| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
(Received for publication, November 9, 1994; and in revised form, December 30,
1994) From the
Entamoeba histolytica trophozoites initiate pathogenic
colonization by adherence to host glycoconjugates via an amebic surface
lectin which binds to galactose (Gal) and N-acetylgalactosamine (GalNAc) residues. Monovalent and
multivalent carbohydrate ligands were screened for inhibition of E.
histolytica lectin-mediated human red cell hemagglutination,
revealing that: (i) the synthetic multivalent neoglycoprotein
GalNAc
Amebiasis is a parasitic infection caused by Entamoeba
histolytica, that results in 40-50 million cases of amebic
colitis and liver abscess and 40,000-100,000 deaths annually
worldwide (1) . An initial step in parasite colonization of the
large bowel is adherence to the colonic wall. Adherence involves
recognition of colonic mucin and epithelial glycoconjugates by an
amebic cell surface lectin which recognizes nonreducing terminal Gal
and GalNAc residues (reviewed in (2) ). The E.
histolytica adherence lectin is a heterodimer consisting of a
170-kDa heavy chain and a light chain of 31 or 35 kDa linked by
disulfide bonds(3) . The heavy subunit is encoded by a gene
family of which three members (89-95% identical) from strain
HM1:IMSS have been sequenced(4) . Each contains a putative
carboxyl-terminal cytoplasmic and transmembrane domain, and a large
cysteine-rich extracellular domain which may be involved in
carbohydrate binding, since monoclonal antibodies to this domain block
amebic adherence(5) . The light subunits are nearly identical
to each other in amino acid composition and CNBr fragmentation,
although the 31-kDa subunit is unique in that it contains a
glycosylphosphatidylinositol anchor(3) . Previous studies
demonstrated that 5-10 mM galactose or N-acetylgalactosamine, but not other monosaccharides,
inhibited E. histolytica adherence to target
cells(6, 7) . The glycoproteins asialoorosomucoid or
asialofetuin, with polyvalent nonreducing terminal Gal (and GalNAc)
residues, were much more potent inhibitors than monosaccharides, with
half-maximal inhibition of amebic adherence occurring in the micromolar
range(8) . Partially purified large molecular mass (9
The standard
radioligand binding reaction contained 10 mM Hepes buffer pH
7.4, 50 mM NaCl, 2 mM CaCl Comparative
Lectin or lectin/inhibitor combinations were flowed
over the GalNAc
Figure 1:
Figure 2:
Isotherm of
Figure 3:
Effect of calcium concentration on
Figure 4:
Effect of pH on
Figure 5:
Inhibition of
The most potent inhibitor was
GalNAc
Figure 6:
Inhibition of
Porcine stomach mucin inhibited A series of commercially obtained polyvalent
carbohydrate-derivatized linear polyacrylamide polymers with Gal or
GalNAc nonreducing termini were tested as inhibitors. Each co-polymer
contained 75% N-(2-hydroxyethyl)acrylamide, 20%
sugar-derivatized acrylamide, and 5% biotin-derivatized acrylamide, and
was
Figure 7:
Inhibition of
Figure 8:
Real time association and dissociation
kinetics of purified E. histolytica lectin binding to
GalNAc
Infection by E. histolytica requires adherence of
the amoebae via interaction of a cell surface lectin with terminal
GalNAc and/or Gal residues on glycoconjugates on the surface of target
tissues (2) . In both hemagglutination assays and radioligand
binding assays, GalNAc was the preferred carbohydrate determinant when
compared to Gal, requiring one-third to one-tenth the concentration for
effective blocking of lectin-mediated binding, depending on the assay
and the aglycon ( Table 1and Table 4). Saccharides having
other than Gal and GalNAc termini were ineffective inhibitors in all
assays. Carbohydrate recognition by cell surface lectins is often
dependent on precise spatial organization of target carbohydrate
determinants(11) . The same holds true for E. histolytica lectin, as indicated by the relative inhibitory potency of
GalNAc The above pattern of
carbohydrate recognition is similar to that of the mammalian hepatic
lectin (MHL), which binds GalNAc and Gal residues (26) and
binds with high affinity to multivalent glycoconjugates such as
Gal The high affinity of
GalNAc Binding of Detailed
carbohydrate recognition studies based on inhibition of Analysis of purified
lectin binding to GalNAc Steric access to the carbohydrate binding site is an
important factor, as reflected by inhibition of The nature of the endogenous ligand for the E.
histolytica lectin on the colonic epithelium, or on other target
tissues (such as liver) is not known. It has been suggested that mucins
are either high affinity binding sites for the amoeba or may act to
inhibit infection by blocking lectin binding to cell surface sites. The
current study supports mucin as a potential target of E.
histolytica binding, since it is a likely source of multivalent
nonreducing terminal GalNAc residues. Porcine stomach mucin was the
most potent inhibitor of hemagglutination (0.12 µg/ml compared to
0.38 µg/ml for GalNAc Whatever the endogenous carbohydrates are for E. histolytica attachment which lead to establishment of colonic or hepatic
disease, the data presented here establish that: (i) multivalent
terminal GalNAc residues are potent binding ligands for the E.
histolytica lectin; and (ii) the spacing of the multivalent GalNAc
residues required for optimal E. histolytica binding is
distinct from that required for optimal binding to the mammalian
hepatic lectin. This raises the possibility of designing a specific
multivalent carbohydrate inhibitor of E. histolytica adherence.
Volume 270,
Number 10,
Issue of March 10, 1995 pp. 5164-5171
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
BSA (having an average of 39 GalNAc residues linked
to bovine serum albumin) was 140,000-fold more potent an inhibitor than
monovalent GalNAc and 500,000-fold more potent than monovalent Gal; and
(ii) small synthetic multivalent ligands which bind with high affinity
to the mammalian hepatic Gal/GalNAc lectin do not bind with high
affinity to the E. histolytica lectin. Radioligand binding
studies revealed saturable binding of 
I-GalNAcBSA to E. histolytica membranes (K
= 10 ± 3
nM, B
= 0.9 ± 0.08
pmol/mg membrane protein). Maximal binding required the presence of
calcium chloride (300 µM) or sodium chloride (50
mM), and had a broad pH maximum (pH 6-9).
GalNAcBSA was 200,000-fold more potent than monovalent
GalNAc in blocking radioligand binding. Among synthetic
saccharide-derivatized linear polymers, the GalNAc
and
GalNAc
3Gal derivatives were the most potent, with GalNAc
and GalNAc3(Fuc2)Gal derivatives much weaker. The data
support a model in which a unique pattern of spaced multiple GalNAc
residues are the highest affinity targets for the E. histolytica lectin.
10
Da) rat and human colonic mucins, which are highly
polyvalent in nonreducing terminal Gal and GalNAc residues, were very
potent inhibitors of E. histolytica adherence to target cells
(IC
< 10
M)(9, 10) . These data are
characteristic of other carbohydrate recognition systems, where
multiple appropriately spaced nonreducing terminal carbohydrates are
required for high affinity binding to cellular lectins(11) .
The current study evaluates a series of multivalent glycoconjugates
with nonreducing terminal Gal and GalNAc residues as potential
high-affinity binding ligands for the E. histolytica lectin.
One of these, GalNAcBSA, (
)was of sufficiently
high affinity to use as a ligand in binding experiments which allowed
further characterization of the binding requirements and properties of
the E. histolytica lectin.
Materials
Monosaccharides were from Pfanstiehl
(Waukegan, IL) or Sigma. Disaccharides, trisaccharides, p-nitrophenyl glycosides, fetuin, heparin, hyaluronic acid,
chondroitin sulfate, dextran sulfate, and porcine stomach mucin (Type
III) were from Sigma. Glycolipids were from Matreya (Pleasant Gap, PA)
or EY Laboratories (San Mateo, CA). Saccharide-derivatized
polyacrylamide polymers were from Synthesome, GmbH
(München, Germany). Neoglycoproteins were
synthesized by the method of Lee et al.(12) .
Synthetic glycosides and cluster glycosides were synthesized as
described
previously(13, 14, 15, 16, 17, 18, 19) .
Fetuin and porcine stomach mucin were chemically desialylated and
degalactosylated by the methods of Kim et al.(20) .
Phosphate-buffered saline (PBS) was a modification of Dulbecco's
formulation, containing 154 mM NaCl, 8.1 mM Na
HPO
, 2.7 mM KCl, 1.5 mM KHPO, 0.7 mM CaCl,
and 0.5 mM MgCl.Membrane and Lectin Preparation
E. histolytica strain HM1:IMSS trophozoites were grown as reported
previously(21) . Amoebae from 3-4 250-ml flasks were
collected by centrifugation (150 g, 15 min, 4 °C),
resuspended in 75 mM Tris, 65 mM NaCl (pH 7.2),
collected as above, and resuspended in lysis buffer containing 10
mM sodium phosphate (pH 8), 2 mM phenylmethylsulfonyl
fluoride, 5 mM EDTA, 0.1 M 4-(2-aminoethyl)-benzenesulfonyl fluoride, and 2 mMp-hydroxymercuribenzoic acid. After incubation for 5 min
at 37 °C to promote lysis, the suspension was chilled on ice and
sonicated (3-5 10-s bursts, microtip, maximum energy, Sonifier
Cell Disruptor). Membranes were collected by centrifugation (50,000
g, 1 h, 4 °C). The pellets were resuspended in 10
ml of lysis buffer and membranes collected by centrifugation (100,000
g, 1 h, 4 °C). The resulting pellets were stored
for up to 48 h on ice, then were resuspended in 10 ml of ice-cold 10
mM sodium phosphate buffer (pH 7.4) containing 2 mM phenylmethylsulfonyl fluoride, 2 mM EDTA, and 1 mMp-hydroxymercuribenzoic acid and subjected to further
sonication (10 10-s bursts over a 20-min period). Membranes were
collected by centrifugation as above and resuspended in 10 mM Hepes buffer (pH 7.4) containing 1 mg/ml bovine serum albumin
(BSA). Aliquots of the membrane preparation were stored at -20
°C until use. E. histolytica lectin was purified as
described previously(21) . Rat liver plasma membranes were
prepared by the method of Ray(22) .Hemagglutination
Hemagglutination was performed as
described previously(21) . Human Type A blood was drawn into
heparinized tubes, diluted 1:1 with PBS, and 4 ml pipetted over 3 ml of
Ficoll-Paque (Pharmacia Biotech Inc.) in a centrifuge tube.
Erythrocytes were collected by centrifugation at 400 g for 25 min at 4 °C, washed twice in 10 ml of PBS by
centrifugation at 400 g for 5 min at 4 °C, and
diluted to 2.4 10
/ml in PBS for storage at 0 °C
for up to 30 days. For use in the hemagglutination assay, a suspension
of 2 10
cells/ml in PBS containing 2 mg/ml bovine
serum albumin was prepared. Hemagglutination was performed in V-bottom
96-well polystyrene plates (Costar Seroclusters catalog No. 3897). Test
samples containing E. histolytica membranes (
4 µg of
membrane protein) in 100 µl of PBS were placed in replicate
microwells and hemagglutination was initiated by adding 100 µl of
erythrocyte suspension. After gentle agitation, plates were incubated
undisturbed at 4 °C for 6-16 h, after which hemagglutination
was detected as a uniform distribution of erythrocytes in the well,
rather than a point concentration of cells at the well bottom in the
absence of hemagglutination. A membrane concentration curve was
performed for each E. histolytica membrane preparation, and a
membrane concentration 2-fold of that necessary to detect
hemagglutination was used for glycoconjugate inhibition experiments.Radioligand-Membrane
Binding
GalNAcBSA was radioiodinated using
carrier-free NaI and Iodo-Beads (Pierce Chemical Co.),
and the radiolabeled product was purified by gel filtration
chromatography on Sephadex G-25 (Pharmacia). For most experiments, the
specific activity ranged from 250 to 500 µCi/nmol., 5 mg/ml BSA,
and the indicated concentrations of I-GalNAcBSA, E. histolytica membranes, and any potential saccharide inhibitor in a total
volume of 100 µl. Reactions were incubated with gentle agitation
for 4-5 h at 4 °C, then rapidly diluted with 10 mM Hepes buffer pH 7.4 and rinsed onto glass fiber filters (presoaked
in 10 mM Hepes buffer pH 7.4, 1 mg/ml BSA) using a Brandel
Harvester (Brandel, Gaithersburg, MD). Membrane-bound radioligand
remaining on the glass fiber filters was quantitated using a
-radiation counter.I-GalNAcBSA binding to rat liver membranes
was based on the method of Grant and Kaderbhai(23) . The
binding reaction contained 25 mM Tris-HCl buffer (pH 7.8), 150
mM NaCl, 0.02% (w/v) Triton X-100, 20 mM CaCl, 10 mg/ml BSA, 400 pM radioligand, 5
µg of rat liver membrane protein, and potential saccharide
inhibitors. After incubation for 110 min at 4 °C, membranes with
bound ligand were collected on glass fiber filters (as above), washed
with buffer (25 mM Tris-HCl buffer (pH 7.8), 150 mM NaCl, 0.02% (w/v) Triton X-100, 20 mM CaCl),
and radioligand remaining on the filters was quantitated.Biosensor-based Lectin Binding
Real time detection
of E. histolytica lectin binding to GalNAcBSA was
performed using a BIAcore (Pharmacia) biosensor-based analytical system (24) . GalNAcBSA was covalently coupled, via its
primary amines, to a derivatized carboxyl group within a dextran layer
on the sensor chip surface. A 35-µl aliquot containing N-hydroxysuccinimide and
1-ethyl-3-(3-dimethylaminopropyl)carbodiimide was flowed across the
sensor chip under conditions designed to activate 30-40% of the
immobilized carboxyl groups using the manufacturer's reagents and
conditions. This was followed by 40 µl of GalNAcBSA
(45-60 µg/ml) in 10 mM Hepes buffer (pH 6.0)
injected at the same rate. Immobilization of GalNAcBSA was
performed at 25 °C using a flow rate of 5 µl/min, and
resulted in an initial reading of 4000 relative units corresponding
to 500 µM GalNAcBSA within derivatized
dextran layer. Any remaining activated carboxyls were reacted with
ethanolamine hydrochloride as described(24) . The derivatized
sensor was washed by injection of 15 µl of 10 mM Hepes,
150 mM NaCl, 3.4 mM EDTA, and 0.005% surfactant P20
prior to use.BSA-derivatized surface at 25 °C in
freshly prepared running buffer containing 10 mM Hepes, 2
mM CaCl, 50 mM NaCl, and 0.005% P20
surfactant. Initial experiments were performed to determine the
concentrations which would allow measurable binding. On an immobilized
base of 450-500 µM GalNAcBSA,
measurable binding was obtained using 300-600 nM soluble E. histolytica lectin. To determine association kinetics, 40
µl of purified E. histolytica lectin was injected at a
flow rate of 5 µl/min, and data points were collected at 5-s
intervals for 8 min. At the end of each run, the surface was
regenerated using 50 mM HCl. Three graded concentrations of
soluble lectin (288 nM, 385 nM, and 578 nM)
were used. Dissociation curves were obtained using a flow rate of 60
µl/min and an injection of 120 µl of lectin or lectin/inhibitor
combinations (association phase) followed by running buffer only, at
the same flow rate (dissociation phase). Binding was observed over a
period of 17 min, with association occurring during the first 2 min and
dissociation during the following 15 min. The curves suggested a
biphasic dissociation, so were analyzed using a two-site model. Kinetic
data analyses were performed based on published principles (24) using BIAcore software.
Saccharide Inhibition of E. histolytica Lectin-induced
Hemagglutination
An initial screen for saccharides with high
affinity for the E. histolytica membrane lectin was initiated
using human red blood cell agglutination as a rapid assay for
inhibitory carbohydrates and glycosides. Consistent with previous
studies, saccharides with nonreducing terminal Gal or GalNAc residues
inhibited hemagglutination, while similar saccharides with other
nonreducing termini (Glc, GlcNAc, Man, etc.) were noninhibitory. Two
novel results arose from these studies (Table 1Table 2).
Small molecular weight divalent or trivalent ``cluster
glycosides'' of Gal or GalNAc(18, 19) , which are
high affinity inhibitors of binding to the mammalian hepatic lectin
(IC < 1 nM) were relatively poor inhibitors of
the E. histolytica lectin (minimum inhibitory concentration
0.1-1 mM). In contrast, polyvalent Gal- and
GalNAc-derivatized neoglycoproteins were very high affinity inhibitors
of the E. histolytica lectin, with GalNAcBSA
having a minimum inhibitory concentration of 5 nM, which was
>150,000-fold more potent than GalNAc or lactose for inhibition of E. histolytica membrane-induced hemagglutination. The
monosaccharide GalNAc was modestly (4-fold) more potent than Gal in
inhibiting hemagglutination, whereas GalNAcBSA was 33-fold
more potent than Gal
BSA in the same assay, suggesting that
appropriately spaced GalNAc residues are the most potent saccharide
ligands for the E. histolytica lectin.
High Affinity Binding of GalNAc
Based on the initial hemagglutination
screen, GalNAcBSA to E.
histolytica MembranesBSA was chosen as a potential ligand to
further probe the E. histolytica lectin.
GalNAcBSA was radioiodinated and incubated with E.
histolytica membranes, then the membranes were collected by
filtration. At low input radioligand concentrations (1 nM),
specific binding was readily demonstrated (Fig. 1). Specific
binding increased linearly as a function of added membrane protein,
while nonspecific binding (in the presence of 300 nM unlabeled
GalNAcBSA) was relatively low and did not increase
markedly with added membranes. Specific binding was saturable, while
nonspecific binding was characteristically nonsaturable, such that at
higher input radioligand concentrations (>10 nM)
nonspecific binding was a substantial fraction of total binding (Fig. 2A). Binding was sufficiently consistent to
generate an equilibrium binding isotherm. Although binding of the
multivalent ligand to a multivalent membrane is inherently complex (see
``Discussion''), the binding data fit a single-class
receptor-ligand binding equation (Fig. 2B), resulting
in an apparent K
of 9.7 ± 3.1 nM and a B of 0.93 ± 0.08 pmol/mg
membrane protein.
I-GalNAcBSA
binding as a function of E. histolytica membrane
concentration. Radioligand binding was performed under standard
conditions using the indicated concentrations of membrane protein per
assay and a radioligand concentration of 1.1 nM (9.2 Bq/fmol).
Specific binding (filled circles) is the difference between
total binding (open circles) and background binding (squares), the latter measured in the presence of 300 nM unlabeled GalNAcBSA. Data are presented as mean
± S.D. for duplicate determinations.
I-GalNAcBSA binding to E. histolytica membranes. Radioligand binding was performed under standard
conditions using 110 µg of E. histolytica membrane protein
per assay and the indicated concentrations of I-GalNAcBSA (9.8 Bq/fmol). A, total
binding, squares; background binding (in the presence of 300
nM unlabeled GalNAcBSA), circles. B, specific binding was calculated by subtracting background binding
from total binding. The line represents the least squares fit
of the data to a rectangular hyperbola, from which the K (9.7 ± 3.1 nM) and B (102 ± 8.8 fmol/assay) were calculated.
Data are presented as mean for triplicate determinations. Standard
deviations fall within the data points in all cases, so are not
indicated.
I-GalNAcBSA binding to E. histolytica membranes was highly dependent on ionic
conditions. Initial binding studies were performed in PBS, the same
buffer used in hemagglutination experiments. In comparison, binding was
essentially absent in a base buffer of 10 mM Hepes pH 7.4 (Table 3). Binding was largely recovered if either calcium
chloride (0.7 mM) or sodium chloride (154 mM) was
added to the base buffer, while addition of magnesium chloride (0.5
mM) resulted in recovery of less than a third of the specific
binding activity found using PBS. The same maximum level of
GalNAcBSA binding was achieved whether sodium chloride,
calcium chloride, or a combination of the two were used (Table 3). Maximum GalNAcBSA binding required
50-100 mM sodium chloride, with 20 mM NaCl
supporting half-maximal binding (data not shown). In contrast, a free
calcium ion concentration of <300 µM supported maximum
binding; half-maximal binding occurred at 35 µM free
calcium ion (Fig. 3). The pH dependence of I-GalNAcBSA binding was broad, with similar
binding over the pH range 6-9 (Fig. 4).
I-GalNAcBSA binding to E. histolytica membranes. Radioligand binding was performed under standard
conditions, except the binding buffer was 10 mM Hepes pH 7.4
supplemented with 1 mg/ml BSA and various calcium concentrations
ranging from 42 to 840 µM. The introduction of EDTA (from
the membrane preparation) at a constant concentration of 200 µM resulted in the indicated (calculated) free calcium
concentrations. Each assay contained 2.8 nM radioligand (1.8
Bq/fmol) and 40 µg of E. histolytica membrane protein.
Specific binding (filled circles) is the difference between
total binding (open circles) and background binding (squares), the latter measured in the presence of 300 nM unlabeled GalNAcBSA. Data are presented as mean
± S.D. for duplicate determinations.
I-GalNAcBSA binding to E. histolytica membranes. Radioligand binding was performed under standard
conditions except that the binding buffer consisted of 10 mM of sodium acetate (pH 4-5), Mes (pH 6), Hepes (pH
7-8), or Tris (pH 9). Each buffer also contained 1 mM CaCl and 150 mM NaCl. Each assay contained
2.8 nM radioligand (2.1 Bq/fmol) and 40 µg of E.
histolytica membrane protein. Specific binding (filled
circles) is the difference between total binding (open
circles) and background binding (squares), the latter
measured in the presence of 300 nM unlabeled
GalNAcBSA. Data are presented as mean ± S.D. for
duplicate determinations.
Saccharide Inhibition of GalNAc
The affinity of various saccharides
for the E. histolytica lectin was measured indirectly by their
ability to inhibit BSA Binding
to E. histolytica MembranesI-GalNAcBSA binding to E. histolytica membranes. Small saccharides having a
nonreducing terminal Gal or GalNAc inhibited half-maximally in the
mM range ( Fig. 5and Table 4). The presence of a
hydrophobic aglycon on p-nitrophenyl
-N-acetylgalactosaminide increased affinity 10-fold. Both
and p-nitrophenyl glycosides of GalNAc were
inhibitory, with the glycoside having modestly higher affinity (Table 4).
I-GalNAcBSA binding to E. histolytica membranes by saccharides. Radioligand binding was performed under
standard conditions using 2.2 nMI-GalNAcBSA (6.9 Bq/fmol) and 32 µg
of E. histolytica membrane protein. Assays were performed in
the presence of the indicated concentrations of the following
inhibitors: p-nitrophenyl GalNAc (filled circles),
GalNAc (filled squares), lactose (open squares), or
galactose (open circles). Data are presented as mean ±
S.D. for duplicate determinations. Horizontal lines represent
the mean (solid) and S.D. (dotted) of binding in the
absence of inhibitors (upper) or in the presence of 300 nM unlabeled GalNAcBSA (lower).
BSA itself, with half-maximal inhibition at 5 nM (Fig. 6), in good agreement with equilibrium binding data
and with inhibition of hemagglutination. GalBSA was
>100-fold less potent than GalNAcBSA, while the
glycoprotein fetuin and its asialo- and asialoagalacto-derivatives did
not inhibit radioligand binding at concentrations exceeding 10
µM.
I-GalNAcBSA binding to E. histolytica membranes by glycoproteins. Radioligand binding was performed
under standard conditions using 1.1 nMI-GalNAcBSA (8.2 Bq/fmol) and 55 µg
of E. histolytica membrane protein. Assays were performed in
the presence of the indicated concentrations of the following
inhibitors: GalNAcBSA (filled squares),
GalBSA (filled circles), asialofetuin (open
squares), fetuin (open circles), or asialoagalactofetuin (triangles). Data are presented as mean ± S.E. for
triplicate determinations. Horizontal lines represent the mean (solid) and S.E. (dotted) of binding in the absence
of inhibitors.
I-GalNAcBSA binding half-maximally at 2
µg/ml, 5-fold less potent than GalNAcBSA on a weight
basis. Asialo-porcine stomach mucin was slightly less potent than the
parent molecule, and subsequent chemical degalactosylation reduced
inhibition potency >25-fold compared to the parent mucin (Table 4). Hyaluronoic acid inhibited radioligand binding
half-maximally at 10 µg/ml (25-fold less potent than
GalNAcBSA), while chondroitin sulfate and heparin were
only weak inhibitors.30 kDa by gel filtration(25) . The -GalNAc
derivative was a potent inhibitor of I-GalNAcBSA binding to E. histolytica membranes, with an IC of 11 µg/ml, corresponding
to 3.1 µM in sugar residues or 400 nM in
polymer chains (Fig. 7). The -GalNAc derivative was much
less potent, resulting in <50% inhibition at 500 µg/ml. Polymers
bearing glycosides of Gal, Glc, and GlcNAc (all ananomers) were
very weak inhibitors. Among the disaccharide-derivatized polymers
tested, only the GalNAc 1,3-Gal -derivatized polymer was a
potent inhibitor, with an IC of 11 µg/ml (2.5
µM in sugar residues, 370 nM in polymer
chains). The closely related GalNAc 1,3-GalNAc derivative
was a poor inhibitor, as were Gal-terminated derivatives. Polymers
bearing the blood group A and B trisaccharides were weak inhibitors,
with the more potent A-group (GalNAc1,3(Fuc1,2)Gal)
polymer causing half-maximal inhibition at 250 µg/ml (44 µM in sugar residues, >8 µM in polymer chains).
I-GalNAcBSA binding to E. histolytica membranes by multivalent saccharide-derivatized linear
polyacrylamide. Radioligand binding was performed under standard
conditions using 0.88 nMI-GalNAcBSA (15 Bq/fmol) and 29 µg
of E. histolytica membrane protein. Assays were performed in
the presence of the indicated concentrations of polymers derivatized
with the following glycosides (see text for description of polymers). Upper panel, GalNAcpropyl, filled circles;
GalNAcpropyl, open squares; Galpropyl, open
triangles; Glcpropyl, closed squares; and
GlcNAcpropyl, closed triangles. Middle panel, GalNAc3Galpropyl, closed circles;
GalNAc3GalNAcpropyl, open circles;
Gal3GalNAcpropyl, triangles; and
Gal3Galpropyl, squares. Lower panel: GalNAc3(Fuc2)Galpropyl, squares; and
Gal3(Fuc2)Galpropyl, circles. Data are
presented as mean ± S.E. for duplicate
determinations.
Binding of Purified E. histolytica Lectin to
GalNAc
Direct binding of the purified E.
histolytica lectin to GalNAcBSABSA was demonstrated
using a BIAcore biosensor-based analytical system.
GalNAcBSA was covalently immobilized via amide linkage
between the neoglycoprotein (which retains 20 free lysine
-amino groups per molecule) and carboxylated dextran on the sensor
surface. When purified E. histolytica lectin was introduced
into the sensor flow cell, binding was detected (Fig. 8, top). The association and dissociation rate constants were
obtained directly from the relative response plots(24) . The
calculated k
, based on analysis of kinetics using
three different lectin concentrations, was 3.2 10M
s. Dissociation (Fig. 8, bottom) was analyzed according to a two-site
model, resulting in dissociation rates of 2.5 10
s (rapid phase, 13% of bound lectin) and 3.2
10
s (slow phase). This
results in K values of 9.8
10M for the larger fraction of the bound
lectin (slow dissociating) and 7.9 10
M for the smaller fraction (rapid dissociating). Addition of 35
µM asialofetuin to the purified lectin prior to
introduction into the sensor flow cell resulted in a marked decrease in
the k and an increase in the k
(Fig. 8), suggesting that binding of asialofetuin to the
lectin inhibited its subsequent association with the immobilized
GalNAcBSA on the sensor chip. A decrease in the k
and an increase in the k
were also
detected when higher concentrations of low molecular weight inhibitors
were added to the lectin prior to injection (data not shown).
BSA. Top, association kinetics of lectin
(580 nM) binding to immobilized GalNAcBSA (460
µM) in the absence of inhibitors is represented by circles; association of lectin (580 nM) and ligand
(490 µM) in the presence of asialofetuin (35
µM) is represented by squares. Bottom, dissociation kinetics after binding of lectin (580 nM) to
immobilized ligand (760 µM) in the absence of inhibitor is
represented by circles; dissociation of lectin (390
nM) binding to ligand (450 µM) in the presence of
asialofetuin (35 µM) is represented by squares. Zero time in the bottom panel represents the time of peak
binding after which lectin-free running buffer was injected and
dissociation was initiated.
BSA compared to GalNAc. The polyvalent
neoglycoprotein was 150,000-200,000-fold more potent than the
monosaccharide, depending on the assay ( Table 1and Table 4). This corresponds to a 3,500- to 5,000-fold increase in
binding affinity per GalNAc residue, apparently due to appropriate
multivalent spacing. Bovine serum albumin has 60 lysine residues, most
of which are available for saccharide derivatization. Prior studies on
vertebrate lectins suggest that BSA-based multivalent conjugates
represent a variety of multivalent saccharide spacings and thus have a
high potential for mimicking natural ligands.BSA and GalNAcBSA (Table 1).
However, several small synthetic multivalent neoglycoconjugates which
are high affinity probes for the MHL (19) were poor inhibitors
of the E. histolytica lectin. For example, the synthetic
trivalent glycoconjugate YEE(GalNAc)
has a 20,000-fold
higher affinity for the MHL than the corresponding monosaccharide, or
7,000-fold higher based on the concentration of GalNAc termini. In
contrast, it has equal affinity to the monosaccharide, per GalNAc
residue, for the E. histolytica lectin. These data indicate
that high potency polyvalent binding of Gal/GalNAc residues is a key
factor in binding of both lectins to their target glycoconjugates, but
that there are fundamental differences in the precise spacing required
to maximize that binding. Presumably, the inherently heterogeneous
neoglycoproteins (such as GalNAcBSA) can satisfy the
spacing requirements for both mammalian and amebic lectins. The
apparent requirement of the E. histolytica lectin for more
widely spaced ligands than those required for the MHL is analogous to
the requirements of the mannose-binding proteins(27) , for
which the term ``macro-cluster'' was coined to discern wide
spacings from the ``mini-clusters,'' such as those found on
the amino acid-based scaffolds used in the current study. This
observation raises the possibility that by obtaining information on the
spacing of GalNAc binding sites on E. histolytica membranes,
novel and specific high affinity inhibitors of E. histolytica adherence could be designed which would interact poorly with the
MHL. To date such an amebic-selective saccharide ligand has not been
found. Finding better defined and more selective high affinity ligands
will require screening of other synthetic multivalent glycoconjugates,
and/or additional information about the quaternary structure of the
lectin binding sites on the amoeba surface.BSA for the E. histolytica lectin allowed
development of a rapid radioligand binding assay, and further
exploration of the lectin's binding characteristics. The ionic
requirements for binding were unusual, in that binding was not
supported in 10 mM sodium Hepes buffer, but required the
presence of either a low concentration of calcium
(half-maximal binding at 35 µM) or a higher concentration
of sodium chloride (half-maximal binding at 20 mM NaCl). The
observation that the same level of radioligand binding was supported
when NaCl, CaCl, or both were added (Table 3)
suggests that a single lectin has a high affinity site for calcium
ions, the functional requirement for which can be overcome by higher
concentrations of monovalent ions. Current structural data on the
lectin is of insufficient detail to suggest a mechanism for this
observation.I-GalNAcBSA to E. histolytica membranes using a filter binding assay resulted
in well behaved data fitting a single site binding model, even though
the multivalent radioligand was binding to multivalent lectins in the
membranes. This type of data has been reported for binding of
neoglycoproteins to certain vertebrate lectins(28) , and would
be expected if binding was essentially an ``all or nothing''
event with no intermediate binding events or affinities.I-GalNAcBSA binding to E. histolytica membranes (Table 4) generally correlated with the less
precise data obtained from hemagglutination inhibition (Table 1).
Exceptions included the glycoprotein fetuin and porcine stomach mucin,
and their chemically modified asialo- and asialoagalacto- forms, which
were 100-fold more potent at blocking hemagglutination than
blocking I-GalNAcBSA binding mediated by the
same E. histolytica membranes. These data do not fit a simple
binding model for lectin and glycoconjugate targets, but may reflect
the high cooperativity involved in binding to I-GalNAcBSA. The difference in inhibitory
potencies may be due to steric exclusion of the extended fetuin or
mucin structures from the neoglycoprotein binding site. Kinetics may
also be involved, in that binding of the soluble I-GalNAcBSA is expected to be more rapid
than binding between cells and membranes. Since off-rates of
multivalent ligands are characteristically slow (Fig. 8), the
system may not reach true equilibrium in the time of the experiment,
and the relative binding affinities may reflect, in part, the relative
binding rates of the ligands and inhibitors.BSA was readily demonstrated using
a BIAcore biosensor-based analytical system. Binding of purified lectin
to GalNAcBSA, in addition to the membrane binding, lends
strong credence to the concept of targeting adhesion factors to
microbial surfaces using synthetic ligands. Using the BIAcore system, K values of purified lectin binding to
GalNAcBSA approached those of membrane binding. However,
inhibition of binding of the lectin to immobilized
GalNAcBSA in the BIAcore by asialofetuin was achieved at
concentrations less than one-tenth of those needed for inhibition of
membrane binding. This may be due to kinetic factors, since in the
BIAcore system the lectin and inhibitor are premixed, then flowed past
the immobilized GalNAcBSA under nonequilibrium conditions.
The rapid dissociation from the immobilized ligand on the sensor in the
presence of inhibitor suggests that polyvalent binding was reduced by
the inhibitor.I-GalNAcBSA binding to E. histolytica membranes by multivalent carbohydrate-derivatized polyacrylamide
polymers (Fig. 7). The anomer of GalNAc linked directly to
the spacer arm on the polymer was >40-fold more potent at blocking
binding than the corresponding anomer, while an anomer of
GalNAc spaced away from the polymer backbone (GalNAc1,3Gal)
was equipotent to the GalNAc-derivatized polymer. These data may
reflect a preference for the GalNAc-Gal disaccharide, but may also
suggest that the anomer close to the polymer backbone may have
critical hydroxyls which are unavailable for binding due to steric
hindrance. In this light, it is interesting to note that a polymer
containing the A-trisaccharide, which differs from
GalNAc1,3Gal by the additional presence of a fucose residue
on the 2-position of the Gal was >20-fold less potent than the
corresponding disaccharide polymer, suggesting that the fucose
restricts access of the lectin to key GalNAc hydroxyls. A more complete
understanding of the requirements for binding to this lectin will
require additional structural data on the lectin's carbohydrate
binding domain.BSA, Table 1). Since the
preparation is heterogeneous in structure and molecular weight, there
may be substructures with terminal saccharide orientations which have
very high affinity for lectin binding. These structures are likely to
consist of spaced GalNAc residues on the polypeptide backbone, since
desialylation and degalactosylation had little effect on potency (Table 1). This conclusion is supported by data using fetuin
preparations as inhibitors (Table 1). The commercial preparation
of fetuin had 50% desialylated Gal termini based on its previously
published oligosaccharide structures and carbohydrate analysis (data
not shown). Further desialylation did not change its inhibitory
potency, but degalactosylation enhanced inhibition. Fetuin has
three N-linked complex oligosaccharides which would terminate
in noninhibitory GlcNAc residues after degalactosylation and three O-linked oligosaccharides, comprised primarily of
NeuAc2,3Gal1,3GalNAc, which would be converted to GalNAc
termini upon degalactosylation. Based on the other carbohydrate
specificity data presented, it is likely that the spaced GalNAc
residues on fetuin result in the enhanced inhibitory potency.
)
We are grateful to Lauren Ashley for preparation of E. histolytica membranes.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  P. Magnelli, J. F. Cipollo, D. M. Ratner, J. Cui, D. Kelleher, R. Gilmore, C. E. Costello, P. W. Robbins, and J. Samuelson Unique Asn-linked Oligosaccharides of the Human Pathogen Entamoeba histolytica J. Biol. Chem., June 27, 2008; 283(26): 18355 - 18364. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Wildschutte and J. G. Lawrence Differential Salmonella survival against communities of intestinal amoebae Microbiology, June 1, 2007; 153(6): 1781 - 1789. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Frederick and W. A. Petri Jr. Roles for the galactose-/N-acetylgalactosamine-binding lectin of Entamoeba in parasite virulence and differentiation Glycobiology, December 1, 2005; 15(12): 53R - 59R. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Garate, Z. Cao, E. Bateman, and N. Panjwani Cloning and Characterization of a Novel Mannose-binding Protein of Acanthamoeba J. Biol. Chem., July 9, 2004; 279(28): 29849 - 29856. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Espinosa-Cantellano and A. Martinez-Palomo Pathogenesis of Intestinal Amebiasis: From Molecules to Disease Clin. Microbiol. Rev., April 1, 2000; 13(2): 318 - 331. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. R. Pillai, P. S. K. Wan, Y. C. W. Yau, J. I. Ravdin, and K. C. Kain The Cysteine-Rich Region of the Entamoeba histolytica Adherence Lectin (170-Kilodalton Subunit) Is Sufficient for High-Affinity Gal/GalNAc-Specific Binding In Vitro Infect. Immun., August 1, 1999; 67(8): 3836 - 3841. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S.-i. Iida, K. Yamamoto, and T. Irimura Interaction of Human Macrophage C-type Lectin with O-Linked N-Acetylgalactosamine Residues on Mucin Glycopeptides J. Biol. Chem., April 16, 1999; 274(16): 10697 - 10705. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Liang, J. Loebach, N. Horan, M. Ge, C. Thompson, L. Yan, and D. Kahne Polyvalent binding to carbohydrates immobilized on an insoluble resin PNAS, September 30, 1997; 94(20): 10554 - 10559. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Venkataraman, B. J. Haack, S. Bondada, and Y. A. Kwaik Identification of a Gal/GalNAc Lectin in the Protozoan Hartmannella vermiformis as a Potential Receptor for Attachment and Invasion by the Legionnaires' Disease Bacterium J. Exp. Med., August 18, 1997; 186(4): 537 - 547. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. W. van Rossenberg, L. A. J. M. Sliedregt, R. Autar, C. Piperi, A. P. Van der Merwe, T. J. C. van Berkel, J. Kuiper, and E. A. L. Biessen A Structure-Function Study of Ligand Recognition by CD22beta J. Biol. Chem., April 13, 2001; 276(16): 12967 - 12973. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. S. Roseman and J. U. Baenziger The Mannose/N-Acetylgalactosamine-4-SO4 Receptor Displays Greater Specificity for Multivalent than Monovalent Ligands J. Biol. Chem., May 11, 2001; 276(20): 17052 - 17057. [Abstract] [Full Text] [PDF]
|
|
| ||||||
