Glycobiology vol. 7 no 6 pp. 737-743, 1997
Sulfated glycans on oral mucin as receptors for Helicobacter pylori
E.C.I.Veennan1-3, CM.C.Bank1-2, F.Namavar2,
BJAppelmelk 2 , J.G.M.Bolscher1 and A.V.Nieuw
Amerongen1
'Department of Oral Biochemistry, Academic Centre for Dentistry
Amsterdam (ACTA) and 2Department of Medical Microbiology, Vrije
Universiteit, Van der Boechorststraat 7, 1081 BT Amsterdam, The
Netherlands
'To whom correspondence should be addressed
Helicobacter pylori is able to colonize gastric epithelia, causing chronic active gastritis, gastric and duodenal ulcers and
presumably gastric malignancies. Attempts to identify the
natural reservoir for this microorganism other than the
stomach have been unsuccessful. It is suspected that H.pylori can be transmitted orally, since the microorganism has
been detected at various sites of the oral cavity. The aim of
the present study was to determine whether H.pylori can
bind to salivary mucins, which in vivo coat the oral epithelia, and characterize further the interaction. Binding of
salivary mucins and of synthetic oligosaccharides was studied in ELISA and immunoblotting, using specific monoand polyclonal antibodies, and synthetic neoglycoconjugates. H.pylori bound most avidly to a highly sulfated subpopulation of high molecular weight salivary mucins, secreted from the palatine salivary glands, and with less avidity to mucin species secreted by the sublingual and
submandibular salivary glands, which are less sulfated.
Binding was strongly enhanced upon decreasing pH from
6.0 to 5.0. Using synthetic polyacrylamide coupled oligosaccharides it was found that SO 3 -3-Gal and the SO 3 -3Lewis" blood group antigen bound to H.pylori. In contrast,
binding of sialylated Lewis" and Lewisb antigens was much
weaker. This study indicates that sulfated oligosaccharides
on salivary mucins may provide receptor structures for
adhesion of H.pylori to oral surfaces.
cavity is important Human saliva contains several proteins
implicated in the protection of the oral cavity against bacterial
colonization, including slgA, lysozyme, proline-rich proteins,
and the high- and low-M, molecular weight salivary mucins,
MG1 and MG2, respectively (Loomis etal., 1987; Lamkin and
Oppenheim, 1993; Nieuw Amerongen etal., 1995). Mucins are
a family of highly glycosylated proteins (carbohydrate > 80%),
which cover and protect epithelial tissues throughout the human body. High M, salivary mucins are secreted mainly by the
(sero)mucous submandibular, sublingual, and palatine glands.
As main constituents of mucous layers investing the oral tissues, mucins are the first sites of attachment for microorganisms that colonize the oral cavity. On the other hand,
salivary mucins, in particular MG2, acting as soluble receptor
analogs for bacterial adhesins, promote oral clearance of micro-organisms. High Mr mucins contain a wide spectrum of
structurally different oligosaccharide side chains, some of
which carry blood group active antigens (e.g., ABH, Lea, Leb)
(Prakobphol et al., 1993), and may function as receptors for
bacterial adhesins. For example, H.pylori binding to gastric
mucins (Piotrowski et al, 1991; Tzouvelekis et al, 1991) is
mediated by sulfated carbohydrate side-chains, which have not
been further characterized. In addition, a number of other structures have been implicated in the attachment of H.pylori to
gastric epithelial cells, viz. phosphatidylethanolamine (Lingwood et al., 1992), GM3 ganglioside (Slomiany et al., 1989;
Saitoh et al, 1991), sialyl-lactose (Evans et al., 1988), and the
blood group-related Lewisb antigen (Bore"n et al, 1993). In
view of the suggested oral transmission route of H.pylori, we
studied the interaction between salivary components and H.pylori. It was found that H.pylori bound preferentially to highly
sulfated mucins secreted by the palatine glands, and that this
interaction was enhanced upon lowering pH. Furthermore, we
have identified a number of potential oligosaccharide receptors
of H.pylori, including sulfated Lewis0 and galactose-3-sulfate.
Key words: W.py/ori/saliva/sulfomucin/nickel/oligosaccharide
Results
Introduction
H.pylori is a causative agent in chronic active gastritis, gastric
and duodenal ulcers, and presumably gastric malignancies. No
natural reservoir for this microorganism other than the stomach
have been found as yet. H.pylori might be transmitted via the
oral route, since the microorganism has been detected at various sites of the oral cavity, by culture as well as by polymerase
chain reaction (Banatvala et al, 1993; Ferguson et al., 1993;
Nguyen et al., 1993; Li et al., 1995; Namavar et al, 1995).
Because the oral cavity functions as the port of entry to the
gastro-intestinal system, defensive systems in saliva are relevant not just for maintenance of the oral health but for the
entire organism. In this context, defining the factors involved
in the attachment of this pathogen to components of the oral
O Oxford University Press
To find out whether salivary components are involved in binding of H.pylori, we have tested which of die major salivary
proteins can bind in vitro to H.pylori. Of the proteins tested
only the high M,. salivary mucin MG1 was bound. The other
salivary proteins, viz. a-amylase, lysozyme, proline-rich proteins, and cystatin S did not bind to H.pylori, either in ELISA
or on immunoblots (data not shown).
The binding of high M,. salivary mucins (MG1) was further
investigated in ELISA, in Western blots, and in immunoelectron microscopy. In the present study two strains of H.pylori
have been used, H.pylori NCTC 11637 and strain 3B3, isolated
from me subgingival dental plaque of a gastritis patient Although in this article the results are shown that were obtained
with strain NCTC 11637, essentially the same results were
obtained with strain 3B3.
Figure 1 shows die binding of MG1 to Helicobacter pylori
737
E-CLVeennan el al
500
so
SattvatOution
25
OO490
1.5 -
0.5
50
500
5000
Saliva dilution
S a i n (Button
Fig. 1. Binding of high molecular weight salivary mucins (MG1) to
H.pylori at different ionic strengths (A) and pH (B), and in the presence of
metal ions (C). (A) Whole saliva (blood group A, non-secretor) was 2-fold
serially diluted (in duplicate) in 10 mM potassium phosphate, in the
presence of the indicated concentrations of sodium chloride. (B) Whole
saliva was two-fold serially diluted (in duplicate), in 10 mM potassium
phosphate buffers containing 150 mM sodium chloride and 0.5% Tween-20.
pH was varied between 3.0 and 7.0. Note the differences in scale between
x-axis in (A) and (B). (C) Saliva was 2-fold serially diluted (in duplicate) in
10 mM Tris buffer, pH 7.4, containing 0.5% Tween 20. Concentrations of
NiCl 2 , ZnSO 4 , and FeS0 4 were maintained at 1 mM.
in ELISA with mAb F2. The importance of physiological pHand hypotonic conditions in saliva for the binding of the bacterium to the high M, mucins was analyzed by testing binding
at various ionic strength and pH values. After increasing the
738
NaCl concentration from 25 to 500 mM, a gradual decrease in
binding of MG1 was observed (Figure 1 A). On the other hand,
lowering the pH from 6.0 to 5.0 resulted in a profound increase
in binding (Figure IB). These results indicate that under the
physiological pH- and hypotonic conditions present in saliva,
H.pylori can bind to the high molecular weight salivary mucin
species MG1. Similar results were obtained when saliva or
purified salivary mucins from donors with different blood
groups (three blood group A, two secretors and one nonsecretor; one blood group B, secretor) were tested, or when
detection was conducted with another anti-MGl mAb IN A (not
shown). The ionic strength dependency suggested that the
binding was governed by electrostatic interactions. No specific
effect of calcium, magnesium, or sulfate ions, other than due to
ionic strength, was observed (not shown). In contrast, divalent
metal ions in low concentrations (1 mM) profoundly enhanced
the MG1 binding to Helicobacter pylori at neutral pH values,
in the order Ni2+ > Zn 2+ > Fe 2+ (Figure 1C). When binding was
tested at different Zn2+ concentrations, it was found that maximal enhancing effect of Zn2+ ions occurred between 0.05 and
0.1 mM (not shown).
Binding of salivary mucins to H.pylori was confirmed by
immunoblotting with mAbs F2 and E9. Figure 2 shows the
immunoblots of the proteins extracted from H.pylori after incubation with saliva. In confirmation with the results of the
ELISA (Figure 1), it was found that H.pylori bound to high M,.
mucins (Figure 2, lane 2). Probing with anti-mucin mAb E9
revealed that the mucin fraction that bound to H.pylori contained, relative to the starting material, only low amounts of the
E9 epitope, i.e., sialylated Lewis" (compare Figure 2, lanes 4
and 5). These results, therefore, suggested that H.pylori preferentially bound to a mucin subpopulation that is enriched in
F2 epitopes, i.e., sulfo-Lewis". The suggestion that binding of
salivary mucins was mediated by receptors on Helicobacter
pylori recognizing sulfated glycoconjugates was corroborated
by the finding that mucin binding was inhibited best by sulfated biopolymers, including chondroitin sulfate and heparin
(Figure 3). After removal of sulfate from salivary mucins by
acid solvolysis, a 4-fold lowering in binding avidity was noted,
again pointing to a role for sulfate residues in the binding (not
shown). In contrast, sialic acid residues appeared not to be
essential in this respect, since neither oxidation by sodiummefa-periodate, nor mild acid hydrolysis of salivary mucins,
which removed ah1 sialic acid residues from the mucin, had an
effect on binding (not shown).
Immunoelectron microscopy of H.pylori after incubation
with purified mucins and labeling with an anti-MGl mAb, is
shown in Figure 4, illustrating that binding of mucin occurred
to structures present at the periphery of the cell.
To examine further the involvement of sulfated mucins, we
conducted binding experiments with isolated salivary MG1
species with different sulfate contents. ELISAs with several
purified MG1 species, isolated from segregated collected glandular secretions, indicated that the highly sulfated palatine
MG1 species bound with the highest avidity (Figure 5), which
is in agreement with the results of the immunoblots. At pH of
5.0, binding of palatine MG1 was saturated at a concentration
as low as 0.2 fig/ml, whereas under these conditions virtually
no binding of H.pylori to the low sulfated submandibular MG1
species could be demonstrated. However, for all MG1 species
tested, including submandibular MG1, binding increased upon
lowering the pH from 6.0 and 4.0. The order in binding affinities (palatine MG1 > sub lingual MG1 > submandibular MG1)
Helicobacter binding to an oral mudn receptor
OD 490 ran
3
2.5
2
1 5
1
05
0 0001
Fig. 3. Effect of polyanions on binding of salivary mucins to Helicobacter
pylori. Saliva was 2-fold serially diluted (in duplicate) in 50 mM sodium
acetate buffer (pH 5.0), 150 mM NaCl, 0.5% Tween, containing 1 p-g/ml of
the indicated polyanion and incubated with Helicobacter pylori coated onto
polystyrene microtiterplate wells for 2 h at 37°C. Bound mucins were
detected with mAb F2.
tively, while low, if any inhibition was seen upon preincubation with sialylated Lewis" (Figure 7B).
Discussion
1 2
3
4
5 6
Fig. 2. Immunoblotring of salivary mucins extracted from H.pylori after
incubation with saliva. Saliva was 2-fold diluted in 50 mM sodium acetate
buffer, containing 150 mM sodium chloride, pH 5.0; 107 H.pylori cells were
incubated with the diluted saliva at 37°C for 2 h. Adhered salivary
components were extracted, separated by electrophoresis (4-15% gradient
polyacrylamide gels), transferred onto nitrocellulose filters and probed with
mAb F2 Oanes 1-3) and mAb E9 (lanes 4-6) Lanes 1 and 4, whole saliva
(starting material); lane 2 and 5, H.pylori extract, after incubation with
saliva; lanes 3 and 6, Control extract, H.pylori, not incubated with saliva.
Arrow indicates the position of the high Mr salivary mucin MG1.
ran parallel with the order in sulfate-content of the different
mucin species (Table I)- Furthermore, at pH 4.0 binding of
isolated salivary mucins to Helicobacter pylori was compared
with that of a gastric mucin preparation (Figure 6). The latter
bound to Helicobacter pylori with comparatively low avidity,
saturation occurring at a concentration of approximately 10
u,g/ml. As was found with salivary mucins, binding of gastric
mucin decreased upon increasing pH to more neutral values
(not shown).
The chemical nature of the oligosaccharides involved in
binding was further investigated by comparing the binding of
synthetic oligosaccharides (Table H). The results, shown in
Figure 7A, demonstrated that H.pylori bound most avidly to
polymeric sulfated carbohydrates, including SO3-3-galactose,
SO3-3-Lewis", and SO3-6-GlcNAc (the latter not shown). Intermediate binding was observed with Lewis" and the HI antigen, while binding to Lewisb was least. Similar to salivary
mucins, oligosaccharides displayed higher binding upon lowering pH and in the presence of Ni 2+ ions (not shown). Inhibition studies conducted with three selected oligosaccharides
indicated that preincubation with (polymeric) SO3-3-galactose
inhibited subsequent binding of salivary mucins most effec-
A number of studies (Banatvala et al, 1993; Ferguson et al,
1993; Nguyen et al. 1993; Li et al, 1995; Namavar et al.,
1995) have demonstrated the presence of H.pylori at oral epithelia, such as the palate and the cheeks, in saliva and in dental
plaque of patients suffering from gastric infections. These stud-
Fig. 4. Immunogold labeling of mucin bound to H.pylori. After incubation
of Helicobacter pylori with salivary rnucins at pH 5.0, bacteria were washed
and transferred onto grids. Cells were washed three times to remove
nonbound mucin and applied to Formvar carbon coated grids, which were
subsequently incubated with mAb INA for 30 min. The reaction of mAb
with bacterial cells was visualized using a colloidal gold (particle size 6
nm) conjugated goat anti-mouse polyclonal serum (diluted 1:20 in
PBS/glyrine. When incubation with salivary mucins was omitted no
labeling occurred (not shown).
739
E.C.LVeerman et al
OD 490 nm
Table L Characteristics of high-molecular weight salivary mucin
preparations (MG1) from different salivary glands , used in this study
Mucin species
Sublingual
Submandibular
Palatine
Number of residues
per molecule*
Reactivity towards
mAbb
Sialic acid
Sulfate
E9
F2
359
647
194
219
135
1239
++
++
—
++
+
+++
"Bolscher et al (1995), assuming a molecular weight of 1,000,000 Da.
""Veerman et al., 1992.
Donor blood group A, non-secretor.
ies suggested that the oral cavity harbors H.pylori and may be
a source of (re)infection and transmission. In keeping with this,
the present study showed that salivary mucins under the hypotonic conditions operating in vivo, can support binding of the
gastric pathogen H.pylori, and that binding is enhanced upon
lowering pH. Under physiological conditions, the pH in saliva
fluctuates between 5.8 and 7.8 (Jenkins, 1978), but pH values
as low as 4.0 have been measured in carbohydrate metabolizing dental plaque (Jensen and Schachtele, 1983). Thus, salivary
mucins as constituents of mucous layers coating oral surfaces
(Nieuw Amerongen et al., 1987; Bradway et al., 1992) can
provide sites of attachment for H.pylori.
The present study indicates that in vitro H.pylori binds best
to a highly sulfated subpopulation of salivary mucins (Figure
2.5
OO490
1.5
Salivary mucins
Gastric mucins
0.5
Buffer
0.01
1
mudn concentration (|ig/ml)
10
100
Fig. 6. Comparison of gastric and salivary mucins with respect to binding to
H.pylori. Mucins were incubated for 2 h at 37°C in wells of microa'terplates
coated with H.pylori. Incubation was conducted in 50 mM sodium acetate
buffer, containing 150 mM NaCl and 0.5% Tween-20 at pH 4.0. Bound
mucins were detected with mAb F2 (salivary mucins) or with mAb 19-OLE
to H-type 2 (gastric mucins).
0.1
1
mucin concentration (ug/ml)
740
Fig. 5. Binding of isolated glandular salivary MG1 species to H.pylori in
F1..ISA at several pH-values. Mucins were incubated for 2 h at 37°C in
wells of microtiterplates coated with H.pylori. Incubation was conducted in
10 mM potassium phosphate buffer, containing 150 mM NaCl and 0.5%
Tween-20 at the indicated pH. Bound mucins were detected with mAb F2.
Under these conditions, direct binding of mucins to polystyrene did not
occur. PAL-MG1, palatine high M, mucins; SL-MG1, sublingual high M,
mucins; SM-MG1, submandibular high M, mucins.
Hdlcobacter binding to an oral mudn receptor
the other hand salivary mucins from a non-secretor, missing
the Lewis'3 antigen, are readily bound (cf. Figure 1).
The presently found pH-dependency suggests that protonation of weakly acidic groups is critical for binding of salivary
mucins to their receptors on H.pylori. The finding that metal
ions, including nickel and zinc, enhance binding of H.pylori to
salivary mucin points also to a role for cysteine and histidine
residues, which often provide binding sites for these metal ions
(Gilbert et al., 1995). Interestingly, H.pylori expresses several
proteins displaying high affinity for nickel, including urease
(Hawtin et al., 1991), membrane proteins implicated in nickel
transport (Mobley et aL, 1991) and heat shock proteins (Suerbaum et al., 1994) which have been implicated in sulfatide
recognition (Huesca et al. 1996). Based on the present findings, it is speculated that nickel binding proteins play a role in
the binding of H.pylori to its receptors.
OD490nm
Materials and methods
Bacterial strains
H.pylori strain NCTC 11637 and clinical strain 3B3 (Namavar et al., 1995)
were grown in Brucella broth (Difco) supplemented with 5% fetal calf serum
and incubated under microaerophilic condition at 37°C. After 48 h, cultures
were harvested, washed twice with phosphate-buffered saline (PBS, pH 7.2),
and resuspended in PBS to an O D ^ of 0.1 and stored at -80°C until use.
Antibodies
0.0001
0.01
0 001
Saliva dilution
Fig. 7. Binding of synthetic biotinylated oligosaccharides to H.pylori by
F.I .ISA. (A) Synthetic oligosaccharides were diluted in 50 mM sodium
acetate, 150 mM NaCl, 0.5% Tween 20, pH 5.0. Bound oligosaccharides
were detected with streptavidin conjugated to horseradish peromdase.
Control incubations (buffer alone) yielded OD49O values below 0.05. (B)
Effect of preincubation with neoglycoconjugates on binding of salivary
mucins to H.pylori. Preincubation (1 h) of 2 u.g/ml oligosaccharides was
conducted in the same buffer as in (A), followed by incubation of whole
saliva for 1 h in 2-fold serial dilutions. Bound mucins were detected with
mAb F2.
5). The involvement of sulfated oligosaccharide side chains in
the binding is further corroborated by the findings that sulfated
neoglycoconjugates display highest binding and compete effectively with salivary mucins for the receptor on H.pylori
(Figure 7). These results are in line with previous data showing
that H.pylori bind to sulfated biomolecules, including sulfatides (Huesca et al., 1996; Kamisago et al., 1996), heparan sulfate (Ascencio et al., 1993) and a sulfated subpopulation of
gastric mucins (Piotrowski et al, 1991).
From the difference in binding properties of sulfo-Lewis" at
one hand, and sialyl-Lewis* and Lewis" at the other, it can be
inferred that the presence of the small negatively charged sulfate residue at the terminal galactose-residue of the Lewis"
antigen favors binding, and furthermore that sialic acid cannot
substitute for sulfate in this respect In the present study we did
not find any support for a prevailing role of the Lewisb antigen,
which previously has been identified as a receptor of Helicobacter pylori in gastric epithelial cells (Bor6n et aL, 1994),
since H.pylori did hardly bind to synthetic Lewis'5, whereas at
Production and characterization of anti-mucin monoclonal antibodies F2 and
E9 have been described previously (Rathman et aL, 1990; Veerman et aL,
1991). Antibody F2 recognizes (SO3-3)-Lewis* (Veerman et aL, 1997). Antibody E9 is directed to the sialyl-Lewis* antigen (E. Veerman, unpublished
observations). In addition, salivary blood group A mucins were monitored with
anti-MGl mAb INA, directed against the bloodgroup A antigen. Gastric mucins were monitored widi mAb 19-OLE to H-type 2, obtained from Bioprobe,
Amsterdam, The Nemerlands. Production and characterization of the monoclonal antibody to salivary cystatin S has been described previously (Rathman
et aL, 1990). Rabbit polyclonal antibodies to lysozyme and amylase were from
Sigma, St Louis, MO. Polyclonal antiserum against acidic proline-rich proteins was a kind gift from Dr. M. J. Levine, Buffalo, NY). Biotin labeled
streptavidin was a kind gift from Dr. R. Negrini, Brescia, Italy.
Glycoconjugates
Collection of whole saliva, segregated collection of submandibular, sublingual
and palatine saliva, as well as isolation and immunochemical characterization
of the high molecular weight mucin fractions thereof have been described
previously (Veerman et aL, 1992, 1995; 1996, Bolscher et aL, 1995). Partial
(immuno)cbemical characterization of the isolated glandular mucin species are
given in Table I. Gastric mucins were isolated with a method that previously
has been described for isolation of salivary mucins (Veerman et aL, 1989). In
short, using a glass rod the mucous layer was gendy scraped off from antral
tissue obtained during surgery, and incubated in 50 mM Tris buffer (pH 7.2)
containing 8 M urea, 0.5 M NaCl, 0.2 M KSCN, 10 mM EDTA, 10 mM
benzamidine-Ha, and 1 mM PMSF at 4°C for 3 days. Then, undissolved
material was removed by centrifugation at 5000 x g for 20 min. The resultant
supernatant was subjected to ultracentrifugation for 24 h at 100,000 x g. The
obtained pellet was dissolved in water, dialyzed against distilled water, and
freeze dried.
Oxidation of salivary mucin carbohydrates was performed by iiicubation of
purified mucins with 5 mM sodium m«a-periodate (Veerman et aL, 1995).
Periodate treatment resulted in the complete loss of binding to mAb E9,
whereas reactivity toward mAb F2 was not affected (Veerman et aL, 1996).
Mild acid treatment of salivary mucin was conducted by incubation of purified
salivary mucins with 0.05 M HCl at 80°C for 1 h. Carbohydrate analysis
showed that by this treatment virtually all sialic acid residues are removed,
while the epitope recognized by mAb F2 remains intact (Veerman et aL, 1991).
Desulfation of mucins was conducted by acid solvolysis in anhydrous methanol, containing 0.05 M HCl at 32°C for 3 h. To enable subsequent detection in
binding assays, the desuifated preparation and the parent native mucin preparation were labeled with digoxigenin (DIG) in simultaneous parallel incubations using the DIG-glycan labeling kit, according to the manufacturer's in-
741
E-CLVeerman et al
Table IL Synthetic carbohydrates, multivalently bound to polyacrylamide
carriers, used in this study
Name
Structure
HI antigen
Le'
Leb
Sialyl-Lex
Sulfc-Le*
Sulfo-3-galactose
Sulfo-N-acetylglucosamine
Fucal-2Gaipi-3GlcNAcB-R
Ga^ 1 -3[Fuca 1 -^]GlcNAcp-R
Fucal-2GalBl-3[Fucal-^]GlcNAcp-R
Neu5 Aca2-3Gaip 1 -4[Fuca 1 -3}GlcNAcp-R
SO3-3Gaipi-3[Fucal^]GlcNAcB-R
SO3-3Gal£-R
SO3-6GlcNAcp-R
Biotirf
"
I
I
I
CH-ChU
sugar
I
(CH 2 ) 6
CONH
OH
I
(CH 2 ) 2
(CH 2 ) 3
I
I
CONH
CONH
I
I
CH-CH,
CH-CH, -
15
Structure of the synthetic biotinylated polymer (R) to which
oligosaccharides are attached.
struction (Boehringer, Mannheim, Germany). To ascertain that mutual comparison of signals obtained with these DIG-labeled mucin species was allowed,
direct ELISAs were conducted, in which DIG-labeled mucins were coated
directly to polystyrene wells with subsequent immunoenzyman'c detection using anti-DIG pcroxidase. Identical dose-response curves were obtained, justifying mutual comparison of the signals obtained when binding of these mucins
to Helicobacter pylori was examined.
Chondroitin sulfate A and DNA (from calf thymus) were obtained from
Sigma (St. Louis, MO), heparin from BDH (Poole, UK), and dextran sulfate
(sodium salt from Leuconostoc spp) was obtained from Fluka (Buchs, Switzerland).
Binding assays
Binding of salivary components to H.pylori was studied in an enzyme linked
immunosorbent assay (ELJSA) and by immunoblotting essentially as described
previously (Veerman et al., 1995), using a panel of mono- and polyclonal
antibodies to salivary components. In short, wells of microtiterplates (high
avidity 96-wells ELJSA microtiterplates from Greiner, Germany) were coated
with 100 (il bacterial suspensions (OD 700 = 0.1) overnight at 4°C. Plates were
emptied and incubated at 80°C for 45 min to firmly fix the cells onto the
polystyrene. Saliva or purified mucins were added in duplicate, 2-fold serially
diluted in buffers (either phosphate or sodium' acetate), supplemented with
0.5% Tween 20 to prevent direct binding of mucins to polystyrene, and incubated (37°C, 2 h). After rinsing, plates were probed with relevant monoclonal
antibodies (-1 u-g/ml) or polyclonal antibodies (diluted 1:500). After washing,
bound antibodies were detected with horseradish peroxidase conjugated to
either rabbit anti-mouse-immunoglobulins, or to sheep anti-rabbit immunoglobulins, using o-phenylene-diamine (0.4 mg/ml) and H 2 O 2 (0.012%) as substrates.
Binding to H.pylori of biotinylated polymeric neoglycoconjugates, obtained
from Synthesome, Munich, Germany (Bovin et al., 1993) was studied in
ELJSA essentially in the same way, with enzymatic detection using streptavidin-conjugated to peroxidase. The neoglycoconjugates used, shown in Table II,
are linked via three-carbon spacers to a polyacrylamide type matrix of M, of
approximately 30 kDa. In these compounds approximately every fifth amide
group of the polymer chain is N-substituted by the carbohydrate on a spacer
arm - ( C H j ^ . Negative controls, conducted by incubation of samples in uncoated wells, indicated that direct binding of either salivary components or
synthetic oligosaccharides to uncoated polystyrene wells was negligible. Negative controls, in which incubation with salivary components or neoglycoconjugates was omitted, indicated that direct binding of antibodies to H.pylori did
not occur. To ascertain that mutual comparison of signals obtained with these
biotinylated neoglycoconjugates was allowed, direct ELJSAs were conducted,
in which neoglycoconjugates were coated directly to polystyrene wells with
742
subsequent immunoenzymatic detection using streptavidin-peroxidase. With
all synthetic neoglycoconjugates tested, essentially identical dose-response
curves were obtained, justifying mutual comparison of signals obtained in
binding studies.
Competition of salivary mucins and oligosaccharides for H.pylori was studied as follows: oligosaccharides (final concentration 2 u.g/ml in 50 mM sodium
acetate, 150 mM NaCl, 0.5% Tween 20) were preincubated in microtiterplate
wells coated with H.pylori for 1 h. Then saliva was added and 2-fold serial
diluted in the same buffer containing 2 u.g/ml oligosaccharides, and incubation
was continued for another hour. Residual binding of MG1 to H.pylori was then
monitored using mAb F2, as described above.
Immunoblotting of salivary components extracted from H.pylori after incubation with saliva was conducted as follows: saliva was 2-fold diluted in 50
mM sodium acetate buffer, containing 150 mM sodium chloride, pH 5.0. A
volume of 1 ml of diluted saliva was incubated with 107 H.pylori cells al 37°C
for 2 h. After washing (5 x 1 ml PBS), adhered salivary components were
extracted by overnight incubation at room temperature in electrophoresis
sample buffer (10 mM Tris-glycine, pH 6.8, 10 mM dithiothreitol, containing
2% sodium dodecyl sulfate), followed by boiling for 5 min. After electrophoresis (4-15% polyacrylamide gels, PhastSystem, Pharmacia, Uppsala, Sweden)
separated proteins were transferred to nitrocellulose filters and probed with
mono- and polyclonal antibodies to salivary proteins. After washing, bound
antibody was detected using alkaline-phosphatase conjugated to rabbit antimouse immunoglobulins, or to sheep anti-rabbit immunoglobulins (Dakopatt,
Glostrupp, Denmark), using nitroblue-tetrazolium and 5-bromo-4-chloroindolyl-phosphate as substrates.
Immunoelectron microscopy
H.pylori cells grown for 2 days were suspended in PBS to approximately 10"
cells per ml. Cells were collected by centrifugation, and incubated with purified salivary mucin (0.5 ml, 100 |xg/ml) for 1 h. Cells were washed three times
to remove nonbound mucin and applied to Formvar carbon coated grids, which
were subsequently incubated with mAb INA for 30 min. The reaction of mAb
with bacterial cells was visualized using a colloidal gold (particle size 6 run)
conjugated goat-anti-mouse polyclonal serum (diluted 1:20 in PBS/glycine,
obtained from Aurodye, the Netherlands.
References
AscencioJ5., Fransson,L.A. and Wadstrom,T. (1993) Affinity of the gastric
pathogen Helicobacter pylori for the W-sulphated glycosaminoglycan heparin sulphate. J. Med. Microbiol., 38, 240-244.
Banatvala^N., Lopez,C.R., Owen.R., Abdi.Y., Davies, G., Hardie J. and FeldmanJt. (1993) Helicobacter pylori in dental plaque. Lancet, 341, 380.
BolscherJ., VeermanJS., Nieuw Amerongen^A.V., Tulp^A. and VerwoerdJD.
(1995) Distinct populations of high-M, mucin secreted by different human
salivary glands discriminated by density-gradient electrophoresis. Biochem.
J., 309, 801-806.
Boren.T., FalkJ5., Roth,K.A , Larson.G. and Normark,S. (1993) Attachment of
Helicobacter pylori to human gastric epithelium mediated by blood group
antigens. Science, 262, 1892-1895.
Bovin,N.V., KorchaginaJLY., Zemlyanukhina.T.V., Byramova^N.E., Galanina.O.E., Zemlyakov^A.E., Ivanov^A.E., Zubov.V.P. and MochalovaJ^.V.
(1993) Synthesis of polymeric neoglycoconjugates based on N-substituted
polyacrylamides. Glycoconjugate J., 10, 142-151.
Bradway.S.D., BergeyJEJ., Scannapieco,F.A., RamasubbuJV., Zawacki.S. and
Levine,MJ. (1992) Formation of salivary-mucosal pellicle, the role of transglutaminase. Biochem. J., 284, 557-564.
Evans.D.G., MouldsJJ. and Graham.D.Y. (1988) N-Acetylneuraminyllactosebinding fibrillar hemagglutinin of Campylobacter pylori, a putative colonization factor antigen. Infect. Immun., 56, 2896-2906.
Ferguson,D.A. Jr., Li,C, Patel,N.R., Mayberry.W.R., Chi.D.S. and Thomas.E.
(1993) Isolation of Helicobacter pylori from saliva. J. Clin. Microbiol, 31,
2802-2804.
GilbertJ.V., RamakrishnaJ., Sundennan.F.W.Jr., WrightA and Plau^A.G.
(1995) Protein Hpn, cloning and characterisation of a histidine-rich metalbinding polypeptide in Helicobacter pylori and Helicobacter mustelae. Infect Immun., 63, 2682-2688.
HawtinJ'.R., Delves,H.T. and Newell.D.G. (1991) The demonstration of
nickel in the urease of Helicobacter pylori by atomic absorption. FEMS
Microbiol Lett., 61, 51-54.
Huescajd., Borgia,S., HoffmanJ5. and Lingwood.C.A. (1996) Acidic pHchanges receptor binding specificity of Helicobacter pylori, a binary adhesion model in which surface heat shock (stress) proteins mediate sulfatide
recognition in gastric colonisation. Infect. Immun., 64, 2643-2648.
Helicobacter binding to an oral mucln receptor
Jenltins.G.N. (1978) The physiology and biochemistry of the mouth. Blackwell
Scientific, Oxford.
JensenJvI.E. and Schachtele.CF. (1983) The acidogenic potential of reference
food and snacks at interproximal sites in the human dentition. J. Dent. Res.,
62, 889-892.
Kamisago.S., Iwamori,M., Tai.T., Mitamura,K., Yazaki.Y. and Sugano.K.
(1996) Role of sulfatides in adhesion of Helicobacter pylori to gastric cancer
cells. Infect. Immun., 64, 624-628.
LamkinJtf.S and OppenheimJ.G. (1993) Structural features of salivary function. CriL Rev. Oral BioL Med., 4, 251-259.
U,C, MusichJ".R., Ha,T., Ferguson.D.A., Patel,N.R., Chi.D.S. and Thomas^.
(1995) High prevalence of Helicobacter pylori in saliva demonstrated by a
novel PCR assay. J. Clin. PathoL, 48, 662-666.
Lingwood,C.A., Huesca^M. and Kuksis.A. (1992) The glycerolipid receptor
for Helicobacter pylori (an exoenzyme S) is phospbatidylethanolamine. Infect. Immun., 60, 2470-2474.
Loomis,R.E., Prakobphol,A., Levine,MJ., ReddyJvi.S. and JonesJ'.C. (1987)
Biochemical and biophysical comparison of two mucins from human submandibular-sublingual saliva. Archs. Biochem. Biophys., 258, 452-464.
Mobley.H.L., Gamer.R.M. and BauerfeindJ3. (1995) Helicobacter pylori
nickel-transport gene nixA, synthesis of catalytically active urease in Escherichia coli independent of growth conditions. MoL MicrobioL, 16, 97-109.
Namavar F., Roosendaal.R., Kuipers.EJ., de Groot.P., van der Bijl.M.,
Pena^i.S. and de Graaff J . (1995) Presence of Helicobacter pylori in the oral
cavity, oesophagus, stomach and faeces of patients with gastritis. Eur. J.
Clin. MicrobioL Infect. Dis., 14, 234-237.
NguyeivA.M., EngstrandX., Genta^.M., Graham,D.Y. and el-Zaatari,F.A.
(1993) Detection of Helicobacter pylori in dental plaque by reverse transcription polymerase chain reaction. /. Clin. MicrobioL, 31, 783-787.
Nieuw AmerongenAV., BolscherJ.G.M. and Veerman,E.C.I. (1995) Salivary
mucins, their diversity in relation to their function. Gtycobiology, 5, 733—
740.
Nieuw Amerongen,A.V., Oderkerk,C.H. and Driessens,A.A. (1987) Role of
mucins from whole saliva in the protection of tooth enamel against demineralization in vitro. Caries Res., 21, 297-309.
PiotrowskiJ., Slomiany.A., Murty.V.L.N., Fekete.Z. and Slomiany.B.L.
(1991) Inhibition of Helicobacter pylon colonization by sulfated gastric
mucin. Biochem. Int., 4, 749-756.
Prakobphol^A., Leffler.H. and Fisher.SJ. (1993) The high-molecular-weight
human mucin is the primary salivary carrier of ABH., Le* and Le b blood
group antigens. Crit. Rev. Oral BioL Med., 4, 325-333.
Rathman.W.M., van den Keijbus,P.A.M., van Zeyl^lJ., Dopp^E.A., Veerman.E.C.L and Nieuw Amerongen^A.V. (1990) Characterization of monoclonal antibodies to human salivary (glyco)proteins. Cellular localization of
mucin, cystatin-like 14 kDa protein and 20 kDa glycoprotein in the human
submandibular gland. J. BioL Buc., 18, 19-27.
Saitoh.T., Natomi.H., Zhao.W.L., Okuzumi.K., Sugano.K., Iwamori>). and
Nagai,Y. (1991) Identification of glycolipid receptors for Helicobacter pylori by TLC-immunostaining. FEBS Lett, 282, 385-387.
Slomiany3-L., PiotrowskiJ., Samanta^A., van Horn.K., Murty.V.L. and Slomiany^A. (1989) Campylobacter pylori colonization factor shows specificity
for lactosylceramide sulphate and GM3 ganglioside. Biochem. Int. 19, 929936.
Suerbaum.S., ThibergeJ.M., KansauJ., Ferrero^.L. and Labigne, A. (1994)
Helicobacter pylori hspA-hspB heat-shock gene cluster, nucleotide sequence, expression, putative function and immunogenicity. MoL MicrobioL,
14, 959-74.
Tzouvelekis,L.S., MentisAF., Makris.A.M., Spiliadis.C, Blackwell.C. and
WeirJD.M. (1991) In vitro binding of Helicobacter pylori to human gastric
mucin. Infect. Immun., 59, 4252-4254.
Veerman,E.C.I., Valentijn-Benz,M., BankJ?., and Nieuw Amerongen^A.V.
(1989) Isolation of high molecular weight mucins from human whole saliva
by ultracentrifugation. J. BioL Buc., 17, 307-312.
Veerman,E.C.I., Valentijn-Benzjvl., van den Keijbus.P.A.M., Rathman.W.M.,
SheehanJ.K. and Nieuw Amerongen^A.V. (1991) Immunochemical analysis
of salivary mucins using monoclonal antibodies. Arch. Oral BioL, 36, 923932.
Veerman E.C.I., van den Keijbus.P.A.M., Valentijn-Benz,M. and Nieuw Amerongen,A.V. (1992) Isolation of different high molecular weight mucin
species from human whole saliva. Biochem. J., 83, 807-811.
Veerman.E.C.1., van den Keybus.P.A.M., VissinkA. and Nieuw AmerongeivA.V. (1996) Human glandular salivas, their separate collection and
analysis. Eur. J. Oral ScL, 104, 346-352.
Veennan.E.C.L, Ligtenberg.A.J.M., Schenkels.L.C.P.M., WalgreenWeterings^E. and Nieuw Amerongen.A.V. (1995) Binding of human high
molecular weight salivary mucins (MG1) to Hemophilus parainfluenzae. J.
Dent Res., 74, 351-357.
Veerman.E.C.L, BolscherJ.G.M., Appelmelk3J-. Bloemena,E., van den
Berg,TX and Nieuw AmerongeruA.V. (1997) A monoclonal antibody directed against high M, salivary mucins recognizes the SO3-Gaip l-3GlcNAc
moiety of sulfo-Lewis*. A histochemical survey of human and rat tissue.
Glycobiology, 7, 37^14.
Received on October 23, 1996; revised on February 7, 1997; accepted on
February 24, 1997
743