Você está na página 1de 7

Eur. J. Biochem.

104, 147-153 (1980)

Sugar-Lectin Interactions : How Does Wheat-Germ Agglutinin Bind Sialoglycoconjugates ?

Michel MONSIGNY, Annie-Claude ROCHE, Claude SENE, Regine MAGET-DANA, and Francis DELMOTTE Centre de Biophysique Moleculaire, Centre National de la Recherche Scientifique, Orleans (Received April 9, 1979)

The specific binding of N-acetylneuraminic acid to wheat-germ agglutinin is based on configurational similarities between N-acetylneuraminic acid and N-acetylglucosamine. The N-acetamido group and an adjacent hydroxyl group, both in an equatorial position are shown to be the main determinants. The N-acetylneuraminic acid - wheat-germ agglutinin interaction is increased by the removal of the last two carbons Cs and C9. The interaction between wheat-germ agglutinin and glycoconjugates containing N-acetylneuraminic acid is shown to be dependent on a charge effect and on an avidity effect. Succinylated wheat-germ agglutinin which is negatively charged at physiological pH, in contrast with wheat-germ agglutinin which is positively charged, does not bind cell surface glycoconjugates containing N-acetylneuraminic acid but does bind cell surface glycoconjugates containing N-acetylglucosamine. The use of wheat-germ agglutinin and of SUCcinylated wheat-germ agglutinin leads to the determination of the number of cell surface receptors containing N-acetylneuramipic acid

Wheat-germ agglutinin is a plant lectin which agglutinates various types of animal cells, of malignant cells and of protease-treated cells [l-31. The agglutination of cells is inhibited by N-acetylglucosamine and its /3l-4 oligomers [2,4]. In addition, wheat-germ agglutinin has been claimed to bind N-acetylneuraminic acid on the basis of equilibrium dialysis [5] and nuclear magnetic resonance [6] studies. Furthermore, the agglutination of neuraminidasetreated cells usually requires a higher concentration of wheat-germ agglutinin than required for untreated cells [2,7] although in some instances the neuraminidase treatment does not lower the cell agglutinability by the lectin [8]. Immobilized wheat germ agglutinin binds membrane sialoglycoproteins [9,10] but fails to bind the related sialoglycopeptides [9]. In the present paper, we wish to report several data showing that N-acetylneuraminic acid, gangliosides and glycoAbbreviations. Glc, glucose; Gal, galactose, NeuAc, N-acetylneuraminic acid ; NeuGc, N-glycoloylneuraminic acid ; Cer, ceramide; GM3,NeuAca2+3Gal~l+4GlcCer; GMz, GalNAc(3tZaNeuAc)~l+4Gal~1+4GlcCer; GMI, Ga1/31+ 3GalNAc(3+2aNeuAc)~l+4Gal~ -4GlcCer; I GDI,, NeuAca2- 3GalP1+3GalNAc(3t2ctNeuAc)fll+4Gal~1-r4GlcCer;GDlb, Galpl -r 3GalNAc(3+2aNeuAcXt2aNeuAc)~l+4Gal~I + 4GlcCer; GTl, NeuAcn2+ 3Gal/11+ 3GalNAc(3 + 2aNeuAcX t 2aNeuAc)flI 4Gal81 +4GlcCer, Nph, p-nitrophenyl-.

proteins containing N-acetylneuraminic acid do bind to wheat-germ agglutinin under specific conditions. We also wish to propose a general mechanism of N-acetylneuraminic acid - wheat-germ agglutinin interactions [ l l ] allowing the interpretation of the reported discrepancies. MATERIALS AND METHODS Wheat-germ agglutinin, prepared as previously described [12] was purchased from IBF-Reactifs (Villeneuve la Garenne, France). N-Acetylglucosamine, p-nitrophenyl-N-acetylglucosaminide, N-acetylneuraminic acid, N-glycolylneuraminic acid, N-acetyl-Lphenylalanine, N-chloroacetyl-L-phenylalanine, bovine serum albumin, bovine submaxillary mucin were obtained from Sigma Chemical Co. Neuraminidase from Vihrio cholerae was purchased from Behringwerke AG. (Marburg, F.R.G .). Gangliosides and periodic-acidtreated gangliosides were a generous gift from R. Veh (Bochum, F.R.G.). Bovine serum albumin substituted or either with diazophenyl-di-N-acetyl-P-chitobioside with diazophenyl-N-acetyl-am-galactosaminide was prepared as previously described [13]. Phosphatidylcholine was isolated from egg yolk according to [14]. Cholesterol from Prolabo (France) was recrystallized twice from methanol. The sialate content of glyco-

Note. All sugars are of the D-configuration.


Wheat-Germ Aggltuinin and Sialoglycoconjugates

conjugates was determined after acid hydrolysis (0.05 M HzS04, 80 "C for 60 min) by Aminoff s method [15] slightly modified. Instead of using n-butanollhydrochloric mixture [I51 to extract the coloured product, dioxane (1 vol.) was added to the reaction solution (1 vol.) leading to a monophasic solution. This modified procedure is easier than the original one, and its sensitivity is not impaired. Succinylated wheat-germ agglutinin was prepared by two treatments (90 min) of wheat-germ agglutinin (20 mg) with succinic anhydride (6 mg) in a sodium acetate buffer at 4"C, followed by dialysis against distilled water [16].
Red-Blood-cell Agglutination Assays

A two fold serial dilution of the proteins was made in 50 pl phosphate-buffered saline and 50 p1 of a 3 % suspension of horse red blood cells using a microtitrator plate (system Cooke, M 220 4A). After 1 h at room temperature the degree of agglutination was assessed with the use of a light microscope. The inhibition assays were carried out as previously described [17].

diluted by addition of 2 ml of distilled ethyleneglycol. The protein solution received four aliquots (0.250 ml) of the fluorescein reagent solution at 3-h intervals and was stirred overnight at 4 "C. The pH of the solution was lowered to 7.5 by addition of 1 M HC1. Fluorescein-labeled wheat germ agglutinin was purified by affinity chromatography on a column (1.5 x 5 cm) of Ultrogel A4 (IBF RCactifs, Villeneuve la Garenne, France) substituted with p-aminobenzylthio-1-Nacetyl-P-D-glucosamine [ 121. The absorbance ratio ( A 4 A Z s 0 )of the labeled lectin obtained by this method was 1.3 corresponding to 2.5 fluorescein molecules bound per molecule of wheat-germ agglutinin ( M , 36000). The succinylated derivative of wheatgerm agglutinin was prepared by two treatments of fluorescein-labeled wheat-germ agglutinin (20 mg) in 5 ml of saturated acetate with succinic anhydride (6 mg) at 4C for 90 rnin [16].
Quantitative Binding Experiments [ I Y ]

Baby Hamster Kidney fibroblasts (BHK 21 C 13, wild type) were obtained from Dr R. C. Hughes (National Institute of National Research, Mill-Hill, London) and grown at 37C in Glasgow modified minimal-essential medium supplemented with 10% (v/v) foetal bovine serum (Rehatuin, F.S. Armour, Comp., U.S.A.) tryptose phosphate broth, NaHC03 (2 g/l) and gentamycin (50 pg/ml) [18]. Cells were removed from culture vessels by incubation for 2 rnin at 37 "C in 0.02 % (w/v) NazEDTA in phosphate-buffered saline pH 7.4. Before use, suspended cells were washed three times with phosphate-buffered saline containing Ca2+ and Mg2+.Cell viability was evaluated by the trypan blue exclusion method.
Neuraminidase Treatment o f Cells

Untreated or neuraminidase-treated cells (2 x lo6 cells/ml) suspended in phosphate-buffered saline pH 7.4 were incubated in the presence of fluoresceinyl wheat-germ agglutinin or of succinylated fluoresceinyl wheat-germ agglutinin (concentration range : 0.5 100 pg/ml) at 4C for 1 h. The cells were spun down (1000 x g, 5 min) and the fluoresceinyl lectin concentration of the supernatant was determined by fluorescence spectrophotometry (free fluoresceinyl lectin). Labeled cells were washed three times at 4 "C with phosphate-buffered saline. The cell-bound fluoresceinyl lectin was released by incubating labeled cells with 0.3 M N-acetylglucosamine (1 ml) at 4 "C for 1 h. The fluoresceinyl lectin concentration of the supernatant was determined by fluorescence spectrophotometry (bound fluoresceinyl lectin). The concentrations of 'free fluoresceinyl lectin' and 'bound fluoresceinyl lectin' were used to draw Scatchard plots [20].
Interaction of Wheat-Germ Agglutinin with Gangliosides

Cells (107/ml) suspended in 0.15 M NaC1, 0.01 M CaC12, 0.05 M sodium acetate buffer pH 5.5, were incubated in the presence of 10 U/ml of Vibrio cholerae neuraminidase (500 U/ml, sialoyl-lactose as substrate) at 37 "C for 30 min. Cells were washed three times with phosphate-buffered saline pH 7.4.
Fluorescein-Labeled Wheat-Germ Agglutinin

Gangliosides were embedded in egg phosphatidylcholine vesicles. Interactions of wheat-germ agglutinin with gangliosides were assessed by measuring the increase of the absorbance at 500 nm of the vesicle suspension upon adding a given amount of lectin [21].
Periodate Treatment 1221

Fluorescein isothiocyanate (2 mg) (Research Organics, Cleveland, OH, U.S.A.) was dissolved in 0.1 ml of freshly distilled dimethylformamide and diluted in 0.9 ml of distilled ethyleneglycol. Wheat germ agglutinin (20 mg) was dissolved in 3 ml of 0.05 M NaC1/0.1 M sodium borate buffer pH 9.3 and

al-Acid glycoprotein (5 mg/ml) or vesicles containing gangliosides (0.5 mg phospholipids per ml) were incubated in phosphate-buffered saline in the presence of 0.001 M sodium periodate at 4 C for 30 min. The excess of sodium periodate was destroyed by addition of ethylene glycol (50 pl/ml). Finally, the

M. Monsigny, A.-C. Roche, C. Sene, R. Maget-Dana, and F. Delmotte

Table 1. Minimal concentration offree ligands required to inhibit the horse red blood cells agglutination induced by wheat-germ agglutinin 120 Pglmli Me-GlcNAc, MezGlcNAc are the 1-0-methyl-P and 1,3-di-0methyl-/l derivatives Sugars Minimal concentration to inhibit agglutination

aldehyde groups were reduced in phosphate-buffered saline pH 8.0 in the presence of NaBH4 (4 mg/ml) at 20 "C for 30 min.

Agglutination Experiments
The minimal concentration of wheat-germ agglutinin required to agglutinate horse red blood cells was found to be 10 pg/ml. Various compounds have been tested in experiments of the inhibition of agglutination induced by wheat-germ agglutinin (Table 1). N-Acetylneuraminic acid and N-acetylneuraminyl-N-acetylgalactosamine inhibited the agglutination by wheat-germ agglutinin, whereas N-glycoloylneuraminic acid was inactive. NAcetyl-L-phenylalanine was active, but the N-chloroacetyl derivative was not. These results show that the N-acetyl group plays an important role in the binding of an inhibitory molecule. When a hydrogen of the acetyl group is replaced by hydroxyl group or by a chlorine atom, the binding is impaired. The dissacharide NeuAc(ct2 - 3)GalNAc was found to be more active than the free N-acetylneuraminic acid which is mainly in a anomeric state. The p-nitrophenyl derivatives of N-acetylglucosamine, of di-Nacetylchitobiose and of N-acetylgalactosamine were found to be active at lower concentration than their non-glycosidic counterparts. This increased activity would be related to the presence of the hydrophobic aglycone. Furthermore, bovine serum albumin substituted with several (four as an average) p-azophenyldi-N-acetylchitobioside or p-azophenyl-N-acetyl-P-Dgalactosamine was much more active than the free p-nitrophenyl-di-N-acetylchitobioside or p-nitrophenyl-N-acetyl-D-galactosamine, respectively. Bovine submaxillary mucin before and after treatment with neuraminidase was equally active as an inhibitor of wheat-germ agglutinin. Because mucin contains mainly disaccharide units : 'NeuAc-GalNAc', native mucin may interact with wheat-germ agglutinin through the N-acetylneuraminate units and the acid-treated mucin through the N-acetylgalactosamine units.

GlcNAc Me-GlcNAc Mez-GlcNAc GlcNAc(B1 -4)GlcNAc NeuAc NeuAc(r2-6)GalNAc NeuGc GalNAc Nph-PGalNAc Nph-ctGalNAc Nph-BGlcNAc Nph-P-[GlcNAc([j1 - 4)GlcNAcI [GlcNAc(B1- 4)GlcNAc]4 bovine serum albumin [mGalNAc]4 bovine serum albumin Bovine submaxillary mucin Desialylated bovine submaxillary mucin ct,-Acid glycoprotein XI -Acid glycoprotein (104-treated) N - Acetyl-L-phenylalanine N-Chloroacetyl-L-phenylalanine

10 3 > 55" 0.1 60 10 >150" 200 3 1 0.6 0.04

0.002 0.012 0.06b 0.06' 0.4h 0.1 10 > 20"

No inhibition of agglutination up to this concentration. Concentration expressed on the basis of NeuAc content. Concentration expressed on the weight basis of native bovine submaxillary mucin

"'J I
6 0.2



: : 0.1

Interaction o f Wheat-Germ Agglutinin with Gangh i d e s

Wheat-germ agglutinin (used in the concentration range 5 - 200 pg/ml) did not aggregate gangliosidefree vesicles, or vesicles containing either GM, and GM3 monosialo-ganglioside or GDlb disialo-ganglioside. On the contrary, wheat-germ agglutinin (but not succinylated wheat-germ agglutinin) aggregated vesicles containing the GDI, disialo-ganglioside and the GT1 trisialo-ganglioside (Fig. 1) and vesicles contain-

0 0 40
Time (min)


Fig. 1. Change in the absorbance at 500 nm oj'a suspension o f vesicles containing GTI gunglioside ufter addition of wheat-germ agglutinin ( ) and of succinylated wheat-germ agglutinin (----). Final lectin concentrations were 100 pg/ml for wheat-germ agglutinin and 200 pg/ml for succinylated wheat-germ agglutinin. The NeuAc: lecithin molar ratio was 0.125. The absorbance of a suspension of vesicles containing gangliosides was compared to that of a reference containing ganglioside-free vesicles. Arrow : addition of GlcNAc(81-4)GlcNAc at a final concentration of 0.5 mg/ml

A. Fluoresceinyl lectin

Wheat-Germ Agglutinin and Sialoglycoconjugates



0 . 1



0 . 4


Lectin bound , b b g )

0 0
2 4 Lectin bound,b (119)

Fig. 2. Scatchard plots of the spec~rficbinding of f l u o w s c ~ e i u j lwhc~cri-germ agglutinin and oj succinylatedfluoresceinyl wheat-germ agglutinin to untreated and neuraminidase-treated baby hamster kidney cells. h : Amount of the fluoresceinyl lectin bound to the cells and specifically released in the presence of 0.3 M N-acetylglucosamine. f:Concentration of the free fluoresceinyl lectin. Cells (2 x lo6) were incubated at 4 C for 1 h in 1 ml of phosphate-buffered saline containing the fluoresceinyl lectin (concentration range: 0.5-100 Fg/ml). ( 0 , O ) Fluoresceinyl lectin; (0,m) fluoresceinyl succinylated lectin; (0,O) binding to untreated cells; (0,m) binding to treated cells

Table 2. Minimal concentration jpglml) of wheatgerm agglutinin required to induce the aggregation of vesicles containing gangliosides Concentrations of phosphatidylcholine : 0.5 mg/ml; ganglioside/ phosphatidylcholine ratio: 0.12 Ganglioside Wheat-germ agglutinin required Fg/ml GMi GM3 GM3 (104-treated) GDi, GDib GTi

Table 3. Binding of wheat-germ agglutinin and of succinylated wheatgerm agglutinin to untreated and neuraminidase-treated B H K 21 C 13 celk, using jhoresceinyl thiocarbamyl derivatives of the lectin Two independent experiments in triplicate. The number of lectin molecules bound per cell, n, and the values of apparent association constants, K , were evaluated after drawing Scatchard plots; the molecular weight of wheat-germ agglutinin and succinyl wheatgerm agglutinin is 36000 [16] Lectin Experiment number Untreated cells Neuraminidasetreated cells
Pxn 10PxK


> 2006
10 200

> 200"


~OPX1 K 0

1 x mol-' Wheatgerm agglutinin 1


I x mol
15 11 1.3 2.1

No aggregation up to this concentration

39 46

1.2 2.0

ing periodate-treated GM3 ganglioside (Table 2). The aggregation could be reversed by addition of N-acetylglucosamine (10 mM final concentration), suggesting that the binding of gangliosides and of N-acetylglucosamine involves the same binding site.

Succinylated wheatgerm 1 agglutinin 2

4.6 4.6

1.2 1.0

3.7 3.5

1.2 1.3

Binding o f Wheat-Germ Agglutinin to Cells

The binding of wheat-germ agglutinin to baby hamster kidney 21C13 cells followed a complex behaviour (Fig. 2) with at least two phases. High affinity binding sites accounted for about half the bound lectin molecules. Biphasic Scatchard plots were found in all cases, i.e. with wheat-germ agglutinin as well as with succinylated wheat-germ agglutinin acting either on native cells or neuraminidase-treated cells. However, the total number of bound wheat-germ

agglutinin molecules dramatically decreased when neuraminidase-treated cells were used in comparison with untreated cells (Table 3). On the contrary, the total number of bound succinylated wheat-germ agglutinin molecules was almost identical with both untreated and neuraminidase-treated cells.

DISCUSSION The molecules of N-acetylglucosamine, of Nacetylgalactosamine and of N-acetylneuraminic acid

M. Monsigny, A.-C. Roche, C. Sene, R. Maget-Dana, and F. Delmotte



N 0



$ 0



Fig. 3 . Structure o f NeuAc( A ) , GlcNAc( BJ mid GulNAc(C)

(Fig. 3) exhibit similarities. These three sugars are all in the pyranose form. The acetamido group (CH3 - CO - NH) and its vicinal hydroxyl group are both in an equatorial position and are in an identical position with reference to the ring oxygen. We may therefore postulate that the ring oxygen, the acetamido group on Cz and the hydroxyl group on C3 of N-acetylglucosamine (the ring oxygen, the acetamido group on Cs and the hydroxyl on C4 of N-acetylneuraminate) are involved in the binding to wheat-germ agglutinin. These conclusions are also supported by the nuclear magnetic resonance data [23] that directly show that the acetamido group of N-acetylglucosamine is involved in the binding, and by the lack of inhibition by 3-0-methyl-N-acetylglucosamine (Table 1) in agreement with [4]. The lack of inhibitory effect of N-glycoloylneuraminate could be related to the hydrophilic property and/or to the size of the glycoloyl group. The lack of inhibitory effect of N-chloroacetyl-Lphenylalanine could be related to the size and or to the partial negative charge of the chloroacetyl group. N-Acetylglucosamine is the best monosaccharide ligand; its hydroxyl group on C4 which is in an equatorial position could also be involved in the binding process. The hydroxyl group on C4 of N-acetylgalactosamine in an axial position partially impairs

the binding of N-acetylgalactosamine to wheat-germ agglutinin [4,24,25]. N-Acetylneuraminate does not have any hydroxyl groups on C3 (corresponding to C4 of N-acetyl glucosamine) in the contrast to both N-acetylglucosamine and N-acetylgalactosamine. The low affinity of N-acetylneuraminate in comparison with that of N-acetylglucosamine could come from its hydrophilic tail, (hydroxyl groups on C7, CS, and C9); As previously shown with N-acetylglucosamine derivatives [26], the N-acetamido group binds wheat-germ agglutinin in subsite B close to a tryptophan residue which is in subsite C ; subsite C is a hydrophobic site. As suggested above on the basis of the structural similarities between N-acetylneuraminic acid and N-acetylglucosamine, the acetamido group of N-acetylneuraminate should interact with subsite B ; so the N-acetylneuraminate hydrophilic tail (C, - C,) is at the level of subsite C inducing a repulsive effect. After mild periodate treatment, the tail is shortened to C7 and the repulsive effect should be weakened. In agreement with this hypothesis, oxidized sialoglycoconjugates exhibited a higher affinity than their unmodified counterparts (Tables 1 and 2). According to the structure of N-acetylneuraminate and to its orientation in the wheat-germ agglutinin binding site, the substituent (R) on Cz of N-acetylneuraminate does not lie in subsites A or C of the wheat-germ agglutinin binding site. So an oligosaccharide containing N-acetylneuraminate should only interact with the wheat-germ agglutinin binding site through its N-acetylneuraminate residue, in contrast with oligosaccharides containing N-acetylglucosamine which may interact with the three subsites. The binding of wheat-germ agglutinin to membranes glycoconjugates containing N-acetylneuraminate is evidenced by the experiments of the aggregation of vesicles containing GT1 and GD1, gangliosides and on the lectin binding to BHK cells before and after neuraminidase treatment. It is noteworthy that vesicles containing either monosialogangliosides or disialogangliosides are not aggregated by wheat-germ agglutinin, but vesicles containing the GDI, and GTI gangliosides are aggregated. The lack of aggregation with the former gangliosides may come from a steric hindrance by the side chain of the disaccharide GalGalNAc. On the contrary, with the latter gangliosides the terminal N-acetylneuraminate is fully accessible. The lack of aggregation of vesicles containing any types of ganglioside by succinylated wheat-germ agglutinin suggests that the modified lectin which is an acidic protein (isoelectric point, pH 4.2) [16], cannot reach the negatively charged vesicles. On the contrary wheat-germ agglutinin is a very basic protein (isoelectric point, pH 8.5 [16,27]), so that a charge effect is expected to act in a positive way to enhance the specific interaction of wheat-germ agglutinin with N-acetylneuraminate. Such a charge effect could be


Wheat-Germ Agglutinin and Sialoglycoconjugates

compared to the interactions of poly-~-lysinewith negatively charged membrane [28 - 301 or to the interaction of Alcian blue with erythrocytes [31]. An alternative hypothesis to explain the lack of aggregation of vesicles containing gangliosides by succinylated wheat-germ agglutinin could be related to the presence of a succinyl group close to the binding site. Such a succinyl group would not impair the binding of neutral glycoconjugates containing N-acetylglucosamine [16], but could impair the binding of acidic glycoconjugates containing N-acetylneuraminate. The binding constant of N-acetylneuraminic acid is very low (kb z lo2 1 x mol-l, [6]) (see Table 1) and as shown above, the other sugars of the oligosaccharides containing N-acetylneuraminate are not expected to give any binding contribution. So, the aggregation induced by 5.5 pM wheat-germ agglutinin on vesicles containing gangliosides ( M 0.1 mM) cannot be explained by such a low binding constant. The apparent binding constant K could derive from a contribution of the specific interaction of wheat-germ agglutinin whith N-acetylneuraminate kb and from a contribution of the charge effect ki and also from an avidity effect k, such as K = kb x k i x k,. Indeed, the actual concentration of the gangliosides at the surface of the vesicles is much higher than if the gangliosides were free in solution, and a wheat-germ-agglutinin molecule bound to a ganglioside has a much higher probability of meeting another ganglioside on the same vesicle than it could be expected with free N-Acetylneuraminate at a similar concentration, so k, should be % 1. This view is supported by the findings that the concentration of free p-nitrophenyldi-N-acetylchitobioside required to inhibit the agglutination of horse red blood cells by wheat-germ agglutinin is much higher than the concentration of p-azophenyl-di-N-acetylchitobioside bound to bovine serum albumin (Table 1). The charge effect is expected to give a ki $ 1 in the interaction between negatively charged vesicles and wheat-germ agglutinin which is basic, and to give a ki I 1 with succinylated wheatgerm agglutinin which is acidic. So, the apparent binding constant K is high enough to allow the aggregation of vesicles containing gangliosides with wheatgerm agglutinin, but is too small to allow the aggregation with succinylated wheat-germ agglutinin. All the three contributions should occur also in the binding of wheat-germ agglutinin to cells. The lower binding constants which can be found using a lectin at a maximum concentration of 10 pM is about lo5 1 x mol-l. So, when the apparent binding constant is lower than 5 x lo4 I x mol-' the lectin-sugar interactions will not be detected. As the binding constant of free oligosaccharides is in the range of lo4 I x mol-l, wheat-germ agglutinin and succinylated wheat-germ agglutinin are expected to bind cell surface glyco-

conjugates with the help of the avidity effect ( K = kb x k J . Wheat-germ agglutinin, but not succinylated wheat-germ agglutinin, will bind cell surface glycoconjugates with the help of the charge effect K = kb

k i x k,.

This view is supported by the findings (a) that the binding constants of the low affinity binding sites are 1 - 2 x lo6 1 x mol-I with both wheat-germ agglutinin and its succinylated derivative, (b) that the number of wheat-germ-agglutinin molecules bound to the untreated cells is much higher than the number of wheat-germ-agglutinin molecules bound to neuraminidase-treated cells, and (c) that the number of succinylated wheat-germ-agglutinin molecules bound to the untreated cells isclose to the number of succinylated wheat-germ-agglutinin molecules bound to neuraminidase-treated cells and not far from the number of wheat-germ-agglutinin molecules bound to neuraminidase-treated cells. The above hypothesis of interaction of wheat-germ agglutinin with N-acetylneuraminate can explain the fact that sialoglycoproteins bind immobilized wheat-germ agglutinin, and that sialoglycopeptides obtained by protease treatment of these glycoproteins fail to bind immobilized wheatgerm agglutinin [9]. In this case the lack of an avidity effect leads to a decrease of the apparent binding constant of the sialoglycoproteins by a factor l/k,. In conclusion, the binding of glycoconjugates containing N-acetylneuraminate to wheat-germ agglutinin involved three types of interactions. a) A specific interaction between N-acetylneuraminate and the subsite B of wheat-germ agglutinin (kb) because the binding of sialoglycoconjugates to wheat-germ agglutinin can be reverted by addition of N-acetylglucosamine and because of the structural similarities of N-acetylneuraminate and N-acetylglucosamine. b) A charge effect (k,) because wheat-germ agglutinin has a high isoeletric point. c) An avidity effect (k,) because wheat-germ agglutinin does not bind free sialoglycopeptides. Finally, in the quantitative studies of cell surface glycoconjugates, the number of receptors containing N-acetylglucosamine can be deduced from the amount of bound succinylated wheat-germ agglutinin and the number of receptors containing N-acetylneuraminate from the amounts of bound wheat-germ agglutinin and of succinylated wheat-germ agglutinin, because wheat-germ agglutinin binds both N-acetylneuraminate and N-acetylglucosamine residues. After this paper was submitted, Bhavanandan and Katlic [32] reported similar studies on the role of sialic acid in the interactions of wheat germ agglutinin with sialoglycoproteins. These authors pointed out the specific role of the acetamido group, the structural similarities of N-acetylneuraminic acid and N-acetylglucosamine and the influence of the density of binding

M. Monsigny, A.-C. Rochc, C. Sene, R. Maget-Dana, and F. Delmotte

153 13. Kitda, C., Delmotte, F. & Monsigny, M. (1977) FEBSLett. 76, 257 - 261. 14. Singleton, W. S., Gray, M. S., Bown, M. L. & White, J . L. (1965) J . Am. Oil Chem. Soc. 42, 53. 15. Aminoff, D. (1961) Biochem. J . 81, 384-392. 16. Monsigny, M., Sene, C., Obrenovitch, A,, Roche, A. C., Delmotte, F. B Boschetti, E. (1979) Eur. J . Biochem. 98, 39-45. 17. Roche, A. C. & Monsigny, M. (1974) Biochim. Biophys. Acta, 371, 242- 254. 18. Meager, A,, Ungkitchanukit, A. & Hughes, R. C. (1976) Biochem. J . 154, 113-124. 19. Monsigny, M., Sene, C. B Obrenovitch, A. (1979) Eur. J . Biochem. 96, 295 300. 20. Scatchard, G. (1949) Ann. N.Y. Acad. Sci. 51, 660-672. 21. Maget-Dana, R., Roche, A. C. & Monsigny, M. (1977) FEBS Left. 79, 305-308. 22. Suttajit, M. & Winzler, R. J. (1971) J . Biol. Chem. 246, 33983404. 23. Grivet, J. Ph., Delmotte, F. & Monsigny, M. (1978) FEBS Letf.88, 176-180. 24. Goldstein, I. J., Hammarstrom, S. & Sunblad, G . (1975) Biochim. Biophys. Acta, 405, 53-61. 25. Privat, J. P., Delmotte, F. & Monsigny, M. (1974) FEBS Lett. 46, 229 232. 26. Monsigny, M., Dclmotte, F. & Helene, C. (1978) Proc. Nut1 Acad. Sci. U.S.A. 75, 1324-1328. 27. Rice, R. H. B Elzler, M. E. (1975) Biochemistry, 14, 40934099. 28. Katchalsky, A,, Danon, D., Nevo, A. & DeVries, A. (1959) Biochim. Biophys. Acta, 33, 120- 132. 29. Deman, J. J. & Bruyneel, E. A. (1974) Exp. Cell Res. 89, 206216. 30. Jacobson, B. S. & Branton, D. (1976) Science (Wash. DC),195, 302- 304. 31. Geyer, G., Halbhuber; K. J. B Feuerstein H., (1976) Acta Histochem. 55, 317 - 324. 32. Bhavanandan, V. P. & Katlic, A. W. (1979) J . B i d . Chem. 254, 4000 - 4008.
~ ~

sugar residues on the glycoproteins, in good agreement with our findings.

We thank Mrs Marie-Louise Welter and Mr Philippe Bouchard for valuable technical assistance. This work was supported by grants C.L. 7540743 from Institut National de la Recherche et de la SantC, A.C.C. 7770252 and A.C.C. 7871086 from DCltgation GtnCrale d la Recherche Scientifique et Technique. A.C.R. is ChargCe de Recherche of the Institut National de la Recherche et de la SantP, R.M.-D. is ChargCe de Recherche of the Centre National de la Recherche Scientifique.

1 Aub, J. C., Tieslau, C. & Lankester, A. (1963) Proc. Natl Acad. Sci. U.S.A. 50,613-619. 2 Burger, M. M. & Goldberg, A. R. (1967) Proc. Nut1 Acad. Sci. U.S.A. 57, 359-366. 3 Burger, M. M. (1969) Proc. Natl Acad. Sci. U.S.A. 62, 9941001. 4 Allen, A. K., Neuberger, A. B Sharon, N. (1973) Biochem. J . 131, 155-162. 5 Greenaway, P. J. & LeVine, D . (1973) Nut. New Biol. 241, 191 - 192. 6 Jordan, F., Basset, E. B Redwood, W. R. (1977) Biochem. Biophys. Res. Commun. 75, 1015 - 1021. 7 Nicolson, G. L., Lacorbiere, M. & Eckhart, W. (1975) Biochemistry, 14, 172 - 179. 8. Schnebli, H. P., Roeder, C. & Tarcsay, L. (1976) Exp. Cell. Rex 98, 273 - 276. 9. Bhavanandan, U. P., Umemoto, J., Banks, J. R. B Davidson, E. A. (1977) Biochemistry, 16, 4426-4437. 10. Adair, W. L. & Kornfeld,, S. (1974) J . Bid. Chem. 249, 46964704. 11. Roche, A. C. (1978) Thesis (Doctorat es Sciences) Orleans, France. 12. Bouchard, P., Moroux, Y.,Tixier, R., Privat, J. P. & Monsigny, M. (1976) Biochimie (Paris) 58, 1247-1253.

M. Monsigny, A-C. Roche, C. Sene, R. Maget-Dana, and F. Delmotte Centre de Biophysique Moltculaire du Centre National de Recherche Scientifique, 1A Avenue de la Recherche Scientifique, La Source, F-45045 Orleans-Cedex, France