J. gen. Virol. (1986), 67, 2813-2817. Printed in Great Britain
2813
Key words: influenza virus~membrane~fusion
Time and Temperature Dependence of Influenza Virus Membrane Fusion at
Neutral pH
By A N N E
M. H A Y W O O D *
AND B R A D L E Y
P. B O Y E R t
Departments of Pediatrics, Medicine and Microbiology, University of Rochester, Rochester,
New York 14642, U.S.A.
(Accepted 15 August 1986)
SUMMARY
The time course and temperature requirements for fusion of influenza virus
membranes with liposomes at pH 7-5 were found to be consistent with the
requirements for cell entry. At 37 °C, fusion was most rapid during the first 5 min and
then continued more slowly up to at least 1 h. The amount of fusion increased semilogarithmically with increasing temperature up to 50 °C.
For enveloped viruses to initiate infection, the viral nucleoprotein must become accessible to
the contents of the cell, which means that it must be at least partially released from its
membrane. One mechanism by which this can occur is by fusion of the viral membrane either
directly with the host plasma membrane or with the membrane of a vacuole after endocytosis of
the virus (Bukrinskaya, 1982; Dimmock, 1982; Kohn, 1985). Sendai virus membranes can fuse
directly with host plasma membranes (Morgan & Howe, 1968) or with liposomes (Haywood,
1974) at neutral pH.
Influenza virus, on the other hand, is thought to be an example of viruses that require low pH
for membrane fusion (Maeda et al., 1981 ; Sato et al., 1983; Skehel et al., 1982; White et al., 1982,
1983; Yoshimura et al., 1982). Influenza viruses have been shown to be present in both clathrincoated vesicles and uncoated vesicles (Matlin et al., 1981), and it has been postulated that fusion
is triggered by low pH within a coated vesicle. Changes in the conformation of the
haemagglutinin protein of X-31 (H3N2) influenza virus when the pH is lowered have been
extensively studied with the idea that these changes are necessary for membrane fusion (Skehel
et al., 1982). Fidgen & Tisdale (1981), however, gave unequivocal evidence that the membranes
of two strains of influenza virus can fuse with erythrocyte membranes at neutral pH. Later work
showed that at pH 7.5 egg-grown X-31 influenza virus membranes can fuse with liposomes that
contain ganglioside GD1a (Haywood & Boyer, 1985). When liposomes contain no net charge, it is
possible to demonstrate the same low pH requirement for fusion of influenza virus membranes
that has been described by other investigators. Therefore, the difference between these results
with liposomes and those of others (Sato et al., 1983; White et al., 1982) seems to be related to the
liposome composition and not to the virus strain or method of assay.
If the course of viral entry is primarily determined by the ability of the virus membranes to
fuse at neutral pH and does not require active participation by the host, then the characteristics
of the fusion of the viral membranes with liposomes should match the characteristics of viral
entry into a cell. Therefore, the time course and temperature requirements for fusion of influenza
membranes with liposomes at neutral pH were measured.
X-31 (H3N2) influenza virus was obtained from Dr R. G. Webster, grown in embryonated
chicken eggs, labelled with [35S]methionine and purified as previously described (Haywood &
Boyer, 1985). The virus was dialysed against H K N (3 mM-HEPES, 2.7 mM-KC1, 130 mM-NaC1,
t Present address: Eastman Kodak Research Laboratories, Rochester, New York 14650, U.S.A.
0000-7335 © 1986 SGM
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 19:13:39
2814
Short communication
pH 7.5). Bovine serum albumin was added at a final concentration of 0-003~ to prevent the
aggregation that occurs during freezing. The virus was divided into small aliquots before
freezing, so that after dialysis it was thawed only once before use. A modification of the Lowry
procedure (Markwell et al., 1978) was used to measure protein concentration, and the specific
activity of the virus was 254 c.p.m./~tg protein.
Liposomes were made from 0.7~tmol egg phosphatidylcholine (PC), 0.3~tmol egg
phosphatidylethanolamine (PE), 0.66 ~mol cholesterol and 0-03 p.mol ganglioside GDla in
H K N The source of these lipids and methods for assaying purity and concentration have been
previously described (Haywood & Boyer, 1984). Multilamellar liposomes were made as
previously described (Haywood & Boyer, 1984) within a few hours of use.
Influenza viruses (7.9 ~tg of protein) were adsorbed to liposomes at 0 °C for 1 h in a final
volume of 0.3 ml. The time and temperature of incubation were as indicated. During fusion,
viral membrane proteins are incorporated into the liposome membranes and viral internal
proteins are released inside the liposomes. The amount of fusion was measured as previously
described (Haywood & Boyer, 1985) by determining the amount of [35S]methionine-labelled
virus protein that was transferred to liposomes. In brief, after incubation, the virus and
liposomes were chilled and the viruses that were adsorbed but not fused were removed from the
liposomes by adding fetuin (50 mg/ml) for 15 min at 0 °C. (Fetuin is a sialoglycoprotein which
binds influenza virus.) The viruses that had not fused were then separated from the liposomes by
centrifugation, and the amount of labelled viral protein associated with the liposomes was
counted. Micrographs have confirmed that this treatment removes bound viruses except for rare
viruses which are trapped between liposomes and which presumably account for the
background counts at 0 °C (Haywood & Boyer, 1985). Micrographs also show that the viral
proteins can be seen incorporated into the liposome structure (Haywood & Boyer, 1985). The
amount of fusion was calculated as the percentage of counts in viruses transferred to liposomes
during incubation minus the percentage of counts associated with the liposomes after adsorption
at 0 °C (6.5~ of the counts added in the time experiments and 8 ~ of the counts added in the
temperature experiments). The S.E.M. for all the data shown was _+ 1 ( ~ virus fused) or less,
except for the experiments performed at 50 °C where the S.E.M. was --+4.
To measure the time course of fusion of influenza virus membranes with liposomes, the
viruses were adsorbed to the liposomes for 1 h in an ice bucket, and the temperature was then
raised to 37 °C for the indicated times. After incubation the samples were returned to the ice
bucket and the amount of protein transferred to liposomes was measured. The time course of the
fusion of influenza virus membranes with liposomes at 37 °C and at pH 7-5 is shown in Fig. 1.
The amount of fusion increased rapidly during the first 5 min at 37 °C and then increased more
slowly until 1 h when the experiment was ended. A similar time course was found when the
preadsorption step was omitted and viruses and liposomes were warmed to 37 °C and then
mixed (not shown). This time course of fusion of influenza X-31 membranes with liposomes is
consistent with the data on entry of different influenza viruses into M D C K cells (Yoshimura et
al., 1982; Matlin et al., 1981). A study that used resonance energy transfer as an assay (Stegmann
et al., 1985) indicated fusion at neutral pH that is comparable to the data in Fig. 1, although
cardiolipin liposomes were used which made the liposomes unlike the host membrane.
Since, at pH 7.5, only 6 ~ of the influenza viruses have fused after 2 rain and 11 ~ have fused
after 5 rain, measurement of fusion after such short incubations gives a low estimate of the
capacity of the virus to fuse at neutral pH. The low amount of fusion after brief incubations may
be dismissed as negligible when compared to the amount of fusion at low pH in these time
periods. Fusion at low pH, however, only continues over a brief time, whereas fusion at neutral
pH continues to increase with time, similar to viral entry in vivo. It is important to realize that the
rates of fusion at pH 5.0 and 7.4 are markedly different and comparisons of the fusion capacities
after 2 or 5 min incubation are not valid.
The mechanism of membrane fusion at neutral pH has been most extensively studied with the
paramyxoviruses. The fraction of Sendai virions that fuse their membranes with liposomes
(Haywood & Boyer, 1982, 1984) is comparable to the fraction of influenza virions that fuse their
membranes with liposomes at neutral pH. Fusion of Sendai virus with liposomes requires an
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 19:13:39
Short communication
70
I
I
I
I
I
2815
'
'
'
'
'
I
25
6O
20
50
40
15
30
10
20
10;i
i
10
20
I
I
30
40
Time (min)
Fig. 1
1
I
50
60
0
10
20
30
40
Temperature ( ° C)
Fig. 2
50
Fig. 1. Time course of fusion of influenza virus membranes with liposomes. Influenza viruses were
adsorbed to liposomes for 1 h at 0 °C, incubated at 37 °C for the times indicated, and returned to 0 °C.
The percentage of labelled viral protein incorporated into liposomes was determined.
Fig. 2. Temperature dependence of fusion of influenza virus membranes with liposomes. Influenza
viruses were adsorbed to liposomes for 1 h at 0 °C. The temperature was raised to the indicated level for
1 h before returning the samples to 0 °C. The percentage of labelled viral protein incorporated into
liposomes was determined.
active F1 protein (Homma & Ohuchi, 1973; Scheid & Choppin, 1974) and occurs at the leading
edge of a developing vesicle where the bilayer is stressed and where the curvature facilitates
close approach of the bilayers (Haywood & Boyer, 1981 ; Haywood, 1983). The influenza virus
haemagglutinin protein is thought to serve the same role in fusion as the Sendai virus F protein.
Envelopment of Sendai virus by liposomes results from circumferential binding (Haywood,
1975) which is similar to the mechanism postulated for envelopment of influenza virus into
uncoated vesicles (Patterson et al., 1979; Matlin et al., 1981). Thus, if fusion of influenza virus
membranes and fusion of Sendai virus membranes occur by the same mechanism, envelopment
of the virus due to circumferential binding may be the rate-limiting step whether fusion occurs
directly with the plasma membrane or inside a vacuole.
The logarithm of the rates of biological reactions as of other chemical reactions increases
inversely as the reciprocal of the temperature (K) as long as no component is denatured by the
increasing temperature. The haemagglutinin proteins of most of the influenza strains that have
been tested have been shown to be resistant to a temperature of 50 °C (Howe et al., 1961).
Therefore, if the influenza haemagglutinin protein is directing the fusion, the fusion should
increase with temperature in a semi-logarithmic fashion up to 50 °C. The temperature
requirements for fusion of influenza virus membranes at pH 7.5 were measured by adsorbing the
virus to the liposomes at 4 °C and then transferring the samples to a water-bath at the indicated
temperatures for 1 h. The samples were then put back in an ice bucket, and the amount of viral
protein incorporated into the liposomes was measured. The fusion of influenza virus membranes
with liposomes after 1 h at different temperatures is shown in Fig. 2. The amount of fusion
increased as the temperature increased. When the natural logarithm of the amount of fusion in
1 h is plotted against the reciprocal of the absolute temperature (K), the result is a line that is
straight between 15 °C and 50 °C (not shown).
A high rate of fusion at 50 °C is also evidenced by the very high levels of fusion of an
influenza virus mutant at neutral pH at 50 °C even after only 5 min incubation (Wharton et al.,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 19:13:39
2816
Short communication
1986). A flavivirus was also noted to give extensive fusion at neutral pH after incubation for
5 min at 45 °C or 50 °C (Gollins & Porterfield, 1986).
Since the lipids used in these liposomes are fluid at all temperatures used, the effect of
temperature is probably related to the viral proteins as has been described for Sendal virus
membrane fusion with liposomes at neutral pH (Haywood & Boyer, 1982). The temperature
dependence of the segmental and rotational motion of the Sendal virus membrane proteins is
similar to the temperature requirement of virus-induced haemolysis, which suggests that the
temperature requirement lor membrane lusion relates to viral protein mobility (Lee et al., 1983).
Envelopment of the virus by liposomes (Haywood, 1975) also is temperature-dependent as is
envelopment of particles by phagocytic cells (Griffin et al., 1975).
In conclusion, the time course for fusion of influenza virus membranes with liposomes at
pH 7.5 is consistent with the time course for viral entry into cells (Yoshimura et al., 1982; Matlin
et al., 1981). The temperature dependence is also consistent with in vivo viral entry except that
fusion with liposomes occurs at 50°C, which probably reflects the influenza virus
haemagglutinin's resistance to heat.
The time course is also consistent with the time course of the binding of myxoviruses to ceils
(Allison & Valentine, 1960; Matlin et al., 1981). Fusion at low pH, in contrast to fusion at neutral
pH, is very rapid. Fusion of influenza WSN with MDCK cells has been reported to have a halftime of less than 1 min at 37 °C (Yoshimura et al., 1982), and fusion of fowl plague virus with
liposomes is mainly complete by 2 min (White et al., 1982). Since, of course, fusion can only
occur after binding, the rapid course of fusion at low pH indicates that the binding must also be
far more rapid than occurs in vivo. Since it has been postulated that fusion at low pH occurs in
endocytic vesicles, the rapid time course of binding and fusion at low pH does not eliminate low
pH as a factor in in vivo fusion, but it does indicate that low pH causes rapid and efficient
binding which does not correspond to in vivo binding. This is consistent with the suggestion that
one major effect of low pH is to allow the virus to bypass the usual binding requirements
(Haywood & Boyer, 1985).
When influenza membrane fusion is studied over the brief times usually used in studies
relating to fusion at low pH, the low levels of fusion at neutral pH found have often been
dismissed as negligible. This work demonstrates that it is not valid to compare the capacity of
viral membranes to fuse at low and neutral pH by measuring the amount of fusion afer very brief
incubation times at 37 °C. Comparisons between fusion at neutral and low pH should be made
afer 1 h incubation. The slower time course of fusion at neutral pH must be taken into account in
any studies of fusion of influenza virus membranes.
This research was supported by National Science Foundation Grant P C M 8205896.
REFERENCES
ALLISON, A. C. & VALENTINE,R. C. (1960). Virus particle adsorption. II. Adsorption of vaccinia and fowl plague
viruses to cells in suspension. Biochimica et biophysica acta 40, 393-399.
BUKRINSKAYA,A.G. (1982). Penetration of viral genetic material into host cell. Advances in Virus Research 27, 141204.
DIMMOCK,N. J. (1982). Initial stages in infection with animal viruses. Journal of General Virology 59, 1-22.
FIDGEN,K. J. &TISDALE,M. (1981). An 'on grid' electronmicroscopicmethodfor studyingthe interaction and fusion
of influenza A virus with human erythrocyte membranes. Journal of Virological Methods 3, 271-276.
GOLLINS,S. W. ~ FORTERFIELD,J. S. (1986). pH-dependent fusion between the flavivirus West Nile and liposomal
model membranes. Journal of General Virology 67, 157-166.
GRIFFIN, F. M., JR, GRIFFIN, J. A., LEIDER,J. E. & SILVERSTEIN,S. C. (1975). Studies on the mechanism of phagocytosis.
I. Requirements for circumferential attachment of particle-bound ligands to specific receptors on the
macrophage plasma membrane. Journal of Experimental Medicine 142, 1263-1282.
HAYWOOD,A. M. (1974). Fusion of Sendal viruses with model membranes. JournalofMolecular Biology 87,625~628.
HAYWOOD,A. M. (1975). 'Phagocytosis' of Sendal virus by model membranes. Journal of General Virology29, 63-68.
HAYWOOD,g. M. (1983). Virus infection ofliposomes. In LiposomeLetters, pp. 277 287. Edited by A. D. Bangham.
London: Academic Press.
HAYWOOD, A. M. & BOYER, B. P. (1981). Initiation of fusion and disassembly of Sendal virus membranes with
liposomes. Biochimica et biophysica acta 646, 31-35.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 19:13:39
Short communication
2817
HAYWOOD,A. M. & BOYER, B. P. (1982). Sendai virus m e m b r a n e fusion : time course and effect of temperature, pH,
calcium, and receptor concentration. Biochemistry 21, 6041-6046.
HAYWOOD, A. M. & BOYER, B. P. (1984). Effect of lipid composition upon fusion of liposomes with Sendai virus
membranes. Biochemistry 23, 4161-4166.
HAYWOOD,A. M. & BOYER,B. P. (1985). Fusion of influenza virus m e m b r a n e s with liposomes at pH 7.5. Proceedings
of the National Academy of Sciences, U.S.A. 82, 4611-4615.
HOMMA,M. & OHUCHI, M. (1973). Trypsin action on the growth of Sendai virus in tissue culture cells. III. Structural
differences of Sendai viruses grown in eggs and tissue culture cells. Journal of Virology 12, 1457-1465.
HOWE, C., LEE, L. T. & ROSE, H. M. (1961). Collocalia mucoid: a substrate for myxovirus neuraminidase. Archives of
Biochemistry and Biophysics 95, 512-520.
KOHN, A. (1985). M e m b r a n e effects of cytopathogenic viruses. Progress in Medical Virology 31, 109-167.
LEE, P. M., CHERRY, R. S. & B.~.CHI,T. (1983). Correlation of rotational mobility and flexibility of Sendai virus spike
glycoproteins with fusion activity. Virology 128, 65-76.
MAEDA, T., KAWASAKI, K. & OHNISHI, S.-I. (1981). Interaction of influenza virus hemagglutinin with target
m e m b r a n e lipids is a key step in virus-induced hemolysis and fusion at p H 5.2. Proceedingsof the National
Academy of Sciences, U.S.A. 78, 4133-4137.
MARKWELL,M. A. K., HAAS,S. M., BIEBER, L. L. & TOLBERT, N. E. (1978). A modification of the Lowry procedure to
simplify determination in m e m b r a n e and lipoprotein samples. Analytical Biochemistry 87, 206-210.
MATLIN,K. S., REGGIO, H., HELENIUS,A. & SIMONS,K. (1981). Infectious entry pathway of influenza virus in a canine
kidney cell line. Journal of Cell Biology 91, 601-613.
MORGAN, C. & HOWE, C. (1968). Structure and development of viruses as observed in the electron microscope. IX.
Entry of parainfluenza I (Sendai) virus. Journal of Virology 2, 1122-1132.
PATTERSON,S., OXFORD, J. S. & DOURMASHKIN,R. R. (1979). Studies on the m e c h a n i s m of influenza virus entry into
cells. Virology 43, 223-229.
SATO,S. B., r~WASAKI, K. & OHmSm, S.-I. (1983). Hemolytic activity of influenza virus hemagglutinin glycoproteins
activated in mildly acidic environments. Proceedingsof the National Academy of Sciences, U.S.A. 80, 31533157.
SCHEID, A. & CHOPPIN, P. W. (1974). Identification of biological activities of paramyxovirus glycoproteins.
Activation of cell fusion, hemolysis and infectivity by proteolytic cleavage of an inactive precursor protein of
Sendai virus. Virology 57, 475-490.
SKEHEL, J. J., BAYLEY,P. M., BROWN, E. B., MARTIN,S. R., WATERFIELD,M. D., WHITE, J. M., WILSON, I. A. & WILEY, D. C.
(1982). Changes in the conformation of influenza virus hemagglutinin at the p H o p t i m u m of virus-mediated
m e m b r a n e fusion. Proceedings of the National Academy of Sciences, U.S.A. 79, 968-972.
STEGMANN, T., HOEKSTRA, D., SCHERPHOF, G. & WILSCHUT, J. (1985). Kinetics of pH-dependent fusion between
influenza virus and liposomes. Biochemistry 24, 3107-3113.
WHARTON,S. A., SKEHEL,J. J. & WILEY, D. C.(1986). Studies of influenza hemagglutinin-mediated m e m b r a n e fusion.
Virology 149, 27-35.
WHITE, J., KARTENBECK,J. & rmLENIUS, A. (1982). M e m b r a n e fusion activity of influenza virus. EMBO Journal l,
217-222.
WHITE, J., KIELIAN, M. & HELENIUS, A. (1983). M e m b r a n e fusion proteins of enveloped animal viruses. Quarterly
Review of Biophysics 16, 151 195.
YOSHIMURA, A., KURODA, K., KAWASAKI, K., YAMASHINA, S., MAEDA, T. & OHNISHI, S.-I. (1982). Infectious cell entry
m e c h a n i s m of influenza virus. Journal of Virology 43, 284~293.
(Received 7 July 1986)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 19:13:39