Download Characterization of Membrane Components of the Erythrocyte

J. sen. Virol. (1987), 68, 2359-2369.
Printed in Great Britain
2359
Key words: VSV/erythrocyte membrane/haemagglutination/haernolysis
Characterization of Membrane Components of the Erythrocyte Involved in
Vesicular Stomatitis Virus Attachment and Fusion at Acidic pH
By P A O L A M A S T R O M A R I N O , l* C I N Z I A C O N T I , 1 P A O L A G O L D O N I , z
B E R N A R D H A U T T E C O E U R 2 AND N I C O L A O R S I 1
l lstituto di Microbiologia, Facoltgt di Medicina, Universith 'La Sapienza', Piazzale Aldo Moro 5,
00185 Roma, Italy" and Zlnstitut Pasteur, Rue du Dr Roux, 75724 Paris Cedex 15, France
(Accepted 1 June 1987)
SUMMARY
Goose erythrocyte membranes were isolated and tested for their ability to compete
with red cell receptors for vesicular stomatitis virus (VSV) attachment and fusion at
acidic pH. Crude membranes, solubilized with Triton X-100, Tween 80 and octyl-/~-Dglucopyranoside, showed a dose-dependent inhibitory effect on virus binding and
haemolysis. The chemical nature of the active molecules was investigated by enzyme
digestion and by separation of purified components. Only the lipid moiety, specifically
phospholipid and glycolipid, was found to inhibit VSV attachment; a more detailed
analysis of these molecules showed that phosphatidylinositol, phosphatidylserine and
GM3 ganglioside were responsible for the inhibitory activity and could therefore
represent VSV binding sites on goose erythrocyte membranes. Removal of negatively
charged groups from these molecules by enzymic treatment significantly reduced their
activity, suggesting that electrostatic interactions play an important role in the binding
of VSV to the cell surface. Enzymic digestion of whole erythrocytes confirmed the
involvement of membrane lipid molecules in the cell surface receptor for VSV.
INTRODUCTION
During the initial phase of infection, the interaction of viruses with their host cells leads to
several events such as attachment, endocytosis and/or membrane fusion. Attachment to cells is
believed to involve an interaction between viruses and specific components of the plasma
membrane. Most enveloped viruses (Dales, 1973; Lonberg-Holm & Philipson, 1974) enter cells
by adsorptive endocytosis and are subsequently delivered to intracellular vacuoles and
lysosomes. The final and critical step in the penetration of the viral genome into the cytoplasm is
a low pH-induced membrane fusion event between the viral and the lysosomal membranes.
Viral haemagglutination (HA) is equivalent to the first step of infection (attachment) while
haemolysis (He) is a useful model for viral fusion with plasma or lysosomal membranes.
Vesicular stomatitis virus (VSV) binds to goose erythrocyte membrane producing HA, He and
fusion (Mifune et al., 1982) at a mildly acidic pH. The interaction of VSV with the erythrocyte
membrane therefore mimics what happens in the lysosomes where the low pH environment
triggers the fusion. The He process seems to be mediated by a hydrophobic peptide segment of
the G envelope glycoprotein of VSV (Schlegel & Wade, 1984), but the cellular counterpart
involved in this phenomenon has not yet been characterized. For Vero cells supporting viral
replication, Schlegel et al. (1983) demonstrated that the chloroform-methanol fraction of
solubilized cell membranes inhibits the binding of VSV to saturable, high-affinity sites on the
cell surface, and phosphatidylserine appeared to be the most potent lipid tested. Infectivity and
HA of rhabdoviruses are inhibited by phospholipids and by the lipid component of low density
lipoproteins, probably because of a chemical similarity between these compounds and the
receptors for the virus in the cell membrane (Halonen et al., 1974; Seganti et al., 1983 ; Superti et
al., 1984).
0000-7756 © 1987 SGM
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 11 May 2017 07:30:43
2360
p. MASTROMARINO AND OTHERS
W e h a v e s o l u b i l i z e d goose e r y t h r o c y t e m e m b r a n e s a n d s t u d i e d t h e i r i n h i b i t o r y a c t i v i t y o n
e i t h e r V S V a t t a c h m e n t or its f u s i o n w i t h r e d b l o o d cells. I n a d d i t i o n , we i s o l a t e d p r o t e i n , lipid,
phospholipid and ganglioside components of membranes and identified those showing receptor
a c t i v i t y for V S V . T o a s c e r t a i n t h e role o f v a r i o u s c h e m i c a l g r o u p s , we a n a l y s e d t h e effect o f
enzymic modification of whole erythrocytes, solubilized membranes and their components on
t h e b i n d i n g a n d f u s i o n a c t i v i t i e s o f VSV.
METHODS
Virus. VSV, Indiana serotype, was grown in CER (chicken embryo-related) cells, pelleted at 80000g for 2 h and
resuspended in bovine albumin borate saline buffer (BABS) pH 9.0.
Haemagglutination. HA studies were carried out in 96-well (V-shaped bottom) microtitre plates. An equal
volume of 1 ~o goose erythrocyte suspension in 0.2 M-phosphate saline buffer pH 5-0 was added to twofold serial
dilutions of VSV in BABS. In HA inhibition tests twofold dilutions of inhibitor in BABS were mixed with an equal
volume of virus containing 4 HA units (HAU). After 2 h incubation at 4 °C red blood cells were added, and the
plates were kept at 4 °C over crushed ice for 3 h, before titres were recorded.
Haemolysis. 0.5 ml of 1~ goose erythrocytes in phosphate buffer pH 5-0 was added to 0.5 ml of a virus dilution in
BABS. The tubes were incubated for 30 rain at 4 °C, followed by 40 min at 23 °C. The samples were then
centrifuged at 170 g for 10 rain at 4 °C, and the absorbance of the supernatant was measured at 540 nm. The results
were expressed as ~ He compared with the maximal 100K obtained by treatment with 0-1 ~ NP40. One
haemolytic unit (HeU) was the viral dilution giving 5 0 ~ He. In the He inhibition test 250 gl of twofold dilutions of
inhibitor were mixed with 250 gl of virus containing 8 HeU/ml. Goose erythrocytes were added after 2 h
incubation at 4 °C over crushed ice. The results were expressed as percent of the maximum He obtained without
inhibitor.
Enzyme treatment o f erythrocytes. Enzymic treatments were carried out using a 10~ suspension of goose
erythrocytes washed three times in PBS. The cells were incubated at 37 °C for 1 h, in the presence or absence of
various enzymes, solubilized in PBS. They were collected by centrifugation and washed three times with PBS
before being used for virus HA and He studies. Only cells treated with phospholipase Az were incubated for 5 min
at 23 °C with 100 mg/ml bovine serum albumin (BSA) to extract the reaction products (Haest et at., 1981).
Isolation and solubilization o f goose erythroeyte membranes. Goose erythrocyte membranes were prepared by the
method of Ginsberg et al. (1976). Red cells were lysed by hypotonic shock in a buffer containing 8-5 mM-Tris-HC1
pH 7.2, 3.0 mM-NaC1, 1.0 raM-glucose, 0"1 mg/ml BSA, 2 mM-MgC12. The haemoglobin-free pellet obtained after
several washings was homogenized in a Dounce homogenizer (tight pestle). The membrane fraction obtained after
sedimentation at 5000 g for 10 min was purified three times by sedimentation at 15000 g. The purified membrane
fraction, containing only membrane vesicles when viewed by phase-contrast microscopy, was resuspended in PBS
at a protein concentration of 4 mg/ml.
Membrane solubilization was performed according to the method of VandenBerg et al. (1983) by the addition of
4.0 ~ octyl-/~-o-glucopyranoside, 3 ~ Triton X-100, 1~ Tween 80 in PBS, followed by sonication. The suspension
was centrifuged at 100000 g for 1 h, and the supernatant was dialysed overnight at 4 °C against PBS. Residual
detergent was then removed by treatment with Bio-Beads SM-2. The dialysed, adsorbed material was considered
as the solubilized erythrocyte membrane (SEM).
Solubitization o f membrane proteins. Proteins were extracted from isolated plasma membranes using n-butanol
according to the method of Maddy (1966). Membranes (10 mg/ml protein) in hypotonic buffer containing 0.2 mMMgC12 were mixed with 0.75 vol. butanol and shaken thoroughly for 20 s. After incubation for 20 rain on ice, the
samples were centrifuged at 2500 g for 10 rain, the lower aqueous phase was collected and extensively dialysed
against PBS at 4 °C.
Isolation o f membrane lipids. Total lipids were extracted from membranes with chloroform-methanol (C:M;
1 : 1.5 v/v) for 1 h at 4 °C, followed by extraction with C :M (2 : 1 v/v) for 5 min and C : M : H 2 0 (32:64:5 v/v/v) for
5 rain. The two last steps were repeated twice, the first time at 55 °C, and the second time at 37 °C. The extracts
were combined and dried under nitrogen. Lipids were washed again with C : M (2 : 1 v/v), the supernatant was
dried under nitrogen and resuspended in PBS by sonication.
To obtain phospholipids the dried residues were dissolved in C : M (2 : 1 v/v) and partitioned by adding 0.2 vol.
distilled water according to Folch et al. (1957). Under these conditions, phospholipids separated into the lower
phase and were solubilized by sonication in PBS before use.
. Isolation o f membrane glyeolipids. Glycolipids were prepared from goose erythrocyte ghosts obtained according
to Hakomori & Watanabe (1976). The packed ghosts were extracted three times with 500 vol. C :M :H20 (4 : 8:3
v/v/v) (Svennerholm & Fredman, 1980). The extracts were combined and evaporated to dryness, dissolved in a
small volume of C : M (1 : 1 v/v) and centrifuged. The gangliosides in the supernatant were isolated according to
Ladish & Gillard (1985) and separated from neutral glycolipids by ion-exchange chromatography on D E A E Sephadex. Gangliosides were further purified by gel filtration on a column of Sephadex LH-20 equilibrated with
C : M : H 2 0 (5:5:1 v/v/v) (Byrne
et al., 1985).
Downloaded
from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 11 May 2017 07:30:43
Erythrocyte receptor for V S V at acidic p H
2361
The ganglioside fraction was separated into two different classes by column chromatography on Silica gel 60
(Merck) using C : M : H20 (62 : 25 : 4 v/v/v) as the solvent system. High-performance thin-layer chromatography of
purified goose red blood cell gangliosides was performed on precoated Silica gel 60 activated at 120 °C for 1 h, and
developed in C : M : 5 M-NH4OH : 27 mM-CaC12.2H20 (60:40 :4 : 5 v/v/v/v). Gangliosideswere visualized with the
resorcinol-HC1 reagent as purple bands.
Enzymic treatments of solubilized membranesand isolated components. Trypsin and pronase were dissolved in
50 mM-Tris-HCl pH 8-2 and 7-4 respectively. Neuraminidase (from Clostridiumperfringens), phospholipases A2
and C were dissolved in PBS. Solubilized erythrocyte membranes, total lipids, phospholipids and gangliosides
were incubated in the presence or absence of various enzymes at 37 °C for 1 h. The enzymes were then inactivated
by heating the samples to 100 °C for 15 min. To remove from SEM the products of digestion, samples were washed
in an Amicon Centricon-I 0 M icroconcentrator using PBS as exchange buffer. Membranes were washed by three
cycles of 10 : 1 concentration and volume restoration using PBS. The original activity of untreated membrane was
recovered at 100~ after this process.
Chemicaldeterminations. Protein concentration was determined by the method of Lowry et al. (1951) using BSA
as a standard. Lipid concentration was determined by the phosphovanillin procedure described by Frings et al.
(1972) using olive oil as a standard. Sialic acid was determined by the thiobarbituric acid method (Aminoff, 1959);
lipid-bound sialic acid was measured by the resorcinol method described by Svennerholm & Fredman (1980).
Chemicals and enzymes. Octyl-fl-D-glucopyranosidewas purchased from Calbiochem. Triton X-100 and Tween
80 were obtained from Sigma. Neuraminidase (from Vibrio cholerae)was obtained from Behring. Neuraminidase
(from C. perfringens), phospholipases A2 and C, pronase and trypsin were obtained from Sigma. Cholesterol,
phosphatidylinositol, phosphatidylserine, phosphatidylcholine, phosphatidylethanolamine, cerebrosides,
sphingomyelin, diphosphatidylglycerol, phosphoserine and sialic acid were obtained from Sigma. Sphingomyelin
was also purchased from Fluka and Koch-Light Laboratories.
RESULTS
Inhibiting activity o f S E M
Crude membranes of goose erythrocytes were solubilized by using Triton X-100, Tween 80
and octyl-fl-D-glucopyranoside. The detergent mixture was chosen because it can be easily
removed from the solubilized fraction by dialysis and adsorption to polystyrene beads, thus
permitting further purification and isolation of putative binding sites.
The presence of receptor activity for VSV in solubilized m e m b r a n e s was assessed by verifying
their capacity to inhibit viral attachment and fusion at acidic pH, measured as H A and He
inhibiting activity (Fig. 1). Solubilized m e m b r a n e s a p p e a r e d to be more active in inhibiting H A
than He, but it must be noted that 256 H A U were necessary to produce 1009/oo He (data not
shown); 4 H A U and 2 H e U of VSV correspond to 2.2 × 108 and 1.4 × 10 xa p.f.u, respectively.
Plotting of ~ He against S E M concentration gave a slightly sigmoid curve with an almost linear
section between 2 0 ~ and 8 0 ~ He. Complete inhibition of He activity was obtained with
66 ~tg/ml (as lipid) of SEM. In these experiments controls were included which consisted of
goose erythrocytes incubated with m e m b r a n e s for 30 min at 4 °C, then for 40 min at 23 °C before
the addition of virus. Controls gave results similar to untreated cells.
Effect o f enzymic treatments on S E M inhibiting activity
To study the chemical nature of the inhibitor, solubilized m e m b r a n e s were treated with
several enzymes. Results are reported in Table 1. The inhibiting activity was resistant to
denaturing conditions such as heating or freezing and thawing and to trypsin and pronase
digestion. After phospholipase Az treatment S E M activity was enhanced; in contrast
phospholipase C and neuraminidase digestion produced a reduction in the capacity of
membranes to inhibit viral attachment and fusion.
Inhibiting activity o f membrane components
To characterize the role of different m e m b r a n e components protein, lipid, phospholipid and
glycolipid were extracted from erythrocyte membranes and tested for their ability to compete
with whole cells for virus binding (Table 2). M e m b r a n e proteins were able to inhibit virus
attachment and fusion only at very high concentration. Total lipids, phospholipids and
Rlvcolioids, on the other Downloaded
hand, retained
definite inhibitory activity.byRemoval of protein from
from awww.microbiologyresearch.org
IP: 88.99.165.207
On: Thu, 11 May 2017 07:30:43
P. MASTROMARINO AND OTHERS
2362
11
100 ~--,~
I
I
|-\
I
I
I
I
I
I
tl
I
I
?
\
50
\
@
HA
I
I
i
i\
\
50
o
0,
I
100
!
I
I
I
I\e--t
0-5 1 2 4 8 17 33 66 133
SEM (~tg/ml lipid)
+ + . . . . . . .
I
l
I
I
0.5
2
8
31
~\. \£~-;.
125 500 2000
Lipid components (/ag/ml)
Fig. 1
Fig. 2
Fig. 1. Inhibition of VSV attachment and fusion by SEM. SEM were incubated with virus for 2 h at
4 °C to permit attachment to VSV envelope, but not fusion of SEM and viral envelope lipids. After
addition of erythrocytes the temperature was maintained at 4 °C to allow adsorption of the virus to the
cell membrane. HA was read 2 h later; He was determined after 40 rain at 23 °C as described in
Methods.
Fig. 2. Dose-dependent inhibition of VSV haemolysis by lipid components of goose erythrocyte
membranes. VSV was incubated for 2 h at 4 °C with twofold dilutions of the total lipid (I), glycolipid
(O) or phospholipid (A) fractions extracted from goose erythrocyte membranes. Haemolysis assay was
performed as in Methods.
T a b l e 1.
Effect of enzyme treatments on inhibitory activity of solubilized erythrocyte membrane
towards haemagglutination and haemolysis by VSV
SEM
SEM + 100 °C, 15 min
+ freeze-thaw
+ trypsin
-t- pronase
Enzyme
concentration•
0
0
0
0.01
t
1.56
0.01
1.56
1
+ phospholipase A2
+ phospholipase C
+ neuraminidase
(C. perfringens)
+ neuraminidase
(V. cholerae)
50~ H A l t
1.56
1-56
1.56
1.56
0.1
1
5
0.1
1
5
1
10
0.01
0-1
0.25
0.5
1,56
1.56
0.57
0.39
1.56
2.82
3.12
0.78
2.64
1.56
1.56
2.34
3.12
~ HeI
50
54
47
61
47
65
56
55
57
59
45
31
19
58
38
55
45
41
36
* Protease concentrations are in mg/ml. Phospholipase and neuraminidase concentrations are in units/ml.
t Values reported represent lipid concentration ~g/ml) of SEM giving 50~ HA inhibition.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 11 May 2017 07:30:43
Erythrocyte receptor for VSV at acidic pH
2363
Table 2. Inhibiting activity of solubilized erythrocyte membrane and isolated membrane components
towards haemagglutinating and haemolytic activities of VSV*
SEM lipid
SEM protein
Protein
Total lipid
Phospholipid
Glycolipid
50~ HAI
1.6
1.1
400
7-8
31-2
15-6
50~ HeI
7.5
5.3
1600
15-6
280
110
* Values represent concentrations ~g/ml) giving 50~ haemagglutination (HA1) and haernolysis (HeI)
inhibition.
SEM caused a steady decrease in the activity of total lipids, by comparison with whole
solubilized membranes. This reduction could have been a consequence of protein removal or of a
structural rearrangement of lipids after the extraction procedure.
Reconstitution experiments were made to ascertain whether it was possible to restore the
effectiveness of whole solubilized membranes. For this purpose, extracted membrane proteins
were mixed with total lipids, or with isolated phospholipids and glycolipids. Lipid and protein
were mixed in the same ratio present in the whole solubilized membranes. In the first case a 50~o
inhibition of binding was achieved with a mixture containing 5.4 p.g/ml of proteins and 7.8 I.tg/ml
of lipids. The inhibition observed seemed due therefore only to the lipid moiety, because
7.8 p.g/ml of lipids alone give 50~o HA inhibition.
When phospholipids and glycolipids were mixed at their respective highest non-inhibitory
concentrations a 50 ~ inhibition was observed; the inhibitory activity of fractions was therefore
additive. The addition of proteins did not modify the inhibitory potency of the mixture (results
not shown).
The dose-dependent inhibition of VSV He produced by different lipid components is
recorded in Fig. 2 which shows that total lipids were more active than isolated phospholipid and
glycolipid components.
To evaluate further the role of different lipid molecules on VSV receptors on goose
erythrocytes, a number of phospholipids and glycolipids were tested for their ability to inhibit
HA and He by VSV. Gangliosides extracted from erythrocyte membrane were analysed by thinlayer chromatography (Fig. 3). This showed that GM3 (about 96~o of total lipid-bound Nacetylneuraminic acid, NeuNAc) was the major ganglioside in the membrane of goose
erythrocytes, while only traces of components with Rv between G M 1 and G D 1a were detected.
The latter were presumed to belong to the neo-lacto series of gangliosides, but we did not attempt
to characterize them further.
Data reported in Table 3 demonstrated that cholesterol, neutral phospholipids and neutral
glycolipids were devoid of any inhibitory effect. Anionic phospholipids (phosphatidylinositol,
phosphatidylserine, diphosphatidylglycerol) and gangliosides, on the other hand, inhibited viral
adsorption on erythrocyte membranes and subsequent He. The inhibitory activity was due to
direct action on the virus, because preincubation of erythrocytes with lipids did not affect their
susceptibility to VSV-induced HA and He (data not shown).
The half-maximal inhibitory concentration for HA and He is reported in Table 4.
Phosphatidylinositol was the membrane component most active in the inhibition of VSV
attachment and fusion; sphingomyelin showed inhibitory effects only at a relatively high
concentration. These results were obtained with three different standard preparations of
sphingomyelin. These were found to be free of contaminating phospholipids and fatty acids as
verified by thin-layer chromatography. Palmitic and oleic acids inhibited the haemolytic
activity of VSV, without any effect on the attachment capacity of the virus.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 11 May 2017 07:30:43
2364
P. MASTROMARINO AND OTHERS
1
2
3
4
5
Fig. 3. Thin-layerchromatogram ofgangliosides. Goose erythrocyte gangliosides fractions (lanes 1 and
2) were isolated by Silica gel 60 column chromatography. Lane 1, NeuNAc GM3; lane 2, minor
gangliosides (RF between GM1 and GDla); lane 3, NeuNAc GM3 from human liver; lane 4, NeuNGc
GM3 from horse erythrocytes; lane 5, standard gangliosides (from top to bottom): NeuNAc GM3,
NeuNGc GM3, GMI, GDla, GDIb and GTlb.
T a b l e 3. Effect of individual membrane components on haemagglutinating and haemolytic activities
of v s v
Control virus
Phosphatidylinositol
Phosphatidylserine
Sphingomyelin
Phosphatidylethanolamine
Phosphatidylcholine
Diphosphatidylglycerol
Cholesterol
GM3 ganglioside
Minor gangliosides
Neutral glycolipids
Galactocerebrosides type I
Galactocerebrosides type II
Cerebroside sulphate
Phosphoserine
N-Acetylneuraminic acid
Palmitic acid
Oleic acid
HA*
+
+
+
-
~ He
100
3
10
100
97
97
2
+
99
+
+
+
+
+
+
+
+
4
8
100
100
100
100
100
100
75
76
* +, No change in HA titre; - , HA inhibition.
Enzymic modification of inhibiting molecules
Substances s h o w i n g i n h i b i t o r y activity towards a t t a c h m e n t of V S V to erythrocyte
m e m b r a n e s b e a r a n e g a t i v e electric charge. T o verify the role of electrostatic i n t e r a c t i o n in
b i n d i n g o f VSV to the lipid c o m p o n e n t o f the erythrocyte m e m b r a n e lipids, p h o s p h o l i p i d s and
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 11 May 2017 07:30:43
Erythroeyte receptor for VSV at acidic pH
2365
Table 4. Inhibition of VSV-induced haemagglutination and haemotysis by purified lipids
Concentration ~M) for 50% inhibition
A
t
HA
1.1
20.1
513-7
5.3
6.3
17.4
> 3000
> 3000
Phosphatidylinositol
Phosphatidylserine
Sphingomyelin
Diphosphatidylglycerol
N-Acetylneuraminic GM3
Minor gangliosides
Palmitic acid
Oleic acid
He
9
193.3
1340
21.4
25.4
139.6
1468
1773
Table 5. Effect of enzymes on the inhibitory activity of membrane components towards
haemagglutination and haemolysis of VSV
Lipids
Lipids + I00 °C, 15 rain
+ phospholipase C
+ phospholipase A2
Enzyme
concentration*
0
0
0-001
0-01
0.1
1
5
0-1
50~ HAIr
7.8
7-8
7-8
15.6
31.2
62.5
125
7.8
1
GM3
GM3+ 100°C, 15 min
+ neuraminidase
Phosphatidylinositol
Phosphatidylinositol + 100 °C, 15 rain
+ phospholipase C
Phosphatidylserine
Phosphatidylserine + 100 °C, 15 min
+ phospholipase C
Proteins
Proteins + 100 °C, 15 rain
+ trypsin
5
0
0
0.5
0
0
l
0
0
1
0
0
1
1-95
2.45
7.8
7.8
256
0.98
0.98
1-96
15.6
15.6
31-2
400
400
400
~ HeI
50
50
50
37
21
9
5
55
70
74
50
51
7
41
45
21
43
47
30
50
52
53
* Trypsin concentration is in mg/ml. Phospholipase and neuraminidase concentrations are in units/ml.
t Values represent concentration ~g/ml) giving 50~ HA inhibition.
gangliosides were enzymically modified to remove anionic groups. After treatment of lipid,
phosphatidylserine and phosphatidylinositol with phospholipase C and GM3 ganglioside with
neuraminidase (Table 5) a significant reduction in their inhibitory activity towards binding and
fusion of VSV to erythrocytes was observed. The effect was dose-dependent. However, the
isolated anionic groups of these molecules, i.e. phosphoserine and sialic acid at 1 mM were
ineffective in inhibiting HA and He (Table 3).
Interestingly, phospholipase A2 digestion of whole m e m b r a n e lipids caused an e n h a n c e m e n t
of inhibitory activity; this effect could have been due to the strong surface activity of
lysoderivatives and fatty acids produced by the enzyme or to a rearrangement of the lipid
molecules in a form more suitable for interaction with the virus.
Heating or trypsin digestion of extracted m e m b r a n e proteins did not modify their inhibitory
activity.
Effect of enzyme treatment of goose erythrocytes on attachment and fusion of VSV
The results reported in the preceding sections suggested a role for lipids, specifically
gangliosides and phospholipids as part of the m e m b r a n e receptors for VSV. To confirm the
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 11 May 2017 07:30:43
2366
P. M A S T R O M A R I N O
AND OTHERS
Table 6. Effect of enzyme treatments of goose erythrocyte on the attachment and fusion of VSV
Enzyme
Phospholipase A2
Phospholipase C
U/ml
0
2
10
50
125
250
0
0.4
1
Neuraminidase*
Untreated~
+ neuraminidase
+ phospholipase C
+ neuraminidase
phospholipase C
2
0
0-25
0.4
0.8
0
0.25
1
0.25
1
HA titre
32
32
32
24
24
24
32
6
4
0
32
256
384
1024
32
256
6
ND~
~ He
81
81
72
67
66
59
60
41
38
28
67
87
96
100
63
84
46
32
* Sialic acid released from membranes by action of V. cholerae neuraminidase was respectively 0.030, 0.035,
0.038 pmol.
t Erythrocytes were digested with enzyme alone or with neuraminidase and successively with phospholipase C.
The concentrations of enzymes used were the maximum that did not result in any haemolysis of erythrocytes upon
combined treatment.
:~ It was not possible to determine agglutination of erythrocytes by VSV, because removal of N-acetylneuraminic
acid and polar head of phospholipids produced spontaneous agglutination of red cells.
importance of these molecules in whole erythrocytes, cells were digested with neuraminidase
and phospholipases and the sensitivity of modified erythrocytes to VSV attachment and fusion
was studied.
After phospholipase Az/albumin treatment of erythrocytes (Table 6) agglutination by VSV
was unaffected; He was, on the contrary, significantly decreased. Phospholipase C digestion
produced a marked reduction in the binding and fusion of VSV with the erythrocyte membrane,
thus showing that polar heads of surface phospholipids played a fundamental role in the binding
of the virus.
The involvement of the membrane ganglioside GM3 as a VSV receptor and the crucial role of
its sialic acid moiety for activity are recorded in Table 5. We therefore tested the effect of
neuraminidase pretreatment of erythrocytes on subsequent VSV attachment and fusion. As
shown in Table 6, binding and He were greatly increased after removal of sialic acid from the
cell surface. This effect was dose-dependent.
DISCUSSION
Competition binding experiments showed that the lipid components of the goose erythrocyte
membrane, notably phospholipids and glycolipids, were specifically recognized by VSV during
the early phases of virus-cell interaction. Total lipids extracted from erythrocyte membranes
were found to be more effective than isolated components in the inhibition of VSV binding to
erythrocyte surface. However reconstitution experiments, carried out by mixing the fractions
together, demonstrated that it was not possible to restore the effectiveness of whole solubilized
membranes. It is probable that the steric arrangement of various components in solubilized
membrane is more suitable for the interaction with the virus than that obtained in reconstituted
membrane liposomes.
Proteins did not appear to have a role as VSV receptors. While it is known that certain
glycolipids associate tenaciously with membrane proteins or glycoproteins and partition into the
aqueous phase on solvent extraction (Hakomori et al., 1972), we consider it unlikely that the
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 11 May 2017 07:30:43
Erythroeyte receptor for V S V at acidic p H
2367
weak inhibitory activity of membrane proteins was due to the protein itself, but rather to a
putative contamination of the extract by traces of lipid. This is supported by the observation that
the activity of the extracted proteins was unaffected after heating or trypsin digestion (Table 5).
This result is in agreement with the previously reported effect of trypsin treatment of
goose (Seganti et al., 1982) or human erythrocytes (Bailey et al., 1984), which produced an
enhancement of VSV binding and He.
The importance of lipid molecules in erythrocyte receptors f~r VSV is confirmed by the
reduction of agglutination of red blood cells after phospholipase C digestion. Moreover,
treatment of goose erythrocyte solubilized membranes with phospholipase C and neuraminidase
rendered them less effective in the inhibition of VSV binding and fusion, suggesting a role of
phospholipids and sialic acid-containing glycolipids in the inhibitory activity. Schlegel et al.
(1983) reported a similar effect towards the inhibitory activity of saturable VSV binding after
treatment of solubilized Vero cell membranes with phospholipase C. They did not notice any
effect after neuraminidase digestion.
Phosphatidylinositol, phosphatidylserine, diphosphatidylglycerol and gangliosides were able
very efficiently to reduce VSV attachment and fusion, probably by competition with its natural
binding sites on the cell surface. A common characteristic of these substances is their negative
electric charge; it is therefore possible that inhibition is mediated by electrostatic interactions.
The role of polar heads of inhibiting molecules is confirmed by results obtained after enzymic
cleavage of anionic groups. Removal of NeuNAc from GM3 ganglioside and of phosphoserine
and phosphoinositol from phospholipids greatly decreased their activity. Thus electrostatic
interactions seem to play a fundamental role for the inhibiting activity towards VSV of purified
membrane components. However sialic acid and phosphoserine were unable to affect
attachment of virus to cell receptors, indicating that in the active molecules both polar heads and
hydrophobic portions are essential for inhibiting activity.
The inhibitory activity shown by diphosphatidylglycerol, an anionic phospholipid found in
the inner mitochondrial membrane and not present on the erythrocyte surface, demonstrated
that molecules bearing head groups with negative charges and hydrophobic lipid portions can
act as VSV receptors at acidic pH. This suggests that binding of VSV to the cell membrane
involves (i) an electrostatic interaction between positive amino acid sequences of viral G
envelope glycoprotein and an anionic group on the cell membrane and (ii) a tight hydrophobic
linkage between a lipophilic peptide segment of G glycoprotein and a lipid portion of the
receptor on the membrane.
Bailey et al. (1984) have suggested that electrostatic interactions can mediate VSV binding to
the cell surface. It was shown that the poly-cation DEAE-dextran increased the binding of VSV
to BHK-21 cells and erythrocytes of several species; He was equally enhanced. DEAE-dextran
produces an increase in positive charge on the cell surface which results in a reduction of
repulsive forces between virus and cell membrane.
We observed a great enhancement of attachment and fusion of VSV to neuraminidase-treated
erythrocytes. Reduction of surface negative charge by the enzyme facilitates electrostatic
interaction between viral envelope and the anionic components of the cell membrane.
Analysis of GM3 ganglioside extracted from neuraminidase-treated erythrocytes demonstrated that the enzyme was able to digest sialic acid from GM3. This result does not preclude
the possibility that GM3 can be a receptor for VSV; in fact it can be postulated that
neuraminidase treatment inactivates one natural receptor site for VSV, but facilitates the
binding of the viral envelope with the high affinity binding sites containing anionic
phosphatidylinositol and phosphatidylserine. Results obtained after combined treatment of
erythrocytes with neuraminidase and phospholipase C seem to confirm this hypothesis. Indeed
erythrocytes treated with the two enzymes were less sensitive to He by VSV than after the action
of phospholipase C alone (Table 6).
The only neutral lipid showing inhibiting activity was sphingomyelin, although only at high
concentrations. This result is in agreement with the observation made by Superti et al. (1984)
who reported that sphingomyelin was able to reduce the infectivity of VSV on CER ceils and
that sphingomyelinase digestion rendered cells less sensitive to infection by the virus.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 11 May 2017 07:30:43
2368
P. M A S T R O M A R I N O
AND
OTHERS
Phosphatidylinositol, phosphatidylserine and GM3 appeared to represent major binding sites
for VSV on erythrocyte membranes at acidic pH. Schlegel et al. (1983) have proposed a similar
explanation for the role of phosphatidylserine at neutral pH in Vero cells. In this case, however,
no activity of phosphatidylinositol was detected.
Binding and fusion activity of VSV were not strictly correlated; HA was essential but not
sufficient for fusion. In fact, phospholipase A2 treatment of erythrocytes produced a decrease in
their sensitivity to He by VSV, although agglutination was unaffected (Table 6). This reduction,
therefore, does not seem to be due to a cleavage of virus receptor from the erythrocyte surface but
to a rearrangement of membrane lipids after removal of fatty acids and lysophosphatidylcholine
upon the extraction with albumin; fusion of the viral envelope with erythrocyte membrane is
thereby hindered.
After binding of VSV to the cell surface, the virus is delivered to intracellular receptosomes
(Matlin et al., 1982) where the low pH triggers the fusion between the viral and vesicle
membranes and causes the release of nucleocapsid into the cytoplasm. Phosphatidylinositol,
phosphatidylserine and GM3 probably mediate this pH-dependent membrane fusion event
within acidic vesicles.
We wish to thank Dr A. Cantafora for the supply of sphingomyelins and for the analysis of their purity and
Professor R. Perez-Bercoff for helpful suggestions towards improvement of the manuscript. This research was
supported by grants from Istituto Pasteur-Fondazione Cenci Bolognetti and Progetto Finalizzato CNR, Controllo
Malattie da Infezione.
REFERENCES
AMIIqOFF,n. (1959). The determination of free sialic acid in the presence of the bound compound. Virology 7, 355357.
BAILEY, C. A., MILLER, D. K. & LENARD, J. (1984). Effects of DEAE~lextran on infection and hemolysis by VSV.
Evidence that nonspecific electrostatic interactions mediate effective binding of VSV to cells. Virology 133,
111-118.
BYRNE, M. C., SBASCHNING-AGLER,M., AQUINO,D. A., SCLAFANI,J. R. & LEEDEN, R. W. (1985). Procedure for isolation
of gangliosides in high yield and purity: simultaneous isolation of neutral glycosphingolipids. Analytical
Biochemistry 148, 163-173.
DALES, S. (1973). Early events in cell-animal virus interactions. Bacteriological Reviews 37, 103-135.
Foecn, J., LEES,M. & SLOANESTANLEY,G. H. (1957). A simple method for the isolation and purification of total lipids
from animal tissue. Journal of Biological Chemistry 226, 497-509.
FRIYGS, C. S., FENDLEY,T. W., DUYN, R. T. & QUEEr~, C. A. (1972). Improved determination of total serum lipids by
sulfo-phospho-vanillin reaction. Clinical Chemistry 18, 673-674.
GINSBERG, B. H., KAHN,C. R. & ROTH, J. (1976). The insulin receptor of the turkey erythrocyte. Characterization of
the membrane-bound receptor. Biochimica et biophysica acta 443, 227-242.
HAEST, C. W. M., PLASA,G. & DEUTICKE, B. (1981). Selective removal of lipids from the outer membrane layer of
human erythrocytes without haemolysis. Consequences for bilayer stability and cell shape. Bioehimica et
biophysica acta 649, 701-708.
~d~OMORI, S. & WATANABE,K. (1976). Blood group glycolipids of human erythrocytes. In Glycolipid Methodology,
pp. 13-47. Edited by L. A. Witting. Bellefonte: Supelco Inc.
~KOMORI, S., STELLNER, g. & WATANABE,g. (1972). Antigenic variants of blood-group A glycolipid. Examples of
highly complex, branched-chain glycolipid of animal cell-membrane. Biochemical and Biophysical Research
Communications 49, 1061-1068.
HALONEN, P. E., TOIVANEN,P. & NIKKARI,T. (1974). Non-specific serum inhibitors of activity of haemagglutinins of
rabies and vesicular stomatitis viruses. Journal of General Virology 22, 309-318.
LAnISn, S. & GILL~d~D, B. (1985). A solvent partition method for microscale ganglioside purification. Analytical
Biochemistry 146, 220-231.
LONBERG-HOLM, K. & PHILIPSON, L. (1974). Early interaction between animal viruses and cells. In Monographs in
Virology, vol. 9, pp.l-149. Edited by J. L. Melnick. Basel: S. Karger.
LOWRY, O. H., ROSEBROUGH,N. J., FARR, A. L. & RANDALL,R. J. (1951). Protein measurement with the Folin phenol
reagent. Journal of Biological Chemistry 193, 265-275.
MADDY, A. H. (1966). The properties of the protein of the plasma membrane of ox erythrocytes. Biochimica et
biophysica acta 117, 193-200.
MATLIN, K., REGGIO, H., HELENIUS,A. & SIMONS, K. (1982). Pathway of vesicular stomatitis virus entry leading to
infection. Journal of Molecular Biology 156, 609-631.
MIFUNE, K., OI-IUCHI, M. & MANNEN, K. (1982). Hemolysis and cell fusion by rhabdoviruses. FEBS Letters 137,
293-297.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 11 May 2017 07:30:43
Erythrocyte receptor for V S V at acidic p H
2369
$CHLEGEL,R. &WADE,M. (1984). Biological activities of peptides corresponding to the amino-terminus of vesicular
stomatitis virus glycoprotein. In Abstracts of Sixth International Congressof Virology, p. 288. September 1-7,
1984, Sendai, Japan.
SCHLEGEL, R., TRALKA, T. S., WlLLINGHAM, M. C. & PASTAN, 1. (1983). Inhibition of VSV binding and infectivity by
phosphatidylserine: is phosphatidylserine a VSV-binding site? Celt 32, 639q546.
SEGANTI, L., SUPERTI, F., MASTROMARINO,P., SINIBALDI,L. & ORSI, N. (1982). Role of carbohydrates on cell m e m b r a n e
receptors for vesicular stomatitis virus. Bollettino dell'lstituto sieroterapico milanese 61, 294-299.
SEGANTI, L., GRASSI, M., MASTROMARINO, P., PAN./~, A.., SUPERTI, F. & ORSI, N. (1983). Activity of human serum
lipoproteins on the infectivity of rhabdoviruses. Microbiologica 6, 91-99.
SUPERTI, F., SEGANTI, L., TSIANG, H. & ORSI, N. (1984). Role of phospholipids in rhabdovirus a t t a c h m e n t to C E R
cells. Archives of Virology 81, 321-328.
SVErC~ERHOL~t, L. & FRED~tAN, P. (1980). A procedure for the quantitative isolation of brain gangliosides.
Biochimica et biophysica acta 617, 97-109.
VANDENBERG, S. R., ALLGREN, R. L., TODD, g. D. & CIARANELLO, R. D. (1983). Solubilization and characterization of
high-affinity [3H]serotonin binding sites from bovine cortical membranes. Proceedings of the National
Academy of Sciences, U.S.A. 80, 3508-3512.
(Received 9 December 1986)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 11 May 2017 07:30:43