, S.H.J., Mansson, M., Moxon, E.R., Richards, J.C. &
Schweda, E.K.H. (2001) Glycine is a common substituent of the
inner-core inHaemophilusinfluenzae lipopolysaccharide. Glyco-
biology 11, 1009–1015.
10. Hood, D.W., Makepeace, K., Deadman, M.E., Rest, R.F.,
Thibault, P., Martin, A., Richards, J.C. & Moxon, E.R. (1999)
Sialic acid inthelipopolysaccharideofHaemophilus influenzae:
strain distribution, influence on serum resistance and structural
characterization. Mol. Microbiol. 33, 679–692.
11. Hood, D.W., Cox, A.D., Gilbert, M., Makepeace, K., Walsh, S.,
Deadman, M.E., Cody, A., Martin, A., Manson, M., Schweda,
E.K.H.,Brisson,J R.,Richards,J.C.,Moxon,E.R.&
Wakarchuk, W.W. (2001) Identificationofa lipopolysaccharide
a-2,3-sialyltransferase from Haemophilus influenzae. Mol. Micro-
biol. 39, 341–350.
12. Bauer, S.H.J., Mansson, M., Hood, D.W., Richards, J.C.,
Moxon, E.R. & Schweda, E.K.H. (2001) A rapid and sensitive
procedure for the determination of 5-N-acetyl neuraminic acid
in lipopolysaccharides ofHaemophilus influenzae: a survey of 24
non-typeable H. influenzae strains. Carbohydr. Res. 335, 251–260.
13. Pavliak, V., Brisson, J R., Michon, F., Uhrin, D. & Jennings, H.J.
(1993) Structureofthesialylated L3 lipopolysaccharide of
Neisseria meningitidis. J. Biol. Chem. 268, 14146–14152.
14. Kogan, G., Uhrin, D., Brisson, J R. & Jennings, H.J. (1997)
Structural basis ofthe Neisseria meningitidis immunotypes
including the L4 and L7 immunotypes. Carbohydr. Res. 298,
191–199.
15. Van Putten, J.P.M. (1993) Phase variation of LPS directs inter-
conversion of invasive and immuno-resistant phenotypes of
Neisseria gonorrhoeae. EMBO J. 12, 4043–4051.
16. Mandrell,R.E.,Lesse,A.J.,Sugai,J.V.,Shero,M.,Griffiss,J.M.,
Cole, J.A., Parsons, N.J., Smith, H., Morse, S.A. & Apicella, M.A.
(1990) In vitro andin vivo modification of Neisseria gonorrhoeae
lipooligosaccharide epitope structure by sialylation. J. Exp. Med.
171, 1649–1664.
17. Mandrell,R.E.,Kim,J.J.,John,C.M.,Gibson,B.W.,Sugai,J.V.
& Apicella, M.A. (1991) Endogenous sialylation ofthe lipo-
polysaccharides of Neisseria meningitidis. J. Bacteriol. 173, 2823–
2832.
18. Vimr, E., Lichtensteiger, C. & Steenbergen, S. (2000) Sialic acid
metabolism’s dual function inHaemophilus influenzae. Mol.
Microbiol. 36, 1113–1123.
19. Mandrell, R.E., McLaughin, R., Abu Kwaik, Y., Lesse, A.,
Yamasaki, R., Gibson, B., Spinola, S.M. & Apicella, M.A. (1992)
Lipooligosaccharides (LOS) of some Haemophilus species mimic
human glycosphingolipids, and some LOS are sialylated. Infect.
Immun. 60, 1322–1328.
20. Phillips, N.J., Apicella, M.A., Griffiss, J.M. & Gibson, B.W.
(1993) Structural studies ofthe lipooligosaccharides from
Haemophilus influenzae typebstrainA2.Biochemistry 32,
2003–2012.
21. Hood, D.W., Cox, A.D., Wakarchuk, W.W., Schur, M., Schweda,
E.K.H., Walsh, S., Deadman, M.E., Martin, A., Moxon, E.R. &
Richards, J.C. (2001) Genetic basis for expression ofthe major
globotetraose containing lipopolysaccharide from Haemophilus
influenzae strain Rd (RM118). Glycobiology 11, 957–967.
22. Weiser, J.N., Love, J.M. & Moxon, E.R. (1989) The molecular
mechanism of phase variation ofHaemophilusinfluenzae lipopo-
lysaccharide. Cell 59, 657–665.
23. Hood, D.W., Deadman, M.E., Allen, T., Masoud, H., Martin, A.,
Brisson, J R., Fleischmann, R., Venter, J.C., Richards, J.C. &
Moxon, E.R. (1996) Use ofthe complete genome sequence
information ofHaemophilusinfluenzae strain Rd to investigate
lipopolysaccharide biosynthesis. Mol. Microbiol. 22, 951–965.
24. Westphal, O. & Jann, K. (1965) Bacterial lipopolysaccharides.
Extraction with phenol-water and further applications of the
procedure. Methods Carbohydr. Chem. 5, 83–91.
25. Lysenko, E., Richards, J.C., Cox, A.D., Stewart, A., Martin, A.,
Kapoor, M. & Weiser, J.N. (2000) The position of phosphor-
ylcholine on thelipopolysaccharideofHaemophilus influenzae
affects binding and sensitivity to C-reactive protein-mediated
killing. Mol. Microbiol. 35, 234–245.
26. Thibault, P. & Richards, J.C. (2000) Application of combined
capillary electrophoresis electrospray mass spectrometry in the
characterisation of short chain lipopolysaccharides: Haemophilus
influenzae.InMethods of Molecular Biology, Vol. 145 Bacterial
Toxins: Methods and Protocols (Holst O., ed.), pp. 327–344.
Humana Press, Totowa, NJ.
27. Uhrin, D., Brisson, J R., Kogan, G. & Jennings, H.J. (1994) 1D
Analogs of 3D NOESY-TOCSY and 4D TOCSY-NOESY-
TOCSY experiments. Application to polysaccharides. J. Magn.
Reson. B104, 289–293.
28. Cox, A.D., Howard, M.D., Brisson, J R., van der Zwan, M.,
Thibault, P., Perry, M.B. & Inzana, T.J. (1998) Structural analysis
of the phase variable lipooligosaccharide from Haemophilus som-
nus strain 738. Eur. J. Biochem. 253, 507–516.
29. Bozue, J.A., Tullius, M.V., Wang, J., Gibson, B.W. & Munson,
R.S. Jr (1999) Identificationofa novel sialyltransferase gene from
Haemophilus ducreyi and construction of mutants deficient in the
production of sialic acid containing glycoforms ofthe lipooligo-
saccharides. J. Biol. Chem. 274, 4106–4114.
30. Gilbert, M., Watson, D.C., Cunningham, A.M., Jennings, M.P.,
Young, N.M. & Wakarchuk, W.W. (1996) Lipopolysaccharide
4018 A. D. Cox et al. (Eur. J. Biochem. 269) Ó FEBS 2002
[...]... 2002 Sialylated LPS inHaemophilusin uenzae (Eur J Biochem 269) 4019 a- 2,3-sialyltransferase from the bacterial pathogens Neisseria meningitidis and Neisseria gonorrhoeae J Biol Chem 271, 28271–28276 31 Phillips, N.J., Miller, T.J., Engstrom, J.J., Melaugh, W., McLaughlin, R., Apicella, M .A & Gibson, B.W (2000) Characterizationof chimeric lipopolysaccharides from Escherichia coli strain JM109 transformed... transformed with lipooligosaccharide synthesis genes (lsg) from Haemophilusin uenzae J Biol Chem 275, 4747– 4758 32 Wakarchuk, W., Martin, A. , Jennings, M.P., Moxon, E.R & Richards, J.C (1996) Functional relationships ofthe genetic locus encoding the glycosyl transferase enzymes involved inthe expression ofthelacto-N-neotetraose terminal lipopolysaccharidestructurein Neisseria meningitidis J Biol Chem . Identification and structural characterization of a sialylated
lacto-N-neotetraose structure in the lipopolysaccharide
of
Haemophilus in uenzae
Andrew. and ions arising from the inner core and from the addition of the
sialylated tetrasaccharide are indicated.
Ó FEBS 2002 Sialylated LPS in Haemophilus in uenzae