- Email: [email protected]
Evidence for a family of schistosome glycan-binding lectins in Biomphalaria afexandrina
Developmental
and Comparative
Immunology,
Pergamon
Vol. 19, No. 5, pp. 365376, 1995 Else&r Science Ltd Printed in the USA 0145-30x/95 S9.50 + 0.00
EVIDENCE FOR A FAMILY OF SCHISTOSOME GLYCAN-BINDING LECTINS IN Biomphalaria alexandrina Mohamed Department
of Zoology,
H. Mansour
Faculty of Science, Cairo University, Cairo 12613, Egypt
(Submitted April 1995; Accepted June 1995) CiAbstract-A novel family of isolectins that selectively recognize a schistosome-associated fucosyllactose determinant was identified in the hemolymph of Biomphalaria alexandrina, a snail vector of Schistosoma mansoni. Three lectins of this family were purified by serial affinity chromatography on a column of L-fucose and elution with a gradient of 0.1-l M L-fucose (designated BaSII and BaSIII), followed by a column of D-glucose and elution with 0.3 MD-glucose (designated BaSI). Assessment of the structural characteristics by one- and two-dimensional gels indicated that, inspite of similarities in native molecular weights, the three lectins were tetramers of noncovalently-associated subunits that were of different sizes and pIs in BaSI, and of equal size but distinct pIs in BaSII and BaSIII. Comparisons of two-dimensional gels of the glycosylated and deglycosylated forms were consistent with the presence of an invariant a subunit (13.2 kDa, pZ 7.2) constituting the three deglycosylated lectins, which associates with other subunits unique to each lectin, namely a /3 subunit (10.1 kDa, pZ 5.8) in BaSI, an al subunit (13.2 kDa, pZ 6.8) in BaSII and BaSIII, and an a2 subunit (13.2 kDa, ~17.0) in BaSIII. Each of these subunits is subjected to differential post-translational N-linked glycosylations, which accounts for
the additional heterogeneity expressed by the glycosylated lectins. Based on miracidial glycoprotein binding and inhibition assays, the three lectins exhibited optimum binding at similar pH and temperature, but were distinct in their binding affinities towards the fucose moiety constituting the fucosyllactose target. These observations indicate that an oligomorphic family of recognition molecules may have evolved to regulate the snails’ response to schistosomes.
qKeywords-Biomphalaria akxandrina; Schistosoma mansoni; Isolectlns; Fucosyllactose determinant; Oligomorphic gene family; Lectin subunits; N-linked glycans.
Nomenclature Endo-F HA HA1 1-D 2-D SDS-PAGE SR
endo-f%N-acteylglucosaminidase F hemagglutination hemagglutination inhibition one-dimensional two-dimensional Sodium dodecyl sulfatepolyacrylamide gel electrophoresis snail Ringer
Introduction
Address correspondence to Dr Mohamed H. Mansour, Department of Zoology, Faculty of Science, Cairo University, Cairo 12613, Egypt.
Factors governing the success or failure of larval Schistosoma mansoni development within its snail intermediate host are poorly understood (l-3). There have been many reports, however, that this 365
366
dynamic host-parasite interplay involves the capacity of the snail to mount effective internal defense reactions, which in turn may be the primary targets of parasite-mediated interference (4-9). In particular, humoral recognition factors with carbohydrate-binding properties, opsonizing as well as agglutinating activities characteristic of lectins, have been implicated in mediating events following host-parasite contact (l&13). In part, this notion has been strengthened by the differential expression of these humoral factors among strains of snails that are genetically resistant or susceptible to schistosome infection (14-16). Nonetheless, information regarding the molecular characteristics of these snail-derived factors or their putative schistosome-associated oligosaccharide targets is, as yet, not fully explored (17-19). In Biomphalaria alexandrina, the snail vector of the Egyptian strain of S. mansoni, two isolectins that express selectivity in binding a schistosomeassociated fucosyllactose determinant have been recently purified and characterized (20). These lectins have been purified by affinity chromatography on a column of equimolar mixture of Dand L-glucose and sequential elution by D-glucose and L-fucose. In the present study, we have employed a different purification strategy, which resulted in the isolation of an additional lectin expressing both distinct subunit structure as well as binding affinity towards the schistosome oligosaccharide target. Assessment of the structural characteristics of the glycosylated and deglycosylated forms of the three snail-derived lectins by two-dimensional gels was consistent with their constituent subunits being the products of an oligomorphic gene family. This family of recognition molecules may have evolved to regulate the snails’ response towards selective schistosome-associated oligosaccharides.
M. H. Mansour
Materials and Methods Animals and Parasites Naturally-infected and non-infected specimens of the adult (12-18 mm shell diameter) snail B. alexandrina were supplied by the Schistosome Biological Supply Program (SBSP, Theodor Bilharz Research Institute, Cairo, Egypt). Schistosoma mansoni miracidia were prepared by allowing eggs collected from livers and intestines of infected hamsters to hatch under illuminatuion, and were supplied lyophilized and egg shell-free by SBSP.
Collection of B. alexandrina Hemolymph Hemolymph was collected by severely touching the foot of the snail with a Pasteur pipette. As a result, the snail retracts into the shell and exudes 100-250 uL of hemolymph (20). Aliquots collected from individual snails (a total of 4 x 103) were mixed with an equal volume of snail Ringer (SR, 0.2 A4 Tris-HCl, pH 8.0 containing 2.5% NaCl, 0.15% KCl, 0.4% MgC&, 0.65% CaC12), pooled, freed of hemocytes by centrifugation at 1500 rpm for 10 min at 4°C and used immediately or stored at -20°C.
Preparation of S. mansoni Miracidial Glycoprotein-bearing Fucosyllactose Determinants Lyophilized miracidia (2 mL pellet) were solubilized in 5 mL of lysis buffer (20 mM Tris-HCI, pH 8.0, 150 mM NaCl, 2% deoxycholate, 2 mA4 phenylmethylsulfonyl fluoride, 1 mM o-phenanthroline, 1 mM p-chloromercuribenzoic acid and 1 mM iodoacetamine) by sonic treatment followed by three cycles of freezing and thawing. The solubilized protein fraction was separated by centrifugation at 100,000 x g for 1 h and diluted
Biomphalaria alexandrina family of isolectins
10 fold with 20 mM Tris-HCl, pH 8.0. A fraction of the miracidial glycoproteins expressing a fucosyllactose determinant was prepared by affinity chromatography of the miracidial extract on an immobilized B. alexandrina lectin column. For the preparation of this column, a purified fraction of the snail lectin specific to the fucosyllactose schistosome-associated determinant (BaSII) was prepared from the hemolymph by affinity chromatography on a column of equimolar mixture of D- and L-glucose and elution by 0.3 M L-fucose as described previously (20). Purified BaSII (2 mg) was desalted and separated from monosaccharides by chromatography on a column of Bio Gel PlO (1.5 x 20 cm) and coupled to CNBractivated Sepharose 4B (Pharmacia Fine Chemicals, Sweden) at a concentration of 2 mg/mL beads following the manufacturer’s instructions. The coupled gel (1 mL) was equilibrated in lysis buffer containing 0.2% deoxycholate (deoxycholate buffer) and the miracidial extract cycled through the column 6 times at 15°C. Bound glycoproteins were eluted by 0.3 M L-fucose and desalted and separated from the monosaccharide by chromatography on a column of Bio Gel PlO (1.5 x 20 cm) in deoxycholate buffer. Protein was determined by the method of Lowry et al. (21) using bovine serum albumin (BSA) in deoxycholate buffer as a standard.
Target Cells Sheep erythrocytes (SRBCs, VACSERA, Cairo, Egypt) were formalin fixed, tanned and coated with BSA or fucosyllactose determinant-bearing miracidial glycoproteins according to standard procedures (22) and used as target cells in all assays. Coatings were done at the ratio of 1 mL packed SRBCs to 3 volumes of the glycoprotein solution (100 ug/mL) in deoxycholate buffer. Prior to use, coated SRBCs were washed three times in SR,
367
pH 8.0 and resuspended buffer as 1.5% suspension.
in the same
Indirect Hemagglutination and Hemagglutination Inhibition Assays Hemagglutination (HA) assays were performed in microtiter plates at room temperature. Snail hemolymph, postcolumn fractions or purified lectins (25 uL) were serially diluted in SR containing 0.05% BSA and mixed with 25 uL of a 1.5% schistosome glycoprotein-coated SRBCs suspension. HA titers were scored microscopically as the reciprocal of the highest dilution of hemolymph (or purified lectins) giving complete agglutination in 1 h. Assays were performed in triplicate and controlled by the substitution of hemolymph or lectin by SR, and by using BSA-coated SRBCs as test cells. Hemagglutination inhibition OIAI) assays were conducted by mixing serially diluted hemolymph or pure lectins (25 uL) with 25 uL of mono-, di- or oligosaccharides (all sugars were purchased from Sigma Chem. Comp., St. Louis, MO, except 2’-fucosyllactose, lacto-Nfucopentose I and III and lacto-N-tetrose, which were from Calbiochem Corp., San Diego, CA). Sugars were dissolved in SR (at concentrations up to 100 m&f’) and the reaction mixture incubated for 1 h at room temperature followed by the addition of 25 uL of a 1.5% suspension of schistosome glycoprotein-coated SRBCs. Controls were the substitutions of hemolymph or pure lectin by SR and by the substitution of the inhibition solution by SR. Fifty percent inhibition values were interpolated rom plots of percent inhibition versus inhibitor concentration.
Serial Afinity Chromatography with Immobilized Monosaccharide Columns Snail hemolymph in SR (900 mL) was chromatographed on a 0.7 x 14 cm
368
column containing L-fucose-agarose (Lfucose attached by a five carbon spacer, Sigma Chemical Comp., St. Louis, MO). Before use, the column was equilibrated with SR and then the pooled hemolymph was recycled through it three times at 5 mL/h at 15°C. The column was extensively washed with SR followed by SR containing 0.5 A4 NaCl, and specifically adsorbed lectins were eluted at room temperature with a gradient of 0.1-l M L-fucose in SR containing 0.5 M NaCl. The material that passed unretarded through the fucose column was pooled and dialyzed against SR and then recycled three times at 5 mL/h through a 0.7 x 14 cm column containing D-glucose-agarose (D-glucose attached by a five carbon spacer, Sigma Chemical Comp., St. Louis, MO). The column was washed with SR followed by SR containing 0.5 M NaCl, and specifically adsorbed lectins were eluted with SR containing 0.3 M Dglucose and 0.5 M NaCl. To attain maximum stability, specifically eluted fractions from either columns were dialyzed against SR, aliquoted, freeze-dried and stored at -20°C. Anti-schistosome binding activities were monitored during chromatography by HA and HA1 assays miracidial glycoprotein-coated using SRBCs (25 pL of 1.5% coated SRBCs/ well) as indicated above. Column fractions were continuously monitored for protein by absorbance at 280 nm. A unit of activity was defined as the minimum concentration of protein required to give one well of HA.
pH Dependence and Temperature Stability of PuriJied Snail Lectins The pH range over which the purified snail lectins exhibit optimal binding to schistosomal glycoproteins was examined in HA assays by titrating purified lectins in the following buffers: 100 mM sodium acetate/acetic acid (pH 3.7-5.4), 100 mA4 sodium cacodylate/HCl (pH [email protected]), 100
M. H. Mansour
mM Tris-HCl (pH 7.5-8.8), 100 mM glycine/NaOH @H 9.0-10.6). All buffers contained 10 mM CaClz and 10 mM MgC12 and test cells were miracidial glycoprotein-coated SRBCs. The temperature stability of the purified lectins was estimated by incubating 100 uL aliquots of the lectins in SR at different temperatures for 30 min, cooling them to room temperature and titrating them with miracidial glycoprotein-coated SRBCs in HA assays.
One- and Two-dimensional Polyacrylamide Gel Electrophoresis (1-D and 2-D SDS-PAGE) 1-D SDS-PAGE was conducted under non-reducing and reducing conditions using the discontinuous buffer of Laemmli (23). 2-D SDS-PAGE was conducted essentially as described by O’Farrell (24). Lyophilized samples of glycosylated and deglycosylated snail lectins were dissolved in lysis buffer, incubated at 37°C for 4 h, and applied to the isoelectric focusing gels for 18 h (8000 Vh). Samples were resolved in the second dimension with 12.5% slab SDS-PAGE. Samples were visualized by silver staining as described by Wray et al. (25) using BioRad low molecular weight standards and the gels photographed wet using Kodak Tri-X-Pan films.
Glycosidase Treatments Purified fractions of snail lectins (25 pg) were reconstituted in 50 uL of 100 mM sodium phosphate, pH 6.1, 50 mM EDTA, 1% Nonidet P-40 containing 100 mU of Endo-F (Endo-B-N-acetylglucosaminidase F, from Flavobacterium meningosepticum, 600 U/mg, Sigma Chem. Comp. St., Louis, MO). Samples were incubated for 18 h at 37°C precipitated with an equal volume of 20% trichloroacetic acid, washed with cold acetone and
Biorrtphalaria alexandrinafamily of isolectins
369
was resolved by 2-D SDS-PAGE into four equal-sized subunits of 15.6 kDa expressing distinct patterns of charge microheterogeneity in each eluate. The ioselectric points of the individual charge entities constituting the 0.43 M r_-fucose eluate were identical to those previously Results estimated for the lectin isolated on a D-/Lglucose column by elution with r_-fucose Isolation of Three Lectins (pls 5.6, 5.8, 6.2 and 6.4) (20) and was Constituting the Anti-schistosome Binding Activity in B. alexandrina hence given the same designation, BaSII. The 0.55 A4 L-fucose eluate, on the other Hemolymph hand, was constituted by four distinct charge varients of pIs 6.4,6.8,7.0 and 7.2, In a previous report, we have demonstrated the high affinity of B. alexandrina and was designated BaSIII. Comparison hemolymph for L-fucosyl and D-glucosyl of the total activities in the cell-free hemolymph, BaSII, BaSIII and the runresidues constituting a fucosyllactose through fraction indicate that about 9% determinant associated with schistosomes (20). This activity has been resolved into of the loaded activity could not be two lectins, designated BasI and BasII, recovered from the r_-fucose column when purification was attempted by affin- under the employed conditions, perhaps due to high binding affinities. ity chromatography on an immobilized The material that passed unretarded equimolar mixture of D- and L-glucose and sequential elution with D-glucose and through the fucose column was constiL-fucose. Under these conditions, part of tuted by about 24% of the initial aggluthe anti-fucosyl binding activity could tinating activity in the hemolymph. When applied to a D-glucose column, over 87% not, however, be recovered. In the present of this activity was bound and readily study, serial affinity chromatography using a column of L-fucose followed by eluted on addition of SR containing 0.3 a column of D-glucose was employed as M D-glucose and 0.5 M NaCl (Fig. 1 and an alternative purification strategy. As Table 1). This fraction was constituted by shown in Figure 1 and Table 1, a fraction a single component of 67 kDa and constituted by about 55% of the loaded resolved by 2-D SDS-PAGE into four activity in the hemolymph was eluted subunits of 20.4 kDa (PI 6.3), 17.6 kDa from the L-fucose column at 0.43 M L- (pf 6.0), 15.6 kDa (pls 5.7, 5.8 and 6.3) fucose with 0.1% of the protein. An and 13.5 kDa (pls 5.8 and 6.0), corresadditional 12% of the activity was yielded ponding to the a, b, c and d subunits of with 0.02% of the loaded hemolymph BaSI (20), and was hence given the same protein when the molarity of r_-fucose designation. reached 0.55 M. No further displacment of detectable activity was observed with up to 1 M L-fucose, 2 A4 NaCl or when pH Dependence and Temperature 0.2% SDS was included in the eluting Stability of BaSI, BaSII and BaSIII buffer. As judged by 1-D SDS-PAGE (Fig. l), BasI, BasII and BasIII were compared unboiled samples of both the 0.34 A4 and in terms of the pH range over which they 0.55 M L-fucose eluates were each con- exhibit optimum binding to miracidial stituted by a single 62 kDa component glycoprotein-coated SRBCs. Aliquots of when analyzed under either reducing or the purified lectins (BaSI at 52 l.tg/mL non-reducing conditions, This component BaSII at 25 pg/mL and BaSIII at 90 ug/ dried under nitrogen gas before analysis by 2-D SDS-PAGE. Control samples were similarly treated but in the absence of Endo-F.
370
M. H. Mansour
0.43NFUC
Elution
PH Vol. (ml)
J Wash
0.55YFUC
5.4
0.3MGlc
lboo Elution Vol.fml) Flgure 1. Serial affinity chromatography of 6. alexandrina hemolymph on immobilized L-fucose and o-glucose columns. Pooled hemolymph (900 mL) was applied to the L-fucose column at 15% and after washing with SR containing 0.5 M NaCI, bound fractions wereeluted at room temperature with a gradient of 0.1-l M L-fucose. Fractions eluted at 0.43 M and 0.55 M L-fucose were constituted by lectins designated BaSll and BaSIlI, respectively. The run-through fraction was then applied to the o-glucose column and the bound material was eluted by0.3 M o-glucose. The lectin constituting this fraction was designated BaSI. Protein was monitored by absorbance at 280 nm (O), and the collected fractions were dialyzed against SR and their agglutinating activity towards S. mansoni miracidial glycoprotein-coated SRBCs (cross-hatched bars) was determined by HA assays. The insets show silver-stained l- and 2-D SDS-PAGE profiles of BaSII, BaSIlI and BaSl analyzed under reducing conditions with estimates of molecular weights indicated. Constituent subunits of BaSl are indicated by a, b, c and d.
mL) were examined in HA assays conducted in the different buffers given in the Materials and Methods. Within a pH range of 8-10.6, similar binding activities
were observed for the three purified lections (Fig. 2A). At the pH range of 7-8, the activity of BaSI and BaSII remained unchanged, whereas four-fold
Biomphalaria alexandrina family of isolectins
371
reduction in the activity of BaSIII was evident. Below pH 7.0, the activity of the three lectins gradually decreased to reach about g-fold (BaSI and BaSII) and 16fold (BaSIII) reduction at pH 6.0, and became completely inhibited at pH 4.0. As shown in Figure 2B, the three lectins were similar in being stable at a temperature range of 1545°C and in being drastically affected at 60-75°C. It is noteworthy that neither repeated freezing and thawing nor long-term storage (for up to 1 month) at 4°C had any effect on the reactivity of the purified lectins.
Difleerential Binding Afinities of BaSI, BaSII and BaSIII to Fucosyllactose Based on previous results attributing the schistosome-binding activity in B. alexandrina hemolymph to specificities towards a fucosyllactose determinant (20), the nature of the binding of purified BasI, BasII and BasIII to this determinant was investigated by HA1 assays using fucosylated and non-fucosylated lactose derivatives as well as other sugars. As shown in Table 2, comparisons of the 50% inhibition values indicated the similarity of the three lectins in having a basic binding specificity to 2’-fucosyllactose (a trisaccharide having terminal fucose in an al,2 linkage to the galactose residue of lactose). This was confirmed by the effective inhibition of lectin activities observed with lacto-N-fucopentose I, which has a closely-related trisaccharide sequence at its non-reducing terminal end. However, a major distinction in the binding activity of BaSI, as opposed to BaSII and BaSIII, seemed to involve the affinity towards fucosyl residues. The comparable efficiency of fucosylated and non-fucosylated lactose deterivatives in inhibiting BaSI activity was indicative of the minimal involvement of fucose moieties in the high-affinity binding of this lectin to the trisaccharide. On the other
372
M. H. Mansour
B 4
I_
ea-a-a
I_
L-
3
I/
91”
P d P I A-A
a-a-a%, a a \
2
1
a -A
I
I
I
I
I
I
I
2
4
6
6
10
12
15
I 30
I
I
46
60
I
I
75
60
Temp. ‘C
PH
Figure 2. pH dependence and thermostability of BaSl (O), BaSll (A) and BaSilI (0). In (A), the agglutinating activity of purified BaSl (52 ug/mL), BaSll (25 pg/mL) and BaSIlI (90 pg/mL) was assayed against S. mansoni miracidial glycoprotein-coated SRBCs in HA assays conducted in the different buffers given in the Materials and Methods. In (B), BaSI, BaSll and BaSIlI were heated at the indicated temperatures for 30 min, cooled and then assayed in HA assays against the same target cells. Each point on the curves is the mean value of three separate series of experiments. Table 2. lnhlbttlon of BaSI, BaSli and BaSlll blndlng actlvltler by mono- and ollgo-aaccharldes. Ic50 OW
Inhibitor
BaSI’
BaSllt
2’-Fucosyllactose Lacto-IV-fucopentose I Lacto-IV-fucospentose III Lacto-IV-tetrose Lactose N-acetyllactosamine L-Fuc Fucosylamine D-Glc o-Gal
2.25 4.50 9.75 11.25 6.75 13.0 24.0 24.0 20.25 27.5
3.50 4.75 11.50 30.0 28.50 38.75 12.5 12.5 >50.0 >50.0
* Assayed on S. mansoni miracidial mL. t Assayed on S. mansoni miracidial mL. $ Assayed on S. mansoni miracidial mL.
hand, both BaSII and BaSIII exhibited a primary binding preference to fucose and seemed to be sensitive to its position as well as its anomeric linkage in oligosac-
BaSIII$ 5.5 5.75 12.5 >80.0 >80.0 >80.0 13.75 13.75 >50.0 >50.0
glycoprotein-coated
SRBCs at 52 ug/
glycoprotein-coated
SRBCs at 25 ugl
glycoprotein-coated
SRBCs at 90 ug/
charides. This was evident by the two-fold decrease in the inhibitory effect observed with lacto-N-fucopentose III, which differ from lacto-N-fucopentose I in having
Biomphalaria
alexandfina family of isolectins
fucose in subterminal ul,3 linkage, and the marginal or complete lack of inhibition observed with non-fucosylated oligosaccharides. The relatively potent inhibition observed with L-fucose or fucosylamine further supported the selectivity of both lectins in binding fucose, in comparison to galactose and glucose residues constituting the target trisaccharide.
Diflerential N-linked Glycosylation Patterns of BaSI, BaSII and BaSIII Subunits The possible glycosylation of the subunits constituting BaSI, BaSII and BaSIII was investigated by 2-D SDS-PAGE following Endo-F treatments. As shown
373
in Figure 3, treatment of BaSI with EndoF resolved the heterogeneous pattern expressed by the a, b, c and d constituent subunits (Fig. 3A) into two homogeneous spots with a shift in molecular weight as well as charge towards more basic pls. The deglycosylated BaSI was constituted by a prominent 13.2 kDa CCsubunit with a pZ of 7.2 in addition to a 10.1 kDa g subunit with a pZ of 5.8 (Fig. 3B). Deglycosylation of BaSII and BaSIII with Endo-F resulted in a uniformed shift in molecular weight but a differential reduction of the acidic charge microheterogeneity expressed by the subunits constituting each lectin. The deglycosylated BaSII was constituted by a prominent spot that corresponded in terms of molecular weight and pZ to the c1subunit of the deglycosylated BaSI, in addition to PH
4
5i4
6i5
744
si4
95
7.4
Figure 3. Two-dimensional SDS-PAGE of glycosylated and deglycosylated BaSI, BaSll and BaSIlI subunits. Isolated BaSI, BaSll and BaSIlI wereeither untreated (A, C and E, respectively) or treated with Endo-F (B, D and F, respectively) and analyzed by 2-D SDS-PAGE using 12.5% slab gels in the second dimension. Constituent subunits of the glycoylated BaSl (A) are indicatedby a, b, c and d, whereas the arrow-head in (E) marks the most acidic subunit of the glycosylated BaSIlI. Constituent subunits of the deglycosylated lectins are indicated by u, ul, u2 and p, Estimated molecular weights of glycosylated and deglycosylated lectin subunits are also shown.
374
a single minor charge variant of pZ 6.8, designated al (Fig. 3D). Endo-F treatment of BaSIII resulted in the disappearance of the most acidic spot (arrowhead in Fig. 3E) and the persistence of three equalsized charge variants corresponding to the three most basic components in the glycosylated pattern. Based on similarities in both molecular weight and pZ values, these components were equivalent to the c1 and al subunits in addition to a unique ~12 subunit with a distinct pZ of 7.0 (Fig. 3F). Given the known specificity of Endo-F in cleaving linkages in the core of complex and high-mannose N-linked glycans (26), the shift of about 2.4 kDa towards basic pZs observed with BaSII was consistent with the removal of a single complex type glycan unit carrying variable acidic moieties. Similarly, the glycosylation of BaSIII subunits seemed to involve a single glycan unit that was either of the complex type (as evident by the disappearance of the most acidic subunit) or mainly of the high-mannose type (as evident by alterations in size but not charge of the three most basic subunits). The exact glycosylation of BasI subunits, on the other hand, was difficult to gauge and probably involve the addition of 1-3 N-linked oligosaccharide units to either o! and/or p subunits.
Discussion In the present report we expand on earlier experiments (20) and further demonstrate that wild-type B. alexandrina hemolymph contains a family of isolectins that selectively bind to schistosomeassociated fucosyllactose determinants. Under the conditions employed for purification, two members of this family (BaSI and BaSII) have been isolated with yields comparable to those reported previously (20). In addition, a novel lectin accounting for about 12% of the binding activity in the hemolymph and designated BaSIII, is reported. Assessment of the
M. H. Mansour
structural characteristics by l- and 2-D gels indicated that, in spite of similarities in native molecular weights, the three lectins were structurally distinct and exist in their native forms as tetramers of non covalently-associated subunits. The subunits were of different sizes and pZs in BaSI (subunits a, b, c and d) and of equal size but distinct pZs in BaSII and BaSIII. However, comparisons of 2-D gels of the glycosylated and deglycosylated forms were consistent with the presence of an invariant CLsubunit (13.2 kDa, pZ 7.2) constituting the three deglycosylated lectins, which associates with other subunits unique to each lectin, namely the /3 subunit (10.1 kDa, pZ 5.8) of BaSI, the al subunit (13.2 kDa, pZ 6.8) of BaSII and BaSIII and the c~2subunit (13.2 kDa, pZ 7.0) of BasIII. Inasmuch as the glycosylated lectins were not susceptible to alkaline borohydride treatments (not shown), and hence may not express O-linked glycans, and given the known specificity of Endo-F in cleaving linkages in the core of both complex and high mannose N-linked glycans (26), the distinction in pZ values observed among the deglycosylated ~1,cl1 and 012polypeptides reflects limited variations in amino acid sequence and suggests that they are allelic products of an oligomorphic gene family. The differences in size as well as pZ are also indicative of the p subunit being the product of a distinct gene. Evidently, these distinct gene poducts are subjected to differential post-translational N-linked glycosylations, which may be essential for maintaining a selective quaternary configuration for each lectin, and account for the additional heterogeneity of constituent subunits of the glycosylated forms. The structural interrelationship among subunits constituting B. alexandrina isolectins is reminiscent of the complex organization of subunits comprising other families of isolectins (27-29). In each of Bandeiraea simplicifolia (27),
Biomphalaria
alexandrina family of isolectins
Phaseolus vulgaris (28) and Vicia villosa (29) families of isolectins, two unique
subunits of similar primary structures associate in different proportions to give a spectrum of tetrameric structures. Interestingly, the subunits in each of these families also have similar physicochemical properties but differ in their carbohydrate binding specificity. By analogy, BaSI, BaSII and BaSIII were similar in having an optimal binding activity at an alkaline pH range and in being inactivated by above 60°C. However, temperatures based on miracidial glycoprotein-binding and inhibition assays with fucosylated and non-fucosylated derivatives of the blood group antigens, the three lectins were similar in having a primary binding specificity to a 2’-fucosyllactose sequence, but with the binding preference to terminal or sub-terminal fucose moieties being in the hierarchy: BaSIII 2 BaSII % BaSI. It is noteworthy that fucose moieties constitute a common structural denominator that seems to be implicated in various key positions among developmentallyimmunogenic schistosome regulated, glycan epitopes (30-33). Consequently, it could be envisioned that a system of finelytuned recognition molecules, similar to the system proposed earlier by Parish (34) may
375
have evolved in snails to cope with selective oligosaccharide sequences (fucosylated lactose derivatives) expressed by the developing schistosomes. This system seems to be sensitive enough to warrant high- or low-affinity binding to key sugar moieties (e.g. fucose) in different positions. This sensitivity is dictated by the combinatorial association of a diversity of subunits that arise by both genetic and post-translational mechanisms. Evidenced by the selective binding and phagocytic activities exhibited by snail hemocyte populations towards fucosyllactose-bearing targets (35) molecules constituting this system may also be anchored in cell membranes as integral receptors to mediate anti-schistosome cellular defense reactions. It is still to be determined whether a regulatory mechanism accounting for the differential expression of high- or low-affinity schistosome glycan-binding lectins would be descisive in establishing resistance or susceptibility to infection among individual members of the snail population.
AcknowledgementsThis study was supported by a USAID-funded Schistosomiasis Research Project (262-0140.2), grant awards 03-l 1-36 and 07-O l-62.
References 1. Lie, K. J. Survival of Schistosoma mansoni and other trematode larvae in the snail Biomphalariu glabrata. A discussion of the interference hypothesis. Trop. Geogr. Med. 34 : 111-122; 1982. 2. Loker, E. S.; Bayne, C. J. Immunity to trematode larvae in the snail Biomphaluriu. Symp. Zool. Sot. London 56 : 199-200; 1986. 3. Yoshino, T. P.; Boswell, C. A. Antigen sharing between larval trematodes and their snail hosts: how real a phenomenon in immune evasion? Symp. Zool. Sot. London 56 : 221-238; 1986. 4. Loker, E. S.; Bayne, C. J.; Yui, M. A. Echinostoma paraensei: Hemocytes of Biomphalaria glabratu as targets of echinostome mediated interference with host snail resistance to SchLtosoma mansoni. Exp. Parasit. 62: 149-154; 1986.
5. Noda, S.; Loker, E. S. Effects of infection with Echinostoma paraensei on the circulating hemocyte population of the snail host Biomphalaria glabrata. Parasitology 98:3541; 1989. 6. Connors, V. A.; Yoshino, T. P. In vitro effect of larval Schistosoma mansoni excretory-secretory products on phagocytosis-stimulated superoxide production in hemocytes from Biomphalaria glubruta. J. Parasitol. 76 : 895-902; 1990. 7. Fryer, S. E.; Bayne, C. J. Schistosoma mansoni modulation of phagocytosis in Biomphaluriu glubratu. J. Parasitol. 76 : 4552; 1990. 8. Lodes, M. J.; Connors, V. A.; Yoshino, T. P. Isolation and functional characterization of snail hemocyte-modulating polypeptide from primary sporocysts of Schistosoma mansoni Mol. Biothem. Parasitol. 49: l-10; 1991.
376
9. Loker, E. S. On being a parasite in an invertebrate host: A short survival course. .I. Parasitol. 80 : 728-747; 1994. 10. Loker, E. S.; Yui, M. A.; Bayne, C. J. Schistosoma mansoni: Agglutination of sporocysts, and formation of gels on miracidia transforming in plasma of Biomphalaria glabrata. Exp. Parasitol. 58 : S&62; 1984. 11. Boswell, C. A.; Bayne, C. J. Isolation, Characterization and functional assessment of a hemagglutinin from the plasma of Biomphalaria intermediate host of Schistosoma glabrata, mansoni. Dev. Comp. Immunol. 8 : 559-568; 1984. 12. Renwrantz, L. L&ins in molluscs and arthropods: Their occurrence, origin and roles in immunity. Symp. 2001. Sot. London 56: 81-93; 1986. 13. Bayne, C. J. Phagocytosis and non-self recognition in invertebrates. Bioscience 40 : 723-731; 1990. 14. Loker, E. S.; Bayne, C. .I. In vitro encounters between Schistosoma mansoni primary sporocysts and hemolymph components of susceptible and resistant strains of Biomphalaria glabrata. Am. J. Trop. Med. Hyg. 31: 999-1005; 1982. 15. Granath, W. 0. Jr.; Yoshino, T. P. Schistosoma mansoni: Passive transfer of resistance by serum in the vector snail Biomphalaria glabrata. Exp. Parasitol. 58 : 188-193; 1984. 16. Spray, F. J.; Granath, W. 0. Jr. Structural analysis of hemolymph proteins from Schistosoma mansoni (Trematoda)-susceptible and resistant Biomphalaria glabrata. Comp. Biothem. Physiol.-94B:543-553; 1989. _ 17. Snrav. F. J.: Granath. W. 0. Jr. Differential binding of hemolymph proteins from schistosome-resistant and -susceptible Biomphalaria glabrata to Schistosoma mansoni sporocysts. J. Parasitol. 76 : 225-229; 1990. 18. Monroy, F. P.; Hertel, L. A.; Loker, E. S. Carbohydrate-binding plasma proteins from the gastropod Biomphalaria glabrata: Strain specificitv and the effects of trematode infection. Dev. Camp. Immunol. 16 : 335-366; 1992. 19. Monroy, F. P.; Loker, E. S. Production of heterogeneous carbohydrate-binding proteins by the host snail Biomphalaria glabrata following exposure to Echinostoma paraensei and Schistosoma mansoni. J. Parasitol. 79: 416423; 1993. 20. Mansour. M. H.: Neam. H. I.: Saad. A. H.: Talaab, N. I. Charact&&tion of Biomphalaria alexandrina-derived lectins recognizing a fucosyllactose-related determinant on schistosomes. Mol. Biochem. Parasitol. 69: 173-184; 1995. 21. Lowry, 0. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. J. Protein measurement with the folin phenol reagent. J. Biol. Chem. 193 : 265275; 1951.
M. H. Mansour
22. Shioiri, K. The observation of proteins adsorbed on erythrocytes and treated with tannic acid in passive- hemagglutination tests IBovden’s method). Jon. J. EXD. Med. 34: 345353; 1964. ’ 23. Laemmli, U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 : 680-685; 1970. 24. O’Farrell, P. H. High resolution two dimensional electrophoresis of proteins. J. Biol. Chem. 250 : 4007421; 1975. 25. Wray, W.; Boulikas, T.; Wray, V. P.; Hancock, R. Silver staining of proteins in polyacrylamide gels. Analyt. Biochem. 118: 197-203; 1981. 26. Elder, J. H.; Alexander, S. Endo-B-N-acetylglucosaminidase F: Endoglycosidase from Plavobacteruim meningosepticum that cleaves high mannose and complex glycoproteins. Proc. Natl. Acad. Sci. U.S.A. 79:454&4549; 1982. 27. Murphy, L. A.; Goldstein, I. J. Five u-Dgalactopyranosyl-binding isolectins from Bandeiraea simvlicifolia seeds. J. Biol. Chem. 252: 4739-4742;-1971. 28. Leavitt, R. D.; Felsted, R.; Bachur, N. R. Biological and biochemical properties of Phaseolus vulgaris isolectins. J. Biol. Chem. 252 : 2961-2966; 1977. 29. Tollefsen, S. E.; Komfeld, R. Isolation and characterization of lectins from Vicia villosa. Two distinct carbohydrate binding activities are present in seed extract. J. Biol. Chem. 258: 516% 5171; 1983. 30. Ko, A. I.; Drager, U. C.; Ham, D. A. A Schistosoma mansoni epitope recognized by a protective monoclonal antibody is identical to the stage-specific embryonic antigen. Proc. Natl. Acad. Sci. U.S.A. 87:41594163: 1990. 31. Levery, S. B.; Weiss, J. B.; Salyan, M. E. K.; Roberts, C. E.; Hakomori, S. I.; Magnani, J. L.; Strand, M. Characterization of a series of novel fucose-containing glycosphingolipid immunogens from eggs of Schistosoma mansoni. J. Biol. Chem. 267 : 5542-5551; 1992. 32. Srivatsan, J.; Smith, D. F.; Cummings, R. D. The human blood fluke Schistosoma mansoni synthesizes glycoproteins containing the Lewis X antigen. J. Biol. Chem. 267: 2019620203; 1992. 33. Mansour, M. H. Structural characterization of Sm 31: A Schistosoma mansoni adult male glycoprotein expressing a fucosyllactose determinant. (Submitted). 34. Parish, C. R. Simple model for self-non-selfdiscrimination in invertebrates. Nature 267:711-713; 1977. 35. Negm, H. I.; Mansour, M. H.; Mohamed, A. H. Biomphalaria alexandrina: Defense mechanism in adult and juvenile snails towards selective Schistosoma mansoni glycoproteins. (Submitted).
and Comparative
Immunology,
Pergamon
Vol. 19, No. 5, pp. 365376, 1995 Else&r Science Ltd Printed in the USA 0145-30x/95 S9.50 + 0.00
EVIDENCE FOR A FAMILY OF SCHISTOSOME GLYCAN-BINDING LECTINS IN Biomphalaria alexandrina Mohamed Department
of Zoology,
H. Mansour
Faculty of Science, Cairo University, Cairo 12613, Egypt
(Submitted April 1995; Accepted June 1995) CiAbstract-A novel family of isolectins that selectively recognize a schistosome-associated fucosyllactose determinant was identified in the hemolymph of Biomphalaria alexandrina, a snail vector of Schistosoma mansoni. Three lectins of this family were purified by serial affinity chromatography on a column of L-fucose and elution with a gradient of 0.1-l M L-fucose (designated BaSII and BaSIII), followed by a column of D-glucose and elution with 0.3 MD-glucose (designated BaSI). Assessment of the structural characteristics by one- and two-dimensional gels indicated that, inspite of similarities in native molecular weights, the three lectins were tetramers of noncovalently-associated subunits that were of different sizes and pIs in BaSI, and of equal size but distinct pIs in BaSII and BaSIII. Comparisons of two-dimensional gels of the glycosylated and deglycosylated forms were consistent with the presence of an invariant a subunit (13.2 kDa, pZ 7.2) constituting the three deglycosylated lectins, which associates with other subunits unique to each lectin, namely a /3 subunit (10.1 kDa, pZ 5.8) in BaSI, an al subunit (13.2 kDa, pZ 6.8) in BaSII and BaSIII, and an a2 subunit (13.2 kDa, ~17.0) in BaSIII. Each of these subunits is subjected to differential post-translational N-linked glycosylations, which accounts for
the additional heterogeneity expressed by the glycosylated lectins. Based on miracidial glycoprotein binding and inhibition assays, the three lectins exhibited optimum binding at similar pH and temperature, but were distinct in their binding affinities towards the fucose moiety constituting the fucosyllactose target. These observations indicate that an oligomorphic family of recognition molecules may have evolved to regulate the snails’ response to schistosomes.
qKeywords-Biomphalaria akxandrina; Schistosoma mansoni; Isolectlns; Fucosyllactose determinant; Oligomorphic gene family; Lectin subunits; N-linked glycans.
Nomenclature Endo-F HA HA1 1-D 2-D SDS-PAGE SR
endo-f%N-acteylglucosaminidase F hemagglutination hemagglutination inhibition one-dimensional two-dimensional Sodium dodecyl sulfatepolyacrylamide gel electrophoresis snail Ringer
Introduction
Address correspondence to Dr Mohamed H. Mansour, Department of Zoology, Faculty of Science, Cairo University, Cairo 12613, Egypt.
Factors governing the success or failure of larval Schistosoma mansoni development within its snail intermediate host are poorly understood (l-3). There have been many reports, however, that this 365
366
dynamic host-parasite interplay involves the capacity of the snail to mount effective internal defense reactions, which in turn may be the primary targets of parasite-mediated interference (4-9). In particular, humoral recognition factors with carbohydrate-binding properties, opsonizing as well as agglutinating activities characteristic of lectins, have been implicated in mediating events following host-parasite contact (l&13). In part, this notion has been strengthened by the differential expression of these humoral factors among strains of snails that are genetically resistant or susceptible to schistosome infection (14-16). Nonetheless, information regarding the molecular characteristics of these snail-derived factors or their putative schistosome-associated oligosaccharide targets is, as yet, not fully explored (17-19). In Biomphalaria alexandrina, the snail vector of the Egyptian strain of S. mansoni, two isolectins that express selectivity in binding a schistosomeassociated fucosyllactose determinant have been recently purified and characterized (20). These lectins have been purified by affinity chromatography on a column of equimolar mixture of Dand L-glucose and sequential elution by D-glucose and L-fucose. In the present study, we have employed a different purification strategy, which resulted in the isolation of an additional lectin expressing both distinct subunit structure as well as binding affinity towards the schistosome oligosaccharide target. Assessment of the structural characteristics of the glycosylated and deglycosylated forms of the three snail-derived lectins by two-dimensional gels was consistent with their constituent subunits being the products of an oligomorphic gene family. This family of recognition molecules may have evolved to regulate the snails’ response towards selective schistosome-associated oligosaccharides.
M. H. Mansour
Materials and Methods Animals and Parasites Naturally-infected and non-infected specimens of the adult (12-18 mm shell diameter) snail B. alexandrina were supplied by the Schistosome Biological Supply Program (SBSP, Theodor Bilharz Research Institute, Cairo, Egypt). Schistosoma mansoni miracidia were prepared by allowing eggs collected from livers and intestines of infected hamsters to hatch under illuminatuion, and were supplied lyophilized and egg shell-free by SBSP.
Collection of B. alexandrina Hemolymph Hemolymph was collected by severely touching the foot of the snail with a Pasteur pipette. As a result, the snail retracts into the shell and exudes 100-250 uL of hemolymph (20). Aliquots collected from individual snails (a total of 4 x 103) were mixed with an equal volume of snail Ringer (SR, 0.2 A4 Tris-HCl, pH 8.0 containing 2.5% NaCl, 0.15% KCl, 0.4% MgC&, 0.65% CaC12), pooled, freed of hemocytes by centrifugation at 1500 rpm for 10 min at 4°C and used immediately or stored at -20°C.
Preparation of S. mansoni Miracidial Glycoprotein-bearing Fucosyllactose Determinants Lyophilized miracidia (2 mL pellet) were solubilized in 5 mL of lysis buffer (20 mM Tris-HCI, pH 8.0, 150 mM NaCl, 2% deoxycholate, 2 mA4 phenylmethylsulfonyl fluoride, 1 mM o-phenanthroline, 1 mM p-chloromercuribenzoic acid and 1 mM iodoacetamine) by sonic treatment followed by three cycles of freezing and thawing. The solubilized protein fraction was separated by centrifugation at 100,000 x g for 1 h and diluted
Biomphalaria alexandrina family of isolectins
10 fold with 20 mM Tris-HCl, pH 8.0. A fraction of the miracidial glycoproteins expressing a fucosyllactose determinant was prepared by affinity chromatography of the miracidial extract on an immobilized B. alexandrina lectin column. For the preparation of this column, a purified fraction of the snail lectin specific to the fucosyllactose schistosome-associated determinant (BaSII) was prepared from the hemolymph by affinity chromatography on a column of equimolar mixture of D- and L-glucose and elution by 0.3 M L-fucose as described previously (20). Purified BaSII (2 mg) was desalted and separated from monosaccharides by chromatography on a column of Bio Gel PlO (1.5 x 20 cm) and coupled to CNBractivated Sepharose 4B (Pharmacia Fine Chemicals, Sweden) at a concentration of 2 mg/mL beads following the manufacturer’s instructions. The coupled gel (1 mL) was equilibrated in lysis buffer containing 0.2% deoxycholate (deoxycholate buffer) and the miracidial extract cycled through the column 6 times at 15°C. Bound glycoproteins were eluted by 0.3 M L-fucose and desalted and separated from the monosaccharide by chromatography on a column of Bio Gel PlO (1.5 x 20 cm) in deoxycholate buffer. Protein was determined by the method of Lowry et al. (21) using bovine serum albumin (BSA) in deoxycholate buffer as a standard.
Target Cells Sheep erythrocytes (SRBCs, VACSERA, Cairo, Egypt) were formalin fixed, tanned and coated with BSA or fucosyllactose determinant-bearing miracidial glycoproteins according to standard procedures (22) and used as target cells in all assays. Coatings were done at the ratio of 1 mL packed SRBCs to 3 volumes of the glycoprotein solution (100 ug/mL) in deoxycholate buffer. Prior to use, coated SRBCs were washed three times in SR,
367
pH 8.0 and resuspended buffer as 1.5% suspension.
in the same
Indirect Hemagglutination and Hemagglutination Inhibition Assays Hemagglutination (HA) assays were performed in microtiter plates at room temperature. Snail hemolymph, postcolumn fractions or purified lectins (25 uL) were serially diluted in SR containing 0.05% BSA and mixed with 25 uL of a 1.5% schistosome glycoprotein-coated SRBCs suspension. HA titers were scored microscopically as the reciprocal of the highest dilution of hemolymph (or purified lectins) giving complete agglutination in 1 h. Assays were performed in triplicate and controlled by the substitution of hemolymph or lectin by SR, and by using BSA-coated SRBCs as test cells. Hemagglutination inhibition OIAI) assays were conducted by mixing serially diluted hemolymph or pure lectins (25 uL) with 25 uL of mono-, di- or oligosaccharides (all sugars were purchased from Sigma Chem. Comp., St. Louis, MO, except 2’-fucosyllactose, lacto-Nfucopentose I and III and lacto-N-tetrose, which were from Calbiochem Corp., San Diego, CA). Sugars were dissolved in SR (at concentrations up to 100 m&f’) and the reaction mixture incubated for 1 h at room temperature followed by the addition of 25 uL of a 1.5% suspension of schistosome glycoprotein-coated SRBCs. Controls were the substitutions of hemolymph or pure lectin by SR and by the substitution of the inhibition solution by SR. Fifty percent inhibition values were interpolated rom plots of percent inhibition versus inhibitor concentration.
Serial Afinity Chromatography with Immobilized Monosaccharide Columns Snail hemolymph in SR (900 mL) was chromatographed on a 0.7 x 14 cm
368
column containing L-fucose-agarose (Lfucose attached by a five carbon spacer, Sigma Chemical Comp., St. Louis, MO). Before use, the column was equilibrated with SR and then the pooled hemolymph was recycled through it three times at 5 mL/h at 15°C. The column was extensively washed with SR followed by SR containing 0.5 A4 NaCl, and specifically adsorbed lectins were eluted at room temperature with a gradient of 0.1-l M L-fucose in SR containing 0.5 M NaCl. The material that passed unretarded through the fucose column was pooled and dialyzed against SR and then recycled three times at 5 mL/h through a 0.7 x 14 cm column containing D-glucose-agarose (D-glucose attached by a five carbon spacer, Sigma Chemical Comp., St. Louis, MO). The column was washed with SR followed by SR containing 0.5 M NaCl, and specifically adsorbed lectins were eluted with SR containing 0.3 M Dglucose and 0.5 M NaCl. To attain maximum stability, specifically eluted fractions from either columns were dialyzed against SR, aliquoted, freeze-dried and stored at -20°C. Anti-schistosome binding activities were monitored during chromatography by HA and HA1 assays miracidial glycoprotein-coated using SRBCs (25 pL of 1.5% coated SRBCs/ well) as indicated above. Column fractions were continuously monitored for protein by absorbance at 280 nm. A unit of activity was defined as the minimum concentration of protein required to give one well of HA.
pH Dependence and Temperature Stability of PuriJied Snail Lectins The pH range over which the purified snail lectins exhibit optimal binding to schistosomal glycoproteins was examined in HA assays by titrating purified lectins in the following buffers: 100 mM sodium acetate/acetic acid (pH 3.7-5.4), 100 mA4 sodium cacodylate/HCl (pH [email protected]), 100
M. H. Mansour
mM Tris-HCl (pH 7.5-8.8), 100 mM glycine/NaOH @H 9.0-10.6). All buffers contained 10 mM CaClz and 10 mM MgC12 and test cells were miracidial glycoprotein-coated SRBCs. The temperature stability of the purified lectins was estimated by incubating 100 uL aliquots of the lectins in SR at different temperatures for 30 min, cooling them to room temperature and titrating them with miracidial glycoprotein-coated SRBCs in HA assays.
One- and Two-dimensional Polyacrylamide Gel Electrophoresis (1-D and 2-D SDS-PAGE) 1-D SDS-PAGE was conducted under non-reducing and reducing conditions using the discontinuous buffer of Laemmli (23). 2-D SDS-PAGE was conducted essentially as described by O’Farrell (24). Lyophilized samples of glycosylated and deglycosylated snail lectins were dissolved in lysis buffer, incubated at 37°C for 4 h, and applied to the isoelectric focusing gels for 18 h (8000 Vh). Samples were resolved in the second dimension with 12.5% slab SDS-PAGE. Samples were visualized by silver staining as described by Wray et al. (25) using BioRad low molecular weight standards and the gels photographed wet using Kodak Tri-X-Pan films.
Glycosidase Treatments Purified fractions of snail lectins (25 pg) were reconstituted in 50 uL of 100 mM sodium phosphate, pH 6.1, 50 mM EDTA, 1% Nonidet P-40 containing 100 mU of Endo-F (Endo-B-N-acetylglucosaminidase F, from Flavobacterium meningosepticum, 600 U/mg, Sigma Chem. Comp. St., Louis, MO). Samples were incubated for 18 h at 37°C precipitated with an equal volume of 20% trichloroacetic acid, washed with cold acetone and
Biorrtphalaria alexandrinafamily of isolectins
369
was resolved by 2-D SDS-PAGE into four equal-sized subunits of 15.6 kDa expressing distinct patterns of charge microheterogeneity in each eluate. The ioselectric points of the individual charge entities constituting the 0.43 M r_-fucose eluate were identical to those previously Results estimated for the lectin isolated on a D-/Lglucose column by elution with r_-fucose Isolation of Three Lectins (pls 5.6, 5.8, 6.2 and 6.4) (20) and was Constituting the Anti-schistosome Binding Activity in B. alexandrina hence given the same designation, BaSII. The 0.55 A4 L-fucose eluate, on the other Hemolymph hand, was constituted by four distinct charge varients of pIs 6.4,6.8,7.0 and 7.2, In a previous report, we have demonstrated the high affinity of B. alexandrina and was designated BaSIII. Comparison hemolymph for L-fucosyl and D-glucosyl of the total activities in the cell-free hemolymph, BaSII, BaSIII and the runresidues constituting a fucosyllactose through fraction indicate that about 9% determinant associated with schistosomes (20). This activity has been resolved into of the loaded activity could not be two lectins, designated BasI and BasII, recovered from the r_-fucose column when purification was attempted by affin- under the employed conditions, perhaps due to high binding affinities. ity chromatography on an immobilized The material that passed unretarded equimolar mixture of D- and L-glucose and sequential elution with D-glucose and through the fucose column was constiL-fucose. Under these conditions, part of tuted by about 24% of the initial aggluthe anti-fucosyl binding activity could tinating activity in the hemolymph. When applied to a D-glucose column, over 87% not, however, be recovered. In the present of this activity was bound and readily study, serial affinity chromatography using a column of L-fucose followed by eluted on addition of SR containing 0.3 a column of D-glucose was employed as M D-glucose and 0.5 M NaCl (Fig. 1 and an alternative purification strategy. As Table 1). This fraction was constituted by shown in Figure 1 and Table 1, a fraction a single component of 67 kDa and constituted by about 55% of the loaded resolved by 2-D SDS-PAGE into four activity in the hemolymph was eluted subunits of 20.4 kDa (PI 6.3), 17.6 kDa from the L-fucose column at 0.43 M L- (pf 6.0), 15.6 kDa (pls 5.7, 5.8 and 6.3) fucose with 0.1% of the protein. An and 13.5 kDa (pls 5.8 and 6.0), corresadditional 12% of the activity was yielded ponding to the a, b, c and d subunits of with 0.02% of the loaded hemolymph BaSI (20), and was hence given the same protein when the molarity of r_-fucose designation. reached 0.55 M. No further displacment of detectable activity was observed with up to 1 M L-fucose, 2 A4 NaCl or when pH Dependence and Temperature 0.2% SDS was included in the eluting Stability of BaSI, BaSII and BaSIII buffer. As judged by 1-D SDS-PAGE (Fig. l), BasI, BasII and BasIII were compared unboiled samples of both the 0.34 A4 and in terms of the pH range over which they 0.55 M L-fucose eluates were each con- exhibit optimum binding to miracidial stituted by a single 62 kDa component glycoprotein-coated SRBCs. Aliquots of when analyzed under either reducing or the purified lectins (BaSI at 52 l.tg/mL non-reducing conditions, This component BaSII at 25 pg/mL and BaSIII at 90 ug/ dried under nitrogen gas before analysis by 2-D SDS-PAGE. Control samples were similarly treated but in the absence of Endo-F.
370
M. H. Mansour
0.43NFUC
Elution
PH Vol. (ml)
J Wash
0.55YFUC
5.4
0.3MGlc
lboo Elution Vol.fml) Flgure 1. Serial affinity chromatography of 6. alexandrina hemolymph on immobilized L-fucose and o-glucose columns. Pooled hemolymph (900 mL) was applied to the L-fucose column at 15% and after washing with SR containing 0.5 M NaCI, bound fractions wereeluted at room temperature with a gradient of 0.1-l M L-fucose. Fractions eluted at 0.43 M and 0.55 M L-fucose were constituted by lectins designated BaSll and BaSIlI, respectively. The run-through fraction was then applied to the o-glucose column and the bound material was eluted by0.3 M o-glucose. The lectin constituting this fraction was designated BaSI. Protein was monitored by absorbance at 280 nm (O), and the collected fractions were dialyzed against SR and their agglutinating activity towards S. mansoni miracidial glycoprotein-coated SRBCs (cross-hatched bars) was determined by HA assays. The insets show silver-stained l- and 2-D SDS-PAGE profiles of BaSII, BaSIlI and BaSl analyzed under reducing conditions with estimates of molecular weights indicated. Constituent subunits of BaSl are indicated by a, b, c and d.
mL) were examined in HA assays conducted in the different buffers given in the Materials and Methods. Within a pH range of 8-10.6, similar binding activities
were observed for the three purified lections (Fig. 2A). At the pH range of 7-8, the activity of BaSI and BaSII remained unchanged, whereas four-fold
Biomphalaria alexandrina family of isolectins
371
reduction in the activity of BaSIII was evident. Below pH 7.0, the activity of the three lectins gradually decreased to reach about g-fold (BaSI and BaSII) and 16fold (BaSIII) reduction at pH 6.0, and became completely inhibited at pH 4.0. As shown in Figure 2B, the three lectins were similar in being stable at a temperature range of 1545°C and in being drastically affected at 60-75°C. It is noteworthy that neither repeated freezing and thawing nor long-term storage (for up to 1 month) at 4°C had any effect on the reactivity of the purified lectins.
Difleerential Binding Afinities of BaSI, BaSII and BaSIII to Fucosyllactose Based on previous results attributing the schistosome-binding activity in B. alexandrina hemolymph to specificities towards a fucosyllactose determinant (20), the nature of the binding of purified BasI, BasII and BasIII to this determinant was investigated by HA1 assays using fucosylated and non-fucosylated lactose derivatives as well as other sugars. As shown in Table 2, comparisons of the 50% inhibition values indicated the similarity of the three lectins in having a basic binding specificity to 2’-fucosyllactose (a trisaccharide having terminal fucose in an al,2 linkage to the galactose residue of lactose). This was confirmed by the effective inhibition of lectin activities observed with lacto-N-fucopentose I, which has a closely-related trisaccharide sequence at its non-reducing terminal end. However, a major distinction in the binding activity of BaSI, as opposed to BaSII and BaSIII, seemed to involve the affinity towards fucosyl residues. The comparable efficiency of fucosylated and non-fucosylated lactose deterivatives in inhibiting BaSI activity was indicative of the minimal involvement of fucose moieties in the high-affinity binding of this lectin to the trisaccharide. On the other
372
M. H. Mansour
B 4
I_
ea-a-a
I_
L-
3
I/
91”
P d P I A-A
a-a-a%, a a \
2
1
a -A
I
I
I
I
I
I
I
2
4
6
6
10
12
15
I 30
I
I
46
60
I
I
75
60
Temp. ‘C
PH
Figure 2. pH dependence and thermostability of BaSl (O), BaSll (A) and BaSilI (0). In (A), the agglutinating activity of purified BaSl (52 ug/mL), BaSll (25 pg/mL) and BaSIlI (90 pg/mL) was assayed against S. mansoni miracidial glycoprotein-coated SRBCs in HA assays conducted in the different buffers given in the Materials and Methods. In (B), BaSI, BaSll and BaSIlI were heated at the indicated temperatures for 30 min, cooled and then assayed in HA assays against the same target cells. Each point on the curves is the mean value of three separate series of experiments. Table 2. lnhlbttlon of BaSI, BaSli and BaSlll blndlng actlvltler by mono- and ollgo-aaccharldes. Ic50 OW
Inhibitor
BaSI’
BaSllt
2’-Fucosyllactose Lacto-IV-fucopentose I Lacto-IV-fucospentose III Lacto-IV-tetrose Lactose N-acetyllactosamine L-Fuc Fucosylamine D-Glc o-Gal
2.25 4.50 9.75 11.25 6.75 13.0 24.0 24.0 20.25 27.5
3.50 4.75 11.50 30.0 28.50 38.75 12.5 12.5 >50.0 >50.0
* Assayed on S. mansoni miracidial mL. t Assayed on S. mansoni miracidial mL. $ Assayed on S. mansoni miracidial mL.
hand, both BaSII and BaSIII exhibited a primary binding preference to fucose and seemed to be sensitive to its position as well as its anomeric linkage in oligosac-
BaSIII$ 5.5 5.75 12.5 >80.0 >80.0 >80.0 13.75 13.75 >50.0 >50.0
glycoprotein-coated
SRBCs at 52 ug/
glycoprotein-coated
SRBCs at 25 ugl
glycoprotein-coated
SRBCs at 90 ug/
charides. This was evident by the two-fold decrease in the inhibitory effect observed with lacto-N-fucopentose III, which differ from lacto-N-fucopentose I in having
Biomphalaria
alexandfina family of isolectins
fucose in subterminal ul,3 linkage, and the marginal or complete lack of inhibition observed with non-fucosylated oligosaccharides. The relatively potent inhibition observed with L-fucose or fucosylamine further supported the selectivity of both lectins in binding fucose, in comparison to galactose and glucose residues constituting the target trisaccharide.
Diflerential N-linked Glycosylation Patterns of BaSI, BaSII and BaSIII Subunits The possible glycosylation of the subunits constituting BaSI, BaSII and BaSIII was investigated by 2-D SDS-PAGE following Endo-F treatments. As shown
373
in Figure 3, treatment of BaSI with EndoF resolved the heterogeneous pattern expressed by the a, b, c and d constituent subunits (Fig. 3A) into two homogeneous spots with a shift in molecular weight as well as charge towards more basic pls. The deglycosylated BaSI was constituted by a prominent 13.2 kDa CCsubunit with a pZ of 7.2 in addition to a 10.1 kDa g subunit with a pZ of 5.8 (Fig. 3B). Deglycosylation of BaSII and BaSIII with Endo-F resulted in a uniformed shift in molecular weight but a differential reduction of the acidic charge microheterogeneity expressed by the subunits constituting each lectin. The deglycosylated BaSII was constituted by a prominent spot that corresponded in terms of molecular weight and pZ to the c1subunit of the deglycosylated BaSI, in addition to PH
4
5i4
6i5
744
si4
95
7.4
Figure 3. Two-dimensional SDS-PAGE of glycosylated and deglycosylated BaSI, BaSll and BaSIlI subunits. Isolated BaSI, BaSll and BaSIlI wereeither untreated (A, C and E, respectively) or treated with Endo-F (B, D and F, respectively) and analyzed by 2-D SDS-PAGE using 12.5% slab gels in the second dimension. Constituent subunits of the glycoylated BaSl (A) are indicatedby a, b, c and d, whereas the arrow-head in (E) marks the most acidic subunit of the glycosylated BaSIlI. Constituent subunits of the deglycosylated lectins are indicated by u, ul, u2 and p, Estimated molecular weights of glycosylated and deglycosylated lectin subunits are also shown.
374
a single minor charge variant of pZ 6.8, designated al (Fig. 3D). Endo-F treatment of BaSIII resulted in the disappearance of the most acidic spot (arrowhead in Fig. 3E) and the persistence of three equalsized charge variants corresponding to the three most basic components in the glycosylated pattern. Based on similarities in both molecular weight and pZ values, these components were equivalent to the c1 and al subunits in addition to a unique ~12 subunit with a distinct pZ of 7.0 (Fig. 3F). Given the known specificity of Endo-F in cleaving linkages in the core of complex and high-mannose N-linked glycans (26), the shift of about 2.4 kDa towards basic pZs observed with BaSII was consistent with the removal of a single complex type glycan unit carrying variable acidic moieties. Similarly, the glycosylation of BaSIII subunits seemed to involve a single glycan unit that was either of the complex type (as evident by the disappearance of the most acidic subunit) or mainly of the high-mannose type (as evident by alterations in size but not charge of the three most basic subunits). The exact glycosylation of BasI subunits, on the other hand, was difficult to gauge and probably involve the addition of 1-3 N-linked oligosaccharide units to either o! and/or p subunits.
Discussion In the present report we expand on earlier experiments (20) and further demonstrate that wild-type B. alexandrina hemolymph contains a family of isolectins that selectively bind to schistosomeassociated fucosyllactose determinants. Under the conditions employed for purification, two members of this family (BaSI and BaSII) have been isolated with yields comparable to those reported previously (20). In addition, a novel lectin accounting for about 12% of the binding activity in the hemolymph and designated BaSIII, is reported. Assessment of the
M. H. Mansour
structural characteristics by l- and 2-D gels indicated that, in spite of similarities in native molecular weights, the three lectins were structurally distinct and exist in their native forms as tetramers of non covalently-associated subunits. The subunits were of different sizes and pZs in BaSI (subunits a, b, c and d) and of equal size but distinct pZs in BaSII and BaSIII. However, comparisons of 2-D gels of the glycosylated and deglycosylated forms were consistent with the presence of an invariant CLsubunit (13.2 kDa, pZ 7.2) constituting the three deglycosylated lectins, which associates with other subunits unique to each lectin, namely the /3 subunit (10.1 kDa, pZ 5.8) of BaSI, the al subunit (13.2 kDa, pZ 6.8) of BaSII and BaSIII and the c~2subunit (13.2 kDa, pZ 7.0) of BasIII. Inasmuch as the glycosylated lectins were not susceptible to alkaline borohydride treatments (not shown), and hence may not express O-linked glycans, and given the known specificity of Endo-F in cleaving linkages in the core of both complex and high mannose N-linked glycans (26), the distinction in pZ values observed among the deglycosylated ~1,cl1 and 012polypeptides reflects limited variations in amino acid sequence and suggests that they are allelic products of an oligomorphic gene family. The differences in size as well as pZ are also indicative of the p subunit being the product of a distinct gene. Evidently, these distinct gene poducts are subjected to differential post-translational N-linked glycosylations, which may be essential for maintaining a selective quaternary configuration for each lectin, and account for the additional heterogeneity of constituent subunits of the glycosylated forms. The structural interrelationship among subunits constituting B. alexandrina isolectins is reminiscent of the complex organization of subunits comprising other families of isolectins (27-29). In each of Bandeiraea simplicifolia (27),
Biomphalaria
alexandrina family of isolectins
Phaseolus vulgaris (28) and Vicia villosa (29) families of isolectins, two unique
subunits of similar primary structures associate in different proportions to give a spectrum of tetrameric structures. Interestingly, the subunits in each of these families also have similar physicochemical properties but differ in their carbohydrate binding specificity. By analogy, BaSI, BaSII and BaSIII were similar in having an optimal binding activity at an alkaline pH range and in being inactivated by above 60°C. However, temperatures based on miracidial glycoprotein-binding and inhibition assays with fucosylated and non-fucosylated derivatives of the blood group antigens, the three lectins were similar in having a primary binding specificity to a 2’-fucosyllactose sequence, but with the binding preference to terminal or sub-terminal fucose moieties being in the hierarchy: BaSIII 2 BaSII % BaSI. It is noteworthy that fucose moieties constitute a common structural denominator that seems to be implicated in various key positions among developmentallyimmunogenic schistosome regulated, glycan epitopes (30-33). Consequently, it could be envisioned that a system of finelytuned recognition molecules, similar to the system proposed earlier by Parish (34) may
375
have evolved in snails to cope with selective oligosaccharide sequences (fucosylated lactose derivatives) expressed by the developing schistosomes. This system seems to be sensitive enough to warrant high- or low-affinity binding to key sugar moieties (e.g. fucose) in different positions. This sensitivity is dictated by the combinatorial association of a diversity of subunits that arise by both genetic and post-translational mechanisms. Evidenced by the selective binding and phagocytic activities exhibited by snail hemocyte populations towards fucosyllactose-bearing targets (35) molecules constituting this system may also be anchored in cell membranes as integral receptors to mediate anti-schistosome cellular defense reactions. It is still to be determined whether a regulatory mechanism accounting for the differential expression of high- or low-affinity schistosome glycan-binding lectins would be descisive in establishing resistance or susceptibility to infection among individual members of the snail population.
AcknowledgementsThis study was supported by a USAID-funded Schistosomiasis Research Project (262-0140.2), grant awards 03-l 1-36 and 07-O l-62.
References 1. Lie, K. J. Survival of Schistosoma mansoni and other trematode larvae in the snail Biomphalariu glabrata. A discussion of the interference hypothesis. Trop. Geogr. Med. 34 : 111-122; 1982. 2. Loker, E. S.; Bayne, C. J. Immunity to trematode larvae in the snail Biomphaluriu. Symp. Zool. Sot. London 56 : 199-200; 1986. 3. Yoshino, T. P.; Boswell, C. A. Antigen sharing between larval trematodes and their snail hosts: how real a phenomenon in immune evasion? Symp. Zool. Sot. London 56 : 221-238; 1986. 4. Loker, E. S.; Bayne, C. J.; Yui, M. A. Echinostoma paraensei: Hemocytes of Biomphalaria glabratu as targets of echinostome mediated interference with host snail resistance to SchLtosoma mansoni. Exp. Parasit. 62: 149-154; 1986.
5. Noda, S.; Loker, E. S. Effects of infection with Echinostoma paraensei on the circulating hemocyte population of the snail host Biomphalaria glabrata. Parasitology 98:3541; 1989. 6. Connors, V. A.; Yoshino, T. P. In vitro effect of larval Schistosoma mansoni excretory-secretory products on phagocytosis-stimulated superoxide production in hemocytes from Biomphalaria glubruta. J. Parasitol. 76 : 895-902; 1990. 7. Fryer, S. E.; Bayne, C. J. Schistosoma mansoni modulation of phagocytosis in Biomphaluriu glubratu. J. Parasitol. 76 : 4552; 1990. 8. Lodes, M. J.; Connors, V. A.; Yoshino, T. P. Isolation and functional characterization of snail hemocyte-modulating polypeptide from primary sporocysts of Schistosoma mansoni Mol. Biothem. Parasitol. 49: l-10; 1991.
376
9. Loker, E. S. On being a parasite in an invertebrate host: A short survival course. .I. Parasitol. 80 : 728-747; 1994. 10. Loker, E. S.; Yui, M. A.; Bayne, C. J. Schistosoma mansoni: Agglutination of sporocysts, and formation of gels on miracidia transforming in plasma of Biomphalaria glabrata. Exp. Parasitol. 58 : S&62; 1984. 11. Boswell, C. A.; Bayne, C. J. Isolation, Characterization and functional assessment of a hemagglutinin from the plasma of Biomphalaria intermediate host of Schistosoma glabrata, mansoni. Dev. Comp. Immunol. 8 : 559-568; 1984. 12. Renwrantz, L. L&ins in molluscs and arthropods: Their occurrence, origin and roles in immunity. Symp. 2001. Sot. London 56: 81-93; 1986. 13. Bayne, C. J. Phagocytosis and non-self recognition in invertebrates. Bioscience 40 : 723-731; 1990. 14. Loker, E. S.; Bayne, C. .I. In vitro encounters between Schistosoma mansoni primary sporocysts and hemolymph components of susceptible and resistant strains of Biomphalaria glabrata. Am. J. Trop. Med. Hyg. 31: 999-1005; 1982. 15. Granath, W. 0. Jr.; Yoshino, T. P. Schistosoma mansoni: Passive transfer of resistance by serum in the vector snail Biomphalaria glabrata. Exp. Parasitol. 58 : 188-193; 1984. 16. Spray, F. J.; Granath, W. 0. Jr. Structural analysis of hemolymph proteins from Schistosoma mansoni (Trematoda)-susceptible and resistant Biomphalaria glabrata. Comp. Biothem. Physiol.-94B:543-553; 1989. _ 17. Snrav. F. J.: Granath. W. 0. Jr. Differential binding of hemolymph proteins from schistosome-resistant and -susceptible Biomphalaria glabrata to Schistosoma mansoni sporocysts. J. Parasitol. 76 : 225-229; 1990. 18. Monroy, F. P.; Hertel, L. A.; Loker, E. S. Carbohydrate-binding plasma proteins from the gastropod Biomphalaria glabrata: Strain specificitv and the effects of trematode infection. Dev. Camp. Immunol. 16 : 335-366; 1992. 19. Monroy, F. P.; Loker, E. S. Production of heterogeneous carbohydrate-binding proteins by the host snail Biomphalaria glabrata following exposure to Echinostoma paraensei and Schistosoma mansoni. J. Parasitol. 79: 416423; 1993. 20. Mansour. M. H.: Neam. H. I.: Saad. A. H.: Talaab, N. I. Charact&&tion of Biomphalaria alexandrina-derived lectins recognizing a fucosyllactose-related determinant on schistosomes. Mol. Biochem. Parasitol. 69: 173-184; 1995. 21. Lowry, 0. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. J. Protein measurement with the folin phenol reagent. J. Biol. Chem. 193 : 265275; 1951.
M. H. Mansour
22. Shioiri, K. The observation of proteins adsorbed on erythrocytes and treated with tannic acid in passive- hemagglutination tests IBovden’s method). Jon. J. EXD. Med. 34: 345353; 1964. ’ 23. Laemmli, U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 : 680-685; 1970. 24. O’Farrell, P. H. High resolution two dimensional electrophoresis of proteins. J. Biol. Chem. 250 : 4007421; 1975. 25. Wray, W.; Boulikas, T.; Wray, V. P.; Hancock, R. Silver staining of proteins in polyacrylamide gels. Analyt. Biochem. 118: 197-203; 1981. 26. Elder, J. H.; Alexander, S. Endo-B-N-acetylglucosaminidase F: Endoglycosidase from Plavobacteruim meningosepticum that cleaves high mannose and complex glycoproteins. Proc. Natl. Acad. Sci. U.S.A. 79:454&4549; 1982. 27. Murphy, L. A.; Goldstein, I. J. Five u-Dgalactopyranosyl-binding isolectins from Bandeiraea simvlicifolia seeds. J. Biol. Chem. 252: 4739-4742;-1971. 28. Leavitt, R. D.; Felsted, R.; Bachur, N. R. Biological and biochemical properties of Phaseolus vulgaris isolectins. J. Biol. Chem. 252 : 2961-2966; 1977. 29. Tollefsen, S. E.; Komfeld, R. Isolation and characterization of lectins from Vicia villosa. Two distinct carbohydrate binding activities are present in seed extract. J. Biol. Chem. 258: 516% 5171; 1983. 30. Ko, A. I.; Drager, U. C.; Ham, D. A. A Schistosoma mansoni epitope recognized by a protective monoclonal antibody is identical to the stage-specific embryonic antigen. Proc. Natl. Acad. Sci. U.S.A. 87:41594163: 1990. 31. Levery, S. B.; Weiss, J. B.; Salyan, M. E. K.; Roberts, C. E.; Hakomori, S. I.; Magnani, J. L.; Strand, M. Characterization of a series of novel fucose-containing glycosphingolipid immunogens from eggs of Schistosoma mansoni. J. Biol. Chem. 267 : 5542-5551; 1992. 32. Srivatsan, J.; Smith, D. F.; Cummings, R. D. The human blood fluke Schistosoma mansoni synthesizes glycoproteins containing the Lewis X antigen. J. Biol. Chem. 267: 2019620203; 1992. 33. Mansour, M. H. Structural characterization of Sm 31: A Schistosoma mansoni adult male glycoprotein expressing a fucosyllactose determinant. (Submitted). 34. Parish, C. R. Simple model for self-non-selfdiscrimination in invertebrates. Nature 267:711-713; 1977. 35. Negm, H. I.; Mansour, M. H.; Mohamed, A. H. Biomphalaria alexandrina: Defense mechanism in adult and juvenile snails towards selective Schistosoma mansoni glycoproteins. (Submitted).