Effect of blood group determinants on binding of human salivary mucous glycoproteins to influenza virus

127

Biochimica et Biophysica Acta, 540 (1978) 127--133 © Elsevier/North-Holland Biomedical Press

BBA 28486

E F F E C T OF B L O O D G R O U P DETERMINANTS ON BINDING OF HUMAN S A L I V A R Y MUCOUS GLYCOPROTEINS TO I N F L U E N Z A VIRUS

THOMAS F. BOAT, JAMES DAVIS, ROBERT C. STERN and PI WAN CHENG

Department of Pediatrics, Rainbow Babies and Childrens Hospital, Case Western Reserve University, Cleveland, Ohio 44106 (U.S.A.) (Received July 25th, 1977)

Summary We have demonstrated that the inhibitor of influenza B virus hemagglutination in human saliva is inactivated by neuraminidase and is associated with the mucous glycoprotein fraction (blood group substance) of this secretion. Inhibitory activity of saliva was found to be roughly proportional to its sialic acid content (r = 0.456). However, the minimal quantity of salivary sialic acid, neutral sugar, or blood group antigen required to inhibit virus hemagglutination was greater for secretors of A and B than for secretors of H and Le a blood group substances. Removal of terminal galactose from blood group B substance with a-galactosidase markedly decreased blood group B activity b u t increased blood group H and virus hemagglutination inhibitory activities of this glycoprotein. These data suggest that terminal a-linked galactose and, probably, N-acetyl-galactosamine interfere with access of influenza virus to binding sites on oligosaccharide chains of the mucous glycoprotein.

Introduction Sialoglycoproteins in serum and mucous secretions inhibit myxovirus hemagglutination and may be chemical analogues of virus receptors in the red blood cell membrane [1]. Neuraminidase treatment eliminates this virus hemagglutination (V.H.) inhibition activity [2], suggesting that virus binding sites include, and in fact m a y be, the neuraminic acid residues occupying terminal positions on oligosaccharide chains [1]. Virus binding is specific for N-acetylneuraminic acid. Pig submaxillary mucin which contains N-glycosyl neuraminic acid b u t n o t the N-acetyl derivative has no virus hemagglutination inhibition activity [1]. The inhibition activity of these glycoproteins is related to their neuraminic acid content and molecular size [3,4]. The extent to which glycoprotein and viral surfaces are complementary also affects their binding properties [ 5].

128

It was the purpose of this study to characterize the interactions between mucous glycoproteins of human salivary secretions and myxoviruses. We have (I) confirmed that human salivary mucous glycoproteins are potent inhibitors of influenza B virus hemagglutination, and (2) demonstrated that V.H. inhibition activity of human salivary mucous glycoproteins is influenced by the blood group determinants which they carry. Materials and Methods

Materials Influenza B virus, Taiwan strain t V F295) was obtained from American T y p e Culture Collection, Rockville, Md. H um an anti-A and anti-B sera were obtained from Hyland Laboratories, Costa Mesa, Calif. Anti-Le ~ serum was obtained from a blood t y p e O hum an donor. Anti-H lectin was prepared by saline extraction o f Ulex europeus seeds (F.W. Schumacher, Sandwich, Mass.}. Clostridium perfringens neuraminidase. T y p e VI, was obtained from Sigma, St. Louis, Mo. This lot of neuraminidase contained little protease activity, equivalent to less than 0.001% (w/w} of crystallized bovine pancreas r.rypsin (Sigma) as assessed by hydrolysis of Azocoll (Calbiochem). Aipha-Dgalactosidase from anaerobic cultures of Rurninococcus AB [6] was a gift of Dr. Lansing Hoskins, D e p a r t m e n t of Medicine, Case Western Reserve University. Th e galactosidase preparation contained no neuraminidase activity and did n o t h y d r o l y z e Azocoll during a 24-h incubation period. It released t 0 nmot of galactose from 4 mg of lyophilized human saliva for each 2-fold decrease of blood group B titer [6]. Saliva was collected w i t h o u t stimulation from 50 healthy subjects (25 male, 25 female; ages 16--42 years) by e x p e c t o r a t i o n at least 1 h after eating. Saliva samples were immediately centrifuged at low speed to remove particulate matter, and th en frozen. Methods Virus hemagglutination inhibition activity and blood group A, B~ H or Le ~ activity in saliva were assayed by microtitration [7]. Two hemagglutmating units o f virus, blood group antiserum, or H-specific lectin were used in these assays. For V.H. inhibition determinations, the influenza virus was preheated at 56°C for 30 rain. A 2% suspension of chicken red blood cells and h u m a n red blood cells of appropriate type were used for hemaggtutination with virus and blood group antisera or lectin, respectively. T y p e O red cells used for assay of Le a substance were pretreated with ficin [8]. Saliva samples which were used for the d eter min a t i on of blood group titers had been heated at 100°C for 15 min b u t those used for the det er m i nat i on of V.H. inhibition activity were n o t heat-treated. 50 t~l of saliva was titered in all assays. For colorimetric analyses, an aliquot of each whole saliva sample was extensively dialyzed against distilled water, lyophilized, and reconst i t ut ed to original volume with distilled water. Protein was determined by the m e t h o d of L ow ry [9] with bovine serum albumin as standard. Sialic acid was determined by the thiobarbituric acid reaction [ 10], and neutral sugar by the ant hrone procedure [11] with galactose as standard. Lysozyme was assayed by the m e t h o d of

129 Osserman [12] and amylase by the m e t h o d of Bernfeld [13]. Sodium dodecyl sulfate 7.5% acrylamide gel electrophoresis was carried out as outlined by Weber and co-workers [14]. Gels were stained with Coomassie blue and periodic acid-Schiff [15]. Agarose gel chromatography using Bio Gel A-5m was carried out as previously described [16]. 1 ml of saliva diluted 1 : 1 with 0.20 M acetate buffer (pH 5.0) was treated with 0.02 units of C. perfringens neuraminidase for 5 min at 37°C. A sample which was treated similarly, but w i t h o u t neuraminidase, was used as a control. Bot]h samples were boiled for 1.5 min after incubation to destroy neuraminidase activity and then were dialyzed against isotonic phosphate buffer saline for virus hemagglutination inhibition titration. Treatment of whole saliva with a-D-galactosidase was carried out as described by Hoskins and Boulding [6]. Equal volumes of culture filtrate and saliva, both dialyzed against 0.02 M phosphate buffer (pH 6.4), were incubated at 37°C for 4 h. The reaction was terminated by placing the reaction mixture in boiling water for 2 min. Results Virus hemagglutination inhibition activity in human saliva was characterized with respect to neuraminidase susceptibility. Inhibition titers of saliva from four donors, secretors of A, B, H, and Le a blood group substances, decreased by at least six serial dilutions after a 5 rain incubation with C. perfringens neuraminidase. This drop in titer indicates that at least 98% of V.H. inhibition activity in h u m a n saliva is neuraminidase sensitive. Inhibition activity was then purified on a Bio Gel A-5m column (Fig. 1). All detectable V.H. inhibition activity and blood group substances were eluted together at the void volume of the column. This material contained more than 80% of the sialic acid and more than 60% of the neutral sugar in saliva. The high molecular weight sialoglycoprotein fraction was well resolved from salivary lysozyme and amylase. It contained no low molecular weight proteins or glycoproteins when subjected to sodium dodecyl sulfate 7.5% acrylamide gel electrophoresis. The mucous glycoprotein scarcely penetrated the gel, but was detected with periodic acid-Schiff stain in the first 1--2 mm of the gel. These results indicate that virtually all V.H. inhibition activity a n d most of the nondialyzable sialic acid in human saliva reside in the mucous glycoprotein fraction. Subsequent studies were performed on unfractionated saliva samples. The V.H. inhibition titer of saliva from 50 donors was significantly correlated with the sialic acid c o n t e n t of the saliva (r = 0 . 4 5 6 ; P < 0.01). However, a considerable scatter of data was noted and appeared to be related to the blood type and secretory status of the donors. Table I shows that the minimal a m o u n t of saliwLry sialic acid required to inhibit virus hemagglutination is significantly greater ( P < 0.001) for samples from type A and type B secretors than for samples from type H secretors or nonsecretors (Le a secretors). Saliva samples from two AB secretors had the same V.H. inhibition activity as did A and B substances. The mean sialic acid content of dialyzed saliva samples from secretors of the four blood group substances did not vary significantly (Table II). Similar differences of V.H. inhibition potency related to the type of blood group sub-

130 Fractions containing blood group B substance I

n0

-[

Fractions actwlty

containing

Fractions activity.

containing

VH[ -t50 lysozyme

Fractions containing amylase activity.

A

--IZ5

--i2.5 Z © FO ,< kL 75

v.6 E ¢q C~ .4

/

\,

/

200

/

300 ELUTION

\--50

400

::L

d (3

500

VOLUME (rnl ]

F i g . 1. B i o G e l A - 5 m c h r o m a t o g r a p h y o f u n s t i m u l a t e d s a l i v a w h i c h w a s c o n c e n t r a t e d i 0 - f o l d . T h e s a l i v a d o n o r was a secretor of t y p e B b l o o d g r o u p s u b s t a n c e . F o u r m l of t h e c o n c e n t r a t e d saliva, p r e v i o u s l y d i a l y z e d a g a i n s t 5 0 m M NaC1. 1 0 m l T r i s - HC1 ( p H 7 . 0 ) , a n d 1 0 m M E D T A , was e l u t e d b y u p w a r d f l o w a t 2 0 m l / h w i t h 5 0 m M NaC1. t 0 m M Tris. C o l u m n w a s 8 0 × 2 . 5 era. 85---90% o f o r i g i n a l p r o t e i n a n d sialic acid was recovered from the column. Alternate fractions were assayed for blood group substance_ v i r u s h e m a g g l u t i n a t i o n i n h i b i t i o n (V.H.I.) a c t i v i t y , sialic acid, n e u t r a l sugar, lysozyme~ a n d amylase.

stance secreted were found when data were expressed as the minimal quantity of neutral sugar, total protein or blood group substance in saliva required co

inhibit red celt agglutination. Additional evidence for a role of blood group determinants in the variation of inhibition activity of salivary mucous glycoproteins was obtahaed by gtycosidase t r e a t m e n t of the glycoprotein. Salivas from four blood group B secretors were treated with c~-galactosidase. In all four cases, the titer of blood group B substance decreased at least 64-fold, the titer of blood group H substance increased at least 4-fold, and V.H. inhibition titer increased 4--8-fotd I Table Iii). Heat inactivated ~-galactosidase alone did n o t inhibit red cell agglutination by TABLE I GLYCOPROTEIN SIALIC ACID REQUIRED FOR INHIBITION OF RED BLOOD CELL AGGLUTINATION BY TWO HEAMGGLUTINATING UNITS OF INFLUENZA B VIRUS Secretor type

Number of saliva samples

Average minimal amount of sialic a c i d f o r i n h i b i t i o n (/zg ± S . E . )

A B H Le a

15 fl 12 15

0.41 0.49 9.09 0.20

~* 0 . ! 2 -+ 0 , 1 1 -~ 0 . 0 5 ~_ 0 . 0 7

131

T A B L E II SIAL][C A C I D C O N T E N T STANCES

OF SALIVA FROM SECRETORS

O F A , B, H , a n d L e a B L O O D G R O U P

Secretor type

Number of saliva samples

M e a n sialic a c i d c o n c e n t r a t i o n (#g/ml ± S.E.)

A B H Le a

15 8 12 15

35 23 22 40

± + + ±

SUB-

22 16 20 25

TABLE III THE

EFFECT

INHIBITION SUBSTANCE

OF

c~-GALACTOSIDASE

("V.H. INHIBITION")

Donor

ON BLOOD

GROUP

AND

VIRUS

HEMAGGLUTINATION

TITERS OF SALIVA SAMPLES CONTAINING

BLOOD GROUP B

Titers Before enzyme treatment

After enzyme treatment

1 B substance H sabstance V.H. i n h i b i t i o n

1 • 210 1 • 26 1 - 24

1 • 24 1 • 28 1 • 26

2 B substance H substance V.H. inhibition

1 • 28 1 • 26 1 • 23

1 • 21 1 • 28 1 • 26

3 B substance H substance V.H. inhibition

1 . 28 I • 25 1 . 21

1 • 21 1 • 28 1 . 24

4 B substance H substance V.H. inhibition

1 • 28 1 • 25 1 • 21

1 • 22 1 • 28 1 • 24

either anti-B serum or by influenza virus. The alteration of B and H blood group substance titers which we observed is similar to the alteration obtained by treating human gastric mucin with this B-degrading enzyme [6]. We have also studied the effect of different human red blood cell types on the direct hemagglutination of these cells with influenza B virus. Erythrocytes from three blood type A, three blood type B, and three blood type O donors were reacted with serial dilutions of virus. No blood type-related difference of hemagglutination titers was observed. Discussion All virus hemagglutination inhibition activity and 80% of the sialic acid in human saliva reside in the high molecular weight, blood-group-specific glycoproteins of this secretion. These findings permitted the use of crude saliva samples ~Lo study the determinants of inhibition activity in salivary mucous glycoproteins. We have failed to demonstrate a 1 : 1 relationship between the sialic acid content of saliva and inhibition activity. Rather, V.H. inhibition activity is

132

also determined by the type of blood group~specific substance secreted. V oH. inhibition activity is significantly greater for H and Le a blood group substances than for A, B, and AB blood group substances and removal of galactose, the determinant for blood group B substance, enhances inhibition activity of this glycoprotein. The number of glycoprotein sialie acid residues available for attachment to hemagglutination sites of myxoviruses appears to determine the inhibitory potency and binding affinity of the glycoprotein [5]~ Thus, sialic acid content and size of glycoproteins are well known factors which determine their V.H. inhibition properties [ 3] o However, the variability of interaction between different salivary blood group substances and influenza B virus does not appear to be the result of large differences in molecular weight of these glycoproteins. Blood group H substance can be synthesized by adding a single ~-iinked fucosyl resi~ due to the terminal galactose of oligosaccharides in Le a substance. A and B substances are synthesized by adding c~-linked N-acetylgalaetosamine and galactose respectively to the penultimate gaiactose residue of oligosaccharides in H substance [17]. These additions to the oligosaceharide chains would result in only minor changes of gtycoprotein molecular weight and would not alter the numo her of virus receptor sites. Similarly, the data in Table Ii suggest strongly that sialic acid content is not responsible for the differences of V.Ho inhibition activity in blood group A and B substances compared with H and Le a substances. The most plausible explanation for our findings is that the addition of gataetose or N-acetylgalaetosamine to termini of H-type oligosaceharide chains interferes with the access of influenza virus to their sialic acid containing binding sites on these or adjacent chains. The position of sialic acid in human salivalT~7 glycoprotein oligosaccharide chains is n o t known. By analogy to the MN glycoprotein of the erythrocyte membrane or ovine submaxi!lary mucin [1,18] the sialic acid which participates in virus binding may be linked to galactose or N-acetylgalaetosamine residues positioned at the reducing end of the oligosaccharide chains. Credence for this suggestion is also derived from the observation that myxovirus receptor activity cannot be separated from MN antigens [19]. increasing the length or bulk of the sialic acid containing, or adjacent oligosaceharide chains m a y impose a steric hindrance to virus interaction with its glycoprotein receptor. Our observations suggest one mechanism whereby the 'complementariness' [ 5] of glycoprotein and virus surfaces may be altered° Failure to demonstrate differences of influenza virus agglutination of A and B, as compared with O, human red cells is n o t surpmsing in that a substanti~ number of these red blood cell group determinants are associated with ms m~ brahe glycolipids [20] while virus binding sites are located on the membrane MN glycoprotein [18]. The biological significance of these findings is unclear. If, as claimed by Ross and Wallace [21], small children with O blood type are more susceptible to respiratory infections with influenza virus, then enhanced binding of virus ~o respiratory mucous glyeoproteins or epithelial plasma membrane glyeoprotems may be deleterious. This possibility requires further study.

133

Acknowledgements T:he authors wish to thank Dr. Lansing Hoskins for the a-galactosidase used in this study. This work was supported in part by grants HL14844 and HR52957 from the United States Public Health Service and a grant from the Health Fund of Cleveland. References 1 Gottschalk, A., Belyvin, G. and Biddle, F, (1972) in Glycoprotelns: Their Composition, Structure and F u n c t i o n (Gottschalk, A., ed.), pp. 1082--1096, Elsevier, New Y ork 2 Gottschalk, A. (1957) Biochim. Biophys. Acta 2 3 , 6 4 5 - - 6 4 6 3 Springer, G.F., Schwick, H.G. and Fletcher, M.A. (1969) Proc. Natl. Acad. Sci. U.S. 64, 634--641 4 Morawiecki, A. amd Lisowska, E. (1965) Biochem. Biophys. Res. Commun. 18, 606--610 5 Fazekas de St. Groth, S. and Gottschalk, A. (1963) Biochim. Biophys. Acta 78, 248--257 6 Hoskins, L.C. and Boulding, E.T. (1976) J. Clin. Invest. 57, 63--73 7 Kabat, E.A. and Mayer, M.M. (1961) E x p e r i m e n t a l I m m u n o c h e m i s t r y , p. 127, C.C. Thomas, Springfield, Ill. 8 Marcus, D,M. and Grollman, A.P. (1966) J. I m m u n o l . 9 7 , 8 6 7 - - 8 7 5 9 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265--275 10 Warren, L. (1959) J. Biol. Chem. 234, 1971--1975 11 Spiro, R.G. (1966) Methods Enzymol. 8, 4--5 12 Osserman, E.F. and Lawlor, D.P. (1967) J. Exp. Med. 124, 921--951 13 Bernfeld, P. (1955) Methods Enzymol, 1, 149--158 14 Weber, K., Pringle, J.R. and Osborn, M. (1972) Methods Enzymol. 26, 3--27 15 Glossman, H. and Neville~ Jr., D.M. (1971) J. Biol. Chem. 246, 6339---6346 16 Boat, T.F., Cheng, P.W., Iyer, R., Carlson, D.M. and Polony, I. (1976) Arch. Bioehem. Biophys. 177, 95 --104 17 Watkins, W.M. (1966) Science 152, 172--181 18 Kathan, R.H. and Winzler, R.J, (1963) J. Biol. Chem. 238, 21--25 19 Springer, G.F., Huprikar, S.V. and Tegtmeyer, H. (1971) in Glycoproteins of Blood Cells and Plasma (Jamieson, G.A. and Greenwalt, T.J., eds.), pp. 35--49~ Lippincott, Philadelphia 20 Koscielak, J. (1963) Biochim. Biophys. Acta 78, 313--328 21 Ross, C.A.C. and Wallace, J. (1974) Lancet 1 , 3 1 4